The CHANGING 2024 WEALTH of NATIONS Revisiting the Measurement of Comprehensive Wealth This edition of The Changing Wealth of Nations is dedicated to the memory of Kirk Hamilton, whose pioneering work and intellectual leadership have made this work possible. © 2024 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. RIGHTS AND PERMISSIONS The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Design and layout: Clarity Global Strategic Communications www.clarityglobal.net https://reproducibility.worldbank.org A reproducibility package is available for this book in the Reproducible Research Repository at https://reproducibility.worldbank.org/index.php/catalog/194 Table of Contents Acknowledgments 3 Preface 6 Executive Summary 11 I. Tracking Wealth to Monitor Economic Progress 25 1. Wealth as an Indicator of Sustainability 26 2. How the World Bank Measures Comprehensive Wealth 45 3. Global and Regional Trends in Wealth, 1995—2020 63 4. The Role of Limited Substitutability for Measuring Sustainability with CWON 83 II. Measuring Components of Comprehensive Wealth 95 5. The Nonrenewable Wealth of Nations 96 6. Valuation of Hydroelectric Resources 121 7. Forests and Agricultural Lands 143 8. Blue Natural Capital: Marine Fish Stocks, Aquaculture, and Mangrove Coastal Protection Services 160 9. Human Capital Wealth: Global Trends 182 Conclusions 199 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 2 ACKNOWLEDGMENTS Acknowledgments The report was developed by the World Bank’s Environment Institute for Environmental Studies). The review also drew Global Department under the counsel of Juergen Voegele, on a background paper on “Deflation in the CWON database” Vice President for the Planet Vice Presidency, Valerie Hickey, prepared by Robert Inklaar (University of Groningen), Director of the Global Department for the Environment, and Wulong Gu (Statistics Canada), and Erwin Diewert (University Richard Damania, Chief Economist for the Planet Vertical. of British Columbia). Additional technical guidance and Stefanie Onder (formerly World Bank, now American advice on the methodological revisions were provided by University), Diego Herrera, and Grzegorz Peszko provided Alessandra Alfieri (International Monetary Fund, IMF), leadership for this report, with core team members Hyungju Bram Edens (Organisation for Economic Co-operation and Kim, Robert Marks, Alexis Rivera, Catherine Van Rompaey, Development), Alejandro Caparros (Durham University), Matias Piaggio, and Isabel Saldarriaga Arango. The team Moritz Drupp (University of Hamburg), Barbara Fraumeni benefited from the thoughtful guidance and generous support (Central University for Economics and Finance in Beijing, of Christian Peter (World Bank), Raffaello Cervigni (World National Bureau of Economic Research in the United States), Bank), and Bekele Shiferaw (World Bank). The team would Carl Obst (Institute for the Development of Environmental- also like to thank the following World Bank peer reviewers Economic Accounting), P. Bhanumati (IMF), and James for their constructive and helpful comments throughout William Tebrake (IMF). the development of this report: Diego Arias Carballo, Kevin Carey, Louise J. Cord, Pablo Fajnzylber, Vivien Foster, The report includes important contributions and input from Stephane Hallegatte, Martin Heger, Elena Ianchovichina, the following lead authors and collaborators: Aart Kraay, Muthukumara Mani, Craig Meisner, Jason Russ, Giovanni Ruta, Stephane Straub, and Steven N. Schonberger. ■ Executive summary: The executive summary was Emily Lynn Gribbin provided excellent editorial support. prepared by Stefanie Onder with inputs from Diego Clarity Global Strategic Communications provided editorial Herrera, Hyungju Kim, Robert Marks, Grzegorz Peszko, and design support for this publication. Alexis Rivera, Catherine Van Rompaey, and Isabel Saldarriaga Arango. In preparation for this report, a systematic methodological review was conducted to ensure that: (i) the methodology ■ Chapter 1: This chapter was prepared by Eli P. Fenichel follows available and emerging international standards and (Yale School of Environment, Yale University) and Stefanie guidelines from the System of National Accounts (SNA) and Onder, with substantial contributions from Catherine the System of Environmental-Economic Accounting (SEEA) Van Rompaey. The chapter benefited from the thoughtful Central Framework and SEEA Ecosystem Accounting (SEEA- inputs and guidance of Isabel Saldarriaga Arango, EA) wherever possible, and (ii) methodological choices Grzegorz Peszko, and Robert Smith. Maude Gibbins (Yale are applied consistently across all assets. The review was University) provided research assistance. led by Robert Smith (Midsummer Analytics), who was supported by Matthew Agarwala (Bennett Institute for ■ Chapter 2: This chapter was prepared by Stefanie Public Policy, University of Cambridge/Tobin Centre for Onder and Robert Smith. The chapter benefited from Economic Policy, Yale University), Catherine Van Rompaey, the thoughtful inputs and guidance of Eli P. Fenichel, Karen Wilson (former assistant deputy head of Statistics Catherine Van Rompaey, Diego Herrera, and Isabel Canada), and Rintaro Yamaguchi (Japanese National Saldarriaga Arango. Maude Gibbins provided research 3 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH ACKNOWLEDGMENTS assistance. The chapter draws on the recommendations of ■ Chapter 5: This chapter was prepared by Robert Marks. the methodological review, which was led by Robert Smith, Special thanks are extended to Grzegorz Peszko for and benefited from the thoughtful guidance of Matthew the box on the low-carbon transition and Juan Camilo Agarwala, Catherine Van Rompaey, Karen Wilson, and Moreno Ramos (Duke University) and Hyungju Kim, who Rintaro Yamaguchi. supported the data collection and analysis, respectively. The chapter benefited from the thoughtful inputs and ■ Chapter 3: This chapter was prepared by Alexis Rivera and guidance of Alexis Rivera, Chadi Habib (World Bank), and Stefanie Onder. The CWON database used in this analysis Stefanie Onder. was prepared by Robert Marks, Diego Herrera, Hyungju Kim, and Alexis Rivera, and the purchasing power parity ■ Chapter 6: This chapter was prepared by Robert Smith, conversion factors were developed by Rui Costa (World with substantial contributions from Zuzana Dobrotkova Bank), Marko Rissanen (World Bank), Giovanni Tonutti (World Bank), Christophe de Gouvello (World Bank), (World Bank), and Mizuki Yamanaka (World Bank). The Grzegorz Peszko, and Stefanie Onder. Bente Taraldsten chapter benefited from the thoughtful inputs and guidance Brunes (World Bank) and Catherine Van Rompaey of Catherine Van Rompaey, Isabel Saldarriaga Arango, provided helpful and insightful comments. and the participants of the European Commission’s Directorate General Economic and Financial Affairs— ■ Chapter 7: This chapter was prepared by Diego Herrera, Output Gap Working Group (DG ECFIN-OGWG) workshop Matias Piaggio (World Bank), Juha Siikamäki (IUCN), Alexis on natural capital measurement and modelling, including Rivera, and Hyungju Kim, and draws extensively on the chairs Bjoern Doering (DG ECFIN) and Toon Vandyck technical report on non-wood forest ecosystem services (OECD), fellow panelist Edward Barbier (Colorado State authored by Juha Siikamäki, Matias Piaggio, Natalia University), and discussant Christopher Zuber (German da Silva (Universidad de la República), Ignacio Álvarez Council of Economic Advisors). (Universidad de la República), and Ziyan Chu (consultant); the technical report on global vegetation carbon stocks ■ Chapter 4: This chapter was prepared by Francois Cohen authored by Alessio Bulckaen (Basque Centre for Climate (University of Barcelona), Richard Damania, Moritz Change/BC3), Raul Abad Viñas (BC3), Diego Bengochea Drupp, Ben Groom (London School of Economics), Cem Paz (BC3), Ruben Crespo (BC3), and Ferdinando Villa Karayalcin (Florida International University), Harun (BC3); and the technical report on the valuation of carbon Onder (World Bank), Stefanie Onder, Sjak Smulders retention services authored by Alejandro Caparrós (Tilburg University), and Daan van Soest (Tilburg (Durham University), Moritz Drupp, Bram Edens (OECD), University), and builds on technical background papers Martin Haensel (University of Leipzig), Stefanie Onder prepared for this report by Smulders and van Soest (2023), (American University), Alexis Rivera (World Bank), and Damania et al. (2023), Karaylacin and Onder (2023), Catherine Van Rompaey. Keith Fuglie (USDA) supported and Drupp et al. (2023). Participants of the World Bank the revision of the agricultural lands methodology and workshop on strong sustainability and of the EAREA 2023 provided key data inputs, and Orjan Jonsson (FAO) conference provided helpful input and guidance. Robert provided guidance on the timber wealth update. Naikoa Marks produced the experimental estimates using Aguilar Amuchastegui (World Bank), Ines Susana Angulo differentiated discount rates. De Aviles (World Bank), Alessandra Alfieri (IMF), Kenneth REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 4 ACKNOWLEDGMENTS Bagstad (United States Geological Survey), Stefano Balbi Kobayashi (independent consultant), Yumiko Kura (BC3), P. Bhanumati (IMF), Timothy Brown (World Bank), (independent consultant), Bard Misund (University of Raffaello Cervigni (World Bank), Andres Espejo (World Stavanger), Ahmed Nasr-Allah (WorldFish), Stefanie Bank), Stephane Hallegatte, Jia Li (World Bank), Alexander Onder, Ronnie Tan (US Grains Council), and Md. Abdul Lotsch (World Bank), and James William Tebrake (IMF) Wahab (Bangladesh Agricultural University) provided provided helpful guidance and inputs. helpful comments and guidance. ■ Chapter 8: This chapter was prepared by Alexis Rivera ■ Chapter 9: This chapter was prepared by Kenan Karakülah and Juan Camilo Moreno and draws extensively on the (World Bank) with inputs from Hyungju Kim, Stefanie technical report on coastline protection services of Onder, and Diego Herrera. Special thanks are extended mangroves authored by Pelayo Menéndez (Center for to Alexis Rivera, Cristobal Bennett Avaria (University of Coastal Climate Resilience, University of California), Chile), Pedro Nicolas Cayul Pina (University of Chile), Michael W. Beck (Center for Coastal Climate Resilience, Marcel Das (Luxembourg Income Study), Cristian Ignacio University of California), Sheila Abad (Instituto de Jara Nercasseau (University of Chile), Steven Johnston (the Hidráulica Ambiental-IH Cantabria), and Iñigo J. Losada Treasury of New Zealand), Mario Gronert (World Bank), (Instituto de Hidráulica Ambiental-IH Cantabria); the Claudio Montenegro (University of Chile), Tim Ngy (the technical report on marine fish stocks authored by U. Treasury of New Zealand), and Israel Osorio-Rodarte Rashid Sumaila (University of British Columbia), Louise (World Bank), who contributed to data collection and Teh (University of British Columbia), Deng Palomares processing the data before the final analyses. Barbara (University of British Columbia), Lubna Alam (University Fraumeni (Central University for Economics and Finance of British Columbia), and William Cheung (University in Beijing, National Bureau of Economic Research in of British Columbia); and the technical report on the United States), Robert Inklaar, and Stefanie Onder the aquaculture pilot accounts authored by Malcolm provided helpful comments and guidance. Dickson (World Bank), Robert Smith (Midsummer Analytics), Harrison Charo Karisa (World Bank), and ■ Conclusion: This chapter was prepared by Stefanie Onder. Stefanie Onder (American University). Philip Buike Livia Bizikova (International Institute for Sustainable (independent consultant), Gabriel Englander (World Development), Eli Fenichel, Catherine Van Rompaey, and Bank), Borja Gonzalez Reguero (World Bank), Mimako Robert Smith provided helpful comments and guidance. 5 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH PREFACE Preface Gross domestic product (GDP) is widely recognized as an Over the past two decades, the CWON program has updated insufficient measure of economic progress and national and expanded its comprehensive wealth estimates with each “success.” Since GDP is nearly universally available and new edition, as new data sources, measurement techniques, comparable across countries, it is extensively used as a and guidance became available. This 5th edition continues benchmarking and reference statistic—even for purposes this tradition and adopts international best practice in for which it was not designed. GDP measures the level of computing wealth in real terms. With this new approach, domestic productive activity, but it ignores the costs of this changes in real wealth per capita will be driven by (i) the growth in terms of the environmental degradation that depletion or accumulation of assets, (ii) changes in the occurs in the process of production, for example. Sir Partha productivity or relative scarcity of assets, (iii) changing Dasgupta likened this to a soccer team that only measures substitution patterns, and (iv) increasing or decreasing success as goals for and ignores goals against. competition for available assets—all of which are important for analyzing the sustainability of economic progress. Whether economic progress is sustainable can be measured by how real wealth per capita is changing, as this represents This report is intended primarily for a technical audience, changes in future production (and ultimately consumption) including policy advisors, statisticians, and researchers. It opportunities. Wealth in this context encompasses the value first presents the rationale for using wealth as a measure of all the assets of a nation that support economic production, of economic progress (chapter 1), explains in detail the such as its factories and roads (produced capital); forests, CWON methodology (chapter 2), and presents global trends fish stocks, and fossil fuel reserves (natural capital); labor observed in the data (chapter 3). It then discusses how the force (human capital); and net foreign assets. As long as methodology could be further improved to account for real wealth per capita does not decline, future generations the increasing relative scarcity of key assets, most notably will have at least the same opportunities as the current renewable natural capital (chapter 4). The subsequent generation, suggesting that development may be sustainable. chapters present the methodology and trends of the assets of the CWON wealth portfolio that are developed by the World All countries produce GDP estimates, but few measure wealth. Bank: nonrenewable natural capital (chapter 5), hydropower The World Bank’s The Changing Wealth of Nations (CWON) (chapter 6), forests and agricultural land (chapter 7), blue program addresses this gap. The CWON program is one of the natural capital (chapter 8), and human capital (chapter 9). pioneering efforts in measuring wealth, producing the most The final chapter concludes and outlines ways to use the comprehensive, publicly accessible, and reproducible wealth CWON database. database currently available. These monetary estimates draw on internationally endorsed concepts and valuation principles from the System of National Accounts and the System of Environmental-Economic Accounting. This ensures that CWON’s wealth measure is methodologically rigorous and comparable to other metrics of economic progress like GDP. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 6 ACRONYMS AND ABBREVIATIONS ACRONYMS AND ABBREVIATIONS CPI Consumer price index OECD Organisation for Economic Co-operation and Development EEZ Exclusive economic zone ONS UK’s Office for National Statistics EIA US Energy Information Administration PPA Power purchase agreement ERRA Energy Regulators Regional Association PPP Purchasing power parity EWN External Wealth of Nations PWT Penn World Tables FAO Food and Agriculture Organization RVM Residual value method FCV Fragility, conflict, and violence SAU Sea Around Us FERU Fisheries Economics Research Unit SCC Social cost of carbon GDP Gross domestic product SDGs Sustainable Development Goals GEM Global Economic Model SEEA System of Environmental-Economic Accounting GHG Greenhouse gas SEEA-CF System of Environmental-Economic Accounting GMW Global Mangrove Watch Central Framework GNI Gross national income SEEA-EA System of Environmental-Economic Accounting GWh Gigawatt hour Ecosystem Accounting HCI Human capital index SNA System of National Accounts IEA International Energy Agency UK United Kingdom IISD International Institute for Sustainable UN United Nations Development UNEP United Nations Environment Programme ILO International Labour Organization UNFC United Nations Framework Classification for IMF International Monetary Fund Resources IPCC Intergovernmental Panel on Climate Change UNFCCC United Nations Framework Convention on IRENA International Renewable Energy Agency Climate Change IUCN International Union for Conservation of Nature US United States kg Kilogram USD United States dollar km Kilometer WTP Willingness to pay LCU Local currency unit MAFF Ministry of Agriculture, Fisheries and Forestry MER Market exchange rates MW Megawatt NDP Net domestic product NPV Net present value 7 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TABLES AND FIGURES TABLES AND FIGURES Table A2.1: Example of two asset Törnqvist volume index Table 4.1: The elasticity of substitution (σ_2), by ecosystem service types Table 5.1: Percent change in nonrenewable natural capital per capita, by asset, region, and income group, 1995–2020 Table 6.1: Residential-to-producer price conversion factors by country/region Table 6.2: Hydroelectric asset value, installed capacity, generation, and weighted average capacity factor by region, 1995 and 2020 Table 6.3: Ten wealthiest hydroelectric countries, 2020 Table 6.4: Hydroelectric asset value per unit of installed capacity, 10 wealthiest hydroelectric countries, 1995 and 2020 Table 7.1: Land cover for forest and agricultural lands in low-income countries, 1995 and 2020 Table 8.1: Resource rent estimates for Norwegian salmon and trout aquaculture, selected years, 1995–2020 Table 9.1: Human capital as a share of total wealth, 1995–2020 Table 9.2: Human capital as a share of total wealth, by region, 1995–2020 Table 9.3: Potential gains in human capital from gender equity, 1995–2020 Figure ES.1: Distribution of global wealth in nominal terms, by income group, 1995 and 2020 Figure ES.2a: Trends in real wealth per capita, by region, 1995–2020 Figure ES.2b: Nominal wealth shares, by region, 2020 Figure ES.3: Countries with declining and non-declining real wealth per capita, 1995–2020 Figure ES.4a: Trends in global wealth per capita, by asset category, 1995–2020 Figure ES.4b: Nominal wealth shares, by asset category, 2020 Figure ES.5a: Trends in renewable natural capital per capita, by asset, 1995–2020 Figure ES.5b: Nominal wealth shares for renewable natural capital, by asset, 2020 Figure ES.6a: Trends in nonrenewable natural capital per capita, by asset, 1995–2020 Figure ES.6b: Shares in nominal wealth for nonrenewable natural capital, by asset, 2020 Figure ES.7a: Trends in produced capital per capita, by region, 1995–2020 Figure ES.7b: Nominal wealth shares for produced capital, by region, 2020 Figure ES.8a: Trends in human capital per capita, by region, 1995–2020 Figure ES.8b: Nominal shares of human capital, by region, 2020 Figure ES.9: Changes in global real GDP and wealth per capita, 1995–2020 Figure ES.10: Cumulative GDP per capita growth vs. cumulative wealth per capita growth, 1995–2020 Box Figure 1.1.1: GDP vs. wealth in the context of CWON Figure 1.1: Coverage of “natural capital” on the official SNA non-financial balance sheet for OECD countries Figure 1.2: Exchange value vs. welfare value Figure 1.3: Changes in prices and their implications for exchange and welfare values Figure 1.4: The CWON wealth accounts (2024 release) and their relationship to the SNA and SEEA Figure A1.1: Real wealth using different methods to deflate: the Törnqvist volume index, GDP deflator, and CPI REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 8 TABLES AND FIGURES TABLES AND FIGURES (CONT) Figure 3.1: Distribution of global wealth, by income group, 1995 and 2020 Figure 3.2: Real wealth per capita, by region, 1995–2020 Figure 3.3: Countries with declining and non-declining real wealth per capita, 1995–2020 Figure 3.4: Trends in real wealth per capita, by FCV status, 1995–2020 (1995=100) Figure 3.5: Wealth per capita, by asset category Figure 3.6: Change in wealth per capita, by income group and asset class, 1995–2020 Figure 3.7: Trends in renewable natural capital, by region, 1995–2020 (1995=100) Figure 3.8: Renewable natural capital per capita, by asset, 1995–2020 Figure 3.9: Nonrenewable natural capital per capita, by asset, index to 1995, 1995–2020 Figure 3.10: Produced capital per capita, by region, 1995–2020 Figure 3.11: Human capital per capita wealth, by region, 1995–2020 Figure 3.12: Shares of human capital by gender, 1995–2020 Figure 3.13: Changes in global GDP and wealth per capita, indexed to 1995, 1995–2020 Figure 3.14: Cumulative GDP per capita growth vs. cumulative wealth per capita growth, 1995–2020 Figure A3.1: Global wealth shares, 2020 Figure A3.2: GDP and wealth trends per capita in market exchange rates and PPP terms Figure A3.3: Trends in wealth asset classes in market exchange rates and in PPP terms, by income group Figure 4.1: A framework for introducing limited substitutability in production and consumption Figure 4.2: Manufacturing employment by land and elasticity of land-labor substitution Figure 4.3: Distribution of global wealth using differentiated discount rates, 1995 and 2020 Figure 4.4: Trends in wealth per capita using differentiated discount rates, indexed to 1995, 1995 and 2020 Figure 4.5: The value of non-wood forest ecosystem services, by the degree of substitutability Figure 5.1: Global comprehensive wealth, nonrenewable natural capital, and GDP in per capita terms, 1995–2020 Figure 5.2: Change in global nonrenewable natural capital per capita, by asset, 1995–2020 Figure 5.3: Shares of nonrenewable wealth, global, 1995–2020 Figure 5.4: Change in nonrenewable natural capital wealth per capita, 1995–2020 Figure 5.5: Change in fossil fuel and mineral wealth, top 10 countries, 1995–2020 Figure 5.6: Shares of global wealth, selected regions, 1995–2020 Figure 5.7: Shares of nonrenewable wealth, by region, 1995–2020 Figure 5.8: Change in fossil fuel and mineral wealth, by income group Figure 5.9: Change in fossil fuel and mineral wealth, by region group Figure 5.10: Global map of the change in fossil fuel and mineral wealth per capita, 1995 and 2020 Figure 5.11: Global trends in mineral wealth per capita, by selected minerals, 1995–2020 Figure 5.12: Percent changes in mineral wealth, selected countries, by mineral, 1995–2020 9 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TABLES AND FIGURES TABLES AND FIGURES (CONT) Figure 5.13: Changes in global wealth per capita for selected minerals: cobalt, lithium, molybdenum Figure 6.1: Hydroelectric asset value by region, 1995–2020 Figure 6.2: Hydroelectric asset value per capita by region, 1995–2020 Figure 6.3. Share of hydroelectric asset value in total energy asset values Figure 7.1: Land cover for forest and agricultural lands in low-income countries, 1995 and 2020 Figure 7.2: Trends in real wealth per capita in land assets, indexed to 1995, 1995–2020 Figure 7.3: Change in wealth per capita by land asset, 1995–2020 Figure 7.4: Combined predicted value of all recreational services, non-wood forest products, and water services, per hectare in 2020 (2020 US dollars) Figure 7.5: Share of forest ecosystem services by region and asset category, 2020 Figure 7.6: Wealth per capita in non-timber forest ecosystem services, by region and by category of service, 2020 Figure 7.7: Share of national forest wealth in non-timber forest ecosystem services provided by protected areas and country, 2020 Figure 7.8: Change in wealth per capita in land assets by income group, 1995–2020 Figure 7.9: Share of land wealth in total wealth in low-income countries, 1995 and 2020 Box Figure 7.2.1: Total global vegetation carbon stock (in megatons), 2020 Figure 8.1: Key steps and data for estimating the flood protection benefits provided by mangroves Figure 8.2: Differences in global mangrove cover between GMW 2.0 and GMW 3.0 Figure 8.3: Global marine fish wealth, indexed to 1995, 1995–2020 Figure 8.4: Per capita wealth from marine fish stocks, by income and region, 1995–2020 Figure 8.5: Wealth in marine fish stocks, top 20 countries, 2020 Figure 8.6: Countries with the top 20 largest loss of fish biomass per capita, 1995–2020 Figure 8.7: Egyptian tilapia/mullet aquaculture resource rent and sales value, 2010–2020 Figure 8.8: Resource rent estimates in selected countries and species, 1995–2020 Figure 8.9: Value of people and capital stock at risk and receiving risk reduction benefits in current US dollars, indexed to 1995, 1995–2020 Figure 8.10: Per hectare flood risk and mangrove benefits to people and capital stock by income level in nominal terms, 1995–2020 Figure 8.11: Mangrove area and the value of coastline protection services in real terms, indexed to 1995, 1995–2020 Figure 8.12: Mangrove coastline protection services value per capita, by income and region, indexed to 1995, 1995–2020 Figure 8.13: Countries with the largest mangrove coastline protection services value, 2020 Figure 9.1: Annual growth rates of human capital per capita, 1995–2020 Figure 9.2: Human capital per capita, by region, 1995–2020 Figure 9.3: Nominal shares of human capital, by region, 2020 Figure 9.4: Shares of human capital by gender, 1995–2020 Box Figure 9.3.1: Real human capital by education level Box Figure 9.3.2: Real human capital vs. quality-adjusted real human capital, indexed REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 10 EXECUTIVE SUMMARY Executive Summary WHY MEASURING CHANGES IN “Production and consumption” should be understood broadly THE WEALTH OF NATIONS MATTERS here to include things that humans benefit from every day, many of which are “produced” by nature. In this sense, a While many countries across the globe have experienced nation’s wealth includes a large portfolio of assets, ranging strong economic growth and improvements in human from clean air, to a local forest that allows for a walk in the development outcomes over the last quarter of a woods, to medical staff treating the sick in a hospital, to oil century, natural resources continue to be degraded and fields where hydrocarbons are pumped out of the ground. overexploited, calling the sustainability of that growth Changes in real wealth per capita should capture changes into question. The main yardstick typically used to assess in the real value of these assets and the production and economic progress is gross domestic product (GDP), which consumption opportunities they support. measures the level of domestic productive activity. Real GDP grew by more than 50 percent between 1995 and 2020, Strategic decisions to measure national wealth must accompanied by significant reductions in global poverty take account of considerable data and conceptual and improvements in educational and health outcomes.1 constraints. At present, it is not possible to measure all the However, natural resources, such as land, experienced assets that support a nation’s production and consumption widespread overexploitation and degradation during the opportunities. Consider the example of water. Measuring same period. 2 and valuing water is challenging and contentious due to its physical characteristics, the way it is regulated and used The world needs a broader metric to assess the sustainability within the economy, and the fact that it is an essential good.6 of economic development. As natural resources become Moreover, its value may be partly embedded in other assets, scarcer and reach critical levels due to climate change and such as agricultural land or hydropower. It is not feasible biodiversity loss, the growth potential of an economy and to account for the value of water. This is also true for many its resilience to shocks will be adversely affected. Yet GDP 3 other assets. does not measure such sustainability concerns. Sir Partha Dasgupta likened this to a soccer team that only measures In measuring wealth, methodological boundaries need success as goals for and ignores goals against. to be respected, but existing systems limit the extent of the balance sheet. Measures of wealth should be coherent A minimum requirement for sustainable development is and aligned with the internationally accepted accounting that real wealth per capita does not decline.4 If wealth can standards of the System of National Accounts (SNA) and the be measured comprehensively, per capita wealth changes System of Environmental-Economic Accounting (SEEA), will reflect the rise or fall of production and subsequent each of which restrict what can be measured within the asset consumption opportunities passed on to future generations.5 boundary. Natural resources such as land, forests, waters, or 1 See Kharas and Dooley (2022), UNESCO (2020), and data from the World Development Indicator, the Wittgenstein Center, and the UN Statistics Division. 2 See, for example, Cohen et. al (2019), Drupp and Hänsel (2021), and Rad et. al (2021). 3 For example, Smulders and van Soest (2023) show that with limited substitutability in factors of production economic growth is ultimately determined by the slowest-growing factor. Limited substitutability between natural assets and other factors of production, such as labor, can also shape economic resilience against natural shocks. Karayalcin and Onder (2023) show that the magnitude of the original impact (fragility) and the speed of the subsequent recovery (resilience) are determined by the ability of the economy to reallocate inputs between sectors, which will in turn be driven by their degree of substitutability as well as institutional characteristics, such as economic openness and property rights in natural assets. 4 Aligned with the definition of the World Commission on Environment and Development (1987). 5 There is a well-established literature in economics on sustainable development, including Solow (1993), Dasgupta and Maler (2000), Arrow et al. (2004), and Dasgupta (2021). 6 See Vardon et al. (2024) for a detailed discussion of how water could be integrated into wealth accounting. 11 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH EXECUTIVE SUMMARY wild animals are included in balance sheets provided that an Changes in CWON’s real wealth per capita metric can institutional unit exercises effective ownership rights over provide insights on the sustainability of economic them, and economically benefits from them. Resources progress, complementing GDP. CWON’s wealth measure such as the atmosphere, the high seas, or the mineral or fuel includes a broad portfolio of market and non-market assets deposits that cannot be extracted economically at present and can track how this asset base evolves over time. For are excluded. example, real wealth per capita will increase if more workers enter the labor force or if the same workers upgrade their The World Bank’s The Changing Wealth of Nations (CWON) skills, if forests grow, or new commercially recoverable program provides the most comprehensive wealth minerals are discovered. However, it will decline if fish database currently available. It builds on and goes beyond stocks are overfished, machinery degrades, or the reserves the asset boundary of the SNA and SEEA7 to better capture of fossil fuels are depleted. By monitoring per capita trends the changing future opportunities nations face. Most of in real GDP and real wealth together, it is possible to assess the assets covered in CWON—such as factories, intellectual whether growth in GDP is achieved by growing or shrinking property, urban land, and roads (produced capital); fossil the productive base of the economy. fuel, mineral, and metal reserves (nonrenewable natural capital); agricultural land, forests, and fish stocks (renewable This fifth edition builds on CWON’S tradition of incremental natural capital); and net foreign assets—are within the SNA improvements. Each update of its global database improves asset boundary. Others, like renewable energy assets, will the methodology and expands asset coverage. This edition be included in the SNA starting in 2025. Over time CWON8 further aligns nominal wealth estimates for key assets with has expanded the SNA asset boundary to include key SEEA SNA and SEEA concepts and valuation principles,14 and ecosystem accounts,9 such as non-timber forest ecosystem adds more metals and hydropower assets.15 Moreover, the services10 and shoreline protection services provided by real wealth estimates are now computed using a chained mangroves, as well as the value of investments in the labor 11 Törnqvist volume index instead of a price-based deflator in force (human capital). While there is currently not systematic line with international best practice.16 In this approach, the measurement for the latter within the SNA, the lifetime relative changes in the physical assets of a nation, such as income approach used by CWON is one of the statistical 12 the number of workers in the labor force, are weighted by approaches recommended as a wealth extension in the 2025 their relative economic importance as measured by their SNA update. 13 shares in nominal wealth.17 7 The United Nations Environment Programme (UNEP) is the second global initiative that is producing a database of wealth estimates (UNU–IHDP and UNEP 2012; UNU–IHDP and UNEP 2014; Managi and Kumar 2018; UNEP 2023). While there are similarities in country and temporal coverage, there are methodological differences, including in the valuation concepts, assets coverage, estimates of human capital, and assumptions used to estimate wealth. 8 Renewable energy assets are not yet part of the SNA. However, guidelines for their inclusion, which were developed for the previous CWON report (World Bank 2021), were endorsed as part of the 2025 SNA revision process and are available at: https://unstats.un.org/unsd/nationalaccount/aeg/2022/M21/M21_14_WS11_ Renewable_Energy_Resources.pdf. 9 Note that the asset boundary of the SEEA Central Framework is fully aligned with the SNA asset boundary, while the SEEA Ecosystem Accounting goes beyond it. 10 Some of these non-wood forest ecosystem services may be in the SNA (for example, non-wood forest products like mushroom harvesting), while others (such as recreation) are not. 11 The value of shoreline protection services may be partially attributed to property value in the SNA to the extent that housing prices account for climate risks. 12 This approach was developed by Jorgenson and Fraumeni (1989, 1992a, 1992b), and is used by several statistical offices to produce satellite accounts for human capital, such as Canada or the United Kingdom. 13 Guidelines on how the dimensions of education, human capital, and labor could be included in the SNA were developed and endorsed as part of the ongoing 2025 SNA revision process and are available at: https://unstats.un.org/unsd/nationalaccount/RAdocs/ENDORSED_WS4_Labour_Human_Capital_Education.pdf. 14 The goal of the methodological improvements is to align further with the internationally endorsed concepts and valuation principles from the SNA and SEEA. For example, the user cost of capital is now estimated directly in the resource rent calculations for nonrenewable natural capital and rent, and wage forecasts are removed from agricultural land and human capital valuations. 15 This edition adds cobalt, lithium, and molybdenum, increasing the CWON coverage from 10 to 13 metals and minerals. 16 In the national accounts space, all real measures (including real GDP) are in fact volume indexes expressed in the prices of a reference year. The new approach is very similar to the methods currently used for real GDP in advanced economies like Canada or the United States (except that they use a slightly different index form, a chained Fisher volume index). 17 This is true for composite asset portfolios, for example, when computing total real wealth including all assets, or for the main asset categories, but not single homogenous assets. Chapter 2 provides more detail on the implementation of the chained Törnqvist volume index. For more technical details on the properties of the Törnqvist index, refer to Dumagan (2002) and Diewert (1992). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 12 EXECUTIVE SUMMARY GLOBAL TRENDS IN REAL WEALTH Trends in real wealth per capita are crucial for assessing PER CAPITA the sustainability of economic progress. Real wealth per capita has grown in all regions due to significant increases Wealth remains highly concentrated in rich countries, in human and produced capital (Figure ES.2a), which were with significant disparities persisting across income driven by rapid urbanization and the growing number groups. High-income countries make up more than two- of women participating in the labor market. The growth thirds of total wealth in nominal terms in 2020, while trend is particularly pronounced in the Middle East and upper-middle-income countries constitute nearly a quarter North Africa region (97 percent) and Latin America and the (Figure ES.1, Panel b). Caribbean (66 percent). This growth is largely attributed to substantial rises in human capital (82 percent and 61 Low-income and lower-middle-income countries, home to percent, respectively) and produced capital (138 percent half of the world’s population, account for just 7 percent and 83 percent, respectively) in these regions, but has not of global wealth. There is no evidence that this wealth gap resulted in substantial increases in their share in global has been closing: while upper- and lower-middle-income wealth (which stood at 4 percent and 3 percent, respectively, countries were able to nearly double their share in global in 2020). wealth, the wealth share of low-income countries has largely remained below 1 percent since 1995. FIGURE ES.1 Distribution of global wealth in nominal terms, by income group, 1995 and 2020 Panel a: Global wealth, 1995 Panel b: Global wealth, 2020 4% 6% 11% 0% 1%  High income 23%  Upper middle income  Lower middle income  Low income 85% 70% Source: World Bank staff estimates. Note: Wealth in nominal terms is measured in current US dollars. Global wealth per income group is computed as the sum of nominal wealth for all countries in that income group for 1995 and 2020, respectively. Wealth shares are then computed relative to the global sum of wealth across all countries in the sample for 1995 and 2020, respectively, and are reported in percent. 13 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH EXECUTIVE SUMMARY FIGURE ES.2A FIGURE ES.2B Trends in real wealth per capita, by region, 1995–2020 Nominal wealth shares, by region, 2020 (1995=100) 3% 3% 4% capita in chained 2019 USD (1995=100) 200 National comprehensive wealth per 2% 32% 150 27% 100 29% 50 1995 2000 2005 2010 2015 2020  North America  East Asia & Pacific  North America  East Asia & Pacific  Europe & Central Asia  Europe & Central Asia  Latin America & Caribbean  Latin America & Caribbean  South Asia  Middle East & North Africa  South Asia  Sub-Saharan Africa World  Middle East & North Africa  Sub-Saharan Africa Source: World Bank staff estimates. Note: Real wealth per capita is computed using the Törnqvist volume index. For the relative volume changes, chained Törnqvist volume indexes are used for produced capital, nonrenewable and renewable natural capital, human capital, and net foreign assets (for net foreign assets, the nominal asset value is deflated using consumer price index (CPI)). The weights are calculated using their respective nominal asset value relative to nominal wealth. The Törnqvist volume index for wealth is then chained with a base year of 2019 and real wealth is computed using the nominal wealth estimate for 2019. Regional wealth per capita is computed as a sum of real wealth for the countries in the region divided by the regional population. Changes in real wealth per capita for each region are reported relative to 1995 (set equal to 100). Nominal wealth is measured in current US dollars and shares are reported in percent. The accumulation of real wealth in some regions, notably While two-thirds of the 151 countries in the sample Sub-Saharan Africa, has not grown at the same speed as experienced growth in real wealth per capita between 1995 their respective populations. While there were periods of and 2020, 27 countries experienced declines or saw little growth in real wealth per capita in Sub-Saharan Africa, change. The decline in low-income countries signals that such as between 2000 and 2005, there has been a stagnating economic progress is unsustainable. One possible driver of trend since then, with the region making up just 2 percent these observed trends could be conflict, as 40 percent of these of global wealth. countries are also classified by the World Bank as affected by fragility, conflict, and violence (FCV). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 14 EXECUTIVE SUMMARY FIGURE ES.3 Countries with declining and non-declining real wealth per capita, 1995–2020 Per capita change (%) ■ Declining wealth ■ Non-declining wealth ■ No data Source: World Bank staff estimates. Note: Percent changes in real wealth per capita are computed for the 1995–2020 period. This figure distinguishes between countries that experienced declining wealth (percent change < 0) vs. non-declining wealth (percent change >= 0) during this period, but does not compare wealth per capita nominal values directly. As discussed throughout the report, non-declining wealth per capita is a minimum requirement for sustainable economic growth, although not a sufficient condition. WHAT IS DRIVING CHANGES 1995 and 2020. Nonrenewable natural assets slightly declined IN REAL WEALTH PER CAPITA? in per capita terms over the same period. This is the most volatile asset category, affected by changes in the underlying The trends in real wealth per capita are driven by changes asset base, technological innovations, and price fluctuations. in the asset portfolio relative to population growth, with starkly different trends across asset categories. There are Renewable natural capital, which should be able to four main asset categories that make up CWON’s wealth regenerate itself if managed sustainably, has declined measure: produced capital, nonrenewable natural capital, by more than 20 percent in per capita terms over the renewable natural capital, and human capital. Human past quarter of a century. It is important to note that this capital, which makes up the largest share of nominal wealth trend and its 6 percent share in global wealth is likely an (Figure ES.4b), has increased by about 9 percent in per capita underestimate, as data and conceptual concerns limit the terms relative to 1995, as shown in Figure ES.4a. Produced ability to measure and value this component. capital increased by 47 percent in per capita terms between 15 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH EXECUTIVE SUMMARY FIGURE ES.4A FIGURE ES.4B Trends in global wealth per capita, by asset category, Nominal wealth shares, by asset 1995–2020 (1995=100) category, 2020 160 6% 140 in chained 2019 USD (1995=100) 2% Wealth per capita 120 32% 100 60% 80 60 1995 2000 2005 2010 2015 2020  Human capital  Renewable natural capital  Produced capital  Produced capital  Nonrenewable natural capital  Human capital  Renewable natural capital  Nonrenewable natural capital Source: World Bank staff estimates, Note: Real wealth per capita by asset category is computed using the Törnqvist volume index. For the relative volume changes, physical measurements of the assets in each asset category are used. The weights are calculated using their nominal asset value relative to the nominal value of their respective asset category. The Törnqvist volume index for each asset category is then chained with a base year of 2019 and the real value of each asset category is computed using the nominal value of each asset category for 2019. The global real value of each asset category (measured in chained 2019 US dollars) is computed by summing across all countries and dividing by the global population. Changes in real comprehensive wealth per capita for each asset category are reported relative to 1995 (set equal to 100). Nominal wealth is measured in current US dollars and shares are reported in percent. There is a significant variation in the level of decline natural capital components—such as agricultural land, which in renewable natural capital wealth per capita across is the most important component of renewable natural regions and assets. Sub-Saharan Africa and the Middle capital (73 percent of its global value), and non-timber East and North Africa region have experienced the largest forest recreation ecosystem services (12 percent)—have declines (around 40 percent), with South Asia losing about experienced similar though less dramatic declines. Notably, a third. These declines are driven by population growth renewable energy from hydropower—a new addition to and overexploitation across almost all renewable natural CWON—experienced an increase of 23 percent over the same resources included in this report, such as forests, marine fish 25-year period, making up 7 percent of the overall value of stocks, and mangroves. The value of marine fish stocks has renewable natural capital. However, renewable energy assets experienced the most dramatic decline, dropping by more such as solar, wind, and geothermal assets were not included than 45 percent since 1995 (Figure ES.5a). Other renewable in this edition due to data limitations. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 16 EXECUTIVE SUMMARY FIGURE ES.5A FIGURE ES.5B Trends in renewable natural capital per capita, Nominal wealth shares for renewable by asset, 1995–2020 (1995=100) natural capital, by asset, 2020 2% 120 4% 6% in chained 2019 USD (1995=100) 1% 7% 100 Wealth per capita 7% 32% 80 73% 60 1995 2000 2005 2010 2015 2020  Agricultural land  NWFP  Agricultural land  Recreation ES  Water ES  Mangroves  Water ES  Mangroves  Hydropower  Recreation ES  Hydropower  NWFP  Timber  Fish stocks  Timber  Fish stocks Source: World Bank staff estimates. Note: Real wealth per capita for each asset is computed using the Törnqvist volume index. The relative volume changes use physical measurements of agricultural land (in square km), timber (in hectares), mangroves (in hectares), non-wood forest ecosystem services (in square km), hydropower (in GWh), and fish stocks (in tons). No weighting is used for measuring individual assets in real terms. The Törnqvist volume index for each asset is then chained with a base year of 2019 and the real value of each asset is computed using the nominal asset value for 2019. The global real value of each asset (measured in chained 2019 US dollars) is computed as the sum of real wealth for each asset divided by the global population. Changes in the real asset value per capita are reported relative to 1995 (set equal to 100). ES = ecosystem services. NWFP = non-wood forest products. Nominal wealth is measured in current US dollars and shares are reported in percent. Globally, nonrenewable natural capital—spanning oil, Rapid urbanization and industrialization in high-income natural gas, coal, and metals and minerals—decreased by and emerging economies have led to substantial growth 2.5 percent in per capita terms between 1995 and 2020, in produced capital wealth. On average, there is about with a small increase in oil wealth offset by declines in coal, 47 percent more produced capital per capita in the world natural gas, and minerals (Figure ES.6a). The low-carbon than there was in 1995, and it has accumulated faster transition is likely to affect these estimates in the short to than population growth in all the regions (Figure ES.7a). medium term. However, large decreases in carbon-intensive Produced capital per capita in South Asia has experienced fossil fuels (except for coal) have not yet been observed, an astonishing expansion, increasing by nearly 500 percent, which still make up nearly 60 percent of the global value of albeit from a very low level. Most of the produced capital nonrenewable natural capital (Figure ES.6b). assets are concentrated in North America, Europe, and East 17 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH EXECUTIVE SUMMARY FIGURE ES.6A FIGURE ES.6B Trends in nonrenewable natural capital per capita, by Shares in nominal wealth for asset, 1995–2020 (1995=100) nonrenewable natural capital, by asset, 2020 120 8% in chained 2019 USD (1995=100) 14% Wealth per capita 100 43% 80 36% 60 1995 2000 2005 2010 2015 2020  Oils  Coal  Oils  Natural gas  Minerals Nonrenewables  Minerals  Coal  Natural gas Source: World Bank staff estimates. Note: Real nonrenewable natural capital per capita is measured in chained 2019 US dollars. For details on the computation of real wealth per capita per asset refer to the notes for Figure ES.5. The relative volume changes use physical measurements of oil (in barrels), gas (in terajoules), coal (in tons), and minerals (in tons). Nominal wealth is measured in current US dollars and shares are reported in percent. Asia, which make up 94 percent of the global value and have Human capital, which accounted for 60 percent of the experienced lower but steady growth rates. Sub-Saharan world’s total wealth value in 2020, has grown consistently Africa experienced the lowest growth rates. Although the for the past 25 years due to increasing labor force region made significant strides in accumulating produced participation and higher returns to education. The share of capital wealth—matching that of other regions with a 129 human capital in total wealth (measured in nominal terms) percent increase over 25 years—rapid population growth led generally increases as countries achieve higher levels of to a modest 17 percent increase from 1995 to 2020. economic development. The world’s real human capital per capita increased by 9 percent between 1995 and 2020, but there are contrasting trends across regions (Figure ES.8a).19 19 The growth rate in global human capital per capita is so low, since there have been small increases in the share of global human capital in South Asia (from 2.6 percent in 1995 to 3.2 percent in 2020) and Sub-Saharan Africa (from 1.1 percent to 1.8 percent). While these changes are small in absolute terms, they are able to depress the global growth rate in human capital per capita due to significant differences in the value of human capital relative to richer regions. For example, North America has a nominal value of human capital that is 60 times higher than in South Asia and Sub-Saharan Africa. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 18 EXECUTIVE SUMMARY FIGURE ES.7A FIGURE ES.7B Trends in produced capital per capita, by region, Nominal wealth shares for 1995–2020 (1995=100) produced capital, by region, 2020 1% 600 3% in chained 2019 USD (1995=100) 500 1% Produced capita per capita 400 27% 38% 300 200 100 0 29% 1995 2000 2005 2010 2015 2020  North America  North America  East Asia & Pacific  East Asia & Pacific  Europe & Central Asia  Latin America & Caribbean  Europe & Central Asia  South Asia  Middle East & North Africa  Latin America & Caribbean  Sub-Saharan Africa World  Middle East & North Africa  Sub-Saharan Africa  South Asia Source: World Bank staff estimates. Note: Real produced capital per capita is measured in chained 2019 US dollars. For details on the computation of real wealth per capita for produced capital refer to the notes for Figure ES.5. For the relative volume changes, the following data are used: capital stock estimates from the Penn World Table 10.0 and urban land area estimates based on World Bank staff estimates using data from the United Nations Population Division’s World Urbanization Prospects, the Food and Agriculture Organization (FAO), and the Center for International Earth Science Information Network. The weights are calculated using their nominal asset value relative to the nominal value of produced capital. Nominal wealth is measured in current US dollars and shares are reported in percent. Human capital is concentrated in the high- and upper-middle- Estimates show a significant disparity between the male income countries of North America, Europe and Central and female shares of human capital. Unfortunately, little Asia, and East Asia and the Pacific, as shown in Figure ES.8b. progress was made toward greater gender parity in human These regions have experienced modest growth rates, with, capital between 1995 and 2020. Globally, women accounted for for example, 12 percent in North America and 16 percent in only 37 percent of human capital in 2020—only 2 percentage East Asia and the Pacific. In contrast, the Middle East and points up from 1995. Although higher levels of economic North Africa, and Latin America and the Caribbean regions development are generally associated with a higher share show much larger increases of 82 percent and 62 percent, of women in human capital, women account for less than respectively, over the same period, albeit from a much lower 40 percent of human capital at all levels of development. starting point. The differences between regions are even more striking. 19 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH EXECUTIVE SUMMARY FIGURE ES.8A FIGURE ES.8B Trends in human capital per capita, by region, Nominal shares of human capital, 1995–2020 (1995=100) by region, 2020 2% 190 3% 2% 5% 180 in chained 2019 USD (1995=100) 170 34% Human capita per capita 160 150 140 27% 130 120 110 100 27% 90 1995 2000 2005 2010 2015 2020  North America  North America  East Asia & Pacific  Europe & Central Asia  Europe & Central Asia  Latin America & Caribbean  East Asia & Pacific  South Asia  Middle East & North Africa  Latin America & Caribbean  Sub-Saharan Africa World  South Asia  Middle East & North Africa  Sub-Saharan Africa Source: World Bank staff estimates. Note: Real human capital per capita is measured in chained 2019 US dollars. For details on the computation of real wealth per capita for human capital refer to the notes for Figure ES.4. For the relative volume changes, labor force numbers disaggregated by gender from the International Labor Organization (ILO) are used, which are scaled by the human capital index from the Penn World Tables to proxy for the average human capital per worker. The weights are calculated using their nominal asset value relative to the nominal value of human capital. Nominal wealth is measured in current US dollars and shares are reported in percent. For example, in South Asia, women represented 15 percent one have so far been able to offset the erosion of the asset of human capital in 2020, marking a 2 percent decline from base, most notably with regards to renewable natural 1995. In contrast, 44 percent of human capital was attributed capital. It remains an open question how long this trend to women in Latin America and the Caribbean. South Asia’s can continue considering that natural capital continues to large gender gap is mostly caused by a male-dominated labor be overexploited and is becoming scarcer. The extent to force and barriers that prevent women from attaining similar which limited substitutability could be accounted for within economic opportunities as men (World Bank 2023b). CWON through, for example, relative price adjustments or differentiated discount rates is further explored in chapter While the different asset components of wealth have been 4. Indicative estimates suggest that such adjustments would on starkly different trajectories, this has so far not yet substantially increase the share of renewable natural capital acted as a brake on growth. Rising productivity and the in overall wealth. ability to substitute one scarcer asset for a more abundant REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 20 EXECUTIVE SUMMARY TRENDS IN REAL WEALTH A more granular look at the data reveals that global wealth AND GDP PER CAPITA trends mask large and persistent differences across income groups and FCV status. Rich countries are becoming Globally, real wealth per capita increased by about 21 wealthier, while poor and conflict-affected nations are in a percent between 1995 and 2020. This contrasts with the downward spiral of low growth and wealth depletion. This observed increase in real GDP per capita of about 50 percent is further illustrated in Figure ES.10, which maps changes in over the same period (Figure ES.9). Changes in real wealth real GDP per capita to changes in real wealth per capita. While most countries are experiencing growth in both real GDP per capita are driven by changes in the real asset base, and wealth per capita, 15 percent of countries are currently capturing the accumulation and depletion of assets over experiencing positive GDP per capita growth rates while time. That is, real wealth per capita will decline as capital is their real wealth per capita declines. For these countries, it used up, degraded, or destroyed in the process of generating will be critical to continue building their asset base to ensure output. Current GDP, on the other hand, often increases a sustainable growth path. The rest of the countries either when asset depletion accelerates; for example, when forest show declines in both GDP and real wealth per capita or is clearcut and timber is sold. CWON provides researchers did not experience growth in GDP per capita but appear to and analysts with the most comprehensive, transparent, and accumulate real wealth per capita.20 These observed trends rigorous global data time series of the wealth of nations to warrant a more detailed empirical analysis to explore what is conduct such analysis. driving these differences. FIGURE ES.9 Changes in global real GDP and wealth per capita, 1995–2020 (1995=100) 160 Index 1995=100 140 120 100 1995 2000 2005 2010 2015 2020  GDP per capita  Wealth per capita Source: World Bank staff estimates. Note: Real GDP per capita is measured in constant 2015 US dollars from the World Bank’s World Development Indicators database (NY.GDP.PCAP.KD) and real wealth per capita is measured using the Törnqvist volume index and is reported in chained 2019 US dollars. 20 This group of countries includes Jamaica, Kuwait, Madagascar, Oman, the Solomon Islands, and Zimbabwe. 21 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH EXECUTIVE SUMMARY FIGURE ES.10 Cumulative GDP per capita growth vs. cumulative wealth per capita growth, 1995–2020 100 Wealth per capita growth, 1995–2020 (%) 50 0 -50 -50 0 50 100 GDP per capita growth, 1995–2020 (%) Source: World Bank staff estimates. Note: The sample was restricted to growth rates less than 100 percent. The scatter plot shows country codes. Countries in green have increasing GDP and wealth per capita, countries in blue have declining GDP per capita but increasing wealth per capita, countries in orange have increasing GDP per capita but declining wealth per capita, countries in red have declining GDP and wealth per capita. CONCLUSIONS wealth estimates of the CWON database will be publicly released on the World Bank’s website. This unprecedented CWON’s estimates of real comprehensive wealth per access will provide users with the opportunity to use more capita can be used to assess the sustainability of a nation’s granular, country-level input data and modify assumptions as economic progress and complement GDP. By producing needed to support their own sustainability analysis. these estimates, the World Bank addresses an important data gap, as all countries produce GDP, but few produce CWON’s long-term ambition is to support the analysis of wealth estimates. Importantly, this edition offers countries sustainability. At this stage, the extent to which this ambition the possibility to construct customized national-level wealth has been fulfilled or even exceeded is an open analytical estimates by building on the CWON methodology. As part of question. The analysis of sustainability is unavoidably the World Bank’s reproducibility initiative, the entire statistical constrained by which assets are included on the CWON balance code and input data used to generate the nominal and real 21 sheet and the precision with which they are measured.22 21 For licensed data, dummy datasets will be made available. 22 For a detailed discussion of these issues, refer to chapter 1. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 22 EXECUTIVE SUMMARY Over time, the assets included have progressively expanded between progressive expansion and the stability of the and the measurement has considerably improved. While asset boundary. A common challenge for the statistical and it is important to acknowledge that important gaps in economic communities will be to explore viable approaches the coverage remain due to data and measurement to assigning monetary values for assets that provide constraints, the CWON balance sheet is nonetheless the essential economic services to humans in the context of most comprehensive wealth database available today, in heavily distorted or missing markets. terms of coverage of assets, countries, and time series, aligned where possible with the internationally accepted Lastly, it is important to note that CWON provides baseline statistical standards and guidelines in the SNA and SEEA. estimates in line with endorsed statistical guidance. This alignment not only ensures methodological rigor, but This means the wealth estimates reflect the current policy also coherence with standard economic measures like GDP. environment and market expectations and do not account for possible impacts of future policy actions or changes While this edition has implemented several critical in market conditions due to, for example, climate change. methodological innovations, more work is needed to To explore these questions and “what if ” scenarios, advance CWON as a regular, ongoing statistical program. researchers will have the opportunity to adapt the CWON To achieve this, it will be critical to take a more systematic input data and source code to change assumptions or adopt approach to defining the asset boundary for the wealth alternative methodologies in line with their requirements. estimates based on the assets that fall within the SNA and This will enhance use applications, which can range SEEA boundaries, with appropriate extensions (for example, from customized wealth estimates at the country level to for human capital). This will enable the CWON program projections of policy-contingent scenarios, as illustrated in to stay aligned with the SNA and SEEA standards, while CWON 2021 with simulations for fossil fuel and renewable also going beyond them in meaningful and appropriate energy assets. CWON thus serves as a flexible, transparent, ways. Future updates should aim to selectively expand and reproducible database available for stress-testing and the boundaries, for example, to also include the value of empirical research. renewable energy resources, while maintaining a balance 23 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH EXECUTIVE SUMMARY REFERENCES Dasgupta, P. 2001. Human Well-Being and the Natural Environment. New York: UN (United Nations) et al. 2021. System of Environmental-Economic Oxford University Press. Accounting—Ecosystem Accounting (SEEA EA). New York: White cover publication, pre-edited text subject to official editing. Dasgupta, P. 2021. The Economics of Biodiversity: The Dasgupta Review. https://seea.un.org/ecosystem-accounting. London: HM Treasury. UNECE (United Nations Economic Commission for Europe). 2016. European Commission, IMF (International Monetary Fund), OECD Guide on Measuring Human Capital. Geneva: United Nations. (Organisation for Economic Co-operation and Development), UN https://unece.org/statistics/publications/guide-measuring-human-capital. (United Nations), and World Bank. 2009. System of National Accounts 2008. UNEP (United Nations Environment Programme). 2023. Inclusive Wealth Hamilton, K. 1999. “Genuine Savings Rates in Developing Countries.” Report 2023: Measuring Sustainability and Equity. Nairobi: UNEP. The World Bank Economic Review 13 (2): 333–356. https://wedocs.unep.org/handle/20.500.11822/43131#:~:text=The%20 report%20highlights%20how%20inclusive,of%20the%20Inclusive%20 Hartwick, J.M. 1978. “Substitution among Exhaustible Resources and Wealth%20series. Intergenerational Equity.” The Review of Economic Studies 45 (2): 347–354. UNESCO (United Nations Educational, Scientific and Cultural Jorgenson, D.W. 2018. “Production and Welfare: Progress in Economic Organization). 2020. Global Education Monitoring Report – Gender Report: Measurement.” Journal of Economic Literature 56 (3): 867–919. A New Generation: 25 Years of Efforts for Gender Equality in Education. UNESCO: Paris. https://unesdoc.unesco.org/ark:/48223/pf0000374514/ Kharas, H., and Dooley, M. 2022. “The Evolution of Global Poverty, PDF/374514eng.pdf.multi. 1990–2030.” Brookings Global Working Paper #166. Center for Sustainable Development at Brookings. UNU–IHDP and UNEP. 2012. Inclusive Wealth Report 2012 – Measuring Progress Toward Sustainability. Cambridge: Cambridge University Press. Solow, R. 1993. “An Almost Practical Step Towards Sustainability.” http://www.unep.org/pdf/IWR_2012.pdf. Resource Policy 19 (3): 162–172. UNU–IHDP and UNEP. 2014. Inclusive Wealth Report 2014 – Measuring Stiglitz, J.E., Sen, A., and Fitoussi, J.P. 2010. Mis-measuring Our Lives: Progress Toward Sustainability. Cambridge: Cambridge University Press. Why GDP Doesn’t Add Up. New York: The New Press. http://mgiep.unesco.org/wp-content/uploads/2014/12/IWR2014-WEB.pdf. UN (United Nations), EU (European Union), FAO (Food and Agriculture Vardon, M., Balasubramanya, S., Karimi, P., and Onder, S. 2024. Adding Organization of the United Nations), IMF (International Monetary Fund), Water to the Changing Wealth of Nations. OECD (Organisation for Economic Co-operation and Development), and World Bank. 2014. System of Environmental-Economic Accounting 2012—Central Framework. United Nations. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 24 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Tracking Wealth I. to Monitor Economic Progress 25 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS 1 Wealth as an Indicator of Sustainability MAIN MESSAGES ■ CWON aspires to measure wealth comprehensively, and its measurement has gradually expanded with ■ The world seeks a headline indicator to help assess each edition as new data and statistical standards aggregate progress on the 17 Sustainable Development have become available, and more sophisticated Goals (SDGs) and sustainability more broadly. methodologies have been adopted. The current coverage includes key assets from the SNA balance ■ To achieve such a measurement, a robust discussion is sheet and critical ecosystem assets covered by the needed about expanding national economic statistics SEEA Ecosystem Accounts. CWON also goes beyond “beyond GDP” (gross domestic product). Despite GDP’s the current SNA standards to include human capital. important role in measuring production and income, it is widely recognized to be an insufficient measure of progress and national sustainability. INTRODUCTION ■ Whether progress is sustainable, that is, whether There is no more evocative phrase in economics than “the future generations will have at least the same wealth of nations.” Indeed, some would argue these were production and consumption opportunities as the the first words written in the field. Despite Adam Smith current generation, can be assessed by looking coining the phrase nearly 250 years ago, economists continue at changes in real wealth per capita. Constant to struggle to measure the wealth of nations and how it or increasing wealth per capita—measured changes through time. This challenge facing economists comprehensively to include produced, human, and and statisticians today is much like the challenge mariners natural capital, as well as net foreign assets—is thus an once faced in measuring longitude at sea—yet measurement important indicator of sustainability. of where you are, and how your position is changing, is important for “steering the ship.” ■ While almost all countries measure GDP, few countries produce wealth measures. Those that do generally do Decision-makers have long put sustainability and the not measure wealth comprehensively by including possible enhancement of economic welfare at the center of natural and human capital. the policy discourse. This includes the many national and international conversations around the SDGs23 the beyond ■ The Changing Wealth of Nations (CWON) database GDP agenda,24 and the importance of mainstreaming nature aims to fill this data gap by producing a comparable into decision-making processes.25 To inform such decisions and consistent measure of change in real wealth requires evidence and data, which ultimately depend on per capita for 151 countries. These estimates are some form of measurement. aligned, where possible, with the System of National Accounts (SNA) and the System of Environmental- Economic Accounting (SEEA) balance sheets to ensure comparability with other standard macroeconomic measures, such as GDP. 23 https://sdgs.un.org/goals. 24 https://unsceb.org/topics/beyond-gdp. 25 See, for example, the G7 Environment Ministers’ Communiques of 2023 (https://www.meti.go.jp/press/2023/04/20230417004/20230417004-1.pdf) and the G7 Science Ministers’ Communique in 2023 (https://www8.cao.go.jp/cstp/kokusaiteki/g7_2023/230513_g7_communique.pdf), as well as the Global Biodiversity Framework (https://www.cbd.int/gbf/targets/14/). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 26 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS The World Bank’s CWON program is one of the pioneering THE IMPORTANCE OF THE CHANGE efforts in this field, endeavoring to produce a global database IN WEALTH IN THE MEASUREMENT of comparable and consistent estimates of comprehensive OF SUSTAINABILITY wealth26 for nearly two decades. These monetary estimates draw on internationally endorsed concepts and valuation Policy definitions of sustainable development usually principles from the SNA and related standards. They have start with the World Commission on Environment and been updated and expanded with each new edition as new Development (1987) definition: “meeting the needs of the data sources, measurement techniques, and guidance current generation without compromising the ability of have become available. CWON aspires to measure wealth future generations to meet their needs.”28 Economists have as comprehensively as possible with each iteration, while developed two complementary measurement approaches to acknowledging that there are still important forms of wealth put this definition into practice: a wealth- and an income- that have not yet been accounted for. based approach. The first approach focuses on assessing changes in the state of the world by measuring changes The CWON report series has elevated changes in real in stock values like real wealth per capita. As long as real wealth per capita, which helps measure whether economic wealth per capita does not decline, welfare per capita does progress is sustainable, as a complementary metric to GDP. not decline, which is a necessary condition for development Changes in real wealth per capita over time, if measured to be sustainable (Arrow et al. 2004; Arrow, Dasgupta, and comprehensively, are proportional to changes in welfare Maler 2003; Dasgupta 2001). The second approach, well (Arrow et al. 2004; Dasgupta 2001; Polasky et al. 2015).27 A synthesized by Sefton and Weale (2006), focuses instead on minimum requirement for sustainable development is that the flow measures for consumption or real income, which welfare, and thus comprehensive wealth, does not decline. could be thought of as a modified version of per capita net Measuring change in comprehensive wealth is therefore domestic product, or net national income. Provided these an essential building block in measuring and managing flow measures are not negative, welfare per capita does not sustainable economic development. decline either, and development is sustainable. In theory, the flow measures should balance the change in the stock This chapter lays out why it is important to measure changes measure, though this is not necessarily the case in practice in real wealth per capita and the theory of using this metric due to differences in scope and measurement challenges. to assess sustainability. Next, it discusses how progress can be measured within the current context of the SNA and CWON uses the economic measure of capital to assess wealth, the extent to which national statistical offices currently (do emphasizing the importance of balance sheet measures that not) record changes in real wealth. It then discusses the are often missing from official national statistics. Wealth importance of an appropriate asset boundary, and price and in this context can either be an input into production or a value concepts in developing wealth estimates. The final two store for future production or consumption opportunities. sections discuss how the CWON program implements these For example, cash savings can be used in the future to pay in the context of a global wealth accounting exercise and the workers or procure raw materials. Similarly, natural capital, extent to which these choices align with the SNA. It concludes such as land fertility, standing forests, and fish stocks, can with a summary of how the current CWON wealth measure be used today or saved for future production.29 To value can be interpreted. capital or wealth, one needs to compute the expected net present value of future income that this capital generates as comprehensively as possible. For example, harvesting 26 Comprehensive wealth estimates include produced, human, renewable natural, and nonrenewable natural capital as well as net foreign assets. 27 This is even the case when welfare is not maximized in a second-best world (Lipsey and Lancaster 1956). 28 See also Pezzey (1992) and Solow (1993). 29 In the case of renewable resources, the resource stock can also grow on its own, increasing in quantity and potentially total value over time. 27 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS fish is not purely the production from fishing labor and maintained with the same combination of land, labor, and fishing boats30—it also involves production by the natural machinery used today. This definition effectively rules out capital itself, in this case the fish stock. Foregoing a fish substitution opportunities across and within asset classes in harvest today could increase production and, ultimately, the present, or those that may emerge through innovations consumption opportunities for tomorrow. Wild fish stocks in future. From a measurement perspective, a society would should therefore be treated as wealth (Fisher 1906) and only pass a sustainability test if all enumerated capital stocks included in any comprehensive measurement of capital. 31 do not physically decline.32 The other end of the spectrum is known as “weak sustainability,” which is often interpreted The change in wealth is an important sustainable development as accepting the possibility of endless substitution or indicator, as it reflects the rise or fall of production and innovation.33 In that case, development is weakly sustainable, subsequent consumption opportunities passed on to future as long as overall wealth (per capita) increases. For example, generations relative to those of the current generation (Arrow a decline in natural capital can be compensated for through et al. 2004). As long as future production and consumption investments in produced or human capital. opportunities are not diminished relative to those of the present, development may be sustainable. Production and The reality is likely somewhere in between these two definitions, consumption should be understood broadly here to include where some substitution and innovation possibilities exist, the non-market goods and services humans benefit from but where there are also complementary relationships, every day, many of which are “produced” by nature, such as especially between natural and produced capital (see Cohen breathable air, clean water, food, weather regulation, and et al. 2019; Rouhi Rad et al. 2021). Take the example of fish recreation. A nation’s wealth includes a large portfolio of (natural capital) and fishing boats (produced capital). If fish assets, ranging from a local forest, which contributes to the and fishing boats were perfect substitutes, then the scarcity of service of a “walk in the woods,” to the medical staff treating fish would not impede sustainable growth if the production of the sick in a hospital, and the oil field where crude oil is boats increased sufficiently. But, in reality, these are imperfect pumped out of the ground. Changes in real wealth per capita substitutes (and in fact are complementary), which means are expected to capture changes in the real value of these the scarcity of fish would further impede economic growth. assets and the production and consumption opportunities A practical solution is offered by Barbier (2011), who suggests they support. measuring actual substitution or “capital sustainability.” Building on Hicks’ concept of real income, the idea is that Yet, how those production (and, ultimately, consumption) changes in real wealth are measured after accounting for, and opportunities change depends on the exact definition of removing, measured substitution effects. That is, changes in sustainability used and, in particular, the extent to which production or consumption due to substitutions need to be substitution opportunities are assumed to vary over time. separated from those that are truly changes in opportunities, On one end of the spectrum, the “strong sustainability” which should then be recorded as changes in real wealth definition implies that future generations need at least the associated with economic income effects. same physical quantities of capital, at least up to broad asset classes like natural and produced capital, as generations Changes in real wealth per capita, if measured comprehensively, today (Dietz and Neumayer 2007). Under such assumptions, are then proportional to changes in welfare, if appropriate agricultural production, for example, could only be prices are used, as shown by Dasgupta (2001).34 30 Production measurements such as GDP attribute all the value added to the labor of fishermen and the returns of fishing capital like boats. 31 Indeed, when fish stocks are managed, they should be recorded on the internationally agreed SNA non-financial balance sheets (European Commission et al. 2009). However, few countries actually do so. Furthermore, Fenichel and Abbott (2014) show that even when fisheries are not rationalized there is a positive marginal value (price) of a fish stock that can be imputed from fishery rents and ecological information in a way consistent with the traditional production boundary. 32 In the strictest sense, this definition rules out the use of any nonrenewable resource. This creates a contradiction because by eliminating all use of nonrenewables, there is no reason not to use them up. 33 This is often operationalized as a partial equilibrium saving rule that is linear in exogenous prices (Pearce and Atkinson 1993). 34 Dasgupta caveats this by pointing out that per capita measures are important when populations are changing. His approach generally aligns with other attempts to develop theoretical measures of changes in national income (for example, Sefton and Weale 2006). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 28 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Limits to substitution and even complementary relationships when the world realized that it needed a systematic approach are reflected in how prices respond to changing scarcity (Yun for measuring economic cycles and mobilizing industrial et al. 2017). The challenge lies in separating the income and output. This desire has evolved into an elaborate accounting substitution effects to identify the changes in real wealth. In system, the SNA, which features GDP—the market value addition, it is critical to capture substitution patterns over of all final goods and services produced within a country’s time, as substitution opportunities are affected by changes boundaries over a period—as the primary indicator. in the relative scarcity of assets. Whether a change-in-capital approach ultimately implies strong or weak sustainability (or The SNA is conceived as a stock-flow-stock system, somewhere in between) and whether it provides a welfare grounded in a double entry accounting system,36 that measure comes down to the prices used, the capital stocks can comprehensively track production and income flows included, whether enough capital stocks are included to leading to changes in wealth in an internally consistent consider the wealth measurement comprehensive, and how accounting framework. Over time, the objective of national wealth is aggregated. These points are discussed in more accounts has evolved to aid governments with monetary detail in the next section. and fiscal questions, such as the setting of monetary policy, the development of budgetary forecasts, and the projection Regardless of the theoretical framework, only changes in of business conditions. This enabled the development of wealth matter for measuring sustainability—not levels. The statistics, such as GDP. In contrast, the SNA’s prescribed absolute level of total wealth of a nation, which could be dimensions related to wealth have had substantially slower conceived as the net present value of all future production uptake across countries, especially those with limited opportunities, is not a well-defined quantity in this context. statistical capacity. The total value of wealth would be determined by removing all those future opportunities, at which point the nation itself GDP’s success as a headline indicator has led to it often being would cease to exist.35 However, the future opportunities described as providing a “birds-eye view” of the economy. available to a country do change. It is this change that can Because it is nearly universally available and adheres to be measured with changes in wealth, even if the entirety standards assuring international comparability, GDP is of wealth—the total size of future opportunities—cannot be widely used as a benchmarking and reference statistic, even measured. It is intentional that this report series is called for purposes for which it was not designed (Jorgenson 2018). The Changing Wealth of Nations and not The Absolute Wealth Indeed, the international standards acknowledge that GDP of Nations. and the associated gross national income (GNI)37 are not the best measures of national income (European Commission et al. 2009, paragraph 16.51), but are often used out of context MEASURING ECONOMIC PROGRESS as such. Perhaps in part because of this misuse, GDP has BEYOND GDP been frequently criticized for three specific shortcomings that are critical for sustainability (Stiglitz, Sen, and Fitoussi To understand why it is important to go beyond GDP to 2010). First, GDP ignores the loss of capital. Second, it draws measure sustainability, it is helpful to understand the the boundary of the economy too narrowly and excludes current system of economic measurement. The seeds for a important services that are not produced through market comprehensive national approach to measuring production, mechanisms. Third, the supply-use tables used to construct income, and wealth were planted in the wake of the Great GDP may misattribute value. Depression of the 1930s, and solidified during World War II, 35 In fact, zero wealth is never defined in the theory, as this would imply no future opportunity—in a global sense it is the end of the world. It is, of course, true that individuals experience different levels of financial wealth and enjoy different qualities of life, which may speak to meeting the needs of current generations. Yet, a person’s wealth is not truly zero until he or she is dead, which is the only state of no opportunity—but also the state of no person. 36 Since these entries are recorded for both sides of a transaction, it is effectively a double-double entry system. 37 GNI comprises GDP and the net receipts of primary income (compensation of employees and property income) from nonresident sources. It is also summarized from the SNA. 29 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS BOX 1.1 THE RELATIONSHIP BETWEEN GDP AND COMPREHENSIVE WEALTH One approach to understanding the relationship between GDP and comprehensive wealth is to start by spelling out standard accounting relationships underlying economic statistics. Concepts and accounting relationships The SNA framework is underpinned by a stock-flow-stock model, illustrated in Box Figure 1.1.1. Economic flows underlying the change in wealth from one period to the next implicitly include a variety of elements. It is important to note that CWON is restricted to measuring only stocks, that is, wealth, and does not articulate all the flows underlying changes in wealth. The diagram below presents a high-level characterization of a full (theoretical) stock-flow-stock framework. In doing so, it makes explicit the conceptual relationship between GDP flows and wealth stocks. BOX FIGURE 1.1.1 GDP vs. wealth in the context of CWON System of National Accounts CWON Extensions Net additions to Net additions to natural capital human capital PRODUCTION The generation of nominal income Production Production from productive activity Gross domestic product or GDP or NDP, incl. GDP or NDP, incl. Net domestic product (GDP less production of investments in depreciation of produced assets) ecosystem services human capital USE OF INCOME Consumption, saving/dissaving Distribution and Distribution and Saving = Investment for total economy use of income use of income Net investment Net investment Net investment excluding depletion/ excluding depreciation (saving or dissaving) degradation of human capital CWON opening stocks CWON closing stocks WEALTH CHANGE +/- Produced capital; Produced capital; Net saving/dissaving Non-produced assets; Wealth change +/- Wealth change +/- Non-produced assets; Natural capital; Other changes in volume of assets Natural capital; Human capital Revaluations Human capital Wealth at period t Wealth at period t+1 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 30 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Standard SNA concepts of GDP and wealth follow a logical flow of interrelated processes presented in a set of integrated “accounts,” beginning with production and ending with wealth accumulation (or decumulation). GDP represents the flow of productive activity inside a country’s geographic boundaries in a given period. This productive activity generates income as returns to the factors of production, broadly characterized as labor and capital. In the context of standard indicators, natural capital in the form of natural resources and land is also able to be measured. A subsequent account shows how income from production is used, for example, for consumption or saving, and how it is allocated among economic actors. The key result of these processes linking production with wealth accumulation is net saving/dissaving, which at the economy level equals net investment. “Net” means excluding the using up (depreciation, depletion, or degradation) of capital either through its ongoing use in production or obsolescence. Unlike gross measures such as GDP or gross investment flows, net measures like NDP or net investment flows can be directly related to changes in wealth. In addition to net saving or dissaving, the volume of assets and thus wealth can change for other reasons. For example, losses due to climate events such as hurricanes, wildfires, or floods can result in important reductions to the asset base. Additions falling into this category include discoveries of new natural resource reserves, or improvements in technologies for energy and minerals, which make the extraction of previously “uneconomic” reserves more viable. Additions could also stem from changes in management or renewable resources, such as restricting the harvest of managed wild fish. Extended measures in the CWON Standard SNA measures of production (GDP) and national wealth have well-defined boundaries. The latter are limited to produced capital and non-produced capital, such as natural resources, including land. In the context of wealth extensions in CWON, if wealth is extended, integrated measurement would also imply extended measures of GDP and other related statistics in the logical flow of accounts. With the extension of standard wealth measures to include new aspects of natural capital (such as renewable energy and ecosystem assets) and human capital, the difference between gross and net production or investment increases even further. For example, net production or investment should also include the depletion of natural resources, or, in the case of ecosystems, the depletion and degradation of ecosystem assets that are not currently in the GDP/SNA asset boundary. Standard SNA measures do not currently account for these production costs. Similarly, measurements of human capital investments need to be based on the expected returns to education and training expenditures, along with many other less easily quantifiable factors, such as the development of skills and aptitudes on the job or outside the formal education system, since education is often rationed rather than procured in the market. CWON produces extended wealth measures that capture net savings, other volume changes, as well as revaluations for a more comprehensive set of assets than currently considered by GDP. While a full articulation of flows contributing to CWON wealth changes would undoubtedly provide useful insights, developing such a framework would be an ambitious exercise, the feasibility of which has yet to be fully examined. 31 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS For example, if green spaces increase a home’s value, this also capital and changes in wealth. In addition, it was an increase may be accounted for, but it is misattributed as opportunity to revisit the production and asset boundaries coming from sticks and bricks only. used in standard national accounts measures, and to push beyond those boundaries to better capture the changing This focus on restricted production concepts, such as opportunities nations may face. GDP and GNI, is inappropriate if the goal is to assess the sustainability of economic progress. GDP is the amount of This renewed attention on comprehensively measuring all new goods and services produced in a period with a defined asset categories, most notably human and natural capital, beginning and end. Only gross investment matters in this was necessary to provide the full national income or changes context to ensure that consumption plus gross investment in wealth measures required to monitor true progress. equate to production (after adjusting for trade). How capital Eighteen years later, the international statistical community generates value in situ does not matter for GDP. Wealth, has made progress in the development of internationally on the other hand, is the net present value of all future accepted statistical standards and guidelines through the production opportunities. The capital that wealth measures System of Environmental-Economic Accounting (SEEA), is the stock embodying those future opportunities, which is covering natural capital assets (stocks) and ecosystem in turn a function of net investment. Wealth thus captures services (flows) (United Nations et al. 2021; United Nations traditional depreciation (as does net domestic product or et al. 2014). The use of natural capital accounting has NDP), but it goes beyond that to also account for volume increased across the globe, with more than 90 countries changes for other reasons, such as natural disasters or producing SEEA accounts (UN 2023a). Notable examples discoveries of fossil fuel reserves. Wealth also accounts include Australia, Botswana, Canada, Costa Rica, India, for revaluations that reflect changes in the productivity of Indonesia, Lao People’s Democratic Republic, Mexico, a given asset or changing supply and demand dynamics. the Netherlands, the Philippines, South Africa, Uganda, Moreover, wealth typically provides a more comprehensive the United Kingdom (UK), and Zambia. In addition, while measurement of the assets relevant to production, including there is no systematic measurement for human capital in natural and human capital (Box 1.1 outlines in more detail the SNA, alternative approaches are recommended as a the relationship between GDP, NDP, and wealth). wealth extension in the 2025 SNA update.38 These have been The main challenge is that other summary measures from successfully implemented by Canada and the UK. the SNA, such as NDP and wealth, are less developed and country-level data are often lacking. To address this data gap, Moreover, there are growing international calls to find the World Bank launched The CWON program in 2006. The economic and sustainable development measures beyond early CWON work drew heavily on the pioneering empirical GDP. The UN Secretary General has made going beyond GDP work of Hamilton and Clemens (1999) on genuine savings in national economic statistics a priority and has recognized rates. The immediate goal was to produce a metric that comprehensive wealth accounting by the World Bank as one would elevate the importance of environmental assets and of the key initiatives in this area.39 There are also calls for shift the discourse in the national accounting community better accounting for the environment and natural capital toward measuring not only production and income, but in macroeconomic statistics, including from the G7,40 the 38 Guidelines on how the dimensions of education, human capital, and labor could be included within the SNA were developed and endorsed as part of the ongoing 2025 SNA revision process and are available at: https://unstats.un.org/unsd/nationalaccount/RAdocs/ENDORSED_WS4_Labour_Human_Capital_ Education.pdf. 39 https://unsceb.org/valuing-what-counts-united-nations-system-wide-contribution-beyond-gross-domestic-product-gdp. 40 This includes a G7 Communique of 2018 (https://www.international.gc.ca/world-monde/assets/pdfs/international_relations-relations_internationales/ g7/2018-06-09-summit-communique-sommet-en.pdf), the G7 Environment Ministers’ Communiques of 2022 (https://www.bundesregierung.de/resource/ blob/974430/2044350/84e380088170c69e6b6ad45dbd133ef8/2022-05-27-1-climate-ministers-communique-data.pdf?download=1) and 2023 (https://www. meti.go.jp/press/2023/04/20230417004/20230417004-1.pdf), and the G7 Science Ministers’ Communique of 2023 (https://www8.cao.go.jp/cstp/kokusaiteki/ g7_2023/230513_g7_communique.pdf). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 32 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS G20,41 and leading academics,42 and through targets set by the non-financial assets, like forests and wild fish stocks (natural SDGs and the Global Biodiversity Framework. In addition, 43 44 capital), may be inside the asset boundary. The SNA states, many national statistical offices have work programs going “many environmental assets are included within the SNA,” beyond GDP, and there are similar initiatives in the corners of (European Commission et al. 2009, paragraph 1.46) in government responsible for economic planning and progress. reference to these non-produced and non-financial assets. Examples include Canada’s Census of the Environment, the A criterion for these assets to be within SNA bounds is that Dasgupta Review published by the UK’s Treasury (2021), or private individuals, companies, or governments48 seek to the United States’ (2023) National Strategy for Natural Capital manage these assets in a way that will contribute to future Accounting and Associated Environmental Economic Statistics. production. However, this limits the set of environmental assets that can be included, as resources such as the atmosphere or high seas are excluded. Similarly, goods and THE RIGHT STUFF: THE WEALTH services provided by uncultivated forests or the mineral ACCOUNTING BOUNDARY or fuel deposits that cannot be extracted economically at The SNA draws specific boundaries on what should be present are not assigned a value. counted and what should be excluded from its accounts. GDP The SEEA system bolsters the SNA’s measurement of the and national accounts use potential market transactions as environment and natural resources. The SEEA Central the starting point to define the production boundary and then Framework (CF) provides guidance on measuring the value adjust to account for other types of economic transactions, of environmentally related services and natural assets that ultimately including a wide range of goods and services. For are within the SNA production boundary (United Nations et example, GDP includes production traded in the market,45 as al. 2014). Before the SEEA-CF, many of these services and well as production that may not involve a cash transaction, assets had been included in the SNA at least since 1993, but such as bartered goods, and subsistence agriculture, hunting, statistical offices had been unsure how to measure them.49 and fishing.46 Goods and services that cannot be traded For example, the SNA uses wild fish stocks within a country’s after being produced or that were not produced on contract exclusive economic zone and managed by a government on for payment are excluded, even if they are economically behalf of a commercial fishing community as an example important (European Commission et al. 2009).47 of inbounds uncultivated biological assets. The SEEA The SNA guidelines differentiate between the production Ecosystem Accounting (EA) expands the asset boundary boundary and the asset boundary, recognizing that assets to include environmental services, such as environmental may still contribute to production despite not being traded health benefits and recreation, and associated natural assets themselves. The SNA asset boundary includes all the traded (United Nations et al. 2021), as has long been called for, or durable assets of the production boundary, such as machines assumed already done, by the economics literature.50 (produced capital), and recognizes that non-produced and 41 This includes the G20 Data Gaps Initiative of the International Monetary Fund (https://www.imf.org/en/News/Seminars/Conferences/g20-data-gaps- initiative), which has identified several climate change indicators and national accounts distributions as critical data gaps for policy. 42 Prominent examples include Arrow et al. (2004), Fleurbaey (2009), Guerry et al. (2015), Jorgenson (2018), and Hulten and Nakamura (2022). 43 Sub-indicator 15.9.1b of the SDGs tracks the integration of biodiversity into national accounting and reporting systems following SEEA principles (https://unstats.un.org/sdgs/metadata/files/Metadata-15-09-01.pdf). 44 Target 14 of the Global Biodiversity Framework calls for the integration of the multiple values of biodiversity into decision-making at all levels, including in national accounting (https://www.cbd.int/gbf/targets/14/). 45 Examples include manufactured machines, agricultural output, rental services of homes and hotels, books, movies, paid domestic work and paid childcare services, timber harvested from forests, commercially caught fish from the ocean, and services produced by governments and non-governmental organizations. 46 Technically, even cutting firewood or growing vegetables in a home garden is within the production boundary, but these are seldom measured in practice. 47 These include cooking one’s own meals, cleaning one’s own house, and caring for one’s own children, health supporting services provided by the environment, and leisure activities for which individuals do not pay. However, GDP does include one measure of services produced by households: the imputed value of owner-occupied housing services. 48 For example, the SNA clarifies that this includes when a government manages the asset on behalf of a user group. 49 In fact, the SEEA-CF and the SNA’s guidance on natural resources exposed many boundary cases, and substantial confusion remains. 50 See, for example, Weitzman (1976); Sefton and Weale (2006); Dasgupta (2001); Hamilton and Clemens (1999). 33 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 1.1 Coverage of “natural capital” on the official SNA non-financial balance sheet for OECD countries United Kingdom Slovak Republic Czech Republic Netherlands Denmark Germany Australia Norway Sweden Canada Estonia Finland Mexico France Korea Japan Cultivated biological resources Land Mineral exploration and evaluation Natural resources Mineral and energy reserves Non-cultivated biological resources and water resources Other natural resources  Total economy  Sub-sector only  No data United States New Zealand Luxembourg Lithuania Hungary Portugal Slovenia Belgium Austria Greece Ireland Poland Latvia Israel Spain Italy Cultivated biological resources Land Mineral exploration and evaluation Natural resources Mineral and energy reserves Non-cultivated biological resources and water resources Other natural resources Source: Authors’ compilation based on OECD data (https://stats.oecd.org/). Note: Cultivated biological resources and mineral exploration and evaluation are considered produced assets. The remaining asset classes are classified as non-produced assets. OECD countries Chile, Iceland, Colombia, Türkiye, and Switzerland are not included in the chart since they do not report any data. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 34 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS However, in practice, the coverage of SNA assets by national THE RIGHT PRICE: statistical offices is often incomplete, and SEEA accounts are VALUATION CONCEPTS not compiled by countries in ways that align with the rest When measuring changes in wealth it is important to of the economy. Not all countries currently measure non- consider prices and physical quantities of assets. Prices serve financial produced capital and even fewer countries produce as the relative weights for combining different capital stocks natural capital accounts or non-produced asset accounts on into an aggregate wealth summary. Were markets perfect and a regular basis (UN 2023) or compile these assets on national complete, then identifying the correct prices to use, the so- balance sheets.51 For example, 43 countries report balance called shadow prices,56 would be easy and uncontroversial. sheet data to the Organisation for Economic Co-operation Prices would be taken directly from observed market and Development (OECD) going back to 2000.52 The OECD data transactions. Since markets are neither perfect nor complete, cover seven categories of produced (two) and non-produced two concerns need to be addressed to arrive at the right price (five) capital that might be considered natural capital. Most 53 for assets in CWON: the price concept and which future of the 43 countries have missing data, especially for natural opportunities to count. How these concerns are addressed capital (Figure 1.1). Five countries—Chile, Iceland, Colombia, influences the insights one can take from the aggregate Türkiye, and Switzerland—report no natural capital at all. wealth summary. The United Nations Statistics Division’s 2023 Global Assessment It is important to distinguish between value and price. of SEEA implementation54 reports that 90 countries complied In economics, a price represents the set of things for with some components of the SEEA-CF, and 35 countries which an incremental unit of a good can be exchanged. For example, the price for a water bottle could be the complied with some components of the SEEA-EA. However, dollars57 that are exchanged for the bottle. Hence, prices these data have not been compiled into a central database, are expressed in the units dollars per quantity, where the and it is difficult to assess the coverage systematically.55 Most price reflects the relative scarcity of the good—in this case countries only produce physical measurements. Only 22 bottles of water.58 The value is defined with units dollars per countries produce monetary accounts for timber (not forest), quanitity×quantity=dollars,59 much like an area is defined in and this is the largest share to produce monetary accounts. units meters×meters or square meters.60 The next most commonly produced monetary account was for water, with only 15 countries developing a monetary account. In the economic measurement of sustainability, there are two Some countries align their SEEA data with other economic apparently conflicting value concepts: “exchange values” and data, while others do not. “welfare values.” The accounting context starts by assuming 51 There is a challenge of terminology: some asset classes that might be considered natural capital are included as produced capital, such as cultivated biological assets. Modern management of ecosystems often blurs the lines between produced and non-produced assets. For example, a forest regenerated using a shelterwood cut or a “wild” fish stock supported by aquaculture. 52 Other countries may include natural capital on the official balance sheet. However, it is likely that the distribution of official statistics in the OECD reporting is biased in favor of greater production of these official balance sheet statistical series. 53 None of the categories are clearly human capital, though some produced capital could potentially be considered human capital (for example, artistic originals). 54 https://seea.un.org/content/2023-global-assessment-results. 55 For the 90 countries with SEEA accounts and 72 countries had links to data. 56 The term shadow price is common in economics literature but is not always used consistently. In some cases shadow price is used as the implied price of a good or service not exchanged in the market but under prevailing conditions, while other times it implies the welfare maximizing price of a good or service that is not exchanged in the market, that is, a corrective price that would lead to efficient internalization of externalities (Fenichel, Abbott, and Yun 2018). 57 The dollar represents other consumption opportunities and is sometimes called a numeraire. 58 Economic scarcity and physical rarity are different. Physical rarity may lead to economic scarcity, but economic scarcity is availability relative to demand. For example, mobile phones are scarce despite being ubiquitous because people are willing to forgo a substantial amount to obtain one. Venomous insects may be rare in some locales but are never scarce, as people seldom wish there were more of them. 59 Of course, euros, renminbi, or any other currency can be used. Dollars are used here as an internationally understood shorthand. 60 Formally, price is the change in value with respect to a change in quantity, and equivalent to the marginal value. At a mathematical level, value and price are linked through the fundamental theorem of calculus: price is the derivative of value with respect to quantity, and value is the integral of price over quantity changes. 35 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS the transfer of a stock between two parties, exchanged at a the production boundary) as well as the observed market specific (market) price.61 However, what is often observed price P1. The area measure A would thus represent the is only sale value and quantity, so that the price needs to be observed market-based transaction value or exchange value.63 imputed by dividing the sales value by quantity.62 When a The welfare economist, on the other hand, would consider price and quantity measure are available, then the “exchange the welfare value of the provision of the good Q1 relative to a value” can be computed by multiplying the price at the counterfactual, here shown as no good at all, that is, relative prevailing quantity by that prevailing quantity, as illustrated to zero quantity. This is represented by the area under the in Figure 1.2, yielding area A for price P1and quantity Q1. To marginal value curve (generally the demand curve), which is understand what is meant by the term “welfare value” on the area A+B.64 This is clearly different from the exchange value other hand, it is important to recognize that welfare measures recorded by the accountant, who is not interested in the always refer to a difference between two possible conditions, hypothetical counterfactual. While the national accountant that is, a change from liquidating quantity Q1. Thus, to and welfare economist disagree on the value, they can agree measure the “welfare value” associated with quantity Q1, on the observed price and quantity—assuming there is first the change from quantity Q1 to zero needs to be considered, agreement on the accounting boundaries. which is defined by the area A+B in Figure 1.2. It is important to understand to what extent these different These value concepts imply different values of holding perspectives can be reconciled, because CWON aims to quantity level Q1. On the one hand, a national accountant align itself (where possible) with SNA and SEEA standards would observe quantity Q1, which is the units of a good that is and guidelines for measurement, while also looking at produced or held in capital (and whose service is fully within sustainability, which is associated with changes in welfare (Arrow et al. 2004). The convention in the SNA is to use FIGURE 1.2 exchange values. Where possible, valuation is based on Exchange value vs. welfare value observed market prices,65 though the SNA imputes some prices when necessary. The SNA acknowledges that it operates in a world with market failures, which means that the measured price vector may not maximize social welfare. By applying the SNA concept of exchange value, the real CWON wealth measures (in levels) will thus reflect values Price based on the policy environment and market structure, Demand curve including any market failures, but not first-best “welfare B values.”66 Nonetheless, prices measured from actual market transactions generally align with the second-best welfare P1 Quantity concept (Lipsey and Lancaster 1956), enabling change-in- A Q1 welfare interpretations. 61 In fact, the SNA guidance defines exchange value like a market price, with paragraph 3.121 stating that “exchange values in most cases will represent market prices.” However, the term “exchange value” should be, and commonly is, used in the “area” sense to refer to measures consistent with the SNA framework. This point is being clarified further as part of the 2025 SNA revision process. 62 This imposes an assumption of a constant marginal price. 63 This assumes that all the quantity can be traded without affecting price. 64 One common approximate welfare measure is the change in the sum of producer and consumer surplus, which is the area under a demand curve less the costs of provision. In the case of gross measurement, cost of provision is not subtracted. In that case, only the area under the demand curve matters. Welfare economists recognize that consumer surplus, as defined by the Marshallian demand curve, is not the appropriate measure for non-marginal changes and prefer Hick’s compensated surplus measures for welfare analysis (Freeman 2003). 65 It is important to note that for CWON measures of capital prices, it is the prices themselves that are forward-looking and are best thought of as the price of net changes in capital. This means that expenditure-based methods, like those used for public services such as health care, education, public transportation, and public parks and conservation areas are unlikely to align with the marginal real income contribution. There are many reasons this can be the case. First, discount rates used to analyze government expenditures may differ from the social discount rate. Second, seldom do decision analyses account for general equilibrium effects. Third, some public services like parks and schools are established on natural and human capital that are not generated through markets. 66 This is true even if the prices are second-best welfare consistent. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 36 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS However, to analyze the sustainability of progress, one must able to account for changes in price due to the changes in the analyze how real wealth changes over time, that is, how wealth demand for stock holding the demand curve constant, for changes after removing general price (substitution) effects. example, as the amount of the good is reduced from level Q1 to Changes in real wealth need to account for changes in quantity level Q1. The approach of finding an average price is not a good between the two accounting periods and changes in prices, approximation if consumers substitute this good for another, because fluctuating prices reflect changes in substitution leading to shifts in the demand curve itself. In that case it would patterns. These patterns are part of the behavioral responses be necessary to account for some of the curvature properties. that are important for sustainability. This is like the challenge of measuring changes in real (or volumetric) GDP (or NDP). Furthermore, price curves often have the curved shape Prices for stocks of capital change as a result of changes in shown in Figure 1.2 or Figure 1.3, which can lead to errors scarcity of the stocks being measured or because of changes when computing the area under the demand curve between in potential substitutes or complements. Prices can also the opening and closing quantity.69 In such a situation, fluctuate because of institutional or policy changes. Yet, the Diewert (1992) shows that the Fisher ideal index, a so- goal of measuring changes in wealth is to ask how much called “superlative index,”70 provides a good approximation. more aggregate consumption can happen, which means accounting for substitution patterns and not misidentifying them as wealth effects. FIGURE 1.3 The main question is what price to use. Consider the change Changes in prices and their implications for exchange from good level Q1 and price P1 to good level Q2 and P2, as and welfare values illustrated in Figure 1.3. The welfare economist would compute the difference in welfare values, which would yield a change in welfare equal to area A+E+G. An accountant, on the other hand, might compute the difference in nominal exchange values yielding area F+H-A. However, this is not a Price meaningful real value measure in either the SNA or welfare context. A reasonable and intuitive approach for calculating B differences in exchange values is to average prices, P , ̃ 67 P2 P H G instead such that the change in wealth would be calculated D F E Demand curve P1 ̃ (Q1-Q2) equal to area A+E+D. If the marginal price curve as P × C A A were a straight line, then G=D, and this provides a good Q2 Q1 approximation of the size of the change if the demand curve Quantity can be reasonably approximated with a straight line. In this 68 Note: The area shaded in green (C+A) represents the exchange case, the welfare economist and accountant would agree on value of quantity Q1. The area shaded in blue (C+F+H) represents the change in value. However, a simple average price is only the exchange value of quantity Q2. 67 In academic studies of the change in wealth, typically an average price or arithmetic mean is chosen. See, for example, Arrow et al. (2012), Dasgupta (2014), and Yun et al. (2017). The average price approach is equivalent to averaging a Laspeyres and Paasche index. 68 This result follows from standard Euclidean geometry. It is also the case that the smaller the change in Q, then the more likely a straight line is to be a good approximation of the demand curve over the change being considered. 69 To compute changes in wealth one must compute the area under the demand curve between the opening and closing quantity. This follows directly from the fundamental theorem of calculus and the definition of an asset price as the marginal value of an asset (Jorgenson 1963). 70 A superlative index is any price or quantity index that can provide a second order approximation to a welfare change. The Fisher ideal index for a single good is defined as a geometric mean of prices, which is computed as √(P_1^2 P_2^2 ), while the arithmetic means are computed as 1/2(P_1+P_2). More generally, the Fisher ideal index is the geometric mean of the Laspeyres and Paasche indexes. 37 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Furthermore, Diewert shows that superlative indexes are markets and where the services from those missing assets capable of reflecting changes in substitution to isolate the can distort market prices through complementarity and income effects. An equivalent way to approximate area A+E+G substitution effects. The answer is that it is likely better than can be achieved by applying the Törnqvist index. The Törnqvist current alternatives and hopefully gives the correct sign of the index is also a superlative index, which provides the same change. Nonetheless, these changes in wealth should not be result as the Fisher ideal index (up to a small approximation interpreted as changes in welfare. error) even when there are multiple stocks (Dumagan 2002), but is computationally simpler to implement. The Fisher ideal and Törnqvist indexes are regularly used in national accounts WHAT CWON MEASURES: to compute real GDP. This edition of CWON uses a Törnqvist THE ASSET BOUNDARY volume index to compute changes in real wealth (for more The CWON program aims to fill some of the data gaps detail refer to chapter 2). identified above by producing estimates of changes in real Real changes in CWON’s comprehensive wealth estimates comprehensive wealth per capita for 151 countries for the measured using the Törnqvist volume index could thus 1995 to 2020 period. These measurements are intended to approximate real changes in welfare, if the capital stocks be comparable across countries and over time, and cover included were truly comprehensive.72 This is because the five asset classes: produced capital, nonrenewable natural shape and location of the marginal value (price) curve depend capital (fossil fuels and metals and minerals), renewable on the accounting boundary. To illustrate this point, consider natural capital (agricultural land, forests, mangroves, marine a forest. The SNA production boundary includes timber and fish stocks, and renewable energy), human capital, and net fuel wood production but excludes some non-wood ecosystem foreign assets. services that are part of the future production opportunities While the CWON methodology aims to align where possible considered in CWON. Therefore, a national accountant with internationally accepted statistical standards and adhering strictly to the SNA production boundary would arrive guidelines of the SNA and SEEA, the grouping of assets at a different marginal value curve for standing forests than an in CWON differs from the SNA (see Figure 1.4). The SNA analyst using the services enumerated in CWON, which also first divides assets into financial and non-financial (real) include the forest’s role in water provision.73 assets, while CWON is mostly focused on non-financial Although the CWON program has pushed the measurement assets. Within the category of non-financial assets, of wealth beyond the current SNA accounting boundaries to the SNA divides them further into produced and non- account for services believed to be important for sustainable produced assets. Some “natural capital,” such as plantation development, its coverage is still not comprehensive. forests and aquaculture fish stocks, are also included as Nevertheless, the CWON balance sheet remains one of the most cultivated biological assets within the produced asset section, comprehensive measures of non-financial assets available while they are grouped under renewable natural capital today. The question thus becomes whether the approximation for CWON. On the other hand, the SNA treats all land as a is “good enough” for a second-best world, where the observed non-produced asset, while urban land is included under market prices reflect inefficient economies and missing produced capital for the purposes of the CWON database.74 71 The single good case is used to develop intuition. In the applied multi-good setting, the relative importance of a price is weighted by the quantities to which it is attached to create a “volume” measure. This is effectively an approach to approximating the multidimensional integration problem. 72 This also means that the services included in the associated income measures were also comprehensive—this depends on the accounting boundaries. This is likely not the case for CWON, but CWON continues to expand beyond the current SNA production boundary. 73 Hashida and Fenichel (2022) address this challenge by examining actual decisions landowners make about harvesting trees from a forest. They find that on the fastest-growing sites the revealed value aligns very closely with what one would expect from institutions that only considered timber. However, on slower growth sites the decision-makers acted as if there were added value to standing forests beyond their timber value. The revealed value thus also included services provided by their forests outside of the SNA and possibly within the SEEA-EA production boundary. 74 This choice is primarily driven by the methodology used, as urban land is estimated as a share of produced capital, drawing on Kunte et al. (1998). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 38 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 1.4 The CWON wealth accounts (2024 release) and their relationship to the SNA and SEEA Comprehensive Wealth Non- Produced renewable Human Net foreign Renewable natural capital capital natural capital assets capital CW O N Timber Shoreline Machinery Fossil fuels Marine Urban Renewable Agricultural non-wood protection Labor Assets- and oil, gas, coal fish land energy  land ecosystem services of force liabilities structures Minerals stocks services mangroves Non- Mineral cultivated Cultivated Other SNA, SE E A+ Intellectual and biological biological Land natural property energy resources resources resources and water reserves Foreign resources direct SEEA-EA investment positions Produced assets Non-produced assets (natural resources) SNA balance SNA balance sheet for non-financial assets Outside SNA of payments Note: Arrows represent where components of CWON can be found in the SNA system, including the non-financial asset balance sheet and balance of payments accounts or in the SEEA-EA (the SEEA-CF is fully aligned with the SNA). Most of the assets covered in CWON, such as machinery over time CWON has expanded the asset boundary to include and structures, fossil fuels, metals and minerals, urban and key SEEA ecosystem accounts, which includes allowing non- agricultural land, timber, and marine fish stocks, are all clearly timber forest ecosystem services (some of which may be in within the asset boundaries of the SNA and SEEA-CF, which is the SNA, such as non-wood forest products like mushroom fully aligned with the SNA, and are part of the SNA balance harvesting, while others, such as recreation, are not) into sheet of non-financial assets. Renewable energy assets are not forests. It also includes important regulating services such yet part of the SNA. However, guidelines for their inclusion, as shoreline protection services capitalizing into mangroves which were developed for the previous CWON report, were (which may be partially attributed to property value in the endorsed as part of the 2025 SNA revision process. 75 SNA). And, with this update, experimental estimates of climate regulation provided by terrestrial ecosystems are included Changes in this subset of SNA assets do not sufficiently embody (which might be partly captured by the SNA to the extent the changes nations face with respect to future production that asset prices account for climate risks, but its non-market (and ultimately consumption) opportunities, requiring CWON benefits are clearly not captured in the SNA). to go beyond the SNA and SEEA-CF asset boundary. In fact, 75 https://unstats.un.org/unsd/nationalaccount/aeg/2022/M21/M21_14_WS11_Renewable_Energy_Resources.pdf. 39 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS CWON also includes other assets critical for sustainable For some assets there is limited international guidance and development, notably human capital. The SNA does not implementation experience, especially in the context of provide the same entry points for human capital as it does wealth accounting. The development of valuation guidelines for natural capital. It includes intellectual property, such for ecosystem services is ongoing, and several key questions as software and artistic originals, as produced capital. In remain unresolved. One such example is carbon retention addition, some statistical offices have produced satellite services provided by terrestrial ecosystems, such as forests or accounts that address elements of human capital, and there mangroves, which play a key role in climate mitigation efforts. is international cooperation in this area (United Nations While it is possible to produce experimental estimates at the Economic Commission for Europe 2016). Guidelines on how global scale (as done in chapter 7), there is no agreed approach the dimensions of education, human capital, and labor could on how to attribute these estimates at the country level due to be included within the SNA were developed and endorsed its global public good nature. Nor is there unanimity on how to as part of the ongoing 2025 SNA revision process. 76 Still, avoid double-counting when aggregating the value of carbon systematic measurement of human capital remains an even retention with other assets.78 Similar conceptual concerns arise larger gap than that for natural capital, which CWON steps in when accounting for critical assets such as water,79 which are to help close. further compounded by data and modelling constraints (see Box 1.2). For these reasons, several important assets cannot The key limiting factor in expanding the asset coverage yet be included in CWON’s comprehensive wealth estimates, further is data availability. To ensure comparability across but there might be future opportunities as data availability countries and over time, data are primarily sourced from improves and remaining conceptual issues are resolved. global databases. However, these data usually have limited assets, country, and temporal coverage—and often lack the necessary granularity. For example, aquaculture has become WHAT CWON IS MEASURING: an important industry that now matches marine capture THE VALUATION APPROACH AND fisheries in terms of production volume and value (FAO 2022). AGGREGATE WEALTH MEASURES Given this, aquaculture-related ecosystem assets should be The approaches and concepts used by CWON to value assets included in the CWON database, as they, in combination with align where possible with SNA and SEEA statistical standards other inputs, contribute to benefits enjoyed by humans.77 and guidelines and use the national accounting “exchange Yet a comprehensive assessment of the wealth arising from value” concept. The valuation approaches used vary across aquaculture is not possible today due to data limitations. A first asset category (for more detail, see chapter 2). For example, step in that direction was taken with this edition of CWON by for the calculation of physical capital stocks, CWON follows compiling pilot accounts for a subset of species and countries the SNA guidance and international practice in employing (for more detail see chapter 8). A back-of-the-envelope the perpetual inventory method (which assesses value as calculation based on this pilot suggests that aquaculture- equivalent to depreciated past investment rather than expected related natural capital may be four times more valuable than net present value of the flow of services). The valuation that of marine fish stocks. 76 https://unstats.un.org/unsd/nationalaccount/RAdocs/ENDORSED_WS4_Labour_Human_Capital_Education.pdf. 77 These assets include the equipment and installations (pens, ponds, and so on) used to farm fish (produced capital), the breeding stock and inventories of partially grown fish (also produced capital), and the ecosystem inputs to fish farming (natural capital). The value of the produced capital data used in aquaculture is, in principle, already captured in the produced capital values CWON draws from the Penn World Tables (PWT), so it may not need to be added. The value of the natural capital used in aquaculture, on the other hand, is missing from CWON, and methods and data would have to be found to value it if it were to be added. This would require data on prices and production quantities, which are publicly available, and on costs, which are unfortunately rarely collected systematically. 78 It is reasonable to assume that the benefits of carbon retention are at least to some extent already captured in the prices of other assets currently included in CWON. To avoid double-counting, only non-market benefits should be included. 79 The value of water is already partly accounted for in agricultural land and hydropower, but it is difficult to account for that contribution. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 40 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS approach for nonrenewable and renewable natural capital heavily weighted toward oil, such as Nigeria or Saudi Arabia. assets is based on the residual value method (RVM) or net In such cases, the GDP deflator will be driven by the price of present value (NPV) approach recommended by the SNA and oil, but will not reflect the price of the goods that are consumed SEEA-CF, while ecosystem assets are valued separately using in that country, such as vehicles, machinery, or cell phones— approaches recommended by the SEEA-EA. For the valuation many of which are imported. If the goal is to understand the of human capital, CWON uses the lifetime earnings approach evolution of the real purchasing power of a nation’s wealth, a developed by Jorgenson and Fraumeni (1989, 1992a, 1992b), domestic demand deflator should be used instead. Even when which is one of the statistical approaches recommended by applying other price-based deflators, changes in real wealth the ongoing 2025 SNA revision process. 80 will be distorted by short-term price volatilities, which are removed when using a volume-based index. Third, given many The main methodological innovation introduced with the countries compute real GDP with the price weights of a fixed current update of the CWON database is the use of the base period, resulting GDP implicit price deflators82 would not Törnqvist volume index to compute changes in real wealth account for substitution. However, accounting for changes in (for more detail, see chapter 2).81 In this approach, the relative substitution patterns, as the Törnqvist volume index enables, changes in the physical assets of a nation, such as the size of is fundamental to the analysis of sustainability, as substitution its fish stocks or the number of workers in the labor force, are opportunities will change over time with changes in the weighted by their relative economic importance (as measured relative scarcity of assets. by their shares in nominal wealth). Changes in real wealth per capita will be driven by (i) the depletion or accumulation of assets (through relative changes in physical assets), (ii) CONCLUSIONS changes in the productivity, or relative scarcity, of assets GDP is widely recognized to be an insufficient measure (through relative price changes), (iii) changing substitution of progress and national “success.” Whether progress is patterns (through cross-price and quantity effects), and (iv) sustainable can be measured by how real wealth per capita increasing or decreasing competition for available assets is changing, as this represents changes in future production (through demographic pressures); all of which are important (and, ultimately, consumption) opportunities. However, while for analyzing the sustainability of economic progress. almost all countries produce and use GDP, only a few countries Moreover, the use of the Törnqvist volume index bridges produce wealth measures, particularly wealth measures that the notions of change in real exchange value and change in are comprehensive and include natural and human capital. welfare value, making this debate more about what is in the The CWON database aims to fill this data gap by producing accounting boundary than the value concept itself. comparable and consistent comprehensive wealth estimates Past editions of CWON calculated wealth changes in constant for 151 countries for the last quarter of a century. prices using the GDP deflator, which was convenient to While CWON aspires to measure wealth comprehensively, mainstream wealth accounting, but this edition aligns the its coverage is still limited by methodological and data deflation approach with best practice. One key concern when constraints. The CWON measurement has gradually expanded using the GDP deflator is that (just like GDP) it only covers with each edition, as new data and statistical standards have current flows of domestic production (not consumption) become available. The current coverage includes key assets within a given country and excludes all imported goods or from the SNA balance sheet and several critical ecosystem services. Take the example of a country whose production is 80 The endorsed valuation guidance for the SNA 2025 revision process can be found here: https://unstats.un.org/unsd/nationalaccount/RADOCS/ENDORSED_ AI1_Valuation_Principles_Methodologies.pdf. 81 This edition also introduces several other innovations and improvements in methodology to enhance both internal consistency and alignment with the SNA and SEEA standards, and existing as well as emerging guidelines. These changes are discussed in detail in chapter 2 as well as in the dedicated asset- specific chapters. 82 The estimation of real GDP is typically done at a very detailed level. There is thus not one “deflator” used in its derivation. Instead, the overall price change for the products encompassed in GDP (that is, domestically produced goods and services) is derived implicitly by the ratio of nominal to real GDP. 41 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS assets that generate ecosystem services that are covered by also declining due to reduced production possibilities. the SEEA-EA. CWON goes beyond the current SNA standards to Such information is critical for development planning. For include human capital. However, the measurement of wealth example, it can help direct investment toward assets that need is still not comprehensive, as key natural capital assets, such as to be accumulated faster, so they can keep up with population water, aquaculture, and climate regulation services cannot be growth or identify assets that are being degraded and require included due to theoretical and practical concerns. Therefore, more sustainable management. changes in real wealth per capita cannot yet be used to make inferences about changes in welfare, but they can be used to Wealth data can also help inform asset diversification track economic progress. strategies, as countries aim to alleviate the constraints imposed by the physical extent of assets by shifting the allocation of By tracking changes in real comprehensive wealth per capita their overall wealth portfolios. This has, in fact, been a key over time, policy makers can assess the sustainability of a element in the development of many nations. Natural capital nation’s progress. As long as real wealth per capita is not endowments are often drawn down to provide income to fund declining, economic progress may be sustainable. Of course, investments in other assets with greater marginal value. When a nation may choose in the short term to draw down some of done carefully and with a clear understanding of the limits its wealth to support current consumption over and above to which natural capital can be drawn down, this can be an the level that current production would allow—say, during a effective means of increasing welfare. This is a point taken up temporary downturn. However, no nation can allow its real per further in this report in the discussion of weak versus strong capita wealth to continually decline without welfare eventually sustainability (see chapter 4). BOX 1.2 ACCOUNTING FOR WATER IN CWON—CHALLENGES AND OPPORTUNITIES W ater is a vital resource for all countries, but water valuation is challenging for theoretical and practical reasons. These include the physical characteristics of water, the way water is regulated and used within the economy, and the fact that water is an essential good. For these reasons, water valuation is contentious and the observed water prices are seldom a true reflection of the marginal value of water. As part of the ongoing development of CWON, three approaches to water valuation consistent with the valuation concepts and methods of the other assets included in CWON and the SEEA were identified by Vardon et al. (2024): asset-by-asset, use-by-use, and ecosystem service-by-service. ■ Asset-by-asset: This bottom-up approach is based on observed market transactions of water assets within countries. This is problematic because the direct trade of water assets rarely occurs. However, while the water assets themselves are not traded, their value can be determined through the value of trades in water rights, which are a “permit to use a natural resource” and are a financial asset in the SNA (see paragraphs 17.324 and 3.36, respectively). However, while some countries and regions have tradable water rights—for example, Australia, Chile, Iran, South Africa, and parts of the United States (UN 2021)—most countries do not. Concerns have also been raised about the functioning of some water markets (for example, Garrick et al. 2020). Still, Fenichel et al. (2016) shows how an asset-by-asset approach can be used for water when high-quality data are available. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 42 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS ■ Use-by-use: This is another bottom-up method, which uses country-by-country assessments of water use by different industries, such as agriculture, mining, manufacturing, energy, water supply, education, health, and so on. In this, the value of water used is embedded in the value added by each industry, rather than just the price paid per unit volume used. The SEEA water supply and use tables could provide the information necessary to implement this approach, but these are only available for a handful of countries at the level of detail needed. ■ Ecosystem service-by-service: This approach could use either a bottom-up or top-down method. Bottom-up would require country-level ecosystem service accounts, but as very few countries have these accounts, this method is not possible. A top-down method could be used, building on the approach developed to estimate the value of forest ecosystem services in CWON (Siikamäki et al. 2024). In this method, the value of forests is derived from a meta-analysis of the existing academic literature on ecosystem services, including erosion control, flood protection, hydropower, and water services. It is a service-by-service approach that can be extended to cover all ecosystem types (beyond forests) and value water assets. Each approach to water valuation has its advantages and drawbacks, with the service-by-service approach being the most feasible in the short term. The growing uptake of water and ecosystem accounting by nations using the SEEA, combined with new data sources and tools for account production, will make all approaches more feasible in the long term. Confounding the difficulties of water valuation is that the value of water is already embedded in the value of other assets included or intended to be included in CWON. This includes agricultural and forested land (Siikamäki et al. 2024) and renewable energy (Smith et al. 2024). If a separate valuation of water is included in CWON, then it could be presented as a standalone value, independent of asset values, or deducted from the value of other assets. REFERENCES Addicott, E.T., Fenichel, E.P., and Kotchen, M.J. 2020. “Even the Representative Cohen, F., Hepburn, C.J., and Teytelboym, A. 2019. “Is Natural Capital Really Agent Must Die: Using Demographics to Inform Long-Term Social Substitutable?” Annual Review of Environment and Resources 44: 425–448. Discount Rates.” Journal of the Association of Environmental and Resource Dasgupta, P. 2001. Human Well-Being and the Natural Environment. New York: Economists 7: 379–415. Oxford University Press. Arrow, K., Dasgupta, P., Goulder, L., Daily, G., Ehrlich, P., Heal, G., Levin, S., Maler, K.G., Schneider, S., Starrett, D., and Walker, B. 2004. “Are We Dasgupta, P. 2014. “Measuring the Wealth of Nations.” Annual Review of Consuming Too Much?” Journal of Economic Perspectives 18 (3): 147–172. Resource Economics 6: 17–31. Arrow, K.J., Dasgupta, P., and Maler, K.G. 2003. “Evaluating Projects Dasgupta, P. 2021. The Economics of Biodiversity: The Dasgupta Review. and Assessing Sustainable Development in Imperfect Economies.” London: HM Treasury. Environmental and Resource Economics 26: 647–685. Dietz, S., and Neumayer, E. 2007. “Weak and Strong Sustainability in Barbier, E.B. 2011. Capitalizing on Nature. New York: Cambridge the SEEA: Concepts and Measurement.” Ecological Economics 61 (4): University Press. 617–626. 43 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Diewert, W.E. 1992. “Exact and Superlative Welfare Change Indicators.” Lipsey, R.G., and Lancaster, K. 1956. “The General Theory of Second Best.” Economic Inquiry 30: 565–582. Review of Economic Studies 24: 11–32. Dumagan, J.C. 2002. “Comparing the Superlative Törnqvist and Fisher Ideal Muller, N.Z., Mendelsohn, R.O., and Nordhaus, W.D. 2011. “Environmental indexes.” Economics Letters 76 (2): 251–258. Accounting: Methods with an Application to the United States Economy.” American Economic Review 101: 1–30. EC (European Commission), IMF (International Monetary Fund), OECD (Organisation for Economic Co-operation and Development), UN (United OSTP (United States (US) Office of Science and Technology Policy), OMB Nations), and World Bank. 2009. System of National Accounts 2008. (US Office of Management and Budget), and DOC (US Department of Commerce). 2023. National Strategy to Develop Statistics for Environmental- FAO (Food and Agriculture Organization of the United Nations). 2022. The State Economic Decisions: A U.S. System of Natural Capital Accounting and of World Fisheries and Aquaculture 2022. Towards Blue Transformation. Associated Environmental-Economic Statistics. Edited by Office of Science Rome: FAO. doi:10.4060/cc0461en. and Technology Policy, Office of Management and Budget, and Department Fenichel, E.P., and Abbott, J.K. 2014. “Natural Capital from Metaphor to of Commerce. Washington, DC: The White House. Measurement.” Journal of the Association of Environmental and Resource Pearce, D., and Atkinson, G. 1993. “Capital Theory and the Measurement of Economists 1 (1): 1–27. Sustainable Development: An Indicator of ‘weak’ Sustainability.” Ecological Fenichel, E.P., Abbott, J.K., and Yun, S.D. 2018. “The Nature of Natural Capital Economics 1993: 103–108. and Ecosystem Income.” In Handbook of Environmental Economics, edited Pezzey, J. 1992. “Sustainable Development Concepts.” World 1 (1): 45. by V.K. Smith, P. Dasgupta, and S. Pattanayak, 85–142. North Holland. Polasky, S., Bryant, B., Hawthorne, P., Johnson, J., Keeler, B., and Pennington, Fisher, I. 1906. The Nature of Capital and Income. Norwood, MA: Norwood D. 2015. “Inclusive Wealth as a Metric of Sustainable Development.” Annual Press. Review of Environment and Resources 40 (6): 6.1–6.22. Fleurbaey, M. 2009. “Beyond GDP: The Quest for a Measure of Social Welfare.” Rouhi Rad, M., Adamowicz, W., Entem, A., Fenichel, E.P., and Lloyd-Smith, P. Journal of Economic Literature 47 (4): 1029–1075. 2021. “Complementarity (Not Substitution) between Natural and Produced Freeman, A.M. III. 2003. The Measurement of Environmental and Resource Capital: Evidence from the Panama Canal Expansion.” Journal of the Values: Theory and Methods. 2nd ed. Washington, DC: Resources for the Association of Environmental and Resource Economists 8 (6): 1115–1146. Future. Sefton, J.A., and Weale, M.R. 2006. “The Concept of Income in a General Guerry, A.D., Polasky, S., Lubchenco, J., Chaplin-Kramer, R., Daily, G. C., Equilibrium.” Review of Economic Studies 73: 219–249. Griffin, R., ... and Vira, B. 2015. “Natural Capital and Ecosystem Services Solow, R. 1993. “An Almost Practical Step towards Sustainability.” Resource Informing Decisions: From Promise to Practice.” Proceedings of the National Policy 19 (3): 162–172. Academy of Sciences 112 (24): 7348–7355. Stiglitz, J.E., Sen, A., and Fitoussi, J.P. 2010. Mis-measuring Our Lives: Why GDP Hamilton, K., and Clemens, M. 1999. “Genuine Savings Rates in Developing Doesn’t Add Up. New York: The New Press. Countries.” The World Bank Economic Review 13 (2): 333–356. UN (United Nations). 2023. Global Assessment of Environmental-Economic Hashida, Y., and Fenichel, E.P. 2022. “Valuing Natural Capital when Accounting and Supporting Statistics 2022. United Nations Statistics https:// Management Is Dominated by Periods of Inaction.” American Journal of seea.un.org/sites/seea.un.org/files/global_assessment_2022_background_ Agricultural Economics 104 (2): 791–811. doc_v4_clean_1.pdf. Hulten, C.R., and Nakamura, L.I. 2022. “Is GDP Becoming Obsolete? The UN (United Nations), EU (European Union), FAO (Food and Agriculture ‘Beyond GDP’ Debate” (No. w30196). National Bureau of Economic Research. Organization of the United Nations), IMF (International Monetary Fund), Jorgenson, D.W. 1963. “Capital Theory and Investment Behavior.” American OECD (Organisation for Economic Co-operation and Development), and Economic Review 53 (2): 247–259. World Bank. 2014. System of Environmental-Economic Accounting 2012— Central Framework. United Nations. Jorgenson, D.W. 2018. “Production and Welfare: Progress in Economic Measurement.” Journal of Economic Literature 56 (3): 867–919. United Nations et al. 2021. System of Environmental-Economic Accounting— Ecosystem Accounting (SEEA EA). New York: White cover publication, Jorgenson, D.W., and Fraumeni, B.M. 1989. “The Accumulation of Human and pre-edited text subject to official editing. https://seea.un.org/ecosystem- Nonhuman Capital, 1948–1984.” In The Measurement of Saving, Investment, accounting. and Wealth, edited by R.E. Lipsey and H.S. Tice, 227–282. Chicago: University of Chicago Press. Weitzman, M.L. 1976. “On the Welfare Significance of National Product in a Dynamic Economy.” The Quarterly Journal of Economics 90 (1): 156–162. Jorgenson, D.W., and Fraumeni, B.M. 1992a. “The Output of the Education Sector.” In Output Measurement in the Service Sectors, edited by Z. World Commission on Environment and Development. 1987. Our Common Griliches, 303–341. Chicago: University of Chicago Press. Future. New York: Oxford University Press. Jorgenson, D.W., and Fraumeni, B.M. 1992b. “Investment in Education and U.S. Yun, S.D., Hutniczak, B., Abbott, J.K., and Fenichel, E.P. 2017. “Ecosystem Economic Growth.” Scandinavian Journal of Economics 94 (Supplement): Based Management and the Wealth of Ecosystems.” Proceedings of the S51–S70. National Academy of Sciences 114 (25): 6539–6544. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 44 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS 2 How the World Bank Measures Comprehensive Wealth MAIN MESSAGES INTRODUCTION ■ CWON wealth estimates align where possible with The World Bank’s The Changing Wealth of Nations (CWON) the internationally accepted statistical standards and program is a pioneering effort in the measurement of guidelines of the System of National Accounts and wealth, producing a regularly updated global database of the System of Environmental-Economic Accounting. comprehensive wealth estimates for nearly two decades. CWON goes beyond these standards to include These estimates are aligned where possible with the United additional assets critical for sustainable development, Nations’ System of National Accounts (SNA; EC et al. 2009), such as human capital and key ecosystem assets. the System of Environmental-Economic Accounting Central Framework (SEEA-CF; UN et al. 2014), and the System of ■ This edition of CWON implements an important Environmental-Economic Accounting Ecosystem Accounting methodological innovation that affects how real wealth (SEEA-EA; UN et al. 2021) to ensure comparability with other per capita and its changes over time are measured. widely used statistical measures, such as GDP. With each new Real wealth estimates are computed using a chained CWON edition, the measurement of wealth has been updated Törnqvist volume index, where the physical assets of a and expanded as new data sources, methods, statistical nation, such as the size of its fish stocks or the number standards, and guidelines have become available. of workers in the labor force, are weighted by their economic importance. This edition of CWON implements an important methodological innovation that affects how real wealth per ■ Changes in real comprehensive wealth per capita will capita and its changes over time are measured. CWON 2024 be driven by the depletion or accumulation of assets, computes real wealth using a chained Törnqvist volume changes in the productivity or relative scarcity of index in line with SNA guidance for computing capital stock assets, changing substitution patterns, and increasing estimates adjusted for general price effects (SNA, paragraphs or decreasing competition for available assets. All of 15.167–15.172). These are commonly referred to as “real”83 these are important for analyzing the sustainability of asset values—a term used in this report.84 What this means economic progress. is that changes in real wealth per capita will now track how the real asset base of a nation and relative prices change ■ Future efforts should build on the critical over time. Real wealth per capita will increase if capital methodological innovations introduced with this is accumulated, as, for example, more workers enter the edition, and improve the implementation of the labor force, forests are replanted, or hydropower plants are volume index, especially the measurement of physical connected to the grid. It will decline if wealth is depleted volumes. too rapidly, as is the case when fish stocks are overfished or the reserves of fossil fuels are depleted, and the resulting 83 As noted further below (see footnote 88), our use of the term “real” here is not, strictly speaking, aligned with guidance on the compilation of stock measures in the SNA. It is, however, consistent with general practice in the presentation of inflation-adjusted data by statistical agencies. 84 The transition to a volume-based index was recommended as part of a wider methodological review led by Robert Smith (Midsummer Analytics) and supported by a technical advisory committee comprising Matthew Agarwala (Bennett Institute for Public Policy, University of Cambridge, and Tobin Centre for Economic Policy, Yale University); Catherine Van Rompaey (World Bank), Karen Wilson (former assistant deputy head of Statistics Canada), and Rintaro Yamaguchi (Japanese National Institute for Environmental Studies). Similar recommendations were also made as part of a review of deflation methods for CWON wealth measures (Inklaar et al. 2023). The change will ensure consistency in deflation methods across asset types, as real produced capital estimates from the Penn World Tables, used in the CWON database, are also computed as a volume-based index (Penn World Table 10.01 by Feenstra et al. 2015). 45 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS rents are not reinvested in other assets. The weight these consumption possibilities must be maintained to ensure the changes are given in the comprehensive wealth index is sustainability of well-being. This approach had the advantage determined by their nominal value (measured as a share of of familiarity and tractability, with GDP deflators being total value in current US dollars at market exchange rates) readily available for most countries from national accounts and changes in relative prices over time, which are driven by data. However, it also had two main shortcomings (Sefton changes in scarcity of the stocks being measured, changes in and Weale 2006; Inklaar et al. 2023; see annex A1): the availability of potential substitutes or complements, or 1. Wealth as a measure of future production institutional and policy changes. opportunities: Deflation using the GDP deflator (or any price deflator for that matter) implicitly treats all assets The remainder of this chapter proceeds as follows. First, a in the comprehensive wealth portfolio as fungible—or discussion is presented on how a real measure of aggregate interchangeable—stores of value, readily converted to comprehensive wealth can be computed and why a chained money to be spent on goods and services. While this Törnqvist volume index was adopted for this edition. The aligns with the work of many leading theoreticians of chapter then explains which assets are included in CWON. wealth accounting (Arrow et al. 2004; Dasgupta 2001), Next, the chapter discusses the broad valuation approach, it is not well aligned with the view of sustainability and which common assumptions are applied to all valuation taken in previous CWON reports: “A nation’s income approaches (where applicable) to ensure consistency of is generated by its [comprehensive] wealth” (CWON methods and comparability of value estimates across assets. 2021, 25). This view places the emphasis squarely on It then describes briefly how each asset in CWON is valued the sustainability of production (since production is the and the extent to which the valuation approaches align with source of income), which will subsequently ensure the SNA and SEEA standards and guidelines. It concludes with sustainability of consumption. The goal is to measure a summary of the outstanding conceptual, methodological, changes in the real productive base of the economy, and and measurement challenges that future CWON editions thus, a comprehensive set of productive assets, such as should tackle. machinery, forests, and people. 2. Appropriate deflator for the real purchasing power A NEW APPROACH FOR COMPUTING of national wealth: Even if CWON were to use a price REAL COMPREHENSIVE WEALTH deflator, which is troublesome as price-based deflators Even though the final goal is to estimate changes in real introduce confounding price effects and cannot isolate comprehensive wealth per capita over time, the level of substitution effects, the GDP deflator is not the most aggregate comprehensive wealth—that is, the value of all the appropriate choice. One key concern when using assets in a nation’s comprehensive wealth portfolio—must be the GDP deflator is that (just like GDP) it only covers computed first. Aggregate comprehensive wealth is compiled current flows of domestic production (not consumption) in nominal and real terms in CWON. Compilation in nominal within a given country and excludes all imported goods terms is straightforward. Nominal aggregate comprehensive or services. For example, a country with an economy wealth is simply the sum of each of the assets measured in dependent on fossil fuel production for export will nominal terms themselves, using market exchange rates. experience quite different price trends in its production Measuring aggregate real wealth, such that only relative price activities than in its consumption activities. The former changes are accounted for, is more complicated. Doing so will be dominated by international prices of bulk oil, requires a decision on how to remove all other price effects. gas, and coal, while the latter will be dominated by the prices of consumer goods and services, many of which In the past, CWON’s approach was to compile aggregate real will be imported. For any open economy, deflation of comprehensive wealth by deflating the nominal value of each national wealth using the GDP deflator will not result in a asset using a GDP deflator and then adding the deflated asset meaningful measure, whether in terms of a total volume values together. Implicit in this approach was the view that of wealth assets or of a constant purchasing power. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 46 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Given these shortcomings, CWON 2024 has shifted away from that are not bought and sold in the market—has value as an the GDP deflator in favor of an approach focused on assets input into a production process.85 Countries cannot dig up as inputs to production processes. This approach rests on ecosystems and sell them to their neighbors. An exception is constructing an index in which the quantities—or “volumes”— made for financial assets, which are relatively liquid and can of the various assets comprising comprehensive wealth are be converted to cash in the short term. Financial assets can aggregated together using their nominal values as weights. thus be logically deflated with the CPI.86 A “volume” in this context refers to a physical quantity of a given asset, which can be measured in different units. The choice to move to a volume index for CWON 2024 For example, the volume of agricultural land is measured meant that a specific index had to be chosen from among in hectares, while the volume of oil assets is measured in the many possibilities (World Bank et al. 2004); a decision barrels. The volume index can sum across assets measured that was guided by international best practice. A commonly using different units by weighing changes in volumes over used index is the Fisher ideal index,87 which many national time by their time-varying shares in nominal wealth. Volume statistical offices use to express changes in price or volume. changes related to assets that make up a greater share of Advanced economies like Canada and the United States use nominal wealth, such as human capital, are thus given a the Fisher ideal index to estimate GDP in real terms. As greater weight in the index. Such a volume index aligns with discussed in chapter 1, this index also has many desirable the idea that what matters for sustainability is not preserving theoretical properties (Diewert 1976), most notably its assets as stores of value but, rather, preserving them as ability to capture relative price and volume changes, and to entities that, when combined with one another in production account for substitution effects. It has drawbacks in practice, processes, yield the goods and services that are themselves however. Notably, a complex formula is required to derive the object of consumption. the contribution of each element of the index to the overall growth in the index (Chevalier 2003). An index that avoids In this context, wealth is best thought of, and measured, in the this and shares many of the same theoretical qualities is the most concrete terms possible. This is what is demanded when Törnqvist index (Törnqvist 1936; Dumagan 2002). actual forests, mineral deposits, machinery, and the people that make up the workforce are all seen as part of national wealth. HOW REAL COMPREHENSIVE Take human capital as an example. The knowledge, skills, and capacities of workers cannot be bundled together and sold off WEALTH IS COMPUTED to the highest bidder. These characteristics and qualities are CWON 2024 has adopted the Törnqvist index to compile inherent to the individuals who possess them. The value of “real” comprehensive wealth estimates,88 measured as the human capital can thus only be realized when workers choose price-weighted volumes of assets and expressed in monetary to offer (or rent) it to others temporarily in return for wages as terms using “chained” prices. The year 2019 was chosen to part of an employment arrangement, or, alternatively, to use compute the real asset values. The choice of reference year it themselves in carrying out their own production activities. is ultimately arbitrary, but 2019 was chosen for CWON 2024 Similarly, much natural capital—especially ecosystem assets because it is the most recent year in the database that was not 85 In this context, “production” includes the production of non-market goods and services like flood control and recreational opportunities. 86 CPI was chosen over the GDP deflator in CWON 2024, given the concern noted above about the appropriateness of the GDP deflator in the context of wealth accounting. 87 Fisher himself referred to it as the “ideal index,” but it has come to be so associated with him that it now bears his name. 88 The primary objective of any measurement in “real” terms is separating price and volume effects. By removing the general price (inflation/deflation) effect, the real measure will only capture the volume and relative price effect. “Real” values (sometimes referred to as “constant price” values) are, strictly speaking, values that have had the general price removed from them by the application of a price index to their nominal value (see SNA para 2.66). Values that have had the general price effect removed from them by direct consideration of quantities are properly referred to as “volumes.” However, it is common practice in statistics to use the term “real” even when “volumes” are referred to, likely because the term is broadly understood to mean “after removing the influence of changes in the general price level.” Thus, CWON 2024 retains use of the term “real” even though, strictly speaking, it is asset “volumes” rather than “real asset values” that are presented. 47 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS a crisis year (2020, the latest year in the database, was marked ■ Changing substitution patterns both across and within by the beginning of the COVID-19 pandemic). A range of asset classes (through cross price and quantity effects). chained Törnqvist volume indexes are then compiled for the ■ Increasing or decreasing competition for available assets CWON database, including indexes for all individual assets (for example, timber or agricultural land), for each individual (through demographic pressures). asset class (for example, renewable natural capital), as well It is important to note that, in practice, the current method as for aggregate comprehensive wealth (which comprises does not fully reflect changes in the quality of assets over produced capital, nonrenewable and renewable natural time, as it ideally should. For example, if the agricultural capital, human capital, and net foreign assets). Readers land available in a country declines over time by area due interested in additional details are referred to annex A2 of to desertification, and in its productivity due to a loss of this chapter and to the overall methodology report (World nutrients in the soil, both changes should be captured in the Bank 2024). index. However, only the former is reflected in the index due While choosing an index that emphasizes changes in the to data constraints (for more detail see annex A1). The current volume of assets is a major improvement, it is important approach can capture increases in productivity, which in the to note that in doing so the impact of prices on the value case of agricultural land would be reflected in higher nominal of assets and, ultimately, changes in wealth is not lost. As asset values of the inputs to agricultural production, namely volume changes are weighted by nominal shares for adjacent produced and human capital. In other words, higher yields periods, changes in relative prices are accounted for.89 Real due to better tractors or climate-smarter farmers would be wealth per capita will thus increase if capital is accumulated— reflected in higher average and marginal revenue product for example, when more workers enter the labor force, or if of produced and human capital, respectively. Improving the the productivity90 of a given asset increases when the same “quality” measurement of the physical volumes should be a workers upgrade their skills. However, it will decline if priority in future extensions of this work. assets are depleted and becoming harder to substitute. For example, as fish stocks are being depleted, initially more ASSETS INCLUDED IN CWON fishing boats might be able to sustain a given harvest level. But, as overfishing continues, more fishing boats will not The CWON program aspires to measure wealth be able to offset the declines in fish stocks. Real wealth per comprehensively across five asset classes: produced capita will also decrease if the population grows faster than capital, nonrenewable natural capital, renewable natural wealth is accumulated, as there is more competition for a capital, human capital, and net foreign assets (see Figure given set of assets. 1.2). In practice, this aspiration cannot be fully realized, as data constraints limit the coverage of the assets that Changes in real comprehensive wealth per capita will can be valued, especially for renewable natural capital, as directly capture key effects that are important for analyzing discussed in chapter 1. The current update of the CWON the sustainability of economic progress, namely: database covers 151 countries—adding Angola, Guinea- ■ Bissau, Israel, Montenegro (after 2007), New Zealand, São The depletion or accumulation of assets (through relative Tomé and Príncipe, Serbia (after 2007), St. Lucia, and Sudan changes in physical assets). (after 2011)91—for the 1995–2020 period and includes the ■ Changes in the productivity or relative scarcity of assets following assets (assets marked with an asterisk are new in (through relative price changes). this edition): 89 This is the case for comprehensive wealth measures as well as the real value of selected asset classes. However, the real value of an individual asset will only be driven by changes in volumes (not prices). 90 Productivity is defined here as the output per unit of (labor and capital) input, and can increase due to a broad variety of factors, such as good management practices, more efficient production processes, and improved quality of intermediate inputs. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 48 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS ■ Produced capital: Machinery and equipment; buildings; the pilot developed for the previous report (chapter 14 in intangible assets such as intellectual property; and World Bank 2021). However, due to data limitations it was urban land. not possible to include global estimates of solar, wind, and geothermal assets. Similarly, additional data were collected ■ Nonrenewable natural capital: Fossil fuels (oil, gas, to increase the coverage of metals and minerals of growing and hard and soft coal); and minerals and metals economic importance, especially for the renewable energy (bauxite, cobalt*, copper, gold, iron ore, lead, lithium*, sector. This resulted in the addition of cobalt, lithium, and molybdenum*, nickel, phosphate rock, silver, tin, and molybdenum, increasing coverage from 10 to 13 metals and zinc).92 minerals. ■ Renewable natural capital: Agricultural land (cropland and pastureland); forests (timber;93 non-wood forest ecosystem services, including recreation, fishing, and HOW ASSETS ARE VALUED hunting; non-wood forest products; and water services, IN CWON reported by protected area status); mangroves (shoreline In line with CWON’s efforts to align itself where possible protection services); marine capture fisheries (including with the statistical standards and valuation guidance from commercial and artisanal fisheries); and renewable the SNA and SEEA, it uses the national accounting “exchange energy* (hydropower).94 value” concept—the marginal price of an asset times its ■ Human capital: The value of skills, experience, and quantity. Where possible, observed market prices are used effort by the working population over their lifetime by to align with SNA and SEEA guidance. Which valuation gender (male and female). approach is chosen for each asset type ultimately depends ■ Net foreign assets: The sum of a country’s external on the available implementation guidance and practice, as assets and liabilities, such as foreign direct investment outlined further below (for more detail, see World Bank and reserve assets. 2024). Moreover, to ensure the CWON database is consistent, several common assumptions are implemented: The inclusion of new assets in CWON is typically determined ■ A constant uniform discount rate of 4 percent is used, as by their growing economic importance in the emergence in previous CWON reports (following World Bank 2006), of new statistical standards and guidelines. For example, wherever discounting is required. This assumption is not the revised version of the SNA standards, intended to come ideal, since wealth estimates (especially for renewable into effect in 2025, will expand the natural resource asset natural capital and human capital) may be discounted boundary to include renewable energy assets (Smith and over long periods of time, making the values sensitive to Peszko 2022),95 treating them with the same as fossil fuels. the choice of discount rate. Moreover, using a uniform Reflecting this guidance and their importance for the low- discount fails to account for significant country- carbon energy transition, this edition of CWON adds an level differences in economic fundamentals (such as account for hydropower assets (see chapter 6), building on 91 CWON 2024 does not cover four countries that were included in the CWON 2021 panel due to missing or incomplete CPI data required for the deflation of financial assets (Turkmenistan, Bolivarian Republic of Venezuela, and Republic of Yemen), or missing renewable energy data (West Bank and Gaza). 92 According to the SEEA-CF (Table 5.6), for a nonrenewable deposit to be considered an economic asset, it must meet several conditions. First, extraction and sale of material from the deposit must have been confirmed to be economically viable. Second, the feasibility of extraction by a mining operation must have been confirmed. Finally, the extraction must either be actively pursued in an on-production project or anticipated in the foreseeable future in a project approved or justified for development. 93 According to the SEEA-CF (para 5.346), timber resources are those parts of forests where timber harvesting is legally permitted; that are accessible for harvest; and that contain commercially useful species. 94 The SEEA-CF does not provide explicit guidance on defining hydroelectric resources. Elsewhere in this report (see chapter 6), these are defined as hydroelectric resources associated with “viable” hydroelectric generation projects. Viable generation projects are those for which the environmental- socioeconomic viability and technical feasibility has been confirmed and development or operation is currently taking place, or sufficiently detailed studies have been completed to demonstrate the technical feasibility of development and operation. Hydroelectric resources that exist at sites where hydroelectric generation plants do not currently exist and none are well advanced in planning do not qualify as assets and are not in scope for measurement in CWON. 49 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS interest rates, growth patterns, and risk) and social and marine fish stocks. Like the previous edition of CWON individual preferences,96 as well as differences in risk, (World Bank 2021), marine subsidies were not deducted scarcity, or environmental or other externalities across in CWON 2024 to ensure consistency across all natural assets (Dasgupta 2008; Dietz and Asheim 2012; Gollier capital assets within the CWON database. CWON’s 2019; Groom et al. 2022). More research is needed to approach of not considering subsidies is aligned with determine how this might be improved in the future. For statistical practice (though not with official conceptual now, this edition continues to use a constant uniform guidance), as subsidies are generally not accounted for discount rate, but presents experimental estimates based by statistical agencies when they are compiling official on differential discount rates for renewable natural national estimates of resource asset values.98 capital in chapter 4. ■ Future rents are held constant for all assets at current PRODUCED year (2019) values whenever an NPV-RVM approach is CAPITAL used in CWON. This means that future rents used in the NPV calculation reflect 2019 market conditions, For the calculation of produced policies, and expectations about the future. This is capital stocks, CWON follows in line with recommendations in the SEEA-CF on the SNA guidance and international indirect valuation of natural assets (SEEA-CF, paragraphs best practice99 in employing the perpetual inventory method. 5.133 and 5.134.). Given the infinite number of possible This approach requires investment data and information on future trajectories of rents, and the wide range of assets’ service lives and depreciation patterns, among other assets and countries in the CWON database, this is the things. Data on produced capital stocks in nominal and real least subjective and most transparent assumption. terms compiled using this approach are available for most CWON 2024 has implemented this recommendation countries directly from the Penn World Tables (PWT) 10.0.100 consistently across all assets, including for agricultural Since the produced capital estimates of the PWT are already land and human capital, which included rent forecasts in reported in real terms using a chained Törnqvist volume previous editions of CWON. 97 index, the estimates are directly comparable to the other real ■ Subsidies are not accounted for in the implementation measures computed for this edition of CWON. of the NPV-RVM approach in CWON. According to the The nominal value of urban land is estimated as a fixed SNA and SEEA-CF, subsidies paid by governments to proportion of the value of produced capital (equivalent to support natural resource production should be deducted 24 percent), drawing on Kunte et al. (1998). To estimate the from revenues in the calculation of resource rent. volume index for urban land, data on the physical extent of However, data on subsidies are difficult to obtain, and urban land are taken from the Center for International Earth are currently only available in the CWON database for Science Information Network at Columbia University101 95 A guidance note (Smith and Peszko 2022) for the treatment of renewable energy as assets was produced by the CWON team as a follow-up to the 15 pilot renewable energy accounts produced by CWON 2021. This guidance note has been reviewed and was endorsed as part of the 2025 SNA revision process: https://unstats.un.org/unsd/nationalaccount/aeg/2022/M21/M21_14_WS11_Renewable_Energy_Resources.pdf. 96 Groom et al. (2022) find significant differences in social discount rates across countries and international organizations, ranging from 1 percent in Germany to 12 percent used by the Inter-American Development Bank. 97 The assumption to remove rent growth from the estimation of agricultural land is uncontroversial, as it is aligned with valuation guidance for renewable natural capital from the SEEA. However, human capital estimates using the Jorgensen-Fraumeni lifetime income approach (as used in CWON) typically assume wage growth. However, in the context of CWON, alignment of valuation concepts and methods across all assets is critical to ensure that one asset value is not inflated relative to others. 98 For example, Statistics Canada’s estimates of oil assets do not account for subsidies. The same is true of the UK, Australia, Norway, and other major economies. CWON’s approach of not considering subsidies for these assets is thus aligned with official statistical practice. 99 For example, most OECD countries adopt this method to estimate their capital stocks (Bohm et al. 2002; Mas, Perez, and Uriel 2000; Ward 1976). 100 For countries without PWT estimates, produced capital stock is constructed using gross capital formation as a proxy for aggregate investment and a depreciation rate of 5 percent. 101 https://sedac.ciesin.columbia.edu/data/collection/grump-v1. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 50 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS and the urban population from the United Nations (UN) Nominal asset values for nonrenewable natural capital are Population Division’s World Urbanization Prospects.102 Given estimated by drawing on a wide range of publicly available that urban land data are only available for a subset of years, 103 databases from the International Energy Agency (IEA), changes in urban land are proxied by changes in the urban the United States (US) Geological Survey and US Energy land-to-population ratio multiplied by urban population for Information Administration, and the UN Statistics Division, each year in the time series between 1995 and 2020. as well as licensed data sources from Rystad Energy, Wood Mackenzie, and S&P. Proven reserve estimates are used as the physical quantities required for the volume index. While NONRENEWABLE the coverage of proven reserves data for oil, gas, and coal NATURAL CAPITAL is relatively comprehensive, there can be significant data gaps for minerals and metals. Gaps are filled forwards by The SEEA-CF recommends deducting production and backwards by adding production. the NPV-RVM approach for the The backwards approach is likely to be more accurate, as it valuation of sub-soil assets.104 would include any past resource discoveries. The forward- CWON adopts this approach, filling approach is unable to account for any new discoveries measuring the value of a nation’s stock that may have occurred.107 of fossil fuels, minerals, or metals as the present discounted value of the stream of rents expected until the resource is exhausted. Resource rent is calculated each year for RENEWABLE each resource type as the difference between the revenues NATURAL CAPITAL from resource extraction and the cost of that extraction, including intermediate inputs, labor compensation, and the For the valuation of renewable “user cost” of the produced capital used in the extraction natural capital assets, such as process. 105 Previous editions of CWON followed this agricultural land, timber, marine approach, but proxied the user cost of capital with data fish stocks, and renewable energy, on annual investments in fixed assets. While this proxy the same NPV-RVM approach used for may be reasonable in the long term, it will not necessarily nonrenewable natural capital is applied, with one important reflect user costs in the short term. 106 This edition of CWON difference. In the case of nonrenewables, the lifetime of the asset developed direct user cost estimates for all sub-soil assets applied in the NPVcalculation is determined by the ratio of current by constructing a historical time series of capital stocks and production to current reserves, while in the case of renewables, estimating rental and depreciation rates for each asset. the lifetime is simply assumed in all instances to be 100 years.108 102 https://population.un.org/wup/. 103 Center for International Earth Science Information Network urban land estimates are only available for the years 2000 and 2015. An urban land to urban population ratio is then calculated for these two years and is linearly interpolated and extrapolated to fill the time series between 1995 and 2020. 104 The SNA, for its part, does not provide explicit guidance to readers on the valuation of natural resource assets. However, it points readers to the SEEA for additional guidance on the topic, so implicitly also recommends the RVM/NPV approach. 105 User costs of capital are estimated as “normal” returns to fixed assets plus depreciation. 106 In the long run, the user cost of capital and investments in an industry are likely to be similar. However, in the short run this is unlikely to be the case, as owners may choose to invest more than they earn (by either borrowing or drawing down corporate savings) or less than they earn depending on where they are in the development of their businesses. 107 For resources in countries for which production data are available but information on reserves is absent, regional or world averages are used. 108 An exception to the 100-year lifetime is made in the case of timber resources that are not sustainably harvested, for which the lifetime varies depending on current timber stock size and current rates of harvest. For other renewable resources, a 100-year lifetime captures most of the asset value (with a 4 percent discount rate, the present value of any harvest more than 100 years in the future would be, at most, no more than 2 percent of the nominal value). Growing uncertainty regarding the scale of the impact of climate change on both the economy and the environment may be such that an assumed 100-year asset life is no longer valid for CWON. This is a point that will be considered further in future editions. The impact of climate change on both the economy and the environment may be such that an assumed 100-year asset life is no longer valid for CWON. This is a point that will be considered further in future editions. 51 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS This follows the practice of the UK’s Office for National bound estimate of terrestrial protected areas, as it only Statistics (ONS 2020). It is important to note that the resource covers forest ecosystems. rent can be computed from detailed revenue and cost data for marine fish stocks109 and renewable energy,110 but data For shoreline protection services provided by mangroves (especially on costs) are more limited for agricultural land (measured by Global Mangrove Watch),112 the expected benefits and timber. A rental rate is thus applied to the gross value of averted damages to property are estimated annually using of production (as reported by the UN’s FAO) to estimate unit process-based storm and hydrodynamic models to identify the resource rents.111 area and depth of flooding (Menéndez et al. 2024). By running a scenario analysis for different storm frequency events For the valuation of ecosystem assets, the NPV of the annual (such as storms every 5, 25, or 100 years) with and without service flow is estimated using spatial socioeconomic mangroves, a probabilistic distribution of flood damages and and biophysical data in line with SEEA-EA guidance. For avoided damages is derived (for more detail, see chapter 8). example, for non-wood forest ecosystem services, which These data can then be overlaid with spatialized produced include recreation, hunting, and fishing, as well as non- capital from the PWT113 to estimate annual expected benefits wood forest products, and hydrological services, annual and, subsequently, the NPV of the asset. benefits are estimated using spatially explicit regression and machine learning models (Siikamäki et al. 2024). For renewable natural capital assets, data on physical volumes These models draw on value estimates from a systematic needed to compute real wealth measures are generally literature review, as well as complementary spatial data, available. For example, the CWON database includes hectare socioeconomic data at the country level, and spatial data on measures for agricultural land, forest and mangrove areas, ecological and biophysical characteristics. These data are and the generation capacity of hydropower plants. The only further disaggregated by protection status, providing the first physical measures not already included in the CWON database global estimates of the value of protected forest areas. This prior to this edition were estimates for marine fish stocks. is an improvement on previous editions of CWON, which Biomass estimates of major exploited species in a country’s proxied the value of protected areas by assuming their next exclusive economic zones (EEZs) were derived by the Sea best use was in agriculture. This likely overstated the value Around Us initiative, using a suite of methods to reconstruct of many protected areas, especially in large countries with fisheries catches for the 1950–2018 period and complement remote protected areas where no meaningful opportunity them with stock assessments performed previously by others for agriculture exists. The current estimate provides a lower- (Sumaila et al. 2024). 109 The resource rent estimates for wild capture marine fish stocks are produced using data on landed values and prices from FAO and the Sea Around Us reconstruction database (www.seaaroundus.org) as well as the fishing cost database from the Fisheries Economics Research Unit at the University of British Columbia (Lam and Sumaila 2021). 110 The resource rent estimates for hydropower are estimated using generation data from the International Renewable Energy Agency (IRENA; https://www.irena.org/Data/Downloads/IRENASTAT) and the UN Energy Statistics database (http://data.un.org/Data.aspx?d=EDATA&f=cmID%3aEC), price data from the IEA’s Energy Prices database (https://www.iea.org/data-and-statistics/data-product/energy-prices#overview), and cost data from IRENA’s regional investment cost estimates (https://www.irena.org/Publications/2023/Aug/Renewable-Power-Generation-Costs-in-2022). 111 For agricultural land, the rental rate is proxied by country- and decade-specific land cost shares provided by the Agricultural Productivity database of the US Department of Agriculture (https://www.ers.usda.gov/data-products/international-agricultural-productivity/). For timber, the rental rate is proxied for by the ratio of unit rents to the export unit value using adjustment factors based on Applied Geosolutions (2016). 112 www.globalmangrovewatch.org. 113 To estimate spatialized produced capital estimates, the per capita capital stock data are multiplied with the European Commission’s Global Human Settlement Layer (https://ghsl.jrc.ec.europa.eu/ghs_pop2019.php), which records the global distribution of population at a 250 meter resolution. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 52 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS HUMAN CAPITAL labor income consists of both the incomes of the employed and self-employed,119 the latter is estimated using data from There is no single internationally the SNA and PWT. After some gap filling, the survey data are accepted method for the adjusted to population estimates from the UN to make sure valuation of human capital. they represent a country’s population. Subsequently, the However, the lifetime income lifetime income (adjusted by survival rates and a discount approach developed by Jorgenson factor) can be calculated for the representative individual and Fraumeni (1989, 1992a, 1992b) used in CWON is widely (aged 15–65) by age, gender, and education. Lastly, these applied by researchers in the field, and is one of the statistical lifetime income profiles are multiplied by the corresponding approaches recommended by the ongoing 2025 SNA revision number of people in a country to compute the human capital process.114 According to this approach, human capital is stock by age, gender, and education. estimated as the present value of the expected future labor income that could be generated over the lifetime of the To compute human capital in real terms, quality-adjusted women and men currently living in a country (Fraumeni labor force data disaggregated by gender are used as a 2008; Hamilton and Liu 2014).115 The implementation of the “volume” measure. The labor force data are taken from the lifetime income approach requires data from many sources, ILO and adjusted for the changing educational composition including data on population by age and gender from the of the labor force, or “quality,” by multiplying these volumes UN’s World Population Prospects; 116 employment and labor by the PWT’s human capital index (HCI).120 This allows us force participation from the ILO and PWT; survival rates to approximate the average human capital per worker. The from the Global Burden of Disease Study; 117 and education nominal human capital estimates (disaggregated by gender) and earnings profiles from the UN, harmonized household are then used in the Törnqvist volume index to estimate and labor force surveys of the World Bank’s International human capital in real terms. Income Distribution Database and Global Labor Database, as well as the Luxembourg Income Study.118 NET FOREIGN ASSETS Estimating human capital involves several steps. The first Net foreign assets are a measure step is to use the standard Mincer equation to estimate of the foreign assets and private returns per year of schooling using survey data from the International Income Distribution Database, Global liabilities held by a country’s Labor Database, and Luxembourg Income Study, and to residents, which is calculated as construct a matrix of expected earnings by age, gender, and the sum of a country’s external assets education level. However, since these returns include only and liabilities, such as foreign direct investment and reserve wages, they need to be scaled up to account for additional assets. Estimates of net foreign assets are mostly obtained benefits workers earn as part of their overall compensation. directly from the External Wealth of Nations (EWN) Mark II This is done using data from the SNA. Moreover, since total database121 developed by Lane and Milesi-Ferretti (2007, 2018). 114 The endorsed valuation guidance for the SNA 2025 revision process can be found here: https://unstats.un.org/unsd/nationalaccount/RADOCS/ ENDORSED_AI1_Valuation_Principles_Methodologies.pdf. 115 That is, CWON estimates the lifetime income of a representative worker with a given educational background, experience, and gender. Real wages for future earnings for, say, a 30-year-old female worker when she is 50 are set equal to the wages of female workers that are currently 50 with the same educational background. 116 https://population.un.org/wpp/. 117 http://www.healthdata.org/gbd/2019. 118 https://www.lisdatacenter.org/our-data/lis-database/. 119 The economic role of the self-employed can be especially important in many low- and middle-income countries where subsistence agriculture and informal economy are very common. 120 https://www.rug.nl/ggdc/productivity/pwt/?lang=en. 121 https://www.brookings.edu/research/the-external-wealth-of-nations-database/. 53 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS The EWN database provides estimates of net foreign assets adjustments are critical, as they may reinforce or offset from 1970 to 2020 for 214 economies. Lane and Milesi- observed trends in physical volumes. Data constraints Ferretti primarily draw on reported international investment currently limit the ability to make such quality positions data from individual countries’ balance of payment adjustments, but future efforts should aim to collect the statistics, disseminated by the International Monetary Fund required information. (IMF). The sole conceptual difference is the EWN database ■ Constrained discount rate: The current edition uses excludes central bank gold holdings from financial assets a constrained, uniform discount rate. Future efforts (since they are not a claim on another country). Otherwise, should explore the possibility of using differential definitions for each component of net foreign assets are discount rates, which could vary over time and across official definitions taken from the IMF’s Balance of Payments assets and countries. Such a revision should build on the and International Investment Position Manual (IMF 2009). guidance and experience of statistical and government To deflate net foreign assets, CPI is used, as discussed agencies around the world, as well as academic previously. Deflation by CPI is justified since financial assets can be viewed as resources available for consumption. literature on social discounting. ■ Lifetime of renewable resources: CWON 2024 continues past practice of valuing all renewable natural assets CONCLUSIONS using an assumed asset life of 100 years (except The main methodological innovation introduced with this where harvests are known to be unsustainable). This update of the CWON database is the use of the Törnqvist assumption may require revisiting given that climate volume index to compute real wealth values. Changes in change is impacting the environment and the economy real comprehensive wealth per capita measured in this more quickly and seriously than anticipated. It may way provide critical information to policy makers on the be more realistic to assume that renewable resource sustainability of economic progress. They signal whether rents will flow unchanged for a shorter period. Further the productive base of the economy is growing or shrinking research would be required to determine whether a new over time (due to changes in the volume of assets) and the value could be identified and applied to all renewable extent to which the scarcity and productivity of these assets resources or whether different lifetimes might be is changing relative to other assets (through relative price appropriate for different resources. changes). This shift to a volume-based index is in line with ■ Accounting for subsidies: Subsidies paid by international statistical guidance and best practice and is governments to support natural resource production a significant methodological improvement on previous should be accounted for when estimating resource rent. editions of CWON, which used the GDP deflator. It is recommended that a data assessment be conducted However, there are several methodological and measurement to identify available global subsidy data sources for all concerns that could not be addressed as part of this update natural resource assets. due to time and resource constraints. Future efforts should ■ Urban land: Current urban land values are estimated consider addressing the following: by applying a time-invariant, global factor of 0.24 to the ■ Quality adjustments to asset volumes: The current value of produced capital assets from the PWT, based on volume measures used in the Törnqvist volume index are Kunte et al. (1998). A growing body of evidence suggests not corrected for the “quality” of the asset (except for the this assumption is not well supported, as the actual share produced capital estimates from the PWT). Such quality (when measured) fluctuates considerably over time and 122 The APO Productivity Database provides estimates of the value and quantity of land used in production for 25 Asian countries. The OECD also reports land data for 16 out of 37 countries. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 54 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS differs markedly across countries. The recommendation Lastly, it is important to point out that the CWON 2024 wealth is to conduct an assessment to determine for which values provide a conservative baseline estimate in line with countries urban land value data are readily available 122 standard statistical practice, and do not account for future and to develop an approach for estimating comparable policy actions or changes in market conditions due to, for values for the rest of the world. example, climate change. Modelling is needed to explore possible “what if” scenarios. To facilitate the use of the CWON ■ Timber: The current approach to valuing timber assets data in such a modelling exercise, this edition will not only uses rents derived from a rental adjustment factor make publicly available the final wealth estimates in real and based on timber export prices. The recommendation nominal terms, but also, where possible, the input data and is to adopt the NPV-RVM approach instead. This will associated statistical code used to derive the wealth estimates. require a data assessment, including the identification Users of the CWON data will thus be able to modify the input of internationally comparable data on revenues data as well as the assumptions used to derive the wealth from timber production and the associated costs for estimates, including assumptions about future rent growth. intermediate inputs, labor, and fixed assets. Access to the underlying data and code may be valuable ■ Protected areas: The current edition of CWON provides to researchers wishing to compile comprehensive wealth the first global estimates of the value of forest protected estimates for individual countries using national, rather than areas using a meta-regression analysis combined with global, data. The relevant input data and statistical code can geospatial data. Future work should explore how to be accessed through the World Bank’s website. expand these estimates to cover non-forest and marine protected areas. REFERENCES Applied Geosolutions. 2016. “Improving the Forests Database to Support Dasgupta, P. 2008. “Discounting Climate Change.” Journal of Risk and Sustainable Forest Management.” Final report to World Bank, J. Jenkins Uncertainty 37 (2/3): 141–69 (principal investigator), Washington, DC, October 2015. Dietz, S., and Asheim, G.B. 2012. “Climate Policy under Sustainable Discounted Arrow, K., Dasgupta, P., Goulder, L., Daily, G., Ehrlich, P., Heal, G., Levin, Utilitarianism.” Journal of Environmental Economics and Management 63 S., Maler, K.-G., Schneider, S., Starrett, D., and Walker, B. 2004. “Are We (3): 321–335. Consuming Too Much?” Journal of Economic Perspectives 18 (3): 147–172. Diewert, W.E. 1976. “Exact and Superlative Indexes.” Journal of Econometrics 4 Bohm, B., Gleiss, A., Wagner, M., and Ziegler, D. 2002. “Disaggregated Capital (2): 115–145. Stock Estimation for Austria—Methods, Concepts, and Results.” Applied Economics 34: 23–37. Dumagan, J.C. 2002. “Comparing the Superlative Törnqvist and Fisher Ideal Indexes.” Economics Letters 76: 251–258. Chevalier, M. 2003. Chain Fisher Volume Index Methodology. Catalogue no. 13-604-MIE, no. 42. Ottawa: Statistics Canada. EC (European Commission), IMF (International Monetary Fund), OECD (Organisation for Economic Co-operation and Development), UN (United Dasgupta, P. 2001. Human Well-Being and the Natural Environment. New York: Nations), and World Bank. 2009. System of National Accounts 2008. New Oxford University Press. York: United Nations. 123 For licensed data, dummy datasets will be made available. 55 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Fraumeni, B. 2008. “Human Capital: From Indicators and Indices to Accounts.” OECD (Organisation for Economic Co-operation and Development). 2009. Paper presented at the Organisation for Economic Co-operation and Measuring Capital – OECD Manual 2009: Second edition. Paris: OECD Development workshop, “The Measurement of Human Capital,” Turin, Italy, Publishing. November 3–4, 2008. ONS (Office for National Statistics, UK). 2020. UK Natural Capital Accounts Gollier, C. 2019. “Valuation of Natural Capital under Uncertain Substitutability.” Methodology Guide: 2020. London: ONS. Journal of Environmental Economics and Management 94: 54–66 OSTP (Office of Science and Technology Policy), OMB (Office of Management Groom, B., Drupp, M.A., Freeman, M.C., and Nesje, F. 2022. “The Future, Now: and Budget), and DOC (Department of Commerce). 2023. National Strategy A Review of Social Discounting.” Annual Review of Resource Economics 14: to Develop Statistics for Environmental-Economic Decisions: A U.S. System 467–491. of Natural Capital Accounting and Associated Environmental-Economic Statistics. Edited by Office of Science and Technology Policy, Office of Hamilton, K., and Clemens, M. 1999. “Genuine Savings Rates in Developing Management and Budget, and Department of Commerce. Washington, DC: Countries.” The World Bank Economic Review 13 (2): 333–356. The White House. Hamilton, K., and Liu, G. 2014. “Human Capital, Tangible Wealth, and the Pearce, D., and Atkinson, G. 1993. “Capital Theory and the Measurement of Intangible Capital Residual.” Oxford Review of Economic Policy 30 (1): Sustainable Development: An Indicator of ‘Weak’ Sustainability.” Ecological 70–91. Economics 1993: 103–108. Hartwick, J.M. 1978. “Substitution among Exhaustible Resources and Sefton, J.A., and Weale, M.R. 2006. “The Concept of Income in a General Intergenerational Equity.” The Review of Economic Studies 45 (2): 347–354. Equilibrium.” Review of Economic Studies 73: 219–249. IMF (International Monetary Fund). 2009. Balance of Payments and Siikamäki, J., Piaggio, M., da Silva, N., Álvarez, I., and Chu, Z. 2024. “Global International Investment Position Manual. Washington, DC: International Assessment of the Economic Value of Non-wood Forest Ecosystem Services.” Monetary Fund. CWON 2024 technical report, World Bank, Washington, DC. Inklaar, R., Gu, W., and Diewert, E. 2023. Deflation in the Changing Wealth of Smith, R.B. 2023. “Guidance Note on the Treatment of Renewable Energy Nations Database. Technical report. Washington, DC: World Bank. Resources as Assets.” SNA/M4.22/14 WS.11 Renewable Energy Resources. Jorgenson, D.W., and Fraumeni, B.M. 1989. “The Accumulation of Human Sumaila, U.R., Teh, L., Palomares, D., Alam, L., and Cheung, W. 2024. “Global and Non-human Capital, 1948–1984.” In The Measurement of Saving, Ocean Fish Wealth Accounts.” CWON 2024 technical report, World Bank, Investment, and Wealth, edited by R. Lipsey and H. Tice, 227–282. Chicago: Washington, DC. University of Chicago Press, National Bureau of Economic Research. Törnqvist, L. 1936. “The Bank of Finland’s Consumption Price Index.” Bank of Jorgenson, D.W., and Fraumeni, B.M. 1992a. “Investment in Education and U.S. Finland Monthly Bulletin 16 (10): 27–34. Economic Growth.” Scandinavian Journal of Economics 94 (Suppl.): S51–S70. UN (United Nations), EU (European Union), FAO (Food and Agriculture Jorgenson, D.W., and Fraumeni, B.M. 1992b. “The Output of the Education Organization of the United Nations), IMF (International Monetary Fund), Sector.” In The Output of the Service Sector, edited by Z. Griliches, T. OECD (Organisation for Economic Co-operation and Development), and Breshnahan, M. Manser, and E. Berndt. Chicago: National Bureau of World Bank. 2014. System of Environmental-Economic Accounting 2012— Economic Research. Central Framework. New York: United Nations. Kunte, A., Hamilton, K., Dixon, J., and Clemens, M. 1998. “Estimating National UN (United Nations), EU (European Union), FAO (Food and Agriculture Wealth: Methodology and Results.” Environment Department Paper 57, Organization of the United Nations), IMF (International Monetary Fund), World Bank, Washington, DC. OECD (Organisation for Economic Co-operation and Development), and World Bank. 2021. System of Environmental-Economic Accounting— Lam, V., and Sumaila, R. 2021 “A Practical Approach for Estimating Marine Ecosystem Accounting (SEEA EA). New York: White cover publication, Fisheries Asset Value.” CWON 2021 technical report, World Bank, pre-edited text subject to official editing. https://seea.un.org/ecosystem- Washington, DC. accounting. Lane, P., and Milesi-Ferretti, G.M. 2007. “The External Wealth of Nations Ward, M. 1976. The Measurement of Capital: The Methodology of Capital Stock Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, Estimates in OECD Countries. Paris: OECD. 1970–2004.” Journal of International Economics 73: 223–250. World Bank. 2006. Where Is the Wealth of Nations? Measuring Capital for the Lane, P., and Milesi-Ferretti, G.M. 2018. “The External Wealth of Nations 21st Century. Washington, DC: World Bank. Revisited: International Financial Integration in the Aftermath of the Global Financial Crisis.” IMF Economic Review 66: 189–222 World Bank. 2021. The Changing Wealth of Nations 2021: Managing Assets for the Future. Washington, DC: World Bank. Mas, M., Perez, F., and Uriel, E. 2000. Estimation of the Stock of Capital in Spain. Review of Income and Wealth 46: 103–116. World Bank. 2024. The Changing Wealth of Nations 2024: Methods and Data. Washington, DC: World Bank. Menéndez, P., Beck, M.W., Abad, S., and Losada, I.J. 2024. “Building Coastal Resilience with Mangroves: The Contribution of Natural Flood Defenses to World Bank, ILO (International Labour Organization), IMF (International the Changing Wealth of Nations.” CWON 2024 technical report, World Bank, Monetary Fund), OECD (Organisation for Economic Co-operation and Washington, DC. Development), and UN (United Nations). 2004. Producer Price Index Manual: Theory and Practice. Washington, DC: International Monetary Fund. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 56 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Annex A1: Methodological Choice to Measure Real Wealth in CWON 2024 WHY MEASURING WEALTH IN changes in the relative scarcity of assets. This new approach REAL TERMS USING A TÖRNQVIST is aligned with international best practice by national VOLUME INDEX IS APPROPRIATE statistical offices and leading data initiatives, like the PWT. In fact, the PWT uses physical volumes to estimate produced The primary objective of any measurement of activity in capital in real terms using a chained Törnqvist volume “real” terms is separating price and volume effects. By index,125 which has been the main source of data for CWON’s removing the general price (inflation/deflation) effect, the produced capital estimates. Using the same approach for all real measure will only capture the volume and relative price other asset classes ensures consistency in methodological effect. One way of doing this is by using an appropriate price choices across produced, natural, and human capital. index, matching the content of the nominal series, to remove general price changes and isolate volume movements, WHY MEASURING WEALTH IN accounting for changes in relative prices. An even better way REAL TERMS USING A PRICE-BASED is to draw on physical quantity data in volume indexes. This DEFLATOR IS INAPPROPRIATE best practice is adopted by many national statistical offices, like Canada and the United States, where physical volumes The GDP deflator only covers current flows of domestic are used whenever available to estimate, for example, production (not consumption) within a given country and elements of GDP in real terms. excludes all imported goods or services. While it may be appropriate to use an overall inflation indicator to understand This edition of CWON computes real wealth measures using the evolution of the real purchasing power of national a Törnqvist volume index method, where volume changes, wealth (or extended, comprehensive wealth) as a store of whether arrived at via deflation techniques or physical value, a domestic demand deflator would be preferrable or, volume indicators, are weighted by nominal shares in if that is unavailable, a deflator for household consumption adjacent periods. The current approach directly captures:124 expenditure/CPI. The GDP deflator is inappropriate in this context, since it does not match the content of the likely use ■ Changes in relative volumes due to the depletion or for national wealth (final domestic demand). accumulation of assets. Moreover, even when applying other price-based deflators, ■ Changes in relative prices due to changes in productivity there are concerns that they will not be able to capture or relative scarcity. changes of wealth in real terms. If a general indicator of ■ Changing substitution patterns over time through cross inflation is used in deflation, changes in real wealth over price and cross volume affects. time will be distorted by short-term price volatility. These are removed with a chained volume index that continuously The Törnqvist volume index accounts for changing captures relative price changes throughout the time series. substitution patterns, which is fundamental to the analysis Second, given that many countries compute real GDP with of sustainability, as substitution opportunities are affected by the price weights of a fixed base period, resulting GDP 124 For more technical details on the properties of the Törnqvist index, please refer to Dumagan (2002) and Diewert (1992). 125 The methodology is outlined in Feenstra et al. (2015). 57 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS implicit price deflators assume that cross price or cross volume effects are the same at the beginning and end of the accounting period. Such a process cannot remove substitution effects correctly if these patterns are changing over time, which is likely as assets become relatively scarcer. The current approach for estimating wealth in real terms, on the other hand, captures the necessary cross price and volume effects, since it uses a chained index, with an ideal index form (that is, the Törnqvist index).126 To deflate economic indicators (GDP, wealth) in the context of sustainability, physical volumes are ideal and in line with economic measurement principles. The chosen index form for the volume measure is a chained Törnqvist index (which is very similar to the chained Fisher ideal index used, for example, for quarterly GDP estimates in the United States), which continuously accounts for changing substitution patterns and factors in changes in relative prices in the weighting of volume measures in each successive period. REFERENCES Diewert, W.E. 1992. “Exact and Superlative Welfare Change Indicators.” Economic Inquiry 30: 565–582. Dumagan, J.C. 2002. “Comparing the Superlative Törnqvist and Fisher Ideal Indexes.” Economics Letters 76: 251–258. Feenstra, R.C., Inklaar, R., and Timmer, M.P. 2015. “The Next Generation of the Penn World Table.” American Economic Review 105 (10): 3150–3182. 126 Diewert (1992) shows that a superlative index like the Törnqvist index is a second-order approximation to the expenditure function, which means that it allows for changes between periods in the nature of the Slutsky matrix and, hence, substitution patterns. A price-based index like a GDP deflator, on the other hand, only provides a first-order approximation, assuming constant substitution patterns. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 58 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Annex A2: Compilation of the Törnqvist Volume Index The Törnqvist volume index for a given set of assets is ■ is the nominal value of all assets {1,2,…,k} included a weighted geometric mean127 of the so-called “quantity in the index in year t -1, defined as relatives” of each asset included in the index—that is, the where is the nominal value of asset a in year t -1 ratio of the quantity (or volume) of the asset in the current time period and its volume in the previous period—weighted Volume as it is used here should be understood to be a by the arithmetic average of the shares of the asset in the physical quantity (or a proxy for a quantity) of a given asset. total nominal value of all k assets in the current period and As an example, the volume of agricultural land assets is the previous period. The generic formula to compute the measured in hectares and the volume of oil assets is measured Törnqvist volume index is as follows.128 in barrels. The purpose of the volume index is to sum the volumes, even though they are measured in different units. This is accomplished by weighting the “quantity relatives” of each asset by their value shares, as described above, where: rendering them unitless and, therefore, commensurable. ■ q is the volume of asset a in year t, where a={1,2,…,k} a,t Ideally, asset volumes should reflect the underlying quality ■ q is the volume of asset a in year t -1 and a,t-1 of the assets, and how these change over time. As an ■ θ a,t is the weight of asset a in year t for all assets {1,2,…,k} example, the volume of agricultural land assets should be measured in quality-adjusted hectares; for example, if the The weight θa,t of asset a is the arithmetic average of the shares per hectare productivity of agricultural land declines by 1 of asset a in the total nominal value of all assets included in percent annually, the volume of the land also declines by 1 the index in period t and t -1, and is defined as percent annually, even if the total quantity of land (measured in hectares) does not change. As another example, not all workers are equal in terms of their contribution to human capital. Workers that are highly educated generally contribute where: more to human capital than those with fewer qualifications. ■ s is the share of asset a in the nominal value of all a,t Thus, an increase in the number of highly educated workers assets {1,2,…,k} included in the index in year t, defined will benefit human capital more than the same change in as unqualified workers. To reflect this, the “volume” of workers ■ s a,t-1 is the share of asset a in the nominal value of all should be broken down by level of education (among other assets {1,2,…,k} included in the index in year t-1, defined characteristics). The global data available for the compilation as of CWON do not permit these kinds of distinctions, which is why the human capital index from the PWT129 is used to ■ is the nominal value of all assets {1,2,…,k} included in approximate the average human capital per worker. Similarly, the index in year t, defined as where the produced capital estimates derived from the PWT reflect is the nominal value of asset a in year t both quantity and quality changes in the underlying assets. Improving this aspect of the methodology for other types of capital, most notably natural capital, is a goal for the future. 127 A geometric mean of a set of values is the nth root of the product of the values, where n is the number of values in the set. 128 Country notation is suppressed for the sake of clarity in presentation. 59 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS However, the above does not mean the index is not sensitive To make the Törnqvist volume index easier to interpret, it is to the issue of asset quality. If markets are functioning “chained” into a time series by selecting a reference year and properly, changes in the quality of assets will be observed then expressing other years in terms relative to the reference in changes in their relative nominal values. Increases in the year. CWON 2024 chose to use 2019 as the reference year. number of highly educated workers should, other things being equal, increase the value of human capital relative For the reference year (t=2019), the chained Törnqvist volume to other assets. Similarly, losses of prime farmland should index is normalized to 100, that is, decrease the relative value of farmland. Since changes in the volume of assets are weighted in the index by the shares of those assets in the total nominal value of all assets, changes For all the years before the reference year, that is, for t<2019, in relative prices will go some way toward reflecting changes the chained Törnqvist volume index is computed as in asset quality in the index. A Törnqvist volume index can be compiled according to For all the years after the reference year, that is, for t>2019, it the formula above for any period and for any number of is computed as individual asset types. In CWON, the period is the calendar year, and the number of assets depends on the specific index within the CWON accounts. The number of assets included in the CWON volume indexes ranges from 1 (in the case of an To compute the 1995–2020 time series of real values expressed index compiled for a single asset like hydroelectric resources) in “chained prices” or “real terms,” the nominal aggregate to 18 (in the case of the index of all nonrenewable natural wealth of the base year (2019) is multiplied by the chained capital assets). The volume indexes for each asset type— Törnqvist volume index in each year. produced capital, renewable natural capital, nonrenewable Table A2.1 illustrates the calculation of real national natural capital, human capital, and net foreign assets130—are comprehensive wealth in chained 2019 US dollars using a then used as inputs into an overall Törnqvist volume index Törnqvist volume index for a country with just two assets: for aggregate comprehensive wealth. farmland and oil. It should be noted that the growth rates of the Törnqvist indexes for the different asset types cannot simply be added together to arrive at the growth rate of the overall index. This is a drawback compared with previous editions of CWON, where the real values of assets were derived by applying a GDP deflator to nominal asset values and overall comprehensive wealth was simply the sum of the real values of each asset type. However, it is important to note that the grouping of assets into asset categories does not affect the overall growth in the index; any combination of assets will yield the same overall change in the index. 129 https://www.rug.nl/ggdc/productivity/pwt/?lang=en. 130 As mentioned above, net foreign assets are simply deflated using a CPI. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 60 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS TABLE A2.1 Example of two asset Törnqvist volume index FARMLAND Share of Share of previous current Farmland Nominal Weighted Proven oil Quantity year’s year’s Quantity YEAR area asset value Weight quantity reserves relative nominal nominal relative (hectares) (US$ billion) relative (barrels) national national wealth wealth 1994 50,000 25 0.357 281,000 1995 50,056 1.001 25 0.357 0.358 0.358 1.00040 279,169 0.993 1996 50,171 1.002 25 0.358 0.351 0.355 1.00082 276,931 0.992 1997 50,184 1.000 26 0.351 0.341 0.346 1.00009 276,878 1.000 1998 50,189 1.000 27 0.341 0.347 0.344 1.00003 274,977 0.993 1999 50,438 1.005 29 0.347 0.348 0.347 1.00173 274,676 0.999 2000 50,608 1.003 29 0.348 0.334 0.341 1.00114 272,046 0.990 2001 50,790 1.004 30 0.334 0.319 0.327 1.00117 271,670 0.999 2002 50,794 1.000 32 0.319 0.328 0.324 1.00003 271,559 1.000 2003 50,957 1.003 35 0.328 0.337 0.332 1.00106 270,904 0.998 2004 51,196 1.005 35 0.337 0.327 0.332 1.00156 269,910 0.996 2005 51,385 1.004 38 0.327 0.329 0.328 1.00120 268,332 0.994 2006 51,617 1.005 40 0.329 0.317 0.323 1.00146 266,005 0.991 2007 51,854 1.005 43 0.317 0.328 0.323 1.00148 264,761 0.995 2008 52,110 1.005 46 0.328 0.333 0.331 1.00163 263,226 0.994 2009 52,306 1.004 48 0.333 0.325 0.329 1.00123 262,226 0.996 2010 52,458 1.003 48 0.325 0.308 0.316 1.00092 260,943 0.995 2011 52,559 1.002 51 0.308 0.311 0.309 1.00059 258,933 0.992 2012 52,802 1.005 54 0.311 0.314 0.312 1.00144 257,721 0.995 2013 52,993 1.004 58 0.314 0.320 0.317 1.00114 256,469 0.995 2014 53,164 1.003 58 0.320 0.320 0.320 1.00104 255,733 0.997 2015 53,217 1.001 60 0.320 0.307 0.313 1.00031 255,197 0.998 2016 53,320 1.002 61 0.307 0.298 0.303 1.00058 254,402 0.997 2017 53,450 1.002 65 0.298 0.295 0.297 1.00072 253,758 0.997 2018 53,456 1.000 70 0.295 0.295 0.295 1.00003 252,038 0.993 2019 53,660 1.004 71 0.295 0.286 0.290 1.00111 251,033 0.996 2020 53,918 1.005 72 0.286 0.288 0.287 1.00138 250,966 1.000 61 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS OIL WEALTH Share of Share of Real national previous current Nominal Nominal Weighted Unchained Chained comprehen-sive year’s year’s national asset value Weight quantity Törnqvist Törnqvist wealth (billion nominal nominal wealth (US$ (US$ billion) relative volume index volume index chained 2019 national national billion) US$) wealth wealth 2019 = 100 45 0.643 70 1.00000 - 46 0.643 0.642 0.642 0.99581 71 0.99621 104.95 262 47 0.642 0.649 0.645 0.99482 73 0.99563 104.49 261 50 0.649 0.659 0.654 0.99988 76 0.99996 104.49 261 51 0.659 0.653 0.656 0.99549 78 0.99552 104.02 260 54 0.653 0.652 0.653 0.99929 83 1.00101 104.12 260 58 0.652 0.666 0.659 0.99368 88 0.99481 103.58 259 64 0.666 0.681 0.673 0.99907 94 1.00024 103.61 259 66 0.681 0.672 0.676 0.99972 99 0.99975 103.58 259 69 0.672 0.663 0.668 0.99839 103 0.99945 103.52 259 73 0.663 0.673 0.668 0.99755 108 0.99910 103.43 259 78 0.673 0.671 0.672 0.99607 117 0.99726 103.15 258 86 0.671 0.683 0.677 0.99412 125 0.99557 102.69 257 89 0.683 0.672 0.677 0.99683 132 0.99831 102.52 256 92 0.672 0.667 0.669 0.99611 138 0.99774 102.29 256 100 0.667 0.675 0.671 0.99745 148 0.99868 102.15 255 108 0.675 0.692 0.684 0.99665 156 0.99757 101.90 255 112 0.692 0.689 0.691 0.99468 163 0.99527 101.42 254 117 0.689 0.686 0.688 0.99678 171 0.99821 101.24 253 122 0.686 0.680 0.683 0.99668 180 0.99782 101.02 253 124 0.680 0.680 0.680 0.99805 183 0.99908 100.93 252 136 0.680 0.693 0.687 0.99856 196 0.99887 100.81 252 142 0.693 0.702 0.697 0.99783 203 0.99841 100.65 252 155 0.702 0.705 0.703 0.99822 219 1.00000 100.65 252 168 0.705 0.705 0.705 0.99521 239 0.99525 100.17 251 179 0.705 0.714 0.710 0.99717 250 0.99827 100.00 250 179 0.714 0.712 0.713 0.99981 252 1.00119 100.12 250 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 62 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS 3 Global and Regional Trends in Wealth, 1995–2020 MAIN MESSAGES INTRODUCTION ■ A key measure of economic progress is how the real A key measure of progress is how the real wealth of a wealth per capita of a nation—consisting of produced nation—consisting of produced capital, nonrenewable capital, nonrenewable natural capital, renewable natural capital, renewable natural capital, human capital, natural capital, human capital, and net foreign assets— and net foreign assets—changes over time relative to changes over time. population growth. As long as real wealth per capita does not decline, economic development is weakly sustainable ■ Real wealth per capita grew by 21 percent globally (Hartwick 1978; Hamilton and Clemens 1999). As in previous between 1995 and 2020 due to significant increases in editions of CWON, the wealth measurement uses the best human and produced capital. However, while two- possible global data sources available and (where possible) thirds of countries in the CWON database saw an methods aligned with the internationally accepted statistical increase in their real wealth per capita, 27 countries standards and guidelines of the System of National Accounts experienced declines or saw little change. Most of (SNA; EC et al. 2009) and its extension, the System of these countries were low-income nations. Some were Environmental-Economic Accounting (SEEA; UN et al. 2014, also fragile and affected by conflict. 2021). Different to previous editions, changes in real wealth per capita (computed using a volume-based index; see ■ Changes in the composition of the asset portfolio chapter 2) are now driven by changes in the physical volumes and population growth have driven changes in real (for example, a country’s fossil fuel reserves, fish stocks, or wealth per capita. While the accumulation of produced labor force) and relative price changes, reflecting changes in and—to a lesser extent—human capital has kept pace the relative scarcity of the stocks being measured, changes in with population growth, renewable natural capital the availability of potential substitutes or complements, or per capita has experienced dramatic declines across institutional and policy changes. the world. Nonrenewable natural capital, on the other hand, has experienced more volatile trends, driven by The interpretation of changes in CWON’s real comprehensive changes in market conditions, new discoveries, and wealth per capita metric depends on how wealth is measured. technological innovations. If one could measure wealth comprehensively, changes in real wealth per capita could be interpreted as welfare ■ GDP only tells a partial story of economic progress. changes (see chapter 1). However, data constraints limit the While most countries are either getting richer and ability of the CWON work program to include key assets, wealthier, or poorer and more impoverished, more most notably renewable natural capital and ecosystem than 25 percent of countries have experienced positive assets. The CWON database captures most of the productive GDP per capita growth while their real wealth per asset base of the economy because it includes all key assets capita has declined. To ensure their economic growth of the SNA non-financial balance sheet as well as human is sustainable, it will be critical for these countries to capital and key ecosystem services that support economic continue investing in a diversified asset base. production. Therefore, changes in real wealth per capita can proxy for changes in the overall sustainability of economic progress. Increases in real wealth per capita over time would suggest that a country’s economic progress is sustainable, 63 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS while declines would indicate that current production and From a regional perspective, wealth is concentrated in North consumption choices are eroding the productive base and, America (the United States and Canada), Europe (Germany, thus, diminishing the future opportunities available to the France, and the United Kingdom), and East Asia and the country in question. Pacific (China and Japan), accounting for 65 percent of global wealth. Meanwhile, the other four regions, comprising Latin This chapter provides an overview of how the wealth of America and the Caribbean, the Middle East and North nations has evolved from 1995 to 2020. It covers the main Africa, Sub-Saharan Africa, and South Asia, where more trends in real wealth per capita across income groups, given than half of the world’s population lives, only account for 14 that there are large disparities in wealth levels and trends by percent of global wealth. These wealth disparities also persist income. It also reports differences across World Bank regions, if adjustments are made for price-level differences across given the diversity of asset portfolios and development countries; the adjustments mainly reduce the gap between experiences across the globe. Moreover, given that conflicts high- and upper-middle-income countries (as illustrated in can have significant impacts on a country’s wealth (World annex A3). Bank 2023a), trends are also reported separately for countries using the World Bank’s fragility, conflict, and violence (FCV) From a sustainability perspective, trends in real wealth per classification. 131 Next, the chapter analyzes the composition capita are most important. Real wealth per capita has grown of wealth and how the main asset categories have evolved globally due to significant increases in human and produced over time. The final section draws some conclusions on how capital (Panel a, Figure 3.2), driven by rapid urbanization changes in real wealth per capita can provide new insights and growing female participation in the labor market. The on the sustainability of economic progress and complement growth in real wealth per capita relative to 1995 is particularly more standard macroeconomic measures, such as GDP. pronounced for the Middle East and North Africa region (97 percent increase) and Latin America and the Caribbean (66 percent increase). This growth is largely attributed to GLOBAL TRENDS IN WEALTH PER substantial rises in human capital (82 percent and 61 percent, CAPITA BY REGION, INCOME respectively) and produced capital (138 percent and 83 GROUP, AND FCV STATUS percent, respectively) in these regions, but has not resulted Wealth is highly concentrated in rich countries, with high- in substantial increases in their share in global wealth (which income countries making up more than two-thirds of total stood at 4 percent and 3 percent, respectively, in 2020; Panel wealth in nominal terms in 2020 (Figure 3.1). The wealth b of Figure 3.2). However, the accumulation of real wealth disparities are significant: upper-middle-income countries in some regions, most notably Sub-Saharan Africa, has not make up nearly a quarter of global wealth, while the rest of grown at the same speed as their populations. While there the world, where half of the world’s population lives, accounts were periods of growth in real wealth per capita between for merely 7 percent. Moreover, there is no evidence that this 1995 and 2010, real wealth per capita has been declining wealth gap has been closing significantly over the last quarter in Sub-Saharan Africa since then due to depletion and of a century, especially for the poorest countries. While overexploitation of natural capital (Panel c, Figure 3.2). middle-income countries were able to almost double their While two-thirds of the 151 countries in our sample have share in wealth—from 11 to 23 percent for upper-middle- experienced growth in real wealth per capita over the last income countries and 4 to 6 percent for lower-middle-income quarter of a century, 27 countries have experienced declines countries—the share of low-income countries has largely or have seen little change (Figure 3.3). Most countries have remained the same at less than 1 percent of global wealth been able to increase their real wealth per capita, some of since 1995. them substantially, primarily due to growth in human capital. 131 The “FCV” concept groups three issues that are often related: (i) deep governance issues and state institutional weaknesses; (ii) situations of active conflict; and (iii) high levels of interpersonal and/or gang violence (World Bank 2023a). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 64 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.1 Distribution of global wealth, by income group, 1995 and 2020 Panel a: Global wealth, 1995 Panel b: Global wealth, 2020 4% 6% 11% 0% 1%  High income 23%  Upper middle income  Lower middle income  Low income 85% 70% Source: World Bank staff estimates. Note: Wealth is measured in current US dollars. World Bank income classifications groupings are used and kept constant from the latest year in the dataset. Changes in shares reflect changes in wealth for a consistent group rather than classification changes. FIGURE 3.2 Real wealth per capita, by region, 1995–2020 Panel a: Trends in real wealth per capita, by region, 1995–2020 Panel b: Nominal wealth shares, (1995=100) by region, 2020 3% capita in chained 2019 USD (1995=100) 200 3% National comprehensive wealth per 4% 2% 32% 150 27% 100 29% 50 1995 2000 2005 2010 2015 2020  North America  East Asia & Pacific  Europe & Central Asia  North America  East Asia & Pacific  Latin America & Caribbean  Europe & Central Asia  Latin America & Caribbean  South Asia  South Asia  Middle East & North Africa  Middle East & North Africa  Sub-Saharan Africa World  Sub-Saharan Africa 65 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Panel c: Trends in real wealth per capita, decadal, 2000 and 2010 (1995=100) 140 130 Real wealth per capita (indexed) 120 110 100 1995 1999 2000 2009 2010 2020  East Asia & Pacific  South Asia  North America  Sub-Saharan Africa  Europe & Central Asia World  Middle East & North Africa  Latin America & Caribbean Source: World Bank staff estimates. Note: Real wealth per capita is computed using the Törnqvist volume index. For the relative volume changes, chained Törnqvist volume indexes are used for produced capital, nonrenewable and renewable natural capital, human capital, and net foreign assets (for net foreign assets, the nominal asset value is deflated using the CPI). The weights are calculated using their respective nominal asset value relative to nominal wealth. The Törnqvist volume index for wealth is then chained with a base year of 2019 and real wealth is computed using the nominal wealth estimate for 2019. Regional wealth per capita is computed as a sum of real wealth for the countries in the region divided by the regional population. In Panel a, changes in real wealth per capita for each region are reported relative to 1995 (set equal to 100). In Panel b, nominal wealth is measured in current US dollars and shares are reported in percent. In Panel c, the time period is segmented into three decadal periods, with the first assigned 5 years as the total time series is 25 years. As human capital makes up nearly two-thirds of global in better infrastructure, increasing the depletion of natural wealth, any increases in human capital due to higher labor resources, or constraining the potential of accumulating force participation or increasing educational attainment will human capital. In 1995, each person living in FCV countries translate into substantial increases in real wealth per capita. had an average wealth of $15,650, compared to $103,000 in However, several countries across regions and income groups the rest of the world’s countries. This shows a staggering have experienced declining wealth per capita, most notably wealth inequality in the world that continues to worsen. many low-income countries, especially in Sub-Saharan In the following 25 years, wealth per capita in non-FCV Africa (Figure 3.3). Such a decline in real wealth per capita countries reached $128,000, but in FCV countries it only signals that economic progress is unsustainable. increased to about $15,970. This increasing wealth inequality can be observed in Figure 3.4, where non-FCV countries have One possible driver of the declines in real wealth per capita experienced a steady increase in real wealth per capita, while could be conflict, with more than 40 percent of the countries the early gains in FCV countries were offset by subsequent on an unsustainable development path being classified as declines, leading to an average increase of merely 2 percent FCV countries. In fact, fragility, conflict, and violence can between 1995 and 2020. have significant impacts on wealth by limiting investments REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 66 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.3 Countries with declining and non-declining real wealth per capita, 1995–2020 Per capita change (0%) ■ Declining wealth ■ Non-declining wealth ■ No data Source: World Bank staff estimates. Note: Percent changes in real wealth per capita are computed for the 1995–2020 period. This figure distinguishes between countries that experienced declining wealth (percent change < 0) vs. non-declining wealth (percent change >= 0) during this period but does not compare wealth per capita nominal values directly. As discussed throughout the report, non-declining wealth per capita is a minimum requirement for sustainable economic growth, although not a sufficient condition. FIGURE 3.4 Trends in real wealth per capita, by FCV status, 1995–2020 (1995=100) 125 Wealth per capita (1995=100) 120 115 110 105 100 1995 2000 2005 2010 2015 2020  Non-FCV  FCV Source: World Bank staff estimates. Note: Real wealth per capita is computed using the Törnqvist volume index. For details on the computation of global real wealth per capita refer to the notes of Figure 3.2. 67 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS GLOBAL TRENDS IN PER CAPITA a one-third reduction. For nonrenewable natural capital WEALTH BY ASSET CATEGORY per capita, the trends are even more pronounced, with substantial declines for low-income and lower-income The growth in real wealth per capita is driven by changes in countries (60 percent and 17 percent, respectively) but the composition of the asset portfolio relative to population increases at higher levels of development (30 percent in growth, with increases in produced and human capital high-income countries). For produced and human capital, driving change over the last quarter of a century. Four main all income groups experienced growth, but the relative asset categories make up CWON’s wealth measure: produced magnitude of the increases varies, especially for per capita capital, nonrenewable and renewable natural capital, and produced capital. For example, upper-middle-income human capital.132 Human capital, which makes up the largest countries saw the largest relative increase of 234 percent share of nominal wealth at 60 percent in 2020 (Panel b, Figure relative to 1995, followed by low- and lower-middle-income 3.5), has increased by about 9 percent in per capita terms countries (143 percent and 97 percent, respectively). relative to 1995 (Panel a, Figure 3.5). Produced capital is the second-most important asset category, making up one-third of global nominal wealth. Despite the fast-growing world RENEWABLE NATURAL CAPITAL population, produced capital per capita has accumulated Although renewable natural capital wealth is increasing in most rapidly, with an increase of 47 percent in per capita some parts of the world (Panel a, Figure 3.7), no region’s terms between 1995 and 2020. growth was fast enough to keep up with population growth, However, the trends are starkly different for the components leading to drastic declines in per capita terms (Panel b, Figure of natural capital. Nonrenewable natural assets, which make 3.7). Sub-Saharan Africa and the Middle East and North up 2 percent of global wealth, have slightly declined in per Africa have experienced the largest declines in renewable capita terms over the same period. This is the most volatile natural capital wealth per capita since 1995 of about 40 asset category, where years of growth are followed by sudden percent, with South Asia losing about one-third. Regions losses of value, partly driven by changes in the underlying with high-income countries have seen the smallest losses, asset base due to discoveries and technological innovations though these are still substantial, ranging from 8 percent for as well as price fluctuations. Meanwhile, renewable natural Europe and Central Asia to 18 percent for North America. capital, which should be able to regenerate itself if managed The declining value of renewable natural capital per capita sustainably, has declined by more than 20 percent in per is driven by an overexploitation across almost all renewable capita terms over the past quarter of a century. This decline natural resources included in this report—agricultural land, and the capital’s 6 percent share in global nominal wealth forests, marine capture fisheries, and mangroves. The only are likely underestimates, as data and conceptual concerns exception is hydropower (added with this edition), which saw limit the ability to measure and value renewable natural a steady increase in per capita wealth over time. capital comprehensively. The degradation trends vary greatly across different forms These global trends mask significant differences across of renewable natural capital, with the most dramatic income groups. Renewable natural capital per capita has declines observed in marine fish stocks per capita. Marine declined dramatically in real terms across all income groups, fish stocks per capita have dropped by more than 45 but the largest relative declines are observed for the lowest percent in real terms since 1995, almost two times faster levels of development. Low-income countries lost nearly than mangroves and timber (Panel a, Figure 3.8). This has half of their real renewable natural capital per capita over nearly reduced its contribution to the value of renewable the last 25 years, while lower-middle-income countries saw natural capital to zero (Panel b, Figure 3.8). 132 Net foreign assets are not included in this global comparison because, by definition, global external assets equal global external liabilities. That this cannot be shown is an empirical discrepancy due to inconsistent reporting, as discussed by Lane and Milesi-Ferretti (2007). Lane and Milesi-Ferretti found that an underreporting of external assets relative to external liabilities led to a similar magnitude of global discrepancy in net foreign assets as is found for national current accounts (trade) and financial accounts (financial flows). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 68 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.5 Wealth per capita, by asset category Panel a: Trends in real wealth per capita, by asset category, Panel b: Nominal wealth shares, 1995–2020 (1995=100) by asset category, 2020 160 6% in chained 2019 USD (1995=100) 140 2% Wealth per capita 120 32% 100 60% 80 60 1995 2000 2005 2010 2015 2020  Human capital  Renewable natural capital  Produced capital  Produced capital  Nonrenewable natural capital  Human capital  Renewable natural capital  Nonrenewable natural capital Source: World Bank staff estimates. Note: Real wealth per capita by asset category is computed using the Törnqvist volume index. For the relative volume changes, physical measurements of the assets in each asset category are used. The weights are calculated using their nominal asset value relative to the nominal value of their respective asset category. The Törnqvist volume index for each asset category is then chained with a base year of 2019 and the real value of each asset category is computed using the nominal value of each asset category for 2019. The global real value of each asset category (measured in chained 2019 US dollars) is computed by summing across all countries and dividing by the global population. Changes in real comprehensive wealth per capita for each asset category are reported relative to 1995 (set equal to 100). Nominal wealth is measured in current US dollars and shares are reported in percent. FIGURE 3.6 Change in wealth per capita, by income group and asset class, 1995–2020 300  Total wealth  Produced capital Changes in per capita wealth between 1995 234  Renewable natural capital 200  Nonrenewable natural capital  Human capital 143 and 2020 (%) 97 Source: World Bank staff estimates. 100 73 Note: Real wealth per capita is 55 computed using the Törnqvist volume 28 30 28 37 31 30 30 20 index. For details on the computation of 0 real wealth per capita refer to the notes 0 -12 of Figure 3.2. Real wealth per capita -17 -18 -29 by income group is computed as a sum -44 -60 of real wealth for the countries in the -100 income group divided by its population. Changes in real wealth per capita for Low income Lower middle Upper middle High income each income group are reported relative income income to 1995 (set equal to 100). 69 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.7 Trends in renewable natural capital, by region, 1995–2020 Panel a: Renewable natural capital 110 East Asia & Pacific Renewable natural capital (indexed, 1995=100)   Sub-Saharan Africa  Middle East & North Africa 105  South Asia  Europe & Central Asia  Latin America & Caribbean  North America 100 World Source: World Bank staff estimates. Note: Real wealth per capita for each 95 asset is computed using the Törnqvist 1995 2000 2005 2010 2015 2020 volume index. The relative volume changes use physical measurements of agricultural land (in square kilometers), Panel b: Renewable natural capital per capita timber (in hectares), mangroves (in hectares), non-wood forest ecosystem 100 services (in square kilometers), hydropower (in GWh), and fish stocks Renewable natural capital per capita (in tons). No weighting is used for 90 measuring individual assets in real terms. (indexed, 1995=100) The Törnqvist volume index for each asset is then chained with a base year 80 of 2019 and the real value of each asset is computed using the nominal asset value for 2019. Regional renewable natural capital per capita in real terms 70 is computed as a sum of real renewable natural capital wealth for the countries in the region divided by the regional 60 population. Changes in real wealth per capita for each region are reported relative to 1995 (set equal to 100). 1995 2000 2005 2010 2015 2020 In some countries, these declines are even more dramatic. though less dramatic declines. For example, timber per For example, Pakistan has lost more than 66 percent of its capita has lost about 53 percent of its value in Sub-Saharan marine fish stocks per capita wealth due to overfishing, while Africa in just 25 years because the continent’s forest cover Belize has lost about 50 percent of its mangroves per capita has shrunk by about 9 percent. Notably, renewable energy value relative to 1995. from hydropower had a remarkable increase of 23 percent over the same 25-year period, making up 7 percent of the Meanwhile, other renewable natural capital components— overall value of renewable natural capital. It more than such as agricultural land, which is the most important tripled for water-abundant countries like Guinea and component of renewable natural capital (accounting for Vietnam. However, other important renewable energy 73 percent of its global value) and non-timber forest recreation assets such as solar, wind, and geothermal assets could not ecosystem services (12 percent)133—have experienced similar be included in this edition due to data limitations. 133 The non-wood forest ecosystem services included in CWON include water services (7 percent), recreation services (4 percent), and non-wood forest products (1 percent). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 70 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.8 Renewable natural capital per capita, by asset, 1995–2020 Panel a: Trends in renewable natural capital per capita, by asset, Panel b: Nominal wealth shares for 1995–2020 (1995=100) renewable natural capital, by asset, 2020 120 2% 4% in chained 2019 USD (1995=100) 6% 1% 7% Wealth per capita 100 7% 32% 80 73% 60 1995 2000 2005 2010 2015 2020  Agricultural land  Water ES  Agricultural land  NWFP  Hydropower  Water ES  Mangroves  Timber  Hydropower  Recreation ES  Recreation ES  Timber  Fish stocks  Mangroves  NWFP  Fish stocks Source: World Bank staff estimates. Note: Real wealth per capita for each asset is computed using the Törnqvist volume index. The relative volume changes use physical measurements of agricultural land (in square kilometers), timber (in hectares), mangroves (in hectares), non-wood forest ecosystem services (in square kilometers), hydropower (in GWh), and fish stocks (in tons). No weighting is used for measuring individual assets in real terms. The Törnqvist volume index for each asset is then chained with a base year of 2019 and the real value of each asset is computed using the nominal asset value for 2019. The global real value of each asset (measured in chained 2019 US dollars) is computed as the sum of real wealth for each asset divided by the global population. Changes in the real asset value per capita are reported relative to 1995 (set equal to 100). ES = ecosystem services. NWFP = non-wood forest products. Nominal wealth is measured in current US dollars and shares are reported in percent. NONRENEWABLE NATURAL growth in the underlying asset base caused by discoveries CAPITAL and technological innovations followed by sudden declines in wealth driven by increased extraction. The low-carbon In contrast to the large depletion trends observed for transition is likely to affect the nonrenewable natural capital renewable natural capital, the world’s nonrenewable natural wealth estimates in the short to medium term. However, large capital—spanning oil, natural gas, coal, metals, and minerals— decreases in carbon-intensive fossil fuels (except for coal) have has been stable in per capita terms. Globally, nonrenewable not yet been observed, as they still make up nearly 60 percent natural capital per capita decreased by 2.5 percent between of the global value of nonrenewable natural capital (Panel b, 1995 and 2020, with small increases in oil wealth offsetting Figure 3.9). The second-largest share (36 percent) is metals declines in coal, natural gas, and minerals (Panel a, Figure and minerals, with coal making up the remaining 8 percent. 3.9). However, this trend was volatile, with periods of fast 71 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.9 Nonrenewable natural capital per capita, by asset, index to 1995, 1995–2020 Panel a: Nonrenewable natural capital per capita, by asset, Panel b: Shares in nominal wealth for 1995–2020 (1995=100) nonrenewable natural capital, by asset, 2020 120 8% in chained 2019 USD (1995=100) 100 14% Wealth per capita 43% 80 36% 60 1995 2000 2005 2010 2015 2020  Oils  Coal  Oils  Minerals Nonrenewables  Minerals  Natural gas  Natural gas  Coal Source: World Bank staff estimates. Note: Real nonrenewable natural capital per capita is measured in chained 2019 US dollars. For details on the computation of real wealth per capita per asset refer to the notes for Figure 3.7. The relative volume changes use physical measurements of oil (in barrels), gas (in terajoules), coal (in tons), and minerals (in tons). Nominal wealth is measured in current US dollars and shares are reported in percent. PRODUCED CAPITAL per capita has been in emerging economies. Most notably, produced capital per capita in South Asia has increased by Rapid urbanization and industrialization in high-income and nearly 500 percent, albeit from a very low level. The region’s emerging economies have produced substantial growth in share in the nominal value of produced capital was merely produced capital wealth. On average, there is about 47 percent 1 percent in 2020. Similar though less dramatic trends are more produced capital per capita in the world than there observed in the Middle East and North Africa, with a 138 was in 1995, and it has accumulated faster than population percent increase between 1995 and 2020. This increased the growth in all regions (Panel a, Figure 3.10). In 2020, most region’s share to 1 percent in the global value of produced of the produced capital assets were concentrated in North capital. Sub-Saharan Africa, on the other hand, experienced America, Europe, and East Asia and the Pacific, which make the lowest growth rates on a per capita basis. Although the up 94 percent of the global value (Panel b, Figure 3.10). Over region made significant strides in accumulating produced the 1995–2020 period, they experienced steady growth rates, capital wealth in real terms—matching that of other regions with a 129 percent increase over 25 years—rapid population ranging from 48 to 89 percent. growth led to a modest 17 percent increase in produced However, the most substantial growth in produced capital capital per capita from 1995 to 2020. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 72 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.10 Produced capital per capita, by region, 1995–2020 Panel a: Trends in renewable natural capital per capita, Panel b: Nominal wealth shares for by region, 1995–2020 (1995=100) renewable natural capital, by region, 2020 1% 600 3% in chained 2019 USD (1995=100) 1% Produced capita per capita 500 38% 400 27% 300 200 100 29% 0 1995 2000 2005 2010 2015 2020  North America  North America  East Asia & Pacific  East Asia & Pacific  Europe & Central Asia  Latin America & Caribbean  Europe & Central Asia  South Asia  Middle East & North Africa  Latin America & Caribbean  Sub-Saharan Africa World  Middle East & North Africa  Sub-Saharan Africa  South Asia Source: World Bank staff estimates. Note: Real produced capital per capita is measured in chained 2019 US dollars. For details on the computation of real wealth per capita for produced capital refer to the notes for Figure 3.7. For the relative volume changes, the following data are used: capital stock estimates from the Penn World Table 10.0 and urban land area estimates based on World Bank staff estimates using data from the United Nations Population Division’s World Urbanization Prospects, the Food and Agriculture Organization (FAO), and the Center for International Earth Science Information Network. The weights are calculated using their nominal asset value relative to the nominal value of produced capital. Nominal wealth is measured in current US dollars and shares are reported in percent. HUMAN CAPITAL human capital), Europe and Central Asia (27 percent), and East Asia and the Pacific (27 percent), as shown in Panel b of Human capital is a key component of wealth, constituting Figure 3.11. These regions have experienced modest growth about 60 percent of the world’s total nominal wealth value rates, with, for example, 12 percent in North America and 16 in 2020. It has grown consistently for the past 25 years due percent in East Asia and the Pacific. In contrast, the Middle to increasing labor force participation and higher returns to East and North Africa, and Latin America and the Caribbean education (Panel a, Figure 3.11). The share of human capital regions show much larger increases of 82 percent and 62 in total wealth generally increases as countries achieve percent, respectively, over the same period, albeit from a higher levels of economic development. Human capital was much lower starting point (their shares in the nominal value greater than 60 percent of wealth in upper-middle-income of human capital are 2 percent and 5 percent, respectively). and high-income countries in 2020, but only about 50 percent in low-income and lower-middle-income countries. Human The human capital estimates reveal a significant disparity capital is concentrated in the high- and upper-middle-income between the male and female shares of human capital. Little countries of North America (34 percent of the global value of progress has been made toward greater gender parity in 73 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.11 Human capital per capita by region, 1995–2020 Panel a: Trends in human capital per capita, by region, 1995–2020 Panel b: Nominal shares of human (1995=100) capital, by region, 2020 2% 3% 190 5% 180 in chained 2019 USD (1995=100) 2% 34% 170 Human capita per capita 160 150 27% 140 130 120 110 100 27% 90 1995 2000 2005 2010 2015 2020  North America  Europe & Central Asia  North America  East Asia & Pacific  East Asia & Pacific  Europe & Central Asia  Latin America & Caribbean  Latin America & Caribbean  South Asia  Middle East & North Africa  South Asia  Sub-Saharan Africa World  Middle East & North Africa  Sub-Saharan Africa Source: World Bank staff estimates. Note: Real human capital per capita is measured in chained 2019 US dollars. For details on the computation of real wealth per capita for human capital refer to the notes for Figure ES.4. For the relative volume changes, labor force numbers disaggregated by gender from the International Labor Organization (ILO) are used. These are scaled by the human capital index from the PWT to proxy for the average human capital per worker. The weights are calculated using their nominal asset value relative to the nominal value of human capital. Nominal wealth is measured in current US dollars and shares are reported in percent. human capital over 1995–2020. Globally, women accounted TRENDS IN REAL WEALTH AND GDP for less than 40 percent of human capital in 2020 at all levels PER CAPITA of human development (Panel a, Figure 3.12). The differences between regions are even more striking. For example, women GDP has long been used as a yardstick for progress, despite accounted for only 15 percent of human capital in South Asia it not being well suited to assess long-term development in 2020, while 44 percent of human capital was attributed to prospects (Stiglitz, Sen, and Fitoussi 2010; Jorgenson 2018). women in Latin America and the Caribbean (Panel b, Figure In fact, GDP is a short-term measure of the market value of all 3.12). South Asia’s large gender gap is mostly caused by a final goods and services that are produced by a country in a male-dominated labor force and many barriers that prevent given year. It is thus best suited to inform immediate fiscal and women from attaining similar economic opportunities as monetary policy questions. However, since it does not account men (World Bank 2023b). for the costs of that production (for example, when capital is used up, degraded, or destroyed in the process of generating output), it fails to provide a comprehensive picture.134 134 Net domestic product (NDP), on the other hand, was explicitly designed to account for depletion of produced capital (SNA 2008, paragraphs 2.141 and 2.142). However, it does not account for the depletion and degradation of the other assets comprising wealth. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 74 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.12 Shares of human capital by gender, 1995–2020 Panel a: By income group 80 70 60 50 40 30 20 10 0 1995 2020 1995 2020 Male share (% of total) Female share (% of total)  Low income  High income: non-OECD  Lower middle income  High income: OECD  Upper middle income 100 Panel b: By region 90 80 70 60 50 40 30 20 10 0 1995 2020 1995 2020 Male share (% of total) Female share (% of total)  East Asia & Pacific  Middle East & North Africa  South Asia  Europe & Central Asia  North America  Sub-Saharan Africa  Latin America & Caribbean Source: World Bank staff estimates. Note: OECD = Organisation for Economic Co-operation and Development. The measurement of GDP is also limited to produced capital Wealth accounting captures the value of assets that are and non-produced capital (such as natural resources, essential for long-term growth and helps monitor their including land), and fails to account for new aspects of accumulation and depletion over time. This report finds that natural capital (such as new natural resources like renewable real wealth per capita has increased by 21 percent between energy and ecosystem assets) and human capital. Whether 1995 and 2020. This contrasts with the observed increase in progress is sustainable can instead be measured by how the real GDP per capita of about 50 percent over the same period future opportunities of a nation—as measured by its real (Figure 3.13). Changes in real wealth per capita are driven by comprehensive wealth per capita—are changing. changes in the real asset base and thus capture not only the accumulation of assets over time but also their depletion. 75 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS That is, real wealth per capita will decline as capital is used continue investing in building their asset base to ensure they up, degraded, or destroyed in the process of generating can continue along a sustainable growth path. The rest of the output. Current GDP, on the other hand, often increases countries either show declines in both GDP and real wealth when asset depletion accelerates—for example, when a forest per capita or did not experience growth in GDP per capita is clearcut and timber is sold. Moreover, wealth measures but appear to accumulate real wealth per capita.135 These consider a more comprehensive set of productive assets than observed trends warrant a more detailed empirical analysis GDP, including additional natural resources like renewable to explore what is driving these different trends. However, energy and ecosystem assets. CWON provides researchers such an analysis goes beyond the scope of this report. and analysts with the most comprehensive, transparent, and rigorous global data time series of the wealth of nations to While the different asset components of wealth have been on conduct such analysis. starkly different trajectories, this has not yet acted as a brake on growth. Rising productivity and the ability to substitute A more granular look at the data reveals that these global one scarcer factor of production for a more abundant wealth trends mask large and persistent differences across one offset the erosion of the asset base, most notably for income groups and FCV status. Rich countries are becoming renewable natural capital losses. It remains an open question wealthier, while poor and conflict-affected nations are in a how long this trend can continue considering that natural downward spiral of low growth and wealth depletion. This is capital continues to be overexploited and is becoming further illustrated in Figure 3.14, which maps changes in real scarcer. The extent to which limited substitutability could be GDP per capita to changes in real wealth per capita. While accounted for within CWON through, for example, relative most countries are experiencing growth in both real GDP and price adjustments or differentiated discount rates is further wealth per capita, 15 percent of countries are experiencing explored in chapter 4. Indicative estimates suggest that positive GDP per capita growth rates while their real wealth such adjustments would substantially increase the share of per capita declines. For these countries, it will be critical to renewable natural capital in overall wealth. FIGURE 3.13 Changes in global GDP and wealth per capita, indexed to 1995, 1995–2020 160 Index 1995=100 140 120 Source: World Bank staff estimates. Note: GDP per capita is measured in constant 2015 US dollars 100 from the World Bank’s World 1995 2000 2005 2010 2015 2020 Development Indicators database (NY.GDP.PCAP.KD), and wealth per capita is measured in chained 2019  GDP per capita  Wealth per capita US dollars. 135 This group of countries includes Jamaica, Kuwait, and St. Lucia. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 76 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 3.14 Cumulative GDP per capita growth vs. cumulative wealth per capita growth, 1995–2020 Panel a: Global sample 4,000 Wealth per capita growth, 1995–2020 (%) 3,000 2,000 1,000 0 0 200 400 600 GDP per capita growth, 1995–2020 (%) Panel b: Subsample with growth rates less than 100 percent 100 Wealth per capita growth, 1995–2020 (%) 50 0 -50 -50 0 50 100 GDP per capita growth, 1995–2020 (%) Source: World Bank staff estimates. Note: GDP and wealth cumulative percentage growth between 1995 and 2020. The scatter plot shows country codes. Countries in green have increasing GDP and wealth per capita, countries in blue have declining GDP per capita but increasing wealth per capita, countries in orange have increasing GDP per capita but declining wealth per capita, countries in red have declining GDP and wealth per capita. 77 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS CONCLUSIONS country-level input data and modify assumptions as needed to support their own sustainability analysis. CWON’s estimates of real comprehensive wealth per capita CWON’s long-term ambition is to support the analysis of can be used to assess the sustainability of a nation’s economic sustainability. At this stage the extent to which this ambition progress and provide complementary information to GDP. has been fulfilled or even exceeded is an open analytical By producing these estimates, the World Bank addresses an important data gap, as all countries produce GDP, but few question. The analysis of sustainability is unavoidably produce wealth estimates. As the world tackles multiple constrained by how comprehensive the asset base is, that is, economic, environmental, and social crises, the UN Secretary- which assets are included on the CWON balance sheet and General has made “Beyond GDP” a priority. Wealth measures the precision with which they are measured (for a detailed are an important step in that direction and can be used to discussion of these issues refer to chapter 1). The assets assess sustainability trends across countries and over time, included on the CWON balance sheet have progressively and provide critical information to support development expanded and the measurement has considerably improved planning. Importantly, with this new release, it is now possible over time. While it is important to acknowledge that important to construct customized country-level wealth estimates, gaps in the coverage remain due to data and measurement building on the CWON methodology. As part of the World constraints, the CWON balance sheet is nonetheless the most Bank’s reproducibility initiative, which independently verifies comprehensive wealth database available today in terms of the reproducibility packages, the entire statistical code and coverage of assets, countries, and time series, aligned where input data 136 used to generate the nominal and real wealth possible with the internationally accepted statistical standards estimates of the CWON database will be publicly released at and guidelines in the SNA and SEEA. This alignment not http://www.worldbank.org/cwon. This unprecedented access only ensures methodological rigor, but also coherence with will provide users with the opportunity to use more granular, standard economic measures such as GDP. REFERENCES EC (European Commission), IMF (International Monetary Fund), OECD OECD (Organisation for Economic Co-operation and Development), and (Organisation for Economic Co-operation and Development), UN (United World Bank. 2014. System of Environmental-Economic Accounting 2012— Nations), and World Bank. 2009. System of National Accounts 2008. New Central Framework. New York: United Nations. York: United Nations. UN (United Nations), EU (European Union), FAO (Food and Agriculture Hamilton, K., and Clemens, M. 1999. “Genuine Savings Rates in Developing Organization of the United Nations), IMF (International Monetary Fund), Countries.” The World Bank Economic Review 13 (2): 333–356. OECD (Organisation for Economic Co-operation and Development), and World Bank. 2021. System of Environmental-Economic Accounting— Hartwick, J.M. 1978. “Substitution among Exhaustible Resources and Ecosystem Accounting (SEEA EA). New York: White cover publication, Intergenerational Equity.” The Review of Economic Studies 45 (2): 347–354. pre-edited text subject to official editing. https://seea.un.org/ecosystem- Inklaar, R., Gu, W., and Diewert, E. 2023. Deflation in the Changing Wealth of accounting. Nations Database. Technical report. World Bank. 2021. The Changing Wealth of Nations 2021: Managing Assets for Lane, P.R., and Milesi-Ferretti, G.M. 2007. “The External Wealth of Nations the Future. Washington, DC: World Bank. Mark II: Revised and Extended Estimates of Foreign Assets and Liabilities, World Bank. 2023a. “Classification of Fragile and Conflict-Affected Situations.” 1970–2004.” Journal of International Economics 73 (2): 223–250. World Bank Group Policy Brief. Washington, DC: World Bank. https://www. Stiglitz, J.E., Sen, A., and Fitoussi, J.-P. 2010. Mis-measuring Our Lives: Why GDP worldbank.org/en/topic/fragilityconflictviolence/brief/harmonized-list-of- Doesn’t Add Up. New York: The New Press. fragile-situations. World Bank. 2023b. Women, Business and the Law 2023. Washington, DC: UN (United Nations), EU (European Union), FAO (Food and Agriculture World Bank. Organization of the United Nations), IMF (International Monetary Fund), 136 For licensed data dummy datasets will be made available. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 78 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Annex A3: Measuring Wealth Using Purchasing Power Parities (PPPs) The core wealth numbers compiled for this report are general consumption PPPs may be preferable to not controlling based on market exchange rates. That is, they do not control for price levels at all. These experimental estimates in PPP for price-level differences across countries, which can be terms have been produced for the real terms volume-based significant, or market exchange rate fluctuations. There may wealth estimates (using the new volume index methodology) be a case for compiling wealth estimates and components and for nominal wealth estimates (which are used to estimate of wealth controlling for these price-level differences across the weights for assets within the overall volume index and can countries by converting them into units of purchasing power be used to assess shares of wealth). These estimates can be parity (PPP terms). Indeed, for specific uses of the wealth found in the published wealth accounts database on the World data, including cross-country comparisons of specific asset Bank data platform. groups, or comparisons with other economic statistics, data in PPP terms may be desirable. To illustrate the effect of the experimental PPP conversions, shares of global wealth (using the nominal wealth estimates) The previous edition included a set of experimental estimates are shown in Figure A3.1, where the wealth gap between of nominal wealth in PPPs for 2018 (chapter 4 of World Bank middle- and high-income countries is reduced considerably 2021). The global benchmark PPP estimates were produced when controlling for price-level differences. The wealth by the International Comparison Program, which is a shares for both lower- and upper-middle-income countries worldwide statistical initiative led by the World Bank under more than double. However, the wealth for low-income the auspices of the United Nations Statistical Commission. It countries changes only marginally, further widening the aims to provide comparable price and volume measures of wealth gap relative to the rest of the world. GDP and its expenditure aggregates among countries within and across regions. Furthermore, an assessment was conducted to determine the extent to which controlling for price levels changes As part of this edition of CWON, further experimental overall global trends. As general consumption-based PPP estimates of total wealth and components of wealth have conversions do not vary across assets within a country, been developed in PPP terms for the entire time period (1995– trends in real terms using the volume-based wealth index 2020), drawing on a new and expanded set of PPP estimates will be unaffected. However, levels of wealth will be affected, produced by the World Bank’s DECDG team.137 These estimates as wealth is shifted up or down to reflect differing purchasing are still regarded as experimental because for PPPs to capture powers of economies. Levels of wealth may be of interest consumption possibilities stemming from asset values, each in and of themselves, particularly when comparing wealth individual asset should have an asset-specific PPP factor against other economic statistics. In addition, there will be (Inklaar et al. 2023). However, asset-specific PPP factors changes to the trends of wealth and asset classes when they are not yet available. Given the importance of adjusting for are aggregated across groups of countries, as the weights purchasing power differences when analyzing the future given to different countries within the group will change consumption possibilities of a nation, the application of according to country PPP conversion factors. 137 DECDG stands for the Development Economics Vice-Presidency’s Data Group. It hosts the International Comparison Program, which produces PPPs and comparable price-level indexes under the auspices of the UN Statistical Commission. 79 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE A3.1 Global wealth shares, 2020 Panel a: In market exchange rate terms Panel b: In PPP terms 6% 13% 1% 1% 23%  High income  Upper middle income  Lower middle income  Low income 34% 52% 70% Source: World Bank staff estimates. Note: Shares of global wealth are estimated using the nominal wealth estimates in current US dollars in PPP terms. FIGURE A3.2 GDP and wealth trends per capita in market exchange rates and PPP terms Panel a: World GDP and wealth per capita (indexed) Panel b: Wealth per capita, market exchange rates and PPP terms, by income group 180 High income Low income 550 30 500 25 Chained 2019 US$, Thousands 160 20 450 15 Index 1995=100 400 10 140 1990 2000 2010 2020 1990 2000 2010 2020 Lower middle income Upper middle income 60 150 120 40 100 20 0 50 100 1995 2000 2005 2010 2015 2020 1990 2000 2010 2020 1990 2000 2010 2020  GDP per capita, using MER  Wealth per capita, using MER  Wealth per capita, using MER GDP per capita, PPP terms Wealth per capita, PPP terms  Wealth per capita, PPP terms REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 80 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Panel c: Wealth per capita, PPP terms, by region 800 Chained 2019 US$, thousands 600 1995 2020 Source: World Bank staff 400 estimates. Note: The changes in wealth per capita are estimates in 200 real terms using the volume index and are in chained 2019 US dollars in market exchange 0 rates and PPP terms. Panel South Asia North America South Asia North America Sub-Saharan Africa East Asia & Pacific Europe & Central Asia Middle East & North Africa Sub-Saharan Africa East Asia & Pacific Europe & Central Asia Latin America & Caribbean Latin America & Caribbean Middle East & North Africa c estimates are total wealth figures all in chained 2019 US dollars in both market exchange rates and PPP terms.  Wealth per capita, using MER  Wealth per capita, in PPP terms Figure A3.2 provides insights into these effects. Panel a Plotting the experimental PPP terms estimates using shows how trends in GDP per capita and wealth per capita general consumption PPPs against market exchange are affected at a global level for market exchange rate and rates corroborates that trends diverge once countries are PPP terms measures. It shows how both GDP and wealth aggregated by region or income group, but the overall growth per capita have increased growth at the global level once and sustainability assessment has not changed (see Figure purchasing power is adjusted for. This analysis has also A3.3). This is important for the economics of sustainability, shown that there are large effects for low- and middle- which is more generally concerned with long-term changes income countries, but trends are largely unaffected (Panel b). in wealth. For other uses of the wealth accounts, including At a regional level, the shift to PPP also changes the ordering comparisons with wider economic statistics, the large- of the wealthiest regions in the world as of 2020. The Middle level effects are notable across income groups and regions. East and North Africa became richer than East Asia and the Assuming future work may be able to use asset-specific PPP Pacific, while South Asia became richer than Sub-Saharan rates, further work would be required to assess whether these Africa in per capita terms. conclusions hold. Work on wealth accounting continues to evolve the production of statistical-grade wealth PPPs, and the interpretation of wealth accounts in PPP terms is an important area for further research. 81 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE A3.3 Trends in wealth asset classes in market exchange rates and in PPP terms, by income group Panel a: Produced capital Panel b: Nonrenewable natural capital High income Low income High income Low income 400 150 300 100 200 Index 1995=100 Index 1995=100 100 50 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020 Lower middle income Upper middle income Lower middle income Upper middle income 400 150 300 100 200 100 50 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020  Produced capital per capita, using MER  Nonrenewable natural capital, using MER  Produced capital per capita, PPP terms  Nonrenewable natural capital, PPP terms Panel c: Human capital Panel d: Renewable natural capital High income Low income High income Low income 140 100 130 90 120 80 110 70 Index 1995=100 Index 1995=100 100 60 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020 Lower middle income Upper middle income Lower middle income Upper middle income 140 100 130 90 120 80 110 70 100 60 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020 1995 2005 2015 2020  Human capital per capita, using MER  Renewable natural capital per capita, using MER  Human capital per capita, PPP terms  Renewable natural capital per capita, PPP terms Source: World Bank staff estimates. Note: The changes are in per capita terms using the real volume index and are in chained 2019 US dollars in market exchange rates and PPP terms indexed to 1995. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 82 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS 4 The Role of Limited Substitutability for Measuring Sustainability with CWON MAIN MESSAGES INTRODUCTION ■ The current CWON methodology implicitly assumes Measuring whether a country is growing sustainably that all assets are highly substitutable. This means remains an unresolved problem in environmental that a decline in, for example, natural capital can economics. Typically, sustainable development is defined as be compensated for through investments in other “development that meets the needs of the present without assets, such as produced or human capital. If overall compromising the ability of future generations to meet wealth increases, development is considered weakly their own needs” (WCED 1987, paragraph 27; see also Pezzey sustainable. 1992). This implies that one condition for development to be sustainable is that the aggregate per capita value of wealth ■ This assumption is unlikely to hold in the current should not decline over time (Hartwick 1978; Hamilton and context, where natural resources are in limited supply Clemens 1999). A key assumption is that natural capital and and experience widespread overexploitation and other forms of capital can be substituted for each other degradation. The degree of substitutability will likely relatively easily and that perfect complementarity can be vary as natural resources become scarcer and reach ruled out.138 As a result, investment in any form of capital critical levels due to climate change, biodiversity contributes to welfare, since a decline in, say, natural capital loss, tipping points, and the crossing of planetary can be compensated for through investments in, for example, boundaries. produced or human capital. If overall wealth increases, development is thus considered to be weakly sustainable. ■ It is thus necessary to adjust standard economic However, it seems unlikely that a high degree of substitution models for limited substitutability between natural among assets can continue in a world where natural capital and other assets. Simulations show that natural resources, such as land, are in limited supply and experience capital decline can have considerable implications for widespread overexploitation and degradation (Cohen et al. a country’s growth potential as well as its resilience 2019; Drupp and Hänsel 2021; Rad et al. 2021). The degree and fragility to natural shocks. of substitutability will thus likely vary as natural resources ■ Limited substitutability can also be introduced into become scarcer and reach critical levels due to climate wealth accounting either via differentiated discount change, biodiversity loss, tipping points, and the crossing rates or relative price adjustments. Indicative of planetary boundaries. With poor substitution, depletion estimates suggest that such adjustments substantially of resources and threats of critical scarcity or tipping points increase the share of renewable natural capital in would translate into giving a considerably higher weight— overall wealth. However, implementation is hampered typically through a shadow price (Dasgupta 2009)—for the by limited empirical evidence, and more research is depletion of natural capital. needed to determine the best approach for CWON. 138 CWON allows for perfect substitutability as well as low degrees of substitutability. That is, one type of capital can replace another and keep total wealth constant. As one of the assets becomes scarcer, substitution might still be possible, but prices will adjust in favor of the scarce good. 83 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Moreover, limited substitutability of factors of production Conventional approaches to tracking whether an economy has significant implications for the growth potential is on a sustainable trajectory, such as country-level natural of an economy and its resilience to natural shocks. For capital accounting initiatives or the CWON work program, are example, Smulders and van Soest (2023) show that with based on the weak sustainability framework, using observed limited substitutability of factors of production, economic market prices where possible. The main challenge is how to growth is ultimately determined by the slowest-growing extend the framework to account for limited substitutability factor. Limited substitutability between natural assets and by adjusting shadow prices (for example, Smulders 2012). The other factors of production, such as labor, can also shape theory on discounting and non-market valuation, for instance, economic resilience against natural shocks. Karayalcin suggests either using differentiated discount rates or relative and Onder (2023) show that the magnitude of the original (shadow) price adjustments (Weikard and Zhu 2005; Hoel impact (fragility) and the speed of the subsequent recovery and Sterner 2007; Traeger 2011; Gollier 2010; Baumgärtner et (resilience) are determined by the ability of the economy to al. 2015; and Drupp 2018). To date, there is limited empirical reallocate inputs between sectors. This ability will in turn evidence to guide the implementation of either approach. be driven by the inputs’ degree of substitutability as well as institutional characteristics, such as economic openness and This chapter explains why the existing weak sustainability property rights over natural assets. framework is not appropriate today and discusses how limited substitutability can affect both the growth and resilience of an Two key challenges of introducing limited substitutability in economy to natural shocks. Next, indicative wealth estimates the context of CWON are determining the appropriate shadow using both differentiated discount rates and relative price price of natural capital (which would require adjusting adjustments are presented, primarily to demonstrate why these the market prices currently being used) and estimating would be important adjustments. However, further research is the elasticity of substitution both across and within asset needed to find the most suitable approach to integrate limited categories. From a theoretical point of view, shadow prices substitutability in a systematic way in the CWON estimates, for sustainability accounting reflect the contribution which will be further discussed in the conclusion. of capital goods to sustainable welfare when used in combination with other assets in the future. As a result, they WHY A WEAK SUSTAINABILITY reflect whether different capital stocks are good substitutes FRAMEWORK IS NOT APPROPRIATE or not, and whether capital is used efficiently or not (that is, TODAY how externalities, suboptimal resource management, and other second-best issues play a role). Concerns about the depletion of natural resources occur with remarkable frequency. Malthus famously opined that From a practical point of view, translating market and the limited availability of land would lead to mass starvation. imputed prices to shadow prices requires adjustments along Likewise, Jevons (1865) warned about the risks of running these lines. On the one hand, both market prices and imputed out of coal as an energy resource, and the Club of Rome’s values of non-marketed goods need to be adjusted to account Limits to Growth report projected widespread shortages for for market imperfections (Dasgupta 2001). On the other hand, a host of minerals that have yet to materialize (Meadows the value of shadow prices will depend on the extent to which 1972). Market forces, combined with rising productivity and the various capital stocks are complements or substitutes in the ability to substitute one scarcer factor of production for production (Cohen et al. 2019). When inputs are substitutes a more abundant one, has meant that such pessimism has in production, the shadow price of scarce natural capital been unwarranted. This is why weak sustainability is based inputs will be higher, and the degree of substitutability lower on the premise that an economy is sustainable so long as (Smulders and van Soest 2023), requiring further adjustments there is non-declining welfare over time, irrespective of the to the market prices. fate of individual resources. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 84 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS An implicit assumption in the weak sustainability paradigm FIGURE 4.1 is that there are no scarce resources that are essential to life A framework for introducing limited substitutability or economic activity. If there is a resource of a certain quality in production and consumption that is critical, such as water or air, this would imply that there is an absolute limit to substitution between the critical Welfare asset and other assets. When an asset renders a service that is essential and irreplaceable, then losing that asset would entail σ2 much greater loss of welfare than losing a readily replaceable one. A clear implication is that the close substitutability assumption would be inappropriate in such cases. As shown Environmental Other formally by Baumgärtner et al. (2015), as natural resource good good scarcity increases, the willingness to substitute these for other human-made assets also diminishes. Though of less σ1 existential concern, the preferences of consumers also determine substitutability between commodities and, hence, Natural Other Natural whether welfare is continuously increasing as stocks of an capital capital capital asset or its services decline. Scientific evidence suggests that the world has transgressed Source: Smulders and van Soest 2023. (or is at risk of transgressing) several planetary boundaries Note: σ1 is the elasticity of substitution in production between natural and other capital. σ2 is the elasticity of substitution in consumption that are deemed critical for sustaining life and economic between the environmental and other (produced) goods. activity (Richardson et al. 2023). The update of the planetary boundaries framework finds that six of the nine boundaries are transgressed, suggesting that Earth is now well outside a rich set of possibilities with imperfect substitutability of the safe operating space for humanity. A key indicator along a production hierarchy (σ1), as shown in Figure 4.1, as of biosphere functioning that drives the Earth system—net well as imperfect substitutability in consumption (σ2). The primary production—has also been transgressed (Richardson authors derive solutions for these cases in a world of second- et al. 2023). In such contexts, the assumption of high best prices and show that with limited substitutability, substitutability between critical natural assets and human- shadow prices for natural capital are higher. In addition, the produced assets seems both misleading and inappropriate. difference between imputed prices and actual shadow prices increases as substitutability decreases. THE IMPORTANCE OF CONSIDERING LIMITED SUBSTITUTABILITY FOR Moreover, limited substitutability in both production and ECONOMIC GROWTH consumption is relevant in determining the sustainability of economic growth. In Figure 4.1, if natural capital While the problem of imperfect substitutability is often is growing slowly, and is a close enough substitute to recognized, the theoretical consequences and the produced capital, then the scarcity of natural capital will implications for measuring sustainability remain at the not impede sustainable growth, as long as produced capital frontiers of cutting-edge research. Smulders and van Soest increases sufficiently. Conversely, where there is limited (2023) make an important contribution to our understanding substitutability between natural capital and other forms of how shadow prices may alter with limited substitutability. of capital, the scarcity of natural capital would impede The paper considers a second-best world in which both economic growth. An important finding of the paper is that market prices and imputed prices for non-marketed goods with poor substitution, growth is ultimately determined by do not reflect their true shadow prices. The study considers the factor that is in relatively short supply. 85 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS THE IMPORTANCE OF CONSIDERING labor productivity in agriculture. With perfect labor mobility LIMITED SUBSTITUTABILITY between sectors, this effect would prompt labor to move to FOR ECONOMIC RESILIENCE IN manufacturing, where wages are higher. Second, it reduces the supply of food, leading to an increase in the relative price DEVELOPING ECONOMIES of food and wages (as the closed economy cannot import it). Limited substitutability between natural assets and other This factor limits the labor outflow from agriculture. Third, factors of production, such as labor, can also shape economic under a plausible assumption that poorer people allocate a resilience to natural shocks. In this regard, Karayalcin and greater share of their expenditure to food,139 the land shock Onder (2023) explore two main aspects: (i) how natural boosts the relative demand for food, reinforcing the second shocks are propagated by economic mechanisms—in this channel. case, through the reallocation of labor between sectors; and Figure 4.2 illustrates that when land and labor are highly (ii) how institutional characteristics of the economy, such substitutable (Panel a), the second and third channels as economic openness and property rights over natural dominate, and a decrease in land draws labor from assets, can shape such propagation. To address these points, manufacturing into agriculture in net terms. Conversely, the paper employs a structural transformation (general when the substitutability is low—that is, when land and labor equilibrium) model. In this model, one sector (agriculture) are complements in agriculture—the first channel dominates, uses a nature-based asset (land) in conjunction with labor, and a decrease in land pushes labor out of agriculture due while the other sector (manufacturing) uses labor exclusively to lower labor productivity and, hence, wages. The only for simplicity, to derive significant conclusions. exception in this scenario is when the land is already too Consider the economic propagation of a natural shock in a small to begin with, in which case labor is still drawn into closed economy with incomplete property rights over natural agriculture after the shock. This dependence of labor flows assets. A decrease in land (T) influences labor allocation on the asset stock, in this stylized model, land, provides an between sectors through three channels. First, it diminishes additional argument in favor of better measuring the non- marketed wealth in an economy. FIGURE 4.2 Manufacturing employment by land and elasticity of land-labor substitution Panel a: Land and labor are substitutes (σ=1.1) Panel b: Land and labor are complements (σ=0.2) n n 0.400 0.65 0.395 0.60 0.55 0.390 0.50 0.385 0.45 0.380 0.40 0.375 0.35 T T 1.0 1.5 2.0 2.5 1.0 1.5 2.0 2.5 Source: Karayalcin and Onder 2023. Note: n is the fraction of labor employed in manufacturing, T is the endowment of land, and σ is the elasticity of substitution between land and labor. 139 This assumes that consumers have non-homothetic preferences, which is a standard assumption in economic theory. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 86 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Next, consider how these responses and the degree of where obstacles to sustainability may begin to emerge. Such substitutability drive economic resilience in developing estimates can help quantify the economic impacts discussed. countries. According to the model of Karayalcin and Onder (2023), economic resilience (the ability to recover from a PRACTICAL APPROACHES shock quickly) is determined by how swiftly natural assets TO ACCOUNT FOR LIMITED can regenerate after a shock. The speed of such growth SUBSTITUTABILITY is in turn determined by two factors: (i) the stock of land: regeneration speed is higher when the stock is low;140 and Supporting any efforts to account for limited substitutability (ii) harvesting: the more labor is allocated to agriculture, in natural capital and wealth accounting initiatives would the slower the regeneration. The elasticity of substitution also require empirical estimates to adjust shadow prices. between land and labor influences both factors and, in turn, However, such evidence is often scarce (for examples, see an economy’s economic resilience to the natural shock. In Drupp 2018; Cohen et al. 2019; Rouhi Rad et al. 2021; Drupp et comparison to the case with low elasticity of substitution al. 2023a, b). Most government appraisal and environmental- (complements), the case with high elasticity of substitution economic accounting guidance thus has yet to explicitly (substitutes) has important differences: address limited substitutability of non-market goods (Groom et al. 2022). However, two approaches for dealing with ■ The economy has a higher equilibrium land stock. this empirical challenge can be applied in the context of Therefore, other things being equal, when an (identical) environmental-economic or wealth accounting (Weikard and natural shock reduces the land, land tends to recover faster. Zhu 2005; Hoel and Sterner 2007; Traeger 2011; Gollier 2010; Baumgärtner et al. 2015; Drupp 2018): differentiated discount ■ The economy allocates more labor to agriculture after a rates or relative price adjustments. shock, as explained above. Consequently, other things being equal, land tends to recover slower. Differentiated discount rates: This approach uses a lower discount rate for non-market goods and services, such as for In simulations, the former effect tends to dominate, ecosystem services derived from natural capital, than for resulting in faster land recovery after a shock when it is manufactured goods to reflect their increasing scarcity and easily substitutable with labor. Therefore, among otherwise limited substitutability. For example, guidelines by the Asian identical economies, the one with higher land-labor Development Bank, Australia, and Canada suggest using substitutability tends to be more resilient. Nonetheless, the lower discount rates for environmental goods or non-market downside is that the same economy tends to be more fragile benefits (6 percent, 4 percent, and 3 percent, respectively; against natural shocks. With a larger nature-sensitive sector Groom et al. 2022).141 By contrast, CWON applies a uniform that comes with more labor allocation to agriculture, natural discount rate of 4 percent as was done in previous CWON shocks can lead to a more drastic reduction in GDP at the reports (following World Bank 2006). This is twice as high as time of impact. the recently adopted rate of 2 percent in revised guidelines in the United States (OMB 2023), which is also the real social Both theoretical models suggest that it is vital to seek discount rate that receives most support according to expert empirical estimates of the degrees of substitutability of key recommendations (Drupp et al. 2018; Nesje et al. 2022). assets in both the production and utility spaces to identify 140 This is a common assumption in environmental economics, which assumes that the regeneration speed diminishes as the stock increases. 141 Some countries also allow the discount rate to vary over time. For example, the UK Treasury’s Green Book proposes a 3.5 percent rate, which declines to 3 percent after 30 years. See https://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-governent/the-green- book-2020. 87 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Relative price adjustments: This approach explicitly to increasing real incomes, using a forecasted growth rate of considers how relative prices of non-market goods change GDP per capita of around 2 percent per year (Christensen et al. over time compared to market traded consumption goods. 2018; Müller et al. 2022). When natural capital is declining, the Comprehensive consumption equivalents are then computed discount rate of the relative price change adjustment should for each point in time and a single discount rate is used to additionally consider a real scarcity effect, and further subtract compute future comprehensive consumption equivalents. the rate of decline from the 2 percent rate. For relative price However, since country-specific estimates are scarce,142 changes, Drupp et al. (2023b) have generated experimental global-level estimates of relative price changes are typically estimates of non-wood forest ecosystem services adjusted for used to inform governmental policy guidance (Groom and relative price changes, where forest area has been declining Hepburn 2017). For instance, the Netherlands’s discounting globally by around 0.1 percent per year. guidance recommends using a relative price change of 1 percent per year, while the UK’s Department for Environment, INDICATIVE ESTIMATES FOR Food and Rural Affairs “uplifts” the damage costs of air COMPREHENSIVE WEALTH USING pollution by 2 percent per year (Groom et al. 2022). DIFFERENTIATED DISCOUNT RATES For imperfect complements, both approaches are Applying differentiated discount rates—using a 2 percent mathematically equivalent and related in a simple formula discount rate for all renewable natural capital assets adjusted within the standard constant-elasticity-of-substitution for relative growth rates and a 4 percent discount rate for all framework (see Drupp et al. 2023b): other assets—has significant impacts on the CWON real wealth estimates. As the net present value of renewable natural Relative price change= capital is computed over a 100-year time span, lowering the Where rC is the discount rate for market goods, rE is the discount rate effectively means that future streams of benefits discount rate for non-market ecosystem services derived are given more weight, increasing the value of renewable from natural capital, σ2 is the constant-elasticity-of- natural capital wealth. The share of renewable natural capital substitution (see Figure 4.1), and gC and gE are the respective in global wealth thus doubles relative to the main estimates (forecasted) growth rates. that use a uniform discount rate, increasing by 5 percentage points in 1995 and 6 percentage points in 2020 (Figure 4.3). To illustrate how accounting for limited substitutability would affect the CWON wealth estimates, indicative estimates Using a lower discount rate also leads to a more substantial using both approaches were produced. For the limited decline in renewable natural capital per capita, which in substitutability approach using differentiated discount rates, a turn depresses growth in real wealth per capita. Renewable 2 percent discount rate is applied to renewable natural capital natural capital per capita declines by 4 percentage points in line with the latest recommendation of leading experts, more relative to when a uniform discount rate is used, while adjusted for relative growth rates (Drupp et al. 2023a), while all growth in global wealth per capita falls from 21.4 percent in other assets are discounted at the previous 4 percent discount 1995 to 17.7 percent in 2020 (Figure 4.4). The gap between rate. This captures the real income effect according to which the two wealth estimates also widens over time as natural future asset values are increasing in real monetary terms due resource degradation across the world accelerates. 142 Baumgärtner et al. (2015) were the first to estimate relative price changes at the global level, assuming that the elasticity of substitution is constant across all countries and ecosystem service types. Subsequently, they applied national growth rates to arrive at country-level results. Heckenhahn and Drupp (2022) provided the first comprehensive country-specific evidence, estimating growth rates of 15 ecosystem services and the degree of limited substitutability based on a meta-analysis of 36 German willingness-to-pay studies. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 88 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS FIGURE 4.3 Distribution of global wealth using differentiated discount rates, 1995 and 2020 Panel a: Global wealth, 1995 Panel b: Global wealth, 2020 3.7% 4.3% 5% 5.7% 1.5%  2% discount rate wedge 1.8%  Relative growth discount rate wedge  Renewable natural capital 90%  Other 88% Source: World Bank staff estimates. Notes: Wealth is measured in current US dollars. FIGURE 4.4 Trends in wealth per capita using differentiated discount rates, indexed to 1995, 1995, and 2020 120 115 110 105 100 1995 2000 2005 2010 2015 2020  4% discount rate Relative growth rate derived discount rate 2% discount rate Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. 89 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Differentiated discount rates clearly have a substantial (meta-analysis) of the WTP literature.143 This information impact on the real wealth estimates reported in CWON. is then used to estimate the income elasticities of the WTP More research is needed to determine to what extent the for ecosystem services,144 and subsequently, the elasticity of assumption of a uniform discount rate should be modified. If substitution between ecosystem services and human-made a decision is made to change the discount rate, more research goods (cf. Heckenhahn and Drupp 2022). is needed to determine whether the discount rate should vary Their main estimate of the elasticity of substitution suggests across assets, countries, and potentially even over time. a weak degree of substitutability between ecosystem services and human-made goods, with an elasticity of substitution of INDICATIVE ESTIMATES FOR 1.3 (an elasticity greater than one implies substitutability, NON-WOOD FOREST ECOSYSTEM while an elasticity less than one implies complementarity). SERVICES USING RELATIVE PRICE They also find that the degree of substitutability varies ADJUSTMENTS across types of ecosystem services (Table 4.1). It is highest The second set of experimental estimates aims to illustrate for rivalrous ecosystem services, such as recreation and how relative price adjustments affect value estimates for ecotourism—that is, activities for which alternatives renewable natural capital using non-wood forest ecosystem exist. On the other hand, the elasticity is lowest for forest services as a case study. This work was carried out by Drupp ecosystem services that cannot be substituted easily, such et al. (2023b), who have developed the first systematic as water regulation. For example, Damania et al. (2023) find global empirical evidence base to inform relative price that there are complementarities between human health adjustments of ecosystem services. In this context, relative and hydrological services provided by forests (Box 4.1). price changes are measured as the relative change in the Most 95 percent confidence intervals border or overlap valuation of ecosystem services (Hoel and Sterner 2007). Cobb-Douglas substitutability (an elasticity of unity). Thus, This will be driven by both their degree of substitutability while all cases yield mean estimates in the substitutes and changes in their relative scarcity over time, which need domain, it cannot be excluded that ecosystem services may to be estimated separately. be regarded as complements. The first step is to estimate the elasticity of substitution, which The second step in the analysis is to proxy for the global shift Drupp et al. (2023b) estimate indirectly using the inverse of the in the relative scarcity of ecosystem services. These can either income elasticity of the willingness to pay (WTP) for ecosystem be estimated using a historical time series of good-specific services (Ebert 2003). In fact, there is a large literature on growth rates (following Baumgärtner et al. 2015 or Heckenhahn estimating WTPs, where respondents are typically asked how and Drupp 2022) or derived endogenously as part of a global much they value a given ecosystem service or—put differently— integrated climate-economy assessment model (for example, what is the maximum price they would be willing to pay for Drupp and Hänsel 2021). Following Baumgärtner et al. it. These WTP estimates thus depend on the survey context, (2015), Drupp et al. (2023b) compute the difference in growth as well as the respondent’s characteristics, such as income, rates of ecosystem services and human-made goods (proxied which Drupp et al. (2023b) collect through a systematic review for by the growth rate of GDP per capita) over 1993–2016. 143 The authors conducted a meta-analysis of contingent valuation-based WTP studies using a large-scale keyword-based search strategy. The initial search resulted in 2,174 articles to which a range of exclusion criteria were applied to ensure the studies were consistent and comparable (for details see Drupp et al. 2023b). From the final set of 1,165 candidate WTP studies a random sample of 1,000 studies was analyzed in depth to ensure studies were comparable. This resulted in a final sample of 351 studies, which yielded 749 distinct WTP-income pairs. 144 This builds on previous work by Jacobsen and Hanley (2009), Richardson and Loomis (2009), Barrio and Loureiro (2010), Subroy et al. (2019), and Heckenhahn and Drupp (2022). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 90 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS TABLE 4.1 The elasticity of substitution (σ2), by ecosystem service types 97.5 MEAN 2.5 PERCENTILE N PERCENTILE Climate regulation 1.4 0.8 4.0 183 Air quality regulation 1.3 0.9 2.0 257 Water regulation 1.2 0.9 1.7 285 Erosion regulation 1.2 0.9 1.7 195 Regulating services 1.3 1.0 2.0 535 Spiritual and religious values 1.2 0.9 1.7 121 Aesthetic values 1.5 1.1 2.3 416 Recreation and ecotourism 1.7 1.0 4.8 353 Biodiversity preservation 1.3 1.0 1.8 384 Cultural services 1.5 1.1 2.1 515 Forest ecosystem services 1.2 0.9 1.8 244 Non-forest ecosystem services 1.3 1.0 2.0 607 Aggregate 1.3 1.0 1.8 851 Source: Drupp et al. 2023b. Their estimates suggest a sizable shift in the relative scarcity made goods. Similarly, there is a substantial uplift of the of ecosystem services relative to human-made goods. On present value of non-wood ecosystem services by 52 percent aggregate, when taking the arithmetic mean of growth rates, over a 100-year period.145 Figure 4.5 illustrates this further which implies Cobb-Douglas substitutability among ecosystem for the mean and 95 percent confidence interval, where the services, ecosystem services have become relatively scarcer by uplift increases as the degree of substitutability declines— nearly 1 percent per year, while GDP per capita has increased or the complementarity between inputs increases. For annually by almost 2 percent over the same period. aggregate ecosystem services, the uplift in the present value would be around 90 percent. These results clearly highlight Once these estimates are combined, it comes as no surprise the importance of accounting for limited substitutability in that the relative price adjustment, and its implications for wealth accounting and echo similar results for climate policy the value of renewable natural capital, are substantial. The appraisal (Bastien-Olvera and Moore 2021; Drupp and Hänsel authors find that the value of aggregate ecosystem services is 2021; Sterner and Persson 2008).146 increasing by around 2.2 percent per year relative to human- 145 This relative price adjustment of 1.6 percent is equivalent to an ecological discount rate of 2.4 percent. 146 This emerging literature has studied how optimal climate policy is affected by increasing scarcity and limited substitutability of non-market ecosystem services relative to manufactured goods, using good-specific discount rates and relative price changes. 91 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS BOX 4.1 FORESTS AS COMPLEMENTS TO HUMAN HEALTH F orest loss is often a consequence of economic activities that bring market and other benefits. However, at the same time, it can adversely affect the provision of forest ecosystem services and the associated socioeconomic and environmental benefits that rural communities enjoy. Which effect dominates is an empirical question. Damania et al. (2023) focus on the possible complementarity between human and natural capital, particularly the health benefits of clean water provided by an intact forest ecosystem. By linking high resolution deforestation data with health outcomes for 0.7 million children across 46 countries, they analyze how deforestation upstream impacts waterborne disease outcomes for rural households downstream. The results indicate increases in diarrheal disease incidence among children under 5 years old. They also offer new evidence of early-life exposure to deforestation on childhood stunting, which is a well-known indicator of later-life productivity. A series of robustness checks confirms that the transmission mechanism is the water channel, where deforestation upstream impacts water quality and thus health outcomes downstream. Therefore, maintaining natural capital, in this case an intact forest ecosystem, has the potential to generate meaningful improvements to long-run human capital. FIGURE 4.5 The value of non-wood forest ecosystem services, by the degree of substitutability 150 Source: Drupp et al. 2023b. Increase in non-wood forest natural capital value (in %) Note: The degree of complementarity corresponds to the income elasticity of the willingness to pay (WTP) and is the inverse of the degree of substitutability. The red line 100 in this figure shows the estimated increase in CWON’s non-timber forest natural 78% capital value (in percent) relative to the current CWON’s estimate, as a function of the degree of complementarity between 52% forest ecosystem services and human-made 50 goods, measured by the income elasticity 32% of the WTP for forest ecosystem services. The vertical black line indicates the central estimate, while the gray-shaded area indicates its 95 percent confidence interval. 0 Horizontal dashed helplines indicate the 0 0.25 0.5 0.75 1.0 1.25 1.5 corresponding increase in the public natural Degree of complementarity (income elasticity of WTP) capital value (in percent). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 92 TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS CONCLUSIONS AND WAY FORWARD In sum, where substitutability between assets is limited, the effective shadow price needs to be adjusted to account The emerging research has clearly shown that limited for the losses that could ensue from depleting natural substitutability is of critical importance from an economic capital stocks. Theoretically, there are two equivalent development point of view, as it has significant implications ways of achieving this adjustment—through differentiated for the long-run growth potential of an economy as well as discount rates or by relative price adjustments. However, for its resilience and fragility to natural shocks. Failing to the ease of implementation varies greatly across the two account for limited substitutability in wealth accounting also options and experience in using either is limited (Groom provides the wrong signal to policy makers on the importance et al. 2022). Implementing differentiated discount rates is of renewable natural capital as part of their broader straightforward, but more research is needed to determine wealth portfolio. By focusing on assets that are critical to which set of discount rates CWON should use. Relative sustainability and whose depletion brings high economic price adjustments are a more targeted way to adjust shadow costs, natural capital accounting has an opportunity to prices; however, estimating such relative prices for all assets highlight the economic significance of the complementary is a complex task. More research and data are needed to role played by natural assets in sustaining economic growth. implement such adjustments in the medium term. REFERENCES Barrio, M., and Loureiro, M.L. 2010. “A Meta-Analysis of Contingent Valuation Drupp, M.A. 2018. “Limits to Substitution between Ecosystem Services Forest Studies.” Ecological Economics 69 (5): 1023–1030. and Manufactured Goods and Implications for Social Discounting.” Environmental and Resource Economics 69 (1): 135–158. Bastien-Olvera, B.A., and Moore, F.C. 2021. “Use and Non-use Value of Nature and the Social Cost of Carbon.” Nature Sustainability 4 (2): 101–108. Drupp, M.A., Fenichel, E., Freeman, M., Gollier, C., Groom, B., Hänsel, M.C., Heal, G., Millner, A., Nesje, F., Quaas, M.F., Smulders, S., Sterner, T., Baumgärtner, S., Drupp, M.A., Meya, J.N., Munz, J.M., and Quaas, M.F. 2017. Traeger, C., and Venmans, F. 2023a. “Public Consultation Response on “Income Inequality and Willingness to Pay for Environmental Public Proposed Revisions to Circular A-4.” https://www.regulations.gov/comment/ Goods.” Journal of Environmental Economics and Management 85: 35–61. OMB-2022-0014-0130. Baumgärtner, S., Klein, A.M., Thiel, D., and Winkler, K. 2015. “Ramsey Drupp, M.A., Freeman, M.C., Groom, B., and Nesje, F. 2018. “Discounting Discounting of Ecosystem Services.” Environmental and Resource Disentangled.” American Economic Journal: Economic Policy 10 (4): Economics 61 (2): 273–296. 109–134. Christensen, P., Gillingham, K., and Nordhaus, W. 2018. “Uncertainty in Drupp, M.A., and Hänsel, M.C. 2021. “Relative Prices and Climate Policy: How Forecasts of Long-Run Economic Growth.” Proceedings of the National the Scarcity of Nonmarket Goods Drives Policy Evaluation.” American Academy of Sciences 115 (21): 5409–5414. Economic Journal: Economic Policy 13 (1): 168–201. Cohen, F., Hepburn, C.J., and Teytelboym, A. 2019. “Is Natural Capital Really Drupp, M.A., Turk, Z.M., Groom, B., and Heckenhahn, J. 2023b. “Limited Substitutable?” Annual Review of Environment and Resources 44: 425–448. Substitutability, Relative Price Changes and the Uplifting of Public Damania, R., Herrera Garcia, L.D., Kim, H., Viotti, L., Zaveri, E., Onder, S., and Natural Capital Values.” arXiv preprint, arXiv:2308.04400. https://arxiv.org/ Pantoja Vallejos, C.A. 2023. “Is Natural Capital a Complement to Human abs/2308.04400. Capital? Evidence from 46 Countries.” Working paper. Gerlagh, R., and Van der Zwaan, B.C.C. 2002. “Long-Term Substitutability Dasgupta, P. 2001. “Valuing Objects and Evaluating Policies in Imperfect between Environmental and Man-Made Goods.” Journal of Environmental Economies.” Economic Journal 111 (471): C1–C29. Economics and Management 44 (2): 329–345. Dasgupta, P. 2009. “The Welfare Economic Theory of Green National Accounts.” Gollier, C. 2010. “Ecological Discounting.” Journal of Economic Theory 145 (2): Environmental and Resource Economics 42 (1): 3–38. 812–829. Downes, R., and Nicol, S. 2020. “Designing and Implementing Gender Gollier, C. 2019. “Valuation of Natural Capital under Uncertain Substitutability.” Budgeting – a Path to Action.” OECD Journal on Budgeting 20.2. Journal of Environmental Economics and Management 94: 54–66. 93 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH TRACKING WEALTH TO MONITOR ECONOMIC PROGRESS Groom, B., Drupp, M.A., Freeman, M.C., and Nesje, F. 2022. “The Future, Now: “Earth beyond Six of Nine Planetary Boundaries.” Science Advances 9 (37): A Review of Social Discounting.” Annual Review of Resource Economics 14: eadh2458. 467–491. Rouhi Rad, M., Adamowicz, W., Entem, A., Fenichel, E.P., and Lloyd-Smith, P. Groom, B., and Hepburn, C. 2017. “Reflections—Looking Back at Social 2021. “Complementarity (Not Substitution) between Natural and Produced Discounting Policy: The Influence of Papers, Presentations, Political Capital: Evidence from the Panama Canal Expansion.” Journal of the Preconditions, and Personalities.” Review of Environmental Economics and Association of Environmental and Resource Economists 8 (6): 1115–1146. Policy 11 (2): 336–356. Smith, V.K. 2023. “Accounting for Income Inequality in Benefit Transfers: The Hamilton, K., and Clemens, M. 1999. “Genuine Savings Rates in Developing Importance of the Income Elasticity of WTP.” Journal of Environmental Countries.” The World Bank Economic Review 13 (2): 333–356. Economics and Management 118: 102781. Hartwick, J.M. 1978. “Substitution among Exhaustible Resources and Smulders, S. 2012. “An Arrow in the Achilles’ Heel of Sustainability and Wealth Intergenerational Equity.” The Review of Economic Studies 45 (2): 347–354. Accounting.” Environment and Development Economics 17 (3): 368–372. Heckenhahn, J., and Drupp, M.A. 2022. “Relative Price Changes of Ecosystem Smulders, S., and Van Soest, D. 2023. “Natural Capital Substitution: Implications Services: Evidence from Germany.” CESifo Working Paper No. 9656. for Growth, Shadow Prices, and Natural Capital Accounting.” Working paper. Hoel, M., and Sterner, T. 2007. “Discounting and Relative Prices.” Climatic Sterner, T., and Persson, U.M. 2008. “An Even Sterner Review: Introducing Change 84 (3): 265–280. Relative Prices into the Discounting Debate.” Review of Environmental Economics and Policy. Jacobsen, J.B., and Hanley, N. 2009. “Are There Income Effects on Global Willingness to Pay for Biodiversity Conservation?” Environmental and Subroy, V., Gunawardena, A., Polyakov, M., Pandit, R., and Pannell, D.J. 2019. Resource Economics 43 (2): 137–160. “The Worth of Wildlife: A Meta-Analysis of Global Non-market Values of Threatened Species.” Ecological Economics 164: 106374. Jevons, W.S. 1866. The Coal Question; An Inquiry Concerning the Progress of the Nation and the Probable Exhaustion of Our Coal-Mines. London: TNFD (Taskforce on Nature-related Financial Disclosures). 2022. “The Macmillan. TNFD Nature-Related Risk and Opportunity Management and Disclosure Framework [Beta v0.1].” https://tnfd.global/wp-content/ Karayalcin, C., and Onder, H. 2023. “The Fragility and Resilience of Nations.” uploads/2022/03/220321-TNFD-framework-beta-v0.1-FINAL.pdf. Policy Research Working Papers 10362, World Bank, Washington, DC. http:// hdl.handle.net/10986/39542. Traeger, C.P. 2011. “Sustainability, Limited Substitutability, and Non-constant Social Discount Rates.” Journal of Environmental Economics and Meadows, D.H., Meadows, D.L., Randers, J., and Behrens III, W.W. 1972. Management 62 (2): 215–228. The Limits to Growth: A Report for the Club of Rome’s Project on the Predicament of Mankind. New York: Universe Books. The Treasury, New Zealand. 2019. The Wellbeing Budget 2019. Government of New Zealand. https://www.treasury.govt.nz/sites/default/files/2019-05/b19- Müller, U.K., Stock, J.H., and Watson, M.W. 2022. “An Econometric Model of wellbeing-budget.pdf. International Growth Dynamics for Long-Horizon Forecasting.” Review of Economics and Statistics 104 (5): 857–876. United Nations. 2023. “Global Assessment of Environmental-Economic Accounting and Supporting Statistics 2022.” https://seea.un.org/sites/seea. Nesje, F., Drupp, M.A., Freeman, M.C., and Groom, B. 2023. “Philosophers and un.org/files/global_assessment_2022_background_doc_v4_clean_1.pdf. Economists Agree on Climate Policy Paths but for Different Reasons.” Nature Climate Change 13 (6): 515–522. WCED (World Commission on Environment and Development). 1987. Our Common Future. Oxford, UK: Oxford University Press. OMB (Office of Management and Budget). 2023. “Circular A4.” https://www. whitehouse.gov/wp-content/uploads/2023/11/CircularA-4.pdf. Weikard, H.-P., and Zhu, X. 2005. “Discounting and Environmental Quality: When Should Dual Rates Be Used?” Economic Modelling 22 (5): 868–878. Pezzey, J. 1992. “Sustainable Development Concepts: An Economic Analysis.” World Bank Environment Paper no. 2, World Bank Group, Washington, DC. World Bank. 2006. Where Is the Wealth of Nations? Measuring Capital for the 21st Century. Washington, DC: World Bank. Quaas, M.F., Van Soest, D., and Baumgärtner, S. 2013. “Complementarity, Impatience, and the Resilience of Natural-Resource-Dependent Economies.” World Bank. 2021. Global Program on Sustainability: Annual Report. Journal of Environmental Economics and Management 66 (1): 15–32. https://www.wavespartnership.org/sites/waves/files/kc/GPS_ AnnualReport_2020_2021.pdf. Richardson, L., and Loomis, J. 2009. “The Total Economic Value of Threatened, Endangered and Rare Species: An Updated Meta-Analysis.” Ecological Zhu, X., Smulders, S., and de Zeeuw, A. 2019. “Discounting in the Presence Economics 68 (5): 1535–1548. of Scarce Ecosystem Services.” Journal of Environmental Economics and Management 98: 102272. Richardson, K., Steffen, W., Lucht, W., Bendtsen, J., Cornell, S.E., Donges, J.F., Drüke, M., Fetzer, I., Bala, G., Von Bloh, W., and Rockström, J. 2023. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 94 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Measuring Components II. of Comprehensive Wealth 95 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH 5 The Nonrenewable Wealth of Nations MAIN MESSAGES INTRODUCTION ■ Nonrenewable natural capital wealth per capita Nonrenewable natural capital encompasses subsoil geological decreased by 1 percent globally between 1995 and assets, including oil, gas, coal, metals, and minerals. Unlike 2020. The trend was volatile, with periods of fast renewable biological resources, these assets do not grow or growth in the underlying asset base caused by regenerate over a meaningful economic time frame.147 However, discoveries and technological innovations followed each has distinct characteristics that makes them valuable by sudden declines in wealth driven by increased commodities and important inputs into the production and extraction. consumption activities of modern economies. Fossil fuels are primarily used for energy, and metals and ■ Most nonrenewable natural capital minerals have a plethora of uses, including the assets declined in per capita terms production of steel, electronics, chemical between 1995 and 2020, with the production, fertilizers, jewelry, and batteries. exception of oil, which flatlined. A country’s nonrenewable asset endowment In total terms, oil, natural gas, is mostly a factor of geology, and discovered and minerals increased, with oil through exploration. Prominent recent increasing the most, and minerals examples include the exploitation of oil the least, while coal declined and natural gas in shale formations in the 7 percent. United States and the filtration of lithium from underground brine in the deserts of Latin America. ■ Nonrenewable natural capital remains highly Changing preferences, demand, and environmental policy concentrated across countries, particularly in may affect the economic viability of extraction. petrostates. Despite the share of nonrenewable wealth in global total wealth being only 2 percent Alongside the beneficial uses of nonrenewable natural in 2020, 10 countries have more than 20 percent of capital, the mining, combustion, and transformation of these their total wealth in nonrenewable assets and another resources leads to costly pollution and effluents, which must be 11 countries have more than 10 percent. absorbed by the environment and ecosystems. This pollution has contributed to the earth transgressing several planetary ■ Metals and minerals are becoming an increasingly boundaries (Richardson et al. 2023) and is the leading cause valuable asset, increasing as a share of nonrenewable of climate change (IPCC 2023). To address these challenges wealth across all regions over the last quarter century. there are now well-developed international agreements in In 2020, global metals and mineral wealth was worth place, with countries committing to reduce their production as much as global oil wealth. New critical minerals of fossil fuels in the coming decades (Paris Agreement 2015) added to the balance sheet (cobalt, lithium, and and reverse the loss of biodiversity (Kunming-Montreal molybdenum), increased in per capita terms by Agreement 2022). The decarbonization of the global economy 31 percent. means that society is likely to shift from using fossil fuels to renewable energy and new technologies, and at a scale that requires a high volume of primary metals and previously less-exploited critical minerals. 147 Conrad (2010) defines a relevant economic time scale as one over which planning, and management decision are meaningful. Subsoil assets like oil and gas are formed over millions of years. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 96 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH This presents both risks and opportunities. Nations to decarbonize, relying on the continued rents from fossil rich in nonrenewable resources have historically had fuels is unlikely to be a sustainable development pathway. mixed success turning these endowments into economic The possibility that assets in the fossil fuel industry will prosperity (Van Der Ploeg 2011; Frankel 2012; Ross 2015). become stranded148 (Van der Ploeg and Rezai 2020) suggests Sustainable development requires that resource rents that countries should diversify their asset portfolios with from nonrenewable assets be reinvested into other assets greater urgency (Peszko et al. 2020). (Hamilton and Hartwick 2014; Hartwick 1977, 1990). But, although the taxation of these resources contributes to Therefore, whether a policy maker wants to assess the fiscal the economy, many resource-rich countries have failed to sustainability of a fossil-fuel-rich state or is deliberating develop a resource tax base due to under-exploration or an the viability of extracting untapped oil reserves, each of inability to raise proportionate tax revenues from exploited these contexts necessitates an analysis and interest in the subsoil assets (Readhead et al. 2023). As the world seeks changing level and composition of nonrenewable wealth. FIGURE 5.1 Global comprehensive wealth, nonrenewable natural capital, and GDP in per capita terms, 1995–2020 160 140 Index 1995 = 100 120 100 80 1995 2000 2005 2010 2015 2020  Total wealth  Nonrenewables GDP Source: World Bank staff estimates. Note: Wealth is measured in chained 2019 US dollars. Assets that end up as a liability before the end of their anticipated economic lifetime. https://www.lse.ac.uk/granthaminstitute/explainers/what-are- 148  stranded-assets. 97 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Unfortunately, the most prominent metric used to analyze MEASURING NONRENEWABLE economic progress, GDP, does not account for the economic NATURAL CAPITAL damages from exploiting nonrenewable resources or the sustainable reinvestment of economic rents (Hoekstra 2019). The nonrenewable natural capital estimates produced for Comprehensive wealth, and the nonrenewable natural capital this edition of CWON build on a nearly 20-year effort by the component, go some way to addressing these measurement World Bank to develop globally consistent wealth accounts shortcomings (see chapter 1). This is clearly illustrated for nonrenewable natural resources (World Bank 2005, 2011, in Figure 5.1, which shows that global GDP per capita 2021; Lange et al. 2018). The first estimates of subsoil assets increased by over 40 percent in the last quarter century, yet were produced in the World Bank’s Where Is the Wealth of nonrenewable natural capital has barely changed, and the Nations publication, which used an NPV approach to capitalize overall comprehensive wealth per capita has increased by less an assumed flow of future resource rents for nonrenewable than half of global GDP per capita. While the size of the global resources. Each report since has introduced incremental economy has increased, the stock of nonrenewable natural improvements to the accuracy of wealth estimates by capital has slightly decreased. Growth in income, therefore, gathering better-quality and more extensive data on resource has been in part accompanied by a reduction in nonrenewable production and reserves, and improving the valuation natural capital wealth, even if the estimates of nonrenewable approach.149 All previous reports have attempted to align with natural capital presented here are conservative upper the latest globally agreed statistical standards provided by the bounds (Atkinson and Venmans 2024; McGlade and Ekins United Nations SEEA-CF, which outlines how to appropriately 2015). A scenario-based analysis of nonrenewable natural measure resource depletion and asset values. capital wealth exploring how likely price, extraction, and use In the previous edition of CWON, Managing Assets for the changes affect wealth due to decarbonization and net-zero Future, measures of nonrenewable wealth were decomposed commitments would be a worthwhile exercise, but is beyond to analyze the underlying changes in wealth (World Bank the scope of this report. 2021). The report and a technical background paper found that The rest of this chapter first briefly recaps how nonrenewable changes in prices (applying GDP deflators) and production natural capital wealth is measured in CWON. It explains in the current period were the main drivers of changes in the intuition of the volume-based index for nonrenewable nonrenewable natural capital wealth (Hoekstra 2021). Volatile resources and outlines the data sources at a high level. commodity prices can provide counterintuitive signals for Next, the chapter presents the global trends for fossil fuels, long-term sustainability—where prices increase despite a metal, and mineral wealth. It looks at how the value of resource being overexploited, causing an increase in capital each nonrenewable resource is changing over time, and at gains in the sustainability metric. Following a methodological distributions by region and income groups. As nonrenewable review, this edition of CWON makes several significant natural capital wealth is highly concentrated, the analysis methodological improvements: then focuses on the 10 countries that hold more than half ■ Changes in nonrenewable wealth are estimated using of the world’s share in nonrenewables. A further section a volume-based index, where changes in the physical explores metals and minerals, including the wealth in new quantities or “volumes” of each asset are weighted critical minerals added to the data in this report. The final according to its economic value (as measured by their section concludes by discussing the next frontier for global share in nominal wealth). Changes in nonrenewable nonrenewable wealth accounts. natural capital wealth over time are now primarily An example of this was the improvement introduced in Lange et al. (2018), which moved from assuming a fixed lifetime of nonrenewable resources of, say, 149  25 years, to the direct estimation of the remaining lifetime of resources based on proven reserves and resource extraction levels. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 98 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH driven by physical changes in proven reserves of each CALCULATING THE asset (for example, economically recoverable barrels NONRENEWABLE WEALTH of oil, kilojoules of naturas gas, tons of coal, and so on). OF NATIONS The economic value and market price of the resource still affects nonrenewable natural capital wealth in Estimating the nonrenewable wealth of nations is a challenging so far as it affects the relative price of assets, and the task. Unlike financial assets, such as stocks or bonds, where an weighting placed on each asset in the wealth series. This asset class may be homogenous and feasibly sold in its entirety advancement removes potentially misleading signals on a specific auction date, nonrenewable resources are from volatile prices. Removing what are likely to be heterogenous153 and extracted and sold gradually over time. short-term capital gains aligns the methodology more As described in chapter 2, there are two important aspects to closely with the economics of sustainability (Atkinson estimating the real nonrenewable natural capital: the physical and Venmans 2024). volume and the economic value of the asset. With data on volumes and value, a volume-based index can be estimated, ■ The resource rent estimates were improved by (i) which changes over time according to the change in the directly measuring the “user costs of capital” in the physical volume of each nonrenewable asset and is weighted extraction of nonrenewable assets (that is, costs from according to each asset’s economic value. the use of machinery and physical premises); (ii) using For nonrenewable assets, the volumes of each asset are the country-level rents rather than regional average rents proven reserves in units of barrels for oil, kilojoules for natural for all assets where data are available; and (iii) more gas, and tons for coal and metals and minerals. To grasp the precisely estimating operating costs in the mining sector. intuition of the index, suppose only changes in the physical ■ Three new critical metals have been added to the volumes of assets are considered without any economic nonrenewable natural capital account: cobalt, lithium, weighting. This may be an interesting exercise for a geological and molybdenum. These metals have a range of agency interested in the pure scarcity of a given nonrenewable increasingly important uses, from lithium-ion batteries resource, but it is not of much use for a sustainability assessment used in electric vehicles and mobile phones,150 to if the asset has no economic value. Therefore, each asset in the cobalt used in rechargeable battery electrodes,151 to volume index is weighted by its economic value. It is possible molybdenum used as an important additive in steel that once a nonrenewable geological resource is mined, it may manufacturing. 152 be stored as an economic or financial asset in another form. Gold, for example, is often held in reserve by governments These methodological changes lead to marked differences with and investors, or in the form of jewelry, which may hold its past results. However, the new estimates should provide more value. To some extent, extracted geological and mineral intuitive and interpretable results, while continuing to align assets, depending on their uses, will be part of the produced with the SEEA standards and economic theory and making the capital account, such as minerals like copper (electric wiring), nonrenewable asset account coverage more complete. iron (steel structures), and lithium (computer batteries). 150 See the British Geological Survey (2016) profile on lithium. 151 See the cobalt factsheet from the Cobalt Institute (2023). 152  See the Molybdenum Commodity Summary 2023 from the United States Geological Survey (2023). 153 For example, coal can be broadly classified into four different types or ranks anthracite, bituminous, subbituminous, and lignite. In this report, the first three types are classified into a hard coal category, and lignite into a brown coal category. 99 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 5.1 ESTIMATING THE USER COSTS OF CAPITAL FOR NONRENEWABLE NATURAL CAPITAL R esource rents (or royalty rents) are a commonly used concept in resource economics. Economic theory on the optimal extraction of nonrenewable natural resources standardly expresses rents as the price of a resource less the marginal cost of extraction. In CWON, and economic-environmental accounting, rents are used to estimate the value of the stock of the resource in situ using an NPV calculation. As explained in chapter 2, the resource rent is calculated as the difference between revenues from extraction, the cost of extraction from intermediate inputs, labor, and the user cost of produced capital used in the extraction process. The user cost is the sum of both the “normal” returns to fixed assets (for example, the economic returns to machinery used in the extraction process) and depreciation. The opportunity cost of foregone rental returns from using the machinery and the depreciation of this machinery should be deducted from the economic rent (or profits) of extracting the resource. The estimation of user costs for a global database, however, is not an easy task. In the absence of estimates of capital stocks for the fossil fuel and mineral industries in a country, capital stocks need to be estimated from capital expenditure. Then a reasonable estimate for the return on capital and depreciation is needed to estimate the user costs from these estimates of capital stock. For oil and gas, country-level capital expenditure data are available from the Rystad database. For coal and metals and minerals, capital expenditure data are available at the mine level from Wood Mackenzie and S&P, respectively. Following the OECD’s Measuring Capital—OECD Manual (2009), the capital expenditure series was converted into capital stock estimates for the longest historical series available in the data. This was done using estimates for an industry-specific rate of return for capital and depreciation to estimate an initial capital stock for time 0, and then adding capital expenditure year-on-year to complete the series. These completed estimates of capital stocks for each asset can then be multiplied by rates of return plus depreciation rates to get the user costs. The estimation of user costs affects CWON’s results by changing the economic value of nonrenewable assets and thus the weights placed on the volume changes in the wealth series. Including user costs in the rent calculation would increase rents when the user costs are lower than capital expenditure but would decrease the rents when user costs are greater than capital expenditure. Globally, the inclusion of user costs increased rents for oil, gas, and metals and minerals, but reduced rents for coal. This effect, however, varies across regions and may reflect differing levels of capital expenditure and industrial investment. As this is a relatively new addition to CWON, further analysis and data collection are needed to fully assess and improve the estimation of user costs. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 100 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH However, there are likely to be some gaps. Where extracted database and S&P’s Global Market Intelligence database, minerals are used as pure stores of value, they could be which has data on realized prices at the mine level. All cost thought of in the same way as domestically held financial data from minerals come from the S&P database across 13 assets, which CWON does not account for, on the basis that metals and minerals.155 these are pure stores of value to be transferred into other assets in the comprehensive wealth index. GLOBAL TRENDS IN Following the valuation approach recommended by the SEEA NONRENEWABLE WEALTH and used in previous editions of CWON (World Bank 2011, Globally, nonrenewable natural capital per capita declined 2018, and 2021), nonrenewable natural capital assets are by 1  percent between 1995 and 2020, coinciding with a valued using the NPV of the future returns of the resource in global commodity price boom between 2000 and 2014. situ, where future returns are a stream of discounted rents Nonrenewable natural capital wealth was volatile over the for the lifetime of the resource, defined as proven reserves reporting period, and partly reflects the depletion of specific divided by current-year extraction. Within this calculation, assets, but also global population growth, which increased rents from the current year are estimated and held constant for all future years within the NPV formula. Rents are by 36 percent over the reporting period (Figure 5.2). While calculated as revenues less operating costs and the user costs oil wealth per capita flatlined, gas, coal, and minerals all of produced assets used in extraction. declined in per capita terms. In absolute terms, natural capital increased by 35 percent between 1995 and 2020, driven Revenues are calculated as market prices multiplied by mainly by increases in oil (37 percent), gas (34 percent), and the quantity of production of a resource. For each asset, metals and minerals (23 percent). Not all nonrenewable estimates of rents are averages for that resource across all the assets increased in absolute terms. Coal wealth has declined extraction or mining projects in a country. Oil and natural gas by 8 percent globally, a trend that can be observed across all use data from Rystad Energy’s UCube database, which takes regions and income groups. the nearest geographical price benchmarks, such as Brent or West Texas Intermediate, and a combination of cost data Nonrenewable wealth can be volatile, as shown in Figure from company statements, interviews, and modelling. For 5.2, likely reflecting years of significant discoveries coal, Wood Mackenzie’s Global Economic Model (GEM) data or exploration of economically recoverable resource are used. This is a bottom-up, mine-level database of prices reserves. Of the resources, global trends in coal are the and costs, with recent data covering 15 countries and up to smoothest, reflecting a stagnant trend in global proven coal 269 mines. 154 For metals and minerals, price data come from reserves, reductions in exploration, and methodological the World Bank’s Global Economic Monitor Commodities limitations in the data collection process for coal reserves.156 154 This is a small improvement in coverage compared to World Bank (2021). However, as country coverage for coal costs and thus rental rates is limited, the coal accounts rely on a large share of regionally averaging. Future updates to the report may seek to widen the data coverage specifically for coal rents and cost shares. 155  For assets and data sources without complete global coverage such as coal (Wood Mackenzie) and minerals (S&P), a range of gap-filling approaches are used to complete series. If coal cost data are missing for a country, a regional or global average is used. The cost data for three minerals (bauxite, phosphate, and tin) are estimated using an index of costs from seven other minerals as cost data for those minerals were not found in S&P. 156  Obtaining data on global coal reserves is challenging and the main source used in this report, the US Energy Information Administration (EIA), only reports data on coal reserves back to 2005, so the remainder of the series is from alternative sources and gap-filling approaches. Trends in coal reserves assessed from other sources, such as the BP Statistical Review of World Energy, show relatively flat lines for global coal reserves, so the smooth decline observed in the above data may reflect the interacting effects of population growth and stagnant coal reserves. The team is seeking out an improved method to estimate coal reserves. 101 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Changes in the proven reserves of a given resource will be the trends in oil and gas most closely. In contrast, while coal reflected in the overall nonrenewable wealth series, to the has been declining both in absolute and per capita terms, extent that the resource is of significant economic value. As the resource has an average share of 10 percent—though this shown in Figure 5.3, oil and gas account for a high of 87 percent fluctuates from a high of 20 percent in 2010 to a low of 5 percent of nonrenewable wealth in 2000 and a low of 53 percent in in 2000—so declines in coal wealth had a smaller effect on the 2020. Therefore, nonrenewable natural capital trends track overall trend in global nonrenewable wealth. FIGURE 5.2 Change in global nonrenewable natural capital per capita, by asset, 1995–2020 120 100 Index 1995 = 100 80 60 40 1995 2000 2005 2010 2015 2020  Oils  Minerals  Natural gas  Coal Nonrenewables Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. There are large data gaps for coal reserves at the country level and stagnant trends globally, which are causing a smooth decline in coal reserves per capita. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 102 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.3 Shares of nonrenewable wealth, global, 1995–2020 100 9 14 16 21 5 21 80 13 8 40 8 22 20 16 20 20 60 7 11 13 40 65 58 55 51 48 20 40 0 1995 2000 2005 2010 2015 2020  Oil  Gas  Coal  Minerals Source: World Bank staff estimates. Note: Shares of assets are calculated as shares of nominal wealth rather than the volume-based estimates. Shares represent the weights each asset is given in the volume index. Table 5.1 shows the breakdown of changes in nonrenewable had significant increases in oil and gas wealth per capita. wealth per capita by region and income group. Over half Although the Middle East and North Africa had very high of the regions in the world had a decline in nonrenewable growth in absolute terms, nonrenewable wealth declined in wealth per capita between 1995 and 2020. The largest growth per capita terms by 12 percent over the reporting period. At was in East Asia and the Pacific (59 percent), particularly due the same time, several regions struggled to maintain their to increases in mineral wealth (89 percent). This growth is nonrenewable natural capital wealth. Sub-Saharan Africa, largely driven by China, where mineral wealth increased for example, had declines across every nonrenewable asset from $78 per capita to $238 per capita, with discoveries of category (see Figure 5.4), with the exception of Botswana and particularly large reserves of copper, molybdenum, and zinc. Mozambique, which discovered large deposits of natural gas The other region with large gains was North America, which in recent years. 103 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 5.1 Percent change in nonrenewable natural capital per capita, by asset, region, and income group, 1995–2020 NONRENEWABLES OIL GAS COAL MINERALS POPULATION Region East Asia and the Pacific 59 11 37 10 89 22 Europe and Central Asia 4 8 -13 -18 -32 6 Latin America and the Caribbean -21 -63 -54 -28 38 36 Middle East and North Africa -12 -16 13 274 -30 67 North America 38 108 127 -19 -58 25 South Asia -17 -44 -29 9 -33 47 Sub-Saharan Africa -61 -6 -17 -90 -77 95 Income group High income 30 21 20 -3 34 16 Upper middle income 3.4 7.4 0.5 -51 4 24 Lower middle income -17 0.8 8 -4 -56 52 Low income -56 -43 1,093 166 -63 104 World -0.7 1.1 -1.7 -32 -9.9 36 Source: World Bank staff estimates. Note: All figures are the percent change to the nearest whole percentage point calculated from changes in real asset values per capita. Across income groups, there has been a notable global capital wealth. This is an unfortunate story, particularly divergence between 1995 and 2020 in nonrenewable natural from the perspective of improving living standards in low- capital per capita between high-income and low-income income countries and reducing levels of global poverty. countries. While nonrenewable wealth per capita rose by Moreover, this aligns with recent evidence showing a global 30 percent in high-income countries, it fell by 56 percent divergence in income levels across rich and poor countries in low-income countries, which represents a widening of since 2015, following a “golden decade” of convergence international inequities in global nonrenewable natural between 2004 and 2014 (Cust et al. 2024). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 104 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.4 Change in nonrenewable natural capital wealth per capita, 1995–2020 Source: World Bank staff estimates. Note: Wealth per capita in chained 2019 US dollars. Blue represents increasing wealth per capita, red decreasing. COUNTRIES WITH HIGH The world’s nonrenewable natural capital is concentrated NONRENEWABLE NATURAL in specific countries, some of which are resource rich and CAPITAL WEALTH others which are not. In 2020, more than half of the world’s nonrenewable natural capital wealth was concentrated in At a country level, it is common to analyze the extent to which 10 countries (Australia, Brazil, China, India, Iran, Iraq, countries are rich in resources, which implies their economic Qatar, the Russian Federation, the United Arab Emirates, and prospects are likely to be dictated by how well they manage the United States), half of which are resource rich.157 Brazil natural resource endowments. Resource-rich economies experienced the highest nonrenewable wealth per capita are most often defined as countries with a high share of growth at 98 percent, followed by Australia at 92 percent, the nonrenewable resource exports as a share of total export United States at 33 percent, and China at 32 percent. receipts or as a share of total tax revenues (IMF 2012). Recent analysis has found that the number of resource-rich countries The drivers of change varied significantly. In China, Brazil, is growing in certain regions in the world. For example, Cust and Australia, metals and minerals were the primary driver and Zeufack (2023) found that the number of resource-rich of increasing wealth, reflecting increases in proven reserves countries in Sub-Saharan Africa rose from 18 to 26 between across a range of minerals. Fossil fuel reserve increases, on 2004 and 2014. the other hand, were the main drivers of change in large petrostates such as Qatar and Iran, although in Qatar’s case this did not lead to per capita increases. 157  Resource-rich countries within the top 10 nonrenewable natural capital list include Iran, Iraq, Qatar, Russia, and the United Arab Emirates. 105 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.5 Change in fossil fuel and mineral wealth, top 10 countries, 1995–2020 Panel a: Fossil fuels per capita Panel b: Metals and minerals per capita 50 200 150 0 % Change 100 % Change 50 -50 0 -100 Qatar -50 Brazil China Australia United States Russian Federation India Iraq Iran, Islamis Rep United Arab Emirates Brazil China Australia India United States Russian Federation Iran, Islamis Rep Panel c: Fossil fuels Panel d: Metals and minerals 300 300 200 200 % Change % Change 100 100 0 0 -100 Qatar Australia Brazil Iraq India China United States Russian Federation United Arab Emirates Iran, Islamis Rep Brazil China Australia India United States Russian Federation Iran, Islamis Rep Source: World Bank staff estimates. Note: Percentage changes in chained 2019 US dollars. Selected countries are the 10 countries with the highest current US dollar wealth in nonrenewables in 2020. THE VALUE OF NONRENEWABLE weight, adjusted over time according to relative value, given NATURAL CAPITAL IN TOTAL to nonrenewables as a whole within the comprehensive WEALTH wealth index. Second, it provides the weight given to each nonrenewable asset within the nonrenewable natural capital The value of nonrenewables matters to the economics of series. More broadly, Gil et al. (2014) argue that an important sustainability for at least two reasons. First, it provides the development strategy is to diversify a country’s asset base, REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 106 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.6 Shares of global wealth, selected regions, 1995–2020 Panel a: Middle East and North Africa Panel b: Sub-Saharan Africa 100 100 4 5 5 9 15 14 17 20 29 26 39 37 80 80 27 30 11 29 33 9 7 27 29 8 60 7 6 60 51 44 52 53 39 40 47 40 50 50 41 45 40 51 20 20 25 21 19 23 16 13 12 14 15 15 11 9 0 0 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020  Produced capital  Human capital  Produced capital  Human capital  Renewable natural capital  Nonrenewable natural capital  Renewable natural capital  Nonrenewable natural capital Source: World Bank staff estimates. Note: Shares are calculated as a share of total wealth in current US dollars. including by reinvesting rents from natural resources into Despite the small, albeit growing, economic value of human capital and institutions, which may suggest a declining nonrenewable natural capital globally, it would be a mistake share of nonrenewable natural capital over time, or may to conclude nonrenewables do not matter within the reflect diversifying economies and development success. comprehensive wealth index at a country level, particularly as there are large differences across regions and countries. In the As a share of global wealth, nonrenewable natural capital Middle East and North Africa, for example, nonrenewables is small and therefore has a small weight in the global ranged from 17 percent to 39 percent of wealth between 1995 comprehensive wealth index. Nonrenewable natural capital and 2020, and in Sub-Saharan Africa from 4 percent to 15 has increased in nominal value relative to other capital assets, percent. This is shown in Figure 5.6, where the largest share and therefore the share in global wealth increased from 1 for nonrenewables (assessed at five-yearly intervals) is in the percent of global wealth in 1995 to 2 percent in 2020. The large period 2005 to 2010 for both regions, reflecting the commodity but volatile increase in nonrenewable natural capital over this price boom at this time. period reflects a global commodity price boom between 2004 and 2014, which increased the economic value of both fossil The economic value of each nonrenewable resource will fuels and minerals in nominal value terms (World Bank 2021, determine the extent to which the increase or decrease chapter 11).158 158  ote, these increases in value are different to changes in the volume-based index of wealth. Changes in value affect the weighting given to changes N in the volumes of each asset in the wealth indexes. 107 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH in physical reserves of, say oil, changes the aggregate also Papua New Guinea and the Philippines. Latin America nonrenewable wealth index. The shares of different and the Caribbean is mineral rich across the continent, with nonrenewables are highly variable across space and time. notable countries including Argentina, Brazil, and Chile. Like differences in the proven reserves, this is a factor of Across every region, the share of minerals increases over geological diversity, preferences, demand, and environmental time between 1995 and 2020, which reflects the increasing policy. Figure 5.7 shows, for example, that metals and economic demand for minerals. minerals represent a large share of nonrenewable wealth in East Asia and the Pacific (Panel a) and Latin America and Europe and Central Asia (Panel b), the Middle East and the Caribbean (Panel c). East Asia and the Pacific has some North Africa (Panel d), North America (Panel e), and Sub- notable metal-rich nations, such as Australia and China, but Saharan Africa (Panel g) are relatively richer in fossil fuels. FIGURE 5.7 Shares of nonrenewable wealth, by region, 1995–2020 Panel a: East Asia and the Pacific Panel b: Europe and Central Asia 100 100 12 15 13 16 17 24 25 32 2 31 3 80 45 44 80 12 10 4 5 68 38 4 37 24 60 31 60 25 12 32 8 51 8 44 40 17 21 40 12 9 48 49 52 53 44 20 39 8 20 31 31 27 16 28 15 7 0 0 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020 Panel c: Latin America and the Caribbean Panel d: Middle East and North Africa 100 100 1 1 2 2 3 7 8 11 10 19 17 22 80 0 43 80 47 7 59 60 82 60 1 0 4 4 97 93 92 88 88 40 74 0 40 81 8 74 52 49 20 20 3 32 1 14 0 3 0 0 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 108 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Panel e: North America Panel f: South Asia 100 100 5 9 10 9 16 16 20 21 14 23 27 18 33 80 80 42 15 36 36 16 32 25 0 60 60 19 28 48 35 0 48 9 40 14 22 40 40 15 45 64 56 50 8 7 45 20 38 20 39 30 31 32 22 4 20 10 0 0 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020 Panel g: Sub-Saharan Africa 100 7 5 10 16 14 5 10 2 9 3 80 9 13 9 1 52 4 60 86 82 10 40 74 72 69 9 20 29 0 1995 2000 2005 2010 2015 2020  Oil  Gas  Coal  Minerals Source: World Bank staff estimates. Note: Shares are calculated as a share of total wealth in current US dollars. Europe and Central Asia have the largest share of gas wealth GEOGRAPHIC AND INCOME GROUP through time, with notable reserves in Russia. North America DISTRIBUTION OF NONRENEWABLE also has gas wealth, although this falls to zero as economic NATURAL CAPITAL rents in both Canada and the United States are estimated to have significantly fallen since 2015. These regional variations Assessing the global distribution of nonrenewable natural in value can lead to different distributions of changes in wealth capital can help us understand which groups of countries across regions and countries. These global distributions in and regions are becoming richer in nonrenewable wealth wealth are discussed in the next section. and which are getting poorer. In all income groups, fossil fuel wealth is converging across countries in absolute terms but not in per capita terms (panels a and b in Figure 5.8, respectively). 109 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.8 Change in fossil fuel and mineral wealth, by income group Panel a: Change in fossil fuel wealth Panel b: Change in fossil fuel wealth per capita 20 80 60 10 % Change % Change 40 0 20 -10 0 Lower middle income Upper middle income Low income High income Upper middle income Low income Lower middle income High income Panel c: Change in mineral wealth Panel d: Change in mineral wealth per capita 60 40 40 20 20 0 % Change % Change 0 -20 -20 -40 -40 -60 High income Low income Upper middle income Low income Lower middle income Lower middle income High income Upper middle income Mineral wealth, on other hand, is diverging, both in terms of Pacific and Latin American and the Caribbean, reflecting large absolute levels of wealth and in per capita wealth across income increases in proven reserves of metals and minerals. Mineral groups (panels c and d in Figure 5.8, respectively), with higher- wealth declined in both per capita and absolute terms in Sub- income countries holding a growing share of mineral wealth Saharan Africa. between 1995 and 2020. Fossil fuel wealth per capita decreased overall in all regions of the world, other than in North America The country-level distribution, as shown in Figure 5.10, clearly and in East Asia and the Pacific due to sizable new discoveries indicates where hot spots exist for changes in fossil fuel and and exploitation of proven reserves (see panels a and b in Figure mineral wealth per capita. Panel a shows hot spots for fossil 5.9). However, as discussed previously, this does not mean fuel wealth per capita in Canada, Kazakhstan, Azerbaijan, and absolute wealth did not grow in other regions of the world, it just China. For metals and minerals, there have been large declines failed to grow at the same pace as population growth. Mineral in wealth per capita across most of Sub-Saharan Africa and wealth per capita has been increasing across East Asia and the North America and Europe. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 110 111 % Change % Change -50 0 50 100 150 -40 -20 0 20 40 60 80 East Asia & Pacific FIGURE 5.9 North America Latin America and Caribbean Middle East & North Africa Middle East & North Africa East Asia & Pacific South Asia South Asia Panel c: Change in mineral wealth Panel a: Change in fossil fuel wealth Europe and Central Asia Europe and Central Asia North America Sub-Saharan Africa MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Sub-Saharan Africa Latin America and Caribbean REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH Change in fossil fuel and mineral wealth, by region group % Change % Change -60 -40 -20 0 20 40 -100 -50 0 50 100 East Asia & Pacific North America Latin America and Caribbean East Asia & Pacific Middle East & North Africa Europe and Central Asia Europe and Central Asia Middle East & North Africa South Asia South Asia North America Sub-Saharan Africa Panel d: Change in mineral wealth per capita Panel b: Change in fossil fuel wealth per capita Sub-Saharan Africa Latin America and Caribbean MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.10: Global map of the change in fossil fuel and mineral wealth per capita, 1995 and 2020 Panel a: Fossil fuels Panel b: Metals and minerals REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 112 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH It clearly shows how the emerging hot spots for mineral wealth concentration of about 2 percent. In Chile, exploration starts per capita are in Latin America, East Asia and the Pacific, parts at 0.2 percent and production at around 0.5 percent. In the of Europe and Central Asia (notably Russia), and Australasia. Democratic Republic of Congo, copper is below 1  percent In sum, the global geographic and income distribution of concentration and is therefore left in the ground, leaving nonrenewable natural capital shows the changing regional foregone minerals for the energy transformation and dynamics of this type of wealth. Fossil fuel wealth per capita revenues for the development of the economy. is increasing in many countries across different regions of the world, while mineral wealth per capita is increasing in more In this report, data were collected on 13 minerals, including concentrated hot spots. Both mineral, and to a lesser extent bauxite, cobalt, copper, gold, iron ore, lead, lithium, fossil fuel, wealth per capita is becoming more concentrated molybdenum, nickel, phosphate, silver, tin, and zinc. Globally, the wealth per capita of six minerals has increased, while in the already richer regions of the world. six have been declining.159 The minerals that increased the most in per capita terms were lithium, nickel, phosphate, and SHIFTING COMPOSITION OF copper, while gold, silver, bauxite, and lead had the biggest NONRENEWABLE NATURAL declines. One of the main characteristics of mineral wealth, CAPITAL WEALTH: METALS AND like nonrenewables more broadly, is the year-to-year volatility. MINERALS There are some notable increases and decreases in wealth per capita, particularly for phosphate, molybdenum, and nickel, Part of the changing global distribution in nonrenewable and these are likely to reflect either changes in the quantity of assets may be due to the shifting composition of the proven resources or the accuracy of geological reporting. nonrenewable wealth of nations, including the changing demands for specific primary metals and critical minerals. The five richest countries in minerals as of 2020 were Australia, As the world transitions to a low-carbon economy and the Brazil, China, India, and Russia. Most of these countries had share of energy use shifts from fossil fuels toward renewable general increases across their mineral wealth per capita, with energy, similar shifts might be observed in nonrenewable India being the exception, where mineral wealth per capita natural capital from fossil fuel wealth to mineral wealth (IEA declined across all metals and minerals (see Figure 5.12). Some 2021; World Bank 2017). Proven reserves of fossil fuels may of the largest increases were for minerals important to both decline as they become uneconomical, while increasing digital technologies and electric batteries—cobalt and lithium. demand for primary metals and critical minerals, such as These minerals had wealth per capita increases in Australia copper, lithium, and nickel, may make more geological of over 15-fold and 25-fold, respectively, reflecting new proven reserves economical and incentivize further exploration. The reserve deposits becoming economical. Lithium wealth per absence of geological surveys and broader exploitation risks capita also increased nearly 80-fold in Brazil. China, which has can be further impediments or enablers to new exploration a large share of production in many minerals globally,160 has and extraction from the resource base. For example, in the had significant mineral wealth per capita increases, with both Democratic Republic of Congo, copper exploration starts at molybdenum and phosphate increasing over 10-fold between a concentration of 1 percent and production has an average 1995 and 2020. 159 See Figure 5.11. 160 This has led to several countries’ geological agencies investigating their reliance on minerals imported from China. See, for instance, USGS (2020). 113 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.11 Global trends in mineral wealth per capita, by selected minerals, 1995–2020 nickel 200 phosphate 150 Index 1995 = 100 copper molybdenum cobalt 100 iron zinc lead bauxite silver gold 50 1995 2000 2005 2010 2015 2020 Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. Lithium increased over sevenfold and so could not be shown on this scale. Tin wealth could not be estimated using the volume index due to estimates finding persistently zero resource rents using best available data. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 114 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 5.12 Percent changes in mineral wealth, selected countries, by mineral, 1995–2020 Panel a: Australia Panel b: Brazil 8,000 2,500 2,000 6,000 1,500 % Change % Change 4,000 1,000 2,000 500 0 0 Bauxite Silver Zinc Gold Nickel Cobalt Iron Copper Lead Lithium Silver Copper Bauxite Iron Zinc Lead Gold Lithium Nickel Panel c: China Panel d: Russian Federation 1,500 250 200 1,000 % Change 150 % Change 100 500 50 0 0 -50 Bauxite Silver Zinc Molybdenum Gold Nickel Phosphate Cobalt Iron Copper Lead Lithium Bauxite Silver Zinc Molybdenum Gold Nickel Phosphate Cobalt Iron Copper Lead Panel e: India 20 0 -20 % Change -40 -60 -80 Bauxite Silver Zinc Gold Phosphate Iron Copper Lead Source: World Bank staff estimates. Note: Wealth is measured in chained 2019 US dollars. 115 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH NEW CRITICAL MINERALS ADDED cobalt, lithium, and molybdenum. While of high and IN CWON’S WEALTH ACCOUNTS— increasing economic value globally, these minerals are very COBALT, LITHIUM, AND geographically concentrated. Around 50 percent of the world’s MOLYBDENUM proven reserves of cobalt are in the Democratic Republic of Congo, 60 percent of the world’s molybdenum proven reserves As economic production becomes more complex and are in China, and around 10 countries have notable proven technologically advanced, the economic use case for minerals reserves for lithium,161 with the largest shares in Chile (46 expands. To estimate changes in the mineral wealth of nations, percent) and Australia (29 percent). Since 2001, wealth per it is important to add minerals of notable economic value, capita of cobalt has been declining globally, signaling that where data are available to produce an account. In this version extraction has increased faster than the discovery or increased of CWON, the wealth for three new minerals was estimated: viability of proven cobalt reserves (Panel a, Figure 5.13). FIGURE 5.13 Changes in global wealth per capita for selected minerals: cobalt, lithium, molybdenum Panel a: Cobalt Panel b: Lithium 1.5 .002 Chained 2019 USD Chained 2019 USD 1 .001 .5 0 0 1995 1996 1997 1998 1999 2000 2001 2003 2006 2016 2002 2004 2005 2007 2008 2009 2010 2011 2012 2013 2014 2015 2017 2018 2019 2020 1995 1996 1997 1998 1999 2000 2001 2003 2006 2016 2002 2004 2005 2007 2008 2009 2010 2011 2012 2013 2014 2015 2017 2018 2019 2020 Panel c: Molybdenum 25 20 Chained 2019 USD 15 10 5 0 1995 1996 1997 1998 1999 2000 2001 2003 2006 2016 2002 2004 2005 2007 2008 2009 2010 2011 2012 2013 2014 2015 2017 2018 2019 2020 Source: World Bank staff estimates.  ountries with available data on proven reserves of lithium were Argentina, Australia, Brazil, Chile, the Democratic Republic of Congo, Portugal, 161 C the United States, and Zimbabwe. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 116 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH In contrast, lithium wealth per capita has been increasing, volume-based index, which, particularly for nonrenewable from about $0.2 per capita to $1.2 per capita globally (Panel b, assets, removes most of the effects of commodity price Figure 5.13). Molybdenum wealth per capita has been more volatility that may distort sustainability assessments. On volatile, but peaked at over $20 per capita in 2019 and stood the other hand, the new volume-based approach makes the at about $16 per capita in 2020 (Panel c, Figure 5.13). As the series sensitive to data on the proven reserves of resources. demand for these critical minerals (particularly cobalt and This is an important area for future work to improve the lithium) increases during the clean energy transition (IEA coverage and accuracy of proven reserves data. 2021), it will be important to assess whether these trends continue or change, and what the implications may be for the In addition, while this edition added important minerals to economic sustainability of countries rich in these resources. the accounts, there are many more critical minerals and rare earth minerals that are becoming increasingly valuable for their uses in information and communication technologies, CONCLUSIONS transportation, and clean energy production. Although these minerals may also be highly concentrated in specific The new volume-based measure of nonrenewable natural countries (notably rare earth minerals in China), a future capital per capita has shown that this type of wealth is nonrenewable natural capital account may want to add these declining globally. Coal, natural gas, and metals and to the balance sheet. minerals wealth per capita are in decline, while oil wealth per capita has flatlined over a quarter century. Although The data and associated accounts provided here are intended nonrenewable natural capital wealth only makes up a to provide a starting point for further scenario and policy small share of total global wealth, it plays a significant role analysis, which would go beyond the scope of this report. in countries that are rich in subsoil assets. Furthermore, With many countries around the world committing to net- metals and minerals are becoming an increasingly important zero targets and decarbonization strategies, an interesting component in the nonrenewable wealth of nations, with its avenue for future research would be to assess the extent to share of nonrenewable wealth increasing in economic value which these alter nonrenewable wealth estimates (see Box across all regions. The 10 richest countries in nonrenewable 5.2). It is possible that a share of proven reserves for fossil natural capital have almost half of the global value. Critical fuels are no longer economically viable to extract under such minerals, including lithium and cobalt, are included on the commitments. However, decarbonization strategies would CWON balance sheet for the first time, and are shown to be also affect prices, extraction, and use of both fossil fuels and increasing in value in per capita terms. minerals for clean energy, which suggests some countries may have increased nonrenewable natural capital, while This version of CWON has made several innovations, which others will have less. have improved the estimation of the nonrenewable wealth of nations. The main innovation has been the switch to a 117 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 5.2 THE LOW-CARBON TRANSITION AND NONRENEWABLE NATURAL CAPITAL A ccording to the United Nations Framework Convention on Climate Change (UNFCCC) (2023), global carbon emissions should peak by 2025 at the earliest, then reach net-zero emissions by 2050, and net- zero GHG emissions by the early 2070s to keep to the goal of a 1.5⁰C temperature increase. Individual countries’ pathways to net zero will vary, but globally it implies no further exploration of fossil fuels well ahead of 2030, a global phase-out of unabated coal power generation by 2040, and phasing out all unabated fossil fuels after 2050 (UNFCCC 2023). By the end of 2023, about 80 countries announced net-zero carbon emission pledges through various policy, regulatory, and legal instruments, with deep uncertainty over whether they will be implemented (UNEP 2023; Net Zero Tracker 2024). Should we, then, expect to see declines in nonrenewable natural capital? Furthermore, how much of the current stocks are in fact stranded assets? It is important to stress that pledges and announced targets by themselves do not cause stranded fossil fuel assets or capital gains for minerals. Only implemented and enforced specific policy instruments and market measures, such as taxes, subsidies, cartel actions, or geopolitical decisions can influence market prices and hence rents and recoverable reserves of subsoil assets. As different countries begin to implement the low-carbon transition, the volatility and uncertainty about future volumes and prices will increase even more. This transition risk (Carney 2015; McGlade and Ekins 2015; Van der Ploeg and Rezai 2019; Mercure et al. 2018) will affect fossil-fuel-producing firms and countries differently depending on their exposed endowments, ability to weather external shocks, and own decarbonization strategies (Peszko et al. 2020). While reduced consumption of fossil fuels would trigger a decline in the global fossil fuel industry, some more resilient producers will experience short- and medium-term increases of their reserves, while others may see an accelerated decline of their subsoil assets (Peszko et al. 2020). The low-carbon transition would also increase the value of climate action minerals. These effects are likely to translate into different changes in nonrenewable natural capital across space and time. Economic actors aiming to be production leaders during the low-carbon transition pathway may see increases in proven reserves and nonrenewable natural capital as they continue to pursue exploration and discoveries. Whereas those divesting from fossil-fuel-dependent sectors may see their nonrenewable natural capital decline commensurately. Another channel through which decarbonization strategies may affect nonrenewable natural capital wealth is through the weights placed on the different assets in the volume-based index. These weights will increase or decrease in line with rents and the profile of production (that is, bringing production forward or pushing it backwards in time). There may be future cases where increasing weights coincide with declining stocks, which would cause sharp declines in nonrenewable natural capital, or decreasing value coincides with declining stocks, causing muted responses in nonrenewable natural capital. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 118 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH CWON is not taking any normative position on future price and rent changes but must make some assumptions. The program follows the recommendations of the SNA (2008), SEEA-CF (SEEA-CF, paragraphs 5.133 and 5.134) and SEEA Energy, which state that in the absence of any specific forecasts, future rents are held constant at current-year (2020) values wherever an NPV-RVM approach is used. Given the infinite number of possible future trajectories of rents, and the deep uncertainty about the future policy and market actions related to the low- carbon transition, this is the most neutral and transparent assumption. But it also means that if costs and policies change, the future value of specific nonrenewable natural assets will be different than estimates in this report. To explore transition risks and the upsides to both fossil fuels and minerals, alternative “what if” scenarios can be simulated with macroeconomic and energy models. Such models should be able to simulate how different real-world policy actions, such as carbon taxes, repurposing of subsidies, tariff and non-tariff trade barriers, or production cuts undertaken by different groups of countries, can influence extraction costs and producer prices of different commodities. The models should also calculate how this would change the volumes of commercially recoverable resources and changes in resource rents for different countries and commodities with respect to CWON’s current policy assumptions. An example of such an analysis of transition risk to future fossil fuel producers is provided in chapter 10 of the CWON 2021 report, conducted with a global computational model, ENVISAGE, modified to estimate transition risks for producers of fossil fuels (Peszko at al. 2021). The analysis found that low-carbon transition policies that can be implemented by fossil fuel importers represent a material risk to the value of all fossil fuel assets. In the 2018–2050 period, global fossil fuel wealth may be $4.4 trillion to $6.2 trillion (13 percent to 18 percent) lower than in the reference scenario, depending on the ambition level of global climate policies. CWON 2021 also calculated the distribution of transition risk across fuels, countries, and asset owners, showing major differences across countries depending on their initial conditions, such as the fuel type they depend on, costs of production, market power, and exposure of the rest of the economy to this risk. Level and distribution of stranded assets also differ by policy pathway—whether they are cooperative or not, and whether free riding will meet border carbon adjustment taxes or not. Where possible, CWON 2024 will make publicly available the final real and nominal wealth estimates, which can be used as inputs for such models and associated statistical code used to derive the wealth estimates. This will enable users and researchers to make their own assessment of likely future scenarios. If simulation models are not available, users of the CWON data will also be able to make much simpler back of the envelope assessments. They can, for example, modify the assumptions about future growth or decline of resource rents used to derive the wealth estimates. Access to the underlying data and codes will not just facilitate users modifying the CWON assumptions to reflect their approaches or questions, but also be valuable as a starting point for researchers wishing to compile comprehensive wealth estimates for individual countries using national, rather than global, data. 119 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH REFERENCES Mercure, J.F., Pollitt, H., Viñuales, J.E., Edwards, N.R., Holden, P.B., Chewpreecha, U., Salas, P., et al. 2018. “Macroeconomic Impact of Stranded Atkinson, G., and Venmans, F. 2024. “Wealth, Saving and Sustainability: Fossil Fuel Assets.” Nature Climate Change 8: 588–593. A Discussion of Issues of Alignment between ‘Changing Comprehensive Wealth’ and ‘Adjusted Net Saving’.” World Bank, Washington, DC. Peszko, G., Van der Mensbrugghe, D., Golub, A., and Chepeliev, M. 2021. “Low- Carbon Transition, Stranded Fossil Fuel Assets, Border Carbon Adjustments, British Geological Survey. 2016. “Mineral Profile – Lithium.” https://www.bgs. and International Cooperation.” In Changing Wealth of Nations 2021, World ac.uk/news/mineral-profile-lithium/. Bank. Carney, M. 2015. “Breaking the Tragedy of the Horizon: Climate Change Peszko, G., Van der Mensbrugghe, D., Golub, A., Ward, J., Zenghelis, D., and Financial Stability.” Speech by the Governor of the Bank of England, Marijs, C., Schopp, A., Rogers, J.A., and Midgley, A. 2020. Diversification London, September 29. and Cooperation in a Decarbonizing World: Climate Strategies for Fossil Fuel- Dependent Countries. World Bank. The Cobalt Institute. 2023. “Cobalt Factsheet.” https://www.cobaltinstitute.org/ wp-content/uploads/2023/02/cobalt_institute_fact_sheet_2023.pdf. Readhead, A., Taurus, V., Lassourd, T., Madzivanyika, E., and Schlenther, B. 2023. The Future of Resource Taxation: 10 Policy Ideas to Mobilize Mining Conrad, J.M. 2010. Resource Economics (second edition). Cambridge Revenues. International Institute for Sustainable Development and African University Press. Tax Administration Forum. Cust, J., Collier, P., and Rivera, A. 2024. “Are the Poorest Catching Up?” World Richardson, K., Steffen, W., Lucht, W., Bendtsen, J., Cornell, S.E., Donges, Bank Policy Research Working Paper, World Bank Group, Washington, DC. J.F., Drüke, M., Fetzer, I., Bala, G., Von Bloh, W., Feulner, G., Fiedler, S., Cust, J., and Zeufack, A., eds. 2023. Africa’s Resource Future: Harnessing Natural Gerten, D., Gleeson, T., Hofmann, M., Huiskamp, W., Kummu, M., Mohan, Resources for Economic Transformation during the Low-Carbon Transition. C., Nogués-Bravo, D., and Rockström, J. 2023. “Earth beyond Six of Nine Washington, DC: World Bank Publications. Planetary Boundaries.” Science Advances 9 (37). Frankel, J.A. 2012. “The Natural Resource Curse: A Survey of Diagnoses and Ross, M.L. 2015. “What Have We Learned about the Resource Curse?” Annual Some Prescriptions.” In Commodity Price Volatility and Inclusive Growth in Review of Political Science 18: 239–259. Low-Income Countries, 7–34. International Monetary Fund. UNEP (United Nations Environment Programme). 2023. Broken Record. Gill, I.S., Izvorski, I., Van Eeghen, W., and De Rosa, D. 2014. Diversified Emissions Gap Report 2023. Development: Making the Most of Natural Resources in Eurasia. Washington, UNFCCC (United Nations Framework Convention on Climate Change). 2015. DC: World Bank. Paris Agreement. Hamilton, K., and Hartwick, J. 2014. “Wealth and Sustainability.” Oxford Review UNFCCC. 2023. “Technical Dialogue of the First Global Stocktake.” Synthesis of Economic Policy 30 (1): 170–187. https://doi.org/10.1093/oxrep/gru006. report by the co-facilitators on the technical dialogue. GE.23-16519(E). Hartwick, J.M. 1977. “Intergenerational Equity and the Investing of Rents from United Nations. 2012. System of Environmental-Economic Accounting 2012 Central Exhaustible Resources.” The American Economic Review 67 (5): 972–974. Framework. United Nations. Hartwick, J.M. 1990. “Natural Resources, National Accounting and Economic United States Geological Survey. 2023. Molybdenum Commodity Summary Depreciation.” Journal of Public Economics 43 (3): 291–304. 2023. https://www.usgs.gov/centers/national-minerals-information-center/ Hoekstra, R. 2019. Replacing GDP by 2030: Towards a Common Language for the molybdenum-statistics-and-information. Well-Being and Sustainability Community. Cambridge University Press. Van der Ploeg, F. 2011. “Natural Resources: Curse or Blessing?” Journal of Hoekstra, R. 2021. “Analyzing the Driving Forces of Changes in Natural Capital Economic Literature 49 (2): 366–420. Wealth through Decomposition Analysis.” In CWON 2021 Technical Report. Van der Ploeg, F., and Rezai, A. 2019. “The Risk of Policy Tipping and Stranded https://doi.org/10.5149/northcarolina/9781469664477.003.0003. Carbon Assets.” Working Paper No. 7769, CESifo Economic Studies, Oxford IEA (International Energy Agency). 2021. The Role of Critical Minerals in Clean University, Oxford, UK. Energy Transitions. IEA. Van der Ploeg, F., and Rezai, A. 2020. “Stranded Assets in the Transition to a IMF (International Monetary Fund). 2012. “Macroeconomic Policy Frameworks Carbon-Free Economy.” Annual Review of Resource Economics 12: 281–298. for Resource-Rich Developing Countries.” Policy paper, IMF, Washington, World Bank. 2005. Where Is the Wealth of Nations? Measuring Capital for DC. https:// www.imf.org/external/np/pp/eng/2012/082412.pdf. the 21st Century. IPCC (Intergovernmental Panel on Climate Change). 2023. “Summary for World Bank. 2011. The Changing Wealth of Nations: Measuring Sustainable Policymakers: Synthesis Report.” In Climate Change 2023: Synthesis Report. Development in the New Millennium. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, 1–34. World Bank. 2017. “The Growing Role of Minerals and Metals for a Low Carbon Future.” In The Growing Role of Minerals and Metals for a Low Carbon Future Lange, G.M., Wodon, Q., and Carey, K. 2018. The Changing Wealth of Nations (June issue). Washington, DC: World Bank. 2018: Building a Sustainable Future. Washington, DC: World Bank. World Bank. 2021. The Changing Wealth of Nations 2021: Managing Assets for the McGlade, C., and Ekins, P. 2015. “The Geographical Distribution of Fossil Fuels Future. Washington, DC: World Bank. Unused when Limiting Global Warming to 2°C.” Nature 517 (7533): 187–190. https://doi.org/10.1038/nature14016. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 120 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH 6 Valuation of Hydroelectric Resources MAIN MESSAGES wealth per installed megawatt (MW) of hydroelectric generating capacity. This was far below the global ■ The estimated values of hydroelectric assets are average of $4.6 million per MW in 2020. globally consequential. In total, hydroelectric assets were valued at $3.5 trillion (chained 2019 US dollars) in ■ Estimates of renewable energy asset values beyond 2020, which is comparable to other renewable natural hydroelectric assets were not possible due to data resource assets. For reference, global fisheries were limitations. The main hurdle preventing the inclusion estimated to be worth $21.8 trillion in 2020, while the of solar and wind assets was the lack of a suitable corresponding value for global timber assets time series of producer price data for solar/wind was $2.8 trillion. producers. ■ In China, Brazil, and Canada, hydroelectric assets are also INTRODUCTION comparable in value to fossil For the first time in the World Bank’s fuel assets. In Paraguay, they comprehensive wealth accounting work, represented 47 percent of the CWON 2021 (World Bank 2021) reported country’s total natural capital experimental results for the value of renewable in 2020. Globally, hydroelectric energy assets. In that report, estimates of resources represented 23 percent of the value of solar, wind, and hydroelectric assets total energy assets that year. were presented for 15 countries as part of a pilot study to ■ Sufficient data are available to value hydroelectric demonstrate the feasibility of their valuation (Smith et al. assets in all countries using the method tested in the 2021). The valuation method used in the pilot study was based pilot study of renewable energy asset values in CWON on an approach where asset values were estimated as the 2021 (Smith et al. 2021). NPV of expected future resource rent. Resource rent, for its part, was estimated as the difference between the revenues ■ East Asia—largely driven by China—was the region realized from exploiting renewable energy resources and the with the largest hydroelectric asset wealth in 2020, costs of doing so (RVM). This approach is referred to here as more than tripling its asset value from 1995. However, NPV-RVM. It is consistent with international guidance on the when considered in per capita terms, East Asia valuation of natural resources from the United Nations SEEA- produced just one-third as much hydroelectric asset CF (United Nations 2012) and SEEA-Energy (United Nations wealth as the leading regions (North America and 2019). Latin America and the Caribbean). This chapter builds on the pilot study results to produce the ■ Being wealthy in terms of hydroelectric assets does first global estimates of hydroelectric assets for the CWON not mean that those assets are exploited to their database based on NPV-RVM. Though efforts were made to fullest potential. The United States, which ranked fifth develop global estimates of solar and wind electricity assets, in terms of total hydroelectric asset wealth in 2020, this proved impossible due to limitations in data availability.162 generated just $1.2 million (chained 2019 US dollars) of It remains a goal to present global estimates of solar and wind As explained in more detail later in the chapter, the problem resulted from the absence of price data specific to the sale of electricity generated by wind 162  and solar producers. 121 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH assets—and other renewable energy assets like geothermal et al. 2009) and SEEA-CF asset boundaries. Beginning with and ocean energy—in subsequent editions of CWON. the 2025 edition of the SNA, renewable energy assets will be recognized as natural resource assets in the system thanks, Future work should focus not only on the use of renewable in part, to the efforts of the World Bank to develop and test energy assets to generate electricity but also on direct heat methods for the valuation of renewable energy assets in the delivery. CWON 2021 pilot study (Smith and Peszko 2022).164 The SEEA- CF asset boundary will be similarly expanded at its next The remainder of this chapter proceeds as follows. First, update. This places CWON at the frontier of work in this area the rationale for valuing renewable energy assets in CWON of emerging importance. is presented. Then the methods and data sources used to value hydroelectric assets in this global study are described Of course, not all hydroelectric resources qualify as economic in considerable detail.163 Following this, the results of the assets. In keeping with the general definition of an asset and global valuation of hydroelectric assets are presented. The with the renewable energy specifications of the United Nations chapter ends with a discussion of the results and suggestions Framework Classification for Resources (UNFC; United Nations for future work. Economic Commission for Europe 2020),165 only hydroelectric resources that are viable for use in electricity production RATIONALE FOR VALUING under prevailing technological and economic conditions are HYDROELECTRIC ASSETS considered assets here (see Box 6.1 for further details). The absence of hydroelectric assets (not to mention solar, Recognizing hydroelectric resources as assets means a method wind, and other renewable energy assets) from the CWON must be found to value them. For CWON 2024, the NPV-RVM natural capital accounts in the past was a concern for approach was adopted, which was first applied to renewable assessing the environmental sustainability of economic energy assets in the CWON 2021 pilot study (Smith et al. 2021). activity. Since fossil fuel assets have always been included This method is consistent with guidance in the SEEA-CF and in CWON, there was an imbalance in its treatment of (somewhat less clearly) the SNA, both of which recommend nonrenewable and renewable energy assets. Given the it for valuing natural resource assets in general. In NPV-RVM, climate-related consequences of fossil fuel use, this asset value is taken to be equal to the present value of the imbalance risked sending distorted signals to users of the future stream of rent flowing from the resource (SEEA-CF, CWON database regarding the relative economic importance section 5.4.5). Rent, for its part, is calculated as the difference of carbon-intensive, nonrenewable energy sources versus between resource revenues (less specific subsidies received less climate-damaging renewable sources like hydropower. plus specific taxes paid) and production costs, including This imbalance is now rectified by treating renewable energy returns to labor and produced capital (see annex A4 for further assets, beginning with hydroelectric assets here in CWON discussion of resource rent). 2024, on par with fossil fuels. The validity of NPV-RVM assumes that markets for the The addition of hydroelectric assets to the CWON natural generation and sale of renewable energy approximate long- capital accounts (and the eventual addition of other run competitive equilibrium, as only in such markets will the renewable energy assets) is also aligned with recent guidance difference between revenues and costs be a reliable guide on the expansion of the 2008 SNA (European Communities to estimating resource rent. Markets in many countries—  ince this is the first time hydroelectric assets are included in the CWON global accounts, the methods used to value them are presented in greater 163 S detail than is the case for the other assets discussed in this report. This is done so that readers will have a clear idea of the approach taken and the rationale for its adoption. Subsequent editions of CWON will not present this same level of detail. Readers interested in details of the concepts and methods applied to valuation of the full suite of CWON assets are referred to the overall methodology report (World Bank 2024). 164 For further details on endorsed guidance in the SNA update process, see https://unstats.un.org/unsd/nationalaccount/towards2025.asp. 165 See https://unece.org/fileadmin/DAM/energy/se/pdfs/UNFC/publ/UNFC_ES61_Update_2019.pdf. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 122 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 6.1 DEFINITION OF HYDROELECTRIC ASSETS F ollowing guidance in the UNFC, hydroelectric assets are defined here as resources that are associated with “viable” hydroelectric generation projects. Viable projects are those for which the “environmental-socio-economic viability and technical feasibility has been confirmed” (UNFC, 3). More specifically, economic viability means that: “Development and operation [of the project] are environmentally-socially-economically viable on the basis of current conditions and realistic assumptions of future conditions. All necessary conditions have been met (including relevant permitting and contracts) or there are reasonable expectations that all necessary conditions will be met within a reasonable timeframe and there are no impediments to the delivery of the product to the user or market.” (UNFC, 6) Technical viability means that: “Development or operation is currently taking place or, sufficiently detailed studies have been completed to demonstrate the technical feasibility of development and operation. A commitment to develop should have been or will be forthcoming from all parties associated with the project, including governments.” (UNFC, 6) Based on this, only hydroelectric resources associated with currently existing and operating hydroelectric generation plants or with plants that are well advanced in planning and development qualify as hydroelectric assets. Hydroelectric resources that exist at sites where hydroelectric generation plants do not currently exist and none are well advanced in planning do not qualify as assets and are not within the scope of CWON’s natural capital accounts. This treatment of hydroelectric resources is consistent with the treatment of other natural resources in the SNA and SEEA-CF. For example, the SNA and SEEA-CF recognize timber in a forest (another renewable resource) as an asset only in instances where that timber may be commercially logged at a profit under existing technological and economic conditions. Remote forests with no potential for logging do not qualify as timber assets.166 166  Similar criteria are applied to defining other renewable and nonrenewable natural resources as assets in the SNA and SEEA-CF as well as in CWON. 123 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH especially in the developing world—do not meet this standard where hydroelectric power utilities remain publicly owned, because governments intervene in markets (through the assumption of market equilibrium may not hold. production subsidies, for example). Ricardian and scarcity rents will still arise, but they will be captured by electricity consumers rather than by the owner Data from the OECD’s Product Market Regulation of the resource (government) and their measurement is Indicators  suggest a movement toward competitive made more difficult. Most large reservoirs also serve multiple generation since the beginning of the deregulation of purposes and the revenue from electricity generation is electricity markets, at least in developed countries. often constrained by the needs of flood control, irrigation, According to World Bank analysis (Foster and Rana 2020), residential/industrial water supply, or recreation. Rents only a handful of developing countries have, however, arising from these uses of reservoirs may be captured by implemented liberalized power market models with fully various economic actors. The purpose of this study is not competitive private power generation. Across the developing to value these other uses, however, but only the use of the world, reforms have resulted in hybrid models in which reservoirs for generating hydroelectricity. elements of market orientation coexist with continued state dominance of the sector. On top of this, within the OECD, Despite the challenges posed by energy market regulation, emerging markets and developing countries alike still a strong rationale for adopting NPV-RVM is found in the subsidize renewable energy production and consumption, benefits of methodological consistency. The method is especially nascent renewable energy technologies, like widely applied to natural asset valuation in country practice, solar and wind. This poses a theoretical challenge to RVM including in CWON and by UNEP.167 Users of CWON can and a practical measurement challenge because subsidy evaluate hydroelectric asset values against other natural data are rarely comprehensive and transparent. assets and the assets of the broader comprehensive wealth portfolio. Thus, consistency in valuation across assets is of For mature renewable energy technologies such as considerable importance. hydroelectricity, however, factor markets can be assumed to be close to long-run competitive equilibrium, even in countries where wholesale electricity markets have not METHODS AND DATA SOURCES FOR been fully liberalized. Marked heterogeneity and scarcity VALUING HYDROELECTRIC ASSETS among sites for hydropower implies that hydro projects should earn both Ricardian and scarcity rents (see the annex Estimates of hydroelectric asset values for all countries were to this chapter for rent definitions). Where equilibrium can prepared using NPV-RVM, where either (i) hydroelectric reasonably be assumed, quasi-rents should not exist. In generation accounted for more than 5 percent of total countries where electricity prices remain regulated and national generation in 2020, or (ii) total installed hydroelectric 167 S ee UNEP’s Inclusive Wealth series of reports (UNU–IHDP and UNEP 2012; UNU–IHDP and UNEP 2014; Managi and Kumar 2018; UNEP 2023). These restrictions were introduced to ensure that countries included in the study would have markets for hydroelectric generation sufficiently well 168  established for the global data used in this study to reflect conditions in the country. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 124 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Estimating hydroelectric resource rent It must be stressed that is the rent attributable and asset value to the hydroelectric resource itself as a form of natural capital; that is, it is the return to nature in the generation HYDROELECTRIC RESOURCE RENT process as opposed to the return to the produced capital Equation 1 expresses the version of the RVM used to estimate used to generate the electricity (such as dams, turbines, or rent for hydroelectric assets in a given country and year . generators). In the RVM formula, the rent attributable to the hydroelectric resource is the residual left over after full 1) costs of converting the flowing water to useful electricity— including an estimate of the normal profit for the producer— where: are deducted.  = the residual value estimate of hydroelectric A NOTE ON SUBSIDIES AND TAXES resource rent in year t in the country in question In general, when estimating rent on natural assets, including hydroelectric assets, it is recommended to exclude subsidies on  = total revenue from sales of electricity production170 from the estimation and add taxes on production generated at so-called “renewable” hydroelectric plants 169 (SEEA-CF, section 5.4.5). Subsidies play a particularly important in year t in the country, including any subsidies paid on role in promoting new renewable energy generation, such as generation solar and wind. Consequently, equation 1 should be written as:  = cost for labor, materials, fuel, and other supplies to operate and maintain the produced assets (that is, the dams or other civil infrastructure required to create reservoirs plus the hydraulic turbines and other where is an estimate of the subsidies received equipment needed to generate electricity and transfer by hydroelectric producers. Subsidies might come in any of it from the hydroelectric station to the local power grid) several forms: used to generate hydroelectricity in year t in the country ■ Subsidies paid directly on production; for example, a  = the economy-wide average annual rate of return to subsidy paid to electricity producers per unit of electricity produced capital in the country (a constant) generated.  = the total value of produced capital used to ■ Indirect subsidies that reduce costs of production but are generate hydroelectricity in year t in the country not directly related to it; for example, an increased capital consumption allowance rate permitting producers to  = the annual rate of depreciation of the produced write investments in certain kinds of produced assets off capital used to generatehydroelectricity in the country (a more quickly than normal. constant). 169  Renewable hydroelectric plants are those where water flows through the hydraulic turbines only because of natural forces. These contrast with so-called “pumped storage” plants where the water flowing through the turbines is pumped from a lower reservoir below the turbines back into an upper reservoir to be used again. This pumping usually occurs at night when demand for electric power is low and excess power is therefore available from non-hydro sources. Pumped storage plants were not considered in this study. So-called “mixed” plants are those which include some pumped storage capabilities along with renewable generation. For the purposes here, mixed plants were considered renewable. 170  In some countries, subsidies are also paid on consumption; for example, many governments artificially lower the price of gasoline or electricity by providing households with direct payments that offset some of their spending on these goods or by holding the producer prices of these goods artificially low. Depending on the nature of these consumption subsidy schemes, and how they impact producer process and revenues, they too might have to be taken account of in estimating resource rent. If, for example, the revenue earned by hydroelectric generators, , was derived using price data that reflect subsidized household prices, then that revenue estimate will be too low by an amount equal to the consumption subsidy. In that case, the value of the consumption subsidy would have to be added back in the estimation of resource rent. Furthermore, countries that subsidize household energy consumers often also compensate producers for their foregone revenues through explicit or implicit production subsidies, such as fiscal transfers, tax breaks, or higher charges to industrial and commercial end users. 125 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Subsidies on production should be deducted when estimating FROM RENT TO RESOURCE VALUE rent as they increase net revenue from resource exploitation With hydroelectric rent estimated following equation 1 for (either directly in the case of those paid on production volumes each country and year between 1995 and 2020, the next step or indirectly for those that reduce production costs) and, by was to determine the expected pattern of future rents for consequence, increase resource rent derived via RVM. Since the NPV calculation. This required decisions regarding two subsidies do not represent a return to nature, they should be parameters: the level of rent in future years and the number excluded from resource rent. of years for which rent will flow. Regarding the latter, it was assumed rent will flow for 100 years—in keeping with the In practice, data on subsidies paid to natural resource assumption used in the valuation of other renewable natural companies are difficult to obtain (especially data on indirect resource assets in CWON. As for the former, in keeping with and implicit subsidies), so subsidies are generally not the general approach to renewable natural resource asset accounted for by statistical agencies when they are compiling valuation in CWON and in the SEEA-CF (SEEA-CF, paragraph official national estimates of resource asset values.171 This is the 5.133), it is assumed that future hydroelectric rents will be case in the estimation of rents on fossil fuel assets, for which equal to the rent observed in the period in question. For detailed data on subsidies by country and type of production example, to value hydroelectric assets for 2020, a 100-year are unavailable. series of rental incomes is assumed equal to the estimated 2020 rent in the NPV calculation. According to the SNA and SEEA-CF, subsidies paid by governments to support natural resource production should With the current rent and its expected future pattern be accounted for; that is, deducted from revenues in the determined, estimation of the value of hydroelectric calculation of resource rent. In practice, however, it is difficult assets in each country proceeded according to equation 2. to know to what extent estimates of hydroelectric asset values do include subsidies paid on hydroelectricity generation or 2) consumption. Direct subsidies paid to producers may be missed because they are unlikely to be reflected in end-user where: electricity prices, and, as explained later, the methodology used here relies on residential end-user prices for estimating  = the value of hydroelectric assets in year t in the producers’ revenues and hydroelectric resource rent. country Indeed, this shortcoming was the main reason why solar and  = the resource rent accruing to hydroelectric wind assets could not be included in the study, since direct assets in year t (as defined in equation 1 and including subsidies on production (for example, feed-in-tariffs) play a subsidies) in the country more important role in solar and wind producers’ revenues  = hydroelectric asset life in years (assumed to be 100 than for hydroelectric producers. Deeper understanding of years in all countries) the extent to which the NPV-RVM valuation approach taken here—indeed, taken in the case of any of the natural assets  = future periods from 1 to 100 included in CWON—would require further study. For this reason, subsidies are not accounted for in the valuation  = economy-wide discount rate (assumed, following of hydroelectric assets, which is consistent with all other CWON convention, to be 4 percent for all countries and CWON assets. years).  or example, Statistics Canada’s estimates of Canada’s oil assets do not account for subsidies. The same is true of the United Kingdom, Australia, Norway, 171 F and other countries. CWON’s approach of not considering subsidies for these assets is thus aligned with official statistical practice, if not with international statistical guidance. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 126 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Data sources and assumptions Given the lack of a global database, producer prices were estimated indirectly. This was done starting with the only REVENUES FROM ELECTRICITY GENERATION available database of electricity prices close to global coverage, Two data points were required to estimate revenues from the IEA’s Energy Prices database.174 This database, which is hydroelectricity generation ( from equation 1): only available by paid subscription, contains weekly, monthly, quarterly, and annual end-user (residential, commercial, and ■ The quantity of hydroelectricity generated industrial) electricity prices in nominal local currency units (LCUs) for 140 countries from 1970 to 2022. Since the most ■ The price received by hydroelectric power producers in complete country coverage in this database was for annual each country and year. residential end-user prices, those prices (adjusted to account for delivery charges and other non-production costs—see Obtaining global data on generated quantities of below) were chosen as the basis for estimating the annual hydroelectricity is relatively straightforward. The International producer prices required for this study. For countries not Renewable Energy Agency (IRENA) provides these data covered by the IEA database, regional average residential annually for most countries in the world going back to 2000.172 electricity prices were calculated and used as a proxy for The UN provides similar data covering the period back to the national prices. For countries included in the IEA database 1990s through its Energy Statistics Database.173 The generation but with data missing for certain years, missing data were data used in this study were derived from a combination of estimated from the available data using either backward or these two sources. In general, the two sources agreed exactly forward linear extrapolation. on generation figures for a given country and year. Where they did not, a simple average of the data from the two sources was To estimate producer prices in each country, IEA residential used unless there was clear reason to prefer the figure from prices were multiplied by a time-invariant conversion factor one over the other. that reflects the share of the price expected to be received by hydroelectric producers. These conversion factors were Obtaining data on the average annual prices received by determined empirically for the individual countries listed in hydroelectric power producers (“producer prices” hereafter) Table 6.1 by identifying factors that, when applied to the IEA at the country level proved more difficult, as no globally residential electricity price data, resulted in figures that best complete database of producer prices exists from public or matched the electricity prices used for the country in question private data suppliers. Preparing such a database would be in the pilot CWON study of renewable energy assets (Smith challenging given the difficulties of estimating annual prices et al. 2021). The prices used in the pilot study are considered when energy markets in countries with competitive wholesale accurate because they were derived from country-level electricity markets today include pricing mechanisms that electricity market data. In determining these country-specific adjust to demand and supply on an hourly basis. In addition, in factors, priority was given to finding factors that resulted in competitive, regulated, and hybrid electricity market models prices that best matched pilot study prices for recent years. there are multiple mechanisms for generating revenues by For countries not included in the pilot study, another approach power producers. Besides electricity, electricity producers to determining the conversion was required. also sell capacity readiness and other ancillary services that system operators buy to maintain grid stability and security. 172 See https://pxweb.irena.org/pxweb/en/IRENASTAT?_gl=1*1djkx00*_ga*MTM2ODIxMzA0Mi4xNjk4NjkyMjYx*_ga_7W6ZEF19K4*MTY5ODY5MjI2MC4xLj EuMTY5ODY5MjU0Mi42MC4wLjA. See http://data.un.org/Data.aspx?d=EDATA&f=cmID%3aEC. 173  174 See https://www.iea.org/data-and-statistics/data-product/energy-prices#overview. 127 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 6.1 Residential-to-producer price conversion factors by country/region RESIDENTIAL-TO-PRODUCER COUNTRY/REGION PRICE CONVERSION FACTOR Canada 0.6 United States 0.28 Australia 0.25 Brazil 0.5 China 0.72 Japan 0.46 Russian Federation 0.5 Türkiye 0.6 India 0.9 Europe and Central Asia region (western and central European countries) 0.25 Europe and Central Asia region (other than western and central European countries) 0.61 Rest of world, competitive markets 0.61 Rest of world, non-competitive markets 0.8 Source: World Bank staff estimates. For countries in western and central Europe, the factor the factor was set to 0.8, based on expert judgement. This (0.25) was similarly chosen to best match the prices used value was chosen to reflect the likelihood of government in the pilot study for other European countries (France, subsidization of residential prices in these countries, with Germany, Italy, Spain, Sweden, and the UK). For all other households paying capped electricity prices close to what countries with a competitive electricity market according to power producers themselves receive. the World Bank’s Global Power Markets Structure Database (Akcura, forthcoming), the average conversion factor (0.61) As a test of the reasonableness of the above approach, price of the countries listed in Table 6.1 was used, weighted by data from a commercial database from the Energy Regulators each country’s share of the combined 2020 hydroelectricity Regional Association (ERRA) were analyzed.175 Although generation. This factor implies that transmission and the ERRA database covers only 44 countries and is missing distribution charges and trader/supplier margins, on average, many values, it does offer reasonable coverage of residential account for 39 percent of residential prices. Finally, for other and producer prices for electricity at the country level on a countries deemed not to have competitive electricity markets, quarterly basis for the period 1999 to 2022. 175 See https://erranet.org/erra-tariff-database/. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 128 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH An analysis of the ratio of producer to residential prices in Indicators database).176 Finally, total annual revenues from this database suggests that the ratio of 0.61 applied to most sales of hydroelectric power by country ( ) were countries in this study is appropriate. For the period and obtained by multiplying estimated hydroelectric generation countries covered by the ERRA database, the average ratio of by the estimated producer price for each country and year. producer to residential prices was 0.58. These countries are, for the most part, like those to which the factor of 0.61 was Costs of electricity generation applied; that is, lower- or middle-income countries. Hydroelectricity generation costs are of two types, both of This approach to converting residential prices to producer which had to be estimated indirectly. prices is pragmatic but has several limitations. First, the assumption of a time-invariant relationship between residential User costs of capital: The annual costs of employing the ■  and producer prices within a given country is unlikely to be hydroelectric power plant (including the dam and any held in practice. For example, assuming time-invariance other civil works) in the production process ( implies that electricity transmission/distribution providers in equation 1), comprising the expected annual return are affected equivalently to electricity generators when to the owner of the power plant plus the annual cost of residential electricity prices change, which is not necessarily depreciation of the power plant (the variable in the case. It is more likely that generators and transmission/ equation 1). distribution providers will be affected differently by price changes. Such nuances could not be approximated in the study Operating and maintenance costs: The annual expenses ■  due to the unavailability of a comprehensive global electricity required to operate and maintain the power plant, producer price database. Second, the assumption that the including labor, materials, fuel, and other supplies same time-invariant ratio can be applied across countries is a ( in equation 1). simplification. For example, a lower ratio should be expected in large and sparsely populated countries, where transmission CAPITAL COSTS and distribution costs are a larger share of consumer prices. As with electricity prices, no global database of country-level Still, until a global database of producer prices is compiled, capital costs for hydro generation exists. The closest to this these pragmatic assumptions are the most robust available is a set of regional investment cost estimates available from way to proceed with the estimation. This is discussed further IRENA as part of its annual report on costs of renewable in the final section of the chapter on future work. energy generation (IRENA 2023).177 IRENA presents these estimates as regional averages for two periods (2010–2015 and Following the estimation of producer prices in LCUs for 2016–2021),178 with separate estimates for the costs of installing all countries, conversion from LCUs to US dollars was large and small hydro plants. To render these capital cost data accomplished using the market exchange rate of the reference suitable for use in the study, it was necessary first to extend year (that is, prices in 1995 LCUs were converted to US dollars them to cover the full period (1995–2020)179 and then to convert by applying the 1995 LCU to the US dollar market exchange the figures from constant US dollars to nominal US dollars rate, obtained from the World Bank World Development using the implicit GDP price deflator for the United States.180 176 See https://databank.worldbank.org/source/world-development-indicators. 177 See https://www.irena.org/Publications/2023/Aug/Renewable-Power-Generation-Costs-in-2022#. 178  IRENA’s regional breakdown is Asia, Africa, Central America and the Caribbean, Eurasia, Europe, the Middle East, North America, Oceania, and South America. In addition to these regions, the IRENA capital investment cost data provide specific estimates for three countries: Brazil, China, and India. 179  Annual values for the IRENA capital costs were estimated as follows: annual investment costs during the 2010–2015 period were assumed equal to IRENA’s 2010–2015 average investment cost value, for the 2016–2020 period they were assumed equal to IRENA’s 2016–2021 average value, and annual investment costs prior to 2010 were assumed equal to the average of IRENA’s 2010–2015 and 2016–2021 values. 180 This approach parallels that used by IRENA to derive the constant price values (IRENA, pers. comm.). 129 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH With an annual time series of nominal capital investment costs  is the long-term annual growth of real GDP in the from 1995 to 2020 by country in hand, the next task was to create country, derived from World Bank data. a time series of values of the produced capital stocks used in hydroelectric generation in each country ( ). Estimating The main missing piece of information in equation 3 was was complicated by the fact that considerable the value of investment, , which had to be separately investment in hydroelectric generation infrastructure took estimated. The installed hydroelectric generation capacity place before 1995 in almost all countries. Therefore, an in each country in 1994 was divided by the assumed age of estimate was required of the 1994 produced capital stock the oldest hydroelectric plants in the country182 to derive an value for each country before the 1995–2020 time series could estimate of the annual average quantity of capacity additions be compiled. The 1994 estimate was derived by applying an over the history of the country’s hydroelectric power industry. approach outlined in the OECD manual on measuring capital This quantity was taken to be the addition of new capacity stocks (OECD 2009, section 15.7). According to that approach, a in 1994, which was then multiplied by the estimated 1994 reasonable estimate of the stock of produced capital in any base investment cost derived from the IRENA data to estimate the year may be derived by dividing the value of investment in the value of in nominal US dollars. base year by the sum of the capital’s deprecation rate plus the Once was estimated, an annual time series of long-term growth rate of real GDP in the country in question. hydroelectric produced capital stocks (nominal US dollars) Equation 3 expresses this approach to estimating base year from 1995 to 2020 for each country was estimated using a stocks of hydroelectric power plant produced capital stocks. standard perpetual inventory method approach. That is, produced capital investment was added in each year to the 3) previous year’s stock value and depreciation was deducted. where: The value of investment in each year was then calculated by multiplying the newly installed hydroelectric generating  is the value of the produced capital stock used for capacity in that year183 by that year’s estimated value of capital hydroelectric generation in the base year (1994 in all but a investment costs per unit of installed capacity (in MW). few cases181) in each country OPERATING AND MAINTENANCE COSTS  is the value of investment in produced capital In addition to providing estimates of capital investment costs by used for hydroelectric generation in the base year in the region, the IRENA renewable energy cost report (IRENA 2023) country provides estimates of operating and maintenance costs for hydroelectric plants. These estimates are highly generalized,   is the annual rate of depreciation of produced however, with IRENA simply reporting that, on average, capital used for hydroelectric generation (a constant operating and maintenance costs at hydroelectric plants can of 1.67 percent in all countries and years based on the be assumed to be around 2 percent of the capital cost of the assumption that hydroelectric-generating dams and installed produced capital. In the absence of a better estimate, equipment have universal 60-year service lives) this figure was applied uniformly to all countries and years. 181  our countries had no installed capacity for hydroelectric generation in 1995: Belize, Cambodia, Liberia, and Sierra Leone. For these countries, F the value of produced capital stocks used in hydroelectric generation were simply estimated by cumulating the net investment in produced capital beginning in whatever year the country’s hydroelectric generation began. This age was taken to be 50 years in all countries except Brazil, India and those in North America, Eurasia, Europe, North America, and Oceania, 182  where it was assumed to be 75 years. The newly installed hydroelectric generating capacity was calculated as the difference between the opening and closing stock of installed generating 183  capacity in the year. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 130 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Estimating rent and asset values chapter 2 for further details). Volume indexes offer estimates of real asset values that better reflect the role assets play in With es timates of the revenues generated from hydroelectric production processes, which is central to the assessment of generation and the associated capital and operating and sustainability. Deflation using a price index is appropriate maintenance costs in hand, it was straightforward to estimate when the role assets play as stores of value to fund future the rent attributable to hydroelectric assets in each country consumption is the key issue. As explained in chapter 2, this and year using equation 1. The only additional variable is only in the case of financial assets. required was , the economy-wide average annual rate of return to produced assets. Ideally, country-specific values of To produce the real value of hydroelectric assets in chained would have been used, but such rates are not readily available. 2019 US dollars following the approach outlined in chapter The following annual rates were assumed instead (intended 2, a physical volume measure was needed to represent the to reflect real returns): 4 percent in Europe, North America, quantity of hydroelectric assets available at any given time in and Oceania; 8  percent in Africa, Central America and the each country. The quantity of electricity generated annually Caribbean, Eurasia, the Middle East, and Latin America; and in a country measured in megawatt hours was chosen as 10 percent in Asia. These are the same rates as those used in the physical volume measure.185 An alternative could have the pilot study.184 Once the hydro resource rent was estimated, been the installed generating capacity measured in MW to the final step was to calculate the value of hydroelectric assets represent the physical volume. This was rejected because as the present value of future rents over the assumed lifetime the installed capacity fails to capture the actual volumes of of hydroelectric assets (100 years) using equation 2. valuable electricity generated due to the different operational priorities of multifunctional reservoirs. Furthermore, changes In certain instances, the value of dropped below zero in generated quantities over time implicitly capture quality in a given country and year due to temporary situations with changes in both the hydroelectric asset and the produced respect to electricity prices or electricity generation levels, assets used to capture it, which is a desirable feature for the both of which fluctuate over time. In those instances, the value volume index. Due to aging of equipment and environmental of was set to zero. factors such as sedimentation of reservoirs (Schellenberg et al. 2017), there tends to be a reduction in the capacity use ESTIMATING HYDROELECTRIC ASSET VALUES factor of a given hydroelectric plant to generate electricity IN REAL TERMS over time. Moreover, the changing climate is affecting the In the past, CWON presented asset values in real terms by availability of water resources in varying ways across the deflating nominal asset values using an economy-wide GDP planet, meaning that previous generation levels may become implicit price deflator for each country. This practice has difficult to maintain due to declining water availability.186 The been replaced in the 2024 edition with a volume index (see next section presents the results in more detail.  aking assumptions regarding these rates is less than ideal. However, the impact on the overall results is muted by the fact that the expected returns 184 M to produced assets do not have a large bearing on the value of resource rent. For example, reducing the assumed rate of return on produced assets by 25 percent (from 4 percent to 3 percent) for Canada increases the 2020 estimate of hydroelectric resource value by just 7 percent. In the case of a single, homogenous asset like hydroelectric assets, changes in the volume index over time are driven entirely by changes in the physical 185  quantity of the asset (in this case, by changes in the quantity of hydroelectricity generated over time). Expressing those changes in chained 2019 US dollars is merely a matter of presentation, done to put the changes in terms that are more familiar in an economic context. Changes in the prices of hydroelectric assets relative to other assets do, however, come into play when those assets are aggregated with other assets into the broader volume indexes of natural capital and, ultimately, comprehensive wealth presented in this report (see chapter 2 for details). In those indexes, changes in hydroelectric assets will figure more or less prominently in changes in the broader indexes, depending on whether hydroelectric assets increase or decrease in value relative to other assets over time. This is happening, for example, in the Colorado River basin of the United States, where the Hoover Dam is less and less capable of generating 186  electricity to its full potential because of reduced water levels in its reservoir, Lake Mead. (NASA, no date. See https://earthobservatory.nasa.gov images/150111/lake-mead-keeps-dropping.) 131 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH HYDROELECTRIC ASSET VALUES global average. In Latin America and the Caribbean, where the capacity factor was greatest in 1995, the 2020 factor was Table 6.2 compares hydroelectric asset values in chained 2019 only 80 percent of the 1995 level. Declines in other regions US dollars for 1995 and 2020 for World Bank country regions were also considerable. Reasons for declines in capacity may and shows significant changes in the regional pattern over include aging of generating equipment and dams, insufficient time. Latin America and the Caribbean was the region with investment in repairs and maintenance, and reduced water the greatest hydroelectric asset value in 1995, due mainly to flow due to climatic factors or sedimentation of reservoirs. Brazil, which had the largest hydroelectric asset value of any Surprisingly, East Asia and the Pacific did not see a significant country in 1995 (Table 6.3). The regions of East Asia and the increase in its capacity factor despite more than quadrupling Pacific, Europe and Central Asia, and North America also its installed capacity. A significant addition of new capacity had large values in 1995. By 2020, the regional picture had would usually have a noticeable increase in the capacity factor, changed considerably, however, with East Asia and the Pacific as new plants generally employ improved technologies and more than tripling its 1995 asset value to gain the top spot by enjoy maximum hydraulic flow.189 a small margin over Latin America and the Caribbean. Except for Europe and Central Asia and North America, other regions Figure 6.1 shows the trend in hydroelectric asset value by saw substantial growth in value over the period as well. The region from 1995 to 2020. Again, the growth of East Asia and lack of growth in Europe and Central Asia and North America the Pacific—largely driven by China—stands out clearly here reflects the maturity of the hydroelectric power industries and the significant global increase is equally apparent. The in those regions and the relative lack of suitable sites for same data, normalized by population, are shown in Figure continued expansion. 6.2. Here the story that emerges is somewhat different. East Asia and the Pacific significantly increases hydroelectric asset Table 6.2 also presents data on the installed hydroelectric wealth per capita, but it does so at a much lower level than generating capacity and electricity generation by region North America and the Latin America and Caribbean region, along with a related variable known as capacity factor, a both of which produced about three times more hydroelectric measure of generation efficacy (Bolson et al. 2022; Prado asset wealth per capita in 2020 than East Asia and the Pacific. Jr. and Berg 2011).187 Looking at weighted average capacity Globally, the growth from 1995 to 2020 is much less impressive factors by region, Sub-Saharan Africa emerges as the leader, when considered in per capita terms. So, while hydroelectric being the only region to improve its factor significantly from asset wealth is growing overall, it is not at a rate much faster 1995 to 2020, 188 and one of only two regions to exceed the than population growth.  apacity factor is the ratio between actual generation each year and the theoretical maximum generation if all installed capacity operated at full output 187 C throughout the year. Variables influencing the capacity factor include weather (less precipitation means less waterflow is available to keep reservoirs full); age of generating equipment (equipment tends to decline in efficiency as time goes by); regularity of repair and maintenance of equipment; unplanned outages; lower than expected demand; quality of operations; and limitations on waterflow from competing needs (for example, irrigation, fish migration) or sedimentation of reservoirs. The capacity factor in each country is weighted by the country’s share of regional (global) hydroelectric power generation in calculating the regional (global) weighted average capacity factor. East Asia and the Pacific increased its generational efficacy, but only marginally. 188  189  About 0.5 percent to 1 percent of the total volume of water stored in hydro reservoirs around the world is lost annually because of sedimentation. As a result, global per capita reservoir storage has rapidly decreased since its peak in about 1980. Current storage is equivalent to levels that existed nearly 60 years ago (Schellenberg et al. 2017). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 132 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 6.2 Hydroelectric asset value, installed capacity, generation, and weighted average capacity factor by region, 1995 and 2020 1995 2020 Weighted average capacity factor Weighted average capacity (million chained 2019 US$) (million chained 2019 US$) Hydroelectric asset value Hydroelectric asset value Installed capacity (MW) Installed capacity (MW) Generation (GWh) Generation (GWh) factor Region1 East Asia and the 390,656 99,664 359,286 0.426 1,134,255 433,467 1,606,162 0.430 Pacific (4,1) Europe and Central 401,783 215,228 744,168 0.449 454,350 275,712 890,353 0.424 Asia (3, 4) Latin America and 628,449 106,566 491,777 0.543 961,673 197,485 722,329 0.435 the Caribbean (1,2) Middle East and 1,216 6,498 20,512 0.427 5,382 17,685 40,817 0.363 North Africa (7, 7) North America (2, 3) 509,042 144,886 646,638 0.521 560,621 165,123 673,646 0.478 South Asia (6, 6) 33,799 26,808 98,887 0.429 68,474 59,343 208,846 0.405 Sub-Saharan Africa 103,318 15,966 46,333 0.445 249,591 27,807 118,961 0.527 (5, 5) World 2,068,263 615,616 2,407,602 0.483 3,434,346 1,176,622 4,261,114 0.438 Source: World Bank staff estimates. Note: Numbers in brackets represent the global ranking of the region in terms of hydroelectric asset value in 1995 and 2020, respectively. 133 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 6.1 Hydroelectric asset value by region, 1995–2020 2,000 hydro per capita (real chained 2019 US$) 1,500 1,000 500 0 1995 2000 2005 2010 2015 2020  East Asia & Pacific  Sub-Saharan Africa South Asia   Latin America & Caribbean  Europe & Central Asia World  North America  Middle East & North Africa Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. FIGURE 6.2 Hydroelectric asset value per capita by region, 1995–2020 2,000 hydro per capita (real chained 2019 US$) 1,500 1,000 500 0 1995 2000 2005 2010 2015 2020  East Asia & Pacific  Sub-Saharan Africa  South Asia  Latin America & Caribbean  Europe & Central Asia World  North America  Middle East & North Africa Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 134 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Table 6.3 presents the same variables as in Table 6.2 but for sixfold increase in Chinese generating capacity between the 10 countries with the highest hydroelectric asset values 1995 and 2020. Only one country, Vietnam, ranked in the top in 2020. China had the largest asset value in 2020, though it 10 in 2020 but not in 1995. Vietnam rose from 28th in 1995 to had ranked only fourth in 1995. Brazil and Canada rounded sixth in 2020 due to an investment program that increased out the top three in 2020, having been bumped from first and its hydroelectric generating capacity more than sevenfold second places in 1995, respectively, by China’s emergence over the period. One country, France, ranked in the top 10 in as the world’s leading hydroelectric power producer. This 1995 (in eighth place) but not in 2020 (falling to 16th place). move into first place in 2020 was thanks to the more than TABLE 6.3 Ten wealthiest hydroelectric countries, 2020 1995 2020 (million chained 2019 US$) (million chained 2019 US$) Hydroelectric asset value Hydroelectric asset value Installed capacity (MW) Installed capacity (MW) Generation (GWh) Generation (GWh) Capacity factor Capacity factor Country China (4, 1) 112,061 51,135 186,622 0.417 793,606 340,504 1,321,641 0.443 Brazil (1, 2) 447,673 51,346 253,905 0.564 698,834 109,306 396,355 0.414 Canada (2, 3) 398,369 64,573 335,933 0.594 458,342 81,311 386,506 0.543 Japan (3, 4) 182,157 21,171 84,454 0.455 169,977 28,139 78,807 0.320 United States (5, 5) 110,672 80,313 310,706 0.442 102,278 83,811 287,140 0.391 Vietnam (28, 6) 9,178 2,827 10,582 0.427 63,680 20,972 73,422 0.400 Sweden (7, 7) 53,168 15,725 68,074 0.494 56,538 16,406 72,389 0.504 Italy (10, 8) 43,297 15,868 38,224 0.275 53,863 18,755 47,552 0.289 Paraguay (9, 9) 43,771 6,861 39,712 0.661 51,110 8,785 46,371 0.603 New Zealand (6, 10) 57,377 5,259 27,532 0.598 50,571 5,434 24,266 0.510 Source: World Bank. Note: Numbers in brackets represent the global ranking of the country in terms of hydroelectric asset value in 1995 and 2020, respectively. 135 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Table 6.4 normalizes the values from Table 6.3 by dividing average. Had the United States generated wealth from its them by installed hydroelectric generating capacity to hydroelectric resources at just the global average rate of enable a comparison of how successful the 10 wealthiest $4.5 million per MW, its assets would have been worth hydroelectric countries were in turning their resources close to $400 billion in 2020, quadrupling its actual hydro into wealth. Normalized values better suit such a country wealth. China also stands out for a low rate of conversion of comparison, since a megawatt of installed capacity does not resources into wealth, generating only $2.3 million of value differ greatly from one country to the next, but the wealth per MW of installed capacity. Notably, every country in created from it can. Unit wealth created from renewable Table 6.4 other than China and Italy saw its global ranking energy resources is, then, a useful measure of a country’s in wealth creation per MW slip from 1995 to 2020. relative success in effectively exploiting its hydroelectric The results in Table 6.4 show that most of the 10 wealthiest asset base to generate well-being for its citizens. hydroelectric countries have become less successful in Interestingly, Table 6.4 shows that of the 10 countries with generating wealth from their hydroelectric resources over the greatest hydroelectric wealth in 2020, only New Zealand time. In addition, they are no more successful on average made it into the top 10 countries globally in terms of wealth than other countries in converting hydroelectric resources generation per unit of installed capacity, and only in 1995. into wealth—several of them much less so. Clearly, having New Zealand ranked 15th globally in this regard in 2020, large and valuable hydroelectric assets does not mean these creating $9.3 million (chained 2019 US dollars) per MW of resources are exploited as effectively as they might be. The installed capacity in that year. This was more than twice the reasons for this trend are likely multifaceted, and assessing global average of $4.5 million per MW in 2020. Brazil and them was beyond the scope of this study, but it is clear Japan were the next most effective at turning hydroelectric that the declining efficiency of hydroelectric systems (as resources into wealth in 2020, creating $6.4 million and measured by capacity factor) played a role in the declining $6 million per MW, respectively. Half of the other 10 effectiveness of converting resources into wealth in several wealthiest countries (Brazil, Canada, Japan, New Zealand, countries (Brazil, Japan, the United States, Vietnam, and Paraguay) generated more wealth per MW than the Paraguay, and New Zealand), As shown in Table 6.3, these global average. The United States, at just $1.2 million per six countries all experienced considerable declines in MW, was last among the 10 countries in terms of creating capacity factor from 1995 to 2020. wealth from its hydroelectric resources, far below the global  or example, the figure in the upper righthand corner of Table 6.4 indicates that China created $2.3 million (chained 2019 US dollars) in hydroelectric 190 F asset value per unit of installed generating capacity in 2020. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 136 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 6.4 Hydroelectric asset value per unit of installed capacity, 10 wealthiest hydroelectric countries, 1995 and 2020 1995 2020 CHANGE 1995–2020 Hydroelectric Hydroelectric Country asset value per unit asset value per unit of installed capacity of installed capacity Percent (million chained 2019 (million chained 2019 US$/GWh) US$/GWh) China (64, 63) 2.2 2.3 4.5% Brazil (15, 24) 8.7 6.4 -26.4% Canada (29, 28) 6.2 5.6 -9.7% Japan (16, 25) 8.6 6.0 -30.2% United States (74, 80) 1.4 1.2 -14.3% Vietnam (48, 53) 3.2 3.0 -6.3% Sweden (47, 50) 3.4 3.4 0.0% Italy (57, 54) 2.7 2.9 7.4% Paraguay (27, 26) 6.4 5.8 -9.4% New Zealand (9, 15) 10.9 9.3 -14.7% Source: World Bank. Note: Numbers in brackets represent the global ranking of the country in terms of hydroelectric asset value in 1995 and 2020, respectively. DISCUSSION OF RESULTS AND that cohere with the theoretical expectation that positive FUTURE RESEARCH AGENDA resource rents arise where natural resources are exploited in competitive markets with mature technologies (Smith et Several lessons have emerged from this first effort to develop al. 2021). The results of this study show that hydroelectricity global hydroelectric asset values. First, there are sufficient resources do, indeed, generate positive rents and asset values data available from global and national sources to implement in most countries and years. the NPV-RVM approach for the valuation of hydroelectric assets, although data on electricity prices and the costs of Second, the estimated values for hydroelectric assets are generation are not as robust as those on the quantities of globally consequential. In total, 2020 hydroelectric assets electricity generated or the installed generating capacity. were valued at $3.6 trillion (chained 2019 US dollars). This Further, and more importantly, NPV-RVM produces results places them at the same level as other renewable natural 137 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH resource assets. Global fisheries are estimated to have In China, Brazil, and Canada, the three wealthiest countries been worth $21.8 trillion in 2020. The corresponding value in terms of hydroelectric assets in 2020, hydroelectric assets for global timber assets is $2.8 trillion. These results show are comparable in value to the countries’ fossil fuel assets, that leaving hydroelectric assets off national balance sheets which were valued at $1,421.2 billion, $66 billion, and $162 misses a great deal of natural wealth. Yet, no country today billion, respectively, in 2020.191 Globally, hydroelectric assets includes hydroelectric assets (or any other renewable represented 23 percent of all energy assets192 in nominal energy asset) on its national balance sheet. This omission is terms in 2020. Figure 6.3 illustrates that, when calculating particularly significant for countries with large hydroelectric wealth in current nominal terms, the share of hydroelectric resources. In Paraguay, the estimate of hydroelectric asset assets in total energy wealth is inversely correlated with value represents 47 percent of the country’s total 2020 natural global fossil fuel prices. For example, in 2020 the share of capital. In Japan, the corresponding figure is 13 percent. For hydroelectric asset value increased because fossil fuel prices these countries, the inclusion of hydroelectric assets gives hit bottom due to the COVID-19 pandemic. Similarly, the an entirely different view of the contribution of renewable hydroelectricity share was high in 2016 because of low fossil natural capital to overall wealth. fuel prices in that year due to weakness in financial markets and oversupply of crude oil. FIGURE 6.3 Share of hydroelectric asset value in total energy asset values 40 30 % 20 10 1995 2000 2005 2010 2015 2020 Source: World Bank staff estimates. Note: The share is calculated as the nominal value of hydropower wealth divided by the sum of the nominal value of hydropower, coal, oil, and natural gas wealth. All values calculated using nominal current market (or market-imputed) US dollars. 191 If the contribution of fossil fuel assets to climate change were netted against their value, then hydroelectric assets (which make smaller contributions to climate change) would appear even more valuable in relative terms. 192 Energy assets include fossil fuel, coal, natural gas, and hydroelectric assets. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 138 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH It is possible that hydroelectric assets have not been included accurately reflect the prices received by producers on any nation’s balance sheet to date because international and the extent to which PPAs include subsidies could statistical guidance on wealth accounting (both in the SNA be validated by undertaking detailed studies of the and the SEEA-CF) has been unclear in terms of concepts and hydroelectric industries in specific countries. methods for their valuation (Smith et al. 2021). This will soon change, however, as the updated SNA handbook expected ■ Treatment of subsidies. Subsidies are not accounted for to come into effect in 2025 will include guidance on valuing in the valuation of hydroelectric assets here. This was renewable energy assets using the NPV-RVM method adopted done to ensure consistency in the valuation of all natural in this study. The details of this proposal are available in resources across the CWON natural capital accounts. an endorsed guidance note on the UN website (Smith and As discussed earlier, although international guidance Peszko 2022). on natural resource valuation recommends adjustment for subsidies and taxes in estimating resource rent, Finally, while the validity of NPV-RVM for valuing the practice in national statistical offices is generally hydroelectric assets is supported by the results here, the to leave subsidies in due to the lack of data. Resolving approach could be further stress-tested using alternative this issue would require a global database of subsidies valuation methods. As discussed by Smith et al. (2021), an received by producers and consumers of natural approach known as the least-cost alternative method has resource commodities that could be used to estimate been used to value hydroelectric assets in several other resource rent net of subsidies. Such a database would be studies. Application of the least-cost alternative method challenging to build, but good work in that direction— to countries with substantial hydroelectric resources and at least, for subsidies to fossil fuels—is already being different electricity market structures could aid in validating done by the OECD (2022), the International Institute for the results of the method presented here. In addition to Sustainable Development’s Global Subsidies Initiative further testing to validate the NPV-RVM approach, the points (no date), and the IEA (2023). below could also be investigated in future. ■ Non-electricity income. Some water reservoir operators ■ Producer prices. As discussed earlier, the approach to are remunerated for services not related to electricity estimating prices received by hydroelectric producers generation, for example, irrigation, water supply, here has important limitations. A means to improve this recreation, and flood control. The extent to which data would be to compile a global database of average the costs associated with these services are included annual wholesale electricity market prices. Such prices in available data on costs associated with electricity are likely to be close to the prices received by production requires further investigation. In principle, hydroelectric producers, at least in countries with open only costs associated with electricity production should and competitive electricity markets. Such a database be included in valuing hydroelectricity assets. would also have shortcomings, however, as annual average electricity prices would fail to capture the ■ Initial produced capital stock estimates and dynamics of electricity markets, which can be important depreciation profiles. The approach used to estimate in countries using spot market pricing. Wholesale the value of the initial (1994) stock of produced market prices will also not reflect the prices received by hydroelectric generating infrastructure could be hydroelectric producers in cases where those producers improved. This is especially important in countries have negotiated long-term power purchase agreements where a high percentage of installed generating capacity (PPAs) that allow them to receive steady revenue was already in place in 1994. A related matter is the streams regardless of price trends in wholesale markets. depreciation profile applied to the produced capital The extent to which average annual wholesale prices stock. A very simple profile was assumed: constant 139 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH linear depreciation over a 60-year lifetime for the producers were not suitable for use in valuing solar/ generating infrastructure. Several aspects of this choice wind assets because those prices were designed merit further investigation. Is the assumed 60-year to reflect prices in wholesale markets. While such lifetime reasonable in all cases? Do hydroelectric prices are an acceptable proxy for the prices received reservoirs and dams typically last longer, for example? by hydroelectric producers, this is not the case for Should all infrastructure be assumed to depreciate solar/wind producers where large subsidies are linearly, or should other profiles be considered? paid directly to producers in the form of feed-in tariffs, technology and site-specific auctions, or ■ Returns to produced capital. Due to a lack of suitable other commercial arrangements under the PPAs. If information on rates of return to produced capital by those prices were used to value solar/wind assets, country, regional rates were assumed. This is not ideal these would have been substantially undervalued and additional research should be undertaken to identify compared to other renewable natural capital assets, country-specific rates for the next edition of CWON. including hydroelectric assets. For geothermal assets, it may well be that the data required to value them ■ Operation and maintenance costs. Due to a lack of are available at the national level in countries where actual operation and maintenance data, such costs were such resources are important (such as El Salvador, assumed to be a constant share of the produced capital Iceland, Kenya, New Zealand, and the Philippines), stock, and these shares were assumed to apply equally but carrying out valuations for individual countries across countries. It would be preferable to have directly using national data is not feasible for CWON, which observed data on operation and maintenance costs, as relies on the availability of global databases to they can be sizable, particularly in countries with low ease the burden on human and financial resources hydroelectricity capacity factors. required for the compilation of comprehensive wealth ■ Valuation of solar, wind, geothermal, and other accounts for 150 countries. The same could likely be renewable energy assets. Inclusion of renewable said for valuing the direct use of solar heat for hot energy assets beyond hydroelectric assets was not water heating, evaporation, and drying. The next step possible here due to data limitations. The main in the valuation of these assets would be a pilot study hurdle preventing the inclusion of solar and wind to test concepts and methods and assess available assets was the lack of a suitable time series of data sources, as was done for hydro, solar, and wind producer price data for solar/wind producers. The electricity assets in CWON 2021. producer prices that were estimated for hydroelectric REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 140 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH REFERENCES Sinner, J., and Scherzer, J. 2007. “The Public Interest in Resource Rent.” New Zealand Journal of Environmental Law 11: 279–295. Akcura, E. Forthcoming. Global Power Markets Structure Database. Washington, Smith, R., Ilas, A., Gustavo Inon, J., and Peszko, G. 2021. “Renewable DC: World Bank. Energy: Unaccounted Wealth of Nations.” In The Changing Wealth of Nations 2021. Washington, DC: World Bank. https://openknowledge. Bolson, N., Prieto, P., and Patzek, T. 2022. “Capacity Factors for Electrical worldbank.org/bitstream/handle/10986/36400/9781464815904_ch14. Power Generation from Renewable and Nonrenewable Sources.” Proceedings pdf?sequence=19&isAllowed=y. of the National Academy of Sciences 119 (52). https://doi.org/10.1073/ pnas.2205429119. Smith. R., and Peszko, G. 2022. Guidance Note on the Treatment of Renewable Energy Resources as Assets. Paper prepared for the Twenty-First Meeting of European Commission, International Monetary Fund, Organisation for the Advisory Expert Group on National Accounts, Washington, DC, October Economic Co-operation and Development, United Nations, and World Bank. 18–20. SNA/M4.22/14. https://unstats.un.org/unsd/nationalaccount/aeg/2022/ 2009. System of National Accounts 2008. Washington, DC: World Bank. M21/M21_14_WS11_Renewable_Energy_Resources.pdf. http://documents.worldbank.org/curated/en/417501468164641001/System-of- national-accounts-2008. United Nations. 2019. System of Environmental-Economic Accounting for Energy: SEEA-Energy. New York: United Nations. https://seea.un.org/sites/seea. Foster, V., and Rana, A. 2020. Rethinking Power Sector Reform in the Developing un.org/files/documents/seea-energy_final_web.pdf. World. Sustainable Infrastructure. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/32335. United Nations Economic Commission for Europe (UNECE). 2020. United Nations Framework Classification for Resources – Update 2019. ECE Energy Hartwick, J.M., and Olewiler, N.D. 1999. The Economics of Natural Resource Use. Series No. 61. Geneva: United Nations Economic Commission for Europe. 2nd Edition. Reading, Mass.: Addison-Wesley. https://unece.org/fileadmin/DAM/energy/se/pdfs/UNFC/publ/UNFC_ES61_ IISD (International Institute for Sustainable Development). No date. “Global Update_2019.pdf. Subsidies Initiative.” https://www.iisd.org/gsi/. United Nations Environment Programme (UNEP). 2023. Inclusive Wealth Report IRENA. 2023. Renewable Power Generation Costs in 2022. Abu Dhabi: 2023: Measuring Sustainability and Equity. Nairobi. https://wedocs.unep. International Renewable Energy Agency. https://mc-cd8320d4-36a1-40ac- org/handle/20.500.11822/43131#:~:text=The%20report%20highlights%20 83cc-3389-cdn-endpoint.azureedge.net/-/media/Files/IRENA/Agency/ how%20inclusive,of%20the%20Inclusive%20Wealth%20series. Publication/2023/Aug/IRENA_Renewable_power_generation_costs_in_2022. United Nations, European Union, Food and Agriculture Organization of pdf?rev=cccb713bf8294cc5bec3f870e1fa15c2. the United Nations, International Monetary Fund, Organization for Managi, S., and Kumar, P., eds. 2018. Inclusive Wealth Report 2018. London: Economic Co-operation and Development, and World Bank. 2014. System Routledge. https://www.taylorfrancis.com/books/e/9781351002073. of Environmental-Economic Accounting 2012 – Central Framework. New York: United Nations. https://unstats.un.org/unsd/envaccounting/SEEA-CFRv/ NASA. No date. “Lake Mead Keeps Dropping.” https://earthobservatory.nasa. SEEA-CF_CF_Final_en.pdf. gov/images/150111/lake-mead-keeps-dropping. UNU–IHDP and UNEP. 2012. Inclusive Wealth Report 2012 – Measuring Progress OECD. 2009. Measuring Capital – OECD Manual. 2nd Edition. Paris: OECD. Toward Sustainability. Cambridge: Cambridge University Press. https://read.oecd.org/10.1787/9789264068476-en?format=pdf. http://www.unep.org/pdf/IWR_2012.pdf. OECD. 2022. Fossil Fuels Consumption Subsidies 2022. Paris: OECD. https://www. UNU–IHDP and UNEP. 2014. Inclusive Wealth Report 2014 – Measuring Progress iea.org/reports/fossil-fuels-consumption-subsidies-2022. Toward Sustainability. Cambridge: Cambridge University Press. http://mgiep.unesco.org/wp-content/uploads/2014/12/IWR2014-WEB.pdf. Prado, F. Jr., and Berg, S. 2011. “Capacity Factors of Brazilian Hydroelectric Power Plants: Implications for Cost Effectiveness.” University of Florida: Wessel, R.H. 1967. “A Note on Economic Rent.” The American Economic Review Public Utility Research Centre. https://bear.warrington.ufl.edu/centers/purc/ 57 (5): 1221–1226. docs/papers/1305_Berg_Capacity_Factors_of.pdf. World Bank. 2021. The Changing Wealth of Nations: Managing Assets for the Rothman, M. 2000 “Measuring and Apportioning Rents from Hydroelectric Future. Washington, DC: World Bank. https://www.worldbank.org/en/ Power Developments.” World Bank Discussion Paper no. 419, World Bank, publication/changing-wealth-of-nations. Washington, DC. https://elibrary.worldbank.org/doi/abs/10.1596/0-8213- 4798-5. Schellenberg, G., Donnelly, C.R., Holder, C., and Ahsan, R. 2017. “Dealing with Sediment: Effects on Dams and Hydropower Generation.” Hydro Review 25 (1). https://www.hydroreview.com/tag/hrw-volume-25-issue-1/. 141 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Annex A4: Concepts of Resource Rent and Their Applicability to Hydroelectric Assets All resource rent concepts share a focus on the benefits In all cases, the fundamental source of rent is scarcity. accruing to a factor of production over and above what is Wessel (1967) considers that Ricardian rent is essentially a required to maintain that factor in the productive process, pure scarcity rent, as it is the scarcity of more productive though they highlight different circumstances by which factors that allows them to earn differential rents. Such these payments come about. The concepts can be roughly pure rents can be taxed or otherwise collected by society, as categorized as follows (Sinner and Scherzer 2007). the collective resource owner, without reducing incentives to invest by energy producers. If scarcity is not permanent, ■ Ricardian/differential rents—rents that accrue to the Marshall’s “quasi-rents” emerge until long-term equilibrium more productive factors of production in homogenous is reached. input markets. In equilibrium, the price at which the least-productive firm is willing to produce clears Not all renewable energy markets193 can be in long-term the market; all firms with marginal costs below this competitive equilibrium, especially not those in the price earn Ricardian (also called “differential”) rents rapidly emerging areas of solar and wind energy. This (Hartwick and Olewiler 1999). Classical economists has implications for the nature and level of rent and its recognized that the location of a resource could be the distribution among factors of production. For example, source of Ricardian rents. Ricardian/differential and scarcity/absolute rents are based on the supposition of market equilibrium. By contrast, ■ Scarcity/absolute rents—rents that arise when demand Marshallian quasi-rents are features of markets that are not exceeds supply in the long term. Since supply cannot in long-term equilibrium. be increased either for natural (fixed physical stock) or arbitrary (regulated entry barriers) reasons, “limits An additional challenge is that the inexhaustible nature of on the supply of a resource allow producers to charge renewable energy resources poses challenges to theories of prices greater than their marginal cost” (Rothman value and thus to theories of rent.194 This is most obvious for 2000, 4). wind and solar resources, though it applies to hydroelectric resources as well. Scarcity and differential rents arise ■ Marshallian short-run/quasi rents—rents that arise locally, however, as a given site can only be used for solar/ in the short term; that is, in the absence of a stable wind/hydro electricity production by one economic unit at long-term equilibrium. Quasi-rents arise when a time and because the resources themselves are variable demand exceeds supply at a fixed point in time and are in quality (wind currents are not the same everywhere, dissipated as the prospect of rent capture encourages the intensity of the sun varies with latitude, and the more entrants to the market. hydraulic characteristics of rivers differ). Scarcity may also be arbitrarily imposed; for example, through legislation granting excludable rights to generate and sell energy on these sites. 193 Renewable energy markets are here understood as more than just markets for electricity produced from renewable sources. They also include markets for producers of renewable energy generation technologies and their supply chains. 194 A related measurement problem arises when the supply of resources is increasing over time (or total expenditures are growing): a declining cost share of the resource is equated with declining productivity in growth accounting, producing a biased view of the contribution to economic growth over time. Santos et al. (2016) explore this issue with regards to structural changes in the energy supply in Portugal. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 142 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH 7 Forests and Agricultural Lands MAIN MESSAGES ■ Wealth per capita in land assets declined globally between ■ In 2020, water protection services represented the 1995 and 2020. Declines were recorded in all land asset greatest share of non-timber forest ecosystem services categories: agricultural lands (24 percent decline), in all regions except in the Middle East and North forest recreation services (18 percent decline), non-wood Africa. The share of water protection services was forest products (27 percent decline), forest water greater than 70 percent in Latin America and the services (26 percent decline), and timber Caribbean, North America, and South Asia. (28 percent decline). This means that investments in land assets have not ■ A key innovation of this edition of CWON compensated for population growth. is that it provides a first global estimate of the value of protected areas based ■ The share of land assets in total on the non-timber forest ecosystem wealth varies greatly across regions services they provide. Their contribution and income groups. In Sub-Saharan is significant, providing 16 percent of the Africa and South Asia land assets total wealth provided by forest non-timber represent the largest shares, about 30 ecosystem services. Their contributions vary percent and 32 percent in 2020, respectively. considerably by world regions, with the highest Agricultural lands and forest wealth represent less than percentage of wealth provided by protected areas in Sub- 10 percent in other regions. Saharan Africa (21 percent). ■ Low- and lower-income countries show the greatest declines in wealth per capita in land assets. Timber INTRODUCTION wealth and non-timber forest ecosystem services fell by more than 50 percent in low-income countries Forests and agricultural lands play a fundamental role in and by more than 30 percent in lower-middle-income fostering economic development. They offer a range of countries. Low-income countries as a group increased benefits, from livelihood opportunities to essential ecosystem their area of agriculture by 4 percent, while losing 5 services that contribute to economic growth and sustainability percent of forest area between 1995 and 2020. However, (World Bank 2021b; FAO 2022; Dasgupta 2021). this has not been reflected in greater wealth per capita Forests are not merely a source of timber or wood products. in land assets, partly because of high population They provide crucial ecosystem services, such as regulating growth rates. water cycles, preventing soil erosion, maintaining biodiversity, ■ Wealth per capita in land assets experienced a decline and opportunities for recreation that support vibrant tourism in all asset categories for most regions. Sub-Saharan industries in many countries (IPBES 2019). Their role in Africa shows the greatest declines, particularly in climate regulation is increasingly recognized as a key asset in forest water services, timber, and non-wood forest the fight against climate change, further contributing to the products (over 50 percent decline), followed by the economic well-being of societies (IPCC 2019, 2023). Middle East and North Africa, and Latin America and Agricultural lands are the foundation of food production the Caribbean (between 30 percent and 40 percent and essential economic development in low- and middle- decline in all assets). Only Europe and Central Asia income countries. They not only provide a vital source of show a small increase in per capita wealth in forest sustenance but also underpin global trade and agribusiness recreation services. 143 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH (Zabel et al. 2019). In addition, the agricultural sector offers the land generates and takes the present value of such rents employment to a substantial portion of the global population in the future. Given that information on land transactions is (FAO 2023), particularly in developing countries, contributing often missing, the second method is used. significantly to their economies. For each country in the database, annual resource rents in Although not studied in this chapter, the synergy between agricultural lands TRt are the sum of the rents Rk,t, for each forests and agricultural lands is also critical. Forests support crop/livestock product k, in year t. Rents are the product of agricultural productivity by regulating ecosystem services price pk,t , quantity produced qk,t , and the average rental and habitat for pollinators and natural pest control (Johnson rate parameter at : et al. 2021). In turn, agriculture can positively influence forests by promoting agroforestry systems, which combine 1) the cultivation of trees with agricultural crops, enhancing land productivity while preserving forest cover (Miller et where: al. 2020). The delicate balance between these land assets is 2) vital for sustainable economic development. Recognizing the importance of conserving and sustainably managing both and k = 1,…,n for the number of crops/livestock products forests and agricultural lands is crucial to maintaining their covered by FAO, t = 1995 to 2020 , or the latest year available, economic contributions while safeguarding the planet’s long- and at is defined by the United States’ International term health. Agricultural Productivity database.195 The rental rate is proxied by land cost shares provided by the International This chapter first describes the main data sources and Agricultural Productivity database for each country and each methods used to estimate the value of agricultural land and decade. For countries where rental rates vary across decades, forests. Next, it presents the main trends in agricultural land annual values are assumed to be constant within each (cropland and pastureland) and forests (timber and non- decade. Because rental rates are not disaggregated for each wood forest ecosystem services) from 1995 to 2020, which crop/livestock product, k, FAO’s gross value of production are disaggregated by region and income group. Additionally, generated in cropland and pastureland is used. for the first time, this edition of CWON estimates the value of forest ecosystem services provided by protected areas. The value of agricultural land is then calculated as the discounted total rents, TRt, where r is the social discount rate MEASURING LAND ASSETS (assumed to be equal to 4 percent). Agricultural land 3) Agricultural land constitutes a considerable portion of total wealth in developing countries, particularly in the low- In this edition of CWON, the estimates of wealth in agricultural income group. For this report’s purposes, agricultural land lands do not consider projections of land degradation or includes cropland and pastureland. There are potentially climate change on land productivity, as was done in World two different methods for estimating land wealth. The first Bank (2021a). Instead, future rents are assumed to be constant method uses information from the sale of land. The second at 2020 values. This assumption is consistently applied across method uses information on the annual flow of rents that all assets. 195 https://www.ers.usda.gov/data-products/international-agricultural-productivity/. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 144 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FORESTS of non-timber forest ecosystem services values, provides a time series of ecosystem services values from 1995 to The value of forests is estimated for two asset categories: 2020, and develops a method to estimate the contribution timber resources and non-timber forest ecosystem services. of protected areas to the production of non-timber forest Timber resources are valued according to the present ecosystem services. This is an important departure from discounted value of rents from the production of roundwood the lower-bound approach used in CWON 2021 and earlier over the expected lifetime of standing timber resources. This reports to estimate wealth in protected areas, which relied value, Vt , is given by the following equation: on opportunity cost values. Key datasets used to estimate non-timber forest ecosystem 4) services are data on total forest area from FAO’s Forest Resources Assessment 2020, annual service values per where Rt is the rent for year t; r is the social discount rate hectare of forest—estimated by Siikamäki et al. (2024) as (assumed to be equal to 4 percent), and T is the lifetime of the sum of recreational, non-wood forest products, and timber resources capped at 100 years. Unlike metals and water services—and protected area boundaries from the minerals, timber is a renewable resource, so T depends on World Database on Protected Areas developed by the UN the rate of timber extraction relative to natural rates of forest Environment Programme (UNEP), World Conservation growth and resource replacement. Rents from timber in year Monitoring Center, and IUCN. Siikamäki et al. (2024) analyzed t are calculated as: hundreds of studies of non-wood forest benefits to develop a spatially explicit meta-regression model that predicts service 5) values for 10km x 10km plots of forest around the globe. The annual value of non-timber forest ecosystem services is where πt denotes unit rents, equal to revenues less production estimated by multiplying total forest area by the sum of the costs, and Qt denotes the quantity of roundwood extracted. per hectare monetary values for the three benefit categories. Data on annual roundwood production are obtained from The capitalized value of ecosystem services is equal to the FAO’s FAOSTAT database. The area of timber forest is estimated present value of annual services. The present value of non- by subtracting forests located within protected areas from timber services is given by the following equation: the total forest area, excluding protected area categories that could be used for sustainable timber production (that 6) is, protected areas in International Union for Conservation of Nature (IUCN) categories V and VI) based on FAO’s Forest Resource Assessment 2020. where S is the sum of per hectare service values for the three benefit categories, F is the total forest area, and r is the social The value of non-timber forest ecosystem services is based discount rate of 4 percent. No distinction is made between on the work of Siikamäki et al. (2024), who developed a meta- natural and planted forest. Values are estimated for the analytic predictive model using regression and machine current forest area, assuming no change in forest cover in learning techniques to spatially estimate the value of the the future. following three ecosystem services: (i) recreation, hunting, and fishing (referred to as “recreational”), (ii) non-wood The value of forest carbon retention services is not estimated forest products, and (iii) watershed protection (referred to for this edition of CWON 2024 due to data and conceptual as “water services”). The 2024 study, which builds on the challenges outlined in Box 7.2, and thus not included as part analysis conducted for CWON 2021, updates the database of the wealth accounts. 145 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH LAND ASSETS: GLOBAL AND REGIONAL is agricultural lands, which represented 80 percent of land TRENDS, 1995–2020 wealth globally in 2020, followed by forest water services (8 percent); timber (6 percent); forest recreation, hunting, and Global trends fishing services (4 percent); and non-wood forest services (2 Land assets (agricultural lands and forests) globally represent percent) (Figure 7.1). At the global level, total wealth in land a relatively small share of total wealth, between 4 percent and assets in per capita terms decreased across all categories: 5 percent for 1995–2020, but this share varies significantly agricultural lands (24 percent decline), forest recreation across regions and income groups. In Sub-Saharan Africa services (18 percent decline), non-wood forest products and South Asia land assets represent the largest shares, (27 percent decline), forest water services (26 percent about 30 percent and 32 percent in 2020, respectively. decline), and timber (28 percent decline). Any growth Agricultural lands and forest wealth represent less than 10 observed in total values of land assets did not keep up with percent in other regions. The largest category of land assets population growth (Figure 7.2). FIGURE 7.1 Global wealth in land assets: percent shares by category in 2020  Agricultural lands 8% 4%  Timber 6% 2%  Forest recreation, hunting, and fishing services  Non-wood forest products  Forest water services 80% Source: World Bank staff estimates. Note: The same percentages are observed in 1995. Shares in wealth are computed using wealth measured in current US dollars. FIGURE 7.2 Trends in real wealth per capita in land assets, indexed to 1995, 1995–2020 105 100 95 90 Index 85 80 75 70 1995 1996 1997 1998 1999 2000 2001 2003 2006 2016 2002 2004 2005 2007 2008 2009 2010 2011 2012 2013 2014 2015 2017 2018 2019 2020  Agricultural lands  Non-wood forest products  Timber  Forest water services  Forest recreation, hunting, and fishing services Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 146 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH REGIONAL TRENDS decline), followed by the Middle East and North Africa and Latin America and the Caribbean (between 30 percent and Wealth per capita in land assets experienced a decline in all 40 percent decline in all assets). Only Europe and Central asset categories for most regions. Sub-Saharan Africa shows Asia show a small increase in per capita wealth in forest the greatest declines, particularly in forest water services, recreation services (Figure 7.3). timber, and non-wood forest products (over 50 percent FIGURE 7.3 Change in wealth per capita by land asset, 1995–2020 10% 0% -10% -20% -30% -40% -50% -60% East Asia & Europe & Latin America Middle East & North America South Asia Sub-Saharan Pacific Central Asia & Caribbean North Africa Africa  Agricultural lands  Non-wood forest products  Timber  Water services  Forest recreation, hunting, and fishing services Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. Figure 7.4 shows the combined average per hectare value in low- and middle-income countries also show high values, 2020 of forest ecosystem services for recreational services including parts of Latin America and the Caribbean, and (including hunting and fishing), non-wood forest products, several areas in Asia, such as the densely populated areas and water services. The value is mapped globally in 0.1º by in Indonesia and China. Africa, in general, has relatively 0.1º resolution for forested grid cells based on the approach low predicted values, but some areas in countries like South implemented in Siikamäki et al. (2024). North America, Africa, Côte d’Ivoire, Ghana, and Nigeria show relatively Europe, and countries such as Japan and Australia feature high values. particularly high value estimates. On the other hand, several 147 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 7.4 Combined predicted value of all recreational services, non-wood forest products, and water services, per hectare in 2020 (2020 US dollars) Source: Siikamäki et al. 2024. Note: The services considered include recreational services (including hunting and fishing), non-wood forest products, and water services. For illustration purposes, only grid cells with forest cover larger than 10 percent are shown. In 2020, water protection services represented the greatest water protection services is greater than 70 percent in Latin share of non-timber forest ecosystem services in all regions America and the Caribbean, North America, and South Asia except the Middle East and North Africa. The share of (Figure 7.5). FIGURE 7.5 Share of forest ecosystem services by region and asset category, 2020 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% East Asia & Europe & Latin America Middle East & North South Asia Sub-Saharan Pacific Central Asia & Caribbean North Africa America Africa  Recreation, hunting, and fishing  Non-wood forest products  Water protection Source: World Bank staff estimates. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 148 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Per capita values of non-timber forest ecosystem services Saharan Africa, South Asia, and the Middle East and North vary greatly across regions and services (Figure 7.6). Water Africa show relatively low per capita values for the three non- services show the highest value per capita in all regions, timber forest ecosystem services, due to low-to-medium per except the Middle East and North Africa. North America hectare values, as well as large populations, particularly in and Latin America and the Caribbean show the highest per South Asia. capita values for water services relative to other regions. Sub- FIGURE 7.6 Wealth per capita in non-timber forest ecosystem services, by region and by category of service, 2020 2,500 2019 chained USD per capita 2,000 1,500 1,000 500 0 East Asia & Europe & Latin America Middle East & North America South Asia Sub-Saharan Pacific Central Asia & Caribbean North Africa Africa  Recreation, hunting, and fishing  Non-wood forest products  Water protection Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. Protected areas cover around 15 percent of the world’s Although various ecological and ecosystem contributions terrestrial land and some 7 percent of the ocean (UNEP- of protected areas are widely recognized (King et al. 2023), WCMC and IUCN 2022). 196 Their coverage has steadily information on the economic value of ecosystem services increased over time, and their current global extent provided by protected areas is not comprehensively available. nearly matches Aichi Biodiversity Target 11 of 17 percent Siikamäki et al. (2024) predicted the contribution of protected of terrestrial and 10 percent of coastal and marine areas areas for non-wood forest products, water services, and protected by 2020. As such, protected areas constitute a recreation, hunting, and fishing. These results are the first key policy instrument to ensure the conservation of nature available estimates of the economic value of these ecosystem to support biodiversity and the provision of ecosystem service categories supported by protected areas, generated services to people. by country. 196 Based on Wolf et al. (2021), 15.7 percent of forests are formally protected. 149 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 7.7 Share of national forest wealth in non-timber forest ecosystem services provided by protected areas and country, 2020 Source: Figure 6.1 in Siikamäki et al. 2024. Note: The following non-timber forest ecosystem services are included in the estimation: recreation, hunting, and fishing; non-wood forest products; and water services. Figure 7.7 shows the percentage of the total national wealth of Protected areas provide 16 percent of the total wealth forest non-timber ecosystem services provided by protected contributed by forest non-wood ecosystem services. The share area forests, by country. The contribution of protected of national forest non-wood wealth provided by protected areas to total value is highest in several African countries, areas varies considerably by world region, with the highest Venezuela, Nepal, Uzbekistan, and many countries in Central percentage of wealth provided by protected areas in Sub- Europe. This is a combination of the large share of forest Saharan Africa (22 percent), Latin America and the Caribbean being protected in those countries and the value of non- (22 percent), and South Asia (22 percent), and the lowest timber ecosystem services in the protected areas relative to percentage of wealth provided by protected areas occurring non-protected areas. in the Middle East and North Africa (9 percent), Europe and Central Asia (11 percent), and North America (11 percent). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 150 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 7.1 TRENDS IN LAND ASSETS: COUNTRY EXAMPLES T aking a closer look at trends in land assets for selected countries helps us understand the underlying drivers of change in wealth per capita. Here the cases of Brazil and Ethiopia, large economies with significant land assets and populations, are presented. Brazil has the largest forest area in Latin America and the Caribbean. It covers a large fraction of the Amazon biome, a critical natural asset for the provision of global ecosystem services and biodiversity protection (Brouwer et al. 2022). Between 1995 and 2020, the country’s real wealth in agricultural lands increased by 4 percent, while its real wealth in non-timber forest ecosystem services declined by 13 percent. Over the same period, the area covered by these two asset categories followed a similar trend. Yet the country’s population increased by 30 percent over the same period, leading to a decline in wealth per capita in agricultural lands of 20 percent and non-timber forest ecosystem services of 33 percent. Overall, the country shows an increase in real per capita wealth of 73 percent once all assets (produced capital, human capital, renewable natural capital, and nonrenewable natural capital) are accounted for, which is a minimum requirement for sustainable growth (under weak sustainability). Further exploration of the limited substitutability of renewable natural capital and its implications for future wealth is needed, considering the risks of reaching critical tipping points in ecosystem functions (Flores et al. 2024). Ethiopia, the second most populous country in Africa, has an economy largely dependent on agriculture and land assets (He and Chen 2022). The country’s long history of forest and land degradation represents a significant drag on rural growth and poverty reduction (UNEP 2016). Between 1995 and 2020, the country’s total real wealth in agricultural lands increased by 26 percent, while real wealth in non-timber forest ecosystem services declined by 10 percent, underpinned by a loss in forest cover of a similar magnitude. But because the population doubled over the same period, per capita real wealth in agricultural lands and non-timber forest ecosystem services declined by 38 percent and 56 percent, respectively. Investments in assets other than those in the land sector barely compensated for this rapid population growth, as real per capita wealth considering all assets increased by 5 percent between 1995 and 2020. For a country where land assets represented close to 30 percent of total wealth in 2020, declines in critical sectors like agriculture and the services provided by forests indicate that much greater investments are needed in natural capital and the subset of land assets. 151 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TRENDS BY INCOME GROUPS share of land assets in total wealth fell in most low-income countries between 1995 and 2020, with a few exceptions like Although all income groups experienced declines in per Niger, Guinea-Bissau, and the Central African Republic, capita wealth across land asset categories, this was a which show the opposite trend (Figure 7.9). particular issue in low-income and lower-middle-income countries. Timber wealth and non-timber forest ecosystem In low-income countries as a group, agricultural area services fell by more than 50 percent in low-income countries increased by 4 percent, while the area of forests decreased and by more than 30 percent in lower-middle-income by 5 percent (Table 7.1). Yet, as the trends of wealth per countries (Figure 7.8). High-income countries show a decline capita in land assets demonstrate, this has not increased in per capita agricultural wealth of 27 percent, greater than overall wealth per capita in the land sector. Countries in this the declines observed in the forest categories. group appear to be trading off forests for land in agriculture. However, this cannot be tested directly using the CWON In 2020 land assets made up 41 percent of total wealth in data; this observation requires additional analysis of remote- low-income countries, down from 78 percent in 1995. The sensing data. FIGURE 7.8 Change in wealth per capita in land assets by income group, 1995–2020 10% 0% -10% -20% -30% -40% -50% -60% High income Low income Lower middle income Upper middle income  Agricultural lands  Non-wood forest products  Timber  Forest water services  Forest recreation, hunting, and fishing services Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 152 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 7.9 Share of land wealth in total wealth in low-income countries, 1995 and 2020 Uganda Togo Chad Sierra Leone Rwanda Niger Malawi Mozambique Mali Madagascar Liberia Haiti Guinea-Bissau The Gambia Guinea Ethiopia Central African Republic Burkina Faso Burundi 0% 20% 40% 60% 80% 100% 120%  1995  2020 Source: World Bank staff estimates. Note: Shares in wealth are computed using wealth measured in current US dollars. 153 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 7.1 Land cover for forest and agricultural lands in low-income countries, 1995 and 2020 AGRICULTURAL LAND (SQ KM) FOREST LAND (SQ KM) 1995 2020 % change 1995 2020 % change East Asia and the Pacific 26,500 25,900 -2% 66,834 60,301 -10% Europe and Central Asia 45,820 49,160 7% 4,090 4,238 4% Latin America and the Caribbean 15,900 18,400 16% 3,818 3,473 -9% Middle East and North Africa 375,250 373,730 0% 9,511 10,711 13% South Asia 377,530 383,560 2% 12,084 12,084 0% Sub-Saharan Africa 4,870,350 5,085,426 4% 3,124,240 2,970,657 -5% Total low-income countries 5,711,350 5,936,176 4% 3,220,576 3,061,464 -5% Source: World Bank staff estimates based on data from FAO for agricultural land and forest land and data from the World Development Indicators for protected areas. Note: sq km = square kilometers. CONCLUSIONS on cost-effectiveness criteria, but the low-income group requires stronger forest governance and land tenure security Although total global wealth in land assets (the sum of to attract greater public and private investments in this sector agricultural lands, timber, and non-wood forest ecosystem (World Bank 2024). services) has increased slightly, wealth per capita in land has declined across all regions and for almost all assets. This This chapter does not discuss the interconnections between trend is particularly pronounced in low-income countries as land assets. Further exploration is required to understand a group and countries in Sub-Saharan Africa. the degree to which countries and regions are trading off forest lands for agricultural lands, and the implications for Land assets continue to be a critical component of wealth changes in total wealth per capita. One way to address this in low- and middle-income countries. While this chapter is to move to fully spatially integrated land accounts in the considered only a subset of forest ecosystem services, next CWON edition. More research is also needed to assess most countries, particularly the low-income group, appear carbon retention services as a critical climate regulation to be managing their land assets unsustainably. Greater service provided by forests, as explained in Box 7.2, and to investments and stronger policy signals are urgently needed consider additional data collection and improvements to the to protect and restore forests and reverse long-term trends valuation methodology for agricultural lands and timber to of nature loss and degradation. Low- and middle-income further improve the rent estimation. countries have large amounts of restorable land based REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 154 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 7.2 KEY CHALLENGES IN ACCOUNTING FOR CARBON RETENTION SERVICES IN CWON T he global climate crisis is one of the most pressing sustainability challenges of our time, exacerbating existing social, environmental, and economic issues (IPCC 2022). Terrestrial ecosystems, most notably forests, can play an important role in mitigation efforts (for example, Grassi et al. 2017; Griscom et al. 2017), as they are able to both emit and absorb greenhouse gases, acting as significant stores of carbon as well as carbon sinks (for example, Cook-Patton et al. 2020; Jones et al. 2013). For these reasons, it is important to evaluate to what extent the value of climate regulation services could be included in CWON. The current wealth estimates already include several ecosystem services provided by forests (as presented in this chapter) and mangroves (see chapter 8). Adding climate regulation services would thus be a natural extension of the asset boundary. There is emerging guidance from the System of Environmental-Economic Accounting Ecosystem Accounting (SEEA-EA) on how to measure global climate regulation from terrestrial ecosystems, which can help guide how such a service could be included in CWON (Edens and Caparrós 2023). The current consensus is to measure a single service, consisting of two components: carbon sequestration, which measures the net uptake of carbon by a given ecosystem asset, such as a forest or wetland, and carbon retention, which measures the avoided release of carbon—or put differently, the ability of the ecosystem to retain carbon and thus avoid climate damages. Users are advised to choose the measurement most suitable for their context (NCAVES and MAIA 2022). For a global assessment like CWON, carbon retention is the primary component of climate regulation services, as carbon stocks in forests and wetlands are typically not increasing for most countries (Bulckaen et al. 2024). If a country’s forests are instead experiencing a clear expansion in their carbon stock, carbon sequestration should ideally be reported as an “of which” item. Unfortunately, this is currently not feasible on the global scale (Pugh et al. 2020). Any changes in carbon retention services thus capture changes in carbon stored due to anthropogenic and other factors, as well as sequestration. In addition, carbon retention is the more policy relevant measure, as a loss in carbon due to, for example, deforestation would lower carbon retention services. Similarly, ecosystems with high carbon stocks, such as tropical rainforests, would be assigned higher retention values, signaling the importance of conserving. This is not necessarily the case for carbon sequestration, which only captures changes in the carbon stock, not its level. This can lead to perverse incentives in natural resource management, where carbon sequestration would increase if a rainforest were replaced by, for example, fast-growing bamboo. 155 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Measuring carbon retention in biophysical terms is possible by estimating vegetation carbon stocks annually using the standard methodology developed by Gibbs and Ruesch (2008) and 2006 IPCC default factors (Bulckaen et al. 2024). To estimate vegetation carbon stocks, it is assumed that each terrestrial land cover class, with a few exceptions like water bodies, glaciers, and bare rock or soil, contributes to storing carbon to varying extents. The classes with the highest contribution to carbon stock are forests and wetlands, particularly mangroves, with each contributing proportionally to the total carbon stock of a given area. These estimates can then provide annual snapshots of the stock of carbon for each year, which will change due to land cover changes (and potentially land cover reclassification) as well as the occurrence of fire and anthropogenic factors. Estimates show that the countries with the highest share in the global value of forest carbon are tropical or large countries, such as Brazil, the Democratic Republic of Congo, Russia, Indonesia, the United States, and Canada (Box Figure 7.2.1). Brazil, being both vast and tropical, accounts for the largest share in the total forest carbon value with 18 percent. The other large tropical countries—the Democratic Republic of Congo and Indonesia—make up around 7 percent and 6 percent, respectively. Combined with the three large land area countries—Russia, the United States, and Canada—these six countries make up nearly half of the global value of climate regulation services provided by forests, underscoring the importance of sustainable forest management in these countries. BOX FIGURE 7.2.1 Total global vegetation carbon stock (in megatons), 2020 8,000 6,000 4,000 2,000 0 Source: Bulckaen et al. 2024. Note: All the code to obtain the results is available at https://github.com/integratedmodelling/im.nca.postprocessing/tree/main/aggregation_region. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 156 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH The primary compilation challenge for measuring carbon retention services is to choose a suitable carbon price for the valuation of avoided climate damages. The SEEA-EA does not provide specific guidance (UN et al. 2021, paragraph 9.32), but Edens et al. (2019) and the NCAVES and MAIA (2022) discuss the suitability of using different carbon prices, such as observed market prices, the marginal abatement costs of carbon, or the social cost of carbon (SCC) in the context of national accounting. As the carbon retention framing is based on the idea of avoided damages, the recommendation is to apply a SCC estimate. While market prices would be preferable as they align with the exchange value principle of the SEEA-EA, existing carbon markets are incomplete, making them unsuitable for a global assessment. Marginal abatement cost curves exist at the global scale, but measurement concerns over underestimating costs and double counting limit their use in national accounting. An additional complication in the context of CWON is that the wealth estimates include a range of assets, such as produced capital, agricultural land, and human capital, whose current value is directly supported by the carbon retention services provided by forests and mangroves. That is, their asset values at least to some extent capture the fact that the climate is currently stable, and that productivity is not lost due to climate damages. Adding the avoided climate damages provided by carbon retention services would result in double counting if the SCC estimate captured market benefits (to produced capital, agricultural land, and human capital, among others). To avoid potential double counting, a tailor- made SCC was estimated by Drupp and Hänsel (2023), which only captures the avoided monetized climate damages for the non-market sectors of the global economy. Based on these estimates, the global value of carbon retention services has declined by 1 percent over the last two decades. This change was driven by a dramatic loss in global forest cover equivalent to nearly 36 million hectares, an area the size of Japan or Norway. However, several conceptual challenges around the valuation of carbon retention services still need to be resolved. First, more work is needed to inform the choice of carbon price to be used in the valuation. Second, additional theoretical and empirical research is required to assess the risk of double counting when including climate regulation (as well as other regulating services) in wealth accounts. While there is evidence that climate regulation services are partially captured in the valuation of other assets, more analysis is needed to systematically estimate this share and develop an approach to reattribute this value to ecosystem assets. Third, at this point, it is not feasible to derive country-specific estimates of the value of carbon retention services. To produce country-specific non-market SCC estimates, more refined data is required on the breakdown and distribution of climate damages into market and non-market damages at the country level, which also distinguish between use and non-use value components. It will also be critical to assess which part of the market SCC should capture to further lower the risk of double counting. 157 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH REFERENCES Brouwer, R., Pinto, R., Dugstad, A., and Navrud, S. 2022. “The Economic Value Griscom, B.W., Adams, J., Ellis, P.W., Houghton, R.A., Lomax, G., Miteva, D.A., of the Brazilian Amazon Rainforest Ecosystem Services: A Meta-Analysis ... and Fargione, J. 2017. “Natural Climate Solutions.” Proceedings of the of the Brazilian Literature.” PLoS One 17 (5): e0268425. https://doi:10.1371/ National Academy of Sciences 114 (44): 11645–11650. journal.pone.0268425. He, X., and Chen, Z. 2022. “Weather, Cropland Expansion, and Deforestation Bulckaen, A., Abad Viñas, R., Bengochea Paz, B., Crespo, R., and Villa, in Ethiopia.” Journal of Environmental Economics and Management 111: F. 2023. Global Estimates of Carbon Stocks in the Vegetation and Soils of 102586. https://doi.org/10.1016/j.jeem.2021.102586. Terrestrial Ecosystems. Technical report. Iizumi, T., Furuya, J., Shen, Z., Kim, W., Okada, M., Fujimori, S., Hasegawa, Cook-Patton, S.C., Leavitt, S.M., Gibbs, D., Harris, N.L., Lister, K., Anderson- T., and Nishimori, M. 2017. “Responses of Crop Yield Growth to Global Teixeira, K.J., Briggs, R.D., Chazdon, R.L., Crowther, T.W., Ellis, P.W., and Temperature and Socioeconomic Changes.” Scientific Reports 7 (1): 7800. Griscom, B.W. 2020. “Mapping Carbon Accumulation Potential from Global https://doi.org/10.1038/s41598-017-08214-4. Natural Forest Regrowth.” Nature 585 (7826): 545–550. IPBES. 2019. Global Assessment Report on Biodiversity and Ecosystem Services Dasgupta, P. 2021. The Economics of Biodiversity: The Dasgupta Review. HM of the Intergovernmental Science-Policy Platform on Biodiversity and Treasury. Ecosystem Services. Drupp, M.A., and Hänsel, M.C. 2023. The Non-market Social Costs of Carbon for IPCC (Intergovernmental Panel on Climate Change). 2006. 2006 IPCC the Changing Wealth of Nations. Technical report. Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse Gas Inventories Programme, edited by H.S. Eggleston, L. Edens, B., and Caparros, A. 2023. The Measurement of Global Climate Regulation Buendia, K. Miwa, T. Ngara, and K. Tanabe. Japan: IGES. Service in SEEA EA – Suggestions for CWON. Technical report. IPCC (Intergovernmental Panel on Climate Change). 2019. An IPCC Special Edens, B., Elsasser, P., and Ivanov, E. 2019. “Discussion Paper 6: Defining and Report on Climate Change, Desertification, Land Degradation, Sustainable Valuing Carbon Related Services in the SEEA EEA.” Paper submitted to the Land Management, Food Security, and Greenhouse Gas Fluxes in Terrestrial Expert Meeting on Advancing the Measurement of Ecosystem Services Ecosystems (SRCCL) (World Meteorological Organization). for Ecosystem Accounting, New York, January 22–24, and subsequently revised. Version of March 15, 2019. https://seea.un.org/sites/seea.un.org/ IPCC (Intergovernmental Panel on Climate Change). 2023. Climate Change files/discussion_paper_6_-_valuing_carbon_final.pdf. 2023: Synthesis Report. Contribution of Working Groups I, II, and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate FAO (Food and Agriculture Organization). 2012. The Forest Resources Assessment Change. Geneva, Switzerland: IPCC. https:// doi:10.59327/IPCC/AR6- Programme: Terms and Definitions. Rome: FAO. http://www.fao.org/ 9789291691647. docrep/017/ap862e/ap862e00.pdf. Johnson, J.A., Ruta, G., Baldos, U., Cervigni, R., Chonabayashi, S., Corong, E., FAO (Food and Agriculture Organization). 2022. State of the World’s Forests. Gavryliuk, O., et al. 2021. The Economic Case for Nature: A Global Earth- Economy Model to Assess Development Policy Pathways. Washington, DC: FAO (Food and Agriculture Organization). 2023. Estimating Global and Country- World Bank. Level Employment in Agrifood Systems. Jones, C., Robertson, E., Arora, V., Friedlingstein, P., Shevliakova, E., Bopp, Flores, B.M., Montoya, E., Sakschewski, B., et al. 2024. “Critical Transitions in L., Brovkin, V., Hajima, T., Kato, E., Kawamiya, M., and Tjiputra, J. 2013. the Amazon Forest System.” Nature 626: 555–564. https://doi.org/10.1038/ “Twenty-First-Century Compatible CO2 Emissions and Airborne Fraction s41586-023-06970-0. Simulated by CMIP5 Earth System Models under Four Representative Concentration Pathways.” Journal of Climate 26 (13): 4398–4413. Gibb, R., Redding, D.W., Chin, K.Q., Donnelly, C.A., Blackburn, T.M., Newbold, T., and Jones, K.E. 2020. “Zoonotic Host Diversity Increases Miller, D.C., Ordonez, P.J., Brown, S.E., Forrest, S., Nava, N.J., Hughes, K., in Human-Dominated Ecosystems.” Nature 584: 398–402. https://doi. et al. 2020. “The Impact of Agroforestry on Agricultural Productivity, org/10.1038/s41586-020-2562-8. Ecosystem Services, and Human Well-Being in Low- and Middle-Income Countries: An Evidence and Gap Map.” Campbell Systematic Reviews 16: Gibbs, H.K., and Ruesch, A. 2008. New IPCC Tier-1 Global Biomass Carbon Map e1066. https:// doi:10.1002/cl2.1066. for the Year 2000. United States. https://doi.org/10.15485/1463800. NCAVES (Natural Capital Accounting and Valuation of Ecosystem Services) and Grassi, G., House, J., Dentener, F., Federici, S., Den Elzen, M., and Penman, MAIA (Mapping and Assessment for Integrated Ecosystem Accounting). J. 2017. “The Key Role of Forests in Meeting Climate Targets Requires 2022. Monetary Valuation of Ecosystem Services and Ecosystem Assets for Science for Credible Mitigation.” Nature Climate Change 7 (3): 220–226. Ecosystem Accounting. Interim version, 1st edition. New York: United Nations Department of Economic and Social Affairs, Statistics Division. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 158 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Pugh, T.A., Rademacher, T., Shafer, S.L., Steinkamp, J., Barichivich, J., UN (United Nations), EU (European Union), FAO (Food and Agriculture Beckage, B., ... and Thonicke, K. 2020. “Understanding the Uncertainty in Organization of the United Nations), IMF (International Monetary Global Forest Carbon Turnover.” Biogeosciences 17 (15): 3961–3989. Fund), OECD (Organisation for Economic Co-operation and Development), and World Bank. 2021. System of Environmental-Economic Ray, D.K., West, P.C., Clark, M., Gerber, J.S., Prishchepov, A.V., and Chatterjee, Accounting—Ecosystem Accounting (SEEA EA). New York: White cover S. 2019. “Climate Change Has Likely Already Affected Global Food publication, pre-edited text subject to official editing. Production.” PLOS One 14 (5): e0217148. https://doi:10.1371/journal. https://seea.un.org/ecosystem-accounting. pone.0217148. UNEP. 2016. The Contribution of Forests to National Income in Ethiopia and Reyer, C.P.O., Rigaud, K.K., Fernandes, E., Hare, W., Serdeczny, O., and Linkages with REDD+. United Nations Environment Programme: Nairobi. Schellnhuber, H.J. 2017. “Turn Down the Heat: Regional Climate Change Impacts on Development.” Regional Environmental Change 17: 1563–1568. Wolf, C., Levi, T., Ripple, W.J., Zárrate-Charry, D.A., and Betts, M.G. 2021. “A https://doi:10.1007/s10113-017-1187-4. Forest Loss Report Card for the World’s Protected Areas.” Nature Ecology & Evolution 5 (4): 520–529. Siikamäki, J., Piaggio, M., Da Silva, N., Álvarez, I., and Chu, Z. 2021. “Global Assessment of Non-wood Forest Ecosystem Services: A Revision of a World Bank. 2021a. “Changing Wealth of Nations: Managing Assets for the Spatially Explicit Meta-Analysis and Benefit Transfer.” World Bank, Future.” World Bank, Washington, DC. Washington, DC. World Bank. 2021b. “Unlocking Nature-Smart Development: An Siikamäki, J., Piaggio, M., Da Silva, N., Álvarez, I., and Chu, Z. 2024. “Global Approach Paper on Biodiversity and Ecosystem Services.” World Bank, Assessment of the Economic Value of Non-wood Forest Ecosystem Washington, DC. Services.” World Bank, Washington, DC. World Bank. 2024. “Cost-Effective Forest Restoration in Low- and Middle- Siikamäki, J., Santiago-Ávila, F.J., and Vail, P. 2015. “Global Assessment of Income Countries.” World Bank, Washington, DC. Forthcoming. Non-wood Forest Ecosystem Services: Spatially Explicit Meta-Analysis and Benefit Transfer to Improve the World Bank’s Forest Wealth Database.” Zabel, F., Delzeit, R., Schneider, J., et al. 2019. “Global Impacts of Future World Bank, Washington, DC. https://www.wavespartnership.org/en/ Cropland Expansion and Intensification on Agricultural Markets and knowledge-center/global-assessment-non-wood-forest-ecosystem-services- Biodiversity.” Nature Communications 10: 2844. https://doi.org/10.1038/ spatially-explicit-meta. s41467-019-10775-z. 159 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH 8 Blue Natural Capital: Marine Fish Stocks, Aquaculture, and Mangrove Coastal Protection Services MAIN MESSAGES INTRODUCTION ■ Marine fish stocks play a key role in supporting local Blue natural capital is critical to coastal nations around the economies and food security. However, they have been world. Two of the most critical services provided by coastal overexploited over the last 25 years, leading to dramatic and ocean ecosystems are food production and coastal storm declines in the real asset value of marine fish stocks. The protection, which support lives and livelihoods. There are real value of marine fish stocks has dropped by more than two marine systems for which there are sufficient data to 25 percent since 1995, equivalent to a $70 billion assess trends in national wealth over time: marine decline in chained 2019 US dollars. fish stocks and mangroves. The previous CWON report considered marine fish stocks ■ Mangroves provide a range of ecosystem and mangroves for the first time (World services, one of the most critical being Bank 2021). These analyses are advanced that they protect people and assets along here by producing real wealth estimates the coastline. While the nominal value for both assets using a new methodology of these protection benefits increased that accounts for both changes in the dramatically between 1995 and 2020, in physical quantities—for example, the total real terms, mangrove wealth still declined weight of fish of a specific taxon (species) in due to the continued loss in mangrove cover. a fishery or the hectares of mangrove forest—and ■ More spatially disaggregated and detailed data are needed real price changes (for more detail, see chapter 2). This has to further improve our measurement of these critical significant implications for the real wealth estimates, which assets and help inform management decisions and policy will now directly capture any degradation in blue natural choices aimed at restoring and sustainably managing blue capital. The potential to include aquaculture in future CWON natural capital. reports is also assessed. ■ Aquaculture assets have not yet been included in CWON, Marine fish stocks and fisheries represent an important but they could be, using a resource rent approach. component of renewable natural capital in coastal Asset values from aquaculture are likely to be positive, nations, especially in small island states and places where significant, and growing as production continues to communities rely on healthy oceans for food and livelihoods expand. However, there are major data constraints as (Stuchtey et al. 2023; World Bank 2021). Globally, 1 out of 10 the necessary information on operating and capital people rely on fisheries and aquaculture for their livelihoods costs is rarely available. This could be addressed by (FAO 2016). However, blue natural capital linked to fisheries the establishment of a systematic and integrated data is deteriorating around the world due to multiple factors, but collection system consistent with the System of National the one affecting the wealth of marine fish stocks the most is Accounts (SNA) and the System of Environmental- overexploitation (World Bank 2021; Srinivasan et al. 2010). As Economic Accounting (SEEA) methodologies, and a renewable natural resource, the asset lifetime and stream supported by an aquaculture economics community of future benefits from marine fish stocks are potentially of practice comprising national statisticians, technical infinite if the fish stock is allowed to regenerate naturally specialists, economists, and regulatory authorities. (Sumaila 2021). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 160 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH However, if marine fish stocks are overexploited beyond productive tank or cage-based systems where species such as natural regeneration, the future stream of economic benefits salmon or marine finfish must be fed protein- and energy-rich and resource rent will decline over time as if these were a diets, to pond-based systems growing carps or tilapia where nonrenewable natural resource (Lam and Sumaila 2021; World natural food comprises a high proportion of the nutritional Bank 2021). requirements, as well as unfed systems for filter-feeding mollusks and seaweeds. Consequently, data requirements Part of the solutions for restoring the health of ocean are very high, as detailed data are needed on each of these ecosystems and preventing this resource degradation is to production systems, which vary across countries. For account for the values of marine fish stocks in national wealth example, aquaculture in most Asian countries includes a accounts. Despite their importance, marine fish stocks are diverse range of species groups, whereas countries such as generally measured and assessed based on data that are often Chile, Norway, Ecuador, and Egypt are more specialized, incomplete, inaccurate, or limited. This type of blue natural growing relatively few species. capital includes the value of these assets in the broader macroeconomic framework, which can help decision-makers This chapter discusses potential methodologies for including understand its contribution to wealth and assess the risks of aquaculture in CWON and the likely scale that aquaculture overexploiting marine fish stocks. might comprise in blue natural capital accounts. Including aquaculture assets in CWON is not challenging from a In contrast to stagnating catches from marine capture conceptual point because a residual value method (RVM)/ fisheries, global aquaculture production has grown rapidly net present value (NPV) approach can be readily applied, over the last 25 years, reaching a total nominal value of $281.5 as recommended by the SNA and SEEA. However, there are billion in 2020 (FAO 2022). This included 87.5 million tonnes significant data constraints, as the required information of aquatic animals (such as fish, shrimp, and bivalve mollusks) on operating and capital costs is rarely reported in publicly and 35.1 million tonnes of algae. The growth of aquaculture available datasets. Most of the estimates for the pilot accounts has helped meet growing demand for fish, which has more presented in this chapter are thus based on research results than doubled from an average of 9.9kg per capita in the and information from key contacts. 1960s to 20.2kg per capita in 2020, with total aquatic animal production expected to grow by another 14 percent by 2030 Lastly, mangroves—and the ecosystem services they provide— (FAO 2022). However, despite being an increasingly important are of critical importance for biodiversity and coastal source of food, income, and employment in many parts of the communities at both local and global scales (Leal and Spalding world (Little, Newton, and Beveridge 2016), aquaculture is not 2022; Bunting et al. 2022; Kauffman and Donato 2012; McLeod yet included in CWON. et al. 2011). Globally, they sustain 4.1 million small-scale fishers (Ermgassen et al. 2021), buffer coastlines against storm surges, A key challenge is that aquaculture is a complex industry provide $65 billion per year in flood protection (Hagger et al. that encompasses a broad range of production systems. 2022), and mitigate climate change by storing an estimated 8.5 These include floating cage-based finfish aquaculture, gigatons of carbon (Hagger et al. 2022; Richards et al. 2020). pond and tank-based finfish and shrimp, as well as various These benefits in turn support coastal communities and boost bivalve shellfish and seaweed growing systems. In addition, economies (Spalding et al. 2014). the systems are situated in marine coastal waters, along coastlines, in river estuaries, and in inland freshwaters, This chapter first presents the methods and data sources used such as lakes and rivers, and may use water supplied by to produce value estimates for each asset. Next, it discusses agricultural irrigation systems. Aquaculture is also carried the respective findings and concludes with suggestions for out at a wide range of production intensities, from highly future work and further methodological improvements. 197 The industry is dominated by China, which represented 60.3 percent of global aquaculture value in 2020, followed by India, Vietnam, Indonesia, Chile, Norway, Bangladesh, Japan, Ecuador, Thailand, Egypt, and the Republic of Korea (FAO 2022). 161 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH ESTIMATING THE WEALTH OF vessel prices for each taxon and country and transformed to MARINE FISH STOCKS, AQUACULTURE, 2020 constant US dollars using price deflators. By combining the catch data with the ex-vessel price of each marine AND MANGROVE COASTLINE taxon, the landed values can be estimated for different PROTECTION SERVICES fishing countries at different spatial locations. FAO data Marine fish stocks were used wherever possible, with SAU catch data used to fill gaps. Finally, the FERU’s fishing cost data are arranged To estimate the value of marine fish stocks, the RVM/NPV by year, fishing entity, gear type, and fishing sectors. Similar approach is used, as recommended by the SNA and SEEA to most of the variables or indicators in the SAU database, (for more detail, see chapter 2). To compute the value of an fishing cost data points were obtained from gray literature, asset, such as marine fish stocks, the NPV of the stream of government and consultant reports, FAO, and other sources rents needs to be computed. As a first step, resource rent (Lam and Sumaila 2021). for marine fish stocks needs to be estimated, which requires information on fish stocks, catch volumes by species, Marine fish stocks might generate positive rents but be fishing costs, landed value, and price. Catch, landed value, managed unsustainability, reducing the time horizon over and fishing cost data are obtained from the Sea Around which the resource could generate rents. Rent from marine Us (SAU) reconstruction database,198 as well as the fishing fish stocks depends heavily on the abundance of fish biomass. cost database from the Fisheries Economics Research Essentially, without fish there will be no fisheries and no fish Unit (FERU) at the University of British Columbia. These jobs or fish dollars. If fishing efforts are at the maximum databases build on data from the Food and Agriculture sustained yield, fish stocks will be able to regenerate Organization of the United Nations (FAO) and include naturally, and future rent will not be compromised. estimates of catch and economic indicators not always However, if fishing efforts exceed the maximum sustained reported in the FAO database. SAU and FERU allow for more yield, marine fish stocks are transformed into a resource detailed and granular data by (i) disaggregating catch and much more akin to an exhaustible mineral resource, which landed value into four major categories, while FAO reports will be unable to regenerate and will decline over time only total catch; (ii) including spatialized catch and landed until collapse. For this report, to produce the core wealth value; and (iii) including the cost of fishing to correctly account for fisheries for all countries in the database, the estimate resource rent. economic value of fisheries in 2020 is assumed to be the net present value of rents over 100 years, assuming rents remain Catch data199 in weight are disaggregated into fishing constant at the 2020 rent. Scenario analysis could be applied entities by different taxa, fishing gear types, distant-water to estimate future rents based on the expected sustainability fleets, domestic fleets, catch types (landings and discards), of fisheries in each country or anticipated policy changes. and fishing sector (industrial, subsistence, artisanal, and However, for practicality and consistency with other recreational). Ex-vessel prices refer to the price fishers accounts such an analysis is not applied when computing receive directly for their catch when it enters the supply the real value of fish stocks. Future data users can then use chain (Sumaila et al. 2021). The ex-vessel prices data are the CWON data as inputs in their own scenario analysis as reported in current and constant US dollars for each they aim to answer specific policy questions. exploited marine taxon and country and can be matched to catch data to estimate the landed value. Ex-vessel prices Finally, to compute real wealth estimates of marine stocks are expanded from 2010 to 2018200 by assuming constant ex- using the new methodology, a measure of the fish stock 198 www.seaaroundus.org. 199 The analysis focuses on total catch by a given fishing country, which includes a country’s own exclusive economic zone as well as foreign exclusive economic zones if the country’s fishing fleet fishes in international waters. Total catch is inclusive of reported and unreported catches by the artisanal, subsistence, and industrial sectors intended for direct human consumption and for processing into fishmeal and fish oil. 200 The complete 1995–2018 series is extrapolated to estimate the values for 2019 and 2020. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 162 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH biomass or “volume” is needed to measure its status—for and not included in the FAO database. As explained in example, whether the fish stock is healthy, close to healthy, the technical report (Dickson et al. 2024), a pilot account or in a bad condition. Biomass estimates of major exploited approach was adopted where resource rents from individual species in a country’s exclusive economic zones were countries (Norway, Egypt, Ecuador, Japan, Bangladesh, and derived by the SAU initiative by applying a suite of methods Indonesia) and species (salmonids, tilapia, shrimp, carps, to reconstruct fisheries catches for the 1950–2018 period pangasius, yellowtail, seabream, milkfish, scallops, oysters, and complement them with stock assessments performed and seaweed) were estimated using available information. previously by others. The total volume of fish stocks from all Only Norway provided sufficient official data to estimate exclusive economic zones is then collapsed at the country resource rents over the period 1995 to 2020. In the other level and filtered by the certainty of the reported data, where pilot countries, estimates were based on research studies or only observations with reliability scores of 3 and 4 (meaning on contributions from organizations and consultants with higher data reliability) are used. extensive experience in those countries. Aquaculture Mangrove coastal protection services CWON asset account methodologies, where possible, should Mangroves provide coastal protection by reducing align with the internationally accepted statistical standards flooding and the resulting damage to produced capital and laid out in the 2008 SNA and its extension, the SEEA (United populations that would occur from storms if mangroves Nations 2008; United Nations et al. 2021). While aquaculture were absent (Figure 8.1). The “averted damage” valuation is briefly mentioned in the SNA as a type of production approach is widely used by economists and provides a alongside livestock, it is more specifically referenced at rigorous foundation for estimates of flood risk and habitat several points in the 2021 SEEA-EA publication, where benefits (Barbier 2015; Beck and Lange 2016; Pascal et al. it is included in the SEEA-EA reference list of selected 2016; Van Zanten, Van Beukering, and Wagtendonk 2014). ecosystem services as a biomass provisioning service. SEEA- It also provides an integrated quantitative framework with EA outlines a range of approaches that can be used to value process-based models and statistical tools consistent with ecosystem services including for agricultural production, national accounting to capture risk and national benefits such as RVM/NPV, where the direct operating and input (Figure 8.1). costs associated with producing an agricultural output (such as fuel, fertilizer, labor, and produced assets) must be The averted damages approach combines multiple deducted from the value of the output to isolate the value of commonly used models to assess storm hazards (waves and the ecosystem services. Depending on the scope of the data, surge), flooding impacts, and socioeconomic consequences. the estimated residual value provides a direct value that can The hazard and impacts are estimated using a combined set be recorded in ecosystem accounts. of process-based storm and hydrodynamic models, which are commonly used by engineers and risk modelers (Beck et al. While this indicates that it should be possible to use the 2024). These models identify the area and depth of flooding RVM/NPV to estimate the value of ecosystem services for under cyclonic and non-cyclonic storm conditions and aquaculture, it is an indirect approach and is prone to error can be used to conduct scenario analysis with and without if based on insufficient or inaccurate data (United Nations mangroves for different storm frequency events. Then, the et al. 2021). The main global database for aquaculture is flood maps are overlaid on population maps and produced the FAO system, which receives inputs from each country capital stock data to determine exposure. Lastly, the flood (FAO 2023). This provides information on aquaculture depth and exposure data are combined, considering depth production quantity and value in each country, broken damage functions, to compute the population and capital down by species and environment. An RVM/NPV approach stock at risk of flood damages as well as the avoided damages requires additional information on operating and capital (that is, the benefits provided by habitats in reducing flood costs not collected by most national statistical agencies risk to people and produced capital). 163 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 8.1 Key steps and data for estimating the flood protection benefits provided by mangroves OFFSHORE NEARSHORE HABITAT IMPACTS CONSEQUENCES DYNAMICS DYNAMICS Impact with mangroves Impact without mangroves Offshore Nearshore Onshore Source: Figure 2.1 in Menéndez et al. 2024. This approach follows the same methodology used in mangrove benefits over time, the flood models were rerun CWON 2021 but uses new and updated past global mangrove for all years using the new data, and new assessments of distribution data from Global Mangrove Watch (GMW 3.0; risk and benefits were developed. Bunting et al. 2022). While the past mangrove data did not change greatly (1996–2015), all prior year models were Despite observing changes in the total area of mangroves, rerun. The results provided here represent improvements these changes had limited impact on the results. The crucial on the prior CWON report in addition to the most recent factor driving variations in coastal flooding is the cross- yearly estimates (to 2020). The most significant difference shore width of the mangrove forest, and this distance did in these datasets is that the improved GMW 3.0 shows a not change greatly even when there were changes in total consistently greater global coverage than GMW 2.0 (about mangrove area. 7 percent greater, Figure 8.2). To consistently assess REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 164 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 8.2 Differences in global mangrove cover between GMW 2.0 and GMW 3.0 16,500,000 16,000,000 15,500,000 15,000,000 14,500,000 14,000,000 1990 1995 2000 2005 2010 2015 2020  GMW (v2.0)  GMW (v3.0) Source: Figure 3.1 in Menéndez et al. 2023. Note: Mangrove cover is reported in hectares. GLOBAL TRENDS OF BLUE stocks, the real value of the asset considerably declined NATURAL CAPITAL between 1995 and 2020. Due to the overexploitation of this resource and the negative effects of climate change, the Marine fish stocks biomass of marine fish stocks has declined (Cheung et al. The global value of marine fish stocks represents only a 2021), resulting in declines in fishery revenues and rents fraction (less than 1 percent) of the world’s total renewable (Sumalia et al. 2011, 2019; Lam et al. 2020). Hence, the natural capital in nominal terms. This is in part due to the decline in fish stocks has impacted global wealth,201 which large number of landlocked countries that do not have has dropped from $54 to $29 chained 2019 US dollars per access to this resource and, thus, record no wealth from capita. This means that the world has lost about one-quarter marine fish stocks. However, in smaller coastal countries, of its wealth from marine fish stock in 25 years (Figure 8.3), 3 marine fish stocks can be an important source of wealth. For times faster than mangroves and 10 times faster than timber. example, almost half of Malta’s renewable natural capital Some countries have seen more dramatic declines, where comes from these marine resources. large economies, including Kenya, Pakistan, Saudi Arabia, Senegal, and Tanzania, have lost more than 60 percent of Moreover, for those countries with access to marine fish their per capita value generated by their marine fish stocks. 201 Unless otherwise indicated, all wealth estimates are in real terms and are reported in chained 2019 US dollars. 165 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 8.3 Global marine fish wealth, indexed to 1995, 1995–2020 100 Indexed fisheries wealth (1995=100) 80 60 40 20 0 1995 2000 2005 2010 2015 2020 Source: World Bank staff estimates. Note: Wealth is measured in chained 2019 US dollars. The decline in global wealth from marine fish stocks is not the depletion of marine fish wealth (Panel a, Figure 8.4; see the only concern, with CWON 2024 revealing significant also Englander 2019 and Hilborn et al. 2020). Wealth from inequalities across income groups. People living in high- marine fish stocks has not been depleted at the same rate income countries hold more than double the marine fish across regions. Coastal countries in the Middle East and wealth of people living in lower- and upper-middle-income North Africa and in Sub-Saharan Africa have experienced countries. The situation for low-income countries is even the fastest wealth depletion rate, losing almost two-thirds more concerning. On average, low-income countries have of their wealth per capita from marine fish stocks in just a lost around one-third of their wealth per capita from marine quarter of a century. The situation in other regions is more fish stocks, going from $12 to $4 in chained 2019 US dollars optimistic. Europe and Central Asia and Latin America and between 1995 and 2020. At the same time, inadequate the Caribbean, on average, have about the same wealth fisheries management in lower-middle-income countries from marine fish stocks now as they had in 1995 (Panel b, has led to unsustainable catch practices that accelerated Figure 8.4). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 166 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 8.4 Per capita wealth from marine fish stocks, by income and region, 1995–2020 Panel a: By income 100 Fisheries wealth per capita (1995=100) 80 60 40 20 1995 2000 2005 2010 2015 2020  High income  Upper middle income  Lower middle income  Low income 100 Panel b: By region Fisheries wealth per capita (1995=100) 80 60 40 20 1995 2000 2005 2010 2015 2020  East Asia & Pacific  Europe and Central Asia  Latin America & Caribbean  Middle East & North Africa  North America  South Asia  Sub-Saharan Africa Source: World Bank staff estimates. Note: Wealth per capita is measured in chained 2019 US dollars. 167 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Wealth in marine fish stocks is highly concentrated in a all regions and income groups, from Peru in Latin America few countries. There are 20 countries that hold more than to Malaysia in Asia, and from Morocco in North Africa to two-thirds of the wealth (Panel a, Figure 8.5), with about Norway in Europe. In per capita terms, Iceland has the 50 percent of the value of all the world’s marine fish stocks largest wealth in marine fish stocks available per person, concentrated in four countries: Indonesia, China, the followed by Norway and other coastal nations with smaller United States, and Vietnam. Despite the relatively high territories, including Belize, the Solomon Islands, and Malta concentration of wealth in a few countries, richness in (Panel b, Figure 8.5). marine fish stocks can be found all over the world and in FIGURE 8.5 Wealth in marine fish stocks, top 20 countries, 2020 Panel a: Total wealth 50 (chained 2019 USD, billions) 40 Fisheries wealth 30 20 10 0 Norway Peru Morocco Mexico Rest of the world Indonesia China United States Vietnam Japan India Thailand Canada Korea, Rep Spain Chile France Malaysia Philippines Bangladesh Russian Federation Panel b: Wealth per capita 8,000 Fisheries wealth per capita (chained 2019 USD) 6,000 4,000 2,000 0 Norway Belize Malta Oman Peru Comoros Guyana Senegal Chile Georgia Rest of the world Solomon Islands Mauritania New Zealand Namibia Denmark Canada Vietnam Suriname Bahrain Iceland Source: World Bank staff estimates. Note: Wealth is measured in chained 2019 US dollars. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 168 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Among these countries with vast wealth in marine fish of the top 10 largest amounts of marine fish stock wealth stocks, there are countries whose resources are being per capita in the world, lost more than 70 percent of fish depleted much faster (Figure 8.6). Many of these countries biomass between 1995 and 2020, which puts at risk the are found in Sub-Saharan Africa and in the Middle East and sustainable future of an important part of their renewable North Africa, where renewable natural capital could be natural capital. scarce. For example, Oman and Mauritania, which hold one FIGURE 8.6 Countries with the top 20 largest loss of fish biomass per capita, 1995–2020 0 Biomass per capita change, -20 1995-2020 (%) -40 -60 -80 -100 Qatar Madagascar Kuwait Jordan Bahrain Somalia Oman Iraq Kenya Sudan Gambia, The Mauritania Tanzania Yamen, Rep Liberia Mozambique Saudi Arabia Senegal United Arab Emirates Equatorial Guinea Source: World Bank staff estimates. Note: The red line shows the global average of marine biomass loss of 45 percent. Biomass is measured in tons. AQUACULTURE and value to estimate resource rents for 1995 to 2020 (Table 8.1). The estimates in Table 8.1 indicate that the Norwegian The resource rents were computed for several pilot countries salmonid industry has been generating significant resource to illustrate the applicability of the RVM/NPV approach to rents in recent years, in line with recent research by Greaker aquaculture and gauge its economic importance. The first and Lindholt (2021).202 The profits being generated by pilot country is the salmon and trout farming industry salmon farming companies were noted by the Norwegian in Norway, for which resource rent could be estimated government, and following several months of negotiations, by drawing on detailed information on the aquaculture an extraordinary tax was applied to large companies with industry’s operating and capital costs published by the promises that this would result in increased resources for Norwegian Directorate of Fisheries website. These data were areas of the country where fish farms are located. combined with FAO information on production quantity 202 Greaker and Lindholt (2021) estimate that while resource rent was around zero until 2010, over the most recent 10-year period, resource rent averaged around 14 billion Norwegian kroner ($1.8 billion) per year. 169 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 8.1 Resource rent estimates for Norwegian salmon and trout aquaculture, selected years, 1995–2020 YEAR 1995 2020 2005 2010 2015 2020 Production (t) 276,510 448,267 645,080 994,211 1,376,353 1,484,697 Output ($ x 1,000) 1,022,255 1,374,495 2,099,890 5,040,216 5,760,477 7,282,847 Intermediate consumption (-) 724,653 858,398 1,540,937 3,274,142 3,697,225 4,014,363 Seed ($ x 1,000) 186,264 148,192 197,020 401,977 457,414 729,076 Feed ($ x 1,000) 403,792 532,655 1,007,938 2,154,124 2,429,858 2,463,965 Fuel and transport ($ x 1,000) 134,597 177,552 335,979 718,041 809,953 821,322 VALUE ADDED (=) ($ x 1,000) 297,602 516,097 558,953 1,766,073 2,063,253 3,268,485 Compensation of employees (-) 184,640 118,996 148,883 364,894 413,796 592,208 GROSS OPERATING SURPLUS 112,972 397,100 410,071 1,401,179 1,649,457 2,676,277 Depreciation (-) 9,411 59,569 85,976 217,664 310,340 522,548 Return on fixed assets (-) 2,936 14,452 20,102 49,210 72,295 127,012 RESOURCE RENT ($ x 1,000) 100,615 323,079 303,993 1,134,305 1,266,822 2,026,717 Source: https://www.fiskeridir.no/English/Aquaculture/Statistics/Atlantic-salmon-and-rainbow-trout. Note: All data are reported in current US dollars. The second pilot study was conducted for Egypt, which and does not include carp culture, most of which takes place has developed a significant tilapia and mullet production in government farms or marine aquaculture, both of which industry based on small-scale fish farmers operating earth have very different operating cost structures and take place ponds fed by water from irrigation systems. Resource in areas beyond the main zones. The low levels of resource rents were estimated over the period 2010 to 2020 based rent during 2011–2014 may be related to a period of social on production quantity and value from FAO, combined unrest when there was considerable economic uncertainty. with estimates from WorldFish Egypt (A. Nasr-Allah, pers. Fluctuating exchange rates and shortages of foreign currency comm.). As shown in Figure 8.7, this analysis shows that restricted imports such as soya, an essential ingredient in Egyptian tilapia and mullet aquaculture generated significant aquaculture feeds. The dip in production value from 2016 to resource rents, estimated at about $500 million per year over 2017 was due to a rapid fall in the value of the Egyptian pound the last five years. It should be noted that this is based only against the US dollar. on tilapia/mullet polyculture in the main aquaculture zones REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 170 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 8.7 Egyptian tilapia/mullet aquaculture resource rent and sales value, 2010–2020 2,500 2,000 1,500 1,000 500 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 -500  Resource rent ($ million)  Sales value ($ million) Sources: FAO 2023a and A. Nasr-Allah, WorldFish Egypt (pers. comm.). Note: The resource rent and sales value are reported in current $ million per year. Estimation of aquaculture resource rents in Egypt highlights Ecuador was selected as the third case study, given that it the challenge posed by collecting data from large numbers has played a major role in the global success of saltwater, of smallholder farmers. Most small-scale fish farmers do pond-based shrimp aquaculture. Despite the importance of not keep accurate records and are not required to report shrimp aquaculture to Ecuador, there appears to be almost production data to regulatory authorities. The main no published data on operating or capital costs for shrimp operating cost for Egyptian fish farmers is feed, while farms. A resource rent estimate for shrimp production in labor and seed costs are relatively low. Most fish farms are Ecuador in 2020 was based on production quantity and family-operated, informal businesses that depend on credit value information from FAO (FAO 2023), and operating from feed companies and wholesalers to cover their annual and capital cost estimates from a key contact who worked operating costs. Wholesalers recover their own credit and in the industry for several years (P. Buike, pers. comm.). It credit advanced by the feed companies when they sell the shows strongly positive resource rents of $1.5 billion being fish from each farmer at the end of the growing season. generated by the sector in 2020, equivalent to over $2 per Aquaculture production has grown in Egypt because value kilogram of production. chain actors have been able to invest and make profits (Dickson et al. 2016). This profitability is likely to continue as long as regulations place restrictions on the area available for aquaculture while the domestic market continues to expand. 171 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Japan is a marine aquaculture pioneer, with records of oyster, been very quick to switch between species. Both pangasius marine finfish farming, and seaweed cultivation going back and tilapia are relatively new additions to Bangladeshi fish to the 17th century. The Japanese Ministry of Agriculture, farms. The driving forces behind these switches are the Fisheries and Forestry (MAFF) publishes sample data each development of new farming systems and support services year on the operating costs for aquaculture subsectors (MAFF (fingerlings, feeds, and markets) for those systems, as well 2022). Together with FAO data on Japanese aquaculture as profitability, whether this is to supply local, urban, or production quantity and value (FAO 2023), excerpts from the export markets (Hernandez et al. 2018). MAFF data were used to calculate resource rent estimates for yellowtail, sea bream, scallops, oysters, and seaweeds in 2020. Indonesia has the second-largest aquaculture production These indicate negative resource rents for yellowtail (-$305 industry in the world, worth around $12 billion per year. million), sea bream (-$96 million), and oysters (-$525 million), It has a diverse range of aquaculture systems growing while resource rents for scallops ($287 million) and seaweeds freshwater, marine, and brackish-water species. Preliminary ($619 million) were positive. resource rent estimates were developed for shrimp, tilapia, carp, milkfish, and pangasius based on operating and This preliminary analysis highlights the variable capital cost data supplied by a consultant (R. Tan, pers. performance of aquaculture systems within a country. comm.) and production figures from FAO. These indicated Yellowtail and sea bream have high feed costs, while market positive resource rents for shrimp ($1 billion), milkfish prices are probably affected by imports. Japan imports ($215 million), and carp ($104 million); a very low resource around 40 percent of its seafood, making it the third-largest rent estimate for tilapia ($7 million); and a negative value global importer (Ganapathiraju, Pitcher, and Mantha 2019). for pangasius (-$17 million). This means domestic producers face market competition while Japanese companies have invested in aquaculture Figure 8.8 summarizes the results of resource rent estimates production in other countries. from the pilot countries and species, expressed as resource rent per tonne of production. It indicates highly positive Aquaculture production expanded rapidly in Bangladesh 2020 resource rents for Japanese seaweed ($2,230/tonne), from 2.8 million tonnes in 2010 to 6.3 million tonnes in Ecuadoran shrimp ($2,067/tonne), Indonesian shrimp 2020 through the intensification of production systems as ($1,450/tonne), and Norwegian salmonids ($1,365/tonne, well as the conversion of agricultural land into aquaculture dropping from a peak of $2,368/tonne in 2018). Resource ponds (Hasan et al. 2021; FAO 2023a). Two recent research rent estimates for Indonesian tilapia, carp, and milkfish and papers focused on the economics of carp polyculture (rohu, Bangladeshi pangasius were between $764/tonne and zero, mrigal, catla, silver carp, and common carp) in 2019, as well while estimates for Japanese oysters, yellowtail, sea bream, as tilapia culture and pangasius catfish culture in 2017 (Saha and Indonesian pangasius were negative. et al. 2022; Hossain et al. 2022). Both papers analyzed data from farms that were considered either profitable or not However, these resource rent estimates cannot be used to profitable. The data were scaled up to national production extrapolate to other countries. For example, resource rents levels and used to calculate resource rent estimates of generated by shrimp production in Ecuador were much around $785 million for carp culture in 2019, while the higher than the equivalent figures in Indonesia, and similar figures for pangasius and tilapia resource rents were $300 variations in resource rents were apparent between tilapia million and $129 million, respectively, in 2017. aquaculture systems in Egypt, Bangladesh, and Indonesia, suggesting that location is an important factor. Each country The resource rent calculations extrapolated from the will have specific costs for feeds, seed, energy, labor, Bangladesh research studies appear to fit with national site rental, and maintenance costs, as most aquaculture trends where there has been large-scale investment in production is sold locally, so revenue will vary according to carp, tilapia, and pangasius farming. Fish farmers have local market conditions. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 172 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 8.8 Resource rent estimates in selected countries and species, 1995–2020 3,000 2,000 1,000 0 1995 1996 1997 1998 1999 2000 2001 2003 2006 2016 2002 2004 2005 2007 2008 2009 2010 2011 2012 2013 2014 2015 2017 2018 2019 2020 -1,000 -2,000 -3,000 -4,000 Norway salmonid Egypt tilapia & mullet Ecuador shrimp Japan yellowtail Japan sea bream Japan scallops Japan oysters Japan seaweed Bangladesh carps Bangladesh pangasius Bangladesh tilapia Indonesia shrimp Indonesia tilapia Indonesia carp Indonesia milkfish Indonesia pangasius Source: Dickson et al. 2024, forthcoming. Note: The resource rent estimates are reported in current US dollars per tonne of production. Although direct extrapolation of results from the pilot aquaculture resource rents (but still within the range of studies is not possible, they provide indications of expected estimates from the pilot study), total resource rents could levels of resource rent. Assuming conservative average be as high as $55 billion, equivalent to a total asset value resource rent values, the global aquaculture resource rent of over $1,300 billion. With global aquaculture production might be $20 billion. This is equivalent to an asset value of continuing to grow, driven by increased demand and $500 billion in nominal terms, which is more than double improved efficiency of aquaculture practices, the asset value the estimated asset value for global capture fisheries of has the potential for even greater impacts in the future. $228 billion in 2020. Using less conservative estimates of 173 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH MANGROVE COASTLINE in nominal terms have increased substantially over the PROTECTION SERVICES past 25 years, driven by population and economic growth in coastal areas. Between 1995 and 2020, the flood risk and As described in Figure 8.1, there are two main components mangrove protection benefit to people almost doubled, while for computing mangrove coastline protection services: the the amount of capital stock at risk and receiving protection risk and habitat benefit of protected people and produced benefits more than quadrupled in value (Figure 8.9). capital. Both flood risks and habitat risk reduction benefits FIGURE 8.9 Value of people and capital stock at risk and receiving risk reduction benefits in current US dollars, indexed to 1995, 1995–2020 1,000 800 Index (1995=100) 600 400 200 0 1995 2000 2005 2010 2015 2020  People at risk  People benefits  Capital stock at risk  Capital stock benefits Source: World Bank staff estimates. Note: People and capital stock benefits are in units of current US dollars prior to indexing using valuation methods described in the CWON 2024 technical background report (Menédez et al. 2024, forthcoming). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 174 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH The benefits received by different groups of nations vary On the other hand, the number of people living in coastal greatly. For example, the benefit provided by mangroves areas in low-income countries benefiting from mangroves of protecting the capital stock in upper-middle- and high- on a per hectare basis increased by more than 3,000 percent income countries increased by about 1,000 percent on a per over the same period (Panel a, Figure 8.10). hectare basis over the last 25 years (Panel b, Figure 8.10). FIGURE 8.10 Per hectare flood risk and mangrove benefits to people and capital stock by income level in nominal terms, 1995–2020 Panel a: Benefit to people, per hectare 4 Average benefited people per hectare 3 (population) 2 1 0 1995 2000 1995 2000 1995 2000 1995 2000 Low income Lower middle Upper middle High income income income Panel b: Benefit to capital stock, per hectare Average benefit to capital stock per hectare 1,000 (current US$, thousands) 500 0 1995 2000 1995 2000 1995 2000 1995 2000 Low income Lower middle Upper middle High income income income Source: World Bank staff estimates. Note: Benefit to people and capital stock are measured in current US dollars, as described in the CWON 2024 technical background report (Menédez et al. 2024). 175 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Coastal flood risk has grown dramatically over the past 10 times the size of New York City. In per capita terms, the three decades, with increases in overall global population results show that in 2020, the mangrove area available per and wealth and concentrated development on coastlines. person was 30 percent lower than in 1995, which resulted Concomitantly, the nominal benefits provided by mangroves in a decline in the value of mangrove coastline protection in reducing that risk has grown, even though mangroves services per capita of 37 percent. If this trend continues, our have been lost. However, in real terms, the value of these next generations could have less than half of the mangrove services has dropped by 15 percent due to an alarming coastline protection services value that was available in loss of mangrove area in the past 25 years (Figure 8.11). 1995, leaving them more exposed to flooding and climate According to Menéndez et al (2023), between 1995 and 2020, change impacts. the world lost about 700,000 hectares of mangroves, nearly FIGURE 8.11 Mangrove area and the value of coastline protection services in real terms, indexed to 1995, 1995–2020 100 95 Index (1995=100) 90 85 1995 2000 2005 2010 2015 2020  Mangrove area  Mangrove shoreline protection services Source: World Bank staff estimates. Note: Mangrove area is measured in hectares and mangrove coastline protection services in chained 2019 US dollars. People in more than 100 nations receive flood protection Saharan Africa and the Middle East and North Africa have benefits from mangroves. Mangroves are in greatest abundance experienced the fastest decline, with the mangrove coastline in (sub)tropical countries in East Asia and the Pacific, with protection services per capita having dropped almost 50 over 155 million hectares, followed by Latin America and the percent in real terms between 1995 and 2020. While the rate Caribbean (87 million hectares) and Sub-Saharan Africa (76 of mangrove degradation in the other regions is lower, it is million hectares). Despite the richness of mangroves in these still alarming that the loss of mangrove forests has reduced its regions, the degradation of these natural resources is reducing coastline protection services per person by one-third over the the value of their coastline protection services per capita. Sub- same period (Panel b, Figure 8.12). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 176 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH The depletion rate of mangrove coastline protection services from mangroves, which is now less than 50 percent of what value is different at different income levels. The lower it was in 1995 (Panel a, Figure 8.12). Lack of preservation the income level, the higher the decline in the value of of this important natural resource in low-income countries mangrove coastline protection services per capita. Higher is aggravating the situation. Low-income countries have population growth in low-income countries is accelerating reduced the mangrove forest area by 4 percent, from 1,173 the reduction of the available coastline protection services thousand hectares in 1995 to 1,122 thousand hectares in 2020. FIGURE 8.12 Mangrove coastline protection services value per capita, by income and region, indexed to 1995, 1995–2020 Panel a: Mangrove coastline protection services per capita by income group 100 Mangrove shoreline protection services wealth per capita (1995=100) 80 60 40 1995 2000 2005 2010 2015 2020  High income  Upper middle income  Lower middle income  Low income 100 Panel b: Mangrove coastline protection services per capita by region Mangrove shoreline protection services wealth per capita (1995=100) 90 80 70 60 50 1995 2000 2005 2010 2015 2020  East Asia & Pacific  South Asia  North America  Middle East & North Africa  Latin America & Caribbean  Sub-Saharan Africa Source: World Bank staff estimates. Note: Mangrove coastline protection services in units of chained 2019 US dollars. 177 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH The countries with the largest value of mangrove coastline value per capita, since most of their population lives close protection services are those with large populations living to their coastlines. There are also other smaller countries, in coastal areas and vast capital stock built around their such as St. Lucia and Belize, where the value of mangrove shores. Several of these countries are in East Asia and coastline protection services per capita are within the the Pacific, including China, Vietnam, Australia, and top 10 in the world. For these countries, the preservation Indonesia, but they can also be found in other parts of the of mangroves is even more important, because failing world, such as Mexico in Latin America and the Caribbean to maintain their current state could impact the value of or Bangladesh in South Asia (Figure 8.13). From these, their natural capital and increase the risks of flooding or Australia and Vietnam are the countries with the largest vulnerability to climate change shocks. FIGURE 8.13 Countries with the largest mangrove coastline protection services value, 2020 200 Panel a: Absolute values Mangrove shoreline protection (chained 2019 USD, billions) 150 services wealth 100 50 0 Ecuador Mexico Brazil China Vietnam Australia India Japan Malaysia United States Indonesia Bangladesh Ghana Rest of the world Suriname Philippines Thailand New Zealand Cambodia Saudi Arabia United Arab Emirates 5,000 Panel b: Per capita values Mangrove shoreline protection services wealth per capita 4,000 (chained 2019 USD) 3,000 2,000 1000 0 Qatar Ecuador Mexico St Lucia Vietnam Rest of the world Belize New Zealand Solomon Islands Guyana Haiti Oman United States Jamaica Bangladesh Cambodia Indonesia China United Arab Emirates Australia Source: World Bank staff estimates. Note: Wealth is measured in chained 2019 US dollars. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 178 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH CONCLUSIONS spatial heterogeneity of built-up infrastructure. Such data are essential to help inform decision-making at the local and Traditional methods of measuring wealth and economic country scale to ensure these critical assets are restored and progress account for built capital and overlook the value managed more sustainably into the distant future. of nature’s other goods and services. The economic benefits provided by natural resources and ecosystems are The inclusion of aquaculture assets in CWON is not undervalued, so their management is adversely affected by a challenging from a conceptual point because an RVM/NPV lack of data. This chapter provides an update of the estimates approach can readily be applied, as recommended by the SNA for marine fish stocks and coastline protection services of and SEEA for traded natural capital assets. However, there mangroves, accounting for the blue natural capital of each are significant data constraints, as the required information country. The degradation of stock of these natural capital on operating and capital costs is rarely reported in publicly assets negatively impacts its value, especially the value of available datasets. Most of the estimates for this pilot account marine fish stocks on a per capita basis. Marine fish stocks thus had to be based on research results and information have experienced the fastest decline among renewable from key contacts. natural capital assets, followed by mangrove coastline Resource rent estimates for most of the aquaculture systems protection services. While the value of protection benefits included in this pilot account were positive and significant, for populations and capital stock along the world’s coastlines including for salmon farms in Norway, tilapia farms in continues to grow in nominal terms, the real value of the Egypt, shrimp farms in Ecuador and Indonesia, and a range mangrove coastline protection services on a per capita basis of finfish species in Bangladesh. However, estimates for is constantly declining due to the insufficient preservation finfish aquaculture in Japan were more variable. A back- efforts to stop mangrove forest loss. of-the-envelope extrapolation finds that the asset value of To prevent this natural resource degradation, it is critical aquaculture could reach $1 trillion, which is significantly to include their value in national wealth accounting. more than the value of marine capture fisheries currently CWON provides a methodology to include these assets in included in CWON. To estimate the asset value of aquaculture a macroeconomic framework to guide policy makers on for all major aquaculture-producing countries would assessing the risks and challenges linked to these resources require significant additional data collection and analysis. that could jeopardize economic sustainability. Nonetheless, A key recommendation of this pilot study is to build a for both assets, there is scope to improve measurement. systematic and integrated data collection system consistent For marine fish stocks, data on landed values and assets are with SNA and SEEA methodologies and supported by an limited, and gaps need to be filled through estimation. More aquaculture economics community of practice comprising systematic data collection is needed to fill these key gaps. For national statisticians, technical specialists, economists, the coastline protection services of mangroves, spatialized and regulatory authorities. Data collection and analysis capital stock estimates are not available for the entire period from 12 leading aquaculture-producing countries would covered by the CWON database. More disaggregated data are be sufficient to cover 75 percent of the total value of global needed to provide more accurate estimates of the capital aquaculture production. stock at risk and protected by mangroves, given the great 179 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH REFERENCES Bunting, P., Rosenqvist, A., Hilarides, L., Lucas, R.M., Thomas, N., Tadono, Hernandez, R., Belton, B., Reardon, T., Hu, C., Zhang, X., and Ahmed, A. T., Worthington, T.A., Spalding, M., Murray, N.J., and Rebelo, L.-M. 2022. 2018. “The ‘Quiet Revolution’ in the Aquaculture Value Chain “Global Mangrove Extent Change 1996–2020: Global Mangrove Watch in Bangladesh.” Aquaculture 493: 456–468. Version 3.0.” Remote Sensing 14: 3657. https://doi.org/10.1016/j.aquaculture.2017.06.006. Cheung, W.W., Frölicher, T.L., Lam, V.W., Oyinlola, M.A., Reygondeau, G., Hilborn, R., Amoroso, R.O., Anderson, C.M., Baum, J.K., Branch, T.A., Sumaila, U.R., ... and Wabnitz, C.C. 2021. “Marine High Temperature Costello, C., ... and Ye, Y. 2020. “Effective Fisheries Management Extremes Amplify the Impacts of Climate Change on Fish and Fisheries.” Instrumental in Improving Fish Stock Status.” Proceedings of the National Science Advances 7 (40): eabh0895. Academy of Sciences 117 (4): 2218–2224. Comely, B., and McGuinness, K. 2005. “Above- and Below-Ground Biomass, Hossain, Md. E., Khan, Md. A., Saha, S.M., and Dey, M.M. 2022. “Economic and Allometry, of Four Common Northern Australian Mangroves.” Assessment of Freshwater Carp Polyculture in Bangladesh: Profit Australian Journal of Botany 53: 431–436. Sensitivity, Economies of Scale and Liquidity.” Aquaculture 548 (1): 737552. https://doi:10.1016/j.aquaculture.2021.737552. Dickson, M.W., Charo-Karisa, H., Onder, S., and Smith, R. 2024. Estimating the Global Contribution of Aquaculture towards the Changing Wealth of Hutchison, J., Manica, A., Swetnam, R., Balmford, A., and Spalding, M. 2014 Nations (CWON) 2024 Report. Technical report. “Predicting Global Patterns in Mangrove Forest Biomass.” Conservation Letters 7: 233–240. Dickson, M.W., Nasr-Allah, A., Kenawy, D., and Kruijssen, F. 2016. “Increasing Fish Farm Profitability through Aquaculture Best Management IPCC (Intergovernmental Panel on Climate Change). 2014. 2013 Supplement Practice Training in Egypt.” Aquaculture 465: 172–178. to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: https://doi.org/10.1016/j.aquaculture.2016.09.015. Wetlands. Edited by T. Hiraishi, T. Krug, K. Tanabe, N. Srivastava, J. Baasansuren, M. Fukuda, and T.G. Troxler. Switzerland: IPCC. Donato, D.C., Kauffman, J.B., Murdiyarso, D., Kurnianto, S., Stidham, M., and Kanninen, M. 2011. “Mangroves among the Most Carbon-Rich Forests in Jardine, S.L., and Siikamäki, J.V. 2014. “A Global Predictive Model of Carbon in the Tropics.” Nature Geoscience 4 (5): 293–297. Mangrove Soils.” Environmental Research Letters 9 (10): 104013. Englander, G. 2019. “Property Rights and the Protection of Global Marine Kauffman, J.B., Adame, M.F., Arifanti, V.B., Schile-Beers, L.M., Bernardino, Resources.” Nature Sustainability 2 (10): 981–987. A.F., Bhomia, R.K., Donato, D.C., Feller, I.C., Ferreira, T.O., and Garcia, M.d.C.J. 2020. “Total Ecosystem Carbon Stocks of Mangroves across Broad Global Environmental and Physical Gradients.” Ecological Monographs Ermgassen, P.S.E., et al. 2021. “Reprint of: Fishers Who Rely on Mangroves: e01405. Modelling and Mapping the Global Intensity of Mangrove-Associated Fisheries.” Estuarine, Coastal and Shelf Science 248 (107159). Kauffman, J.B., and Donato, D.C. 2012. “Protocols for the Measurement, Monitoring and Reporting of Structure, Biomass and Carbon Stocks in FAO (Food and Agriculture Organization). 2001. Global Forest Resources Mangrove Forests.” Working Paper 86, CIFOR, Bogor, Indonesia. Assessment 2000. Rome, Italy: FAO. Lam, V.W., Allison, E.H., Bell, J.D., Blythe, J., Cheung, W.W., Frölicher, FAO. 2022. The State of World Fisheries and Aquaculture 2022. Towards Blue T.L., ... and Sumaila, U.R. 2020. “Climate Change, Tropical Fisheries, Transformation. Rome, Italy: FAO. https://doi:10.4060/cc0461en. and Prospects for Sustainable Development.” Nature Reviews Earth & Environment 1 (9): 440–454. FAO. 2023. “FAO FishStatJ – Software for Fishery and Aquaculture Statistical Time Series.” www.fao.org/fishery/en/statistics/software/fishstatj. Lam, V.W.Y., and Sumaila, R. 2021. A Practical Approach for Estimating Marine Fisheries Asset Value. Technical report. Ganapathiraju, P., Pitcher, T.J., and Mantha, G. 2019. “Estimates of Illegal and Unreported Seafood Imports to Japan.” Marine Policy 108 (103439). Leal, M., and Spalding, M.D., eds. 2022. The State of the World’s Mangroves https://doi.org/10.1016/j.marpol.2019.02.011. 2022. Global Mangrove Alliance. Greaker, M., and Lindholt, L. 2021. “The Resource Rent in Norwegian Little, D., Newton, R., and Beveridge, M. 2016. “Aquaculture: A Rapidly Aquaculture 1984–2020.” Discussion Papers No. 962, Statistics Norway, Growing and Significant Source of Sustainable Food? Status, Transitions Research Department. https://doi.org/10.1016/j.fsirep.2023.100099. and Potential.” Proceedings of the Nutrition Society 75 (3): 274–286. https://doi:10.1017/S0029665116000665. Hagger, V., Worthington, T.A., Lovelock, C.E., et al. 2022. “Drivers of Global Mangrove Loss and Gain in Social-Ecological Systems.” Nature McLeod, E., Chmura, G.L., Bouillon, S., Salm, R., Björk, M., Duarte, C.M., Communications 13 (6373). Lovelock, C.E., Schlesinger, W.H., and Silliman, B.R. 2011. “A Blueprint for Blue Carbon: Toward an Improved Understanding of the Role of Vegetated Hasan, N.A., Heal, R.D., Bashar, A., Bablee, A.L., and Haque, M.M. 2021. Coastal Habitats in Sequestering CO2.” Frontiers in Ecology and the “Impacts of COVID-19 on the Finfish Aquaculture Industry of Bangladesh: Environment 9 (10): 552–560. A Case Study.” . 130 (104577). https://doi.org/10.1016/j.marpol.2021.104577. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 180 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Menéndez, P., Beck, M.W., Abad, S., and Losada, I. 2024. Building Coastal Stuchtey, M.R., Vincent, A., Merkl, A., Bucher, M., Haugan, P.M., Lubchenco, Resilience with Mangroves: The Contribution of Natural Flood Defenses to J., ... and Haugan, P.M. 2023. “Ocean Solutions that Benefit People, Nature the Changing Wealth of Nations. Technical report. Washington, DC: World and the Economy.” In The Blue Compendium: From Knowledge to Action Bank. for a Sustainable Ocean Economy 783–906. Cham: Springer International Publishing. Polidoro, B.A., et al. 2010. “The Loss of Species: Mangrove Extinction Risk and Geographic Areas of Global Concern.” PLOS One 5 (4). Sumaila, U.R. 2021. Infinity Fish: Economics and the Future of Fish and Fisheries. Academic Press. Richards, D.R., Thompson, B.S., and Wijedasa, L. 2020. “Quantifying Net Loss of Global Mangrove Carbon Stocks from 20 Years of Land Cover Change.” Sumaila, U.R., Cheung, W.W., Lam, V.W., Pauly, D., and Herrick, S. 2011. Nature Communications 11 (4260). “Climate Change Impacts on the Biophysics and Economics of World Fisheries.” Nature Climate Change 1 (9): 449–456. Rovai, A.S., Twilley, R.R., Castañeda-Moya, E., Midway, S.R., Friess, D.A., Trettin, C.C., ... and Riul, P. 2021. “Macroecological Patterns of Forest Sumaila, U.R., Tai, T.C., Lam, V.W., Cheung, W.W., Bailey, M., Cisneros- Structure and Allometric Scaling in Mangrove Forests.” Global Ecology and Montemayor, A.M., ... and Gulati, S.S. 2019. “Benefits of the Paris Biogeography 30 (5): 1000–1013. Agreement to Ocean Life, Economies, and People.” Science Advances 5 (2): eaau3855. Rovai, A.S., Twilley, R.R., Castañedpela-Moya, E., Riul, P., Cifuentes-Jara, M., Manrow-Villalobos, M., ... and Pagliosa, P.R. 2018. “Global Controls on Sumaila, U.R., Teh, L., Palomares, D., Alam, L., and Cheung, W. 2024. Global Carbon Storage in Mangrove Soils.” Nature Climate Change 8 (6): 534–538. Ocean Fish Wealth Accounts 2024. Technical report. Washington, DC: World Bank. Saha, P., Hossain, Md. E., Prodhan, Md. M.H., Rahman, Md. T., Nielsen, M., and Khan, Md. A. 2022. “Profit and Loss Dynamics of Aquaculture Farming.” United Nations. 2008. System of National Accounts, 2008. New York: United Aquaculture 561: 738619. https:// doi:10.1016/j.aquaculture.2022.738619. Nations. https://unstats.un.org/unsd/nationalaccount/docs/SNA2008.pdf. Sanderman, J., Hengl, T., Fiske, G., Solvik, K., Adame, M.F., Benson, L., ... United Nations et al. 2021. System of Environmental-Economic Accounting— and Landis, E. 2018. “A Global Map of Mangrove Forest Soil Carbon at 30 m Ecosystem Accounting (SEEA EA). White cover publication, pre-edited text Spatial Resolution.” Environmental Research Letters 055002. subject to official editing. https://seea.un.org/ecosystem-accounting. Schile, L.M., Kauffman, J.B., Crooks, S., Fourqurean, J.W., Glavan, J., and World Bank. 2021. The Changing Wealth of Nations 2021: Managing Assets for Megonigal, J.P. 2017. “Limits on Carbon Sequestration in Arid Blue Carbon the Future. Washington, DC: World Bank. Ecosystems.” Ecological Applications 27 (3): 859–874. Simard, M., Fatoyinbo, L., Smetanka, C., Rivera-Monroy, V.H., Castañeda- Moya, E., Thomas, N., and Van der Stocken, T. 2019. “Mangrove Canopy Height Globally Related to Precipitation, Temperature and Cyclone Frequency.” Nature Geoscience 12 (1): 40–45. 181 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH 9 Human Capital Wealth: Global Trends MAIN MESSAGES Economic theory did not put enough emphasis on human capital before the 1950s, when economists started to consider ■ Human capital—estimated as the present value of human capital to account for income and growth differentials. future earnings for the labor force, employed and self- In 1961, Theodore W. Schultz brought forward the idea employed—is the largest asset across all income groups, of treating humans as a form of capital in his work called constituting 60 percent of total wealth in 2020, slightly “Investment in Human Capital,” and developed the human lower than in 1995. capital theory of economic development. In 1962, Gary Becker developed the theoretical framework for ■ Significant disparity between male and human capital in his work called “Investment female human capital persists across in Human Capital: A Theoretical Analysis.” In most regions and income groups, with this work, he also introduced the economic great variation among regions: in concept of human capital and for the first 2020 women accounted for 44 percent time examined links between education of human capital in Latin America and incomes. Since the 1980s, human capital and the Caribbean, but only 15 percent has become one of the key components of the in South Asia. neoclassical growth accounting frameworks, as well as endogenous growth models. Today, human ■ Setting aside the impact that an influx of capital is accepted as the most important endowment of a women in the labor market might generate, closing the country, making up about 60 percent of the total wealth of gender gap could substantially improve human capital nations in 2020. wealth. If gender parity were globally achieved, this could increase global human capital wealth by about 21 According to Gary Becker’s understanding, human capital is percentage points. a broad concept. It includes not only education and training but also other additions to knowledge and health, such ■ Current CWON estimates are using a simple as accumulated work experience and health habits, even approximation for the changing educational composition including harmful addictions, such as smoking and drug of the work force. Future efforts should focus on use (Becker 1962, 1993a, 1994b). Human capital wealth is improving the quality adjustments to the volume essentially defined in this report as the present value of the measures by compiling a global labor force that future flow of wages and other labor earnings of the current disaggregates workers by gender and educational levels. working population, including both employed and self- employed workers. That is, it focuses on the economic benefits INTRODUCTION that a well-educated and healthy workforce generates. Human capital—the knowledge, skills, competencies, and CWON estimates human capital based on expected lifetime attributes embodied in individuals that facilitate the creation earnings. The first global estimates using expected lifetime of personal, social, and economic well-being—plays an earnings were produced for the 2018 edition of CWON, increasingly central role in the economic success of nations drawing on household surveys for 141 countries over two and individuals (OECD 2001). Human capital is the main decades, from 1995 to 2014 (Lange, Wodon, and Carey driver of sustainable growth and poverty reduction because 2018). This was a significant innovation to previous editions more human capital is associated with higher earnings for (World Bank 2006, 2011), which measured human capital individuals, higher income for countries, and stronger indirectly as a component of the unexplained residual, cohesion in societies (World Bank 2020). called “intangible capital.” REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 182 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH This edition builds on the human capital methodology ESTIMATING HUMAN CAPITAL established in CWON 2018 by expanding coverage to 151 countries from 1995 to 2020 and introducing a volume-based CWON estimates human capital by following the lifetime index to estimate human capital in real terms. income approach developed by Jorgenson and Fraumeni (1989, 1992a, 1992b). According to this approach, human The chapter is organized as follows. First, the estimation of capital is estimated as the total present value of the expected human capital is summarized. The detailed methodology is future labor income that could be generated over the lifetime included in CWON’s methodology document (World Bank of the current working population. There are different 2024). The next section provides an overview of trends in approaches to measuring human capital (see Box 9.1 for human capital at the global, income group, and geographic more detail), but here human capital is considered to be an region levels. This is followed by a more detailed look at asset that generates a stream of future economic benefits. trends in gender disparity and the importance of human The same conceptual approach is applied to other assets in capital in the informal sector. the wealth accounting framework. BOX 9.1 DIFFERENT APPROACHES TO MEASURE HUMAN CAPITAL H uman capital consists of the knowledge, skills, and health that people accumulate over their lives. In addition to its intrinsic importance, human capital is a key driver of sustainable growth and poverty reduction. There are two broad approaches to measuring human capital. The first is an indicators-based approach and the second is a monetary measure-based approach. The indicators-based approach estimates human capital based on measures of population characteristics, such as years of schooling, educational attainment, and test scores (Boarini, Mira d’Ercole, and Liu 2012). Single indicators cannot capture the various dimensions of human capital, and some indicators- based measures—like the UN Development Programme’s Human Development Index or the World Bank’s HCI—combine multiple components to produce more comprehensive human capital indexes. The monetary measure-based approach calculates the total stock of human capital either indirectly or directly. The indirect approach estimates human capital residually, as the difference between the total discounted value of each country’s future consumption flows (which is taken as a proxy for total wealth) and the sum of the tangible components of that wealth, that is, produced capital and the market-component of natural capital (Boarini, Mira d’Ercole, and Liu 2012). This is a useful method, but it has some drawbacks. First, since it is measured residually, estimates for human capital may be biased by measurement error in all the terms entering the accounting identities. Second, it does not consider the nonmarket benefits of the various capital stocks (Liu 2011). 183 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Direct monetary approaches to calculating the stock of human capital include the cost-based approach (for example, Kendrick 1976 and Eisner 1985) and the income-based approach (for example, Jorgenson and Fraumeni 1989, 1992a, 1992b). The cost-based approach considers all the costs that are incurred when producing human capital. Therefore, the human capital wealth stock is the stream of past investments in human capital. Even though the cost-based approach is easy to apply, it relies only on production costs and does not account for demand and supply (Boarini, Mira d’Ercole, and Liu 2012). The income-based approach considers future earnings that human capital investment generates, and human capital wealth stock is a function of these future earnings. While the cost-based approach measures human capital wealth stock from the input side, the income-based approach measures the stock of human capital from the output side (Boarini, Mira d’Ercole, and Liu 2012). The lifetime income approach for measuring human capital as childcare, food preparation, and home repair. The SNA stock brings together a broad range of factors that shape accounts for household production of goods, such as food the stock of the working population. These factors include for own consumption, but does not include household the total population and population structure, but also the production of services.203 As a result, the human capital expected lifespan of people (a measure that reflects health associated with the production of household services is not conditions), their educational attainment, and their labor measured—an omission that disproportionately affects the market experiences in terms of employment probabilities measure of women’s human capital. and earnings. All these factors are disaggregated by gender to capture both possible differences in access to education This concept of human capital differs from that of human and healthcare, as well as the gender pay gap. The lifetime development or human capabilities and complements the income is then computed for a representative male or female World Bank’s Human Capital Project, which compiles a wide worker with a given educational background and experience. range of nonmonetary indicators of human capital (see Box Real wages for future earnings for, say, a 30-year-old male 9.2 for more detail). While these indicators are computed for worker when he is 50 are set equal to wages of male workers the entire population, CWON’s measures of human capital that are currently 50 with the same educational background. focus only on the current working population. Moreover, this approach emphasizes the role of human capital in generating An additional advantage of the lifetime income approach income through wages and earnings, but does not recognize is that it allows changes in human capital to be described other essential benefits from investments in human in terms of investment and stocks. These can include such development, such as the intrinsic value of a good education things as formal and informal education; depreciation, such and good health. However, for wealth accounting purposes, as deaths; and revaluation, such as changes in the labor the focus needs to remain strictly on the monetary estimates market premiums of education (Liu 2011). of wealth associated with human capital. Therefore, human capital estimates in CWON are an underestimate because However, because this approach builds on the concepts and they leave out positive externalities, such as the public good measurement of market labor earnings in the SNA, CWON benefits of an educated population, and restrict the measure human capital estimates have a major omission: human to the current working population not the total population. capital that produces nonmarket household services, such Note that regularly accounting for households’ unpaid service work is a recommended extension in the new SNA 2025, which would be compiled 203 and released periodically. However, it will not be part of standard measures. For more information on the 2025 revision of the SNA see: https://unstats.un.org/unsd/nationalaccount/towards2025. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 184 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 9.2 THE HUMAN CAPITAL INDEX AND CWON’S HUMAN CAPITAL T he World Bank’s HCI is an international metric measuring the human capital that a child born today can expect to attain by their 18th birthday, given the risks of poor health and poor education prevailing in their country. The HCI incorporates key dimensions of human capital: health (child survival, stunting, and adult survival rates) and the quantity and quality of schooling (expected years of school and international test scores). Using global estimates of the economic returns to education and health, these components are combined into an index that captures the expected productivity of a child born today as a future worker, relative to a benchmark of complete education and full health (World Bank 2020). In CWON, human capital is measured as the expected future earnings of the entire labor force. It is estimated as the total present value of the expected future labor income that could be generated over the lifetime of the current working population. In other words, human capital is considered an asset that generates a stream of future economic benefits. CWON’s measure of human capital focuses on the economic benefits that a well-educated and healthy workforce generates. The HCI uses a broader concept of human capital than CWON, incorporating several nonmonetary indicators of health and education outcomes. Conceptually, however, the two measures have much in common, as both are anchored in the development-accounting literature and measure human capital in terms of expected future earnings. The main difference between the two measures is that the HCI measures expected future earnings of a child born today while CWON estimates expected future earnings of the current labor force. In addition, CWON reports estimates in monetary terms, while the HCI is expressed relative to a benchmark of complete education and full health. A child born in a country with an HCI value of 0.5 will be only half as productive as a future worker as they would be if they enjoyed complete education and full health. The CWON measure of human capital complements the HCI, using outcomes that derive indirectly from factors such as educational attainment and health to provide an understanding of the current stock of human capital in countries. CWON also accounts for labor market outcomes, such as the probability of employment and labor market premiums across countries. While the HCI does not include labor market outcomes, the 2020 update introduced the utilization-adjusted index. This analytical extension accounts for the underuse of human capital, based on the fraction of the working age population that is employed, or in the types of jobs that might better enable them to use their skills and abilities to increase their productivity. Sources: World Bank 2018, 2020; the Human Capital Project. 185 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH DATA SOURCES AND METHODS Survival rates are not readily available from the data sources. To calculate survival rates, death rates obtained from the To compute human capital as the discounted value of Global Burden of Disease Study 2019207 are used. The shares expected future labor income, data on the population, of compensation of employees and the self-employed in the employment, annual earnings, survival rates, GDP, and national accounts are retrieved from the PWT 9.1 to control labor shares are needed from different data sources. The the estimated wages. Finally, employment data from the ILO International Income Distribution Database, a unique are used for controlling and scaling up total employment database developed by the World Bank containing more from the consolidated database. The data and methods than 1,500 harmonized household surveys, is used for are described further in CWON’s methodology document calculating annual earnings, educational attainment, and (World Bank 2024). employment rates. The survey information from the World Bank’s International Income Distribution Database204 was Because the survey data do not capture the entire world complemented in this update with harmonized labor force population, the data from the surveys are adjusted to surveys from the World Bank’s Global Labor Database, 205 population estimates from the UN’s World Population the Luxembourg Income Study,206 and the New Zealand Prospects208 to ensure that the estimates are representative. Treasury. Thus, the country coverage for this edition of In addition, the earnings profiles are not compatible with CWON increased to 151 countries, by adding Angola, the published data from the SNA because the profiles from Guinea-Bissau, Israel, New Zealand, and St. Lucia. the surveys do not include any benefits other than wages, including social security payments and other wage-related The consolidated database of household and labor force payments. Hence, the estimated earnings profiles from surveys is used to estimate the private returns per year the surveys are benchmarked to the compensation of of schooling. This requires collecting data from the employees and self-employed that is obtained from the PWT. harmonized surveys on the number of people, their Expected labor earnings from the surveys are scaled up to earnings, school enrollment rates, and employment rates, the labor earnings in the UN’s National Accounts database. and then to estimate the Mincerian coefficients using the In addition, no adjustments are made for the future wage Mincerian wage regressions. Based on these results, a forecasts in line with the international guidance (SNA/ matrix of expected earnings is constructed, which accounts SEEA), which calls for resource rents to be held constant in for labor earnings of the population by age, gender, and the NPV calculation. In other words, no future wage growth education level. is assumed for all countries. The lifetime for working is assumed to be a maximum of Using this in-formation, the lifetime income (adjusted by 50 years, starting at age 15 and ending with retirement at survival rates and a discount factor) can be calculated for age 65, for all countries. All individuals younger than 15 are the representative individual (aged 15–65) by age, gender, assumed to be in school. Individuals between the ages of and education. Subsequently, these lifetime income profiles 15 and 24 are enrolled in school or part of the labor force. are multiplied by the corresponding number of people in a Individuals in the labor force are then expected to work until country to compute the human capital stock by age, gender, age 65, after which labor income is assumed to be zero. In and education. Summing the stocks of human capital across calculating the NPV, a uniform discount rate of 4 percent is all classified categories generates an estimate of the aggregate used for human capital to ensure consistency across assets. value of the human capital stock for each country. 204 1,147 surveys were used from the I2D2 database. 205 52 surveys were used from the Global Labor Database (https://worldbank.github.io/gld/). 206 382 surveys were used from the Luxembourg Income Study (https://www.lisdatacenter.org/). 207 https://www.healthdata.org/research-analysis/gbd. 208 https://population.un.org/wpp/. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 186 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH To compute human capital in real terms, quality-adjusted Human capital by income group labor force data disaggregated by gender are used as Human capital is a critical component of a nation’s wealth, a “volume” measure. The labor force data are taken accounting for the largest share of wealth for most countries. from the ILO and adjusted for the changing educational On average, human capital constitutes about 60 percent of composition of the labor force or “quality” by multiplying total wealth at the global level, dropping from 61 percent in these volumes by the HCI from the PWT.209 This allows us 1995 to 60 percent in 2020 (Table 9.1). The share of human to approximate the average human capital per worker. This capital in total wealth changes steadily with the level of is a simple adjustment for the quality changes in the labor development—human capital’s share of total wealth generally force over time, since the HCI assumes constant, global increases as countries achieve higher levels of economic Mincerian coefficients and does not account for the quality development. Human capital was greater than 60 percent of of education. Ideally, the labor force data should have more wealth in upper-middle-income and high-income countries detail, disaggregating workers by their gender, years of in 2020, but only about 50 percent in low-income and lower- education, and experience. Experimental estimates for a middle-income countries. subset of countries for which such data exist suggest that this is a reasonable first approximation (for more detail Trends in human capital differ over time between different see Box 9.3). Future efforts should focus on building the income groups. On average, the share of human capital in necessary detail in the labor force data to use as improved high-income countries (including both OECD and non-OECD volume measures. countries) plateaued during 1995–2020, while it substantially increased in all other income groups. This can be explained in part by the share of labor earnings in GDP, which anchors GLOBAL TRENDS IN HUMAN the human capital estimates. Labor earnings as a share of CAPITAL WEALTH GDP and per capita human capital grew rapidly in the 1990s, This section presents the estimates of human capital across but much more slowly since 2000 because of technological countries and trends in human capital over the 1995–2020 change, stagnating wages, and, in many countries, a reduction period. The estimates of human capital are summarized at in the share of the population in the labor force due to aging. the global, income, and regional levels, with an additional But, in many middle- and low-income countries, educational discussion on the self-employed portion of human capital. attainment and returns to education are still growing, and human capital is rapidly increasing. TABLE 9.1 Human capital as a share of total wealth, 1995–2020 1995 2000 2005 2010 2015 2020 World 61.2 60.8 57.5 58.7 62.7 60.2 Low income 30.2 38.9 37.1 37.5 44.5 49.7 Lower middle income 44.3 47.0 47.1 52.4 56.1 54.3 Upper middle income 52.8 56.5 53.6 58.6 65.7 60.9 High income: OECD 63.3 62.0 58.8 59.3 62.4 60.5 High income: Non-OECD 60.8 60.0 55.8 59.6 63.8 64.9 Source: World Bank staff estimates. Note: Shares are computed using human capital wealth measured in current US dollars and are reported as percent. OECD = Organisation for Economic Co-operation and Development. 187 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Inequality in total wealth across income groups extends Human capital per capita in real terms tends to grow faster to human capital as well. Real human capital per capita in as countries’ income increases (Figure 9.1). High-income high-income OECD countries in 2020 was about 65 times that non-OECD countries experience the highest growth rate at in low-income countries, and about 32 times that in lower- 2.33 percent, followed by lower- and upper-middle-income middle-income countries. In high-income OECD countries, countries (1.27 percent and 1.08 percent, respectively) and human capital per capita was close to $331,418 (in chained low-income countries (1.05 percent). The main exception 2019 US dollars), while it was only $5,215 in low-income is high-income OECD countries, which experience countries. This significant disparity between low- and high- considerably lower growth rates at only 0.75 percent due to income countries reflects the difference in incomes. stagnating wages and aging populations. FIGURE 9.1 Annual growth rates of human capital per capita, 1995–2020 Low income 1.05 Upper middle income 1.08 Lower middle income 1.27 High income: OECD 0.75 High income: Non-OECD 2.33 0.0 0.5 1.0 1.5 2.0 2.5 Source: World Bank staff estimates. Note: Human capital wealth is measured in chained 2019 US dollars. OECD = Organisation for Economic Co-operation and Development. REGIONAL TRENDS IN HUMAN share of human capital in total nominal wealth increased CAPITAL from 1995 to 2020 in most regions. For example, South Asia and Sub-Saharan Africa experienced notable growth in the Ranging from about half to two-thirds of total wealth, human proportion of human capital within total wealth. In 1995, this capital makes up the largest share of total nominal wealth in share stood at approximately 48 percent for South Asia and all regions. While human capital’s share is about half in Sub- 44 percent for Sub-Saharan Africa, and rose to 56 percent Saharan Africa and the Middle East and North Africa regions, and 50 percent, respectively, by 2020. In contrast, the share of human capital constitutes 65 percent of Latin America’s human capital in total wealth dropped substantially in North wealth and 63 percent of Europe and Central Asia’s wealth America (by 7 percentage points) and to a lesser extent in (Table 9.2). Human capital’s share in total wealth is between East Asia and the Pacific (by 3 percentage points). 55 percent and 60 percent in other regions. In addition, the REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 188 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 9.2 Human capital as a share of total wealth, by region, 1995–2020 1995 2000 2005 2010 2015 2020 East Asia and the Pacific 59.3 56.2 53.9 54.8 59.6 56.5 Europe and Central Asia 58.0 56.4 54.9 59.1 64.6 62.8 Latin America and the Caribbean 62.1 62.5 61.9 69.1 71.8 65.4 Middle East and North Africa 52.3 51.5 44.3 48.5 57.1 54.8 North America 68.9 69.0 63.8 61.2 63.5 62.0 South Asia 48.0 54.4 55.9 55.5 60.0 56.2 Sub-Saharan Africa 44.4 39.1 40.4 50.6 50.2 50.1 Source: World Bank staff estimates. Note: Shares are computed using human capital wealth measured in current US dollars and are reported as percent. Human capital per capita in real terms has grown Pacific (27 percent), as shown in Figure 9.3. These regions consistently for the past 25 years due to increasing labor have experienced modest growth rates, with 12 percent in force participation and higher returns to education. The North America and 16 percent in East Asia and the Pacific. world’s real human capital per capita increased by 9 percent In contrast, the Middle East and North Africa, and Latin between 1995 and 2020, but there are contrasting trends America and the Caribbean regions show much larger across regions (Figure 9.2). 210 Human capital is concentrated increases of 82 percent and 62 percent, respectively, over in the high- and upper-middle-income countries of North the same period—albeit from a much lower starting point America (34 percent of the global value of human capital), (their shares in the nominal value of human capital are 2 Europe and Central Asia (27 percent), and East Asia and the percent and 5 percent, respectively). 210 The growth rate in global human capital per capita is so low, since there have been small increases in the share of global human capital in South Asia (from 2.6 percent in 1995 to 3.2 percent in 2020) and Sub-Saharan Africa (from 1.1 percent to 1.8 percent). While these changes are small in absolute terms, they are able to depress the global growth rate in human capital per capita due to significant differences in value of human capital relative to richer regions. For example, North America has a nominal value of human capital that is 60 times higher than in South Asia and Sub- Saharan Africa. 189 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH FIGURE 9.2 Human capital per capita, by region, 1995–2020 (1995=100) 190 180 chained 2019 USD (1995=100) Human capital per capita in 170 160 150 140 130 120 110 100 90 1995 2000 2005 2010 2015 2020  East Asia & Pacific Sub-Saharan Africa   South Asia  Latin America & Caribbean Europe & Central Asia  World  North America Middle East & North Africa  FIGURE 9.3 Nominal shares of human capital, by region, 2020 3% 2% 5%  North America 2% 34%  East Asia & Pacific  South Asia 27%  Sub-Saharan Africa  Europe & Central Asia  Latin America & Caribbean 27%  Middle East & North Africa Source: World Bank staff estimates. Note: Real human capital per capita is measured in chained 2019 US dollars. Shares in nominal wealth are measured in current US dollars. GENDER AND HUMAN CAPITAL as their access to economic opportunities and education, are captured in the estimated Mincerian coefficients. Therefore, Since human capital wealth is estimated using the current a difference in human capital across gender and a gender wage, education, and experience profiles for a stock of gap indicates the disparity between men and women in human capital, the productivity differences between men terms of both monetary and non-monetary aspects of human and women are captured by the Mincerian parameters. capital. Unfortunately, the human capital estimates reveal a In addition, the extent to which the non-monetary aspects significant disparity between the male and female shares of human capital impact wages, education, and work of human capital. Moreover, little progress has been made experience, such as the health and skills of workers as well toward greater gender parity in human capital over 1995–2020. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 190 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Although higher levels of economic development are in lower-middle-income countries, they make up less than generally associated with a higher share of women in human one-third of human capital in high-income non-OECD capital, women accounted for just 36 percent of human countries and about one-third of low-income countries’ capital in 2020 (measured in nominal terms), which was only human capital wealth. The share of women is slightly greater 4 percentage points greater than in 1995. than one-third of human capital in upper-middle-income and high-income OECD countries, compromising about 37 While women account for a quarter of human capital wealth percent of human capital wealth. FIGURE 9.4 Shares of human capital by gender, 1995–2020 Panel a: By income group 80 70 60 50 40 30 20 10 0 1995 2020 1995 2020 Male share (% of total) Female share (% of total)  Low income  High income: non-OECD  Lower middle income  High income: OECD  Upper middle income 100 Panel b: By region 90 80 70 60 50 40 30 20 10 0 1995 2020 1995 2020 Male share (% of total) Female share (% of total)  East Asia & Pacific  Middle East & North Africa  South Asia  Europe & Central Asia  North America  Sub-Saharan Africa  Latin America & Caribbean Source: World Bank staff estimates. Note: Shares are computed using human capital wealth measured in current US dollars. OECD = Organisation for Economic Co-operation and Development. 191 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH The differences between regions are even more striking. To capture the magnitude of gender-based disparities As shown in Figure 9.4, women accounted for only 15 in human capital over time, Table 9.3 provides a simple percent of human capital in South Asia in 2020, while 44 measure of the gender gap in human capital, defined as the percent of human capital was attributed to women in Latin ratio of the human capital of women divided by that of men America and the Caribbean. The share of women in Europe in a country. In 2020, the global gender gap in human capital and Central Asia’s human capital wealth was 40 percent was 57 percent, meaning that the remaining gap to close is and slightly below 40 percent in North America and Sub- 43 percent. Although there was progress from 1995 to 2020, Saharan Africa. Women account for less than one-third of global progress was not satisfactory: only 10 percentage human capital wealth in East Asia and the Pacific, and the points. In lower-middle-income and high-income non- Middle East and North Africa. OECD countries, the gender gap ratio is particularly low, below 50 percent. In other words, women’s presence and These results indicate that women’s role in human capital contribution to human capital is still extremely limited tends to increase as countries achieve higher levels of at these levels of economic development. In countries at economic development. This is an expected outcome higher levels of economic development, the gender gap because higher educational attainment, better quality of ratio is higher, but still well below parity. Interestingly, education, higher participation of women in the labor only high-income OECD countries made progress toward force, and more competitive wages are associated with gender equality over 1995–2020, narrowing the gap by 14 economic development. However, as the results suggest, percentage points. In contrast, the gender gap worsened in there is still substantial gender disparity between men and countries at all other levels of development. One possible women, even in high-income countries and regions. There reason why the gender gaps are widening outside of high- are several other factors causing the gender disparity in income OECD countries could be that women’s wages tend human capital, including: to be lower than men’s wages even as women’s labor force participation is increasing. However, further research is ■ Careers that are interrupted for childbearing. needed for a full explanation. ■ Penalties for childcare, as women work part-time to The gender gap in human capital across regions is even meet family needs and as employers question the more noticeable. The gender gap ratio has a wide range, commitment of women to their career. from 18 percent in South Asia to 78 percent in Latin America ■ Preferences on the part of women for occupations that and the Caribbean. South Asia’s large gender gap is mostly may be lower paid, an effect that is often reinforced caused by a male-dominated labor force and many barriers by preferences for fields of study that lead to such that prevent women from attaining similar economic occupations. opportunities as men (World Bank 2023). In contrast, female labor force participation is higher in Latin America and the ■ Barriers that prevent women from attaining similar Caribbean. Although the gender gap ratio is higher in North economic opportunities as men. America and Europe and Central Asia compared with other regions, it is still far from parity, at below two-thirds. ■ A lack of women in leadership positions in the workforce. Gender discrimination fosters and reinforces many of these negative influences on women’s earnings (World Bank 2023; Georgieva, Sayeh, and Sahay 2022). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 192 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH TABLE 9.3 Potential gains in human capital from gender equity, 1995–2020 Gender gap ratio (x100) Potential gain from gender equity (ratio of human capital wealth by gender) (percentage increase from base) 1995 2000 2005 2010 2015 2020 1995 2000 2005 2010 2015 2020 World 47 48 53 56 56 57 27 26 24 22 22 21 INCOME GROUP Low income 48 50 51 51 52 54 26 25 24 24 24 23 Lower middle income 52 36 32 32 34 33 24 32 34 34 33 34 Upper middle income 65 69 64 64 60 58 18 16 18 18 20 21 High income: Non-OECD 45 44 43 41 41 43 28 28 28 29 30 28 High income: OECD 45 46 52 56 57 59 28 27 24 22 21 20 REGION East Asia and the Pacific 30 32 35 42 47 47 35 34 33 29 26 26 Europe and Central Asia 52 53 59 62 63 65 24 23 20 19 18 17 Latin America and the Caribbean 64 72 72 77 76 78 18 14 14 12 12 11 Middle East and North Africa 34 34 33 32 35 37 33 33 34 34 33 31 North America 53 54 58 64 60 63 23 23 21 18 20 19 South Asia 15 14 16 17 17 18 43 43 42 42 42 41 Sub-Saharan Africa 123 93 65 52 65 63 -12 3 17 24 18 18 Source: World Bank staff estimates. Note: OECD = Organisation for Economic Co-operation and Development. The gender gap in human capital can be used to conduct simple human capital (NG) can be estimated as NG = (100 – gender simulations of the gains that could be achieved from greater gap ratio) × 50/100. As shown in Table 9.3, human capital equity in earnings and thereby human capital by gender. worldwide could increase by 21 percentage points with gender Assume that the working age population is equally divided parity. In low-income, lower-middle-income, and high- between men and women, each with a 50 percent share. Then, income non-OECD countries where the gender gaps in human if the earnings of women were on par with those of men, capital are more pronounced, the gains from gender equity women’s human capital would rise considerably. Assuming no would be larger. Meanwhile, countries at all levels of economic decrease in the human capital of men, the resulting gains in development benefit from gender equity. 193 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Because the gender gaps are substantially larger in some The focus in this chapter was solely on human capital as a regions, the gains from gender equity in these regions productive asset that produces a stream of benefits: future are substantial. The region with the largest difference in wages. This is not to deny that education, good health, and human capital by gender is South Asia. If gender parity knowledge are sources of well-being in and of themselves, were achieved in South Asia, this could increase human or that doing a job well is one of the great human pleasures. capital nationally by about 41 percentage points (Table 9.3). Development is about building human capital—this requires These simple simulations do not account for the general direct investment, such as education, but it also requires equilibrium impact that an influx of women in the labor broader investment in a healthy environment, water, market might generate, and thereby tend to overestimate sanitation, and clean air. the benefits that could result from gender equity. Still, the estimates show that major gains in human capital per capita In future work, the methodology could be further could be achieved if women were able to work more and earn improved, most notably the volume measures used in the more, and that deeper analysis is needed on the components estimation of human capital in real terms. The weights driving women’s human capital compared to men. in the current methodology are already incorporating productivity improvements that are reflected in higher nominal values of human capital. However, the volume CONCLUSIONS measures are currently adjusted with a simple correction factor that is not able to approximate the changing This chapter provided a set of comparable estimates of educational composition of the labor force in more detail. human capital based on a time series of household surveys More detailed labor force data are necessary, but global for 151 countries throughout 1995–2020. Human capital coverage of such data is currently limited. Experimental accounts for about 60 percent of total global wealth. On estimates for a subset of countries are presented in Box average, the share of human capital increases with higher 9.3, which illustrates that, while the simple adjustment is levels of development and is highest in high-income and an important first step, more insights can be gained from upper-middle-income countries. Estimates by gender using more detailed labor force data. demonstrate the continued, significant disparity between men’s and women’s human capital, which is greater in some regions than others. Globally, the female share in human capital is slightly above one-third, and progress in closing the gender gap has been slow over the 1995–2020 period. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 194 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 9.3 EXPERIMENTAL ESTIMATION OF QUALITY-ADJUSTED HUMAN CAPITAL T his edition of CWON estimates the value of human capital using the number of people aged 15–65 and their lifetime income profiles based on age, being male or female, and education (as detailed in chapter 2). For the computation of the real human capital estimates in chained 2019 US dollars, the quantity relatives are derived from the number of male and female workforce reported by the ILO’s Department of Statistics adjusted for by the HCI of the PWT.211 This adjustment proxies for the changing educational composition of the labor force or its “quality” by approximating average human capital per worker. However, the HCI is a simple adjustment, as it does not allow Mincerian coefficients to differ across countries and by gender. It also does not account for the quality of education, by simply using years of education in the construction of the HCI, which provides an overestimate of educational attainment, particularly in low-income countries where the quality of education is low.212 To better reflect increased education attainment of the labor force, future editions of CWON could collect detailed labor force data disaggregated by gender and education levels. The advantage of such an approach is that it would directly account for changes in the composition of the workforce and would assign differential weights to each worker based on their educational attainment (with larger weights attributed to higher levels of education as determined by the nominal human capital wealth estimates). For example, if the share of higher-educated workers in the labor force increases over time, the Törnqvist volume index would assign a greater weight to these workers, reflecting their higher level of nominal human capital. This is indeed the approach chosen by the statistical offices of Canada213 and the UK,214 which are the only countries publishing experimental human capital estimates. Unfortunately, there is currently no global database available to provide labor force data disaggregated by gender and education level for all the countries in the CWON database. However, some countries have a complete ILO labor data series. To illustrate how this sophisticated quality adjustment compares to the current estimates, experimental estimates of human capital in real terms were produced for Indonesia and the United States using ILO working age male and female populations disaggregated into four aggregated education levels—lower than basic, basic, intermediate, and advanced quality. These data are used to compute the quantity relatives for eight human capital components (male with lower than basic education, female with basic education, and so on) that are inserted into the Törnqvist volume index formula to compute human capital in real terms. 211 https://www.rug.nl/ggdc/productivity/pwt/?lang=en. 212 The value of human capital in low-income countries is constrained by a lack of produced capital broadening and deepening (the low-income trap) and while they have experienced increases in years of education this translates into modest gains in human capital due to low-quality education and low returns to education. In fact, while years of education have increased in these countries, learning poverty (children under 10 who cannot read) has risen as well. 213 https://www150.statcan.gc.ca/n1/en/catalogue/11F0027M2010062. 214 https://www.ons.gov.uk/peoplepopulationandcommunity/wellbeing/datasets/humancapitalestimatessupplementarytables. 195 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Indonesia and the United States were chosen because they have a similar size of population and labor force, but with a different composition in education attainment. In 2022, the US population with basic education earned 1.25 times more than people with lower than basic education. The population with intermediate and advanced education earned on average 1.4 and 2.6 times, respectively, more than the population with lower than basic education. Indonesia had similar earnings ratios for its different education levels, and these levels are matched with ILO aggregate education definitions. Following the Törnqvist volume index formula, the total human capital nominal value is divided into each education group and weighted by average earnings ratios. The United States, which has a larger and increasing share of population with advanced education levels compared to Indonesia (Figure 9.3.1), has a higher growth rate in real human capital per capita when the more detailed labor force data are used (Figure 9.3.2). On the other hand, for Indonesia, which has a larger and growing share of a working age male population with intermediate and basic education, the human capital growth is revised downwards when using the more detailed labor force data, since it capture more directly the lower returns to education by these types of workers. BOX FIGURE 9.3.1 Real human capital by education level United States 50  Female, advanced Trillion chained 2019 US$  Female, intermediate 40  Female, basic  Female, less than basic 30  Male, advanced 20  Male, intermediate  Male, basic 10  Male, less than basic 0 1995 2000 2005 2010 2015 2020 Indonesia .8  Female, advanced Trillion chained 2019 US$  Female, intermediate .6  Female, basic  Female, less than basic .4  Male, advanced  Male, intermediate .2  Male, basic  Male, less than basic 0 1995 2000 2005 2010 2015 2020 Source: World Bank staff estimates. Note: Human capital wealth is measured in chained 2019 US dollars. These estimates are based on data from ILO, BLS and BPS. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 196 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX FIGURE 9.3.2 Real human capital vs. quality-adjusted real human capital, indexed 140 United States 130 Index (1995=100)  Human capital (HCI-adjusted) 120 Human capital (labor force detail) 110 100 1995 2000 2005 2010 2015 2020 180 Indonesia 160 Index (1995=100)  Human capital (HCI-adjusted) 140 Human capital (labor force detail) 120 100 1995 2000 2005 2010 2015 2020 Source: World Bank staff estimates. Note: Human capital wealth is measured in chained 2019 US dollars. These results suggest that more nuanced insights into trends in human capital can be gained from using more detailed labor force data. However, further research is needed to develop a global labor force database with the necessary detail. There is scope to further improve the computation of the the large differences between men and women’s human capital human capital estimates, including increasing the number of are also important, especially for policy makers. Nevertheless, surveys used and improving the gap-filling approach between even with the data now available, additional analysis as well as surveys. Further research and analysis on the factors driving simulations can be undertaken to inform policy. 197 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH REFERENCES Becker, G.S. 1962. “Investment in Human Capital: A Theoretical Analysis.” Jorgenson, D.W., and Fraumeni, B.M. 1992b. “Investment in Education Journal of Political Economy 70 (5): 9–49. and U.S. Economic Growth.” Scandinavian Journal of Economics 94 (Supplement): S51–S70. Becker, G.S. 1993a. “Nobel Lecture: The Economic Way of Looking at Behavior.” Journal of Political Economy 101 (3): 385–409. Kendrick, J.W. 1976. The Formation and Stocks of Total Capital. New York: Columbia University Press. Becker, G.S. 1993b. Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education. University of Chicago Press. Lange, G.-M., Wodon, Q., and Carey, K., eds. 2018. The Changing Wealth of Nations 2018: Building a Sustainable Future. Washington, DC: World Bank. Boarini, R., Mira d’Ercole, M., and Liu, G. 2012. “Approaches to Measuring the Stock of Human Capital: A Review of Country Practices.” OECD Liu, G. 2011. “Measuring the Stock of Human Capital for Comparative Statistics Working Paper 2012/04, Organisation for Economic Co-operation Analysis: An Application of the Lifetime Income Approach to Selected and Development, OECD Publishing, Paris. Countries.” OECD Statistics Working Paper 2011/06, Organisation for Economic Co-operation and Development, OECD Publishing, Paris. EC (European Commission), IMF (International Monetary Fund), OECD (Organisation for Economic Co-operation and Development), UN (United OECD (Organisation for Economic Co-operation and Development). 2001. Nations), and World Bank. 2009. System of National Accounts 2008. New The Well-Being of Nations: The Role of Human and Social Capital. OECD York: United Nations. Publishing, Paris. Feenstra, R.C., Inklaar, R., and Timmer, M.P. 2015. “The Next Generation of Schultz, T.W. 1961. “American Economic Association.” The American Economic the Penn World Table.” American Economic Review 105 (10): 3150–3182. Review 51 (1): 1–17 Fraumeni, B.M. 2008. “Human Capital: From Indicators and Indices to World Bank. 2006. Where Is the Wealth of Nations? Measuring Capital for the Accounts.” Paper presented at the Organisation for Economic Co- 21st Century. Washington, DC: World Bank. operation and Development workshop, “The Measurement of Human Capital,” Turin, Italy, November 3–4. World Bank. 2011. The Changing Wealth of Nations: Measuring Sustainable Development in the New Millennium. Washington, DC: World Bank. Georgieva, K., Sayeh, A.M., and Sahay, R. 2022. “How to Close Gender Gaps and Grow the Global Economy.” IMF Blog. International Monetary Fund. World Bank. 2018. The Human Capital Project. Washington, DC: World Bank. Hamilton, K., and Liu, G. 2014. “Human Capital, Tangible Wealth, and the Intangible Capital Residual.” Oxford Review of Economic Policy 30 (1): World Bank. 2020. The Human Capital Index 2020 Update: Human Capital in 70–91. the Time of COVID-19. Washington, DC: World Bank. Jorgenson, D.W., and Fraumeni, B.M. 1989. “The Accumulation of Human World Bank. 2021. The Changing Wealth of Nations: Measuring Sustainable and Nonhuman Capital, 1948–1984.” In The Measurement of Saving, Development in the New Millennium. Washington, DC: World Bank. Investment, and Wealth, edited by R.E. Lipsey and H S. Tice, 227–282. Chicago: University of Chicago Press. World Bank. 2023. Women, Business and the Law 2023. Washington, DC: World Bank. Jorgenson, D.W., and Fraumeni, B.M. 1992a. “The Output of the Education Sector.” In Output Measurement in the Service Sectors, edited by Z. World Bank. 2024. The Changing Wealth of Nations 2024: Methods and Data. Griliches, 303–341. Chicago: University of Chicago Press. Washington, DC: World Bank. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 198 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH Conclusions MAIN MESSAGES decisions, evidence and data are needed, especially a headline indicator that can help assess aggregate progress ■ Real comprehensive wealth per capita estimates provide toward the SDGs and sustainability more broadly. an important indicator of sustainable development. However, most countries do not produce the necessary Whether progress is sustainable—that is, whether future underlying data to develop country-level real generations will have at least the same production and wealth estimates. consumption opportunities as the current generation—can be assessed with changes in real wealth per capita, measured ■ To fill this data gap, the World Bank’s CWON program comprehensively to include produced, human, and natural produces comparable and consistent estimates of capital, as well as net foreign assets. Changes in real wealth real comprehensive wealth per capita for more than per capita capture how future production and, ultimately, 150 countries for the 1995–2020 period. These estimates consumption opportunities of a country change over time. are aligned with internationally accepted statistical They provide key insights into the sustainability of current standards and guidelines and are complementary to development patterns—as measured by changes in real wealth headline indicators, such as GDP. per capita and its components—and the policy environment driving them. Constant or increasing real wealth per capita is ■ To transition the CWON program into a regular thus an important indicator of sustainability. statistical program, several conceptual and practical challenges remain. These range from decisions around A key challenge in developing real wealth per capita the appropriate discount rate and lifetime of renewable estimates is that, even when the necessary data are natural capital resources, to improved measurement collected, they are not organized in a way that facilitates of the quality of volume measures, prices, and reporting by national statistical offices on comprehensive selected assets. wealth. For example, the physical measures and economic data necessary for valuation are not easily integrated. A NEW HEADLINE INDICATOR Organizing data, building data-sharing mechanisms, and FOR SUSTAINABILITY: CHANGES providing administrative support are challenges in many IN REAL WEALTH PER CAPITA countries. In other cases, the necessary data are simply not collected. This is especially the case in countries with low Decision-makers are increasingly placing sustainability statistical capacity. While there have been growing calls for and the enhancement of well-being at the center of policy better accounting for the environment and natural capital in discourse. This includes national and international policy macroeconomic statistics, including from the UN Secretary dialogues around the SDGs, 215 the beyond GDP agenda, 216 General,217 the G7218 and G20,219 and leading academics,220 and the importance of mainstreaming nature into decision- implementation is lagging. Though more than 90 national making processes. To help steer these dialogues and policy statistical offices develop natural capital or ecosystem 215 https://sdgs.un.org/goals. 216 https://unsceb.org/topics/beyond-gdp. 217 https://unsceb.org/valuing-what-counts-united-nations-system-wide-contribution-beyond-gross-domestic-product-gdp. 218 This includes by the G7 Communique of 2018 (https://www.international.gc.ca/world-monde/assets/pdfs/international_relations-relations_ internationales/g7/2018-06-09-summit-communique-sommet-en.pdf), the Environment Ministers’ Communiques of 2022 (https://www. bundesregierung.de/resource/blob/974430/2044350/84e380088170c69e6b6ad45dbd133ef8/2022-05-27-1-climate-ministers-communique-data. pdf?download=1) and 2023 (https://www.meti.go.jp/press/2023/04/20230417004/20230417004-1.pdf), and the G7 Science Ministers’ Communique in 2023 (https://www8.cao.go.jp/cstp/kokusaiteki/g7_2023/230513_g7_communique.pdf). 219 This includes the G20 Data Gaps Initiative of the International Monetary Fund (https://www.imf.org/en/News/Seminars/Conferences/g20-data-gaps- initiative), which has identified several climate change indicators and national accounts distributions as critical data gaps for policy. 220 For example, Arrow et al. (2004), Fleurbaey (2009), Guerry et al. (2015), Jorgenson (2018), and Hulten and Nakamura (2022). 199 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH accounts (UN 2023), their coverage is often limited, even while recognizing that those guidebooks reflect what is for OECD countries (see chapter 1). Practically no country internationally agreed as a statistical standard and not measures human capital. necessarily what is needed to fully assess sustainability. It would also allow the program to report summary changes The World Bank’s CWON program aims to fill this data in wealth aligned with each boundary along with a single gap. The CWON database offers the most comprehensive summary (see, for example, OSTP, OMB, and DOC 2023). and consistent wealth estimates currently available, and these estimates are comparable with and complementary The question of what set of assets is appropriately included to other headline macroeconomic indicators, such as GDP. within the CWON measurement boundary remains an area The current update of the CWON database implements of active discussion. Future wealth assessments must resolve an important methodological innovation that affects this to stabilize the asset boundary and focus the work on how real comprehensive wealth is measured. With this updating the database rather than expanding it. This will edition, a Törnqvist volume index is used to compute enhance the usability and comparability of the data across real comprehensive wealth estimates. Changes in real editions. Of course, there will remain a need to occasionally comprehensive wealth per capita measured in this way expand the asset boundary, as understanding of the assets signal whether the productive base of an economy (that is, that contribute to well-being deepens and as the economy, its assets) is growing or shrinking over time due to changes society, and environment evolve. Such expansions will, in the quantity—or volume—of assets available. It also however, be undertaken occasionally rather than continually reflects how the scarcity and productivity of different assets and in a way that maintains maximum comparability of the change relative to others (through relative price changes) as database across time and countries.221 time goes by. This shift to a volume-based index is in line with international statistical guidance and is a significant In addition, several methodological and measurement improvement relative to previous editions of CWON, which concerns remain that need to be addressed—drawing on the used the GDP implicit price index for deflation. However, latest research, international guidance, and implementation there remain several methodological and practical experience—to further improve the real comprehensive challenges to be addressed before the CWON program can wealth per capita estimates. Regarding methodology, more mature into a statistical program producing regular and work is needed to determine the appropriate discount consistent updates of real wealth measures. rate for CWON to address concerns around the limited substitutability of assets and intergenerational equity in access to wealth. In addition, the assumption of a 100-year THE WAY FORWARD FOR THE lifetime for renewable natural resources requires revisiting CWON PROGRAM given the evidence that climate change is affecting the As done in previous editions, the current CWON release has environment and the economy more quickly and seriously expanded the asset boundary, adding assets for which data than anticipated. There are also several measurement sources and methodological approaches are available. To challenges to be addressed, including (i) adjusting asset mature to a regular statistical program, the CWON program volumes in the Törnqvist index to reflect not just changes in would need to be more systematic, defining measurement quantities but also in quality; (ii) producing real estimates boundaries conceptually rather than pragmatically. of wealth and its components in purchasing parity power These boundaries should match those in the SNA and terms; (iii) improving measurement of urban land, SEEA as closely as possible, with selected extensions as timber, renewable energy, aquaculture, and water assets; required to permit the assessment of sustainability (for (iv) improving spatial integration and delineation of the example, for human capital). This would enable the CWON contribution of protected areas for land assets; and (v) program to maintain alignment with the SNA and SEEA, collecting subsidy data for all assets. 221 For all editions of CWON so far including this update, it is not possible to compare databases given that both methods and asset coverage have changed across editions. REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 200 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH THE USES OF THE CWON DATABASE be provided. This unprecedented access will provide users with the opportunity to build on the CWON database and code CWON data are already being used widely by World Bank to customize wealth estimates for specific countries. This task teams, the academic community,222 the private sector,223 may involve using more granular, country-level input data and other multilateral actors. The main use of the CWON and modifying the assumptions used to estimate the nominal data is to conduct sustainability analyses by comparing and real wealth estimates. Several examples of country-level the trends in real comprehensive wealth per capita and its wealth estimates using a similar methodology to CWON have components over time and across countries. As shown in recently been produced (see Box 10.1), showing the feasibility this report, both trends in real comprehensive wealth per of such country-level exercises. capita and its components can provide key insights into the sustainability of current development patterns, as well as The opening up of the CWON input data and code will also into the evolution of the underlying asset portfolio. This facilitate policy and scenario modelling by allowing users analysis can reveal which assets are being degraded too to decide for themselves how to treat key aspects of the fast relative to population growth and which sectors require methodology. For example, in line with standard statistical targeted investment to continue building aggregate real practice, the CWON database provides conservative baseline wealth per capita. Since the CWON database primarily draws estimates of wealth that hold future rents constant. This on global data sources, it is best suited for analysis of trends means the estimates reflect the current policy environment both within and across countries. It lacks the necessary and market expectations and are blind to the possible effects granularity to inform the design of specific interventions at of future policy actions or changes in market conditions the subnational scale. due to, for example, climate change. To explore “what if ” scenarios, researchers will now be able to adapt the CWON With this new release of CWON data, it is possible, for the input data and code to change this assumption—and any first time, to construct customized country-level wealth other aspect of the methodology they wish—in modelling estimates. Previous editions of CWON only made available the exercises to derive their own, policy-contingent, wealth final nominal and real wealth estimates for comprehensive estimates. Several examples of such analysis are under wealth and its components. As part of the World Bank’s way within the World Bank using computable general reproducibility initiative, the entire statistical code used to equilibrium models. Many similar analyses and modelling generate the estimates will be publicly released on the World frameworks are available that users can explore to answer Bank’s website together with the detailed input data. The only their unique policy questions using the latest release of the exception is licensed datasets for which dummy datasets will CWON database. 222 A recent example includes an analysis of the unequal climate impacts on global values of natural capital by Bastien-Olvera et al. (2023). The Institute for Global Sustainability at Boston University also frequently draws on the nonrenewable natural capital data as part of its Visualizing Energy project (https://visualizingenergy.org/). 223 One key example is the Sovereign Environmental, Social, and Governance data portal of the International Finance Cooperation (https://esgdata. worldbank.org/?lang=en), which curates a wide range of data for policy makers, financial market participants, and academic researchers. This database includes CWON data as one of its primary sources of natural capital data. 201 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH MEASURING COMPONENTS OF COMPREHENSIVE WEALTH BOX 10.1 IMPLEMENTATION EXPERIENCE FROM COUNTRY-LEVEL WEALTH ESTIMATES T he International Institute for Sustainable Development (IISD) has played a leading role in promoting the measurement of comprehensive wealth for nearly a decade. Its efforts began with two reports on comprehensive wealth for Canada (IISD 2016, 2018). These reports painted quite a different picture of Canada’s development than that offered by GDP alone. While Canada did well with GDP as the gauge of success, comprehensive wealth analysis showed that success to rest on a shaky foundation. According to the IISD’s figures,224 human capital—Canada’s greatest asset225—was stagnant in per capita terms from 1980 to 2015. At the same time, natural capital was declining, produced capital was overly concentrated in fossil fuel extraction and residential housing, and financial capital was too reliant on holding gains on foreign assets. Building on what it learned from the two Canadian studies, the IISD next tackled the more complex challenge of measuring comprehensive wealth in countries with less advanced statistical systems. With funding from the Canadian International Research Development Centre, the IISD worked from 2020 to 2024 with researchers and experts in Ethiopia, Indonesia, and Trinidad and Tobago to compile comprehensive wealth estimates for those countries. A primary motivation for the project was to determine if comprehensive wealth estimates could be compiled in countries with limited statistical resources using mainly nationally sourced data, following methods and time frames like those used in CWON.226 Although the results of the project remain to be published, three findings are clear. First, it proved possible to compile comprehensive wealth figures for the study countries using data available from national sources (mainly the national statistical offices and central banks). This suggests that comprehensive wealth accounts, while by no means straightforward to compile, should be feasible in almost any country. Second, while the IISD’s results are not identical to those in CWON, there is generally good agreement between them. This suggests that CWON’s global methodology yields credible results for individual countries. Finally, as in Canada, comprehensive wealth analysis reveals features of development in the three countries that are not apparent from GDP data alone: decline of natural capital, highly concentrated produced capital, and lack of investment in human capital. This suggests that comprehensive wealth accounting is worth the effort, as it provides a missing—and urgently needed— perspective on national progress that GDP cannot. 224 The IISD drew the figures for its analysis directly from Statistics Canada, Canada’s national statistical agency, so the results are considered to be as robust as possible given current concepts, methods, and data. 225 Human capital is not just Canada’s greatest asset, but the most important asset in every country. 226 The main exception is the method the IISD used to value human capital, which was a simplified approach based on national accounts labor compensation data. This choice was made not because the lifetime income approach used in CWON was deemed impracticable, but because project resources did not support its application (it is analytically intensive). REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH 202 MEASURING COMPONENTS OF COMPREHENSIVE WEALTH REFERENCES Arrow, K., Dasgupta, P., Goulder, L., Daily, G., Ehrlich, P., Heal, G., Levin, IISD (International Institute for Sustainable Development). 2018. S., Maler, K.G, Schneider, S., Starrett, D., and Walker, B. 2004. “Are We Comprehensive Wealth in Canada 2018—Measuring What Matters in the Consuming Too Much?” Journal of Economic Perspectives 18 (3): 147–172. Long Term. Winnipeg, Canada: International Institute for Sustainable Development. https://www.iisd.org/publications/report/comprehensive- Bastien-Olvera, B.A., Conte, M.N., Dong, X., Briceno, T., Batker, D., wealth-canada-2018-measuring-what-matters-long-term?q=publications/ Emmerling, J., ... and Moore, F.C. 2023. “Unequal Climate Impacts on comprehensive-wealth-canada-2018-measuring-what-matters-long-term. Global Values of Natural Capital.” Nature 1–6. Jorgenson, D.W. 2018. “Production and Welfare: Progress in Economic EC (European Commission), IMF (International Monetary Fund), OECD Measurement.” Journal of Economic Literature 56 (3): 867–919. (Organisation for Economic Co-operation and Development), UN (United Nations), and World Bank. 2009. System of National Accounts 2008. OSTP (US Office of Science and Technology Policy), OMB (US Office of Management and Budget), and DOC (US Department of Commerce). Fleurbaey, M. 2009. “Beyond GDP: The Quest for a Measure of Social 2023. National Strategy to Develop Statistics for Environmental-Economic Welfare.” Journal of Economic Literature 47 (4): 1029–1075. Decisions: A U.S. System of Natural Capital Accounting and Associated Environmental-Economic Statistics. Edited by Office of Science and Guerry, A.D., Polasky, S., Lubchenco, J., Chaplin-Kramer, R., Daily, G.C., Technology Policy, Office of Management and Budget, and Department Griffin, R., ... and Vira, B. 2015. “Natural Capital and Ecosystem Services of Commerce. Washington, DC: The White House. Informing Decisions: From Promise to Practice.” Proceedings of the National Academy of Sciences 112 (24): 7348–7355. UN (United Nations). 2023. Global Assessment of Environmental-Economic Accounting and Supporting Statistics 2022. United Nations Statistics Hulten, C.R., and Nakamura, L.I. 2022. “Is GDP Becoming Obsolete? Division. https://seea.un.org/sites/seea.un.org/files/global_ The ‘Beyond GDP’ Debate” (No. w30196). National Bureau of Economic assessment_2022_background_doc_v4_clean_1.pdf. Research. UN (United Nations) et al. 2021. System of Environmental-Economic IISD (International Institute for Sustainable Development). 2016. Accounting—Ecosystem Accounting (SEEA EA). New York: Comprehensive Wealth in Canada—Measuring What Matters in the White cover publication, pre-edited text subject to official editing. Long Run. Winnipeg, Canada: International Institute for Sustainable https://seea.un.org/ecosystem-accounting. Development. https://www.iisd.org/publications/report/comprehensive- wealth-canada-measuring-what-matters-long-run?q=publications/ UN (United Nations), EU (European Union), FAO (Food and Agriculture comprehensive-wealth-canada-measuring-what-matters-long-run. Organization of the United Nations), IMF (International Monetary Fund), OECD (Organisation for Economic Co-operation and Development), and World Bank. 2014. System of Environmental Economic Accounting 2012— Central Framework. United Nations. 203 REVISITING THE MEASUREMENT OF COMPREHENSIVE WEALTH The CHANGING WEALTH of NATIONS