Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses Jobs Group | World Bank | 2021 Elizabeth Ruppert Bulmer, Kevwe Pela, Andreas Eberhard-Ruiz, Jimena Montoya Photo Credits Photo Credits Cover Photo Thossaphol / iStock Pg 00, 11, 17, 48, 125, 128, 144 Tanmay Kothari Pg 39 Andresr / iStock Pg 53 Prusaczyk / iStock Pg 70 i-Stockr / iStock Pg 97 Syah / iStock Attribution—Please cite the work as follows: Elizabeth Ruppert Bulmer, Kevwe Pela, Andreas Eberhard-Ruiz and Jimena Montoya. 2021. “Global Perspective on Coal Jobs and Managing Labor Transition out of Coal.” World Bank, Washington, DC. License: Creative Commons Attribution CC BY 3.0 IGO. © 2021 International Bank for Reconstruction and Development / The World Bank Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses Jobs Group | World Bank | 2021 Elizabeth Ruppert Bulmer, Kevwe Pela, Andreas Eberhard-Ruiz, Jimena Montoya Acknowledgments This study was carried out by Elizabeth Ruppert Bulmer (Team Leader, Lead Economist), Kevwe Pela (Economist), Andreas Eberhard-Ruiz (Economist) and Jimena Montoya (Consultant), Jobs Group, World Bank. Excellent research assistance was provided by Ami Shrestha, consultant for the Jobs Group. The work was commissioned under the Global Support to Coal Regions in Transition (P171194), led by Michael Stanley (Lead Oil and Gas Specialist, Infrastructure, Energy and Extractive Industry Group). Strategic guidance was provided by Rohit Khanna (Practice Manager, Infrastructure, Energy and Extractive Industry Group) and Ian Walker (Manager, Jobs Group), under the senior leadership of Demetrios Papathanasiou (Global Director, Energy and Extractives) and Michal Rutkowski (Global Director, Social Protection and Jobs). The team expresses particular thanks to the many country team colleagues in Indonesia, South Africa and India who supported this extensive research effort. The team is grateful for thoughtful comments by Sheoli Pargal (IEEXS), Harshit Agrawal and Balada Amor (IEEXI), Abhinav Goyal (ISAE1), Franz Gerner, Mariano Salto and Frederic Verdol (IAEE3), Catrina Godinho (SCCDR), and Michal Hetmanski (energy.instrat. pl). The final report benefited from insightful feedback by peer reviewers Grzegorz Peszko (Lead Economist, SENGL), Achim Schmillen (Practice Leader, HEADR), Zuzana Dobrotkova (Senior Energy Specialist, IEEES) and Erik Caldwell Johnson (Senior Social Development Specialist, SCASO). Financial support from the Energy Sector Management Assistance Program (ESMAP) and the Extractives Global Programmatic Support (EGPS) Multi-Donor Trust Fund is gratefully acknowledged. Through the World Bank Group (WBG), ESMAP works to accelerate the energy transition required to achieve Sustainable Development Goal 7 (SDG7) to ensure access to affordable, reliable, sustainable, and modern energy for all. The EGPS Multi-Donor Trust Fund helps resource-dependent developing countries manage their oil, gas, and mining resources to support poverty reduction and boost inclusive, sustainable growth and development. The findings, interpretations, and conclusions expressed in this work are those of the authors, and do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. Abbreviations ALMP Active Labor Market Policy MOSPI Ministry of Statistics and Programme Implementation ARC Appalachian Regional Commission MT Million tonnes BP British Petroleum NDC National Determined BPS Badan Pusat Statistik Contribution CO2 Carbon Dioxide NEET Not in Employment or in COP 24 Conference of the Parties 24 Education or Training DMO Domestic Market Obligation OECD Organization for Economic ECA Europe and Central Asia Cooperation and Development EIA Energy Information OHS October Household Survey Administration PEG-CPR Prayas Energy Group and Centre EU European Union for Policy Research EUS Employment – Unemployment PEP Poland Energy Policy Survey PGE Polska Grupa Energetyczna GDP Gross Domestic Product PGG Polska Grupa Gornicza GLD Global Labor Database PLFS Periodic Labor Force Survey GHG Green House Gas PLN Polish currency (zloty) GTAP Global Trade Analysis Project PPU Power Production Unit GWh Gigawatt Hour PSE Polskie Sieci Elektroenergetyczne HIC High-Income Country QLFS Quarterly Labor Force Survey IC Intermediate consumption RUEN National General Plan on Energy IEA International Energy Agency SA South Africa IESR Institute of Essential Services STEM Science, Technology, Engineering Reform and Mathematics ILO International Labour TJ Terajoule Organization TWh Terawatt Hour ILOSTAT International Labour Statistics UK United Kingdom IPCC Intergovernmental Panel UMIC Upper Middle-Income Country on Climate Change UNFCC United Nations Framework IPP Independent Power Producer Convention on Climate Change Ktoe Kilotonne of oil equivalent US United States LFS Labor Force Survey WDI World Development Indicators LIC/LMIC Low-Income Country/Lower- WWF World Wildlife Fund Middle Income Country MEMR Ministry of Energy and Mineral Resources Glossary of Terms Coal mining sector The industry category established in the UN’s International Standard of Industrial Classification of All Economic Activities (ISIC) Revision 4, classified under code 05, the 2-digit category for coal and lignite mining, which falls under the 1-digit category for mining and quarrying. Coal mining job Any type of employment (formal or informal) within the coal and lignite ISIC sector classification. Coal sector Economic activity along the coal value chain, including coal and lignite mining, coal-fired power plants, coal transport, steel production, etc. Direct coal mining job Job at a mine or for a mining company whose activity falls under the ISIC sector classification. Includes mining occupations and non-mining management, administrative and support occupations within a mining company. Indirect jobs linked to coal This category comprises: (i) jobs related to the coal supply chain, such as transporters of coal, or jobs that provide goods or service inputs for the extraction of coal or its downstream industrial uses, including in power plants; and (ii) jobs induced by coal activity, such as jobs that produce goods and services consumed by coal mine workers and their families (often referred to as “induced” jobs). Table of Contents Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 01 Chapter 1: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Chapter 2: Global Coal Trends: A Mixed Picture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.1 Coal Production and Consumption Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.2 The Winds of Change Affect the Pace of Transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.3 Snapshot of Coal Employment Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Chapter 3: Coal's Role in Structural Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.1 Coal Consumption Patterns Change with Economic Development . . . . . . . . . . . . . . . . . 41 3.2 Coal's Direct and Indirect Use in Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Chapter 4: Labor Market Implications of Coal Production in Five Country Case Studies . . . . . . . 54 4.1 Direct and Indirect Effects of Coal Demand on Coal and Non-coal Jobs . . . . . . . . . . . . 55 4.2 Poland: Lessons From a Long Transition Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 4.3 United States: Slow Convert to Post-coal Transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.4 Indonesia: Crowding Into the Export Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 4.5 South Africa: Holding It's Ground . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 4.6 India: Producing to Meet it's Own Massive Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 Chapter 5: Policies for Managing the Transition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 5.1 Key Findings on the Magnitude of the Challenge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 5.2 Lessons from Past Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 5.3 Policy Framework for Managing the Labor Impact of Coal Transition . . . . . . . . . . . . 136 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Annex 1: Using Input-Output tables to estimate the coal content of sectors . . . . . . . . . . . . . . . . . . 154 Annex 2: Technical Results for Indonesia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Annex 3: Technical Results for South Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Annex 4: Technical Results for India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Executive Summary The widely-shared objective of transition to cleaner energy and reduced dependence on coal presents tremendous challenges, not only to coal sector producers and workers, but because of the broader implications for other sectors in coal-producing nations. A large proportion of energy infrastructure is built around coal-fired power plants (even in non-coal producing countries), economic production structures are energy-intensive, and coal value chains are long. In regions where coal mining takes place, the effects of transition cut very deeply, especially in small, remote mining communities where the local economy depends on coal. The transition can create multiple disruptions: to jobs – both direct and indirect, to household incomes, to local economies heavily tied into the coal supply chain, to community well-being and social capital, and to local and regional government capacity and fiscal solvency. 01 This issues paper analyzes the status of coal related to displaced mine workers to consider the phase-out around the world, the magnitude wider implications for local labor markets and and character of coal mining jobs and sustainable recovery of regional economies. The their spillovers in local economies, and the policy framework articulated in this global report challenges associated with future labor is intended to guide future country-specific transition. The analysis exploits differences engagements through which detailed policy in transition stages to draw lessons from recommendations could be developed to address countries that have experienced coal mine a particular country or sub-regional context. closures in the past, and uses these lessons to inform policy responses in the context of future At the global level, coal-based energy decarbonization, with particular attention production has risen steadily over the past to facilitating the transition of directly and 40 years, to a large degree driven by rising indirectly affected workers – whether formal energy demand in the industrializing or informal – into alternative employment. economies of the world. Many countries undergoing rapid structural transformation This report is part of a broader multi-sector since 1991 depend on coal. As former coal effort by the World Bank to support coal powerhouses in Europe as well as the U.S. regions confronting the realities of transitioned away from coal and shifted their decarbonization and help lay the groundwork priorities toward alternative sources of power for achieving a just transition for all. The World generation, they have been replaced by rapidly Bank framework of support comprises three scaling coal extraction in other regions of the pillars: institutional governance, people and world. Increased electricity consumption is the communities, and environmental remediation main component of this energy demand, and and repurposing land and assets. By focusing on coal is the largest fuel source for electricity pillar two, this paper deepens existing analysis worldwide. The developing world more than and extends the policy discussion beyond issues doubled its per capita electricity consumption since 1990. Figure 0.1 Energy consumption by source (1985-2019) 1 Consumption 0.8 Non-fossil Fu l 0.6 Sourc of Tot l En r Oil & G s 0.4 Co l 0.2 0 19 19 19 02 02 02 85 85 85 94 94 94 11 11 11 20 20 20 20 20 20 20 20 20 19 19 19 19 19 19 LIC/LMIC UMIC HIC Note: Country income classification on the basis of 1991 WB classification Source: Author's calculations based on BP Statistical Review of World Energy Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 02 Inexpensive coal-based energy has played and an accelerating shift to cleaner and more a prominent role in many countries’ sustainable sources of energy and electricity economic development, especially in the generation. Coal meets nearly half of low early stages of structural transformation. and lower-middle income countries’ energy Structural transformation occurs as jobs shift needs and more than half of their electricity from low-productivity primary sectors into consumption, but coal-intensity declines as higher productivity industry and ultimately country incomes rise. into more skilled services sectors. As low and lower-middle income countries industrialized, The world’s increasing demand for coal they increased both their coal consumption is being met by a shrinking pool of large and their coal dependence. Part of this is due to coal producers. China is dominant – it higher electricity demand and the prevalence accounts for about half of global production of coal-fired power generation, but part stems and consumption – but other countries are from the use of coal-derived products other than increasingly exploiting their coal deposits, electricity in many manufacturing subsectors, and have ramped up coal production activities. such as the steel industry. And in countries that Six countries supply four-fifths of the world’s are coal producers, these effects are magnified, annual coal consumption, marking a dramatic suggesting that access to inexpensive energy change since 1980, when the U.S., Germany, helps to accelerate industrialization. In Poland and Former Soviet Union countries upper middle-income and especially high- were much bigger suppliers. income economies that are in more advanced stages of structural transformation, we This shift in coal production is reflected in observe a decline in coal dependence, due to heterogeneous patterns at the country-level, increasingly services-centered economies and is the result of various factors. There are countries that rapidly expanded coal production, others that saw tepid contraction, Figure 0.2 Global coal production shares 100% Sh r of Co l Production 80% 60% 40% 20% 0% 1980 1985 1990 1995 2000 2005 2010 2015 2020 Chin Unit d St t s Austr li South Afric Oth r Top Countri s Indi Indon si Russi G rm n R st of th World Source: BP Statistical Review of World Energy June 2020 03 and still others that experienced periods in Figure 0.3), partial transitioners (dotted of sharp fluctuations in both directions. green), accommodators of rising domestic Some coal producers faced stiff competition demand (dotted red), and expanding coal from oil and gas, or headwinds from tighter exporters (solid red). Some countries have government regulations to curb carbon phased out of coal mining, or at least to a emissions. Some countries were motivated by significant degree, reflecting a commitment technology-induced productivity increases, or to transition (with the caveat that strategic national objectives related to energy “commitment” may not be perfect or security or local employment preservation. may experience setbacks or fluctuating Some countries pursued new export markets political will). This group includes the United as the global coal landscape shifted. Some Kingdom, Germany, Poland, Czech Republic, countries expanded production of coking coal and Ukraine. Other countries have more used in steel production and other chemical recently moved in the direction of a cleaner manufacturing processes. energy mix, notably Romania, Canada, Greece, and the U.S. The reasons for the The world’s top 20 coal-producing countries delayed shift appear linked to internal rather share some common features, and can be than external factors, including recent categorized into 4 groups: advanced coal declines in domestic coal demand. The transitioners (denoted by a solid green line tremendous production increases in China Figure 0.3 Coal production trends in the top 20 coal producers 1980-2020 (million tonnes) Chin Indi Unit d St t s Indon si Austr li 4000 800 1100 600 500 3000 600 1000 400 400 900 2000 400 300 800 200 1000 200 700 200 0 0 600 0 100 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Russi n F d r tion South Afric G rm n K khst n Pol nd 450 250 500 140 300 400 400 120 250 200 350 300 100 200 150 300 200 80 150 250 100 100 60 100 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Turk Colombi C n d Vi tn m C ch R public 80 100 80 50 120 80 40 70 100 60 60 30 60 80 40 20 40 20 50 10 60 20 0 40 0 40 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Gr c Ukr in Rom ni Th il nd Unit d Kin dom 70 200 70 25 150 60 150 60 20 100 50 15 50 100 40 10 50 40 50 30 5 30 0 20 0 0 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Source: BP Statistical Review of World Energy June 2020 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 04 and India were primarily driven by the rising indirect demand for local goods and services energy needs of their large and fast-growing by coal mine workers and their families domestic economies, whereas Indonesia and (often referred to as induced effects). On the Australia, among others, have been motivated other hand, the high wages earned by mine by export opportunities. employees – much higher than most other sectors, both on average and when controlling The total number of workers directly for individual characteristics – can distort local engaged in coal and lignite mining is wages in other sectors, effectively crowding currently 4.7 million globally, accounting out economic activity and depressing labor for a very small and declining share of total demand. In addition, the boom and bust cycles employment, even within the major coal typical of extractives industries tend to limit producing countries. Despite expanding coal economic diversification in coal regions, making production, coal jobs are being shed; over 2 local economies vulnerable to large demand million coal mining jobs have been lost in the swings that undermine long-term growth. last decade. This aggregated picture reflects These natural resource curse effects are well- coal phase-out in some countries, expansion in documented in the literature, and are illustrated others, and sector productivity gains in most in this report’s country-level analysis. Evidence countries, as extraction technology has become from Indonesia shows the distorting impact of more capital-intensive. Not surprisingly, China coal mining jobs, namely that well-paid coal accounts for the largest number of jobs in the jobs spurred job creation in other sectors and coal mining sector, numbering around 3.2 pulled up their wages to some degree, but at million in 2018, more than double the sum of the same time these positive spillovers were coal mine jobs in all other countries combined. in fact smaller in very coal-intensive districts, India is the next largest coal employer, at which also experienced relatively slower wage 416,000 direct coal mining jobs, followed growth in non-coal sectors. by Indonesia (240,000) and Russia (150,000). Several countries’ coal employment levels are The report examines five countries in detail in the range of 75,00-110,000 – specifically to understand how their coal production South Africa, Poland, Vietnam, and Ukraine – patterns link to coal employment patterns, while Australia, Colombia, Turkey, and the U.S. and some of the factors behind the observed each employ nearly 50,000. Note that these data country-level differences. These deep-dives do not reflect employment in the coal sector examine the effects of coal jobs on local labor value chain beyond mining. markets and in the broader national labor market context, exploring the extent to which Whereas the level of coal mining jobs is coal employment contributes to or works modest, they generate significant indirect against better job outcomes and stronger jobs across economic sectors and have a economic development. The analysis sheds disproportionate influence on local labor light on the complexities associated with markets. Although not easily measured past and present coal production and using available data, coal mining jobs have a employment outcomes in different country positive impact through high job spillovers contexts. The selected countries – Poland, in other sectors due to increased economic U.S., Indonesia, South Africa and India – activity along the coal supply chain (e.g., in represent the four different categories of complementary activities) as well as through our typology of coal producers. 05 Figure 0.4 Five country studies si Po l n do nd In Exp ndin Adv nc d So Export rs Tr nsition rs ut h Af r ic Un it dS di t In t Dom stic D m nd P rti l s Accommod tors Tr nsition rs The country case studies illustrate that many segment of coal employees, there are many coal mining jobs are of good quality, but not informal own-account and micro-enterprise all. The types of occupations, contract terms, workers engaged in the sector who lack compensation and working conditions can written contracts or other protections, earn vary widely between formally and informally very low incomes and are highly vulnerable employed coal mine workers. Formal coal to demand fluctuations. The case study on mining jobs tend to be highly paid and well- India highlights this segment of informal coal regulated, due to their hazardous nature, sector workers. The significant segmentation and in some countries are highly unionized evident in coal sector employment implies and/or in the public sector. They tend to quite disparate outcomes with respect to involve semi-skilled production and machine job quality, and calls for differentiated operation occupations, which in other sectors policy interventions in the context of future are remunerated substantially less. Even transition associated with coal phase-out. large formal mining companies employ workers on informal contracts, however; Two-thirds of the world’s top coal producing these could be deemed semi-formal from the countries shed coal mining jobs in the last perspective of occupation or pay, even if they decade, including countries with rising coal do not benefit from labor code protections, output. Similar to the heterogeneity observed union representation, or access to severance/ in coal production patterns, coal employment pension benefits or social insurance. manifests disparate trends across countries Indonesia’s coal sector saw a proliferation and over time. Differences in coal type, of small mine operations concentrated in extraction methods and technologies affect rural districts with limited opportunities for the size and skills-mix of the coal sector labor waged employment; coal mining jobs were a force in each country. Non-coal factors also relatively attractive option. In addition to this affect the size and nature of coal sector jobs, Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 06 such as the skills composition, wages and Past episodes of coal transition in Poland availability of alternative work opportunities and the U.S. provide some useful lessons in other economic sectors, mining operators’ for policymakers and local authorities who agility to adjust to demand fluctuations, anticipate future coal phase-out. Although the relative mobility of workers to shift many of these experiences were negative, they between different jobs, and governments’ are nevertheless informative. policy stance toward transitioning away from coal. It is notable that even in countries that Transition takes a long time. When many •  aggressively expanded coal production – for workers, businesses and communities are example, China and India – productivity implicated, fundamental change to an gains in the coal industry have resulted in industry cannot happen quickly, even with significant labor shedding. the best advance planning and post-closure transition policies in place. Figure 0.5 Coal mining employment trends in 6 countries Pol nd Unit d St t s Chin 500,000 180,000 7,000,000 450,000 160,000 6,000,000 400,000 140,000 350,000 5,000,000 120,000 300,000 100,000 4,000,000 250,000 80,000 3,000,000 200,000 60,000 150,000 2,000,000 100,000 40,000 1,000,000 50,000 20,000 0 0 0 00 09 06 09 03 07 01 01 89 85 18 19 19 95 15 13 93 12 20 20 20 20 20 20 20 20 20 20 20 20 20 20 19 19 19 19 Indon si Indi South Afric 300,000 1,000,000 100,000 900,000 90,000 250,000 800,000 80,000 700,000 70,000 200,000 600,000 60,000 150,000 500,000 50,000 400,000 40,000 100,000 300,000 30,000 200,000 20,000 50,000 100,000 10,000 0 0 09 08 04 03 07 01 10 18 18 16 19 15 13 13 13 11 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Note: Employment level measured on the y axis. Employment data includes formal and informal workers employed in the coal and lignite mining sectors. Sources: Poland data from energy.instrat.pl; US Bureau of Labor Statistics; China Coal Technology & Engineering Group; Indonesia LFS (Sakernas); India EUE and PLFS; South Africa LFS. 07 Transition requires a comprehensive •  The advantages of inducing voluntary •  approach with complementary initiatives, job separations through generous policies and incentives to sway the many compensation packages are offset by the actors along the coal value chain, including risk of inflicting long-term damage on those with vested interests like utility local economies. High reservation wages monopolies and manufacturers of dampen local labor demand and economic mining equipment. recovery through diversification, which can undermine public fiscal health. The timing and speed of transition are •  subject to political economy dynamics. Severe social dislocation and local •  Uncertainty around commodity prices economic viability may pass a point of no makes it difficult for communities to return. The risk is higher where long-term transition because prices affect both dependence on coal has delayed acceptance willingness and capacity to diversify toward of transition. other industries. Where actors are public (e.g., Poland), governments have the power Economic diversification is essential and •  to act quickly but risk the future support requires help from both local and higher of the electorate. Where actors are private level government with respect to planning but unions are strong and/or regulatory and financial resources. Advance planning, authority is weak or captured by private investment in infrastructure, addressing interests (e.g., the U.S.), boom/bust cycles can environmental degradation and attracting be exacerbated, which could create obstacles private investment are key ingredients to both the design and implementation of of economic diversification, requiring effective transition policies. significant local and regional institutional capacity and coordination. Transition assistance programs targeting •  formal mine workers fall short of Recent developments in the coal industries meeting the needs of informal workers of Indonesia, South Africa and India share in and around the mines. Even large some common themes, and especially some mine operators employ a significant share common factors affecting the path and of their workforce on temporary and/or speed of transition. These include: rising informal contracts. Informal coal sector market demand for coal – whether domestic workers are at greater risk than their formal (India) or external (South Africa, Indonesia) – counterparts and less equipped to weather to meet electricity needs; costly replacement income shocks. of coal-based technologies with renewable sources; limited economic diversity in coal Remoteness and small market size are •  communities; weak regulation and capture by mutually reinforcing impediments to vested interests (including within the public transition. When communities are not sectors in India and Indonesia, and unions in connected to larger markets, workers cannot South Africa); political economy pressures that access jobs elsewhere and local businesses are limited by their small local client base. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 08 shape government decision-making; and the chain – fall beyond the reach of many policies. potentially disruptive impact on livelihoods The World Bank’s three-pillar framework for and the economic viability of coal communities. supporting energy transition in coal regions Even among countries committed to articulates labor policies to help displaced coal transitioning away from coal, the marginal mine workers navigate the lay-off process cost of continued coal extraction to power and access retraining and other assistance to electricity generation is much lower than the ease the transition to alternative employment. cost to replace installed generation capacity. In the present paper, we use the lessons from past transitions together with the case The outsized impact of coal mining jobs in study findings on coal-related labor market small and/or remote communities makes challenges in Indonesia, South Africa and India them vulnerable to significant dislocation to motivate the design of a comprehensive, in the event of mine closure, which poses a multi-channel policy framework for managing risk of destabilizing local economies. Energy coal transition. The policy framework presented transition in coal regions will impact workers here extends the World Bank (2018a) framework directly engaged in mining operations and by incorporating a broader group of affected along the coal supply chain, but also workers workers, such as informal coal mine workers, with indirect connections to coal activity, such those employed in coal supply chains, and as retail, restaurants, and recreation service those within coal communities that may suffer providers to coal miners and their families. negative economic shocks due to mine closure. In this context, government planning will be essential to mitigate the negative effects To achieve an effective and just transition on livelihoods and the sustainability of for all, it will be necessary to address the local economies. Where coal is an important informal and formal segments of the employer, political considerations can affected workforce through a combination delay the energy transition and resulting of local and national policies and programs. mine closures, but delays may in fact The concept of “just transition” extends to increase existing distortions and exacerbate national priorities of inclusive, sustainable and segmentation, making future transition even broad-based economic growth. Understanding more challenging. the potential welfare losses by workers is only part of the challenge; weighing the trade-offs Addressing these challenges effectively and risks of prioritizing some stakeholders requires a solid understanding of the scope over others is the fundamental task of and nature of the potential impacts of strategic policy design. Given the complex transition. Policymakers need to understand systems of implicit- and cross-subsidy of the ways in which a future transition away energy generation and its links to industrial from coal may affect the livelihoods of sector production and jobs, it is important both coal and non-coal workers and their to understand who currently benefits from surrounding communities, in order to these existing systems, and the economic and implement policies and programs for managing fiscal costs and benefits associated with these transition effectively. Policy design is further systems. A just transition is one in which the complicated by the fact that informal workers – costs and benefits are shared more equitably. an important segment of the coal sector value 09 Traditional labor policy instruments that There are five main channels through which support the transition of displaced workers public policies and programs can facilitate to new jobs are necessary but not sufficient. workers’ transition: In addition to extending the World Bank’s coal transition policy framework to address (i)  emporary income support (e.g., employer T all types of affected workers, this paper also severance pay, national social safety net) incorporates complementary policies for ensuring a sound environment that fosters Increasing workers’ capacity to qualify (ii)  economic diversification. Income support is for jobs in new sectors (e.g., through skills an effective tool for smoothing consumption or entrepreneurship training) in households affected by job loss; it also helps to sustain demand for local goods and  onnecting workers to potential (iii) C services and the businesses that provide employers (e.g., through job search them. Temporary income support such as assistance, mobility grants) through the national safety net should be the minimum policy response for affected informal Stimulating private sector labor (iv)  workers. Although income support can address demand and local or regional business immediate and short-term needs, longer-term development (e.g., through investment interventions are needed to help workers move incentives aligned with strategic into alternative employment – whether local national, local and/or regional priorities, or elsewhere – and to create an environment matching grant programs); and conducive to business development and private job creation. Ensuring the business environment and (v)  labor regulations are conducive to private sector investment and job creation. Figure 0.6 Five policy channels to support transition Incom Support P ssiv LMPs Compl m nt r Incr sin Work rs’ C p cit Sust in bilit R ul tions n Jobs ALMPs Busin ss & L bor — Ensur includin s f t n ts, busin ss clim t Consist nc Gr Conn ctin Work rs to Jobs ALMPs Stimul tin Priv t Job Cr tion Supportiv infr ., duc tion curriculum, public-priv t -CSO p rtn rships Coordin t cross multipl l v ls of ov rnm nt, priv t s ctor, CSOs, communiti s Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 10 11 A sustainability lens could be added to these Phase 3 begins with the announcement policy channels to ensure that workers displaced of mine closure and layoffs, and requires from coal sector jobs do not simply transition communicating the various types of assistance to alternative but equally unsustainable that will be made available to workers and sectors. Introducing sustainability criteria providing support services such as benefit would also support the parallel objective of eligibility advice or career counseling, with stimulating green economic transition. the goal of empowering individual workers to prepare for and shape their own post-layoff These policy channels are relevant across transitions. different phases of the transition; the policy framework developed in this report is Phase 4 comprises the delivery of post- organized into four phases, ranging from layoff assistance including temporary before the mine closure decision is taken income support to displaced workers and through to the period following layoffs and implementation of active labor market closure. The motivating objectives of this programs. A key aspect will be monitoring framework are to enhance the welfare of program take-up and effective job placements affected workers and promote the medium- to enable timely program adjustments to term viability of local and/or regional improve effectiveness. economies. Government’s role in the transition process Phase 1 focuses on broader economic needs to be multi-faceted and proactive. development planning to lay the groundwork A well-planned and systematic process of for absorbing the negative economic shock coal mine closure and layoffs is essential of mine closure. This entails measures to for supporting the reallocation of affected enhance the capacity and resilience of the workers to alternative jobs and at the same local economy through diversification toward time mitigating the economic, social and new economic sectors and occupations, and political costs of transition. Governments do requires upstream planning, significant not have to deliver everything themselves, but investment, close coordination with national they do need to provide strategic direction and authorities, and partnership with a range of leadership, coordinate across stakeholders, local, regional and national CSOs and private arbitrate competing interests, and mobilize sector organizations. adequate financing that represents an investment in transition. Phase 2 comprises pre-closure analysis of the labor situation, including the number and profiles of workers likely to be affected, and assessing existing programs available to affected workers, including safety net coverage and qualifying criteria for passive and active labor market policies. Any safety net or ALMP program adjustments or regulatory reforms need to be implemented prior to the announcement of layoffs. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 12 Figure 0.7 Policy framework for managing labor transition Fost r div rsific tion throu h busin ss clim t r forms, ntr pr n urship initi tiv s Inv st in supportiv ph sic l nd di it l infr structur incl. nh nc d conn ctivit 1 Ali n duc tion nd tr inin curricul tow rd m r in conomic s ctors (n tion l not just loc l; STEM plus soft skills) Economic Est blish p rtn rships nd str t ic ov rsi ht bodi s comprisin n tion l/r ion l nd loc l D v lopm nt ov rnm nt offici ls, c d mic institutions, priv t s ctor ssoci tions, civic or ni tions Str t nd oth r communit roups (Pr -closur d cision) Str n th n ov rsi ht of nd p rtn rship with min op r tors (to minimi disruption post closur r : work r prot ctions nd l nd/infr structur r purposin ) P rtn r with unions nd/or mplo r or ni tions to d si n nd d liv r st ff r -skillin for post-co l conom Id ntif work rs lik l to b dir ctl nd indir ctl ff ct d b th min closur proc ss, profil xistin skills nd pot nti l mism tch with curr nt l bor d m nd 2 R vi w soci l prot ction nd ALMP pro r ms nd l bor r ul tions r l t d to l offs Ass ss li ibilit for un mplo m nt b n fits nd oth r soci l s rvic s (distin uishin b An l sis work r ch r ct ristics such s , hous hold incom ), id ntif cov r ps, stim t pot nti l n ds for ssist nc & Pl nnin R vis incom support pro r ms, ALMPs nd productiv inclusion int rv ntions to (Pr -closur ) ccommod t s st mic shock nd build c p cit of s rvic provid r nci s B in communic tions/communit consult tions, ccomp ni d b positiv cultur l si n lin round n w soci l contr ct not c nt r d on co l Issu dv nc notific tion of l off Inform min work rs nd communit m mb rs of ssist nc options, off r p ck st r t d 3 to min mplo s nd to oth r work rs ( . ., productiv inclusion, public works) to ncour s lf-s l ction into th b st fit to f cilit t r lloc tion into n w jobs. R quir s ov rnm nt to sc l up its outr ch Announc H lp work rs cl rif th ir b n fit ntitl m nt with min op r tor, union L offs Est blish n twork of work r dvoc t s to promot ssist nc nd st r work rs to & Assist nc ppropri t pro r ms B in provision of c rt in s rvic s ( . ., c r r couns lin , ps cho-soci l outr ch, job s rch ssist nc ) Provid t mpor r incom support 4 Impl m nt ctiv l bor m rk t polici s in ph s s/b s d on scr nin ; ALMPs c n includ job s rch/t chnic l/softskills/ ntr pr n urship tr inin , job s rch r nts, w subsid , Post-l off busin ss incub tor, mobilit r nts Assist nc Monitor ssist nc t k -up nd job pl c m nt; djust pro r m p r m t rs to improv ff ctiv n ss Consid r uxili r s rvic s in r spons to communit n ds Source: Authors’ extension of the (formal) labor policy approaches developed in Fretwell (2017), World Bank (2018a) and Cunningham and Schmillen (2021) 13 CHAPTER 1 Introduction The objective to move toward a cleaner energy mix is widely shared by policymakers, civil society organizations and households worldwide. There are multiple motivating factors behind this objective, including the increasingly urgent climate crisis, national and local pollution concerns, declining competitiveness of the coal industry, fiscal efficiency considerations, and the long-term viability of fossil fuel-dependent jobs. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 14 Despite broad support for this objective, yet as coal-dependent regions across the globe the best path to achieve it is unclear, nor is are waking up to the realities of the climate there agreement on the optimal speed of crisis and governments are committing to travel. The transition away from fossil fuels, mine closures, policymakers and communities and coal in particular, will create dislocation are eager for guidance on the best ways to and require adjusting current structures of approach this complex and sensitive agenda production. Understanding the scope and scale to achieve sustainable transition. of the transition challenge is an essential first step but is in itself daunting. The complexity At the global level, the rationale for of the challenge contributes to inertia transition is well recognized and accepted. by policymakers, as do the large implied Environmental and health concerns stemming economic and social costs. But the costs of from mining activities – toxic for workers, inaction are likely to be much greater in the community residents, and natural assets medium term. – have long been acknowledged, although the local nature of the most easily observed The challenges implicate all sectors in impacts are easy to ignore at the national coal-dependent nations, because of energy level. The accumulating scientific evidence on infrastructure built around coal-fired power human-induced climate change and extreme plants, economic production structures that weather has finally crystallized international are energy-intensive, and national supply attention, but policy action has lagged. The chains related to coal use. For coal-producing 2007 Intergovernmental Panel on Climate regions themselves, the effects of transition Change (IPCC) report made the case that cut very deeply, especially when coal mining human actions were the main contributor to regions are small, remote and dependent global warming (IPCC 2007). The 2014 IPCC on coal. The transition can create multiple report laid out in stark terms the urgency of disruptions: to jobs – both direct and indirect, drastically reducing CO2 and other greenhouse to household incomes, to local economies gas (GHG) emissions, which would require heavily tied into the coal supply chain, to fundamental changes to the way we live community well-being and social capital, and the way we do business (IPCC 2014). and to local and regional government The rationale for decarbonization becomes capacity and fiscal solvency. clearer every time an extreme weather event damages infrastructure or physical assets Although some economies have moved or disrupts economic activity or livelihoods. away from coal, notably the United Kingdom, The frequency of these events is increasing. Spain, South Korea and to a lesser degree Even without the urgency of climate change, Poland and the US, there has been limited the economic costs of carbon-dependence are success in addressing the associated large and projected to become untenable. regional labor market disruptions. Moreover, these disruptions had persistent Mine closure can have potentially large negative impacts on social, human and (negative) demand spillovers in surrounding institutional capital that in some cases communities and regional economies. undermined local economic sustainability. The Energy transition in coal regions will perceived risks related to future mine closures impact workers directly engaged in mining are slowing the decarbonization process. And operations and along the coal supply chain, 15 but also workers with indirect connections especially at the local level, which in the case to coal activity, such as retail, restaurants, of coal regions, tend to be rural and not well and recreation service providers to coal captured in national labor force surveys. miners and their families. In this context, government planning will be essential to The analysis in this report exploits mitigate the negative effects on livelihoods differences in transition stages to and the sustainability of local economies. draw lessons from countries that have Where coal is an important employer, experienced coal mine closures in the political considerations can delay the past, and uses these lessons to inform energy transition and resulting mine policy responses in the context of future closures, especially in settings with decarbonization, with particular attention high union membership. to facilitating the transition of affected workers into alternative employment. This This paper analyzes the status of coal report is part of a broader multi-sector effort mining phase-out around the world, by the World Bank to support coal regions describes the magnitude and character confronting the realities of decarbonization of jobs in the coal mining sector 1 – both and help lay the groundwork for achieving globally and in five detailed country a just transition for all. The World Bank studies, and identifies key challenges framework, elaborated in the 2018 report associated with future labor transition. “Managing Coal Mine Closure: Achieving a Structured as an issues paper, it takes a global Just Transition for All” (World Bank 2018a), perspective on recent coal sector trends comprises three pillars: institutional and the associated coal mining jobs created governance, people and communities, and or destroyed, considers the drivers of coal environmental remediation and repurposing production in various country contexts, and land and assets. Focusing on pillar two, the implications of past and future transition the analysis below deepens existing work and coal mine closures on workers and on and extends the policy discussion beyond local labor markets. Although the extent of issues related to displaced mine workers coal activity may be relatively limited within to consider the wider implications for local the overall national economic context, it labor markets and sustainable recovery of may still have significant direct and indirect regional economies. The policy framework effects on local economies. Understanding developed here is intended to guide future the size and nature of these effects will country engagements through which therefore be important for managing an detailed policy recommendations would effective economic transition following coal be developed to address a specific country mine closures. The measurement challenge is or sub-regional context. non-trivial, however, given data limitations, 1  ote that throughout this paper, the term “coal mining job” refers to employment in the coal and lignite mining sector N as classified under International Standard Industrial Classification (ISIC) Rev.4 industry code 05. This category is also defined to include non-mining occupations (e.g., management or support functions) within a mining company, and excludes employment in coal-fired power generation and other industry categories that may be part of the coal value chain (e.g., coal transport, steel production). These latter activities are considered indirect coal sector jobs. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 16 The remainder of this report is organized processes involving coal and its derivative. as follows. Chapter 2 provides a description Chapter 4 presents five detailed country of coal sector trends at the global level, studies examining the evolution of coal characterizing the top coal producing mining employment within their country- countries based on their patterns of coal specific labor market settings. And Chapter 5 production and consumption, and describing concludes with some lessons from past coal the associated levels of coal mining transitions and a proposed policy framework employment in coal-producing countries. for managing labor transitions in the context Chapter 3 explores the role of coal in economic of future mine closures. structural transformation during the last two decades through a demand-side lens, related to both energy demand and manufacturing 17 CHAPTER 2 Global Coal Trends: A Mixed Picture Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 18 2.1 Coal Production and Consumption Trends Global energy production has climbed steadily 1.8 percent annually between 1990 and 2018, and over the past four decades, and coal has played global per capita energy supply climbed from an increasing role. As global GDP rises and the 1.7 to 1.9 toe in the same period. After oil, coal population grows, economies consume more (including lignite)2 accounts for the second largest energy in their economic activities and as well share of global energy supply (Figure 2.1). Global in their household activities. Rising per capita production and consumption of coal posted robust incomes compounds this effect through higher growth over this period – especially between living standards, driving increased residential 2000 and 2010, during which its share of the demand for electricity, heat and air conditioning. total energy market surpassed 28 percent, before The global demand for energy grew by an average ebbing slightly in the past decade (Figure 2.2). Figure 2.1 Global energy supply from combustible sources (ktoe) 5,000,000 16,000,000 14,000,000 Co l 1,000,000 12,000,000 N tur l G s 4,000,000 10,000,000 Nucl r 8,000,000 2,000,000 6,000,000 Biofu ls 4,000,000 Oil 1,000,000 2,000,000 Tot l combustibl s 0 0 (ri ht xis) 1990 1995 2000 2005 2010 2015 2018 Note: Data excludes electricity and heat trade Figure 2.2 Source: IEA data Global coal production and consumption 160 140 Ex joul s Glob l Production 120 100 Glob l Consumption 80 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 Source: Authors’ calculations based on BP Statistical Review of World Energy 2  Throughout this report, we use the general term “coal” to include lignite as well. 19 Economic activity and industrialization agriculture. The last four decades have seen a across the developing world have been the slowdown in OECD coal demand, but a rapid key impetus behind rising coal demand. expansion of coal use by non-OECD industrial In most countries, the industry sector is the sectors (Figure 2.3). By 2018, industrial largest consumer of coal (final use, excluding demand for coal had increased by 75 percent electricity), accounting for 60 percent on globally, and more than doubled in non-OECD average in 1990, compared to 20 percent for countries, while coal demand from other residential use and significantly less for other sectors declined. sectors such as commerce, public services and Figure 2.3 OECD vs. Non-OECD final coal consumption (ktoe) 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 - 1990 1995 2000 2005 2010 2015 2018 OECD Industr OECD Non-industr Non-OECD Industr Non-OECD Non-industr Source: IEA data Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 20 Electricity represents a major part of With respect to electricity, coal has been total energy consumption, and coal is the traditional fuel source for power the largest fuel source for electricity plants, and remains dominant, despite worldwide. According to IEA data, nearly recent inroads from natural gas, and two-fifths of global final energy consumption to a lesser degree hydro, wind and solar is attributable to oil products, followed PV generation (Figure 2.4). by electricity (one-fifth), natural gas (16 percent), bio-fuels and coal (10 percent each). Figure 2.4 Global electricity generation by source (GWh) 30,000,000 25,000,000 Co l 20,000,000 GWh Oil 15,000,000 N tur l G s 10,000,000 Nucl r 5,000,000 H dro Wind 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Co l N tur l G s W st H dro Sol r PV Wind Oth r Sourc s Oil Biofu ls Nucl r G oth rm l Sol r Th rm l Tid Source: IEA data 21 Developing countries – and particularly factor driving the global results, given China’s China – have been the main drivers of very large economy and population of 1.44 rising electricity consumption. Average billion. India – population 1.37 billion – also global per capita electricity consumption has played an outsized role, as it more than tripled increased by 60 percent since 1990, although its per capita electricity consumption. The less developed countries experienced much US, by contrast, has much higher electricity faster gains (Figure 2.5). China’s per capita consumption on a per capita basis, four times consumption increased nine-fold in the space the global average and 2.6 times higher of four decades, reaching 4.91 MWh per person than China (IEA data). Together, these three in 2018, compared to an average 8 MWh per countries dominate electricity consumption capita in the OECD This remarkable expansion and production (Figure 2.6). in both access and demand is the primary Figure 2.5 Ratio of per capita electricity consumption 2018:1990 10 14 9 12 8 7 10 El ctricit /c pit (MWh) R tio 2018:1990 6 8 5 6 4 3 4 2 2 1 0 0 World OECD Non OECD Chin Indi US Indon si South Pol nd Afric R tio 2018:1990 (l ft xis) El ctricit /c pit (ri ht xis) Source: IEA data Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 22 Figure 2.6 A few countries dominate electricity production and coal consumption 5,000 CHN 4,000 n r tion (billion kWh) 3,000 USA Fossil fu ls l ctricit n t 2,000 IND 1,000 RUS DEU UKR AUT TUR IDN ZAF GBR CAN 0 ROU GRC CZE KUZ POL 80% 85% 90% 95% 100% Co l Consumption (p rc ntil r nkin ) Source: BP Statistical Review of World Energy June 2020 23 Coal production is shifting its regional profile Looking at individual countries reveals and becoming increasingly concentrated in a significant heterogeneity within regions small number of countries. Whereas China is a and over time. Figure 2.8 provides a snapshot big part of the global coal story, other countries of the wide-ranging patterns of country- are also pushing into the extraction and export specific coal output since 1980, reflecting markets with enthusiasm. The ECA region and cases of rapid expansion, tepid contraction, North America, home to the dominant coal as well as instances of sharp fluctuations producers of the last century such as Germany, in both directions in some countries. Some Poland, UK, and US, account for a shrinking coal producers encountered periods of share of global production, squeezed out by stiff competition from natural gas or faced East Asia and the Pacific – notably China, headwinds from tighter regulation as but also Australia, Indonesia, Vietnam and governments responded to negative pollution Thailand – and by India. Only ten countries externalities or global warming concerns. account for 90 percent of global production, Some countries expanded production of reflecting an increasing concentration of the coking coal used in steel production. In coal market (Figure 2.7). many countries, coal mining expanded Figure 2.7 World coal production is dominated by few countries 100% R st of th World Oth r Top Countri s G rm n South Afric 80% Russi Austr li Indon si Sh r of Co l Production 60% Unit d St t s Indi 40% Chin 20% 0% 1980 1985 1990 1995 2000 2005 2010 2015 2020 Chin Unit d St t s Austr li South Afric Oth r Top Countri s Indi Indon si Russi G rm n R st of th World Source: BP Statistical Review of World Energy June 2020 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 24 during periods of surging global demand, induced productivity increases, strategic such as during the oil-crisis of the late 1970s national objectives around domestic industry when coal prices became more attractive, targets3 or the coal export market4 or local or as coal deposits were identified. Other employment preservation5, weak institutional motivating factors include technology- settings, or a combination of these factors. Figure 2.8 Coal production trends in the top 20 coal producers 1980-2020 (million tonnes) Chin Indi Unit d St t s Indon si Austr li 4000 800 1100 600 500 3000 600 1000 400 400 900 2000 400 300 800 200 1000 200 700 200 0 0 600 0 100 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Russi n F d r tion South Afric G rm n K khst n Pol nd 450 250 500 140 300 400 400 120 250 200 350 300 100 200 150 300 200 80 150 250 100 100 60 100 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Turk Colombi C n d Vi tn m C ch R public 80 100 80 50 120 80 40 70 100 60 60 30 60 80 40 20 40 20 50 10 60 20 0 40 0 40 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Gr c Ukr in Rom ni Th il nd Unit d Kin dom 70 200 70 25 150 60 150 60 20 100 50 15 50 100 40 10 50 40 50 30 5 30 0 20 0 0 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 1980 2000 2020 Note: Production data from BP Statistical Review of World Energy June 2020. Country classifications defined by the authors as follows: green solid line indicates an advanced stage of transition to coal phase-out; red solid line indicates expanding coal exporters; dashed line means production responds to local demand; red dashed line means increasing production to meet rising domestic demand; green dashed line indicates partial transition that either stalled or reflects very recent transition in response to declining local demand. Source: BP Statistical Review of World Energy June 2020 3  E.g., China’s massive domestic industrialization agenda. 4 E.g., Indonesia’s and Australia’s entry into the rapidly expanding East Asian and South Asian markets.  5  E.g., state-owned coal mining firms in Eastern Europe that have scaled back production without significant job cuts, or politically connected private coal companies like Murray Energy in the US, which lobbied for government support in exchange delaying mine closure (insideclimatenews.org 2019). 25 Despite differences, production trends production increases in China and India were in the top 20 coal producing countries primarily driven by the rising energy needs nevertheless share some common of their large and fast-growing domestic features. Coal producing countries can be economies, whereas Indonesia and Australia, categorized into 4 general groups: advanced among others, have been motivated by coal transitioners, partial transitioners, export opportunities. accommodators of rising domestic demand, and expanding coal exporters. Figure 2.8 Coal export patterns are especially revealing uses color coding to denote each category: in explaining coal production trends. solid green denotes advanced transitioners, Whereas China and India have sharply solid red denotes expanding exporters, and expanded their coal production, both have dashed lines indicate those falling in the become large net importers in response to intermediate categories. This organizing their rapidly increasing internal electricity framework may not perfectly capture each demand. Vietnam and Thailand also consume country’s experience, and sometimes the far more than they produce, and continue distinctions between categories are fuzzy, to expand coal production to accommodate but the framework provides insight into domestic demand. These patterns contrast key drivers of country-level trends. Some sharply with the small but powerful group of countries have effectively phased out of coal expanding exporters, most notably Australia, mining, or at least to a significant degree, Indonesia, and Russia, which together account reflecting a commitment to transition (with for seven-tenths of global exports. Between the caveat that “commitment” may not 2000 and 2019, Australia doubled its coal be perfect or may experience setbacks or exports, Indonesia’s coal exports quadrupled, fluctuating political will). This group includes and Russian coal exports grew by a factor of the United Kingdom, Germany, Poland, five (Figure 2.10). South Africa meets about Czech Republic, and Ukraine. Other countries a quarter of the Africa region’s total import have more recently moved in the direction demand, but focuses more intensively on of a cleaner energy mix, notably Romania, the large India market. Less dominant but Canada, Greece, and the US. The reasons for still important in export markets are the the delayed shift observed in these countries US (exporting mainly to Europe), Colombia appear linked to internal rather than external (exporting to Europe and Latin America) factors, including relatively recent declines in and Canada (targeting Asia). domestic coal demand associated with energy efficiency gains and increasing adoption of renewable energy (Box 2.1 describes the range of competing market-related factors underlying coal production patterns in the US over the past hundred years). The tremendous Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 26 Box 2.1 A Century of Coal Production in the US: Market Drivers of Gradual Change Coal’s dominance in the US energy mix a century ago has undergone a series of changes over time, reflecting both headwinds and tailwinds that affected coal demand and production levels. During the Great Depression through the mid-1950s, coal production contracted as demand shifted toward less expensive oil. Appalachian coal – primarily from the Pennsylvania/Ohio/West Virginia/Virginia/Kentucky corridor – gained competitive advantage during this period due to falling transport costs and proximity to population centers on the east coast. The economic boom period of the late 1950s through the 1960s saw a sustained increase in energy demand and coal production, and this was followed by a rapid shift toward mining in the western US, concentrated in the Powder River Basin in Wyoming and Montana. Technology advances, higher-quality coal, and the lower production costs associated with western strip mining provided the main impetus behind the geographical shift away from Appalachia, and a significant and steady decline in coal mining employment (Figure 2.9). Wyoming coal mines had 8 times the labor productivity of the average Appalachian mine (Lobao er al. 2021). Moreover, coal from the Powder River Basin is less polluting than Appalachian coal, due to its lower sulfur, mercury and arsenic content, making it a more appealing fuel source under the environmental regulations introduced during this period, notably the 1972 Clean Water Act, the 1970 Clean Air Act, and the 1990 Clean Air Act. By the mid-2000s, the emergence of fracking contributed to a natural gas boom and very low energy prices, disrupting the coal market as consumers and power generators substituted toward natural gas. This coincided with the end-of-life phase-out of many coal-fired power plants, although expanding global demand for coal helped to sustain US mining jobs. US coal exports grew three-fold between 2002 and 2012, as the share of US coal exports in total US coal production rose from 5 to 15 percent (BP Statistical Review of World Energy 2020). By the mid-to-late 2010s, renewable energy had become increasingly accessible and affordable, and is likely to soon surpass coal-fired electricity (EIA 2020). Figure 2.9 US Coal Mining Employment and Production: 1919-2017 1,200,000 1,000,000 800,000 600,000 400,000 200,000 0 1919 1929 1939 1954 1963 1972 1982 1992 2002 2012 2017 US Co l Emplo m nt US Production (1000 tons) Source: U.S. Census Bureau, Census of Mineral Industries, as presented in Lobao et al. (2021) 27 Figure 2.10 Coal exporters (2019) Mon oli +++ C n d + US + C n d Oth r Afric Columbi ++ Indon si +++ Europ US Austr li Columbi Chin Russi +++ Europ Indon si Russi Mon oli Chin -- Oth r Asi P cific Oth r CIS + Oth r CIS South Afric + South Afric R st of World Austr li ++ Note: + indicates increased exports between 2000 and 2019; ++ indicates more than doubled coal exports between 2000 and 2019; +++ indicates more than quadrupled coal exports between 2000 and 2019; -- indicates reduced exports by more than half between 2000 and 2019. Source: Data from BP Statistical Review of World Energy June 2020 2.2 The Winds of Change Affect the Pace of Transition The situation is changing, but not to the global total). Indonesia has steadily ramped extent needed to mitigate the intensifying up its generating capacity over the last decade climate crisis. Although coal consumption and a half, adding 27,000 MW. The US also has expanded in most regions since 1981, ECA installed new power plants – especially and North America saw net contractions, and between 2009 and 2013, adding over 25,000 the remaining regions at least slowed the rate MW of coal-fired generation through 2015 but of increase in the most recent decade (Figure zero thereafter. South Korea and Japan were 2.11). In contrast to these positive trends, not far behind, adding 25,000 and 24,000 however, is evidence of expanding coal-fired MW respectively, including new construction power generation, even among countries as recently as 2019 and 2020 (Global Energy signaling a commitment to transition away Monitor). It is important to note that the rate from coal dependence. China and India are at which old power plants are being retired has by far the biggest actors, together adding 1.15 accelerated over the last decade, such that in million MW of coal-fired power generation net terms, coal power additions are declining, between 2000 and 2020 (four-fifths of the although still positive (Figure 2.12). Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 28 Figure 2.11 Coal consumption growth by region EAP ECA 1981-1990 LAC 1991-2000 MENA 2001-2010 NA 2011-2019 SA SSA World -10% 0% 10% 20% 30% Source: BP Statistical Review of World Energy June 2020 Figure 2.12 Coal power plant additions and retirements 100K 80K 60K 40K w tts Av r 20K M Additions 0K R tir m nts -20K -40K 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 Note: Line represents net added MWs. Source: Globalenergymonitor.org 29 Coal activity is a main contributor to today. Air pollution from Europe’s coal-fired greenhouse gas (GHG) emissions and power plants together are estimated to cause pollution that exacerbate climate change nearly 23,000 premature deaths6 per year and damage human health. Nearly half of all across Europe, stemming disproportionately CO2 emissions stem from coal, two-thirds of from Poland and Germany, and with which through electricity and heat generation. significant cross-border effects (WWF CO2 emissions have climbed steadily over European Policy Office et al. 2016). Additional the past three decades (Figure 2.13), during health impacts are measured in increased which time the emissions from coal-fired incidence of chronic bronchitis, asthma electricity and heat generation doubled. attacks in children, and pollution-related China earns top billing as the world’s largest hospital admissions. There are numerous emitter, accounting for a quarter of global GHG other negative externalities associated with emissions (Climate Watch data). The US and EU coal mining that are not addressed here, (27) rank second and third, followed by India, such as land degradation and destabilization, Russia, Japan, Brazil, Indonesia, Iran and and ecosystem disruption, among others. Canada (Figure 2.14). Despite the significant Together, these inflict serious and lasting reduction in coal mining activity in much of damage on human and environmental health the EU, and especially Poland and Germany, that will be costly to remediate. the negative health effects remain evident Figure 2.13 Global CO2 emissions by energy source (Mt) 40,000 Oil 35,000 30,000 25,000 Co l 20,000 Mt CO2 15,000 Oil 10,000 5,000 N tur l G s 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Co l Oil N tur l G s Oth r Source: IEA data 6 Air pollution – and specifically particulate matter – increases deaths attributed to stroke, heart disease, lung cancer and respiratory diseases. WHO (2016) estimates that 4.5 million deaths per year are due to ambient (outdoor) air pollution. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 30 Figure 2.14 Top 10 GHG emitters contribute over two-thirds of global emissions (2018) Note: Preliminary global greenhouse gas emissions data for 2018 excludes land-use change and forestry (LUCF). Source: climatewatchdata.org (graphic by Johannes Friedrich) 31 Future prospects for global coal production Despite the heightened awareness of are uncertain but are unlikely to change the negative environmental and health quickly. In the short-term, coal consumption externalities associated with coal production levels are not likely to decline significantly, and coal-fired energy generation, various given the country-level consumption and factors impede a faster pace of transition. export patterns described above. China and These include the high upfront costs to replace India’s growing economies will continue to or retrofit installed coal-fired power plants demand coal, whether for electricity or as a with renewable energy, and the incentive to manufacturing input (e.g., steel), and many wait until the lifespan end of existing power new coal mines are either planned or under plants. Strong global coal demand has spurred construction in both countries (Figure 2.15, countries like Australia, Indonesia, and Russia left panel). Russia, Australia and South Africa to invest in additional coal production to are also actively planning for additional coal increase their export market shares in the mines, which will exacerbate CO2 emissions still-strong European market (as the supply into the future (Figure 2.15, right panel). of European-sourced coal wanes) and the Figure 2.15 Proposed new coal mines and their annual CO2-equivalent emissions Numb r of n w min s ( s of J nu r 2021) Annu l CO2 quiv l nt missions (Mt) Chin Chin Russi Austr li Austr li Russi Indi Indi South Afric South Afric C n d Mo mbiqu Unit d St t s Indon si Indon si C n d Mo mbiqu Unit d St t s Mon oli Mon oli Turk B n l d sh Botsw n Pol nd Pr -construction Co l combustion Colombi K khst n Construction M th n Pol nd Turk 0 20 40 60 80 100 120 140 0 500 1000 1500 Note: Methane emissions based on 20-year horizon. Source: Globalenergymonitor.org Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 32 rapidly expanding East and South Asian and competitive pricing of renewable energy markets. Producers of steel and other coal- technologies have made green alternatives derived products have to balance domestic accessible to a large and increasingly global industry interests within the broader context market. Policies and regulations that of a globally declining coal market. Energy sanction industrial pollution – including security considerations may be inducing some power plant emissions – are being adopted countries to adopt a longer timeframe for more widely, although their effectiveness is their national transition strategies.7 Policies undermined by weak compliance and large that explicitly or implicitly subsidize coal or implicit subsidies. The intensification of the electricity provide a competitive advantage climate crisis is stimulating more aggressive over other energy sources and induce over- policy effort, but countries’ carbon reduction consumption and over-production. The targets reflected in their voluntary nationally direct and indirect costs associated with determined contributions (NDCs) under the these subsidies – calculated as the difference Paris Agreement are neither sufficient nor between the price that consumers pay for coal are they being met, according to the UN’s and for electricity and the real supply cost recent NDC Synthesis Report (UNFCCC 2021). of coal and electricity, taking into account Financing of carbon-intensive or carbon- environmental costs and foregone tax revenue dependent investments continues, despite – are extremely high, globally accounting nominal commitments by international for over US$2 trillion in 2015 (equivalent financial institutions and global investment to 2.8 percent of global GDP; Coady et al. banks to end future financial support. The 2019). Whereas the direct “pre-tax” coal and widespread economic disruption caused electricity subsidies (using IMF terminology) by the COVID pandemic lockdowns in 2020 are negligible in most countries, the “post- and 2021 highlighted the vulnerability of tax” subsidies are orders of magnitude economies dominated by informal jobs or greater; and for coal, the largest share of these low-productivity service sector employment, indirect costs is due to air pollution, followed as significant shares of workers lost their by global warming. Coady at al. (2019) estimate jobs, small businesses went bankrupt, and that China’s coal subsidies approached US$1.3 global economic production fell by up to 10 trillion in 2015, compared to around US$200 percent. The resulting abrupt albeit temporary billion in Russia, the US, and India. reduction in pollution and congestion during lockdown raised awareness of the damaging Forces for change are gaining traction. effects of our carbon-dependent economic Each coal-consuming country faces different structures; it is possible that this shock may constraints and choice sets shaped by a range create momentum and political will to allocate of domestic and external factors, and this significant public resources to sustainable and heterogeneity in turn affects the timing and resilient recovery. pace of coal transition in each. Some common threads are starting to emerge, however. Market forces in terms of innovation, scaling 7 Peszko et al. (2020) explore potential climate strategies for countries dependent on fossil fuels.  33 2.3 Snapshot of Coal Employment Trends range of 75,00-110,000 – specifically South Africa, Poland, Vietnam, and Ukraine – while Coal production is an important source Australia, Colombia, Turkey, and the U.S. each of employment in the top coal-producing employ close to 50,000 (figures reflect 2019 or countries, although modest compared to most recent data available; Table 2.1). other economic sectors. On the basis of national employment data for the largest Over the last decade, 2.4 million coal mining coal producers , the total number of workers jobs have been lost worldwide in net terms, currently engaged in the coal mining sector reflecting coal phase-out in some countries, is 4.7 million globally (see Box 2.2 for a expansion in others, and sector productivity discussion of data sources and challenges).8 gains in most. Downscaling due to significant This level represents a very small share of productivity gains in China resulted in 1.8 total employment, averaging 0.24 percent million lost coal mining jobs between 2008 and across the 20 top coal producing countries. Not 2018, and coal mining jobs in India declined surprisingly, China accounts for the largest by half. On the other hand, coal mining number of coal mining jobs, numbering around employment increased in many of the other top 3.2 million in 2018, more than double the sum of coal producing countries, notably in Indonesia, coal jobs in other countries combined. India is but also in Australia and South Africa. Even the next largest coal employer, at 416,000 coal in countries that significantly expanded coal mining jobs, followed by Indonesia (240,000) production, employment growth did not and Russia (150,000) (Figure 2.16). Several keep pace, reflecting large productivity gains countries’ coal employment levels are in the (denoted by the black data points in Figure 2.16). Box 2.2 Data challenges for measuring coal mining employment over time Global datasets on employment do not have a very long period of coverage compared to datasets on economic production and population statistics. As governments increasingly standardized their household-level survey instruments to the international standards established under the leadership of the ILO, and as countries carried out more frequent labor force surveys, more complete coverage became available. The ILOSTAT database contains sector-level employment information for most countries beginning in 1991, but aggregated at the 1-digit ISIC industry code level. This means that coal and lignite activities are reported in combination with other mining and quarrying activities. Beginning around 2008, ILOSTAT data reports coal and lignite mining activities separately from other mining and quarrying. Coal-sector disaggregated data for European countries became available in the mid-2000s, reported by the European Commission’s Euro Stat, and subsequently also collected by UNIDO, but these data are not perfectly consistent with ILOSTAT figures, even in later years. In order to consider longer time trends, it is necessary to rely on country-specific micro datasets on labor outcomes (as we have done for Indonesia, South Africa and India; see Chapter 4) or on alternative administrative or secondary sources, which tend to be available only for the historically large coal producers. 8  his figure captures coal and lignite mining employment in the top 20 coal-producing countries; data are for 2019 or T most recent available (2018 for China and Indonesia; 2017 for Czech Republic, India, Ukraine and Vietnam; 2016 for Kazakhstan and Thailand). Employment data include formal and informal workers. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 34 Figure 2.16 Coal mining employment and sector productivity change by country (2008-2019) 5,000,000 200% 4,500,000 150% 4,000,000 3,500,000 100% 3,000,000 Emplo m nt l v l Productivit ch n 2,500,000 50% 2,000,000 0% 1,500,000 1,000,000 -50% 500,000 0 -100 ic n nd n d li d in bi bl n c di si n i tio ic l n r k in dom US m r n Ch m pu In n s t ol fr i r tn us t ol o rm Gr do k h P m r A T u k U K in A C C R G In Ro d ut h Th Vi ch F d K n So t C s i U ni s Ru 2008 2019 Productivit ch. (ri ht xis) Note: Data are from 2008 and 2019 or closest year available. Productivity measured as coal production (in thousand tonnes) per coal sector worker; percent change compares 2019 to 2008. Sources: Labor data from Australian Bureau of Statistics, Statistics Canada, Colombia’s Gran Encuesta Integrada de Hogares, MINSTAT, ILO, India EUE and PLFS, Indonesia LFS (Sakernas), Poland data from energy.instrat.pl (coal mining company employment), ROSSTAT (Russian Federation Federal State Statistics Service), South Africa LFS, UK Department for Business, Energy and Industrial Strategy, US Bureau of Labor Statistics; production data from BP Statistical Review of World Energy 2020 35 Coal accounts for a declining share of total Table 2.1). Extractive industries expanded their mining activities, even in countries with share of total output in the mid-to-late 2000s, rapidly expanding coal production. Total as coal gained importance in the mineral- mining and quarrying employment in our coal-natural gas-oil extractives mix. But in 20-country sample was around 12 million the years since, coal’s share in total extractives in 2019, but coal’s share of these mining and output returned to the historical levels of the quarrying jobs fell from 24 percent in 2008 mid-1990s (WDI data). to only 17 percent today (excluding China9; Table 2.1 Coal employment in top 20 coal-producing countries Coal & Lignite/ Coal & Mining & Quarrying Mining & Lignite/Total Coal & Lignite Employment Employment Quarrying Employment 2008 2019 2008 2019 2019 2019 Australia 30,142 50,368 180,812 251,659 20% 0.4% Canada 5,095 7,845 55,105 74,245 11% 0.0% China 5,000,000 3,209,000 5,400,00 4,140,000 78% 0.4% Colombia 34,620 44,338 167,512 182,293 24% 0.2% Czech Republic 24,024 15,064 0.3% Germany 47,626 14,932 107,460 71,607 21% 0.0% Greece 6,852 3,496 16,953 11,064 32% 0.1% India 795,176 416,240 2,849,133 1,828,969 23% 0.1% Indonesia 108,210 240,041 1,077,800 1,690,150 14% 0.2% Kazakgstan 34,035 29,686 201,990 279,531 11% 0.3% Poland 136,608 92,601 229,227 206,086 45% 0.5% Romania 38,143 27,055 102,076 64,356 42% 0.3% Russian Federation 168,800 150,100 1,331,573 1,651,398 9% 0.2% South Africa 66,206 74,827 339,833 418,994 18% 0.5% Thailand 861 59,995 73,242 1% 0.0% Turkey 51,950 47,955 113,478 152,607 31% 0.2% Ukraine 305,867 110,822 618,132 448,384 25% 0.6% United Kingdom 6,157 699 132,235 135,088 1% 0.0% US 86,300 47,700 626,656 623,717 8% 0.0% Vietnam 102,541 86,399 0.2% Note: China data from 2008, 2018; Colombia 2009, 2019; Czech Republic 2010, 2017; India 2009, 2017; Kazakhstan 2012, 2016; Russian Federation 2010, 2019; Thailand 2016; Turkey 2009, 2019; Ukraine 2012, 2017. Sources: Coal employment data as Figure 2.17; mining and quarrying employment data (unless available in original source listed in Figure 2.17) from ILO; total employment data from ILO (except Colombia) 9  oal accounts for about three-quarters of employment in China’s mining and quarrying sector (mining and quarrying C employment data is from China’s National Bureau of Statistics; coal employment data is from China Coal Technology & Engineering Group). Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 36 Countries that have transitioned away from employs less than a thousand in the sector, coal production – namely, the advanced Germany’s coal mining employment is under transitioners – experienced significant coal 15,000, and in Poland, where mining activities mining job losses long before the recent are ongoing, total coal mining employment contraction in the last decade. Looking back is around 93,000. The US trajectory has been to the 1980s, coal mining employment was relatively gradual, dictated more by market over 416,000 in Poland, 365,000 in Germany, forces, disperse private ownership, and new and 172,000 in the UK; in these countries, open-pit investments in the western region governments took aggressive phase-out which displaced labor-intensive underground measures to close mines and shed a significant mines in Appalachia. The move to alternative share of workers (Figure 2.17). Today, the UK energy sources has not heretofore generated Figure 2.17 Long-term coal mining employment trends in selected countries Pol nd Unit d Kin dom Chin 500,000 200,000 7,000,000 450,000 190,000 6,000,000 400,000 160,000 350,000 140,000 5,000,000 300,000 120,000 4,000,000 250,000 100,000 3,000,000 200,000 80,000 150,000 60,000 2,000,000 100,000 40,000 1,000,000 50,000 20,000 0 0 0 99 81 11 17 87 05 93 00 06 09 03 07 01 89 18 19 95 15 13 12 20 20 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 19 19 Indi Austr li South Afric 1,000,000 70,000 100,000 900,000 90,000 60,000 800,000 80,000 700,000 50,000 70,000 600,000 40,000 60,000 500,000 50,000 30,000 400,000 40,000 300,000 20,000 30,000 200,000 20,000 10,000 100,000 10,000 0 0 0 88 84 96 04 92 00 18 08 12 08 04 03 07 01 18 10 16 19 13 13 20 20 19 19 20 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 Sources: Australian Bureau of Statistics, China Coal Technology & Engineering Group, India EUE and PLFS, Poland data from energy.instrat.pl, South Africa LFS, UK Department for Business, Energy and Industrial Strategy 37 significant absorption of displaced coal workers, This rather mixed snapshot of coal mining in large part due to skills and geographical employment at the global level suggests mismatch between coal mining activities and significant country-level heterogeneity; wind/solar/natural gas supply and generation. coal employment dynamics therefore need to be analyzed at the country level – and For countries that continue to expand coal even at the sub-national level – to gain a production, some have had productivity more comprehensive understanding. Like improvements that reduced demand for coal production trends, coal employment labor inputs, while others increased labor manifests disparate trends across countries demand. The degree to which the demand for and over time. Heterogeneity in types of coal labor in coal and lignite mines has increased or and extraction technologies affects the size decreased varies by country, as does the labor- and skill-composition of the coal sector labor intensity associated with different types of force in each country. Other non-coal factors coal and the extraction methods used. China, are also determinant, such as the composition for example, added two million coal mining of economic sectors and the distribution of jobs between 2000 and 2013, but subsequently employment across these sectors, the size of shed nearly 3 million jobs while tripling its the coal sector relative to other sectors, and productivity. Coal mining jobs in India fell governments’ commitment to and progress from 890,000 in 2004 to 416,000 by 2017, and toward a post-coal transition. In Chapter 3, we productivity more than doubled. In Australia, consider the role of coal in observed patterns by contrast, a period of steady decline in the of structural economic transformation, and in 1980s and 1990s reversed direction in 2000, Chapter 4, we present detailed analysis of coal after which nearly 34,000 jobs were added to mining employment within specific country the coal mining sector. In South Africa, coal labor markets for five country examples. sector contraction in the early 2000s gave way to robust job creation; over the past 15 years, 43,000 coal mining jobs have been added and productivity declined by 50 percent. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 38 39 CHAPTER 3 Coal’s Role in Structural Transformation Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 40 3.1 Coal Consumption Patterns Change with Economic Development Coal’s share of global energy consumption income countries not only increased their coal has remained steady in recent decades, but consumption; they also increased their coal increased strongly in emerging economies. dependence. Figure 3.2 illustrates that coal And the pattern is similar with respect to meets nearly half of low and lower-middle coal-fired electricity. Coal consumption income countries’ energy needs, compared to patterns were relatively flat from 1985 to less than twenty percent in richer countries. around 2000, both in aggregate and across Much of this is due to LIC/LMIC investments countries at different levels of development. in increasing coal-fired electricity generation But the last two decades saw a sharp rise in capacity to meet rising electricity demand of coal consumption by low income and lower- their large and expanding populations and middle income countries concurrent with their growing economies. At the same time, rapid GDP growth rates,10 while higher income higher income countries have been phasing coal demand stagnated or declined (Figure 3.1). out coal power plants in favor of alternative As upper middle-income and especially high- sources; UMICs have largely shifted toward income economies transitioned to cleaner oil and gas sources, and HICs are increasingly and more sustainable sources of energy and investing in non-fossil fuel power generation electricity generation, low and lower-middle (Figure 3.3). Figure 3.1 Coal consumption by country income group (1985-2019) 120 100 80 Ex joul s 60 40 20 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 LIC/LMIC UMIC HIC Note: Country income classification on the basis of 1991 WB classification Source: Authors’ calculations based on BP Statistical Review of World Energy 10  ote that because country income classifications are based on 1991 incomes, countries such as China, Chile, Ecuador, N India, Malaysia, Romania, Peru and Turkey fall under the LIC/LMIC category. In total there are 18 countries that had transitioned from LIC/LMIC status to UMIC or HIC status by 2019. 41 Figure 3.2 Source of energy consumption by country income group (1985-2019) 1 consumption 0.8 Non-fossil Fu l 0.6 Oil & G s Sh r of tot l n r 0.4 Co l 0.2 0 19 19 19 02 02 02 85 85 85 94 94 94 11 11 11 20 20 20 20 20 20 20 20 20 19 19 19 19 19 19 LIC/LMIC UMIC HIC Note: Country income classification on the basis of 1991 WB classification Source: Authors’ calculations based on BP Statistical Review of World Energy Figure 3.3 Electricity fuel sources by country income group (1985-2019) 1 n r tion 0.8 Non-fossil Fu l 0.6 Sh r of tot l l ctricit Oil & G s 0.4 Co l 0.2 0 19 19 19 02 02 02 85 85 85 94 94 94 11 11 11 20 20 20 20 20 20 20 20 20 19 19 19 19 19 19 LIC/LMIC UMIC HIC Note: Country income classification on the basis of 1991 WB classification Source: Authors’ calculations based on BP Statistical Review of World Energy Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 42 Countries experiencing faster GDP coal-based energy. In several fast-growing growth since 2000 tended to increase economies – namely Vietnam, Indonesia, their dependence on coal during the same and Pakistan – coal consumption grew faster period. We observe a positive (non-causal) than GDP. Note that China is an outlier both correlation between GDP growth rate and for its rapid GDP growth and because its rising coal intensity, and this relationship coal intensity fell between 2000 and 2017 holds for both coal producing countries as oil intensity increased. There are many and non-coal producers (Figure 3.4). India, factors underpinning the observed positive Vietnam and Indonesia were the fastest- correlation between GDP growth and coal- growing coal producing economies between intensity, which we do not analyze here; 2000 and 2017 (when we exclude China), and rather, we examine how the role of coal each saw an intensification of coal energy changes during the process of structural dependence. Greece, Spain, Great Britain, economic transformation, and the concurrent and the US posted the lowest GDP growth impact on jobs. during this period while shifting away from Figure 3.4 Correlation between GDP growth and coal-intensity of energy consumption 2 IDN PHL MYS mix, 2000-2017 1 VNM JPN CHL KOR PAK BGD IND MEX EST LTU NLD ARG THA 0 CHE NOP NZI IRN EGY ITA FRA GRE RSE ISL BAR in co l sh r in n r LGM COL PRT AUT LUX DZA SVK DEU SVN ZAF LVA ROU CAN RUG MAR TUR ESP HRG HUN IRL BEL USA CHN -1 DNK GRC GBR AUS CZE POL -2 0 1 2 3 4 2017 GDP incr s r l tiv to 2000 l v l Non-Co l Produc r Co l Produc r Source: Authors’ calculations based on BP Statistical Review of World Energy and WDI data 43 In broad terms, structural transformation better quality12, and higher productivity jobs occurs as jobs shift from low-productivity pay more. The emergence of increasingly primary sectors into higher productivity sophisticated government services and industry and services sectors. As economies regulations adds the formality dimension, develop from low per capita income levels and which brings labor protections, better working low-productivity structures of production conditions and social insurance coverage in to higher value-added activities, they shift formal wage employment. away from primarily agriculture-based production in unpaid or own account work Most economies that experienced rapid toward more capital-intensive production structural transformation in the period based in firms and employing waged labor. since 1991 were relatively coal-intensive. Subsistence producers become more market- For a sample of 91 countries, we compare oriented, selling their surplus production and/ the rate of economic transformation by or transforming it into processed goods sold decomposing labor productivity gains into to consumers. As firms specialize and become within-sector productivity gains and across- more productive, they expand operations to sector productivity gains as employment meet a wider client base, invest in capital and shifts into higher productivity activities. hire more labor that is increasingly specialized We consider the rate of change over three and skilled. Industrial activity requires separate periods: 1991-2000, 2000-2009, services as inputs to production, and at the and 2009-2018. Some countries made faster same time, wage workers in industrial sectors across-sector productivity gains (shown in the consume services that they did not require bottom right quadrant of Figure 3.5), others as subsistence producers. As the share of the made faster within-sector gains (upper left labor force engaged in agriculture declines, the quadrant), and those that had both types share of wage employment rises. of gains (in the upper right quadrant) are considered to be the fastest transformers, Transformation also occurs as jobs within benefiting from the dynamic interaction of the the same sector become more productive. two. There are more growth episodes falling Firms upgrade production technology and into this upper right quadrant, especially product quality, while new firms may enter among coal-intensive countries (defined as the same industry, introducing innovations. those where coal contributes over 20 percent Both of these are examples of within-sector of total energy needs). Some of these are productivity gains. Labor mobility is essential coal producing countries that benefited from for workers to change sectors and move the availability of cheaper energy, but many between firms within the same sector, and are not coal producers. When we consider this labor flow drives down productivity only LIC countries13, the pattern is even differences.11 Firm-based wage jobs are of stronger: coal-intensive countries (denoted 11  ee Jobs Diagnostic background note “Structural change, growth and labor market dualism in developing economies” S (Jobs Group, World Bank, forthcoming) for a more extensive discussion. 12  hat is, when traditional measures of quality are used, such as compensation and worker protections; T this may not be the case when large negative environmental or social externalities are present. 13  Country income category defined based on 1991 status. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 44 by light blue dots in Figure 3.6, panel a) had coal producers14 experienced relatively faster more high-high growth episodes compared structural transformation (Figure 3.6, panel b). to non coal-intensive-countries. A similar Note that we also observe many coal-intensive pattern emerges when comparing structural and non coal-intensive countries in the low- transformation in coal producing countries to low category. non-coal producers; LICs that were also large Figure 3.5 Decomposing structural transformation patterns in 91 countries, 1991-2018 12 hi h within hi h within & low cross & hi h cross 10 dd d p r work r 08 06 d rowth in v lu 04 02 Within-s ctor nnu li 0 -02 -04 low within low within & low cross & hi h cross -06 -02 -01 0 01 02 03 04 Across-s ctor nnu li d rowth in v lu dd d p r work r Non Co l-Int nsiv Co l-Int nsiv Note: High/low thresholds defined on the basis of above/below median productivity gains for within- and across- components respectively between 1991 and 2018. Countries defined as ‘coal-intensive’ when coal contributes more than 20% of total energy needs at start of episode. Note: Each dot corresponds to a country growth episode, with episodes defined over the following periods: 1991-2000, 2000-2009, 2009-2018. Source: Authors’ calculations based on WDI data and BP Statistical Review of World Energy. 14 Those among the top 20 coal producing countries identified in Chapter 2.  45 Figure 3.6 Decomposing structural transformation patterns in 28 low-income countries, 1991-2018 ( ) Co l-int nsiv vs. non co l-int nsiv 12 hi h within hi h within & low cross & hi h cross 10 rowth in v lu dd d p r work r 08 Within-s ctor nnu li d 06 04 02 0 -02 -04 low within low within & low cross & hi h cross -06 -02 -01 0 01 02 03 04 Across-s ctor nnu li d rowth in v lu dd d p r work r Non Co l-Int nsiv Co l-Int nsiv (b) Co l produc rs vs. non co l produc rs 12 hi h within hi h within & low cross & hi h cross 10 rowth in v lu dd d p r work r 08 Within-s ctor nnu li d 06 04 02 0 -02 -04 low within low within & low cross & hi h cross -06 -02 -01 0 01 02 03 04 Across-s ctor nnu li d rowth in v lu dd d p r work r Not L r Co l Produc r L r Co l Produc r Note: High/low thresholds defined on the basis of above/below median productivity gains for within- and across- components respectively between 1991 and 2018. Countries defined as ‘coal-intensive’ when coal contributes more than 20% of total energy needs at start of episode. Note: Each dot corresponds to a country growth episode, with episodes defined over the following periods: 1991-2000, 2000-2009, 2009-2018. Source: Authors’ calculations based on WDI data and BP Statistical Review of World Energy. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 46 Countries undergoing more rapid structural productivity growth (bottom of Figure 3.7). change also experienced higher growth We would expect more high-income in energy consumption per worker. economies to have low across-sector It is interesting to note that the fastest productivity growth (that is, being transformers (high within, high across) in an advanced stage of structural increased their per-worker coal consumption, transformation), so this finding is consistent while countries with low across-sector with evidence that advanced economies productivity gains reduced their coal are transitioning away from coal toward consumption during the period, especially alternative energy sources. those with low within and low across Figure 3.7 Change in energy consumption by fuel source and pattern of structural transformation hi h within & low cross hi h within in & hi h cross T p of l bor productivit low within & hi h cross low within & low cross -0.01 0 0.01 0.02 Annu li d % ch n in v . n r consumption p r work r Co l Oil & G s Non-fossil Fu l Note: High/low thresholds defined on the basis of above/below median productivity gains within- and across- components respectively between 1991 and 2018. Source: Authors’ calculations based on WDI data and BP Statistical Review of World Energy 47 3.2 Coal’s Direct and Indirect Use in Manufacturing The observed linkages between coal and was associated with an influx of labor to the labor productivity are largely driven by manufacturing sector from less productive growth patterns in the manufacturing agriculture or services activities (such sector. All countries that experienced as own account production or informal moderate-to-rapid productivity gains on personal services or retail jobs). This account of labor reallocations into industry finding is underscored by comparing the and/or within industry improvements manufacturing employment share across intensified their coal-based electricity use countries of different income levels and or coal-fired combustion, while productivity different resource intensities. We expect to gains on account of the services sector were observe a concave relationship, namely rising less strongly associated with increases manufacturing share in the early stages of in coal-based electricity intensity. Most economic development followed by declining manufacturing activities require energy manufacturing share as high-income inputs. Growth in manufacturing employment economies shift increasingly into skilled and coal consumption move hand in hand15 services. This “graduation” pattern indeed (Figure 3.8), and especially growth in labor- holds for both non coal and coal-intensive intensive low-productivity manufacturing. countries, albeit with higher manufacturing This suggests that the rapid structural employment shares in the latter (Figure 3.9). transformation in coal-intensive countries 15 This positive correlation also holds for other sectors.  Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 48 Figure 3.8 Correlation between coal consumption and manufacturing employment 8 6 s BGD 2017 co l consumption incr VNM r l tiv to 2000 l v l 4 IDN PHL PAK 2 MEX IND CHN CHL THA LTU KOR VPG MAP TUR HKG EST BRA COL EGY NIG IRN ABC EUR IRN 0 SWE FIN THE ISU SHEIRV SVK POL GAN PAN FAN USA HUN CZE RCU LUX DNK GRC DZA GBR BEL LVA CYP -2 -0.1 -0.05 0 0.05 0.1 in mpl. sh r of m nuf cturin s ctor, 2000-2017 Fitt d v lu s Source: Authors’ calculations based on BP Statistical Review of World Energy and WDI data Figure 3.9 Both coal-intensive and non coal-intensive countries eventually shift away from manufacturing 0.2 pr dict d mplo m nt in m nuf cturin s sh r of tot l mplo m nt 0.15 0.1 0.05 0 4 6 8 10 12 p r c pit GDP (2010 const nt USD), lo . sc l Non Co l-int nsiv Co l-int nsiv Note: Countries defined as ‘coal-intensive’ when coal contributes more than 20% of total energy needs at start of episode. Source: Authors’ calculations based on WDI data, World Bank Cross-Country Database of Sectoral Labor Productivity (Dieppe and Matsuoka, 2020), and BP Statistical Review of World Energy 49 Coal’s role in electricity generation is not As economies grow and transition to the only driver of this manufacturing link; higher-value production activities, coal’s coal and its derivatives are also used as importance in manufacturing wanes. inputs to production in many manufacturing Historical patterns of early-stage economic sub-sectors, such as steel, chemical and metal transition from agriculture-based production products, and even in light manufacturing. to labor-intensive light manufacturing, Analyzing data from input-output tables for resource-intensive heavy manufacturing and 121 countries in the Global Trade Analysis low-productivity services eventually give Project (GTAP) database (methodology way to more sophisticated manufacturing described in Annex 1) indicates that as an and services that are more human capital- intermediate good, the utilities sector is by far intensive, rely less on coal, and add more the largest consumer of coal, namely for power direct and indirect value to the economy. generation. But coal is also a direct input for The pattern of declining coal inputs to many manufacturing subsectors, notably in manufacturing holds across both coal- and heavy manufacturing dominated by chemicals non-coal producers, although the shares of and metals industries, and an indirect input coal-inputs are higher in coal-producing to many other manufacturing subsectors – countries (illustrated in Figure 3.11). e.g., paper products, processed rice, textiles, wood products, motor vehicles and parts, and processed foods (Figure 3.10). Figure 3.10 Sectors’ use of coal as a direct or indirect input to production (% of intermediate consumption; excludes electricity) Min r l products n c P trol um, co l products F rrous m t ls Ch mic l, rubb r, pl stic products P p r products, publishin M t l products M nuf ctur s n c M t ls n c Proc ss d ric T xtil s M chin r nd quipm nt n c Wood products B v r s nd tob cco products W rin pp r l Motor v hicl s nd p rts V t bl oils nd f ts Tr nsport quipm nt n c El ctronic quipm nt Su r L th r products Food products n c D ir products Bovin m t products M t products n c 0 0.005 0.01 0.015 v r sh r of co l in IC, djust d for indir ct co l consumption (w/o l ctricit ) Source: Authors’ calculations based on GTAP I/O tables Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 50 Figure 3.11 Comparison of indirect coal inputs to manufacturing by country income group . Co l-produc rs LIC/LMIC UMIC HIC P trol um, co l products P trol um, co l products P trol um, co l products Min r l products n c Min r l products n c Min r l products n c Ch mic l, rubb r, pl stic products F rrous m t ls F rrous m t ls F rrous m t ls Proc ss d ric M t ls n c P p r products, publishin Ch mic l, rubb r, pl stic products Ch mic l, rubb r, pl stic products B v r s nd tob cco products M t ls n c Su r Wood products M t l products P p r products, publishin M nuf ctur s n c Su r T xtil s Motor v hicl s nd p rts P p r products, publishin M t l products T xtil s M nuf ctur s n c Proc ss d ric W rin pp r l Food products n c Food products n c M t l products V t bl oils nd f ts B v r s nd tob cco products M t ls n c M chin r nd quipm nt n c Bovin m t products Tr nsport quipm nt n c Wood products D ir products L th r products T xtil s V t bl oils nd f ts M chin r nd quipm nt n c D ir products M nuf ctur s n c Food products n c Bovin m t products M t products n c Bovin m t products B v r s nd tob cco products Wood products Su r Tr nsport quipm nt n c M chin r nd quipm nt n c D ir products Motor v hicl s nd p rts L th r products Proc ss d ric M t products n c W rin pp r l El ctronic quipm nt El ctronic quipm nt Motor v hicl s nd p rts V t bl oils nd f ts L th r products Tr nsport quipm nt n c M t products n c W rin pp r l El ctronic quipm nt 0 0.01 0.02 0.03 0.04 0 0.01 0.02 0.03 0.04 0 0.01 0.02 0.03 0.04 b. Non co l-produc rs LIC/LMIC UMIC HIC Min r l products n c Min r l products n c Min r l products n c P trol um, co l products P trol um, co l products P trol um, co l products F rrous m t ls F rrous m t ls Proc ss d ric Ch mic l, rubb r, pl stic products Ch mic l, rubb r, pl stic products F rrous m t ls M t l products V t bl oils nd f ts Ch mic l, rubb r, pl stic products M nuf ctur s n c M nuf ctur s n c P p r products, publishin P p r products, publishin M t l products M t l products M chin r nd quipm nt n c M t ls n c Food products n c El ctronic quipm nt T xtil s M chin r nd quipm nt n c M t ls n c Tr nsport quipm nt n c V t bl oils nd f ts V t bl oils nd f ts P p r products, publishin B v r s nd tob cco products T xtil s M chin r nd quipm nt n c Su r W rin pp r l B v r s nd tob cco products Tr nsport quipm nt n c Motor v hicl s nd p rts Su r Wood products Tr nsport quipm nt n c Wood products M t ls n c L th r products Food products n c T xtil s Wood products W rin pp r l D ir products B v r s nd tob cco products L th r products M nuf ctur s n c D ir products Motor v hicl s nd p rts Motor v hicl s nd p rts Su r D ir products Bovin m t products Food products n c Proc ss d ric M t products n c Bovin m t products El ctronic quipm nt L th r products Proc ss d ric Bovin m t products W rin pp r l M t products n c M t products n c El ctronic quipm nt 0 0.01 0.02 0.03 0.04 0 0.01 0.02 0.03 0.04 0 0.01 0.02 0.03 0.04 Note: X axis represents average share of coal in IC, adjusted for indirect coal consumption (w/o electricity) Source: Authors’ calculations based on GTAP I/O tables 51 The link between coal-intensive that countries with the least productive manufacturing and job quality is complicated. mix of manufacturing activities (those in In static terms, average productivity levels in the lowest quintile) use more coal in their heavy manufacturing sub-sectors within LIC/ manufacturing sectors than countries with LMICs tend to exceed those in apparel and food higher manufacturing productivity. The and beverage manufacturing activities and upward trajectory of productivity and job pay higher relative wages. This means better- quality inherent to economic structural quality jobs in heavy manufacturing compared transformation aligns with broader objectives to alternatives in light manufacturing or in of transitioning away from economic the low-productivity agriculture and service dependence on coal and other fossil-fuels. jobs prevalent in most developing economies. On the other hand, heavy manufacturing is The degree to which coal-intensive more capital-intensive and therefore creates manufacturing patterns in emerging fewer jobs.16 In dynamic terms, as economies economies matter for advancing up the value shift up the value chain and increase their chain to more sophisticated, higher-value reliance on more sophisticated ICT-intensive production is unclear. But when economies manufacturing and high-skilled services have a large number of indirectly coal-linked and reduce their reliance on coal inputs and manufacturing jobs in addition to direct coal less-skilled labor, the manufacturing sector jobs associated with coal-fired electricity becomes more productive and generates better and coal extraction, a coal-centric structure quality jobs. This pattern is confirmed when of economic production may in fact slow the comparing manufacturing productivity levels diversification of economies toward higher and manufacturing coal-intensity across the productivity activities. This theme will be global GTAP database; Figure 3.12 indicates explored in the next chapter. Figure 3.12 Coal intensity and labor productivity in manufacturing L st productiv Most productiv co l int nsit in m nuf ct. s ctor 0.004 0.003 0.002 0.001 0 1 2 3 4 5 L bor productivit quintil s in m nuf cturin (in 2017 PPP US$) Note: Coal intensity in manufacturing sector calculated as average intensity across all GTAP manufacturing sub-sectors in each country on the basis of measure (iii) of Annex 1 Source: Authors’ calculations based on GTAP I/O tables and World Bank Cross-Country Database of Sectoral Labor Productivity (Dieppe and Matsuoka, 2020) 16 N  ote that the net impact on aggregate productivity of creating a small number of highly productive jobs in heavy manufacturing versus a large number of less productive jobs in light manufacturing may be positive or negative. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 52 53 CHAPTER 4 Labor Market Implications of Coal Production in Five Country Case Studies Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 54 4.1 Direct and Indirect Effects of Coal Demand on Coal and Non-coal Jobs Global coal mining employment trends belie Five country case studies help to unravel the significant heterogeneity at the country production-employment link. In this chapter, level. Rising coal production may or may we explore these questions through the lens not be accompanied by an increase in coal of specific country experience, illustrating mining employment; the analysis in Chapter some of the complex realities behind past and 2 presents a rather mixed picture, even when present coal production and employment in comparing employment trends within the five countries. These examples reflect a wide largest and fastest growing coal producers. array of experiences and provide the basis To what extent are countries’ coal production for considering the future job implications patterns linked to their coal employment of changing demand for coal. Countries were patterns? In other words, are there common selected to represent the different categories patterns among the various categories of of coal producers, noted in Figure 4.1, with two coal producers defined above, namely the examples to illustrate the expanding exporter (i) advanced coal transitioners, (ii) partial category – Indonesia and South Africa – due transitioners, (iii) accommodators of domestic to their distinct labor market contexts. demand, and (iv) expanding exporters? Figure 4.1 Four categories of coal producers, five country examples si Po l d on nd In Exp ndin Adv nc d So Export rs Tr nsition rs ut h Af r ic Un it dS di t In t Dom stic D m nd P rti l s Accommod tors Tr nsition rs 55 Job quality is at the core of economic well, significantly more than local alternatives development and well-being. As discussed in agriculture or low-skilled services, and in the previous chapter, economies undergo typically more than similar occupations in structural transformation from largely the construction and manufacturing sectors. agricultural or primary production-based This high wage premium – and in many activities performed by self-employed or settings, early retirement eligibility – reflects low-wage workers using unsophisticated compensation for the hazardous nature of production technologies (manual rather than mining work. Underground mining jobs mechanized) with low-quality inputs and can mean difficult, dangerous or unhealthy serving local markets on the one hand, to more working conditions, while surface mining industrialized activities centered in firms with can be highly mechanized, involving heavy many workers, specialized by task, and using machinery. Many coal mining jobs are formal more capital inputs to generate more added and therefore subject to labor code protections value and selling to larger (urban, national and covered by social insurance.17 or external) markets. Wage employees who engage in more productive work earn a higher Coal production and its associated return (i.e., salary) than workers in low- employment make a positive short-term productivity activities. For individuals, higher contribution to local economic development. earnings facilitate increased consumption Coal mine investments – similar to other and may facilitate household savings, extractive activities – bring jobs and economic enabling investment in physical assets and/ stimulus and their associated tax revenue to or human capital, as well as access to social otherwise small, remote and under-funded insurance (e.g., in a formal wage job) and/or districts, many of which have above-average an old-age pension. These investments, in poverty rates. The creation of coal mining jobs turn, facilitate household human capital and spurs labor demand within coal supply chains income gains not only within the family but as well as in other local sectors, as coal workers also intergenerationally, enabling even low- spend their wages on local goods and services, endowment families to transition eventually generating taxable transactions that can out of poverty to middle-class status. contribute to government coffers. Estimates of the size of the multiplier effect of added coal Are coal mining jobs of high quality? What mining jobs vary, with evidence from advanced counts as a good job? Different criteria can be economies ranging from very small local used to rate job quality, such as compensation, spillovers (0.174 multiplier estimated by Black et working conditions, social externalities, or al. (2005a) in four US coal states during a boom environmental and economic sustainability period) to modest spillovers (0.99 multiplier considerations. These criteria are likely to estimated by Moritz et al. (2017) in northern vary between a worker’s perspective, his/her Sweden), to moderate spillovers (1.74 multiplier family’s perspective, and society’s perspective. estimated by Farren and Partridge (2015) Job quality is also a relative rather than across 3 counties in the US state of Virginia, absolute concept. Coal mining jobs tend to pay based on I/O modeling), to highly variable 17 F  ormal employment status can be defined in different ways. The most widely used criteria is social insurance coverage. In some settings – often as a result of data limitations – formal work status is defined based on whether or not the worker has a written contract, regardless of the benefits specified in that contract. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 56 spillover effects by sector (not significant in employees lose their jobs. This includes miners some sectors, but multipliers of 0.4 in real engaged in production occupations, as well as estate services, 1.2 in wholesale trade, and 1.5 the non-production staff performing support in accommodation and food, estimated by functions such as coal sorting, grounds and Fleming and Measham (2014) in Australia). machinery maintenance, administration and management. Coal is part of a supply chain Evidence on the longer-term effects of that includes inputs used to extract the coal, coal mining employment is quite negative, such as machinery and processing chemicals, linked to boom and bust cycles that can and the goods and services associated with the undermine economic growth. Coal can be downstream use of coal, such as transportation characterized as a natural resource curse to the end-user, which could be a power plant, because it distorts the local economy by or port services and shipping for exported driving up wages and potentially crowding coal. The mine closure reduces demand for out other economic activity. When job seekers these goods and services along the supply are aware that coal mine jobs pay well-above chain, which in turn reduces labor demand in the prevailing alternatives, they may only be these sectors. The lost income associated with induced to accept jobs with similar pay. Wage coal mine job losses and coal supply chain job distortions can have persistent dampening losses reduces families’ purchasing power, effects on labor demand across multiple causing them to curtail consumption of local sectors in the local economy, ultimately retail, entertainment and restaurant services, constraining economic growth. There is a and even essentials like health services and wide literature documenting these types of food. As a result, these firms lose business and effects. For example, Van der Ploeg’s (2011) therefore earnings, which curtails their own literature review finds lower long-run growth spending and leads to reductions in operations rates in natural resource-intensive locations and layoffs (these effects are often referred to when averaged over boom and bust cycles. as induced job losses). In this way, a negative Black et al. (2005a) find positive job spillovers shock to coal demand gets transmitted during booms, but larger negative job impacts through multiple channels. during busts. According to Freudenburg (1992), communities “over-adapt” to extractive When a negative shock is large relative to industries by assuming that booms are the other economic activity in the communities long-term norm, while busts are temporary. near the closed mine, local economies can Haggerty (2014) concludes that longer periods be severely disrupted. As with any sector of natural resource specialization (in this case, downsizing, there can be significant collateral oil and gas) result in lower average incomes. damage, for both people and communities. Glaeser et al. (2015) and Betz et al. (2015) find When retrenched workers’ incomes plummet evidence that the presence of coal mines and and are not quickly replaced through coal employment crowd out business start-ups alternative employment or other cash benefits and entrepreneurship. such as severance pay or unemployment insurance, households no longer frequent local A large demand shock to coal has multiple businesses and risk losing their savings and transmission channels, both direct and often their housing. Some families migrate for indirect. Closing a coal mine means that mine better work opportunities, accelerating the 57 shrinking economic base. This in turn puts The five country deep-dives that follow pressure on the tax base, as governments are examine the impact of changing patterns unable to collect income or profit taxes from in coal production and coal mining unemployed workers or insolvent businesses, employment within specific country labor and at the same time are burdened with markets. This approach enables a better higher spending on social assistance. The understanding of the complex short- and financial stress on families can contribute to longer-term labor market outcomes that new negative social behaviors such as substance coal jobs can bring, and that destroyed coal abusive and within-family violence (Lobao jobs can engender. Each country study begins et al. 2021). In coal-dependent regions, mine with a description of coal production and closure can result in deep and prolonged consumption trends over recent decades, and economic recession as the initial shock then examines in greater detail the effects of ultimately spills over into the housing coal mining jobs on local labor markets as well market and reduced government investment as within the broader national labor market in health and education services, which context, exploring the extent to which coal in turn risk undermining human capital, mining employment contributes to or works weakening social capital, and contributing against better job outcomes and stronger to increased outmigration. economic development. Figure 4.2 Loss of coal mine jobs creates wide-ranging ripple effects Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 58 4.2 Poland: Lessons from a Long Transition Poland’s coal sector has been in gradual under increasing financial stress, leading the but steady decline since 1990. Annual government to close mines (Baran et al. 2018). production of coal fell by more than half The number of operating coal mines declined between 1988 and 2019, driven by reductions from 70 in 1990 to 24 today (Kapetaki et al. in hard coal (Figure 4.3). Poland is the EU’s 2021), and more are scheduled for closure second-biggest producer after Germany, in 2022.18 The industry continues to suffer when considering both hard coal and lignite from weak market conditions and elevated together, and currently accounts for 30 production costs, partly due to surplus labor, percent of total EU production and 95 percent and partly due to the low quality of coal of the EU’s hard coal production. Already produced. In the past seven years alone, inefficient and unprofitable before 1989, coal production fell by 31 Mt. the coal mining sector subsequently came Figure 4.3 Poland’s coal production and coal mining employment 450 450,000 400 400,000 350 350,000 300 300,000 Production (Mt) Emplo m nt 250 250,000 200 200,000 150 150,000 100 100,000 50 50,000 - 0 1981 1983 1985 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 H rd co l Li nit H rd co l mpl. H rd co l nd li nit mpl. Note: Data excludes lignite mining employment. Sources: OECD; energy.instrat.pl 18  ard coal mines that remain in operation are located in the Upper Silesian Coal Basin (with the exception of Bogdanka H in the Lublin region) and are either owned or controlled by the state (Baran et al. 2018; Polish Geological Institute 2021). Lignite mining is located in Belchatow (Lodzkie) and Turow (Dolnoslaskie) (both owned by the state’s power company PGE), Konin-Turek región (Wielkopolska)(privately owned by ZE PAK), and Sieniawa (Lubuskie) (a small family-held mine). 59 Poland has extremely high per capita coal Despite declining coal production, coal consumption, even though it has fallen by remains the primary source of electricity half over the past three decades. Historically, generation. Power generation in Poland Poland’s per capita consumption of coal-based was almost exclusively based on coal until energy was four to six times the world average, the early 2000s when alternative power and as much as 70 percent higher than the generation sources began to emerge (mostly average for Europe (Figure 4.4). Poland’s natural gas). Renewable energy sources did energy consumption has declined very not achieve scale until 2018. By 2019, coal rapidly since 1990, to 1.2 toe per capita in 2019, still accounted for 74 percent of total power although this is still more than double the generation (down from 92 percent in 2017), global per capita level. Most of Poland’s hard while renewables19 (excluding hydropower) coal production is consumed domestically. and natural gas accounted for 14 percent and Prior to 2004, the industrial sector was the 9 percent, respectively. This persistence of dominant consumer (final consumption), coal-fired electricity – among the highest but its sharp decline in demand was replaced in Europe (Figure 4.5) – is partly the result by rising residential demand, which has of massive installed capacity, two-thirds of remained relatively unchanged for over a which is older than 30 years (most were built decade. Final consumption of electricity has between 1960 and 1980) and need replacement however continued to increase, not only in the by 2050 (Bogdan et al. 2015) and many even residential sector but also by industry and by sooner.20 The efficiency of Poland’s power commercial and public services (IEA data). plants is lower than others across Europe, resulting in higher production costs and CO2 Figure 4.4 Energy consumption per capita from coal (toe) 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 World Europ Pol nd Source: Authors’ calculations based on BP Statistical Review of World Energy, WDI data 19 Note that 25 percent of renewables used in electricity generation is biomass co-fired with coal in power plants.  20  Rogala (2021) reports the average age of power plants is 47 years. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 60 Figure 4.5 Europe’s electricity generation mix in 2018 EU-28 3,272 Estoni ** 12 Pol nd 170 C chi 88 Bul ri 47 G rm n 642 Gr c 53 Slov ni 16 N th rl nds 114 Rom ni 65 Tot ls, TWh D nm rk 30 Portu l 60 Hun r 32 Finl nd* 70 Sp in 274 Ir l nd* 31 Slov ki 27 It l 289 Cro ti 14 UK 331 Austri 69 B l ium 75 Fr nc 581 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Co l Oil G s Nucl r H dro N w RE W st Note: * denotes that coal includes peat; ** denotes that coal includes oil shale. Source: Eurostat data (graphic by Euracoal 2021) emissions (Alves Dias et al. 2018).21 Globally, thousand MW in new generating capacity that Poland has the 6th highest number of coal- was approved and under construction was fired power plant units with over 30 MW recently converted to a gas-fired power plant capacity, at 156 operating units spread across (Carpenter 2020). About four-fifths of total 50 power stations, and ranks 10th in terms of generation capacity is either state-owned or generating capacity, at 30,200 MW (End Coal controlled, and the transmission grid is owned 2021). Between 2010 and 2020, 48 generating and operated by state-owned Polskie Sieci units were retired (nearly 6,000 MW), another Elektroenergetyczne (PSE).22 23 were cancelled, and 2 were mothballed. One 21  ost power generation companies struggle financially, given that the generation cost of one megawatt-hour of M electricity is higher than the revenue from its sale (Czyżak and Wrona 2021). 22 Note that the electricity generation market is controlled by four state-owned companies: Polska Grupa Energetyczna  (PGE), Tauron Polska Energia, Energa, and Enea. These companies also have some control and/or ownership over coal mining companies. Both Polish coal mining and electricity generation are dominated by state presence. 61 A major driver of high residential demand furnace, reflecting installed capital that would for coal is its use for heating. Two-thirds be extremely costly to replace at national scale. of Poland’s residential heating comes from coal, double the share in European neighbors The negative environmental effects Czech Republic and Bulgaria; most of Europe stemming from coal production as well as uses other sources for 80-90 percent of consumption are considerable. Poland’s heating needs (Bertelsen et al. 2020; Figure CO2 emissions originate mostly from coal, 4.6). Poland is the only country in Europe to although CO2 emissions have been declining use more coal-derived energy for heating since the early 1990s (Figure 4.7). In addition today than it did in 1990 (The Economist to the legacy of environmental damage to land 2021). Much of this is associated with coal- and water resources, methane leakage from fired district heating systems, but direct coal operations – 659 kilotonnes in 2018, most burning of hard coal by small consumers also from underground mines – is a major source contributes (Badiani-Magnusson et al. 2019). of GHG emissions (Kasprzak 2020).23 Several of About one-half of households have a coal-fired Poland’s hard coal mines are deep underground Figure 4.6 Fuel sources of residential heating across Europe (2015, TWh) Austri B l ium Bul ri Cro ti C prus C ch R public D nm rk EU-28 Estoni Finl nd Fr nc G rm n Gr c Hun r Ir l nd It l L tvi Lithu ni Lux mbour M lt N th rl nds Pol nd Portu l Rom ni Slov ki Slov ni Sp in Sw d n Unit d Kin dom 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Co l N tur l G s Nucl r Oil R n w bl s Biom ss W st , Non-R n w bl Note: 2015 level of coal consumption for residential heating was 135 TWh in Poland, 62 TWh in Germany and 23 TWh in the UK. Source: Bertelsen et al. (2020). 23 T  he impact of one tonne of methane emissions is equivalent to 86 tonnes of CO2 emissions, when considered for a 20- year time horizon (Kasprzak 2020). Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 62 and have high methane content. Coal, EU average, but in winter, it us up to three times especially hard coal, is by far the largest source greater (The Economist 2021). About 44,000 of greenhouse gas emissions and air pollution, annual premature deaths are attributable to posing serious threats to public health and smog in Poland (European Environment the physical environment within Poland and Agency 2016)24, half of which are due residential beyond national borders by exacerbating coal stoves (Ministry of Development 2018). Most climate change (Alves Dias et al. 2018). countries have banned this fuel for individual heating. In recognition of the adverse health Poland’s air quality is among the worst in effects, the Government of Poland offered Europe. Coal-only emissions are measured by subsidies to replace coal-powered heaters under levels of PM10, SOx, NOx, and more general the Clean Air Priority Program (CAPP); this ambient air quality is measured by levels proved effective in eliminating coal-burners of PM2.5; all are estimated to have high in large public, commercial, and residential concentrations in Poland. Upper Silesia and buildings. The city of Krakow was the first to Malopolska regions are home to Poland’s ban residential burning of coal and wood (The highest mining-related polluting cities. In Economist 2021). Eleven out of 16 Polish regions fact, two-thirds of Europe’s most polluted have imposed emissions standards for heating cities are located in Poland, according to appliances in single-family homes, with the World Health Organization (Mortkowitz implementation dates ranging from January and Martewicz 2016). Coal-based residential 2022 to January 2027, after which it will become heating is also a major pollutant. In summer, illegal to use heating appliances that are not Poland’s PM2.5 level is only slightly above the compliant with the standards. Figure 4.7 Poland’s CO2 emissions by fuel source (Mt of CO2) 400 350 300 250 200 150 100 50 0 1990 1992 1994 1996 1978 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Co l G s Oth r Industr Oil Fl rin C m nt Note: Includes emissions from combustion as well as methane leakages from coal mines. Source: Our World in Data (Ritchie et al. 2021) 24 In 2018, the numbers of premature deaths rose to 46,300, according to European Environment Agency (2020).  63 Emissions from coal-fired power plants are Coal sector subsidies and bailouts represent blamed for serious health risks that translate enormous direct cost to taxpayers. Indirect into significant economic costs. In the period subsidies of coal-related activities come in 2010-2019, CO2 emissions from coal-fired the form of environmental, health and labor plants were 130 Mt annually (data from Europe productivity losses (described above), as well Beyond Coal 2021). In 2016, an estimated 2,500 as the cost of wages to mine workers employed premature deaths and over two thousand in the state’s unprofitable mines and power hospitalizations were attributed to emissions plants. Stoczkiewicz et al. (2020) estimate that from Poland’s coal-fired power plants, between 2013 and 2018, the state spent €6.8 according to civil society alliance Europe billion propping up the power sector. Some Beyond Coal based on data from 35 power subsidies have been touted as promoting the plants (Table 4.1). Additional health impacts transition to renewables or supporting the include chronic bronchitis among adults transition of workers and communities. In and asthma symptoms in children, giving 2017, coal mining subsidies were allocated rise to additional health costs. An estimated to mine decommissioning, rehabilitation 776,000 days of work were lost in 2016 as a and the support of former miners through result of coal plant pollution (albeit less than reemployment in other sectors, compensatory in Germany). The estimated economic costs pensions and social security benefits (Whitley associated with these pollutants were €7,5 et al. 2017). Devoting substantial public billion in 2016 (Europe Beyond Coal 2021).25 resources to sustain the coal sector displaces Table 4.1 Health impacts from coal power plants in 2016 Modelled health impacts (caused by power plants in country/region) Chronic Asthma Health costs Health costs Premature Hospital Lost working bronchitis symptom days (Million Euro (Million Euro deaths admissions days (adults only) in children 2016) (median) 2016) (high) EU (not including the UK) 12,243 5,627 9,396 3,707,296 238,388 18,613 35,583 Germany 4,238 1,700 3,124 1,308,036 69,761 6,338 12,205 Poland 2,596 1,106 2,093 776,559 42,402 3,934 7,536 Note: Based on methodology in Jones (2018) Source: Europe Beyond Coal (2021) 25 E  stimation methodologies vary by source. A similar analysis by Jones et al. (2016) found that in 2013, Poland’s coal power plants were to blame for 1,100 premature deaths in Poland and another 4,700 premature deaths (combined) in nearby Slovakia, the Czech Republic and Hungary, as well as Italy, Greece and France. Poland itself also suffered an additional 700 premature deaths due to coal plant emissions from outside its borders. The study also estimates total annual health costs at €8 - €16 billion. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 64 climate-friendly financing and potentially to fully give up on coal” (Brauers and Oei undermines the transition process. 2020). Poland’s current commitments to energy transition by 2030 are inconsistent Recent impetus to reduce Poland’s coal with EU targets (Czyżak and Wrona 2021). The dependence is coming from many quarters, provisions of the Poland Energy Policy for both internal and external. In spite of the 2040 (PEP 2040) assume a decrease in coal’s high direct cost of continued subsidy as share of energy generation from 75 percent to well as government’s acknowledgment of 56 percent by 2030, far above the EU target of 2 serious environmental threats and their percent.27 In addition, the PEP 2040 projects a high associated costs to health and labor 32 percent share of renewable energy sources productivity (not to mention the costs related in electricity generation in 2030, while the EU to climate change), the pace of transition average is expected to be 68 percent. Together, is slow. The country’s high degree of coal these projected reductions will be insufficient dependence explains some of this inertia; the to reach the EU’s 55 percent GHG emissions scale of infrastructure replacement or re- reduction target by 2030 (Kasprzak 2021; configuring to shift electricity and heating to Czyżak and Wrona 2021). Recent discussions renewable sources is massive and costly. The of coal phase-out between the government, government has allocated resources to energy mine operators (such as PGG), generation efficiency initiatives and renewable energy companies (such as PGE) and unions appear investments, reflected in the rising foothold of to be taking place without any consideration renewables in electricity generation. Over the of PEP 2040.28 Moreover, the PEP 2040 years, however, Poland has been reluctant to projections themselves seem highly unlikely, reach consensus on the EU’s commitment to because they require a massive uptake of achieve climate neutrality in all EU member nuclear (where little has happened so far) and states by 2050.26 hydrogen energy sources, increases in biomass use in co-combustion,29 as well as significant The government continues to provide mixed generation by offshore wind energy (from its signals on how – and how fast – to complete current capacity of zero; Van Renssen 2021).30 the coal transition. At COP 24 in Katowice, More recently, the government has proposed Poland (December 2018), the coal industry was post-COVID-19 recovery investments that proudly showcased, and President Andrzej are not moving away from business-as-usual Duda confirmed that “there is no plan today (Van Renssen 2021). 26 At the time of writing, Poland is the only EU member state that has not signed up.  27  ven in Poland’s ambitious scenario, coal is expected to account for 30 percent of net electricity generation in 2030 E (Czyżak and Wrona 2021). 28 In September 2020, the Polish government and trade unions reached an agreement to halt operations of two thermal  coal mines of state-owned Polska Grupa Gornicza (PGG) and to close all of PGG’s coal mines (8 mines) by 2049 (IEA 2020). The last mines to be closed in 2049 will be Chwalowice and Jankowice in the town of Rybnik, considered PGG’s most efficient. The agreement is the first time Poland has put a timeline on ending coal, and is in line with EU's climate targets of net-zero carbon emissions by 2050. However, the agreement is conditional on the European Commission's consent for new state aid to ensure the stability of the hard coal mining companies. 29  ased on this plan, a recent analysis indicates that Poland would need to import waste biomass in the future (Mankowska et al. 2021). B 30  n January 2021, Poland passed a historic Offshore Wind Act which paves the way to 4GW in 2030 and 28GW I capacity in 2050. Czyżak and Wrona (2021) argue that unblocking investments in onshore wind farms is crucial for accelerating Poland’s coal phase-out and aligning with EU targets. 65 To achieve the challenging coal transition European Commission technical assistance targets, Poland will need to consolidate and and financing is being provided to Poland’s accelerate its approach. Poland’s coal regions various coal regions to prepare Territorial Just have prepared or are updating their regional Transition Plans, which will include project development plans that incorporate green proposals to accelerate coal communities’ objectives, but a national coal phase-out plan transition to alternative economic activities. that meets the EU’s climate net zero target Under the European Green Deal’s Just by 2050 or complies with Paris Agreement Transition Fund, Poland is slated to receive EUR commitments of an EU-wide coal phase-out 3,5 billion to support economic diversification by 2030 is still lacking (Czyżak and Hetmański and labor transition (World Bank 2021a). 2020). The European Commission is offering a range of incentives for transition, including The number of coal mining jobs has fallen through regulation, financial support, advisory sharply since the beginning of the transition support and capacity building. Ongoing period. From a height of 444,000 coal mining Figure 4.8 Scale of employment in Silesia’s hard coal mines Source: Lewandowski et al., 2020. Based on data from Industrial Development Agency, Branch Office in Katowice (2018) Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 66 jobs31 in 1989, the aggressive mine closures percent work in associated processing plants initiated in 1990 triggered massive layoffs and 2 percent in mining administration (Czerwińska 2002); by 2002, coal mine jobs had (Lewandowski et al. 2020). Four-fifths work in fallen by nearly two-thirds to 164,000. This production occupations (Lewandowski et al. was followed by a slower pace of contraction, 2020). Coal mining workers are relatively low- largely due to attrition rather than layoff skilled compared to many other sectors and (Baran et al. 2020). By 2019, coal mining tend to have lower than average educational employment had fallen to 92,600, equivalent attainment. In 2014, 6 percent of coal mining to 0.5 percent of total employment in the Polish workers had primary education and 37 percent economy. Ninety percent of hard coal mining had basic vocational education, which is 16 pp. is located in Upper Silesia, and 80 percent is in higher than the national rate. Only 16 percent underground mines (Figure 4.8; Lewandowski of coal mining workers had tertiary education, et al. 2020). Employment in the lignite sector about half the average for all workers (Baran is much smaller than in hard coal; as of 2019, et al. 2018). Men account for 80 percent of the there were 9,300 people working in lignite coal mining workforce. mines (energy.instrat.pl 2021). Coal mining jobs pay high wages with Although only a small share of the generous benefits, which distorts local labor labor force, coal mining jobs play a markets and impedes labor reallocation disproportionate role, especially in Upper to other sectors. The hard coal and lignite Silesia. The coal mining industry has a large sector's average monthly wage is about twice presence in Silesia, where it directly accounts the prevailing wage in manufacturing (Baran for 4 percent of total employment and over et al. 2018). Part of this higher wage can be 7 percent of male employment (2019 data explained as a hazard premium, due to the from Energy.instrat.pl and Statistical Office health32 and other risks associated with mine in Katowice 2020). In addition to direct coal work, especially in underground mining mining jobs, many jobs in the region are activities. Mine workers are also eligible for indirectly related to the sector. It is estimated early retirement. The high sectoral returns that one job in the mining industries gives rise are especially large for less educated workers, to between 1.16 and 1.35 jobs in other economic whose non-coal employment alternatives pay sectors (IBS 2020). About one-fifth of all significantly lower wages. Black et al. (2005b) mining and quarrying jobs within the EU-27 provide evidence from U.S. coal regions that are located in Silesia (2017 data). educational attainment declined in periods when demand for coal workers was strong. Hard coal mining operations are labor The impact of high mining wages on the local intensive, dominated by semi-skilled labor market is three-fold: (i) it drives up the production occupations, and employ mostly reservation wage of mine workers, reducing men. The vast majority of employees in the their willingness to take other jobs; (ii) it sector work in mines (94 percent), while 4 squeezes labor demand in competing sectors; 31 Includes hard coal and lignite.  32 E  very year, an average of 400 new cases of pneumoconiosis – caused by coal dust inhalation – are diagnosed among current and former mine workers (WUG 2017). 67 and (iii) it pushes education choices toward 2002-2015, much higher than those who exited mining-oriented technical training that has manufacturing jobs during the same period little relevance outside of the increasingly (Figure 4.9). The effect is especially large for obsolete mining sector. workers over age 45, which can be explained by the generous early retirement policy in the The large magnitude of coal mining jobs Polish mining sector.31 These early retirement lost during the first stage of transition provisions create a strong incentive to remain generated persistent economic challenges in the sector until pension eligibility. Winkler in mining communities. The post-separation (2021) finds evidence of persistent negative employment outcomes for displaced mine employment effects in municipalities that workers depend on the profile of workers, the experienced at least one mine closure, incentives to transition, and the availability including lower long-run employment of alternative work. Older mine workers may rates for men (8 percentage points lower, be less inclined to take a new job if available considering employment in non-micro-sized work is low paid or if their degraded health firms only). limits their work options. Baran et al. (2020) find evidence of very high rates of inactivity among workers who left mining jobs during Figure 4.9 Destination of workers exiting mining and manufacturing jobs (2002-2015 average) 100% 90% 80% 71% 70% Emplo m nt in oth r s ctors 60% 50% 45% B c m 39% un mplo d 40% 37% 35% 34% 33% 33% 30% 24% B c m in ctiv 20% 21% 20% 9% 10% 0% A 15+ A 15-44 A 15+ A 15-44 M nuf cturin M nuf cturin Minin Minin Note: “Age 15+” category reflects all employed workers in the sector, whereas “Age 15-44” restricts the sample to younger workers. Mining defined broadly but using data only from regions with active hard coal mines. The figure shows probabilities of transition in the year following job exit. Source: Baran et al. (2020). 33  The regular retirement age for men is 65, whereas underground miners can retire up to 15 years earlier. Specifically, they can retire at age of 50 (55) if they have worked at least 25 years, including at least 15 (10) years working underground. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 68 Until around 1996, the unions successfully Given the high cost of the Mining Social resisted proposals to reduce wages or cut Package, the impacts on post-separation employment, but in 1998, government outcomes were rather mixed. The total introduced the Mining Social Package to funds spent on assistance under the Mining accelerate voluntary layoffs while protecting Social Package during 1998-2002 amounted against a significant deterioration in to 5.38 billion PLN, equivalent to 0.75 percent miners’ standard of living. Despite its of Poland’s GDP over that period (Turek and success in facilitating mine closures, the Karbownik 2005, Baran et al. 2018). An impact generous terms incentivized labor force assessment of miners who took redundancy exit rather than transition to new sectors payment found that 54-65 percent found a job of work. Between 1998 and 2002, a total of within several months of leaving the mining 67,000 mining workers benefited from the sector (Turek and Karbownik 2005), while package. Nearly 37,000 received miners’ leave the rest exited the labor force. Apart from assistance, paid at a level equal to 75 percent one optional retraining course, there was no of their monthly mining wage. Workers comprehensive job-search support under the within five years of retirement were made program. There is evidence that the average eligible. Individuals on the miners’ leave economic status of beneficiary households were not permitted to receive other forms worsened systematically after job separation; of assistance, but they were allowed to take Karbownik (2005) found that in 2001, 6 percent employment outside of mining, in which case of beneficiaries declared they could not afford their benefit was reduced by a half. The 29,700 to cover expenditures for basic needs (food, workers ineligible for miners’ leave assistance electricity, clothes), and by 2004, this share instead received conditional redundancy had risen to 13 percent. payments offered to miners who voluntarily left the mining sector and found a job within Government ownership of mines can in 24 months. It was paid at the moment the theory simplify transition planning, but it individual started his or her new employment. also complicates the landscape, with the If the worker started the new job by the end effect of slowing progress. Because most of of 1999, the payment was equivalent to 14.4 Poland’s mines are under government control, months of the average wage in the mining they fall victim to competing objectives of sector. Payment was gradually reduced to commercial viability on the one hand, and about 7.2-monthly average wages for those job creation and economic stability in lagging workers starting a new job by the end of 2002. regions on the other hand. A recent IMF study An unconditional redundancy payment was asserts that although Poland has successfully also available, equivalent to 24-monthly transitioned out of its communist-era legacy, average mining sector wages, but recipients the footprint of SOEs remains significant; were required to give up other forms of SOEs account for one-eighth of total assistance. The last program offered – which employment (Richmond et al. 2019). Whereas was taken up by just 419 individuals – was a the initial phase of mine closures was part of monthly welfare allowance equal to 65 percent a broader economic restructuring effort, the of the pre-layoff wage, paid for a maximum of current environment for announcing the next 24 months and intended to assist those taking phase of closures is more fraught, partly due retraining courses. to the negative and in some places lingering effects of high unemployment and economic 69 losses. The state must therefore balance a communities as current coal sector workers. number of political objectives; and these are Finding the right policies and programs complicated by the presence of unions, which to facilitate the transition of both local exert tremendous influence in the sector.34 economies and workers towards sustainable alternatives will not be easy. It will be Progressing to a complete phase-out critical to avoid exacerbating labor market of coal-fired power generation and the distortions via too-generous separation attendant reductions in coal production are packages to mine workers who already essential for meeting emissions reduction benefit from higher (publicly-financed) targets, but this will necessitate labor wages compared to workers in other sectors; adjustment in communities where coal has widening existing labor market distortions been a dominant economic player. Shifting risks slowing the pace of adjustment and local to renewable energy, green technologies in economic recovery. Significant political will other sectors and other greening services is needed to accommodate the competing will create alternative work opportunities, interests of various stakeholders in a way that but not necessarily in the same numbers, or balances the broader objectives of the state with the same skills profiles, or in the same with the needs of local communities. 34  ccording to Szpor et al. (2018), coal mining trade unions are among the strongest in Poland, their strength having A originated in the 1970s and 1980s when Solidarność—both a trade union and a social movement—became the driving force of the country’s democratic transformation. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 70 4.3 United States: Slow Convert to Post-Coal Transition35 From the early 19th century, Appalachia of Pennsylvania and Ohio had no periods of was the primary U.S. coal producing region, sustained coal employment increase from but over the course of the last century, 1919 to 2017, whereas the coal sector in central coal jobs declined and Appalachia’s coal Appalachian states experienced mini-booms, industry ultimately faded from market with employment increasing in many counties dominance. Recall from Figure 2.9 that US during 1963-1982 and 2000-2010. But the coal production experienced a steady rise alternating bust-periods vastly overwhelmed from the 1950s until the 2000s, whereas coal positive employment effects during coal mining employment has been declining for a booms. The central Appalachia coal industry century, from a height around 800,000 in 1919 ultimately followed northern Appalachia, but to less than 50,000 today. The displacement with a lag of 20-30 years. The “bust” years of Appalachian-sourced coal by less labor- of the 1980s triggered more mine closures, intensive surface mining of less polluting which intensified in subsequent years with coal in the Powder River Basin of Wyoming the tightening of environmental regulations and southeastern Montana in the western U.S. and competition from the rapid emergence exerted downward pressure on coal sector of fracking beginning around 2005.36 By labor demand in Appalachia, especially from the 2012-2017 period, central and northern 1980 onward (Figure 4.10). Appalachian coal mining employment each tracked downward. All five Appalachian Appalachia’s transition from sector states experienced a long-term collapse dominance to dwindling relevance took of coal employment: over the last century, a very long time, and spanned periods of West Virginia lost 85 percent of coal mining localized boom and bust, with new mine jobs, compared to 78 percent for Virginia, 85 openings and mine closures across the percent for Kentucky, and 98 percent in both many counties of Appalachia. Decline in Ohio and Pennsylvania (Figure 4.11).37 north Appalachia began immediately after World War I, while the central Appalachian coal industry experienced more volatile ebbs and flows. The northern Appalachian states 35 This case study draws from the more detailed 2021 study entitled “Socioeconomic Transition in the Appalachia  Coal Region: Some Factors of Success” by L. Lobao, M. Partridge, O. Hean, P. Kelly, S. Chung and E. Ruppert Bulmer, produced for the World Bank under the Global Support to Coal Regions in Transition project. 36 S  cores of U.S. coal companies have filed for bankruptcy in recent years, including giants Peabody Energy, Cloud Peak Energy, Arch Coal, Murray Energy, and Alpha Natural Resources (see https://en.wikipedia.org/wiki/Coal_mining_in_ the_United_States (downloaded April 30, 2020)). 37  eginning in the 1920s through the early 1960s, key reasons for declining coal demand were the growing use B of electricity, oil, natural gas and other fuels; substitution of diesel locomotives for steam locomotives; and the general rise of trucking over rail. See http://explorepahistory.com/story.php?storyId=1-9-B&chapter=0 and http:// explorepahistory.com/story.php?storyId=1-9-18&chapter=0 (downloaded May 1, 2020). 71 Figure 4.10 Drivers of US coal production and employment 1960-2015 Emissions st nd rds str n th n d in St d conomic rowth. 1990 Cl n Air Act Am ndm nts Co l is import nt for pow r n r tion 1000M Growth in surf c minin M tric Tons 800M Stron cl n n r polici s. R n w bl stron rowth Sh l G s nd imm ns comp tition for 600M 250 co l. Pow rs ctor shiftin w from co l 400M 200 PRODUCTION s X 1,000 150 M ch ni tion 100 EMPLOYMENT Emplo Mod st ssist nc for App l chi n co l communiti s 50 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Source: World Bank (2018a), based on US Energy Information Agency sources and from Coal Transition in the United States, Irem Kok (IDDRI and Climate Strategies, 2017) Figure 4.11 Coal mining employment by state (selected states) 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 1919 1929 1939 1954 1963 1972 1982 1992 2002 2012 2017 PA Co l Emplo m nt OH Co l Emplo m nt KY Co l Emplo m nt Oth r App l chi n Co l Emplo m nt WV Co l Emplo m nt VA Co l Emplo m nt AL Co l Emplo m nt Source: Lobao et al. (2021) Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 72 Domestic U.S. coal consumption has sources such as natural gas and renewables, diminished in recent decades. Industry which have accelerated over the past decade. demand for coal inputs (excluding coal-based Coal-fired power generation capacity has electricity) has significantly declined since waned in recent years, due to a combination 1990 (Figure 4.12), while ever-rising electricity of plant retirements and cancelled demand is increasingly met by non-coal construction (Figure 4.13). Figure 4.12 U.S. coal final consumption by sector (ktoe) 50,000 40,000 30,000 kto 20,000 10,000 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Industr R sid nti l Non-sp cific A ricultur /for str Comm rci l & public s rvic s Source: IEA data for the US Figure 4.13 Coal-based power generation capacity in the U.S. 400K 300K C nc l d w tts 200K R tir d M Op r tin 100K 0 2014 2015 2016 2017 2018 2019 2020 Source: globalenergymonitor.org 73 The declining fortunes of Appalachia’s hinder transition and recovery in Appalachia.38 coal industry exposed many communities The principal lessons are as follows. to severe negative economic shocks. The coal economy-based Appalachian region • Mining and coal-dominated communities has historically been poorer than much fair worse on welfare outcomes such as of the U.S. For decades, it had some of the poverty, incomes, employment, population lowest per-capita income levels and highest growth, and other measures of well- poverty rates in the country (Lobao et being (Betz et al. 2015; Black et al. 2005a; al. 2016). During the coal industry’s rise, Cook 1995; Douglas and Walker 2017; communities across Appalachia became Freudenburg and Wilson 2002; Lobao et al. dependent on the coal economy and coal 2016). This finding might seem obvious, but employment, whether in coal mining or in these impacts do not necessarily hold for associated coal supply chains. But as coal employment shocks in other industries such production began to shift west, the ensuing as service sectors. decline of the coal sector in Appalachia was severe, requiring communities to adjust to • Coal mining effects on community new market realities. wellbeing tend to be positive in times of price upswings and negative when prices The extent and timing of the adjustment are low (Betz et al. 2015; Black et al. 2005a; varied from one county to the next, but Lobao et al. 2016). The uncertainty makes it in most regions, the resulting economic difficult for communities to adjust for the outcomes were similar: deep and sustained “natural resource curse.” Communities over- economic dislocation, high unemployment, adapt to extractive industries (Freudenburg sharp declines in household income, and 1992). Local receptivity to diversification deteriorating measures of community well- or alternative industries is cyclical, and being. The negative impact of mine closures depends on commodity price movements. was likely exacerbated by the fact that the U.S. coal industry consists of a large number • Three key sets of barriers across coal and of privately-owned mines; as private mining communities exist, related to: (a) operators responded to market conditions, geography and degree of remoteness from the result was sharper upswings and cities (Douglas and Walkers 2017; Haggerty downswings, with limited coordination with 2019; Haggerty et al. 2018; Snyder 2018); local governments on the labor impacts. (b) availability of alternative economic opportunities (Carley 2018; Deaton and A literature review of 37 U.S.-focused studies Niman 2012; Haggerty 2014; Haggerty et addressing transition in coal communities al. 2018); and (c) population vulnerability, or in communities that share similar reflected by low educational attainment characteristics to Appalachian coal (Douglas and Walker 2017; Haggerty et al. communities was carried out by Lobao et al. 2018) or an aging workforce (Haggerty 2019). (2021) to identify factors that facilitate or These structural barriers limit workforce upgrading and occupational mobility. 38 See Lobao et al. (2021) appendix A for a detailed description of the publications reviewed and their main conclusions.  Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 74 • Local environmental degradation hampers • Rural communities are not homogeneous, future development and the ability to attract and the benefits/costs of transition tourism and other non-extractive industries vary by population group. Appalachian (Appalachian Citizens’ Law Center et al. 2019; communities with a greater share of coal Haggerty et al. 2018; Kelsey et al. 2016). Places employment tend to have a lower share of with higher quality of life – including natural sole proprietors, higher disability rates, amenities as reflected in climate, topography, and a higher share of poor people (Betz et and water area – are more likely to attract al. 2015). Much has been written about the migrants, especially retirees (Isserman et al. uneven impacts of natural gas expansion, 2009; Partridge and Olfert 2011). with communities divided among those who benefit and those who do not. Extractive • Transition-affected populations are aware industries employ a higher proportion of the challenges their communities face. of men, and women tend to have fewer In Appalachia, residents’ concerns about local employment opportunities. Declines their community shifting to non-coal in extractive employment affect family employment include fear of potential job structure, and may increase the share of loss and business closures, detrimental female-headed households (Cook 1995). effects on schools and retail as families leave, lack of affordable housing elsewhere, • Communities with higher levels of social and high attachment to the community capital – that is, strong inter-group (mobility barriers) (Carley et al. 2018). relationships within the community – tend to be more resilient. In the case of • The prevalence of a “coal culture” plant closures, Besser et al. (2008) find that impedes transition in Appalachia. Carley et residents report less negative overall quality al. (2018) note that coal mining generational of life where social capital is higher. employment has fostered a community bond and identity with the industry. Haggerty et • Low capacity of local governments is al. (2018) point out that residents’ resistance a barrier to transition. Lack of local to change can arise when populations blame government capacity in rural and small restrictive environmental regulations, U.S. communities has long been noted while they ignore the larger role of markets (Johnson et al. 1995; Lobao and Kelly 2020).39 and price competition from, for example, The overriding problem is replacing and natural gas. Carley et al. (2018) note that stabilizing income streams (Haggerty et promising return of coal jobs discourages al. 2018). Haggerty (2019) summarizes the community and individual efforts to adapt barriers small or rural governments face to change. Long-term dependence on coal when coal plants close, including limited delays acceptance of transition, but in any administrative leadership capacity (for case, populations find it difficult to move example, little or no planning staff); limited elsewhere (Haggerty 2019). staff and ties to state or regional actors, which limits ability to apply for federal and state assistance; and low fiscal autonomy, 39 This has been documented in studies of transition planning, but there are few empirical analyses of actual transitions.  75 which limits local budgeting authority. and case studies of the region (Billings and Haggerty et al. (2018) and Haggerty (2019) Blee 2000, Duncan 2014, and others) explain stress that small communities often lack how coal mining weakened local government access to a dedicated transition fund, and institutional capacity to address residents’ requiring them to substitute other funds or needs. A majority of studies examining secure external funding. community-level impacts (positive or negative) of coal or other mining employment • Conventional policies to attract business focused on a single or specified point(s) investment, retain local businesses, in time, rarely addressing the long-term and develop the workforce have consequences of transitioning from coal (with variable effects. Benefits from economic the exception of case studies). development policies typically appear to be modest (Daniels et al. 2000), and strategies The study by Lobao et al. (2021) addresses tend to be overly focused on retaining or these knowledge gaps to deepen our attracting a single large employer (Haggerty understanding of the factors that contribute et al. 2018). No policies or programs work to communities’ ability to transition everywhere, and successful models appear successfully from reliance on coal. The difficult to replicate. study draws on the wider social sciences literature regarding the factors that contribute Whereas Appalachian coal communities to community well-being and successful are well-studied, knowledge gaps remain. transition, and then empirically tests the For example, the existing literature tends to relevance of these factors for Appalachia. Their focus at the regional multi-state level rather regression methodology exploits county-level than the community level, missing important differences in socioeconomic outcomes across heterogeneity. Much analysis is case-study 99 coal-dependent communities (within a based, although some key quantitative studies broader sample of 420 Appalachian Regional estimate employment spillovers during boom Commission (ARC) counties plus 650 counties and bust (for example, Black et al. 2005) and within 100-mile buffer zone) to identify the long-run natural resources curse (Deaton counties that have successfully transitioned and Niman 2012). Factors that contribute to away from coal at some point during the positive local outcomes in the wake of coal period from 1950 to the present. “Successful decline are not well-studied. Betz et al. (2015) transition” is defined as a county that was find that Appalachian communities with a dependent on coal mining sometime in the larger share of coal employment had lower past, is no longer dependent on coal mining entrepreneurship rates, which limits the today, and sustained above the sample- degree to which communities can transition average population growth during 1980-2018.40 to other economic activities. Historical studies 40  Population growth is the most common metric for assessing regional success in U.S. studies. Betz et al. (2015) found that population growth highly negatively correlates with intensity of local coal mining employment, and they found a statistically significant causal link between coal mining employment and population decline. Other variables associated with economic prosperity —per-capita income, poverty rates, and growth in median household income— positively correlate with population growth. Moreover, population growth directly captures the movement of people in and out of these regions based on economic as well as socioeconomic factors. Population data is also readily available at the necessary disaggregated levels. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 76 Based on these selection filters, only four Thirdly, the prevalence of non-structural “successful” transition counties emerge out impediments reinforces poor economic of 222 mining-intensive counties: Sequatchie outcomes and reduces local economic County, Tennessee; Laurel County, Kentucky; resilience. Whether or not structural Athens County and Noble County, both in Ohio impediments are present, historically coal- (shown in Figure 4.14, among the 420 counties dependent communities exhibit less economic within ARC boundaries). It is notable that these diversification, modest manufacturing activity, successful counties transitioned at different problematic patterns of “boom and bust” times, and for different reasons. Robustness cycles, and low levels of entrepreneurship. checks and qualitative analysis confirm the results and yield six key conclusions. The fourth main finding is that institutional capacity and social capital have helped The first and perhaps most striking some counties transition more successfully. conclusion from this quantitative analysis Local government institutional capacity and is that very few Appalachian counties social capital are generally low across the have managed a positive transition from ARC region compared to national averages. In coal dependence, and the level of success the four counties, local government capacity is modest. Only four counties managed to to design, finance, and implement economic transition and remain economically viable development initiatives in collaboration communities with sustained population with local civil society and regional planning growth. And traditional measures of economic authorities appears to have helped sustain well-being are not particularly strong. While transition impetus. the four counties have grown in population and diversified their production, and most have In fifth place, infrastructure investment is a experienced significant poverty reduction, common theme in the successful counties, average household incomes remain low and but is not sufficient to guarantee successful poverty rates exceed national and ARC averages. transition from coal. Much of Appalachia has received significant investment, at least with Secondly, severe economic structural respect to road infrastructure, and yet most impediments across Appalachia constrain counties failed to remain viable or thrive. That growth. Being small and remote, most ARC said, improved roads helped our four successful counties have limited access to labor markets counties by increasing connectivity to larger with more and diverse job opportunities. metropolitan areas and manufacturing chains, ARC counties have low levels of physical and increasing access by larger urban centers capital, especially infrastructure, and high to tourism, recreational opportunities and transportation costs. Human capital is also affordable housing. low, with lower educational attainment and lower quality education and health services. 77 Figure 4.14 County-level population growth rates within the ARC region 2000-2018 Note: Dark red denotes counties with the fastest population growth between 2000 and 2018. Black line denotes the Appalachian Regional Commission (ARC) boundary, which encompasses 420 counties. Source: Lobao et al. (2021) Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 78 Finally, the transition paths of each of the The checkered transition experience of four counties have unique features and Appalachia’s coal-dependent communities success factors distinct from one another, provides some broad policy takeaways making it difficult to generalize approaches. for governing authorities and other But each has some kind of economic anchor stakeholders seeking better transition that stimulated both direct and indirect outcomes in their own coal regions. The job creation. very long nature of the transition meant that it spanned multiple changes in local, • Athens County’s economic development regional and national governments and has centered around its large public Ohio their associated policies. And because the University which supports direct and indirect period of study was long enough to coincide jobs and generates local social capital. with countless human and physical capital investment initiatives at the local, regional • Noble County was able to attract a large and national levels, it is impossible to isolate public investment to build a state prison, which policies were effective in supporting which has served as an economic driver. or facilitating transition, and whether these worked in isolation or jointly. There are some • Laurel County became a regional hub broad lessons that emerge, however, including following investment to construct two major the importance of enhancing connectivity highways – including Interstate-75, which of remote coal regions with larger economic linked the area to northern manufacturing centers, investing in human capital to centers – as well as a regional airport, improve labor mobility, diversifying business a hydroelectric power dam, piped water activities to foster resilience to boom and bust supply, and industrial parks. cycles, building local institutional capacity, especially regarding economic planning, • Sequatchie County benefited from and coordinating across multiple levels investments in highways to access of government and other stakeholders for nearby Chattanooga, a large metropolitan strategic longer-term economic development. market offering diverse job opportunities. The county became a main “bedroom” community for daily commuters into Chattanooga. 79 4.4 Indonesia: Crowding into the Export Market Coal production has grown rapidly in Indonesia’s total export basket in 2000 to Indonesia, directly contributing to the nearly 10 percent by 2018 (Harvard Atlas of economy through increased output and Economic Complexity). Half of Indonesia’s exports. Expanding steadily since the early coal exports are destined to China and 1990s, Indonesia’s coal sector took off in India, each having vastly increased their the mid-2000s, subsequently averaging 10 import demand compared to other markets percent annual growth in production (Figure (Figure 4.16). During this period, Indonesia’s 4.15). Between 2000 and 2010, coal exports petroleum products lost market share: nearly tripled to 6.2 Exajoules, and reached domestic oil production fell by nearly half, 9.2 Exajoules in 2019 (BP Statistical Review and oil exports declined from 20 percent of of World Energy 2020). In terms of export total exports in 2000 to 8 percent in 2018. share, coal grew from a modest 2 percent of Figure 4.15 Indonesia’s rising coal production (ktoe) 350,000 Oil 300,000 250,000 Oth r 200,000 bituminous co l kto 150,000 100,000 Sub-bituminous co l 50,000 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Anthr cit Cokin co l Oth r bituminous co l Sub-bituminous co l Source: IEA data for Indonesia 41 IEA data reports Indonesia coal exports of 154,511 ktoe in 2010 and 235,987 ktoe in 2018. 42 Philippines, Malaysia and Thailand are significant exceptions. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 80 Figure 4.16 Shift in Indonesia’s coal export destination markets (thousands of tons) 450,000 400,000 Oth rs T iw n 350,000 Sp in Kor 300,000 250,000 Th il nd J p n 200,000 Indi Philippin s Indi 150,000 100,000 M l si Chin Chin 50,000 Hon Kon 0 2009 2019 Source: 2019 Handbook of Energy and Economic Statistics of Indonesia, Ministry of Energy and Mineral Resources In addition to rising external demand for access to electricity, and households consumed coal, domestic demand for energy has more electricity, disproportionately sourced surged, driving the coal sector’s expansion. from coal-fired power plants. Between 2009 Three-quarters of domestic coal is consumed and 2019, the share of households with access by power plants.43 Low production costs for to electricity increased from 94 to 99 percent extracting Indonesia’s abundant coal deposits (World Bank 2021b), and the total amount of translated into relatively cheap energy inputs, electricity sold to households by Indonesia’s especially when combined with Government- state electricity company45 nearly doubled.46 imposed price controls44, spurring demand Electricity consumption by industry increased by households and by industry. Indonesia’s at a similar pace, followed by commercial strong GDP growth over the last two decades and public services (Figure 4.18). Indonesia’s – averaging 5 percent per year – was electricity consumption per capita rose by 67 accompanied by significantly higher per capita percent over the last decade, outpacing the incomes and declining poverty (Figure 4.17). As regional average for Asia Pacific, although domestic incomes grew and living standards lagging the lightening pace observed in China rose, more Indonesian households gained (98 percent).47 43  he 2020 Ministry of Energy and Mineral Resources (MEMR) data indicate that iron, steel and metallurgy activities T accounted for nearly 11 percent of domestical coal demand, while the cement industry accounted for another 9 percent. 44 M  EMR sets the price of domestic coal used for coal-fired power generation and sets a quota for domestic coal consumption – the Domestic Market Obligation (DMO). 45 PLN - Perusahaan Listrik Negara.  46  Based on data from the Directorate General of Electricity and PLN (Perusahaan Listrik Negara) Statistics, reported in the 2019 Handbook of Energy and Economic Statistics of Indonesia. 47 IEA data.  81 Figure 4.17 Robust GDP growth and declining poverty and vulnerability 0.7 10% 0.6 5% 0.5 Sh r poor or vuln r bl 0% 0.4 GDP rowth 0.3 -5% 0.2 -10% 0.1 0 -15% 09 05 03 07 01 99 95 93 15 13 97 17 91 11 20 20 20 20 20 20 19 20 20 20 19 19 19 19 Poor or Vuln r bl GDP rowth (ri ht xis) Source: World Bank “Pathways to Middle-Class Jobs in Indonesia” (2020a), using the Susenas national poverty line. Figure 4.18 Upward trajectory of electricity consumption across sectors (ktoe) 25,000 20,000 Industr 15,000 10,000 R sid nti l 5,000 Comm rci l & public s rvic s 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Industr R sid nti l Comm rci l & public s rvic s A ricultur / for str Source: IEA data for Indonesia Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 82 Coal increasingly dominates the electricity 30 percent in 1990 to nearly 60 percent in 2019 mix in Indonesia, even as other countries are (IEA data). Figure 4.19 shows that oil-based shifting to greener alternatives. Electricity electricity has been squeezed out by natural generation grew by 9 times between 1990 and gas over the last decade, and renewables are 2019, and coal-fired electricity expanded even just starting to pick up. faster, as coal’s share of electricity grew from Figure 4.19 Indonesia’s electricity generation by source (GWh) 300,000 250,000 200,000 Co l 150,000 Oil 100,000 N tur l s 50,000 H dro G o Bio 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Co l Oil N tur l G s H dro G oth rm l Biofu ls Wind Sol r PV W st Source: IEA data for Indonesia 83 Indonesia’s total carbon dioxide emissions 2010 to 2018 finds that most of Indonesia’s quadrupled over the same period, the greatest increase in CO2 emissions was due to its rising share of which is attributed to electricity and per capita incomes and growing population, heat producers, followed by transport and followed by the expanded share of coal in then industry (Figure 4.20). Guan et al. (2021) energy production (Figure 4.21). None of these examine the many factors that drive emissions factors bodes well for reversing the upward increases in Indonesia and around the world. emissions trajectory in the near term Their decomposition exercise for the period or achieving Indonesia’s NDC target.48 Figure 4.20 Indonesia’s CO2 emissions by sector (Mt CO2) 600 500 400 El ctricit & h t produc rs 300 Industr 200 100 Tr nsport R sid nti l 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 El ctricit & h t produc rs Oth r n r industri s Industr Tr nsport R sid nti l Comm rci l & public s rvic s A ricultur Fin l consumption not ls wh r sp cifi d Source: IEA data for Indonesia 48  ndonesia’s energy intensity, defined as energy consumption per unit of GDP, did at least decline between 2010 I and 2018, which helped to constrain the upward pressure on CO2 emissions (Cui at al. 2020). Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 84 Figure 4.21 Factors driving Indonesia’s rise in CO2 emissions (2010-2018) +30 +10.3 +2.7 +46.6 -0.4 -12.6 +66.6 -19.1 +145.4 541.8 -117.7 390 Popul tion co l s CO2.2018 CO2.2010 oil Industr : prim r CO2 int nsit Int nsit Industr : t rti r Industr : s cond r GDP/c pit n tur l En r En r En r En r Source: Guan et al. (2021) The slow adoption of alternative energy and effectively subsidizing any cost sources is largely explained by coal’s low differences (Atteridge et al. 2018, Susanto production cost and the monopoly power of 2017). PLN’s grid-based energy purchases Indonesia’s state electricity company PLN from IPPs and PPUs are still highly (Perusahaan Listrik Negara), which plays an concentrated in coal-fired generation influential role with respect to Indonesia’s (accounting for about two-thirds), but the energy strategy. PLN’s mandate to return portion from renewable resources is finally a profit was a main reason why it resisted increasing (albeit from a very low base).49 renewable energy alternatives that are more expensive than coal; this disconnect was In addition to rising consumer demand, addressed in a 2017 regulation requiring other factors contributed to the explosive the state electricity company to purchase rate of coal mining growth since 2000. any surplus renewable-source electricity According to a report by Stockholm generated by Independent Power Producers, Environment Institute (2018)50, a confluence 49  Indonesia’s state electricity company PLN has purchased hydro and geothermal power from IPPs and PPUs for at least a decade, but solar and wind generation represents a new source (based on data from PLN (Perusahaan Listrik Negara) Statistics and Electricity Statistics, Directorate General of Electricity, reported in 2019 Handbook of Energy and Economic Statistics of Indonesia). 50 A  . Atteridge, M. Thazin Aung and A. Nugroho, “Contemporary Coal Dynamics in Indonesia”, SEI working paper 2018- 04, Stockholm Environment Institute. 85 of institutional and governance factors fueled 2018). The recently issued Omnibus Law on the proliferation of new mining permits, Job Creation (Undang-Undang Cipta Kerja, which surged from a modest number of 2020) includes provisions to revoke local mining firms that were directly contracted government authority over mining permits by the national government to a much higher (unless stipulated in specific implementing number following the decentralization of regulations) but the Ministry’s limited licensing decisions to the district government capacity for local oversight and monitoring level. This shift in permitting authority was may impede the law’s effectiveness. part of the broader decentralization reform in 1999. Between 2001 and 2008, the number The coal sector in Indonesia has generated of mining permits grew more than eight- significant employment over the past decade. fold, reaching 8,000 by 2008, and increasing According to national labor force survey data, further to 11,000 by 2014, two-fifths of coal mining employment more than doubled which were for coal mining, mostly small from around 90,000 jobs in 2007 (the first year and medium-sized mines (Hayati 2015 and for which sufficiently representative data are HFW 2014, cited in Atteridge et al. 2018). The available) to a peak of almost 264,000 in 2014, fee structure for coal permits, under which before receding to 215,000 jobs by 2015 (Figure coal producers pay a royalty on production 4.22).52 Coal mining employment expanded as well as a land rent on the area covered by thereafter, reaching nearly 240,000 jobs in the permit, incentivizes district governments 2018. Indonesia’s coal mining workers account to issue more permits. Atteridge et al. for only 0.2 percent of Indonesia’s 120 million (2018) also point to weak governance and employed workers but represent 5 percent of the oversight, illustrated by examples where local global coal workforce (recall from Chapter 2). politicians profit directly or indirectly from mining activity. Government has introduced Indonesia’s labor market is steadily a series of measures to increase oversight modernizing. Although still a largely and exert more control over what it deems a rural labor force, over the past decade national resource, albeit with mixed effect.51 employment has increasingly shifted away It is notable that, despite the relative ease of from agriculture-based work toward services acquiring a permit, significant illegal mining and, to a lesser degree, manufacturing takes place – whether in non-permitted areas (Figure 4.23). The share of agriculture jobs or in protected forest reserves where permits fell from 40 percent in 2007 to 27 percent a should not have been issued – and illegal decade later, as workers increasingly moved exports are also a problem (Atteridge et al. into alternative employment in the services, 51  n 2009, limits on the share of coal available for export were introduced via the Domestic Market Obligation I established as part of the 2009 Mining Law and MEMR Regulation No. 34/2009. Regulations were added in 2010 requiring mine owners to prepare a post-closure mine reclamation plan and deposit an upfront reclamation guarantee, but these rules are often ignored. Since its 2015 inception, the Clean and Clear program is a certification process that reviews mining firms’ compliance with a range of obligations. 52 T  he data used for this analysis comes from Indonesia’s National Labor Force Survey, Sakernas, conducted by BPS- Statistics Indonesia. Since its inception in 1976, Sakernas has undergone a series of improvements in geographical coverage and type of labor market information collected. It is the largest source of employment data in Indonesia, representative at the district level beginning in 2007. We utilize the August wave of Sakernas because it provides the necessary level of industrial sector disaggregation to differentiate coal mining activities from non-coal mining and quarrying. Note that this level of disaggregation is not available for 2016, 2017 or 2019. As in the rest of this report, using industry classifications to measure coal mining employment captures direct coal mining jobs but not indirect employment in supporting activities. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 86 Figure 4.22 Coal mining employment 2007-2018 Indon si 300,000 250,000 200,000 150,000 100,000 50,000 0 09 08 07 10 18 16 15 13 12 14 17 11 20 20 20 20 20 20 20 20 20 20 20 20 Source: Sakernas data 2007-2018 manufacturing and construction sectors. As high commodity prices started to ease in This pattern reflects continued structural the early 2010s, manufacturing employment transformation of the economy, but with picked up again, but labor productivity growth a shift in focus to natural resource-based in most sectors of the economy has been slower export activities in the aftermath of the Asian this decade compared to last, and most jobs are financial crisis, compared to earlier periods in low-productivity sectors and/or in relatively of dynamic industrialization focused on unskilled occupations (World Bank 2020a). labor-intensive manufacturing exports. The Note the concentration of current employment World Bank’s (2020a) jobs assessment asserts (bubble size denotes employment share) that Indonesia’s declining competitiveness and decadal job creation in low-paid low- in the 2000s led to a contraction in the productivity sectors in the left half of Figure manufacturing employment share to around 4.24, especially in commerce, manufacturing 13 percent while low-value added services and construction. absorbed the greatest share of jobseekers. 87 Figure 4.23 Shifting sectoral employment shares (% of total employment) 100% Communit & 90% p rson l (incl. public dmin) 80% Fin nc & busin ss 70% s rvic s 60% Tr nsport tion & communic tions 50% Whol s l & r t il, r st ur nts 40% Construction 30% M nuf cturin 20% Minin & qu rr in 10% A ricultur 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: Sakernas data 2007-2019 Figure 4.24 Job creation during 2000-2018 concentrated in relatively low-productivity sectors 40 Whol s l nd r t il tr d , v hicl Public dministr tion, d f ns , nd compulsor 30 soci l s curit Accomod tion nd food s rvic ctiviti s M nuf cturin 20 Fin nci l nd insur nc El ctricit nd s ctiviti s tion Oth r s rvic ctiviti s Construction 10 Busin ss ctiviti s Educ tion ctiviti s Sh r of Job Cr 0 R l st t Inform tion nd W t r suppl , s w , communic tion -10 w st m n m nt, nd ctiviti s Minin nd r m di tion qu rr in A ricultur , for str , huntin Hum n h lth nd soci l -20 work ctiviti s nd fish r Tr nsport tion nd stor -30 -40 0 50,000 1,000,000 1,500,000 2,000,000 2,500,000 3,000,000 3,500,000 4,000,000 4,500,000 M di n r lw 2018 Note: Bubble size denotes share of employment in 2018 Source: from World Bank (2020b) using Sakernas data 2000-2018 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 88 Formal employment – defined here as permanent wage jobs from 29 percent in 2007 permanent wage work – is becoming more to 41 percent in 2019 was a key contributor prevalent, although informal work status to the increasing share of middle-class jobs remains widespread. Nearly every sector observed over the past decade (World Bank increased its share of permanent wage jobs, an 2020a). Nevertheless, informal53 work status is indication of the labor market’s shift to more still predominant, with 56 percent of workers formal structures of production and more engaged in self-employment, casual work, or formal employment contracts. The rise in unpaid family work in 2019 (Figure 4.25). Figure 4.25 Breakdown of work status (% of total employment) 100% 90% P rm n nt w mplo 80% Emplo r with p rm n nt mplo s 70% Emplo r with t mp. mplo s C su l w mplo , 60% non- ricultur C su l w mplo , 50% ricultur Own ccount 40% (non- ricultur ) F rm r Inform l 30% Unp id f mil work rs 20% 10% 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: Sakernas data 2007-2019 53  arious criteria can be used to define informality. In this analysis, we define informal as those in casual wage V employment, own account work including farmers, unpaid work, and employers of temporary employees. 89 Coal mining jobs are of higher average engage in production occupations or as quality and pay more than most other machine operators, at rates similar to those sectors in Indonesia’s economy. Over 95 observed in manufacturing, construction, and percent of coal mining jobs are formal, on transport and communications. Coal mining par with the government sector, and employ jobs pay higher wages than most sectors, more workers with above-average education, mostly than double the average agriculture wage, 86 with secondary school qualifications (Figure percent higher than the average construction 4.26). Coal mining workers are relatively job, and 59 percent more than the average young, nearly all male, and predominantly manufacturing wage (2018 data; Figure 4.27). Figure 4.26 Educational attainment levels by sector 100% 14% 13% 90% 80% 70% 60% 49% 50% 72% 40% 30% 25% 20% 10% 11% 14% 0 s on s s l l l , r v ss rs & ti & on ur in in ur i l nt on t Co ti ic ic r t in ur lt To t p it s in uc un o cu rM ct un st r s m sp tr bu & ri m m n ns f th nu & m A l Co co Tr O r Co s nc M l ho n W Fi L ss th n prim r compl t Prim r compl t Som or compl t s cond r Post S cond r Source: Sakernas data 2018 Figure 4.27 Comparison of average sectoral wages (2018, constant 2007 rupiahs) 16,000 12,000 8,000 4,000 - n l on ur il, rv s s s ti & s l rs & ur in in s ns nt on Co ti ic ic r t on st r t in ur lt M p it uc un o si cu rM ct un m sp bu & tr ri m n m ns f & l th nu m A co Tr Co s r O Co nc l M ho n W Fi Co l Provinc s All Provinc s Source: Sakernas data 2018 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 90 Indonesia’s coal activities are geographically coal intensity varies by district. Figure concentrated in Kalimantan, and to a 4.29 shows the district-level variation in lesser extent Sumatera. South Kalimantan, coal employment shares across Indonesia, East Kalimantan and North Kalimantan which are clearly concentrated in the three provinces54 account for the bulk of coal provinces, but to different degrees from one employment, followed by South Sumatera district to the next, ranging from negligible (Figure 4.28). Within these “coal provinces”, or zero coal jobs up to 15 percent. Figure 4.28 Coal employment high and rising in South Kalimantan and East and North Kalimantan (coal employment levels in 2007, 2018) Source: Sakernas data 2007, 2018 54  ote that although East Kalimantan and North Kalimantan became two separate provinces in 2012, the analysis below N considers them together to ensure consistent treatment across the entire period of analysis, namely 2007-2019. 91 Figure 4.29 Coal employment intensity by district (% of total employment, 2018) 18 16 14 12 10 8 6 4 2 - Ri u K p Ri u B li B nt n C ntr l K lim nt n South K lim nt n E st K lim nt n North K lim nt n North Sul w si C ntr l Sul w si South Sul w si st Sul w si Goront lo W st Sul w si North M luku W st K lim nt n M luku NAD J mbi B n kulu W st J v C ntr l J v E st J v L mpun B n k -B litun DI Yo j k rt P pu DKIJ k rt r W st P pu Sum t r Sum t r Sum t r Nus T n Source: Sakernas data 2018 South Despite its small share in total employment, that are vulnerable to boom and bust cycles coal plays an outsized role, especially in – can have large spillover effects for the local coal-intensive districts. During the period of economy. As coal production expands and new rapidly expanding coal production – notably coal jobs are created, labor demand increases between 2007 and 2012 – the economies through two main channels: (i) within coal of South Kalimantan and East and North supply chains (e.g., linked to mining, mining Kalimantan added 726,000 net jobs in total, operation inputs, coal transport, and coal- nearly 110,000 of which were coal mining jobs, fired power generation), and (ii) in non-coal reflecting a 21 percent annual average growth sectors within local economies. Regarding the rate in coal mining employment (Figure latter, as more coal workers spend their higher 4.30; details in Annex 2 Table 1). Whereas the earnings on local goods and services and local scale of total job creation in non-coal sectors tax revenues rise, increased aggregate demand dwarfs the number of coal jobs created, coal and government spending induce additional mining jobs – like other extractive activities job creation in other sectors. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 92 Figure 4.30 Sectoral job growth in South Kalimantan and East and North Kalimantan from 2007 to 2012 200,000 25% 20% 150,000 15% 100,000 10% 50,000 5% - 0% (50,000) -5% on ti & rv & l rv t & l ur t i s s s s* r r in s ur Ut r i n in nt on ic r t ti Co Q in ic ti S c lt uc un po ic i & tu ss n & rM st R ili l S un cu tr m s c R & n F in on m ri m r n u ns f th rs om & l nu Co A Co T O s C M l ho si Bu W P N t jobs dd d Annu l job rowth (ri ht xis) * Includes public administration. Source: Sakernas data 2007, 2012 The Indonesia data shows that the presence Annex 2 Table 2). This positive and strongly of coal mining jobs significantly affects significant correlation also holds when we local labor market outcomes related to restrict the estimate to the two main coal both wage levels (positive) and wage provinces, South Kalimantan and North and growth (negative). Regression analysis on East Kalimantan (Annex 2 Table 3). Looking the correlates of wages indicates that coal at the local impact of higher coal wages over wages are higher than all other sectors even time paints a mixed picture. Applying the when accounting for education level and methodology in Black et al. (2005)55, we find other characteristics (Figure 4.31 shows the evidence that the increase in well-paid coal wage regression coefficient values; details in jobs pulled up wages in other sectors, but it 55  imilar to Black et al. (2005), we restrict our sample to include only the wages of men aged 25-45 to reduce bias from S any changes in the composition of the labor force. We estimate the following equation: Ln(Wage i ) = β 0 + β 1 Coal i + β 2T + β 3 (Coal i * T) + Xβ 4 + U i where Coal is a binary variable that equals one if an individual is in our treatment coal-intensive district group; T is a time variable representing year 2007, 2012 or 2018; and X controls for age, age-squared, urban location, province and educational attainment level. β1 is interpreted as the differential wage between the treatment and comparison district groups and β3 is interpreted as the differential wage growth in a particular time period (2007-2012 or 2012-2018) between the treatment and comparison districts. 93 may have squeezed labor demand and wage observed in manufacturing and to a lesser growth in other sectors where firms had to extent construction, suggesting a degree compete for workers to fill local vacancies. of crowding out in these sectors, which For the 2007-2012 period, non-coal wage employ similar types of workers.56 There is no growth within coal-intensive districts was significant effect observed in these sectors in 23 percent slower than in districts with lower the subsequent 2012 to 2018 period when the coal intensity (although it should be noted number of coal jobs declined, although there that wage growth is measured relative to a is some evidence of modest upward pressure higher initial wage; Annex 2 Table 4). This on agriculture wages (results reported in result holds across all sectors considered. Annex 2 Table 5). The strongest negative wage effects are Figure 4.31 Correlates of real hourly wage (2018 data) A A squ r d Compl t Prim r Incompl t or Compl t S cond r Post-S cond r M l Urb n Co l s ctor Oth r minin M nuf cturin El ctricit , s, w t r Construction Whol s l & r t il, r st ur nts Tr nsport & communic tions Fin nc & busin ss s rvic s -20% 0 20% 40% 60% 80% 100% Note: Coefficient values from OLS regressions on the correlates of real hourly wages reported by wage employees in 2018 (converted to 2007 rupiahs). Independent variables include age, age squared, and dummy variables for education level, male gender, and urban location, and with sector and province fixed effects. Full regression results are reported in Annex 2 Table 2, column 2. Source: Authors’ estimates using 2018 Sakernas data. 56  his result contrasts with findings by Black et al. (2005) for the US showing separate impacts on tradable sectors T (i.e., manufacturing) and non-tradable sectors (i.e., construction and services). Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 94 Coal sector growth stimulates job public investment and sub-par public services creation in other sectors, but too many due to the lower tax revenues generated by coal jobs risks crowding out employment a struggling economy. This interpretation in non-coal sectors; our analysis finds both accords with findings by Edwards (2015)59 of positive employment spillovers as well as inferior health and education outcomes in evidence of crowding out in Indonesia’s Indonesia’s mining-intensive communities coal-intensive districts. Recall that analysis compared to neighboring districts. on the US Appalachia region shows severe natural resource curse effects in isolated Past episodes of global demand fluctuations regions and/or when compounded over time have impacted coal mining jobs in Indonesia. (Lobao et al. 2021). We test for employment For example, the negative price shocks in 2015 spillover effects in districts within Indonesia’s and 2016 affected local mine production and two main coal provinces, South Kalimantan jobs, especially in South Sumatera and Banten, and East and North Kalimantan, using the where smaller mines have higher production methodology articulated in Black et al (2005) costs; many of these mines were forced to cease and summarized in Annex 2. Our regressions production, at least temporarily (Atteridge et indicate a strongly positive correlation57 al. 2018). The Sakernas data indicates that coal between the level of coal employment mining employment in South Sumatera fell and employment in other sectors (highest from 15,500 in 2012 to 10,000-11,000 in 2014 in manufacturing (0.56) and lowest in and 2015, before rebounding to over 21,000 by construction (0.36)), but in districts that are 2018. Banten province suffered even greater especially coal-intensive , we find evidence of production shocks during this period, leading crowding out, namely that the spillovers are to permanent job losses; coal employment not as big compared to less intensive coal58 exceeded 15,000 in 2012, fell to 1,400 in 2015, districts (results reported in Annex 2 Table and only slightly recovered to 3,200 by 2018. 6). The effects appear strongest vis-à-vis the manufacturing sector. The boom and bust cycles associated with resource extraction are a challenge for The disproportionate influence of coal sustained economic development. Lobao mining jobs in local labor markets may et al. (2021) and Black et al (2005) provide stem from their relatively high wages which significant evidence on the challenges create distortions in the local economy. The facing coal communities vulnerable to price presence of high-paying coal jobs distorts both fluctuations that exacerbate local boom and the labor supply decisions of job seekers and bust effects. The decline in coal consumption the hiring decisions of employers, generating in most advanced economies and the persistent dampening effects on labor demand associated transition out of coal-fired energy across multiple sectors in the local economy in favor of cheaper and/or cleaner alternatives and ultimately constraining economic growth. has inflicted sometimes catastrophic damage Other channels of persistence include lower on coal-dependent economies. Recall that 57  ote that OLS regressions do not indicate causality, only correlation, although there is little reason to think that fast- N growing agriculture, manufacturing, construction or services employment would be driving an increase in coal jobs. 58 Districts are designated as coal-intensive if the coal share of employment is at least 4 percent.  59 R . Edwards (2015). “Mining Away the Preston Curve”, World Development Vol. 78, pp. 22–36.  95 in the US Appalachian region, districts that local economic shock may be more intense were heavily dependent on coal suffered job and the subsequent recovery prolonged. In losses and severe economic dislocation that a national survey of Indonesian employers had persistent effects; very few affected conducted in 2015, respondents reported counties have successfully transitioned away that the main obstacle to hiring unskilled from coal and still maintained a viable local production workers was applicants’ excessive economy (Lobao et al. 2021). With respect to and untenable wage expectations (Gomez- Indonesia’s mining sector, Bhattacharyya Mera and Hollweg 2018). and Resosudarmo (2015) find evidence that increasing coal mining employment has The current energy landscape – both in terms no significant effect on poverty, but when of energy supply and projected demand – mining activity accelerates, poverty increases, suggests that Indonesia’s transition away implying a serious negative dynamic effect of from coal may be gradual. Even if change is coal mining booms. Edwards (2017) cites case not imminent because of the rising trajectory study evidence that coal mining in Indonesia of Indonesia’s electricity demand, global export crowds out agriculture activity, which is the markets will ultimately dry up and Indonesia main sector of employment in rural areas, will shift its priorities away from the coal especially among the low-skilled. industry. The 2017 National General Plan on Energy (RUEN) caps annual coal production at Future prospects for the coal sector and coal 400 MT from 2019 to 1950, and targets a rise mining jobs in Indonesia are uncertain, in the renewable energy share from its 2017 at least in the medium term. If we see level of 6.2 percent to an ambitious 23 percent continued growth in coal production and local by 2025 and 31 percent by 2050. But this target economic specialization centered around coal depends partly on biomass for co-firing, which mining, there is a risk of coal dependence that is highly polluting. Moreover, promoting the increases districts’ vulnerability to demand use of biomass in energy generation is highly shocks. When the energy transition away problematic because expanding palm-oil from coal finally gets on track, what will it plantations can be devastating for forests and mean for Indonesia’s 240,000 coal mining biodiversity. Government currently mandates a workers and the many others employed in biodiesel blend of at least 30 percent (B30), but coal-dependent activities? Winding down insufficient refinery infrastructure impedes the mining operations and cutting coal sector jobs even higher target of 40 percent minimum (B40) may take on the characteristics of a “bust” for 2021. Meeting the Government’s ambitious cycle in which employment and wages in other targets will be difficult given the current pace of sectors are pulled down faster in the coal- transition as well as institutional weaknesses. intensive district than in non-coal intensive The Institute for Essential Services Reform districts, consistent with Black et al. (2005). If (IESR) projects that the country will likely displaced miners delay taking up alternative fall well short of its renewable energy target, jobs due to lower wage offers, the size of the forecasting a 15 percent share by 2025.60 Recent 60  he National Energy Board (DEN) estimated the 2019 renewable energy share at 9.2 percent. Source: “RI to T break long-term green energy promises at current pace: IESR”, The Jakarta Post, October 1, 2020. https://www. thejakartapost.com/news/2020/09/30/ri-to-break-long-term-green-energy-promises-at-current-pace-iesr.html Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 96 recommendations in a 2020 IEA report focus on Government of Indonesia’s CO2 emissions regulatory reforms to attract private investment reduction commitments. Indonesia’s strategic into renewable energy production to meet focus on palm-oil mixed biodiesel may domestic market needs.61 generate jobs in the palm oil value chain, but at an excessive environmental cost. With Policies that can facilitate transition away respect to mitigating the negative local effects from coal need to consider the interests of coal-related job losses, transitional support and incentives of the various agents engaged mechanisms could be implemented to help in or affected by coal production – coal workers adjust to the new market context. sector workers, local governments, small Chapter 5 lays out a policy framework for private investors, large energy suppliers and facilitating the transition to a post-coal world. IPPs, state electricity company PLN, and the 61 Source: “Attracting private investment to fund sustainable recoveries: The case for Indonesia’s power sector”, World  Energy Investment – Country Focus, Country report July 2020, Energy Supply and Investment Outlook (ESIO) Division of the Directorate of Sustainability, Technology and Outlooks, International Energy Agency. 97 4.5 South Africa: Holding its Ground Mineral discoveries in the late 19th century 1984). The strong and persistent demand set South Africa on its path to becoming the for mine workers and supporting activities most industrialized economy in Africa today. contributed to South Africa’s economic Discoveries of diamond, gold and platinum transition away from a patchwork agrarian deposits in the late 19th century stimulated economy to a more industrial economy large-scale national and international (Turok 2012). This long structural transition migration, especially to the main mining areas was accompanied by significant changes in around Kimberley (in Free State province) the labor market. As the economy gradually and Witwatersrand (in Gauteng province; diversified into manufacturing activities, labor Figure 4.32). The mining boom during the became increasingly specialized and wage early part of the 20th century required employment became the norm. an extremely large workforce, especially in diamond and gold mining (Yudelman Figure 4.32 Mining areas in South Africa by product Source: Council for Geoscience South Africa Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 98 South Africa’s rich natural resource base and mineral intensive – was developed for the the exploitation of these resources spurred transportation and export of minerals. the development of many other economic Geographic clusters that formed around sectors. The scale and variety of mineral mining activities generated spillovers in resources enabled the establishment of construction, forestry, and financial services. multiple supporting sectors. For example, coal, In addition to industrial diversification, South manganese, and iron ore resources gave rise Africa’s high-value agriculture and tradable to a robust steel industry, and manufacturing services – notably citrus and wine products, activities including the production of metals tourism, transport, and ICT – made important such as electrolytic manganese, chemicals and growing contributions to the economy and for explosives, and mineral fuels (Harris 1977; job creation. Majozi 2015). Machinery linked to extractive industries (e.g. hydraulic technology, The mining sector’s role in the South underground locomotives and mining African economy has steadily diminished fans) not only served the domestic market in recent decades. Mining and quarrying but became a robust part of the country’s as a percentage of GDP fell from 20 percent export base (IGF 2018). The transformation in 1970 to 8 percent in 2018 (Dessus and of minerals and precious metals into higher- Hanusch 2018). Mining and quarrying value value products brought higher profits and added experienced volatile swings, including earnings. Railway infrastructure - itself steep declines in 2008-2009, 2012 and 2016 Figure 4.33 Mining and quarrying value-added (constant 2010 ZAR, millions) 250,000 6.0% 245,000 4.0% 240,000 ZAR millions (const nt 2010) 235,000 2.0% 230,000 225,000 0.0% 220,000 -2.0% 215,000 210,000 -4.0% 205,000 200,000 -6.0% 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Minin nd qu rr in VA Annu l rowth (ri ht xis) Source: Stats SA, GDP P0441 Annual, quarterly and regional fourth quarter 1994-2019 99 Figure 4.34 Coal and gold production trends (constant 2010 ZAR, millions) 60,000 140,000 120,000 50,000 Gold VA (2010 ZAR millions) Co l VA (2010 ZAR millions) 100,000 40,000 80,000 30,000 60,000 20,000 40,000 10,000 20,000 0 0 1994 1996 1994 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Co l Gold Source: Stats SA, GDP P0441 Annual, quarterly and regional fourth quarter 1994-2019 with intermittent recovery (Figure 4.33). The per year on average, subsequently slowing to sector’s contribution to government revenue 1.3 percent annual growth in the 2000s, and declined from as high as 29 percent in 1981 stagnating since 2015 (BP Statistics). Coal’s to 2.5 percent in 2014 (Haddad et al. 2019), contribution to the economy has hovered in driven by especially sharp contractions in the range of R48 – R49 billion (constant 2010 gold mining (Figure 4.34). Gold production prices) since 2006, equivalent to about 0.5 fell by more than half between 2004 and percent of GDP (compared to over 2 percent 2016; by 2019, South Africa’s once significant for other mining and quarrying). In terms share of global gold production had fallen to of mineral sales, coal is the biggest earner, 4 percent (Chamber of Mines 2016; Minerals accounting for 27 percent of 2015 total sales, Council 2019). Mining nevertheless continues followed by precious gold and metals (21 to be important for South Africa’s balance of percent), iron ore (16 percent) and gold (13 payments. In 2018, the sector accounted for percent) (SA Stats 2015). 15 percent of private-sector fixed investment, 10 percent of total fixed investment, and 27 The rise in coal production was driven partly percent of total exports (Minerals Council by increasing domestic demand, especially 2019; South Africa Chambers of Mines 2017). rising demand for electricity. Nearly nine- tenths of South Africa’s electricity generation Coal has played a relatively minor but comes from coal (IEA data). Currently, about growing role compared to other mining and 70 – 75 percent of coal production by volume quarrying activities. During the 1980s and (and about 65 percent by value) is consumed 1990s, coal production increased by 3 percent domestically, as South Africa exports relatively Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 100 higher grade coal (Burton et al. 2018a). State- become more competitive.62 In 2018, more than owned Eskom, South Africa’s largest electricity half of South Africa's exports was destined company, relies on coal-fired generation, and to India. Other key export destinations are most of its electricity is sourced from coal-fired Pakistan, South Korea, Europe, and Africa power stations located in Mpumalanga province. (Figure 4.35). In terms of electricity distribution, 42 percent goes to the different municipalities, 52 percent Ageing and inefficient rail infrastructure is sold directly to industrial, mining, transport, is making access to external markets more commercial and residential clients, and 6 difficult. Nearly all South Africa’s export coal percent is exported (Winkler et al. 2020). Sasol is transported via rail from the central coal Ltd., a coal-to-liquids energy company, is also a basin to Richards Bay on the East Coast. While main consumer of coal (Burton et al. 2018a). port capacity at Richards Bay has increased, coal exports are constrained by rail capacity External demand for coal has also been a (as of 2010, rail capacity was below 68 Mtpa; persistent motivation behind expanding coal Eberhard 2011). Transnet, South Africa’s production. Since the mid-1980s, South Africa railway operator, has invested significantly in has steadily expanded its coal exports, which recent years to expand railway infrastructure, accounted for nearly three-tenths of total in part to facilitate coal transport. However, coal production in recent years. South Africa’s a 2017 EIA report flags the risk that weaker share of the coal export market averaged 8-9 global demand for coal, lower international percent for much of the 1990s and 2000s, coal prices, and regulatory uncertainties before moderating to around 6 percent in the together undermine the rationale for these rail last decade as Australia, Russia and Indonesia infrastructure investments (EIA 2017). Figure 4.35 South Africa's coal export destination markets in 2018 12% 8% Indi South Kor 9% 48% Europ Afric P kist n Oth rs 13% 10% Source: Richards Bay Coal Terminal in Nicholas and Buckley (2019) 62  Source: “Attracting private investment to fund sustainable recoveries: The case for Indonesia’s power sector”, World Energy Investment – Country Focus, Country report July 2020, Energy Supply and Investment Outlook (ESIO) Division of the Directorate of Sustainability, Technology and Outlooks, International Energy Agency. 101 Future prospects for South African coal Resource Plan (IRP 2019) articulates plans for are uncertain. With respect to exports, a more diversified energy mix with increased South Africa’s reliance on Indian demand in reliance on renewable energy and natural gas. particular has served it well in the past, but this may change. The cancellation of more South Africa’s dependence on coal as its than 50 percent of planned coal-fired power primary energy source contributes the plants in India suggests the possibility of largest share of national greenhouse gas a long-term stagnation in export demand, emissions. South Africa is the world’s 14th which would intensify as India transitions to largest emitter of greenhouse gases (GHGs) alternative energy sources. Domestic demand and has the second lowest share of renewables is also facing challenges, especially as a result in the G20 (after Saudi Arabia). Carbon dioxide of recent price increases. Eskom’s cost of coal emissions from electricity have increased by increased ninefold over the last two decades, 64 percent since 1990. Today, electricity and from R42,79/ton in 1999 to R393/ton in 2017 heat producers account for the greatest share (Eskom 1999; 2017b).63 New renewable capacity of carbon dioxide emissions, followed by other such as wind and solar generation is now energy industries, transportation and industry considerably less costly (about 40 – 50 percent (Figure 4.36). Changes in total energy use since less) than the Eskom coal-fired power plants 2010 were largely due to increased economic under construction (CSIR 2016; Steyn et al. activity, although some energy efficiency 2016; Garg et al. 2017; Burton et al. 2018a). The savings between 2014 and 2018 helped offset Government of South Africa’s 2019 Integrated the aggregate gain (Figure 4.37). Figure 4.36 South Africa’s CO2 emissions by sector (Mt CO2) 500 400 300 El ctricit nd h t produc rs Mt CO2 200 Oth r n r industri s Industr 100 Tr nsport R sid nti l 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Source: IEA data for South Africa 63  rice increases were partly due to increased production costs as easier-to-access mine deposits have been depleted, P and partly due to Eskom’s weak management, inefficient coal purchasing and uncompetitive practices, all of which have undermined its financial position (Burton et al. 2018a; Baker et al. 2015). Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 102 Figure 4.37 Decomposition of increased energy use in South Africa between 2010-2014 and 2014-2018 4 3.54 3.5 -0.03 -0.07 3 2.54 Us (EJ) 2.5 2.33 +1.1 -0.06 +0.05 +0.22 2 En r 1.5 1 0.5 0 10 14 18 ur ur t t nc nc vi vi 20 20 20 ct ct ti ti ci ci ru ru Ac Ac ffi ffi St St l l c c ni ni ch ch T T Source: IEA data for South Africa Despite Government efforts to diversify its closure plans, which are primarily driven by energy sources, coal retains its influential ‘end of design life considerations’ (IRP 2019), position. The state’s target to create an the Government has committed to new coal- additional 26 GW of electricity generation fired generation capacity to meet projected capacity from renewables and natural gas by demand during the transition to low-carbon 2030 (in addition to the current 5.6 GW from energy.64 Coal’s continued dominance is hydro, wind and solar) is coupled with plans linked to its long history as an abundant and to reduce coal-fired generation capacity by low-cost energy source – which facilitated decommissioning 10.5 GW of coal-fired power energy-intensive economic development and capacity by 2030 and another 20 GW by 2050 the emergence of multiple coal-dependent (compared to current coal-fired generation industries – as well as the state’s direct capacity of 42 GW; IRP 2019). Eskom’s 2017 intervention in the sector and the presence announcement of plant shutdowns was met of politically-connected unions, inter alia. with protests and opposition by multiple Together, these aspects of South Africa’s coal coal-related unions, eliciting calls for better sector render decisions about future coal planning and stakeholder consultations phase-out highly politicized (Baker et al. 2015). (Burton et al. 2018a). Concurrent with plant 64  further 8 GW of coal-fired generation capacity is either under construction or has been announced (End Coal 2021), A although some of these projects may encounter financing challenges (IRP 2019). 103 Figure 4.38 Employment trends in coal and other mining and quarrying 500,000 450,000 441,092 443,129 437,783 445,987 422,041 418,994 406,159 400,000 374,192 346,551 350,000 339,833 325,721 328,916 300,000 250,000 200,000 150,000 100,000 88,963 84,512 79,999 74,827 66,206 64,503 64,670 69,805 70,049 65,354 69,081 55,191 50,000 - 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Co l All Minin nd Qu rr in Source: South Africa’s QLFS data 2008-2019 3rd quarter, authors’ calculation While coal production has risen in past growth over the last decade, peaking in 2017 decades, coal mining employment trends at 446,000 jobs, equivalent to only 3 percent have been more variable. From an average of total employment. To understand the of 65,000 jobs in 2008-2010, coal mining labor impact of South Africa’s coal industry employment expanded thereafter to peak in recent decades and what it might portend at 89,000 in 2013, subsequently declining for the future, especially in a context of to 55,000 in 2016 and recovering thereafter diminishing coal demand, it is essential to (Figure 4.38). The broader mining and understand the broader labor market context quarrying sector posted above average job within the country. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 104 The most notable features of South Africa’s The large anticipated influx of youth into labor market are its large and rapidly the labor market creates an opportunity to growing youth population that is not being accelerate economic growth and raise living adequately absorbed into employment – standards. South Africa is not an outlier in good jobs or otherwise, and a high degree this regard; much of the Africa region is in a of segmentation that creates frictions to similar position. By 2050, Africa will be home worker mobility into and between different to a billion youth, on its way to becoming the types of jobs and work status.65 continent with the largest number of young people, nearly double the youth population • South Africa is in the midst of a of South Asia, Southeast Asia, East Asia, and demographic transition, which creates Oceania combined (World Economic Forum pressure on the labor market. South Africa’s 2020). For countries to achieve the potential youth population (ages 15 – 35) numbers 20 “demographic dividend”, however, youth million, making up 36 percent of the country's need to be absorbed into productive paid work. total population. The working-age population expanded by nearly 7 million between 2008 • Only a small share of South African youth and 2019 and UN projections suggest that finds work, and an increasing number the working-age share of total population are choosing to remain outside the labor will remain around 66 percent until 2030. force. Unemployment rates are extremely Figure 4.39 Trends in labor force participation and unemployment status 45,000 40% popul tion (’000) 40,000 35% 35,000 30% Un mplo m nt (%) 30,000 25% 25,000 20% 20,000 15% 15,000 10,000 10% Workin 5,000 5% 0 0% 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Emplo d Un mplo d Out of L bor Forc Un mplo m nt r t (tot l, ri ht xis) Un m. r t (m l outh, ri ht xis) Un m. r t (f m l outh, ri ht xis) Source: South Africa’s QLFS data 2008-2019 3rd quarter, authors’ calculation 65  he data used for this analysis is South Africa’s Quarterly Labor Force Survey (QLFS) conducted by Statistics South T Africa (Stats SA). The data is representative at the provincial and metropolitan levels. Since 1993, Stats SA has collected labor market information with the October Household Surveys (OHS) conducted annually between 1993 and 1999, and the Labor Force Survey (LFS) conducted biannual between 2000 and 2007. Due to methodological and sampling issues, the QLFS was introduced to replace the LFS in 2008. The QLFS is available quarterly from 2008 – 2020. We have utilized the 3rd wave of the QLFS for all available years excluding 2020 due to a drop in the sample size as a result of COVID-19 mobility restrictions. Note that households without telephones were excluded from the sampling framework. 105 high by international norms and have risen graduates seeking work. Between 2008 and persistently over the last decade (Figure 2019, the labor force expanded from about 4.39). Youth unemployment rates in 2019 19 million to 23 million but employment reached 37 percent for males and 44 percent increased more slowly, from 14.5 million for females. Broken down by age cohort, the to 16.4 million, driving up unemployment unemployment rate among 15 to 25 year-olds and pushing youth out of the labor force. is 58 percent, compared to 36 percent for The difficulty of finding work, especially 25 to 35 year-olds. These statistics indicate work that meets youth’s expectations in severe stagnation in the labor market, as the line with their education levels, discourages economy fails to create enough jobs to absorb youth from participating in the labor force. the existing pool of unemployed workers Nearly two in ten young people are not or the annual inflow of school leavers and in employment, education, training, or Figure 4.40 School-to-work transitions by age (2019, 2008) M l , 2019 M l , 2008 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 15 17 19 21 23 25 27 29 31 33 35 15 17 19 21 23 25 27 29 31 33 35 F m l , 2019 F m l , 2008 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 15 17 19 21 23 25 27 29 31 33 35 15 17 19 21 23 25 27 29 31 33 35 Schoolin S lf- mplo d A ricultur Inform l w work r Form l W mplo NEET Unp id f mil work rs S lf- mplo d non-A ricultur Emplo r, oth rs, unsp cifi d Un mplo m nt Note: This graph shows a static plot of male and female youths’ work status by age, and does not capture dynamic transitions. Formal employment status is defined as having a written contract. Patterns are similar when formality is defined as having access to a pension, but with lower formality shares. Source: South Africa’s QLFS data 2008-2019 3rd quarter, authors’ calculation Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 106 unemployed (NEET 66), and the female NEET • South Africa’s labor market is segmented rate exceeds that for males (23 percent along multiple lines – formal versus compared to 15 percent; Stats SA 2019). These informal; public versus private; union versus data reflect a significant deterioration since non-union; white, black, mixed race and 2008, especially for males (Figure 4.40). others (Figures 4.41-4.43). These categories are sometimes overlapping, and sometimes • Youth typically struggle to access formal they compound the segmentation. For employment immediately after school but example, significant racial disparities in gain some access with age. Female youth work status compound income disparities, have achieved improved access to formal given that non-white South Africans – and employment over the past decade, while especially black South Africans – are less their male counterparts have lost access likely to access well-paid formal work. over time. Despite this progress by female Even though white South Africans account youth, a higher share of male youth are in for only 9 percent of the population, they formal work compared to female youth. On a account for 30 percent of the formally positive note, female youth are remaining in employed (Figure 4.45). school longer compared to a decade ago. Figure 4.41 South Africa’s labor force is segmented between formal and informal work status Unp id f mil work rs S lf- mplo d A ricultur Form l w mplo S lf- mplo d non-A ricultur (p nsion) Inform l w work r Inform l w work rs Emplo r, oth rs, unsp cifi d Form l w mplo (p nsion) Inform l Note: Formal status defined here based on pension access. Source: South Africa’s QLFS data for 2019 66 N  EET is defined as those outside the labor force (and therefore neither employed nor unemployed and looking for work) and not in school or other types of education or training. Note that South Africa’s Department of Higher Education and Training includes those who are unemployed (and therefore inside the labor force) in their definition of NEET (“Fact Sheet on ‘NEETs’”, Department of Higher Education and Training, 2017. https://www.dhet.gov.za/Planning%20Monitoring%20 and%20Evaluation%20Coordination/Fact-sheet-on-NEETs-Final-Version-27-Jan-2017.pdf) 107 Figure 4.42 Formality by race, public versus private sectors 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0 Inform l Form l Inform l Form l Inform l Form l Inform l Form l w w w w w w w w work r mplo work r mplo work r mplo work r mplo Public s ctor (2019) Public s ctor (2008) Priv t s ctor (2019) Priv t s ctor (2008) Afric n Colour d/mix Asi n Whit Note: Formal status defined here based on pension access. Source: South Africa’s QLFS data for 2008, 2019 Figure 4.43 Share of union membership by sector 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0 r s s s l on hi , r v nd on on s l in in t od in fi s ur t n ic ic ic ti Co To ic ur ti ti rv m r rv rv uc n d u lt w un r rr m to ct t nc ic tr nd r ric s ls rs m ,s qu cc , c s ns nt f nu co or t ss si n om nd s Co on st A th i d bu F m nd p n O rs M rn nd Tr ns , in P t ov ci Tr in r fo ri M l ct r El n G Union Non-union Do not know Source: South Africa’s QLFS data for 2019 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 108 South Africa experienced tepid economic per- (such as cleaning, accounting, secretarial, and formance over the past decade, accompanied security services; Burton et al. 2018a).67 68 Most by weak job growth as employment shifted services sector growth was in low productivity increasingly to the services sectors. South sectors such as personal services and construc- Africa is an upper middle-income economy in tion, as well as in moderate productivity sectors the late stages of structural transformation. Over transport and finance (Figure 4.44). Government the last 10 years, labor continued to shift away services and utilities also expanded at a robust from manufacturing and towards the services pace. Between 2008 and 2019, the manufacturing sectors and construction. The fastest growing sector contracted, declining from 16 percent to 12 services subsector was finance and business percent of total employment during the period, services, which has moderate labor productivi- while the wholesale and retail trade sector – by ty levels on average, employing a mix of skilled far the largest – grew only marginally. In terms and unskilled workers. In fact, much of this of number of jobs created, personal services and sector’s growth was driven by the expansion of financial and business services together account- temporary employment agencies, which act as ed for 78 percent of the 1.9 million net jobs added third-party contractors, placing employees in to the economy over the last decade, while the temporary positions across various occupations manufacturing sector shed nearly 300,000 jobs. Figure 4.44 Job growth in low- and high-productivity sectors (2008 - 2019) 700,000 Hi h v . productivit Gov rnm nt s rvic s dd d p r work r in 2019 (ZAR millions) 600,000 Minin & qu rr 500,000 El ctricit , s&w t r Tr nsport tion & 400,000 communic tion M nuf cturin 300,000 Fin nc & Low v . productivit busin ss s rvic s 200,000 100,000 Whol s l & r t ilin P rson l s rvic s 0 A ricultur Construction -100,000 V lu -2.0% -1.0% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% Annu l job rowth r t , 2008-2019 Notes: Bubble size reflects sector employment in 2008. Source: South Africa’s QLFS data 2008-2019 3rd quarter and SA STAT GDP P044 and P0441, authors’ calculation 67 Note that these Temporary Employment Services “employees” would be more accurately classified into other services sectors.  68 T  his trend is consistent with the observed increase in formal employment (defined as workers with a written contract), which does not reflect a significant improvement in average job quality, but rather a more modest improvement. 109 Most of the new jobs added to the economy in rates at which youth transition to formal in the last decade were either informal or wage work depicted in Figure 4.40 above quasi-formal temporary jobs, pointing in fact reflects the emerging prevalence of to labor market rigidities. For instance, temporary contracts.69 The pressure from the number of temporary jobs increased by this stress found different outlets. Some 30 percent from about 1.4 million to about youth gave up and became NEETs. Some 1.8 million between 2008 and 2019, while invested in more schooling to compete for permanent jobs increased at a much slower public sector jobs. Many took up temporary pace. Net employment among youth (ages work, either to earn a basic livelihood or as a 15-35) declined during this period by 546,000 pathway to a permanent contract. The result, jobs, while those aged 35-45 accounted unfortunately, is increased segmentation for three-fifths of the new jobs created. and rigidity. This not only further reduces Youth struggle to access permanent formal the opportunities for outsiders (e.g., youth) to employment, especially when we define access better jobs, but it also makes it harder “formal” as having access to a pension for policymakers to correct. (Figure 4.45). Much of the “improvement” Figure 4.45 Temporary versus permanent employment by age group (2019) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 A roup P rm n nt w mplo T mpor r w mplo (3 rs nd bov ) T mpor r w mplo (1 - 3 rs) T mpor r w mplo (l ss th n 1 r) Unsp cifi d Source: South Africa’s QLFS data 2019 3rd quarter 69 Note that temporary contracts tend to be written, so that employers can clearly stipulate the time-bound terms of employment.  Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 110 South Africa is a highly unequal society, Coal mining workers are largely permanent not coincidentally, and poverty rates workers (73 percent), although this is much have deteriorated in recent years. Nearly lower than during the late-1980s, when 38 percent of the population fell under the coal mines began to shift increasingly to $3.20 per day threshold in 2014/15 (up from contract workers (Burton et al. 2018a, citing 36 percent in 2010/11) and 19 percent were in Baartjes 2009). Workers in the coal mining extreme poverty (World Bank 2020c). Racial sector tend to have above-average levels and gender inequality is pervasive. Our of education: 70 percent have completed at regression analysis on labor market outcomes least a secondary education, compared to shows that: being a black South African 56 percent of manufacturing workers, 34 reduces the likelihood of being employed percent in construction, and less than 20 compared to other races; females have worse percent in agriculture (Figure 4.46). Six in employment outcomes than males; and low ten coal mining workers are union members educational attainment reduces the likelihood (above the national average of 30 percent), of being employed (Annex 3 Table 1). Large enabling collective bargaining to negotiate wage disparities are prevalent between compensation and working conditions. Other the different labor market segments. Even mining and quarrying employment have controlling for individual characteristics even higher union membership (80 percent) and sectors, regression analysis suggests and collective bargaining, similar to the significant implicit discrimination along public sector (Figure 4.47).72 Coal mining racial and gender lines, and a large wage workers are mostly black South Africans (82 premium for those in formal employment.70 percent in 2019), in prime working age, and Throughout the course of their working disproportionately male (83 percent). Most lives, workers must navigate these racial/ engage in production and machine operation gender/formality/union barriers to find occupations (Figure 4.48). the best-quality jobs. Where do coal sector jobs fit within this labor market context? Ninety-nine percent of coal mining jobs are formal (employed with a written contract), on par with the public sector (which includes general government services and electricity, gas and water71), and 77 percent of coal mining workers have access to some form of pension benefit – nearly double the national average of 41 percent. 70 Results from OLS regression estimates of the correlates of hourly wages show that black South Africans earn less  than other races even when accounting for education level and controlling for other characteristics. The gender wage gap is estimated at 18 percent, all else being equal, and formal workers earn 34 percent more than informal workers (not controlling for sector, due to extreme segmentation). Details of regression results are in Annex 3 Table 2. 71  Electricity, gas and water utilities are majority owned by the government (i.e., SOEs). 72 Note that coal mines are privately owned, similar to other mining and quarrying mines in South Africa.  111 Figure 4.46 Education breakdown by sector 100% Oth r 80% Post-s cond r 60% S cond r compl t 40% S cond r incompl t 20% Prim r 0 compl t r on s l s s on w s hi , l ls c s & tin , rv & on r v nt & spo l tr & m ,s d n in n r r in l Co t G ic ic ic t r un r ti ti ri L ss th n s t h ltu To qu in rn r un r i s c t l rv rv uc tu m nd , fis m to ss n ic & M ov n s cu t c tr i prim r ci rs n in r rs s l ns co r t ri ho f ri si F nu th ct r Co th W A O O M n El r bu fo Tr P Figure 4.47 Wage bargaining power by sector (2019) 100% Oth r 80% No r ul r nnu l s l r incr s 60% Emplo r d t rmin s 40% N oti tion b tw n work r 20% & mplo r B r inin council 0 w it , r on on s s s on hi , & N oti tion l l ls c s & tin , on r v nt rv & n in in r r in l t G ic ic ic t r un r Co ti ti ti ur nd ric s t h ltu To qu Min rn r m in i s c un r rv rv uc m b tw n union fis m to ss n od ic ct r ov n s ct cu tr m ,s rs om t n in rs s & mplo r ns ri f El co r t si F nu cc , c & Co th & spo A d O M r n bu Tr fo P Tr Figure 4.48 Occupation by sector (2019) 100% Oth r M chin op r tors 80% Production, cr ft work rs, l bor rs 60% Skill d 40% ricultur l S rvic 20% &s l s Cl rks 0 T chnici ns r on s s s on w s l hi , s & tin , rv & r v nt il & d n in in r r in l t Co ic ic ic r un r ti ti ic ur s t h ltu qu in r tr un or s c t l rv rv uc s m nd , fis ss n ic ct & rM ov n m t Prof ssion ls r s rn cu t tr m ,s ci ls rs n in G l ns t ri ho f ri & s nior offici ls co or si F nu th ct Co on th W A & sp O O rs M El n r bu fo Tr P Source: South Africa’s QLFS data 2008-2019 3rd quarter Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 112 The spillovers from a coal job can raise in the coal mining sector are higher than household incomes and reduce poverty, all other sectors, according to regression at least for the miner and his family. Jobs estimates that control for factors such as in the coal mining sector, like jobs in other education level, gender and race, among types of mining, pay significantly more on others.73 A recent World Bank study on the average than most sectors in the economy: 3.7 poverty effects of job creation in different times the average agriculture wage, 2.2 times sectors using South Africa data finds that the average construction wage, 2.7 times the mining and agriculture have the highest average wholesale and retail trade wage, and poverty-reducing impact; one additional job in 1.9 times the average manufacturing wage in these sectors is estimated to lift 1.3 people out 2019 (Figure 4.49). Only predominantly public of poverty (World Bank 2017). sectors pay more. The returns to working Figure 4.49 Nominal hourly wages by sector (2019 Rand) 120 100 80 60 40 20 0 r on s s s s l on w s r v nt & st , il & rv & d n in in fis r rr in Co t or r ic ic ic ic ti hi ti , f ltu ur qu in tr s m s c t l un r rv rv uc nd , m to ss n & rM ic ct rn r ls in icu t tr m ,s ci ls rs n in ns ho f ri ov nt r co r t si F th nu ct Co on th W hu A & spo O l O rs M El n r bu P n Tr G All provinc s Co l Provinc (Mpum l n ) Source: South Africa’s QLFS data 2019 3rd quarter 73 Refer to wage regression results in Annex 3 Table 2.  113 Figure 4.50 Coal mining employment by province (2008-2019) 80,000 io d Bu Mpum l n m p r st B oo 70,000 p rio W st rn C p d 60,000 G ut n 50,000 Limpopo 40,000 Kw Zulu-N t l Fr St t 30,000 E st rn C p 20,000 North rn C p 10,000 North W st 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: South Africa’s QLFS data 2008-2019 3rd quarter Coal activity is concentrated in The growing presence of coal mining jobs Mpumalanga province, with more modest helped Mpumalanga’s economy grow, mining activity in Free State, Gauteng and but these positive growth spillovers were Kwazaulu-Natal provinces (Figure 4.50). undercut by the distorting effects of high Mpumalanga is relatively rural and is home coal wages. The economics literature has to 93 percent of South Africa’s coal mining documented the negative dynamic effects employment. Coal mining jobs account for 6 associated with boom and bust extractive percent of Mpumalanga’s total employment industries, as discussed above, and illustrated and are most highly concentrated in four in the case studies on the US and Indonesia. municipalities: eMalahleni (26 percent), Is the situation different in South Africa? Steve Tshwete (17 percent), Msukalingwa We answer this question by testing for (14 percent) and Govan Mbeki (11 percent; statistically significant spillover effects within TIPS 2020). In 2015, coal contributed 19 Mpumalanga province using a methodology percent of the region’s gross value added similar to Black et al. (2005) and described in (Strambo et al. 2019). Annex 3. Our regression results suggest that, over the last decade, coal mining jobs had Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 114 positive spillover effects in terms of spurring for coal workers. Since the average miner more job creation in Mpumalanga’s non- in Mpumalanga supports three dependents mining sectors – particularly in the services (Strambo et al. 2019), coupled with the fact that sector (also in manufacturing, albeit with alternative jobs are likely to pay significantly lower explanatory power, and no significant lower wages, displaced miners and their spillovers in agriculture or construction).74 families will experience large income shocks The effects were not uniform over the even if alternative work is found quickly. The entire period, however, because the sector degree to which lost coal jobs exacerbate experienced a mini-boom between 2008 inequality in the long run is unclear, and 2014, and a mini-bust from 2015 to 2017, however, given coal workers’ above-average after which coal employment rebounded. We incomes and, for most, formal employment exploit this variation to test whether coal jobs status, union representation and pension crowded out other sectors during the boom access. In the post-transition context, wage period, which would be reflected by faster distortions created by coal sector jobs will no non-mining employment growth during the longer be present to skew local labor market bust period. The regression findings indeed opportunities in Mpumalanga’s principal point to accelerated job growth in all non- coal centers. mining sectors during 2015-2017, suggesting a potential crowding out (full regression results Policies to ease the transition for in Annex 3 Table 3). The rise in agriculture displaced coal workers have an essential employment during the coal bust could partly role to play. Although South Africa’s be the result of displaced coal workers being national dialogue on just transition is well absorbed into farm work, but this does not advanced (WRI 2020)75, it is encumbered by appear to be the case in construction. a very challenging labor market context and complex institutional setting. Any future transition away from coal production will leave coal regions and their communities vulnerable to negative employment effects, at least in the short run. Mpumalanga’s coal-intensive municipalities may be especially hard hit, given their high shares of coal employment. A more diversified economy that offers alternative local employment opportunities is crucial for absorbing future declines in labor demand 74 Note that other mining employment is not significant, implying that the observed distortion effects come from coal  rather than other resource extraction jobs. This is consistent with the fact that other mining employment accounts for a very small employment share in the province. 75 I  n September 2020, the Government established a Presidential Climate Change Coordinating Commission to ‘coordinate and oversee the just transition’ across government ministries and public agencies (Department of Forestry, Fisheries and the Environment media release September 13, 2020). 115 4.6 India: Producing to Meet its Own Massive Demand India’s coal production has grown GWh, and 60 percent of these plants were built steadily over the past 40 years, increasing within the last nine years.76 Although new nearly five-fold. In 2019, India accounted plant construction continues, many plants for 9 percent of global coal production were cancelled over the past decade (395 coal- (BP Statistics). Most of India’s extracted fired plants with a total 565 GWh capacity), coal is non-coking bituminous coal, and and a small number were retired. is consumed (in final use) primarily by the industrial sector (Figure 4.451). The The coal sector’s long history in India and industry sector also significantly increased its transformation from its roots under its electricity consumption in the last two colonial ownership pre-Independence to decades (largely coal-fired), a time of robust being a strategic national priority provide and sustained economic growth. Between insight into coal production patterns 1980 and 2019, GDP growth averaged 6 observed over the past half-century.77 percent per year in real terms, – accompanied Coal played an important part in India’s by similar per capita income gains. During industrialization, both as a raw input to this period, India’s population grew by 670 production as well as for power generation. million, representing a massive increase in With rising residential electrification as potential energy consumers. well as households switching to coal-fired cooking ovens in the 1960s, the demand Coal dominates the electricity mix, and for coal rose sharply. Until the mid-1990s, even expanded its share in the last decade. India’s rising coal consumption was met by Electricity generation grew five-fold between domestic production, but thereafter, coal was 1990 and 2019, and coal-fired electricity increasingly imported (Figure 4.53). By 2019, expanded even faster (Figure 4.52). In over 30 percent of total coal consumption was 2019, coal accounted for 71 percent of total supplied by external sources, predominantly electricity supply. Natural gas and hydro by Indonesia and South Africa (BP Statistical power made some headway beginning in Review of World Energy 2020). 2000, while wind and solar have only recently picked up, but none as a replacement to coal. According to the Global Energy Monitor, India’s 281 operational coal-fired power plants have annual generation capacity equal to 229 76 More than half of these plants risk being stranded after 2030 if India were to pursue policies in line with the Paris  Agreement (Malik et al. 2020). 77  Lahiri-Dutt (2016) provides a rich historical perspective of the economic, political and socio-cultural drivers of coal activity. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 116 Figure 4.51 Coal final consumption by sector (ktoe) 120,000 100,000 Industr 80,000 Tr nsport R sid nti l kto 60,000 Industr Industr Comm rci l 40,000 nd public s rvic s 20,000 Non-sp cifi d Non-sp cifi d 0 ’90 ’92 ’94 ’96 ’98 ’00 ’02 ’04 ’06 ’08 ’10 ’12 ’14 ’16 ’18 Source: IEA data for India Figure 4.52 India’s electricity generation by source (GWh) 1,800,000 1 El ctricit consumption/c pit P r c pit consumption (MWh) n r tion (GWh) 0.8 W st 1,300,000 Biofu ls 0.6 Sol r PV 800,000 Wind 0.4 H dro Nucl r El ctricit 300,000 0.2 N tur l G s Oil -200,000 0 Co l 1990 1995 2000 2005 2010 2015 2019 Source: IEA data for India Figure 4.53 Coal demand outstrips supply 20 18 16 Exc ss d m nd 14 m t b imports Ex joul s 12 10 8 Consumption 6 4 Production 2 0 ’81 ’83 ’85 ’87 ’89 ’91 ’93 ’95 ’97 ’99 ’01 ’03 ’05 ’07 ’09 ’11 ’13 ’15 ’17 ’19 Source: BP Statistical Review of World Energy June 2020 117 Figure 4.54 India’s CO2 emissions by sector (Mt CO2) 2,500 2,000 1,500 El ctricit & h t produc rs Mt CO2 1,000 Industr 500 Tr nsport R sid nti l 0 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 El ctricit & h t produc rs Oth r n r industri s Industr Tr nsport R sid nti l Comm rci l & public s rvic s A ricultur Fin l consumption not ls wh r sp cifi d Source: IEA data for India Figure 4.55 Factors behind India’s rising CO2 emissions (2010-2018) +48.4 +25.6 +66.9 +112.2 -0.1 +171.6 -39 +805.7 -214.5 2271.2 -235.3 1529.7 oth r Popul tion oil co l s CO2.2010 CO2.2018 Industr : s cond r CO2 int nsit Industr : t rti r Int nsit GDP/c pit n tur l En r En r En r En r En r Source: Guan et al. (2021) Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 118 State control of the coal sector has shaped India’s rich coal deposits are a long way its trajectory and central role in India’s from being depleted. The Ministry of Coal energy policy. Coal mining was accorded reports that even during the past five years, priority status by the state in 195778, and coal significant new reserves have been identified. mining was nationalized in the early 1970s, The 2020 figures indicate coal reserves of 344 ostensibly to ensure energy security in the billion tonnes – 162 billion tonnes proved, context of global oil price shocks and increase 151 billion tonnes indicated, and another 31 the efficiency of and control over diffuse billion tonnes inferred. Jharkhand, Orissa and inefficient production practices (Coal and Chhattisgarh states have the largest India Ltd.). Coal India Ltd., the state-owned reserves, together accounting for 70 percent enterprise created to manage all coking and of total national reserves (MOSPI 2021). non-coking mines, is the dominant producer today, accounting for over four-fifths of total Coal-based power generation has large coal production (2019 data from Coal India negative health and environmental effects Ltd). Coal India Ltd. operates 364 mines79, of on the Indian population. In line with rising which 166 are underground, 180 are opencast coal consumption, India’s carbon dioxide and 18 are mixed mines (IEA 2020). While emissions have quintupled since 1990 (Figure much of India’s economy has undergone 4.54), although its contribution to global liberalization, the state remains heavily emissions has been relatively modest (IEA involved in the sector and Coal India Ltd. has 2021a). Electricity and heat producers are the retained a quasi-monopoly position. The coal largest emitters, followed by the industrial sector, and the Ministry of Coal specifically, and transport sectors. Emissions from coal- are influential in India’s national energy fired power plants result in between 80,000 policy, which helps explain the country’s and 115,000 premature deaths per year reluctant shift to alternative energy sources, (Urban Emissions et al. 2013). Decomposing even when they are cheaper (Spencer et al. CO2 emissions into its contributing factors 2018). Montrone et al. (2021) point out several shows that for the period 2010 to 2018, the obstacles to reforming the sector, including increase in CO2 emissions was due in greatest the politically popular subsidy of electricity measure to rising per capita incomes and through low tariff rates (also noted in the growing population; these pressures Mahadevan 2019 and Min and Golden 2014; were partially offset by declines in energy note that power generation and distribution intensity and energy use by the industrial are also state-controlled), as well as job- sector (Figure 4.55).81 creating public investment projects along the coal supply chain, especially in the railway sector.80 78 Coal Bearing Areas (Acquisition and Development) Act 1957, cited in Lahiri-Dutt (2016).  79  Out of a total 459 operational mines (Pai and Zerriffi 2021). 80  amboj and Tongia (2018) and Tongia and Gross (2019) argue that coal provides an important revenue stream for central K and regional governments and non-coal SOEs, notably Indian Railways (India’s largest employer; Montrone et al. 2021). 81  India’s energy intensity, defined as energy consumption per unit of GDP, declined from 1990 to 2016, which helped to constrain the upward pressure on CO2 emissions (Worldindata 2021). The reduction in industrial demand for coal from 2010 onward coincides with the flattening consumption curve in Figure 4.6.1. 119 India’s coal sector has long been considered 2015; Chikkatur et al. 2009). Between 2000 and an important source of employment, but 2014, the sector's labor productivity increased labor demand is declining. According to by an estimated 6.6 percent annually (Spencer estimates using the national Employment- et al. 2018). As coal mining becomes less labor- Unemployment Survey (EUS) and Periodic intensive – even if coal production levels remain Labor Force Survey (PLFS) data82, the coal unchanged – the sector offers fewer formal mining sector83 accounted for 888,000 direct and/or well-paid employment opportunities, jobs in 2004, following which the sector’s including for India’s large pool of young workers. employment steadily contracted, falling by more than half to 416,000 in 2017 (denoted by Formal coal mining jobs – defined here blue bars in Figure 4.56). Other estimates of coal as having a written contract84 – are mining employment are much higher. Lahiri- concentrated in but not limited to Coal India Dutt (2016) puts the number at nearly double Ltd. Employment in Coal India Ltd. accounts these figures, suggesting that the EUS and PLFS for the largest share of coal mining jobs – do not fully capture informal coal workers, between one-half and three-fourths – and especially in remote rural communities. Pai this share has grown over the last decade, and Zerriffi (2021) estimate total coal mine implying that Coal India Ltd. shed jobs at a employment at 745,000 in FY2020, based on slower pace compared to the rest of the sector. a new database they compiled using labor- It is important to note that these figures intensity factors. exclude the large number of sub-contract workers performing mining-related services The decline in coal mining employment for Coal India Ltd. but hired on informal observed since 2004 was particularly driven terms by “job-companies” (Lahiri-Dutt 2016). by a reduction in formal rather than informal The past three years have seen more rapid coal jobs as Coal India Ltd. shifted increasingly shedding of direct employees by Coal India to informal contract labor (Figure 4.57). This Ltd., where employment fell to 260,807 at end reduction coincided with improvements in labor December 2020 (Coal India Ltd.). Figures 4.56 productivity associated with mechanization and 4.57 illustrate that other types of mining and other efficiency improvements including and quarrying provide significantly more those related to increased reliance on opencast employment than coal mining, and that these rather than underground mining (Henderson non-coal jobs are most likely to be informal. 82  ndia’s employment data comes from the World Bank Jobs Group’s Global Labor Database (GLD), which harmonizes national I and subnational surveys across a set of labor market variables. The GLD pulls data from India’s Employment-Unemployment Survey (EUS) and the Periodic Labor Force Survey (PLFS), both conducted by the National Sample Survey Organization (NSSO). Data are representative at the state level. Since 1972, NSSO has collected labor market information with the EUS every five years. In 2017, the PLFS was launched to replace the EUS and provide higher-frequency data on broad labor-force indicators. The EUS and PLFS are the most comprehensive household-level sources of employment data in India, albeit with well-documented sectoral and geographical coverage limitations. For example, neither data set reports any workers in subsistence coal mining activities, although field work by Lahiri-Dutt and Williams (2005) documents significant informal employment in the coal sector. Srinivasan (2006) provides details on coverage limitations in EUS. The change in states’ geographical composition between survey rounds complicates consistent state-level comparisons over time; we addressed this by creating a new state ID that re-constitutes the original state-district mapping. For example, Telangana split from Andhra Pradesh in 2014, but in our analysis we retain Andhra Pradesh’s district composition before the split and apply it to the 2017 survey data. Finally, we adjusted the survey weights of each state to align with more recent state-level population estimates by the Office of the Registrar General and Census Commissioner (https://www.censusindia.gov.in/). 83 Coal mining sector includes hard coal as well as lignite. 84 The PLFS does not include information on pension or social insurance coverage, criteria typically used to define formal employment. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 120 Figure 4.56 Mining sector employment trends 3,000,000 2,849,133 2,714,098 2,499,290 2,500,000 2,000,000 1,828,969 1,500,000 1,000,000 887,969 795,176 593,760 493,061 500,000 419,214 390,243 416,240 313,829 0 2004 2009 2011 2017 Co l Indi Ltd Co l All Minin & Qu rr in Note: Coal India Ltd data are administrative data, whereas the EUS and PLFS are survey-based, so the relative magnitudes may not be consistent. Sources: Coal India Ltd.; Global Labor Database (India), World Bank Jobs Group (forthcoming) Figure 4.57 Formality breakdown of coal and other mining jobs 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0 Co l Oth r Minin Co l Oth r Minin & Qu rr in & Qu rr in 2017 2004 Inform l w d Unp id Own ccount Form l Note: Formal defined as having a written contract. Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation 121 India’s coal sector is segmented; parallel degree to which the EUS and PLFS accurately to the state’s direct engagement in coal capture these small artisanal or subsistence mining, coal is also mined by a range of coal producers; most likely, they are private producers, at both large (formal) underrepresented in the labor market data.88 and small (informal) scales. Beginning in 1976, coal mining was allowed by private Finding formal paid work in any sector is a firms producing iron and steel (World challenge across India’s economy, especially Bank 2021c). And since 1993, mining leases for the growing youth segment of the labor have been granted to private “captive” force. Over nine-tenths of India’s workers are operators that supply thermal power plants informally employed, whether in farming, (on concessionary rather than competitive own account work, or as informal wage contracts); these operations – which workers (Figure 4.58). Moreover, the rate of account for an estimated 6 percent of total informality seems to be increasing; between coal production – report relatively low 2004 and 2017, 99 percent of the 23 million net formal employment, are owned by Indian jobs added to the economy were informal. In entrepreneurs, and at least some have net terms, 26 million fewer youth aged 15-35 significant foreign investment (Lahiri-Dutt were employed in 2017 than in 2004. With 2016).85 Artisanal and subsistence coal mining 1.37 billion people, India is the world's second also takes place, whether on privately-owned most populous country, and its population land or rural commons, both of which are is expected to rise to 1.5 billion by 2030. In unregulated and thus illegal.86 A fourth terms of demographic structure, India has category of mine producers identified by one of the world’s youngest populations, Lahiri-Dutt (2016) comprises private land with more than 54 percent under 25 years old owners and local indigenous communities (Figure 4.59; Sharma et al. 2019). Despite its in the remote state of Meghalaya, which was young population age structure and rising granted special status for political economy educational attainment, only 24 percent of the reasons. Coal producers in Meghalaya working age population in 2017 had completed therefore fall outside the regulatory secondary schooling (albeit a marked framework. Own account and subsistence coal improvement over 2004, when the share was workers (sometimes called coal collectors) only 14 percent). have much lower productivity than large mining operators87, and tend to sell in small quantities to local consumers. It is unclear the 85 Note that captive coal mining is also permitted to support cement production (World Bank 2021c).  86 T  here is a large literature on the impact of extractives on the residents of local rural communities in India, including in coal regions where many have been displaced from traditional farming and forestry activities on their own (non- deeded) land or community commons (Lahiri-Dutt and Williams 2005; Padel and Das 2010). Indigenous and tribal communities have been particularly affected (Lahiri-Dutt 2016). 87 L  ahiri-Dutt (2016) estimates labor productivity among the larger producers (Coal India Ltd and the private collieries supplying power plants) at 1200 tonnes/worker annually, 33 times the estimated production of an informal subsistence coal collector delivering locally with his bicycle (36 tonnes/worker annually). 88 A  ccording to PLFS 2017, 86 percent of coal sector workers are public employees (and therefore mostly work for Coal India Ltd.), and 90 percent report a firm size over 20 workers. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 122 Figure 4.58 Labor force breakdown by work status (ages 15 - 64), 2017 2% Unp id f mil work 8% 14% F rm r Own ccount 23% Inform l w ( ricultur ) 33% Inform l w (non- ricultur ) Emplo r 19% Form l w 1% Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation Figure 4.59 Population by five-year age groups, 2030 projection 100+ 90-94 80-84 70-74 60-64 roup 50-54 A 40-44 30-34 20-24 10-14 0-4 100,000 80,000 60,000 40,000 20,000 0 20,000 40,000 60,000 80,000 100,000 1,000 p opl M l (m dium f rtilit sc n rio) F m l (m dium f rtilit sc n rio) Source: Jobs Group Demographic Tool, using WDI data 123 This difficult labor market environment in unpaid family work compared to older provides limited opportunities for youth, age groups. Male youth are most likely to who are increasingly opting to remain find informal wage work – often after a outside the labor force. Female youth in spell of unemployment – and their access to particular have very low participation rates, formal work improves when in their late 20s and these have deteriorated over the past 15 and early 30s. These patterns of school-to- years. About 30 percent of young people are work transition represent an improvement neither in employment, nor in education or compared to 2004, when larger shares of male training, nor unemployed (NEET). The NEET youth were farming or self-employed. Female situation is far worse for young females, 62 youth, by contrast, are less attached to the percent of whom are NEET compared to 3 labor force today; on the positive side, more percent of male youth (Figure 4.60). Youth who female youth are accessing informal wage jobs. do enter the labor force are disproportionately Figure 4.60 School-to-work transition by age (2017, 2004) M l , 2017 M l , 2004 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 15 17 19 21 23 25 27 29 31 33 35 15 17 19 21 23 25 27 29 31 33 35 F m l , 2017 F m l , 2004 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 15 17 19 21 23 25 27 29 31 33 35 15 17 19 21 23 25 27 29 31 33 35 Schoolin Un mplo m nt F rm r Inform l w ( ricultur ) Emplo r NEET Unp id F mil work Own ccount Inform l w (non- ricultur ) Form l W Note: This graph shows a static plot of male and female youths’ work status by age, and does not capture dynamic transitions. Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 124 125 Coal mining jobs are of relatively high commerce or construction sector workers, and quality on average compared to most 70 percent compared to the manufacturing other sectors. Jobs in the coal mining sector. This positive and strongly significant sector are more likely to be formal (only the correlation is even larger when we restrict government sector has a higher share of the estimation to the six main coal states, formal employment) and they pay higher highlighting coal’s relative importance within wages on average (Figures 4.61 and 4.62). these regional labor markets (Annex 4 Table 2). Workers formally employed by Coal India Ltd. benefit from union representation and public Coal mining waged jobs – whether formal or sector wage setting and other job protections. informal – are mostly taken up by men (94 Regression analysis on the correlates of wages percent), and tend to engage workers with for wage earners captured in the PLFS (that relatively less education and in relatively is, excluding subsistence workers and own lower-skilled occupations. According to the account workers and possibly undercounting PLFS 2017 data, coal mining workers have informally employed wage workers in the lower than average educational attainment, on coal mining sector) indicates that coal wages par with the wholesale and retail trade sector are sharply higher than all other sectors, and the transport sector; 69 percent have controlling for education level and other less than a complete secondary degree. About characteristics (regression results reported a third of coal mining employees are craft in Annex 4 Table 1). Even controlling for workers, 23 percent are machine operators and formality, coal mining workers earn a wage 28 percent engage in elementary occupations. premium around 80 percent compared to Figure 4.61 Work status breakdown by sector (2017) 100% Form l w 90% Emplo r 80% 70% Inform l w (non- ricultur ) 60% 50% Inform l w ( ricultur ) 40% 30% Own ccount 20% F rm r 10% 0% Unp id s s il on n d ti & rv & s ur in in l tio on u in ic ic t s ti ti t fi ur Co lt Q in c ili ci uc rr un r n R cu tr & M m po ct ut tr & is p n in ri S r m ns ns ic ns in f si F s th nu l bl A s m ,U Co Co r s O Pu Ad T rs M l ho ic Bu th bl W Pu O Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 126 Figure 4.62 Nominal weekly wages by sector (2017 Rupees) 9,000 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0% s s il on n d ti & rv & s ur in in l tio on u in ic ic t s ti ti t fi ur Co lt Q in S c ili r ci uc rr un r ss n R cu tr & rM m po ct ut Av tr & is p n in ri m ns ns ic ns in f si F th nu l bl A m ,U Co Co r s O Pu Ad T rs M l ho ic Bu th bl W Pu O All St t s All Co l St t s Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation Coal mining activities are geographically among older mines and underground mines concentrated in six states: Andhra Pradesh,89 that were no longer profitable; moreover, Jharkhand, West Bengal, Orissa (also about half of the mines currently operating referred to as Odisha), Chhattisgarh and in Jharkhand are not profitable (World Bank Madhya Pradesh, hereafter referred to as 2021c). Despite a significant presence of coal “coal” states.90 Jharkhand alone is home to mining activity, the share of coal mining jobs a quarter of India’s coal reserves. Each state in total state employment was only 2 percent has experienced fluctuations in coal mining in Jharkhand and even smaller elsewhere, employment since 2004, but a declining trend suggesting limited coal dependence at the on net (Figure 4.63). Jharkhand experienced state level (although with some highly coal- significant coal mine closures, especially dependent districts). 89  ote that Telangana split off from Andhra Pradesh in 2014, but we retain them in a combined state for this analysis to N enable comparisons over time. 90 Other states in which significant coal production takes place include Maharashtra (by volume) and Gujarat (by employment).  127 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 128 Figure 4.63 Coal mining employment in 6 states (2004, 2009, 2011 and 2017) 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 2004 2009 2011 2017 Jh rkh nd W st B n l Andhr Pr d sh Oriss Chh ttis rh M dh Pr d sh Note: Data for Andhra Pradesh includes Telangana. Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation These “coal” states face greater-than- rates – in communities that experienced average economic development challenges, coal mine closures, specifically Bokaro, and are likely vulnerable to coal-related Jamtara, Hazaribagh and Ramgarh districts shocks. The “coal” states are characterized in Jharkhand. The survey data do not allow us by weaker socio-economic performance to test for district-level economic spillovers of compared to national averages, notably with coal employment (similar to the analysis for respect to wage levels (Chhattisgarh and West Indonesia) because data are not representative Bengal are particularly low) and education, at the district level. Nevertheless, it is likely among others (Figures 4.64 and 4.65). Four that future coal mine closures risk upending of these “coal” states – Chhattisgarh, workers’ livelihoods, not only for those Jharkhand, Madhya Pradesh and Orissa – directly employed in mines, but also their rank among the 10 poorest states in India families and those engaged in downstream (Reserve Bank of India 2019). Bhushan et al. industries reliant on coal, such as coal (2020) find large negative economic outcomes washeries, steel and cement plants – namely higher unemployment and poverty (PEG-CPR Roundtable 2021). 129 Figure 4.64 Nominal weekly mean wage in coal states (2017 Rupees) 3,000 2,500 2,000 1,500 1,000 500 0% rh s s nd sh l sh s ris t t n d d h t t is O rk lS lS B Pr Pr tt st Al Jh o h hr lC W Ch dh Al An M Note: Data for Andhra Pradesh includes Telangana. Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation Figure 4.65 Educational attainment in coal states (2017) 100% 90% 80% 70% Post-s cond r 60% S cond r compl t 50% 40% S cond r incompl t 30% Prim r compl t 20% 10% L ss th n prim r 0% rh nd s sh l sh s t ris n d d h is t rk O lS B Pr Pr tt st Al Jh h hr W Ch dh An M Note: Data for Andhra Pradesh includes Telangana. Source: GLD (India), World Bank Jobs Group (forthcoming), authors’ calculation Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 130 The future speed of adjustment in the Coal’s trajectory in India has created energy sector is uncertain, complicated paradoxical effects. State ownership of by multiple governance-related factors. reserves has facilitated coal’s continued Based on India’s “Stated Policies Scenario” centrality in India’s energy policy even in the developed by IEA consistent with current post-liberalization period. The increasing energy policies, energy demand is likely to energy needs of India’s rapidly growing grow by 35 percent by 2030, driven not only economy, together with the government’s by increased (non-coal) transport activities economic and energy security objectives, are but also increased use of air conditioning used to justify continued reliance on coal. (IEA 2021a). The energy needed to keep up Coal production decisions are influenced with India’s projected population growth by political objectives including sustaining over the next two decades will require added direct and indirect public employment. A very power generation equivalent to the EU ‘s high level of implicit subsidy is propping up total current electricity generation capacity. unprofitable activities in India’s large public But even with increasing demand for coal, sector, which also includes power generation, India’s coal sector faces productivity and power distribution networks, and Indian efficiency challenges that in many cases Railways, among others.91 Taken together, render production unprofitable. And yet, the these factors have delayed the shift to cleaner adoption of alternative energy sources has energy sources, despite the environmental been slow. This is partly due to technological and economic rationale for doing so. obstacles linked to storage capacity and the Paradoxically, the positive spillover effects of need to connect renewable energy to the grid coal mining jobs are offset by crowding out of and the need for interstate grid integration alternative industries, limiting job creation (CIF 2021), but it is also explained by the state’s and diversification of local economies and direct engagement in the coal sector and coal- exacerbating dependence on coal. According fired power generation. Vested interests and to Lahiri-Dutt (2016), political patronage that spill beyond the coal sector create pressure to maintain the status “The need for coal…is often couched in the quo (CIF 2021). Coal-related taxes and levies language of nation-building in India. This also represent an important source of states’ addiction to coal drains local livelihoods and revenues (World Bank 2021c). The landscape degrades the environments but remarkably, is beginning to change, however, reflected at the same time, creates coal dependent in more ambitious Government targets for livelihoods in the same areas.” p. 10. renewable energy capacity (227 GW by 2022 compared to 96 GW as of May 2021; IBEF 2021) and a number of recent Government regulatory and fiscal initiatives (e.g., such as an incentive scheme to support solar equipment manufacturing). 91 Tongia and Gross (2019) note that coal accounts for close to half of Indian Railways’ freight revenues.  131 CHAPTER 5 Policies for Managing the Transition Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 132 5.1 Key Findings on the Magnitude of the Challenge At the global level, coal-based energy plants. Even among countries committed to production has risen steadily over the past 40 transitioning away from coal, the marginal years, to a large degree driven by rising energy cost of continued coal extraction to power demand by the industrializing economies of electricity generation is much lower than the the world. Increased electricity consumption cost to replace installed generation capacity. is the main component of this energy In addition to the enormous cost implications, demand, and coal is the largest fuel source for other factors impede the transition away from electricity worldwide. The developing world coal, such as technical and logistical challenges more than doubled its per capita electricity to convert to new electricity sources, energy consumption since 1990, converging toward security concerns, foregone export revenues the high consumption levels prevalent in from coal and its derivatives, and the desire advanced economies. by governments to avoid dislocating current producers and affected workers along the Even as many of the former coal powerhouses coal value chain. In addition to coal’s use in in Europe as well as the U.S. are transitioning electricity generation, it is an input into many away from coal and shifting their priorities manufacturing supply chains, including toward alternative sources of power generation, as a direct or indirect input into metal and they have been replaced by rapidly scaling chemical manufacturing (e.g., steel), paper coal extraction in other regions of the world. and wood products, construction materials, China is by far the largest coal producer today, textiles and food processing, among others. The meeting not only the rising electricity needs manufacturing sectors in emerging economies of its massive population but also fueling tend to be more concentrated in coal-intensive its industrial sector, the engine of China’s sub-sectors, but as countries progress toward remarkable growth story. India similarly took an advanced stage of structural transformation, advantage of its coal resources to facilitate coal plays a relatively small role in increasingly industrialization through inexpensive energy, services-based economies. enabling energy-intensive firms to be more competitive and stimulating household Globally, coal mining jobs number 4.7 million, demand for electricity. Indonesia’s more recent accounting for less than one percent of scale-up of coal activity was motivated not employment even in the main coal producing only by rapid growth in domestic demand, but countries. Over 2 million coal mining jobs also in response to the flourishing coal export have been lost over the last decade, reflecting market. In a similar vein, Australia and South coal phase-out in some countries, expansion Africa have aggressively expanded their coal in others, and sector productivity gains in production, incentivized to a significant degree most countries, as extraction technology by potential export revenues. has become more capital-intensive. Despite a modest role in national labor markets, coal Despite a rising awareness of the destructive jobs disproportionately affect local labor effects of coal mining and coal combustion markets through positive spillover effects in exacerbating climate change, many that at the same time limit or crowd out countries rely on coal for a large share of their economic activity in other sectors because of electricity needs through coal-fired power wage distortions that depress labor demand. 133 Moreover, the boom and bust cycles associated electricity needs; limited economic diversity with extractives industries in particular tend in coal communities; weak regulation and to limit economic diversification, making local sometimes regulatory capture; political economies vulnerable to large demand swings economy pressures that shape government that undermine long-term growth. decision-making; and the potentially disruptive impact on livelihoods and the The five case studies presented in this report economic viability of coal communities. illustrate these effects, albeit to different degrees, given heterogeneous country Policymakers need to understand the ways in settings. Multiple factors affect the observed which a future transition away from coal may coal employment patterns, but some common affect the welfare of both coal and non-coal features emerge that impede transition. workers and their surrounding communities, These include: rising market demand for coal in order to create the policies and programs to – whether domestic or external – to meet manage transition effectively. 5.2 Lessons from Past Transitions The experiences of past episodes of coal • Transition requires a comprehensive transition in the U.S. and Poland, and to a approach with complementary initiatives, lesser degree in India, provide some lessons policies and incentives to sway the many for policymakers and local development actors along the coal supply chain, from and planning authorities who anticipate producers at the top (i.e., mine operators, future coal phase-out. There is no recipe for power plants) all the way to consumers success, unfortunately; many of the transition (buyers of coal for heating stoves or experiences described in this report as well household electricity). Policies need to be as those in other countries facing similar designed with the many stakeholders in transitions were quite negative. The following mind, including those with vested interests lessons nevertheless provide insights and like utility monopolies and manufacturers guidance for planning more effective and less of mining equipment and coal stoves. costly transitions. • The timing and speed of transition are subject • Transition takes a long time. Most coal to political economy dynamics. Uncertainty sectors developed over many decades, around commodity prices makes it difficult cultivating links across national economies. for communities to adjust for the “natural When many workers, businesses and resource curse” because prices affect communities are implicated, adjusting to a both willingness and capacity to diversify fundamental change in one industry cannot toward other industries. Where actors are happen quickly, even with the best advance public – such as in state-owned mines and planning and post-closure transition power plants (e.g., in Poland and India) – policies in place. governments have the power to act quickly but risk the future support of the electorate. Where actors are private but unions are strong (e.g., Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 134 South Africa) and/or government capacity and • The advantages of inducing voluntary job regulatory authority are weak (e.g., Indonesia) separations through generous compensation or captured by private interests (e.g., U.S. to miners are offset by the risk of inflicting coal communities captured by an “elite” of long-term damage on local economies if surviving coal-connected families), boom/bust prolonged income support further distorts cycles can be exacerbated, creating obstacles local wages or ex-miners permanently to both the design and implementation of exit the labor force. High reservation effective transition policies. wages dampen local labor demand and economic recovery through diversification, • Transition assistance programs targeting and premature labor force exit by a large formal mine workers fall short of meeting component of the population (as observed in the needs of informal workers in and around Poland) reduces the demand for local goods the mines. Informal mine workers are at and services and can directly undermine greater risk than their formal counterparts for public fiscal health if affected households several reasons: they lack severance rights and qualify for long-term social assistance. other basic labor protections such as advanced notice of layoff; they are ineligible for social • Severe social dislocation and local insurance programs such as unemployment economic viability may pass a point of benefits; and they earn much lower incomes no return. The risk is higher where long- and are therefore less able to weather income term dependence on coal has delayed shocks. Even large mine operators employ acceptance of transition. When local job a significant share of their workforce on losses – whether directly or indirectly the temporary and/or informal contracts. The result of coal mine closure – are significant risks are likely even greater for informal to the point of stimulating human capital workers in the coal value chain or in other flight while stranding those with the lowest local sectors that are indirectly sustained by capacity to find alternative work, the mine employees’ spending. The displacement localized economic malaise can spill over of informal workers dependent on the coal to persistent labor productivity and welfare sector for their livelihoods is particularly losses, deteriorating public services harmful to low-income households. and outmigration. • Remoteness and small market size are • Economic diversification is essential and mutually reinforcing impediments to requires help from both local and higher transition. When communities are not level government with respect to planning connected to larger markets, workers and financial resources. Advance planning, cannot access jobs elsewhere and local investment in infrastructure, addressing businesses are limited by their small local environmental degradation and attracting client base. “Bonding” social capital (what private investment are key ingredients of binds a community together; Lobao et al. economic diversification. These in turn 2021) may be strong, but “bridging” social require local and regional institutional capital (which fosters connectedness across capacity and coordination. A large negative groups) is needed to build cooperation and shock requires financial support beyond collaboration among local institutions, local government capacity. businesses, and governments. 135 5.3 Policy Framework for Managing the Labor Impact of Coal Transition The five country cases illustrate the capacity to absorb all potentially displaced significant and wide-ranging challenges workers? How well communities adjust to the posed by coal transition, and highlight the shock will depend on many factors including the need for effective policies to address them. size of the shock relative to the local economy, We use the lessons from past transitions in workers’ capacity to access alternative jobs, the Poland and the US, together with the case local economy’s ability to attract investment in study findings on coal sector and labor market alternative business ventures, and connectivity dynamics in Indonesia, South Africa and India, to larger markets, inter alia. to motivate the design of a multi-channel policy framework for managing the impact of A comprehensive policy approach must coal transition on workers. be multifaceted, multi-stakeholder, and span several layers of government and Large-scale coal mine closures risk wide, several government ministries – and all of deep and prolonged negative effects on local this requires planning, coordination, and communities and their economic viability; strategic, risk-informed decision-making in addressing these challenges effectively advance of the mine closure and throughout requires a solid understanding of the scope the closure process. Developing an effective and nature of the potential impacts. As we policy framework is further complicated by the consider the future prospects for coal-related reality that informal workers – an important jobs in the context of eventual downscaling of segment of the coal sector value chain – fall coal production to mitigate the effects of the beyond the reach of many policies. Herein lies climate crisis, it is essential to identify who a fundamental challenge. may be impacted and the potential magnitude of their loss. Governments need to understand Achieving an effective and just transition how many workers are directly employed by for all necessitates addressing the informal the coal mine, and whether their skills profile and formal segments of the affected enables them to move easily into alternative workforce through a combination of local jobs. More difficult is to understand how many and national policies and programs. The additional jobs and businesses will suffer concept of “just transition” under a broad income losses. Are alternative employment conceptualization should extend to national opportunities available locally or within easy priorities of inclusive, sustainable and broad- commuting distance, and how does the wage based economic growth. Coal transition – in alternative jobs compare to the lost job? similar to other sector adjustments driven Are retrenched coal-mine workers entitled to by technology or productivity gains that severance, health or early retirement benefits ultimately replaced obsolete production paid by the mine company? Will the existing structures and jobs – represents an existential safety net support all affected workers, or will threat to some segments of the economy some be left out? Are existing employment that, although small, have potentially wide- services such as job search assistance and reaching impacts. Understanding the potential training programs effective in matching welfare losses by workers is only part of the jobseekers to vacancies, and do they have the challenge, albeit a big challenge regarding Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 136 informal workers. Weighing the trade-offs and workers or artisanal/own-account miners can risks of prioritizing some stakeholders over also suffer severe income shocks. Temporary others is the fundamental task of strategic income support through the national safety net policy design. Risks include poor economic should be the minimum policy response, with outcomes in terms of deeper crisis and slower a view to accommodating a long adjustment recovery, costly transition support programs period where shocks are systemic in nature, that yield inadequate economic stimulus such as in settings where the local economy and job creation, worker transitions into is dependent on the closing mine. Income unsustainable jobs or activities with negative support not only smooths consumption; it environmental or other externalities, and also helps to sustain demand for local goods potential derailment of transition due to and services and the businesses that provide vested interests or powerful interest groups, them. Whereas income support can address resulting in minimal abatement of CO2 immediate and short-term needs, longer-term emissions, for example. Given the complex interventions are required to help workers systems of implicit- and cross-subsidy of move into alternative employment and to energy generation and its links to industrial create an environment conducive to business sector production and jobs, it is important diversification and private job creation. to understand who currently benefits from these existing systems, and the economic and There are five main channels through which fiscal costs and benefits associated with these public policies and programs can facilitate systems. A just transition is one in which the workers’ transition. Some policies target costs and benefits are shared more equitably. workers, some target firms: When implicit environmental costs of coal- linked activity are added to the equation, (i) temporary income support (e.g., employer the cost-benefit analysis is likely to favor a severance pay, national social safety net) realignment of public resources and policies toward more socially inclusive and sustainable (ii) increasing workers’ capacity to qualify structures of economic production. for jobs in new sectors (e.g., through skills or entrepreneurship training) Traditional labor policy instruments that support the transition of displaced workers (iii) connecting workers to potential to new jobs are necessary but not sufficient employers (e.g., through job search in this context. Past experience illustrates assistance, mobility grants) what can go wrong when, for example, there is insufficient labor demand in remote or lagging (iv) stimulating private sector labor regions, or when transition packages distort demand and local or regional business incentives to work, or when training is not development (e.g., through investment aligned with private sector needs, or when only incentives aligned with strategic some workers receive support while others national, local and/or regional priorities, – perhaps even the majority – struggle to matching grant programs); and make ends meet. Ensuring that informal mine employees can access the active and passive (v) ensuring the business environment and labor market programs offered to formal mine labor regulations are conducive to private employees is an important step, but non-mine sector investment and job creation. 137 Figure 5.1 Five policy channels to support transition Incom Support P ssiv LMPs Compl m nt r Incr sin Work rs’ C p cit Sust in bilit R ul tions n Jobs ALMPs Busin ss & L bor — Ensur includin s f t n ts, busin ss clim t Consist nc Gr Conn ctin Work rs to Jobs ALMPs Stimul tin Priv t Job Cr tion Supportiv infr ., duc tion curriculum, public-priv t -CSO p rtn rships Coordin t cross multipl l v ls of ov rnm nt, priv t s ctor, CSOs, communiti s A sustainability lens could be added to these The first phase – long before the closure policy channels to ensure that workers itself – should focus on broader economic displaced from coal sector jobs do not development planning to lay the groundwork simply transition to alternative but equally for absorbing the negative economic shock. unsustainable sectors. Jobs in environmentally This entails measures to enhance the capacity sustainable activities are likely to be more and resilience of the local economy through resilient to shocks and generate other positive diversification toward new economic sectors externalities, for example related to worker and new occupations requiring different skills. and community health. Moreover, introducing Given the wide-reaching and complex nature sustainability criteria into these policy of economic development and the many actors channels – such as building workers’ capacity involved, developing and coordinating the to qualify for green sector jobs, supporting various elements of an effective strategy is green entrepreneurship, or incentivizing green extremely difficult, and requires a combination investments that promote private sector job of leadership and partnership across local, creation – would support the parallel objective regional and national governments, private of stimulating green economic transition. sector, CSOs and communities. These five policy channels are relevant The policy framework presented here is across different phases of the transition, organized into four phases ranging from including before the transition begins. before the mine closure decision is taken Income support and active labor market through to the period following layoffs and policies typically come in the last phases of the closure. Even focusing only on the labor mine closure and worker transition process. aspects of coal transition – the objective of Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 138 this report – calls for a broad approach to toward emerging sectors at the national level address the needs of affected workers and (e.g., IT, green construction) and increasing local communities. The proposed framework, focus on STEM and soft skills will build local summarized in Figure 5.2, builds on the labor human capital and prepare future graduates policy approach developed by Cunningham for higher productivity occupations. The and Schmillen (2021) and Fretwell (2017) for policies and programs implemented in this addressing the transition of formal mine first phase are essential to providing the employees. Our framework incorporates two necessary impetus to local demand, especially additional dimensions. Firstly, it covers a in remote and/or lagging regions. Experience wider universe of affected workers, formal and from earlier transition episodes illustrates informal, whether direct mine employees on that local governments and institutions temporary informal contracts, for example, or cannot manage alone, requiring regional and workers in the coal value chain or businesses sometimes national support – with planning, that meet the consumption needs of coal mine policy coordination, and financing – as well employees and their families. And secondly, it as the cooperation and expertise of non- incorporates policies that go beyond labor to government organizations (e.g., charitable target private sector incentives and capacities organizations, cultural and academic to create jobs. The framework is informed by institutions, community entities and partners the strategic lessons presented in World Bank in the private sector). (2018a) as well as the lessons from past coal transitions described above. The framework’s There are significant risks of inadequate motivating objectives are to enhance the preparation or insufficient investment in welfare of affected workers and ensure that upstream mitigation efforts, and these viable medium-term employment outcomes can ultimately prove costly. Policymakers emerge, whether in the local economy or beyond. and local stakeholders need to recognize the many and varied costs and take the necessary Phase 1: Economic development strategy preventative steps. In addition to the costs before the mine closure decision is taken. imposed on workers and firms, prolonged In advance of the decision to close operations, economic recession leads to high fiscal outlays policymakers should be considering measures for social assistance, low fiscal revenues to bolster local economic prospects in case and reduced investment in schools, health of mine closure, which may not be imminent centers and infrastructure maintenance; but is ultimately expected. Taking actions in the result is weakened local institutions, advance to diversify the types of businesses diminished government effectiveness, and and jobs available would help cushion the loss of community confidence and cooperation. negative shock of mine closure. Efforts to Delayed transition can give rise to even higher improve the business environment and foster costs and more political pressure. Authorities entrepreneurship can increase profitability need to consider the long time horizon of and attract new firms. Investment in physical transition; even without inertia and delay, infrastructure and physical and digital it takes a long time for communities and connectivity will enhance the appeal for economies to change. In addition to cross- investors by reducing transport and other cutting measures to enhance local economic operational costs. Revising and reorienting opportunity, governments need to increase curricula in schools and training centers their engagement with and oversight of the 139 Figure 5.2 Policy framework for managing labor transition Fost r div rsific tion throu h busin ss clim t r forms, ntr pr n urship initi tiv s Inv st in supportiv ph sic l nd di it l infr structur incl. nh nc d conn ctivit 1 Ali n duc tion nd tr inin curricul tow rd m r in conomic s ctors (n tion l not just loc l; STEM plus soft skills) Economic Est blish p rtn rships nd str t ic ov rsi ht bodi s comprisin n tion l/r ion l nd loc l D v lopm nt ov rnm nt offici ls, c d mic institutions, priv t s ctor ssoci tions, civic or ni tions Str t nd oth r communit roups (Pr -closur d cision) Str n th n ov rsi ht of nd p rtn rship with min op r tors (to minimi disruption post closur r : work r prot ctions nd l nd/infr structur r purposin ) P rtn r with unions nd/or mplo r or ni tions to d si n nd d liv r st ff r -skillin for post-co l conom Id ntif work rs lik l to b dir ctl nd indir ctl ff ct d b th min closur proc ss, profil xistin skills nd pot nti l mism tch with curr nt l bor d m nd 2 R vi w soci l prot ction nd ALMP pro r ms nd l bor r ul tions r l t d to l offs Ass ss li ibilit for un mplo m nt b n fits nd oth r soci l s rvic s (distin uishin b An l sis work r ch r ct ristics such s , hous hold incom ), id ntif cov r ps, stim t pot nti l n ds for ssist nc & Pl nnin R vis incom support pro r ms, ALMPs nd productiv inclusion int rv ntions to (Pr -closur ) ccommod t s st mic shock nd build c p cit of s rvic provid r nci s B in communic tions/communit consult tions, ccomp ni d b positiv cultur l si n lin round n w soci l contr ct not c nt r d on co l Issu dv nc notific tion of l off Inform min work rs nd communit m mb rs of ssist nc options, off r p ck st r t d 3 to min mplo s nd to oth r work rs ( . ., productiv inclusion, public works) to ncour s lf-s l ction into th b st fit to f cilit t r lloc tion into n w jobs. R quir s ov rnm nt to sc l up its outr ch Announc H lp work rs cl rif th ir b n fit ntitl m nt with min op r tor, union L offs Est blish n twork of work r dvoc t s to promot ssist nc nd st r work rs to & Assist nc ppropri t pro r ms B in provision of c rt in s rvic s ( . ., c r r couns lin , ps cho-soci l outr ch, job s rch ssist nc ) Provid t mpor r incom support 4 Impl m nt ctiv l bor m rk t polici s in ph s s/b s d on scr nin ; ALMPs c n includ job s rch/t chnic l/softskills/ ntr pr n urship tr inin , job s rch r nts, w subsid , Post-l off busin ss incub tor, mobilit r nts Assist nc Monitor ssist nc t k -up nd job pl c m nt; djust pro r m p r m t rs to improv ff ctiv n ss Consid r uxili r s rvic s in r spons to communit n ds Source: Authors’ extension of the (formal) labor policy approaches developed in Fretwell (2017), World Bank (2018a) and Cunningham and Schmillen (2021) Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 140 coal mine operator to ensure it meets its occupations requiring similar skills will be key obligations to workers (in terms of severance inputs for designing public policy responses. or health benefits, for example) and to the local Given the extent and depth of impact on community in terms of pollution remediation workers, and the different types of affected to facilitate future land repurposing. workers – ranging from informal unskilled coal collectors to semi-skilled workers in the Unions and employer organizations may coal supply chain to micro-entrepreneurs try to impede transition, but could be and small business owners providing services effective partners if engaged early. If unions to coal mine workers and their families to or employer representatives believe that formal mine employees themselves – a wide maintaining the status quo is the best way to range of programs will be needed. A first protect their members, then when the coal step is to assess whether the existing safety operation ultimately closes, displaced mine net and employment services can meet workers are in reaction mode, scrambling identified needs, and if not, governments need to find alternative jobs. If, instead, they are to introduce program changes or additions invited into the planning process to design and to address the gaps. This diagnostic should deliver re-skilling for the post-coal economy inform planning around the timing and speed (including in dynamic sectors like renewable of closure, and the public resources likely to energy), then their members will be prepared be needed to finance the various passive and to move –even before mine closure – and have active labor market policies being offered. access to better jobs or business opportunities at the outset rather than waiting until after Proactive reform of passive and active labor new dynamic firms have already attracted market programs and systems should be younger or more skilled workers, and miners implemented well in advance of layoffs. face greater competition. Safety net programs such as social assistance or unemployment benefits may provide basic Phase 2: Pre-closure analysis and planning income support for a certain time period, starts with diagnostics and program review. but the level or coverage period or eligibility When the likelihood of future mine closure rules may be inadequate to meet the scale becomes clear, national and local authorities of layoffs, especially when these types of need to begin more specific planning to line up support are designed to protect against the right programs and make any necessary temporary idiosyncratic shocks rather than a policy changes. Pre-transition diagnostics systemic and persistent labor demand shock. of the local labor market will be useful for High coal wages or partial income losses identifying the number of workers likely to be by coal supply chain workers may restrict affected – direct coal mine employees, mine access when income qualification thresholds sub-contractors, local or regional workers in are low. Union members may be entitled to the coal supply chain, and indirectly affected receive some support from the mine operator, workers – as well as their skills profiles and while non-union members are not. ALMPs occupations, which can then be compared to designed to connect job seekers to vacancies, existing employment opportunities in the such as through job search assistance, job local and regional labor markets. The size and search training, technical training or wage nature of any identified mismatch or implied subsidies, may not have the systems in place wage differences between coal and non-coal to handle the likely high demand for services 141 following mine closure. On the basis of coming, presenting the range of support identified coverage needs and gaps, program options available, and providing preparatory parameters may need to be revised to address support. During this phase, mine workers the anticipated demand, or new support are given official advanced layoff notice, and programs may need to be introduced, perhaps the targeted packages of various support to address spatial-related issues such as labor options are offered, potentially in sequenced mobility constraints. Informal and especially rounds, in ways that encourage self-selection low-skilled workers in the coal value chain by workers into the best option for them to may need offerings such as public works transition to their preferred job, whether and productive inclusion programs. Service in a similar occupation or new occupation, delivery agencies may require upgraded local or elsewhere, or in their own start- systems or other capacity building efforts. up enterprise. Once workers receive layoff Governments need to provision budget notice, they need to clarify what separation resources to ensure the existing (or improved) benefits, health and pension benefits they safety net and ALMPs can meet the demand. are entitled to from the mine operator and/or the union. It would be helpful for workers to Armed with this information on skills be able to seek guidance on how to navigate mismatch, coverage gaps and new/revised the various program options; this could be support programs, governments need to facilitated by establishing a network of worker build a communications strategy and begin advocates – e.g., under a partnership between communications outreach to workers local government and community/non- and broader communities. Consultations governmental organizations – to help steer with community groups will be an essential workers to appropriate programs. Finally, pre- component. In settings where coal culture is layoff assistance such as counseling services embedded in social values, government efforts (career or psycho-social counseling), job could include introducing revised cultural search training and job search advice should signaling of “new economy” or “green” be rolled out to mine workers, with a view to alternatives and stimulating debate around extending support to non-mine workers who a more inclusive social contract not centered may also be affected. on coal. In addition to these steps to diagnose the nature and scope of the jobs challenges Phase 4: Post-layoff assistance comprises and design and roll out the tools to address delivery of temporary income support to them, there are other measures not related to displaced workers and implementation labor that are also important for effectively of active labor market policies. ALMPS managing mine closures; World Bank (2018a) may include various types of training (e.g., lays out guidelines for complementary job search training, technical, softskills or planning related to, for example, stakeholder entrepreneurship training), job search grants, mapping, regulatory requirements for firm targeted wage subsidy programs, business exit and property title transfer, and pollution incubator support, and mobility grants to remediation, inter alia. connect workers to jobs in other regions. The set of ALMPs can be organized as a menu Phase 3: Announcement of layoffs and of options, or offered in a phased approach assistance involves informing mine workers or on the basis of applicant screening to and the broader community that layoffs are ensure good fit. Program offerings – whether Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 142 passive or active – need to be monitored stakeholders, arbitrate competing interests, with respect to take-up and job placement and provide leadership and motivation for rates. If programs are not working effectively, transition. Governments’ most fundamental their parameters should be adjusted or responsibility is to mobilize adequate financing redesigned to deliver better outcomes. Regular that represents an investment in transition, monitoring and evaluation as well as broader rather than simply reacting to a systemic assessment of community well-being will labor demand shock by applying band-aids. enable stakeholders to identify and respond The speed of transition will determine its to emerging crises sooner rather than later, ultimate cost. Each of the four stages of this before they become intractable. policy framework is integral to and designed to facilitate an effective labor transition. Government’s role is multi-faceted and complicated, but more effective if proactive. The accumulating forces for change – A well-planned and systematic process of including increasing recognition of the coal mine closure and layoffs is essential climate crisis and the urgent need to reduce for supporting the reallocation of affected carbon emissions – are creating momentum workers to alternative work, and – equally for transition. This momentum takes importantly – mitigating the economic and different forms, but is more and more evident social and political costs of transition. This within transitioning countries as well as wide scope of impact requires coordination among some of the biggest coal players, such across sectors and across various levels as China and India. Country context related of government. Moreover, the long time to the nature and extent of coal reliance helps period of policy design and implementation, explain country motivation to accelerate or which spans the four phases described here, delay or avoid transition. The patterns of coal requires particular attention to intertemporal production, consumption and employment policy coherence. Governments alone do documented in this report offer insights for not have to deliver everything; they can future decisions to reduce coal production and provide strategic direction, coordinate across coal-dependent employment. 143 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 144 References Alves Dias P.,K. Kanellopoulos, H. Medarac, Z. Kapetaki, Betz, M, M. Partridge, M. Farren and L. Lobao. 2015.“Coal E. Miranda-Barbosa, R. Shortall, V. Czako, T. Telsnig, Mining, Economic Development, and the Natural C. Vazquez-Hernandez, R. Lacal Arántegui and Resources Curse.” Energy Economics 50:105-116. W. Nijs. 2018. “EU coal regions: opportunities and challenges ahead”, European Commission, Joint Bhattacharyya, S. and B. Resosudarmo. 2020. “Growth, Research Centre (JRC) EUR 29292 EN, JRC112593. Growth Accelerations, and the Poor: Lessons from Indonesia”, World Development Vol. 66, February Appalachian Citizens’ Law Center, Appalachian Voices, 2015, pp. 154-165. Coalfield Development Corporation, Rural Action, and Downstream Strategies. 2019. “A New Horizon: Bhushan, C., S. Banerjee and S. Agarwal. et al. 2020. Innovative Reclamation for a Just Transition.” Just Transition in India: An inquiry into the challenges Reclaiming Appalachian Coalition. and opportunities for a post-coal future. Sustainable Innovations and Advisories Pvt. Ltd., New Delhi. Badiani-Magnusson R., J. Franco, F. Kochan and K. Sander, 2019, “Air Quality in Poland, what are the Billings, Dwight B. and Kathleen M. Blee. 2000. The Road issues and what can be done?”, The World Bank, Brief to Poverty: The Making of. Wealth and Hardship in Fall 2019. Appalachia. New York: Cambridge University Press. Baker, L., J. Burton, C. Godinho and H. Trollip. 2015. Black, D., T. McKinnish, and S. Sanders. 2005a. “The “The political economy of decarbonization: Exploring Economic Impact of the Coal Boom and Bust.” The the dynamics of South Africa’s electricity sector”, Economic Journal 115 (April): 449-476. Energy Research Center, University of Cape Town, Cape Town. Black, D., T. McKinnish, and S. Sanders. 2005b. “Tight Labor Markets and the Demand for Education: Baran, J., P. Lewandowski, A. Szpor and J. Witajewski- Evidence from the Coal Boom and Bust.” Industrial Baltvilks. 2018. “Coal transitions in Poland - Options and Labor Relations Review vol. 59, issue 1: 3-16. for a fair and feasible transition for the Polish coal sector”, IDDRI & Climate Strategies. Bogdan, W., D. Boniecki, E. Labaye, T. Marciniak and M. Nowacki. 2015. “Poland 2025: Europe’s new growth Baran J., A. Szpor and J. Witajewski-Baltvilks. 2020. engine”, McKinsey & Company. “Low-carbon transition in a coal-producing country: A labour market perspective”, Energy Policy, 147, BP Statistical Review of World Energy. 2020. https:// p.111878. www.bp.com/content/dam/bp/business-sites/en/ global/corporate/pdfs/energy-economics/statistical- Bertelsen, N. and B. Vad Mathiesen. 2020. “EU-28 review/bp-stats-review-2020-full-report.pdf residential heat supply and consumption: Historical development and status”, Energies, 13(8), p.1894. Brauers H. and P.Y. Oei. 2020. “The political economy of coal in Poland: Drivers and barriers for a shift away Besser, Terry L., Nicholas Recker, and Kerry Agnitsch. from fossil fuels”, Energy Policy, 144, p.111621. 2008. “The Impact of Economic Shocks on Quality of Life and Social Capital in Small Towns.” Rural Sociology 73 (4): 580–604. 145 Burton, J., T. Caetano and B. McCall. 2018a. “Coal Coady, D., I. Parry, N. Le, and B. Shang. 2019. “Global transitions in South Africa: Understanding the Fossil Fuel Subsidies Remain Large: An Update implications of a 2 C-compatible coal phase-out plan Based on Country-Level Estimates”, IMF Working for South Africa”, Cape Town: Energy Research Centre. Paper WP/19/89, Fiscal Affairs Department, IMF, Washington, DC. Burton, J., T. Lott and B. Rennkamp. 2018b. “Sustaining Carbon Lock in: Fossil Fuel Subsidies in South Africa” Coal India Ltd. Website, https://www.coalindia.in/ in J. Skovgaard and H. van Asselt (eds) The Politics of Fossil Fuel Subsidies and Their Reform, Cambridge: Cook, A. 1995. “Increasing Poverty in Timber- Cambridge University Press Dependent Areas in Western Washington.” Society & Natural Resources 8 (2): 97–109. Carley, Sanya, Tom P. Evans, and David M. Konisky. 2018. “Adaptation, Culture, and the Energy Transition Cunningham, W. and A. Schmillen. 2021. “The Coal in American Coal Country.” Energy Research & Social Transition: Mitigating Social and Labor Impacts”, Science 37 (March): 133–39. Social Protection and Jobs Discussion Paper, World Bank. Carpenter, S. 2020. “Polish firms suspend financing for new coal plant, in latest sign that King Coal is Czerwińska E. 2002. “Restructuring of hard coal mining slipping”, Forbes.com, February 15, 2020; https:// in Poland”, Bureau of Studies and Expert Opinions. www.forbes.com/sites/scottcarpenter/2020/02/15/ Chancellery of the Sejm. Information no. 891. polish-firms-suspend-financing-for-new- coal-plant-latest-sign-that-king-coal-is- Czyżak, P. and M. Hetmański. 2020. “2030: Analysis slipping/?sh=649f76d41937. of the border phase-out year in the energy sector in Europe and Poland”, Instrat Policy Paper 01/2020. Chamber of Mines. 2016. “Mine SA 2016: Facts and Figures Pocketbook”, Johannesburg: Chamber of Czyżak, P. and A. Wrona. 2021. “Achieving the goal. Mines of South Africa. Departure from coal in the Polish power sector”, Instrat Policy Paper 01/2021. Chikkatur, A.P., A.D. Sagar and T.L. Sankar. 2009. “Sustainable development of the Indian coal sector”. Daniels, S. E., C. L. Gobeli, and A. J. Findley. 2000. Energy, 34(8), pp.942-953. “Reemployment Programs for Dislocated Timber Workers: Lessons from Oregon.” Society & Natural CIF. 2021. “Supporting Just Transitions in India”, Just Resources 13 (2): 135–50. Transition Case Study Climate Investment Funds, March 2021. Deaton, J.B., Niman, E., 2012. An empirical examination of the relationship between mining employment and Climate Investment Funds (CIF). 2021. “Supporting poverty in the Appalachian region. Appl. Econ. 44, Just Transitions in India”, Just Transition Case Study 303–312 Climate Investment Funds, March 2021. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 146 Department of Forestry, Fisheries and the European Environmental Agency (EEA). 2020. “Air of Environment. 2020. “Cabinet approves climate quality in Europe – 2020 report”, EEA Report No change, emissions reduction and waste management 09/2020. plans”, media release by the Department Forestry, Fisheries and the Environment, Republic of South European Environment Agency (EEA). 2016. “Air of quality Africa, September 13, 2020. in Europe – 2016 report”, EEA Report No 28/2016. Dessus, S. and M. Hanusch. 2018. “South Africa economic Farren, M. and M. Patridge. 2015. “Reevaluation of the update: Jobs and inequality”, World Bank Group Impact of Coal Mining on the Virginia State Budget”, Douglas, Stratford and Anne Walker. 2017. "Coal Ohio State University Department of Agricultural, Mining and the Resource Curse in the Eastern United Environmental, and Development Economics, C. States.” Journal of Regional Science 57(4): 568-590. William Swank Program in Rural-Urban Policy. Duncan, Cynthia. 2014 Worlds Apart: Poverty and Politics Fleming D.A., and T.G. Measham. 2014. “Local job in Rural America. Yale University Press. multipliers of mining”, Resources Policy 41:9–15. Eberhard, A. 2011. “The Future of South Africa Coal: Fretwell, D. 2017. “Mitigating the Social Impact of Market, Investment and Policy Challenges”, Program Economic Change During Enterprise Restructuring on Energy and Sustainable Development, pp.1-44. and Privatization.” Unpublished background paper. World Bank: Washington, DC. Edwards. R. 2017. “Has Resource Extraction Reduced Poverty?”, Inside Indonesia Oct-Dec 2017, https://www. Freudenburg, W. 1992. “Addictive Economies: Extractive insideindonesia.org/has-resource-extraction-reduced- Industries and Vulnerable Localities in a Changing poverty-3 End Coal. 2021. “Global Coal Plant Tracker”, World.” Rural Sociology 57(3):305–32. https://endcoal.org/global-coal-plant-tracker/ Freudenburg, W. and L. Wilson. 2002. “Mining the Data: End Coal. 2021. Global Coal Plant Tracker, https:// Analyzing the Economic Implications of Mining for endcoal.org/global-coal-plant-tracker/ Nonmetropolitan Regions.” Sociological Inquiry 72 (4): 549–75. Energy Information Administration (EIA). 2017. “South Africa”, https://www.eia.gov/international/analysis/ Garg, A. and J.C. Steckel. 2017. “Bridging the Gap: country/ZAF Phasing out Coal-The Emissions Gap Report 2017 Chapter 5”, The Emissions Gap Report 2017: A UN Energy.instrat.pl. 2021. “Employment in coal sector in Environment Synthesis Report. Poland", https://energy.instrat.pl Glaeser, E., S. Pekkala Kerr and W. Kerr. 2015. Energy Policy Tracker. 2021. “Poland”, https://www. “Entrepreneurship and Urban Growth: An Empirical energypolicytracker.org/country/poland Assessment with Historical Mines”, Review of Economics and Statistics Vol. 7, no. 2. Euracoal. 2021. “Euracoal Statistics”, https://euracoal. eu/info/euracoal-eu-statistics/ Gomez-Mera, L. and C. Hollweg. 2018. “Firm performance and constraints in Indonesia”, mimeo, World Bank. Europe Beyond Coal. 2021. “Coal Plant Database”, https://beyond-coal.eu/database/ 147 Guan, D., C. Cui, D. Wang, V. Chemutai, P. Brenton, S. IEA. 2021b. “Poland”, https://www.iea.org/countries/ Zhang, Q. Zhang and S. Davis. 2021. “Global Mitigation poland Efforts Should Prioritize Support to Emerging Emitters”, Nature (forthcoming). IEA.2020. “Coal 2020”, https://www.iea.org/reports/ coal-2020 Haddad, M., S. Karimjee, D. Khalifa and P. Noumba Um. 2019. “Creating Markets in South Africa: Boosting IEA. 2019. “South Africa Energy Outlook”, analysis from Private Investment to Unlock South Africa’s Growth Africa Energy Outlook 2019, https://www.iea.org/ Potential”, World Bank Group. articles/south-africa-energy-outlook Haggerty, J. 2014. “Long-Term Effects of Income Inside Climate News. 2019. “Coal Giant Murray Energy Specialization in Oil and Gas Extraction: The U.S. Files for Bankruptcy Despite Trump’s Support”, West, 1980-2011.” Energy Economics 45: 186–95. October 29. 2019. https://insideclimatenews.org/ news/29102019/coal-bankruptcy-bob-murray- Haggerty, Julia H., Mark N. Haggerty, Kelli Roemer, and energy-chapter-11-trump-regulations-rollback/ Jackson Rose. 2018. “Planning for the Local Impacts of Coal Facility Closure: Emerging Strategies in the U.S. Instrat. 2021. Hard coal and lignite mining database, West.” Resources Policy 57 (August): 69–80. energy.instrat.pl. Haggerty, Mark. 2019. “Communities at Risk from Integrated Resource Plan (IRP). 2019. Department of Closing Coal Plants.” Working Paper. Bozeman, MT: Energy, Republic of South Africa, October 2019. http:// Headwaters Economics. www.energy.gov.za/IRP/2019/IRP-2019.pdf Harris, K. 1977. “The production of electrolytic manganese Intergovernmental Forum on Mining, Minerals, Metals in South Africa”, Journal of the Southern African and Sustainable Development (IGF). 2018. “South Institute of Mining and Metallurgy, 77(7), pp. 137-142. Africa Horizontal Linkages: Building Expertise by Overcoming Country Specific Constraints (Case Henderson, C. 2015. “Coal-fired power plant efficiency Study)”, IGF Guidance for Governments: Leveraging improvement in India”. London: IEA Clean Coal Centre. Local Content Decisions for Sustainable Development, Winnipeg: International Institute for Sustainable IBEF. 2021. “Renewable Energy Industry in India”, Development (IISD). India Brand Equity Foundation (IBEF) article posted September 3, 2021. https://www.ibef.org/industry/ IPCC. 2007. Climate Change 2007: Synthesis Report. renewable-energy.aspx Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental IBS. 2020. “Silesia at the Turning Point of Panel on Climate Change [Core Writing Team, Transformation”, Institute of Structural Research Pachauri, R.K. and Reisinger, A. (eds.)]. IPCC, Geneva, (Instytut Badan Strukturalnych (IBS)). Switzerland, 104 pp. IEA. 2021a. “India Energy Outlook 2021: World Energy Outlook Special Report”, IEA. https://www.iea.org/ reports/india-energy-outlook-2021 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 148 IPCC. 2014. Climate Change 2014: Synthesis Report. Kasprzak M. 2021. “Disappointing lack of ambition in Contribution of Working Groups I, II and III to the Poland's Energy Policy until 2040”, Ember. Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K. Kasprzak M. 2020. “Poland's Second Bełchatów”, Pachauri and L.A. Meyers (eds.)]. IPCC, Geneva, Ember, https://ember-climate.org/project/polands- Switzerland, 151 pp. second-belchatow/ Isserman, Andrew, Edward Feser, and Drake Warren. Kelsey, Timothy, Mark Partridge, Nancy White. 2016. 2009. “Why Some Rural Communities Prosper While “Unconventional Gas and Oil Development in the Others Do Not.” International Regional Science Review United States: Economic Experience and Policy 32 (3): 300-342. (An earlier more detail version Issues.” 2016. Applied Economic Perspectives and Policy was published in the 2007 Report to the Office of 38 (2): 191-214. Rural Development, USDA. http://www.jj0955. com/PdfFiles/IssermanFeserWarren_070523_ Kok, Irem. 2017. “Coal Transition in the United States.” RuralProsperity.pdf) Paris: IDDRI and Climate Strategies, 2017. Johnson, Kenneth, M. Johnson, John P. Pelissero, Lahiri-Dutt, K. 2016. “The diverse worlds of coal in David B. Holian and Michael T. Maly. 1995. “Local India: Energising the nation, energising livelihoods”, Government Fiscal Burden in Non-Metropolitan Energy Policy vol. 99 (December), pp. 203-2013. America.” Rural Sociology 60 (3):381-398. Lahiri-Dutt, K. and D. Williams. 2005. “The coal cycle: A Jones D. 2018. “Last Gasp: The coal companies making small part of the illegal coal mining in eastern India”, Europe sick”, Ember (formerly known as Sandbag), Journal of Resources, Energy and Development vol. 2, Beyond Coal, Greenpeace Central and Eastern Europe, no. 2, pp. 93-105. and the European Environmental Bureau. Lewandowski P., M. Antosiewicz, J. Frankowski, Kamboj, P. and R. Tongia. 2018. “Indian railways and coal: J. Mazurkiewicz and J. Sokolowski. 2020. “The An unsustainable interdependency”. Brookings India influence of the 2050 carbon neutrality scenarios on the labour market in Upper Silesia”, Instytut Badan Kapetaki, Z., P. Alves Dias, A. Conte, K. Kanellopoulos, Strukturalnych (IBS). G. Mandras, H. Medarac, W. Nijs, P. Ruiz, J. Somers, D. Tarvydas. 2021. “Recent trends in EU coal, peat and oil Lobao, Linda, Mark Partridge, Minyu Zhou, and Michael shale regions”, European Commission, Joint Research Betz. 2016. “Poverty, Place and Coal Employment in a Center (JRC) Science for Policy Report EUR 30618 EN, New Economic Era.” Rural Sociology 81 (3):343-386. JRC123508. Lobao, Linda and Paige Kelly. 2020. “Local Governments Karbownik, A. 2005. “Zarządzanie procesem across the Rural-Urban Continuum: Findings from dostosowawczym w górnictwie węgla kamiennego a Recent National County Government Study.” State w świetle dotychczasowych doświadczeń” and Local Government Review (forthcoming). [Management of the adjustment process in hard coal mining in the light experiences so far: Collective work]. Wydawnictwo Politechniki Śląskiej, Gliwice. 149 Lobao, L., M. Partridge, O. Hean, P. Kelly, S. Chung and Ministry of Statistics and Programme Implementation E. Ruppert Bulmer. 2021. “Socioeconomic Transition (MOSPI). 2021. “Energy Statistics 2021”, http:// in the Appalachia Coal Region: Some Factors of www.mospi.nic.in/sites/default/files/reports_and_ Success”. Produced for the World Bank, under the publication/ES/Chapter-1%20Reserves%20and%20 Global Support to Coal Regions in Transition project. potential%20for%20generation.pdf Mahadevan, M. 2019. “The price of power: Costs of Montrone, L., N. Ohlendorf and R. Chandra. 2021. political corruption in Indian electricity”. Department “The political economy of coal in India–Evidence of Economics, University of Michigan. from expert interviews”. Energy for Sustainable Development, 61, pp.230-240. Majozi, T. & Veldhuizen, P. 2015. The chemicals industry in South Africa. American Institute of Chemical Moritz, T., T. Ejdemo, P. Söderholm and L. Wårell. Engineers (AIChE), pp. 46-51. 2017. “The local employment impacts of mining: an econometric analysis of job multipliers in northern Mańkowska, M., M. Pluciński and I. Kotowska. 2021. Sweden”, Miner Econ 30:53-65. “Biomass Sea-Based Supply Chains and the Secondary Ports in the Era of Decarbonization”, Mortkowitz L. and M. Martewicz. 2016. “As Europe Energies, 14(7), p.1796. Drops Coal, Poland Embraces It”, Bloomberg (June 2016), https://www.bloomberg.com/news/ Marais, L., F.M. McKenzie, L. Deacon, E. Nel, D. van articles/2016-06-24/as-europe-drops-coal-poland- Rooyen and C. Jan. 2018. “The changing nature of embraces-it mining towns: Reflections from Australia, Canada and South Africa”, Land Use Policy, Volume 76, pp. MOSPI (India’s Ministry of Statistics and Programme 779-788. Implementation). 2021. http://www.mospi.nic.in/ Min, B. and M. Golden. 2014. “Electoral cycles in Nicholas, S. and T. Buckley. 2019. “South African Coal electricity losses in India”. Energy Policy, 65, Exports Outlook: Approaching Long-Term Decline”, pp.619-625. Institute for Energy Economics and Financial Analysis. Minerals Council. 2018. “Facts and Figures 2017”, Organisation for Economic Co-operation and http://www.mineralscouncil.org.za/downloads/ Development (OECD). 2015. “South Africa Policy send/18-current/634-facts-and-figures-2017 Brief”, OECD. Minerals Council. 2019. “Facts and Figures 2018”, Padel, F. and S. Das. 2010. Out of this Earth: East India https://www.mineralscouncil.org.za/industry-news/ Adivasis and the Aluminium Cartel. Orient Blackswan, publications/facts-and-figures Hyderabad. Ministry of Development. 2018. “Ogromne koszty Pai, S. and H. Zerriffi. 2021. “A Novel Dataset zdrowotne i finansowe smogu z niskiej emisji – for Analysing Sub-National Socioeconomic MPiT przedstawiło raport”, https://www.gov.pl/ Developments in the Indian Coal Industry”, web/rozwoj-praca-technologia/ogromne-koszty- IOPSciNotes 2 014001. zdrowotne-i-finansowe-smogu-z-niskiej-emisji- mpit-przedstawilo-raport Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 150 Partridge, Mark D. and M. Rose Olfert. 2011. “The Sharma, M., J. Sahni and S. Kumar. 2019. IAS Mains Winners' Choice: Sustainable Economic Strategies for Chapterwise Solved Papers General Studies 2020. 3 ed. Successful 21st Century Regions.” Applied Economic New Delhi: Arihant Publications (India) Limited. Perspectives and Policy. (33): 143-178. Snyder, Brian F. 2018. “Vulnerability to Decarbonization PEG-CPR Roundtable. 2021. “Summary of Discussion: in Hydrocarbon-Intensive Counties in the United 2nd PEG-PCR Roundtable on Managing a Fair States: A Just Transition to Avoid Post-Industrial Transition away from Coal in India”, Prayas (Energy Decay.” Energy Research & Social Science 42 Group) and Centre for Policy Research. (August): 34–43. Peszko, G., D. van der Mensbrugghe, A. Golub, J. Ward, South Africa Chambers of Mines. 2017. “Integrated D. Zenghelis, C. Marijs, A. Schopp, J. A. Rogers and A. Annual Review 2017: Making Mining Matter”, Midgley. 2020. Diversification and Cooperation in a Chambers of Mines of South Africa. Decarbonizing World : Climate Strategies for Fossil Fuel-Dependent Countries. Climate Change and Spencer, T., R. Pachouri, G. Renjith and S. Vohra. 2018. Development. Washington, DC: World Bank. “Coal transition in India”, TERI Discussion Paper https://openknowledge.worldbank.org/ (New Delhi: The Energy and Resources Institute). handle/10986/34011 License: CC BY 3.0 IGO. Statistical Office in Katowice. 2020. “Labour market Polish Geological Institute, 2021, “Mineral Resources in Slaskie Voivodship in the years 2018-2019”, of Poland”, National Research Institute, http:// Statistical Office in Katowice, katowice.stat.gov.pl. geoportal.pgi.gov.pl/surowce/energetyczne/ wegiel_kamienny Statistics South Africa (SA STAT). 2015. “Mining: Production and sales”, SA Stats, 1 April, pp. 1-13. Reserve Bank of India, 2019. Reserve Bank of India. https://m.rbi.org.in/scripts/PublicationsView. Statistics South Africa (SA STAT). 2019. “Stats SA”, aspx?id=18810[Accessed April 21, 2021] http://www.statssa.gov.za/publications/P0211/ Presentation%20QLFS%20Q4_2019.pdf Richmond C., D. Benedek, E. Cabezon, B. Cegar, P. Dohlman, M. Hassine, B. Jajko, P. Kopyrski, M. Stoczkiewicz M. and A. Śniegocki (ed.). 2020. “Subsidies: Markevych, J.A. Miniane and F.J. Parodi. 2019. A driving force or obstruction for the Polish energy “Reassessing the Role of State-Owned Enterprises transition? Analysis of State aid for the power sector in Central, Eastern, and Southeastern Europe”. in Poland”, ClientEarth. International Monetary Fund, European Department, No. 19/11. Strambo, C., J. Burton and A. Atteridge. 2019. “The End of Coal? Planning a "Just Transition" in South Africa”, Ritchie H. and M. Roser. 2021. “Poland: CO2 Country Sweden: Stockholm Environment Institute. Profile”, Our World in Data. Szpor A. and K. Ziółkowska. 2018. “Transformation of Rogala. B. 2021. “Old power industry in Poland: Most of the the Polish coal sector”, International Institute for power plants are over 50 years old”, 300GOSPODARKA, Sustainable Development (IISD). https://300gospodarka.pl/news/energetyka- konwencjonalna-wiek-elektrowni-w-polsce. The Economist. 2021. “Midnight Sky”, The Economist Print Edition (January 30th, 2021). 151 Tongia, R. and S. Gross. 2019. Coal in India, Brooking India. Winkler, H., S. Keen and A. Marquard. 2020. “Climate finance to transform energy infrastructure as part of a Trade & Industrial Policy Strategies (TIPS). 2020. just transition in South Africa”, Cape Town: Research “Sector Jobs Resilience Plan: Coal Value Chain”, report for SNAPFI project. Pretoria: TIPS World Bank. 2017. “South Africa Economic Update: Turek, M. and A. Karbownik. 2005. “Evaluation of Private Investment for Jobs”, World Bank Group. Mining Social Package in restructuring employment in mining”, Zeszyty Naukowe Politechniki Śląskiej, World Bank. 2018a. “Managing Coal Mine Closure: no. 1681, 7-14. Achieving a Just Transition for All”, World Bank Group. Turok, I. 2012. “Urbanisation and development in South World Bank. 2018b. “Overcoming Poverty and Inequality Africa: Economic imperatives, spatial distortions and in South Africa: An Assessment of Drivers, Constraints strategic responses”, London: Human Settlements and Opportunities”, World Bank Group. Group. International Institute for Environment and Development. World Bank. 2020a. “Pathways to Middle-Class Jobs in Indonesia”, World Bank. United Nations Environment Programme (UNEP). 2019. “Emission Gap Report 2018”, UNEP. World Bank. 2020b. “The Demand Side of Jobs in Indonesia: Plant-Level Analysis in the Medium and United Nations Framework Convention on Climate Large Manufacturing Industry”, World Bank. Change (UNFCCC). 2021. “Nationally Determined Contributions under the Paris Agreement”, February World Bank. 2020c. “Poverty and Equity Brief for South 26, 2021, https://unfccc.int/sites/default/files/ Africa”, World Bank, https://databank.worldbank. resource/cma2021_02_adv_0.pdf org/data/download/poverty/33EF03BB-9722-4AE2- ABC7-AA2972D68AFE/Global_POVEQ_ZAF.pdf. Urban Emissions, Conservation Action Trust and Greenpeace India. 2013. “Coal Kills: An Assessment World Bank. 2021a. “World Bank and the European of Death and Disease cause by India’s Dirtiest Energy Commission to Support Poland to Transition Out of Source”. Coal” The World Bank, Press Release 2021/ECA/62 (January 27th, 2021). Van der Ploeg, F. 2011. Natural Resources: Curse or Blessing? Journal of Economic Literature 49 (2): World Bank. 2021b. “Tracking SGD 7 - The Energy 366–420. Process Report - Country Profile: Indonesia”, World Bank. https://trackingsdg7.esmap.org/country/ Van Renssen S. 2021. “Poland sows the seeds of a post- indonesia Covid future beyond coal”, Energy Monitor. World Bank. 2021c. “Support to the Indian Energy Winkler, H. 2021. “When coal leaves town: Can local Transition: Coal Mine Closure, Achieving a Just governments help?”, World Bank (mimeo). Transition for All” Concept Note, Energy and Extractives Unit, World Bank. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 152 World Economic Forum (WEF). 2020. “The Children’s Continent: Keeping up with Africa's Growth”, World Economic Forum, 13 January. WHO. 2016. “Health topics: Air pollution”. https://www. who.int/health-topics/air-pollution#tab=tab_1 World Resources Institute (WRI). 2020. “South Africa: Strong Foundations for a Just Transition”, World Resources Institute, https://www.wri.org/just- transitions/south-africa (downloaded August 31, 2021). WWF European Policy Office, Sandbag, Climate Action Network Europe, and Health and Environment Alliance (HEAL). 2016. “Europe’s Dark Cloud: How Coal-Burning Countries Are Making Their Neighbors Sick”, Brussels, June 2016. XMP Consulting. 2013. “Review of the South African Coal Mining Industry”, https://cer.org.za/wp-content/ uploads/2017/12/Annexure-P.pdf Yudelman, D. 1984. “The emergence of modern South Africa: State, capital, and the incorporation of organised labour on the South African gold fields”, Canadian Journal of History, 18(3), pp. 435-437. 153 ANNEX 1 Using Input-Output tables to estimate the coal content of sectors Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 154 Input-output tables provide a comprehensive into account the consumption of coal picture of how sectors intertwine in an along the sectors upstream value chain.92 economy through backward and forward This measure can be used to identify value chain linkages. For a given number of sectors that are not themselves directly sectors, they record how the production of a dependent on coal but depend on inputs sector’s output depends on another sector’s that are coal dependent. input. If sufficiently disaggregated at the sector level, they can be a useful tool to analyze how (iii) A coal dependency measure that different sectors across an economy rely on the considers direct and indirect effects as availability of a particular input, such as coal. in (ii) but that excludes indirect input- output links to the coal sector when A global, standardized repository of input- these emanate from coal being an input output tables for 121 countries, each covering in electricity generation. This measure 57 sectors (among them coal), is available from is calculated as in (ii) but without the the Global Trade Analysis Project (GTAP). We electricity sector. The measure seeks to analyzed each to generate three measures of abstract from the fact that indirect coal coal dependence for every sector in a given content in sectors will be particularly country. high in countries that rely on coal for electricity generation. It therefore allows (i) A coal dependency measure that one to gauge sector’s indirect reliance on considers only direct input-output links coal other than through its importance as with the coal sector: This measure is a source of energy. calculated as a sector’s share of coal in total intermediate consumption. The measure provides insight into which sectors are direct consumers of coal in the production process. In most countries these sectors will include the electricity sector and the metals industry. (ii) A coal dependency measure that considers direct input-output links with the coal sector as well as indirect input- output links with the coal sector due input-output relationships along value chains: This measure is calculated by re-weighting the sector’s share of coal in total intermediate consumption to take 92  In practice, this is achieved by calculating the Leontief inverse of the input-output matrix in every country and re- weighing coal input requirement accordingly. 155 ANNEX 2 Technical Results for Indonesia Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 156 Methodology for estimating coal employment spill-overs to local non-coal sectors in Indonesia We rely on the methodology developed by Black et al. (2005) to estimate the impact of coal sector jobs on non-coal employment in relatively coal-intensive regions, and we test whether the degree of coal-intensity affects the magnitude of this elasticity. For Indonesia, where data is representative at the district level, we estimate the following equation: Ln(Ys,t,i )=β 0 + β1 Ln (X t,i ) + β 2 Z i + β 3 Pt-1,i + α v + π t + Ut,i where Ys,t,i is the number of jobs in non-coal sector s at time t in district i ; X t,i is the number of jobs in the coal sector at time t in district i ; and Z i is a dummy variable for whether the district is in the treatment group of coal-intensive districts, which we define as having a coal share of employment equal to at least 4 percent in 2007. We include lagged population size “Pt-1,i ” to account for agglomeration effects, and include province “α v” and year “π t ” fixed effects to control for potential structural variations across the different time periods and provinces. Ut,i is the error term. Therefore, β1 is the elasticity of the number of jobs in the non-coal sector to the number of jobs in the coal sector, and β 2 captures the difference in employment in the non-coal sector between the treatment districts and non-treatment districts. We run this for non-coal sectors aggregated together, and subsequently for each non-coal sector separately. We follow Black et al. (2005) and Moritz et al. (2017) by restricting our sample to districts within coal provinces because they share similar institutional and geographic characteristics, which reduces confounding factors. The sample includes only coal provinces South Kalimantan and East and North Kalimantan, and within these, excludes districts that have zero or negligible coal employment (i.e., less than 0.5 percent). For robustness, we test on a wider sample of coal provinces (namely by adding South Sumatera and Banten) and find similar results. 157 Annex 2 Table 1 Net job creation in Indonesia Employment level Net job creation 2007 to 2012 Net job creation 2012 to 2018 o/w % Annual job o/w % Annual job ALL PROVINCES 2007 2012 2018 # of jobs # of jobs male growth rate male growth rate Agriculture 38,540,867 36,844,412 32,885,651 (1,696,455) 63% -1% (3,958,761) 56% -2% Coal 90,075 260,162 240,041 170,087 96% 24% (20,121) 105% -1% Other Mining & Quarrying 894,193 1,324,460 1,187,376 430,267 97% 8% (137,084) 83% -2% Manufacturing & Utilities 12,321,481 15,571,799 18,679,363 3,250,318 63% 5% 3,107,564 51% 3% Construction 5,220,599 6,802,997 8,173,464 1,582,398 98% 5% 1,370,467 99% 3% Wholesale & Retail & Restaurants 19,970,919 22,879,060 30,282,118 2,908,141 44% 3% 7,403,058 45% 5% Transportation and 5,901,586 4,992,125 6,258,773 (909,461) 91% -3% 1,266,648 77% 4% Communications Finance & Business Services 1,382,738 2,660,271 3,834,138 1,277,533 71% 14% 1,173,867 71% 6% Community and Personal Services* 11,821,820 17,069,477 18,361,441 5,247,657 43% 8% 1,291,964 -5% 1% Total 96,144,278 108,404,763 119,902,365 12,260,485 55% 2% 11,497,602 49% 2% Employment level Net job creation 2007 to 2012 Net job creation 2012 to 2018 SOUTH KALIMANTAN AND EAST AND NORTH o/w % Annual job o/w % Annual job 2007 2012 2018 # of jobs # of jobs KALIMANTAN male growth rate male growth rate Agriculture 1,062,615 1,174,134 1,062,381 111,519 70% 2% (111,753) 39% -2% Coal 69,678 179,196 166,575 109,518 96% 21% (12,621) 85% -1% Other Mining & Quarrying 48,931 67,325 66,922 18,394 98% 7% (403) -447% 0% Manufacturing & Utilities 216,642 234,292 349,240 17,650 169% 2% 114,948 54% 7% Construction 126,121 201,080 192,749 74,959 99% 10% (8,331) 114% -1% Wholesale & Retail & Restaurants 557,201 722,767 1,026,035 165,566 50% 5% 303,268 44% 6% Transportation and 151,662 139,903 209,905 (11,759) 73% -2% 70,002 76% 7% Communications Finance & Business Services 40,149 94,869 128,650 54,720 87% 19% 33,781 53% 5% Community and Personal Services* 369,703 555,510 701,220 185,807 45% 8% 145,710 39% 4% Total 2,642,702 3,369,076 3,903,677 726,374 70% 5% 534,601 49% 2% Note: * Includes public administration. Source: Sakernas data 2007-2018. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 158 Annex 2 Table 2 OLS regressions on the correlates of real hourly wages (2018; all provinces)   VARIABLES (1) (2) (3) (4) Age 0.0372*** 0.0384*** 0.0377*** 0.0395*** (0.0012) (0.0009) (0.0009) (0.0009) Age squared -0.000352*** -0.000367*** -0.000368*** -0.000372*** (1.55e-05) (1.13e-05) (1.13e-05) (1.13e-05) Complete Primary 0.136*** 0.113*** 0.106*** 0.111*** (0.0076) (0.0073) (0.0072) (0.0073) Incomplete or Complete Secondary 0.400*** 0.382*** 0.339*** 0.369*** (0.0075) (0.0073) (0.0073) (0.0074) Post-Secondary 0.956*** 0.933*** 0.755*** 0.906*** (0.0103) (0.0099) (0.0111) (0.0100) Male 0.291*** 0.271*** 0.279*** 0.265*** (0.0052) (0.0050) (0.0049) (0.0050) Urban 0.189*** 0.112*** 0.111*** 0.104*** (0.0046) (0.0045) (0.0045) (0.0046) Coal sector 0.542*** 0.457*** 0.409*** (0.0238) (0.0248) (0.0248) Other mining 0.191*** 0.178*** 0.157*** (0.0175) (0.0172) (0.0171) Manufacturing 0.125*** 0.123*** 0.0873*** (0.0076) (0.0076) (0.0078) Electricity, gas, water 0.106*** 0.106*** 0.0734*** (0.0245) (0.0240) (0.0238) Construction 0.0441*** 0.0737*** 0.0656*** (0.0075) (0.0074) (0.0074) Wholesale & retail, restaurants -0.00769 -0.0132* -0.0219*** (0.0072) (0.0071) (0.0071) Transportation & communications 0.0302*** 0.00793 0.00166 (0.0102) (0.0099) (0.0098) Finance & business services 0.294*** 0.267*** 0.231*** (0.0135) (0.0131) (0.0135) Community & personal -0.0109 -0.00219 -0.0417*** (0.0083) (0.0081) (0.0085) Administrative & Managerial 0.426*** (0.0215) Clerical 0.0843*** (0.0107) Sales -0.162*** (0.0117) Services workers -0.249*** (0.0134) Skilled agricultural -0.194*** (0.0118) Production & Machine operators -0.147*** (0.0109) Others -0.0850*** (0.0162) Permanent wage employees 0.0909*** (0.0053) Constant 7.046*** 7.220*** 7.463*** 7.168*** (0.0237) (0.0229) (0.0251) (0.0232) Observations 193,977 193,977 193,977 193,977 R-squared 0.197 0.24 0.244 0.242 Notes: Table reports results for Ordinary Least Squares regressions estimating the correlation of individual characteristics with log real hourly labor earnings of wage employees using data from Indonesia’s Sakernas 2018 dataset. Real wages are expressed in constant 2007 rupiahs. Reference categories are: less than primary complete education, agriculture sector, and professional or technical occupations. Columns 2, 3 and 4 include province controls. Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Province controls are included in columns 2, 3 and 4. Source: Authors’ estimates. 159 Annex 2 Table 3 OLS regressions on the correlates of real hourly wages in South Kalimantan and East and North Kalimantan provinces (2018)   VARIABLES (1) (2) (3) (4) 0.0546*** 0.0540*** 0.0571*** 0.0564*** Age (0.00462) (0.00461) (0.00459) (0.00463) -0.000536*** -0.000530*** -0.000582*** -0.000550*** Age squared (593e-05) (593e-05) (589e-05) (592e-05) 0.130*** 0.124*** 0.116*** 0.126*** Complete Primary (0.0304) (0.0304) (0.0304) (0.0303) 0.393*** 0.374*** 0.343*** 0.362*** Incomplete or Complete Secondary (0.0292) (0.0292) (0.0293) (0.0292) 0.852*** 0.835*** 0.662*** 0.808*** Post-Secondary (0.0371) (0.0371) (0.0415) (0.0372) 0.273*** 0.270*** 0.313*** 0.262*** Male (0.0198) (0.0197) (0.0199) (0.0196) 0.0263 0.0098 0.0006 0.0056 Urban (0.0180) (0.0181) (0.0179) (0.0181) 0.385*** 0.380*** 0.335*** Coal sector (0.0312) (0.0313) (0.0321) 0.0720 0.0688 0.0424 Other mining (0.0578) (0.0571) (0.0571) -0.109*** -0.0961*** -0.115*** Manufacturing (0.0342) (0.0343) (0.0345) 0.172* 0.184* 0.1540 Electricity, gas, water (0.1010) (0.0994) (0.0974) 0.0120 0.0206 (0.0000) Construction (0.0315) (0.0315) (0.0315) -0.0705*** -0.0600** -0.0584** Wholesale & retail, restaurants (0.0269) (0.0270) (0.0269) (0.0316) (0.0192) (0.0277) Transportation & communications (0.0405) (0.0404) (0.0404) 0.134*** 0.137*** 0.106** Finance & business services (0.0452) (0.0449) (0.0452) -0.0587** -0.0515* -0.0863*** Community & personal (0.0294) (0.0294) (0.0303) 0.365*** Administrative & Managerial (0.0686) (0.0172) Clerical (0.0372) -0.180*** Sales (0.0412) -0.357*** Services workers (0.0501) -0.182*** Skilled agricultural (0.0409) -0.169*** Production & Machine operators (0.0383) -0.172*** Others (0.0516) 0.102*** Permanent wage employees (0.0221) 6.963*** 7.040*** 7.167*** 6.945*** Constant (0.0896) (0.0899) (0.0958) (0.0927) Observations 10,023 10,023 10,023 10,023 R-squared 0.193 0.197 0.194 0.199 Notes: Table reports results for Ordinary Least Squares regressions estimating the correlation of individual characteristics with log real hourly labor earnings of wage employees using data from Indonesia’s Sakernas 2018 dataset. Real wages are expressed in constant 2007 rupiahs. Reference categories are: less than primary complete education, agriculture sector, and professional or technical occupations. Columns 2, 3 and 4 include province controls. Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Province controls are included in columns 2, 3 and 4. Source: Author's estimates. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 160 Annex 2 Table 4 OLS regressions comparing non-coal real wage growth in coal-intensive and non-coal intensive districts in South Kalimantan and East and North Kalimantan provinces, 2007 to 2012 Non-Coal employment Agriculture Manufacturing Construction Services (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) 0.228*** 0.157*** 0.0318 -0.0227 0.477*** 0.432*** 0.322*** 0.293*** 0.263*** 0.190*** Coal-intensive district (0.0233) (0.0267) (0.0409) (0.0482) (0.0683) (0.0748) (0.0530) (0.0619) (0.0368) (0.0406) -0.00273 -0.0071 -0.024 -0.0281 -0.0108 -0.0109 -0.0896* -0.0884 0.0277 0.0213 2012 (0.0224) (0.0223) (0.0388) (0.0388) (0.0586) (0.0586) (0.0543) (0.0538) (0.0340) (0.0340) 2012* -0.234*** -0.229*** -0.178** -0.172** -0.560*** -0.559*** -0.291*** -0.292*** -0.185*** -0.178*** Coal-intensive district (0.0409) (0.0409) (0.0705) (0.0705) (0.1520) (0.1520) (0.0919) (0.0914) (0.0647) (0.0648) 0.0816*** 0.0799*** 0.0458 0.044 -0.0246 -0.0211 0.0777 0.0762 0.111*** 0.109*** Age (0.0213) (0.0212) (0.0360) (0.0360) (0.0570) (0.0577) (0.0521) (0.0523) (0.0326) (0.0323) -0.000910*** -0.000889*** -0.000573 -0.000553 0.000709 0.000654 -0.000947 -0.000929 -0.00126*** -0.00124*** Age squared (0.0003) (0.0003) (0.0005) (0.0005) (0.0008) (0.0008) (0.0008) (0.0008) (0.0005) (0.0005) 0.0871** 0.0884** 0.107** 0.108** -0.00248 0.00717 -0.0451 -0.0443 0.115* 0.116* Complete Primary (0.0365) (0.0364) (0.0540) (0.0538) (0.1280) (0.1310) (0.0977) (0.0982) (0.0693) (0.0684) Incomplete or 0.240*** 0.227*** 0.196*** 0.194*** 0.346*** 0.338*** 0.0944 0.0865 0.240*** 0.220*** Complete Secondary (0.0348) (0.0350) (0.0536) (0.0535) (0.1170) (0.1200) (0.1010) (0.1020) (0.0627) (0.0620) 0.852*** 0.837*** 0.745*** 0.740*** 1.440*** 1.442*** 0.594*** 0.579*** 0.801*** 0.778*** Post-Secondary (0.0482) (0.0481) (0.1420) (0.1400) (0.1990) (0.1990) (0.1730) (0.1760) (0.0699) (0.0690) -0.160*** -0.174*** -0.163*** -0.182*** -0.230*** -0.236*** -0.188*** -0.195*** -0.197*** -0.205*** Urban (0.0199) (0.0201) (0.0579) (0.0562) (0.0670) (0.0671) (0.0437) (0.0462) (0.0330) (0.0330) -0.121*** -0.0824** -0.0906 -0.0558 -0.136*** South Kalimantan (0.0245) (0.0397) (0.0625) (0.0642) (0.0367) 6.781*** 6.897*** 7.661*** 7.755*** 8.432*** 8.431*** 7.116*** 7.181*** 6.224*** 6.338*** Constant (0.3640) (0.3620) (0.6080) (0.6060) (0.9480) (0.9520) (0.8690) (0.8750) (0.5660) (0.5600) Observations 6,990 6,990 1,969 1,969 643 643 773 773 3,325 3,325 R-squared 0.052 0.054 0.353 0.357 0.144 0.146 0.155 0.161 Notes: Table reports results building on Black et al. (2005) methodology estimating the change in real hourly wages in non-coal sectors between 2007 and 2012 by regressing log wages in sector i on dummies for coal-intensive district, end year, and the interaction term. District is coal-intensive if coal employment share is at least 4%. The sample is restricted to males aged 25-45 to reduce bias due to changing composition of the workforce. Reference category is less than primary complete education. A negative coefficient value on the interaction variable 2012*coal-intensive district indicates slower wage growth in coal-intensive districts. Data from Indonesia's Sakernas 2007 and 2012. Standard errors are reported in parentheses where *** , ** and * indicate significance at 1, 5 and 10 percent respectively. Source: Author's estimates. 161 Annex 2 Table 5 OLS regressions comparing real wage growth in coal-intensive and non-coal intensive districts in South Kalimantan and East and North Kalimantan provinces, 2012 to 2018 Non-Coal employment Agriculture Manufacturing Construction Services (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) 0.0039 (0.0101) -0.144** -0.118* (0.0802) (0.0565) 0.0590 0.1100 0.0954* 0.0607 Coal district intensity (0.0335) (0.0354) (0.0571) (0.0641) (0.1350) (0.1370) (0.0761) (0.0762) (0.0536) (0.0553) 0.194*** 0.193*** 0.164*** 0.162*** 0.268*** 0.270*** 0.365*** 0.376*** 0.162*** 0.160*** 2018 (0.0251) (0.0250) (0.0461) (0.0461) (0.0677) (0.0669) (0.0549) (0.0528) (0.0376) (0.0374) 2018*Coal district 0.0160 0.0174 0.158** 0.159** 0.1870 0.1870 (0.1440) (0.1570) (0.0872) (0.0844) intensity (0.0474) (0.0474) (0.0786) (0.0787) (0.2020) (0.2020) (0.1240) (0.1230) (0.0736) (0.0735) 0.100*** 0.101*** 0.0725* 0.0747* (0.1010) (0.1060) 0.0789 0.0770 0.123*** 0.124*** Age (0.0240) (0.0240) (0.0416) (0.0417) (0.0774) (0.0780) (0.0540) (0.0536) (0.0355) (0.0354) -0.00117*** -0.00117*** (0.0010) -0.000980* 0.0016 0.0017 (0.0009) (0.0009) -0.00141*** -0.00143*** Age square (0.0003) (0.0003) (0.0006) (0.0006) (0.0011) (0.0011) (0.0008) (0.0008) (0.0005) (0.0005) 0.0676* 0.0678* 0.0565 0.0531 (0.1300) (0.1230) (0.0208) (0.0209) 0.129* 0.128* Complete Primary (0.0382) (0.0382) (0.0587) (0.0586) (0.1360) (0.1370) (0.0781) (0.0781) (0.0735) (0.0736) Incomplete | Complete 0.233*** 0.230*** 0.257*** 0.258*** 0.1720 0.1840 0.0586 0.0674 0.245*** 0.232*** Secondary (0.0362) (0.0365) (0.0590) (0.0589) (0.1200) (0.1220) (0.0787) (0.0804) (0.0651) (0.0654) 0.700*** 0.697*** 0.697*** 0.701*** 0.893*** 0.891*** 0.346** 0.371*** 0.704*** 0.691*** Post-Secondary (0.0473) (0.0474) (0.1890) (0.1910) (0.2070) (0.2070) (0.1430) (0.1420) (0.0707) (0.0707) -0.0946*** -0.0982*** -0.157*** -0.149*** (0.1310) (0.1220) (0.0574) (0.0398) -0.0858** -0.0931*** Urban (0.0215) (0.0217) (0.0529) (0.0522) (0.0796) (0.0790) (0.0456) (0.0478) (0.0356) (0.0356) (0.0255) 0.0449 0.0508 0.0883 -0.0710* South Kalimantan (0.0257) (0.0465) (0.0705) (0.0591) (0.0380) 6.434*** 6.447*** 7.176*** 7.108*** 10.02*** 10.07*** 6.894*** 6.858*** 5.925*** 5.955*** Constant (0.4130) (0.4130) (0.7250) (0.7290) (1.3200) (1.3280) (0.9340) (0.9260) (0.6110) (0.6090) Observations 4,992 4,992 1,313 1,313 430 430 540 540 2,516 2,516 R-squared 0.13 0.13 0.087 0.088 0.212 0.213 0.154 0.159 0.14 0.141 Notes: Table reports results building on Black et al. (2005) methodology estimating the change in real hourly wages in non-coal sectors between 2012 and 2018 by regressing log wages in sector i on dummies for coal-intensive district, end year, and the interaction term. District is coal-intensive if coal employment share is at least 4%. The sample is restricted to males aged 25-45 to reduce bias due to changing composition of the workforce. Reference category is less than primary complete education. A negative coefficient value on the interaction variable 2018*coal-intensive district indicates slower wage growth in coal-intensive districts. Data from Indonesia's Sakernas 2012 and 2018. Standard errors are reported in parentheses where *** , ** and * indicate significance at 1, 5 and 10 percent respectively. Source: Author's estimates. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 162 Annex 2 Table 6 OLS regressions estimating coal to non-coal employment elasticities Non-Coal employment Agriculture Manufacturing Construction Services (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) 0.408*** 0.387*** 0.545*** 0.514*** 0.591*** 0.565*** 0.363*** 0.355*** 0.414*** 0.399*** Coal employment (0.0408) (0.0367) (0.106) (0.103) (0.0978) (0.0956) (0.0704) (0.0705) (0.0522) (0.0510) -0.726*** -0.580*** -0.686*** -0.469** -1.198*** -1.014*** -0.537*** -0.482*** -0.735*** -0.634*** Coal-intensive district (0.0786) (0.0737) (0.203) (0.206) (0.188) (0.192) (0.135) (0.141) (0.101) (0.102) 2.63e-06*** 2.80e-06*** -1.78e-06*** -1.52e-06** 3.98e-06*** 4.19e-06*** 3.89e-06*** 3.95e-06*** 3.78e-06*** 3.89e-06*** lag total pop (2.62e-07) (2.36e-07) (6.77e-07) (6.60e-07) (6.27e-07) (6.15e-07) (4.51e-07) (4.53e-07) (3.35e-07) (3.28e-07) 0.0507 0.0641 0.173 0.193 -0.233 -0.216 0.108 0.113 -0.00566 0.00364 2008.year (0.116) (0.104) (0.299) (0.290) (0.277) (0.270) (0.200) (0.199) (0.148) (0.144) -0.0758 -0.0547 0.156 0.187 -0.552** -0.525* -0.0505 -0.0425 -0.147 -0.132 2009.year (0.115) (0.104) (0.298) (0.289) (0.276) (0.270) (0.199) (0.199) (0.148) (0.144) 0.0525 0.0340 0.0572 0.0298 -0.136 -0.160 0.0588 0.0518 0.115 0.102 2010.year (0.113) (0.101) (0.293) (0.284) (0.271) (0.264) (0.195) (0.195) (0.145) (0.141) 0.0133 0.00775 0.0607 0.0525 -0.444* -0.451* 0.149 0.146 0.0652 0.0613 2011.year (0.111) (0.0996) (0.287) (0.278) (0.266) (0.259) (0.191) (0.191) (0.142) (0.138) -0.0176 -0.0210 0.0210 0.0159 -0.331 -0.336 0.221 0.220 0.0377 0.0353 2012.year (0.110) (0.0989) (0.285) (0.276) (0.264) (0.257) (0.190) (0.190) (0.141) (0.137) 0.0184 0.00494 0.0670 0.0470 -0.273 -0.290 0.0913 0.0862 0.0900 0.0806 2014.year (0.109) (0.0981) (0.283) (0.274) (0.262) (0.255) (0.189) (0.188) (0.140) (0.136) 0.132 0.0970 0.104 0.0516 -0.155 -0.199 0.376* 0.363* 0.270* 0.246* 2015.year (0.116) (0.105) (0.301) (0.292) (0.279) (0.272) (0.201) (0.201) (0.149) (0.145) 0.200* 0.134 0.179 0.0818 0.325 0.243 0.164 0.139 0.359** 0.314** 2018.year (0.119) (0.107) (0.307) (0.298) (0.284) (0.278) (0.205) (0.205) (0.152) (0.148) East & North -0.333*** -0.494*** -0.419*** -0.126 -0.231*** Kalimantan (0.0493) (0.138) (0.128) (0.0947) (0.0685) 7.945*** 8.198*** 6.495*** 6.870*** 3.580*** 3.898*** 4.845*** 4.941*** 6.743*** 6.919*** Constant (0.280) (0.253) (0.723) (0.708) (0.670) (0.660) (0.482) (0.487) (0.358) (0.352) 0.408*** 0.387*** 0.545*** 0.514*** 0.591*** 0.565*** 0.363*** 0.355*** 0.414*** 0.399*** Observations 194 194 194 194 194 194 194 194 194 194 R-squared 0.799 0.839 0.142 0.199 0.624 0.645 0.652 0.656 0.791 0.803 Notes: Table reports regression results based on Black et al. (2015) methodology to estimate the correlation between log coal employment and log non-coal employment in coal provinces and whether the degree of coal-intensity matters. Province and year fixed effects. The coefficient on “coal employment” is the elasticity of the number of jobs in the non-coal sector to the number of jobs in the coal sector. The coefficient on “coal-intensive district” – a dummy variable equal to 1 for districts with a coal employment share of at least 4% – captures the difference in employment between coal-intensive districts and non- coal-intensive districts; a negative coefficient value indicates lower non-coal employment in coal-intensive districts. Sample restricted to districts in South Kalimantan and East and North Kalimantan and excludes districts that have zero or negligible coal employment (i.e., less than 0.5 percent). Data from Indonesia’s Sakernas 2007-2018. Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Source: Author's estimates. 163 ANNEX 3 Technical Results for South Africa Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 164 Methodology for estimating coal employment spill-overs in South Africa Building on Black et al. (2005), we estimate the impact of coal sector jobs on non-coal employment within Mpumalanga province, and test the impact during periods of rapid coal sector expansion (“coal boom”) and contraction (“coal bust”) to detect any differential effects. We estimate the following equation: Ln(Ys,t )=β 0 + β1 Ln (X t ) + β 2 P + β 3 M t + Ut,i where Ys,t is the number of jobs in non-coal sector s at time t ; X t is the number of jobs in the coal sector at time t ; P is a dummy variable representing the coal bust period between year 2015 and 2017 and M t controls for other mining and quarrying activities at time t . Ut,i is the error term. Therefore, β1 is the elasticity of the number of jobs in the non-coal sector to the number of jobs in the coal sector, and β 2 captures the difference in employment in the non-coal sector between the boom period and the bust period. We run our equation for non-coal sectors aggregated together, and subsequently for each non-coal sector separately. For robustness, we test by excluding other mining and quarrying activity control and find similar results. We follow Black et al. (2005) and Moritz et al. (2017) by restricting our sample to districts within coal provinces because they share similar institutional and geographic characteristics, which reduces confounding factors. The sample includes only coal provinces South Kalimantan and East and North Kalimantan, and within these, excludes districts that have zero or negligible coal employment (i.e., less than 0.5 percent). For robustness, we test on a wider sample of coal provinces (namely by adding South Sumatera and Banten) and find similar results. 165 Annex 3 Table 1 Probit regressions on labor market outcomes (likelihood of being employed) Employed/Unemployed 2019 2008 (1) (2) 0.0934*** 0.0992*** Age (0.00643) (0.00613) -0.000775*** -0.000890*** Age square (8.20e-05) (8.10e-05) 0.332*** 0.312*** Married (0.0252) (0.0247) 0.0879*** 0.0344 Urban (0.0271) (0.0254) -0.136*** -0.229*** Female (0.0209) (0.0215) 0.158*** 0.150*** Colored/mixed race (0.0533) (0.0496) 0.462*** 0.361*** Asian (0.107) (0.0942) 0.673*** 0.823*** White (0.0882) (0.0808) 0.0966 0.249*** Female* Colored/mixed race (0.0697) (0.0597) -0.222 0.102 Female * Asia (0.168) (0.137) 0.0114 0.136 Female * White (0.123) (0.108) -0.0143 -0.0732 Primary complete (0.0617) (0.0487) -0.131*** -0.0288 Secondary incomplete (0.0396) (0.0314) 0.124*** 0.0369 Secondary complete (0.0406) (0.0346) 0.674*** 0.778*** Post-secondary (0.0626) (0.0808) -2.003*** -1.563*** Constant (0.126) (0.113) Observations 24,710 31,598 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1. Notes: Table reports results for probit regressions estimating the probabilistic correlation of individual characteristics with labor market outcomes (labor market outcome is a binary variable were 1 is being employed and 0 is being unemployed) using data from South Africa QLF 3rd quarter, 2019 and 2008 dataset. Reference categories are less than primary complete education, male, and Black South African. The results include province controls. Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Source: Authors’ estimates. Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 166 Annex 3 Table 2 OLS regressions on the correlates of real hourly wages (All provinces, 2019) (1) (2) (3) (4) (5) VARIABLES OLS OLS OLS OLS OLS 0.0459*** 0.0432*** 0.0493*** 0.0436*** 0.0331*** Age (0.00612) (0.00608) (0.00595) (0.00582) (0.00610) -0.000388*** -0.000357*** -0.000415*** -0.000361*** -0.000259*** Age square (7.45e-05) (7.40e-05) (7.26e-05) (7.10e-05) (7.46e-05) -0.182*** -0.168*** -0.178*** -0.176*** -0.180*** Female (0.0204) (0.0201) (0.0200) (0.0196) (0.0186) 0.284*** 0.216*** 0.234*** 0.247*** 0.204*** Urban (0.0216) (0.0237) (0.0233) (0.0228) (0.0219) 0.142*** 0.152*** 0.117*** 0.0885** 0.148*** Colored /Mix (0.0304) (0.0361) (0.0346) (0.0343) (0.0356) 0.226** 0.252** 0.139 0.114 0.277*** Asian (0.108) (0.107) (0.0967) (0.0923) (0.102) 0.783*** 0.762*** 0.535*** 0.525*** 0.864*** White (0.0593) (0.0578) (0.0577) (0.0570) (0.0586) 0.0401 0.0339 0.0411 0.0371 0.0315 Primary complete (0.0448) (0.0441) (0.0446) (0.0447) (0.0438) 0.192*** 0.176*** 0.190*** 0.143*** 0.154*** Secondary incomplete (0.0306) (0.0303) (0.0307) (0.0305) (0.0296) 0.515*** 0.487*** 0.464*** 0.357*** 0.407*** Secondary complete (0.0350) (0.0348) (0.0354) (0.0349) (0.0336) 1.488*** 1.446*** 1.124*** 1.008*** 1.298*** Post-secondary (0.0478) (0.0475) (0.0500) (0.0494) (0.0459) 0.275** 0.206 0.114 0.0938 0.251* Other education (0.137) (0.138) (0.132) (0.137) (0.145) 0.338*** Formal sector worker (0.0225) 0.633*** Union member (0.0249) 1.008*** 0.993*** Coal (0.133) (0.134) 0.889*** 0.882*** Other Mining & quarrying (0.0617) (0.0616) 0.177*** 0.181*** Manufacturing (0.0415) (0.0411) 0.540*** 0.564*** Electricity, gas & water (0.0904) (0.0915) 0.0793* 0.108** Construction (0.0428) (0.0429) 167 Annex 3 Table 2 — OLS regressions on the correlates of real hourly wages (All province, 2019) (continued) (1) (2) (3) (4) (5) VARIABLES OLS OLS OLS OLS OLS -0.0424 -0.0289 Trade, catering & accommodation (0.0366) (0.0357) Transport, storage & 0.0172 0.0254 communication (0.0523) (0.0518) 0.174*** 0.170*** Finance and business services (0.0398) (0.0393) 0.734*** 0.756*** General government services (0.0625) (0.0622) 0.0509 0.0752** Personal Service (0.0373) (0.0370) -0.00847 0.000315 Other Services (0.0377) (0.0370) -0.329*** -0.336*** Technicians (0.0641) (0.0633) -0.556*** -0.564*** Clerks (0.0636) (0.0629) -0.956*** -0.896*** Service & sales (0.0586) (0.0582) -0.955*** -0.876*** Skilled agricultural (0.0634) (0.0630) -0.740*** -0.710*** Production, Craft workers, laborers (0.0609) (0.0603) -0.715*** -0.678*** Machine operators (0.0638) (0.0622) 1.175*** 1.326*** 2.132*** 2.047*** 1.550*** Constant (0.126) (0.128) (0.140) (0.137) (0.126) Observations 7,288 7,288 7,288 7,288 7,213 R-squared 0.434 0.446 0.453 0.474 0.474 Notes: Table reports results for Ordinary Least Squares regressions estimating the correlation of individual characteristics with log real hourly labor earnings of wage employees using data from South Africa QLFS 3rd quarter 2019 dataset. Reference categories are male workers, urban location, less than primary completed education, agriculture sector, and professional or technical occupations. Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Columns 2, 3, 4 and 5 include province controls. Source: Authors’ estimates Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 168 Annex 3 Table 3 OLS regressions on the correlates between coal employment and non-coal employment (Mpumalanga only, 2019) Non-Mining Agriculture Manufacturing Electricity Construction Services (1) (2) (3) (4) (5) (6) VARIABLES OLS OLS OLS OLS OLS OLS 0.274*** 0.0929 0.235** 0.481 0.221 0.302*** Coal employment (0.0699) (0.105) (0.114) (0.322) (0.145) (0.0662) 0.0683*** 0.0513** 0.0536** 0.183** 0.109*** 0.0629*** 2015-2017 (0.0151) (0.0226) (0.0246) (0.0695) (0.0313) (0.0143) -0.0175 0.0317 -0.0223 -0.108 0.00511 -0.0241 Other mining and quarrying (0.0248) (0.0371) (0.0404) (0.114) (0.0515) (0.0235) 10.99*** 10.06*** 9.072*** 5.802 8.942*** 10.39*** Constant (0.818) (1.225) (1.333) (3.769) (1.699) (0.775) Observations 48 48 48 48 48 48 R-squared 0.488 0.112 0.202 0.276 0.292 0.514 Notes: Table reports regression results building on Black et al. (2005) to estimate the correlation between log coal employment and log non-coal employment in Mpumalanga and the direct impacts of the coal bust period. The coefficient on “coal employment” is the elasticity of the number of jobs in the non-coal sector to the number of jobs in the coal sector. The coefficient on “coal bust period” – a dummy variable equal to 1 during the period when coal employment declined between 2015 and 2017 – captures the difference in employment between bust and boom periods; a positive coefficient value indicates higher non-coal employment in the bust period. Sample restricted to Mpumalanga province. Data from South Africa QLFS 2008-2019. Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Source: Authors’ estimates 169 ANNEX 4 Technical Results for India Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 170 Annex 4 Table 1 OLS regressions on the correlates of real weekly wages (all states, 2017) (1) (2) (3) (4) (5) VARIABLES OLS OLS OLS OLS OLS 0.0416*** 0.0388*** 0.0379*** 0.0352*** 0.0393*** Age (0.00227) (0.00215) (0.00212) (0.00250) (0.00264) -0.000406*** -0.000370*** -0.000368*** -0.000320*** -0.000360*** Age square (2.95e-05) (2.81e-05) (2.77e-05) (3.31e-05) (3.49e-05) -0.452*** -0.459*** -0.503*** -0.507*** -0.449*** Female (0.0108) (0.0107) (0.0103) (0.0122) (0.0129) 0.214*** 0.202*** 0.230*** 0.182*** 0.212*** Urban (0.00899) (0.00876) (0.00790) (0.00822) (0.00887) 0.420*** 0.375*** Formal (0.0124) (0.0128) 0.0700*** 0.0931*** 0.0887*** 0.105*** 0.107*** Primary education (0.0136) (0.0130) (0.0126) (0.0146) (0.0160) 0.200*** 0.198*** 0.183*** 0.184*** 0.244*** Secondary incomplete (0.0110) (0.0105) (0.0102) (0.0118) (0.0127) 0.436*** 0.424*** 0.350*** 0.336*** 0.470*** Secondary complete (0.0153) (0.0148) (0.0148) (0.0156) (0.0164) 0.951*** 0.944*** 0.716*** 0.650*** 0.922*** Post-secondary (0.0152) (0.0149) (0.0165) (0.0168) (0.0157) 0.138* 0.0945 0.122* 0.177*** 0.196*** Others Unspecified (0.0783) (0.0673) (0.0717) (0.0502) (0.0670) 1.249*** 1.356*** 0.999*** Coal (0.0893) (0.0991) (0.0839) 0.414*** 0.391*** 0.356*** Other mining and Quarrying (0.0379) (0.0382) (0.0523) 0.364*** 0.349*** 0.301*** Manufacturing (0.0142) (0.0136) (0.0354) 0.622*** 0.604*** 0.419*** Public Utilities (0.0354) (0.0337) (0.0479) 0.207*** 0.208*** 0.220*** Construction (0.0123) (0.0117) (0.0349) 0.222*** 0.197*** 0.178*** Wholesale & Retailing (0.0176) (0.0165) (0.0361) 0.534*** 0.510*** 0.457*** Transport & Communications (0.0178) (0.0171) (0.0365) 0.517*** 0.486*** 0.401*** Financial & Business Services (0.0208) (0.0206) (0.0380) 0.816*** 0.783*** 0.569*** Public Administration (0.0232) (0.0227) (0.0401) 0.372*** 0.352*** 0.236*** Others, unspecified (0.0162) (0.0159) (0.0362) -0.0482* -0.135*** Professionals (0.0282) (0.0251) -0.176*** -0.298*** Technicians (0.0294) (0.0265) 171 Annex 4 Table 1 — OLS regressions on the correlates of real weekly wages (all states, 2017) (continued) (1) (2) (3) (4) (5) VARIABLES OLS OLS OLS OLS OLS -0.179*** -0.308*** Clerks (0.0285) (0.0265) -0.498*** -0.538*** Service & market sales workers (0.0276) (0.0249) -0.720*** -0.500*** Skilled agricultural (0.0358) (0.0566) -0.517*** -0.506*** Craft workers (0.0276) (0.0248) -0.350*** -0.383*** Machine operators (0.0280) (0.0254) -0.693*** -0.612*** Elementary occupations (0.0278) (0.0250) 6.037*** 5.940*** 6.903*** 6.871*** 5.601*** Constant (0.0433) (0.0480) (0.0543) (0.0588) (0.0607) Observations 69,794 69,794 69,795 55,218 55,218 R-squared 0.474 0.509 0.507 0.519 0.471 Notes: Table reports results for Mincer-type ordinary least squares regressions estimating the correlation of individual characteristics with log real weekly labor earnings of wage employees using data from India PLFS 2017-18 dataset. Reference categories are male workers, rural location, less than primary completed education, agriculture sector, and senior officials, Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Column 2, 3 4, 5 includes state controls. Source: Authors’ estimates Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 172 Annex 4 Table 2 OLS regressions on the correlates of real weekly wages (coal states, 2017) VARIABLES (1) (2) (3) (4) (5) 0.0454*** 0.0430*** 0.0424*** 0.0378*** 0.0427*** Age (0.00411) (0.00397) (0.00391) (0.00487) (0.00501) -0.000471*** -0.000436*** -0.000436*** -0.000358*** -0.000412*** Age square (5.32e-05) (5.11e-05) (5.07e-05) (6.44e-05) (6.60e-05) -0.384*** -0.426*** -0.460*** -0.524*** -0.445*** Female (0.0185) (0.0181) (0.0173) (0.0223) (0.0249) 0.236*** 0.202*** 0.224*** 0.185*** 0.221*** Urban (0.0168) (0.0160) (0.0141) (0.0153) (0.0171) 0.553*** 0.480*** Formal (0.0207) (0.0223) 0.0662*** 0.122*** 0.115*** 0.111*** 0.0726** Primary education (0.0230) (0.0222) (0.0211) (0.0273) (0.0299) 0.197*** 0.218*** 0.181*** 0.185*** 0.216*** Secondary incomplete (0.0197) (0.0188) (0.0183) (0.0220) (0.0235) 0.451*** 0.470*** 0.382*** 0.325*** 0.431*** Secondary complete (0.0279) (0.0267) (0.0262) (0.0294) (0.0307) 1.038*** 1.028*** 0.767*** 0.653*** 0.918*** Post-secondary (0.0262) (0.0254) (0.0285) (0.0318) (0.0291) -0.118 -0.0690 -0.110 0.0488 0.00788 Others Unspecified (0.170) (0.166) (0.181) (0.0912) (0.0900) 1.402*** 1.452*** 1.186*** Coal (0.0655) (0.0686) (0.0843) 0.306*** 0.334*** 0.292*** Other mining and Quarrying (0.0693) (0.0735) (0.0882) 0.198*** 0.233*** 0.240*** Manufacturing (0.0259) (0.0252) (0.0621) 0.428*** 0.454*** 0.263*** Public Utilities (0.0657) (0.0645) (0.0859) 0.107*** 0.136*** 0.232*** Construction (0.0205) (0.0198) (0.0601) 0.0977*** 0.108*** 0.167*** Wholesale & Retailing (0.0282) (0.0262) (0.0630) 0.474*** 0.458*** 0.479*** Transport & Communications (0.0328) (0.0315) (0.0635) 0.344*** 0.368*** 0.272*** Financial & Business Services (0.0415) (0.0411) (0.0688) 0.637*** 0.670*** 0.421*** Public Administration (0.0443) (0.0432) (0.0719) 173 Annex 4 Table 2 — OLS regressions on the correlates of real weekly wages (coal states, 2017) (continued) VARIABLES (1) (2) (3) (4) (5) 0.186*** 0.225*** 0.153** Others, unspecified (0.0271) (0.0259) (0.0627) -0.0548 -0.0198 Professionals (0.0573) (0.0580) -0.215*** -0.218*** Technicians (0.0544) (0.0575) -0.143** -0.185*** Clerks (0.0569) (0.0592) -0.529*** -0.391*** Services & market sales workers (0.0530) (0.0574) -0.727*** -0.343*** Skilled agricultural (0.0634) (0.102) -0.487*** -0.293*** Craft workers (0.0525) (0.0570) -0.291*** -0.155*** Machine operators (0.0537) (0.0581) -0.680*** -0.443*** Elementary occupations (0.0521) (0.0575) 5.943*** 5.933*** 6.813*** 6.656*** Constant (0.0787) (0.0809) (0.0934) (0.112) Observations 18,649 18,649 18,649 13,447 13,447 R-squared 0.459 0.496 0.495 0.554 0.513 Notes: Table reports results for Mincer-type ordinary least squares regressions estimating the correlation of individual characteristics with log real weekly labor earnings of wage employees using data from India PLFS 2017-18 dataset. Reference categories are male workers, Rural location, less than primary completed education, agriculture sector, and senior officials, Standard errors are reported in parentheses where *** , **, and * indicate significance at 1, 5, and 10 percent respectively. Column 2, 3 4, 5 includes coal state controls. Source: Authors’ estimates Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses 174 175 Global Perspective on Coal Jobs and Managing Labor Transition out of Coal Key Issues and Policy Responses Jobs Group | World Bank | 2021 Elizabeth Ruppert Bulmer, Kevwe Pela, Andreas Eberhard-Ruiz, Jimena Montoya