Multidimensional Poverty and Spatial Disparities in Solomon Islands 2009-2019 © 2025 The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved. This work is a product of The World Bank. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, links/footnotes and other infor- mation shown in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The citation of works authored by others does not mean The World Bank endorses the views expressed by those authors or the content of their works. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Attribution—Please cite the work as follows: “World Bank. 2025. Multidimensional Poverty and Spatial Disparities in Solomon Islands. Washington, DC: World Bank. © World Bank.” Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e‑mail: pubrights@worldbank.org. Cover photo: © World Bank. Cover design: Laframboise Design. and Spatial Disparities Multidimensional Poverty  inSolomon Islands Acknowledgments The report was written by Shohei Nakamura (Senior Economist and Task Team Leader, EEAPV), Ryoko Sato (Consultant, EEAPV), and Dhanu Thamarapani (Consultant, EEAPV). The team would like to thank the following colleagues for their comments and inputs: Renee Berthome (Energy Specialist, IEAE1), Alice Calder (Consultant, SEAS1), Shona Fitzgerald (Senior Water Supply and Sanitation Specialist, SEAW1), Virginia Horscroft (Senior Public Sector Specialist, EEAG1), Paul Lachlan McDonald (Consultant, EEAM2), Darian Naidoo (Economist, EEAPV), Tevi Obed (Senior Disaster Risk Management Specialist, SEAU1), Alison Ofotalau (Operations Officer, EACSB), Reshika Singh (Senior Economist, EEAM2), Ifeta Smajic (Senior Social Development Specialist, SEAS1), Lodewijk Smets (Senior Economist, EEAM2), Sharad Tandon (Senior Economist, EEAPV), and Son Thanh Vo (Senior Agricultural Specialist, SEAAG). Mildred Gonsalvez (Program Assistant, EEAPV) and Manju Venkiteswaran (Program Assistant, EAPCF) provided administrative support. The report was prepared under the guidance of Lalita Moorty (Regional Director for East Asia and the Pacific); Stephen Ndegwa (Division Director for Papua New Guinea and Pacific Countries); Annette Leith (Resident Representative for Solomon Islands and Vanuatu); Ralph van Doorn (Program Leader); Rinku Murgai (Former Practice Manager, EEAPV); and Benu Bidani (Practice Manager, EEAPV). The report benefitted from detailed peer review from Lodewijk Smets (Senior Economist, EEAM2), Sailesh Tiwari (Lead Economist, EEAPV), Aude-Sophie Rodella (Lead Water Economist, SEAW1), and many colleagues at the World Bank who provided insights, suggestions, and improvements to the report process. Brenan Gabriel Andre (Consultant, GOHMA) and Bruno Bonansea (Program Manager, GCSIM) supported the preparation of maps. Angela Takats edited the report, with graphic design by Laframboise Design. This report was produced as a joint work between the Solomon Islands National Statistics Office (SINSO) and the World Bank. The team especially thanks Samson Kanamoli, Government Statistician, and the late Douglas Kimi, Former Government Statistician. The team would also like to thank Mark Johnston (Institutional Development Advisor, Ministry of Agriculture and Livestock), who was the former Rural Development Program (RDP) International Advisor and provided detailed project beneficiary information. 3 Table of Contents List of Figures.....................................................................................................................................................................5 List of Tables.......................................................................................................................................................................6 List of Boxes........................................................................................................................................................................6 List of Abbreviations........................................................................................................................................................7 Acknowledgments.........................................................................................................................................................3 Executive Summary......................................................................................................................................................8 Chapter 1. Introduction..........................................................................................................................................12 1.1. Context......................................................................................................................................................................12 1.2. Measuring multidimensional poverty............................................................................................................14 1.3. Potential key drivers for poverty reduction.................................................................................................16 Chapter 2. Poverty and Spatial Inequality...................................................................................................21 2.1. Solomon Islands’ multidimensional poverty level is higher than the average of lower-middle-income countries.................................................................................................21 2.2. Multidimensional poverty is more prevalent and concentrated in rural areas.............................23 2.3. Multidimensionally poverty incidence declined across the board between 2009 and 2019........................................................................................................................26 2.4. Improvements in living standard dimensions, particularly housing and water, drove multidimensional poverty reduction..........................................................................................................30 Chapter 3. Migration, Urbanization, and Jobs............................................................................................37 3.1. Urbanization has been driven by migration to Honiara..........................................................................38 3.2. Migration to Honiara was selective, with migrants’ characteristics better than rural stayers and similar to Honiara residents.........................................................................................41 3.3. Internal migration likely accounted for a significant share of multidimensional poverty reduction achieved during the decade..................................................................................................42 3.4. Strategic urban planning and service delivery is essential to harness urbanization for poverty reduction.....................................................................................................................................................47 Chapter 4. Government Policies and Investments Towards Rural Areas....................................48 4.1. CDF likely contributed to multidimensional poverty reduction..........................................................48 4.2. The government heavily invested in household water access, though water deprivation was reduced only moderately...............................................................................51 4.3. Reducing concentrated rural poverty requires rural investments to be more efficient and effective...........................................................................................................................52 References.......................................................................................................................................................................54 Annex A: Technical Note..........................................................................................................................................56 A1. Multidimensional poverty measurement......................................................................................................56 A2. Definitions of urban area and Honiara Urban Area....................................................................................59 A3. Measuring labor force status in Population and Housing Censuses 2009 and 2019................60 Annex B. Additional Figures and Tables.........................................................................................................61 Annex C. Maps...............................................................................................................................................................68 4 and Spatial Disparities Multidimensional Poverty  inSolomon Islands List of Figures Figure ES1. Multidimensional poverty rates in Solomon Islands and other countries..........................9 Figure ES2. Multidimensional poverty rates by location, 2009 and 2019.............................................. 9 Changes in multidimensional poverty dimensions and indicators, Figure ES3.  2009 and 2019................................................................................................................................... 10 Figure 1. Solomon Islands achieved only moderate economic growth from 2009 to 2019............................................................................................................................ 13 Figure 2. Multidimensional poverty consists of health, education, and living standard dimensions.................................................................................................... 14 Figure 3. Multidimensional and monetary poverty are positively correlated at the ward level.................................................................................................................................. 16 Figure 4. Low levels of GDP per capita and urbanization are associated with a high share of agricultural workers.................................................................................. 19 Figure 5. CDF Budget has increased significantly since 2009........................................................... 20 Figure 6. Solomon Islands’ multidimensional poverty rate is relatively high............................... 22 Figure 7. Solomon Islands is lagging in many non-monetary outcomes....................................... 22 Figure 8. Wards with high multidimensional poverty are concentrated in Guadalcanal and Malaita............................................................................................................. 24 Figure 9. A group of wards are characterized by high poverty incidence and populations..... 25 Figure 10. One-third of the multidimensional poor are children under age 10, while there is no gender difference............................................................................................ 26 Figure 11. Nearly two-thirds of the multidimensionally poor population live in Honiara, Guadalcanal, and Malaita......................................................................................... 27 Figure 12. MPI reduction in poorer wards resulted in a convergence between 2009 and 2019............................................................................................................... 29 Figure 13. Almost all multidimensionally poor people are deprived in living standards, while being deprived in all three dimensions is rare............................................................ 30 Figure 14. Deprivations in housing, water, and, to a lesser extent, sanitation and cooking fuel improved........................................................................................ 30 Figure 15. Nearly 90 percent of working-age adults did not complete secondary education............................................................................................. 32 Figure 16. Mobile phone ownership increased, but asset ownership remains low....................... 35 Figure 17. Living standards are clearly associated with household heads’ education levels............................................................................................. 35 Figure 18. Population changes at the ward level, 2009-2019.............................................................. 39 Figure 19. People with higher education and/or living in poor areas tended to migrate to Honiara........................................................................................................................ 41 Figure 20. Unlike rural stayers, about 75 percent of the recent migrant households in Honiara Urban Area rely on wages and salaries................................................................ 41 Figure 21. In terms of multidimensional poverty status, recent migrants and non-migrants are similar in Honiara Urban Area unless they still rely on agriculture as the main income source............................................................................................................. 42 Figure 22. Internal migration accounts for nearly 30 percent of multidimensional poverty reduction...................................................................................... 43 Figure 23. Multidimensional poverty rates are higher among agricultural households.............. 44 Figure 24. No structural shift took place at the national level, with the agricultural worker share increasing in urban areas.......................................... 44 5 Figure 25. Only 11 wards experienced more than 10 percentage point increases in non-agricultural worker shares............................................................................ 45 Figure 26. Multidimensional poverty reduction was faster among subsistence farmers........... 46 Figure 27. Multidimensional poverty incidence is lower among households whose heads are wage workers............................................................................ 46 Figure 28. Female, low-skilled, recent migrants are more likely to be agricultural workers in Honiara Urban Area.................................................................. 47 Figure 29. About 40 percent of CDFs was spent on housing................................................................. 49 Figure 30. Multidimensional poverty improved more in places where more people positively evaluated the CDFs............................................................................ 50 Figure 31. RDP (I & II) and CRISP invested in household water access in the majority of rural wards......................................................................................................... 52 List of Tables Table 1. Solomon Islands experienced rapid urbanization, particularly around Honiara, between 2009 and 2019....................................................... 17 Table 2. Neither strong job growth nor a structural shift took place between 2009 and 2019..................................................................................................... 18 Table 3. Multidimensional poverty is more prevalent and concentrated in rural areas.......... 23 Table 4. Multidimensional poverty incidence moderately fell between 2009 and 2019....... 27 Table 5. Number of multidimensionally poor individuals (thousands).......................................... 28 Table 6. In contrast to the declining multidimensional poverty incidence, there was no change in the intensity between 2009 and 2019............... 29 Table 7. Other than educational attainment in rural areas, there was no significant improvement in education deprivation........................................................ 31 Table 8. Many rural households shifted to improved water sources for drinking, including private water taps................................................................................. 33 Table 9. Changes in access to basic services by province, 2009-2019........................................ 34 Table 10. Nearly 90 percent of the urban population growth occurred in Honiara Urban Area.................................................................................................... 38 Table 11. More people left than arrived in places except for Honiara and Guadalcanal........... 40 Table 12. Labor force participation and employment rates increased, and so did unemployment rates................................................................................................... 46 Table 13. People reported that CDFs contributed to housing and solar energy.......................... 49 List of Boxes Box 1. Multidimensional poverty measures........................................................................................... 15 Box 2. Potential population undercount in the 2009 Population Census................................ 16 Box 3. Estimating changes in the number of multidimensionally poor individuals between 2009 and 2019............................................................................... 28 Box 4. Definition of deprivation in water, sanitation, and electricity........................................... 34 6 and Spatial Disparities Multidimensional Poverty  inSolomon Islands List of Abbreviations ADB Asian Development Bank CDF Constituency Development Fund CLFG Commonwealth Local Government Forum CRISP Community Resilience to Climate and Disaster Risk in Solomon Islands Project FSM Federated States of Micronesia GDP Gross Domestic Product HCI Human Capital Index HFPS high-frequency phone surveys HIES Household Income and Expenditure Survey ILO International Labor Organization LMIC lower-middle-income country MHMS Ministry of Health and Medical Services MP Member of Parliament MPI Multidimensional Poverty Index NGO non-governmental organization OECD Organization for Economic Co-operation OPHI Oxford Poverty and Human Development Initiative PCDF Provincial Capacity Development Fund PIC Pacific Island Country PPP purchasing power parity RDP Rural Development Program SBD Solomon Islands Dollar SDG Sustainable Development Goals SI Solomon Islands SIART Solomon Islands Agriculture and Rural Transformation Project SIG Solomon Islands Government SINSO Solomon Islands National Statistics Office SIWA Solomon Islands Water Authority UNDP United Nations Development Programme WASH Water, Sanitation, and Hygiene WB World Bank 7 Executive Summary Solomon Islands is a lower-middle-income country with poverty levels in Solomon Islands with other countries, low human development, constrained by economic examines changes in multidimensional poverty during the geography and fragility challenges. Solomon Islands’ per decade with an emphasis on spatial aspects, and discusses capita GDP was US$2,211 (2017 purchasing power parity) in what needs to be done to accelerate poverty reduction and 2022, well below the average of Pacific Island Countries improve spatial equality in the future. (PICs). Low incomes are also reflected in the country’s low This report highlights three key findings. Firstly, the Human Development Index, which ranked 156th out of 193 reduction in multidimensional poverty in Solomon countries in 2022. The country’s economic growth has been Islands has been slow, with significant spatial disparities slowing down during the past decade or so, with the annual between urban and rural areas, as well as across per capita GDP growth rate averaging around 1.5 percent provinces. The multidimensional poverty level remains between 2009 and 2019. As a remote archipelago with a considerably higher than in other comparable countries. small population of 721,000 dispersed over 90 inhabited Secondly, to effectively reduce concentrated rural islands across 1.6 million square kilometers of ocean, poverty, rural investments need to be more efficient and Solomon Islands faces a challenge to realize economies of impactful. Although substantial investments in housing and scale and provide quality public services and infrastructure. water access have contributed to rural poverty reduction, Compounding such economic geography challenges are accelerating this progress requires enhanced efficiency and limited state capacity, institutional and social fragility, effectiveness. Lastly, strategic urban planning and and political economy dynamics that constrain the improved service delivery are essential to leverage rapid implementation of effective public policies.1 urbanization for poverty reduction. The increasing This report attempts to provide a snapshot of the living migration to Honiara has already put a strain on the city’s standards of the population from multidimensional infrastructure and basic services. poverty perspectives and assess the progress made The multidimensional poverty level in Solomon Islands between 2009 and 2019. Given the outdated official is higher than in other comparator countries. About 35 monetary poverty data in the 2012/13 Household Income percent of the population in Solomon Islands is estimated to and Expenditure Survey (HIES), this report utilizes more face multidimensional poverty in 2019; in other words, recently collected data from the Population and Housing one-in-three Solomon Islanders are multidimensionally poor. Census 2019, and compares it to data in the previous one in This level is far higher than other PICs—such as Kiribati 2009, to investigate non-monetary well-being of Solomon (19.8), Samoa (6.3), Fiji (1.5), and Tonga (0.9)—and the Islanders. Multidimensional poverty, measured by applying average of lower-middle-income countries (24.3) (Figure a global methodology to the population censuses, is a ES1). Among the three dimensions constituting composite index of the nine indicators capturing three multidimensional poverty, the deprivation in education and dimensions of welfare: health, education, and living living standards is particularly severe in Solomon Islands. For standards. Child mortality is the indicator of health; school example, only 17 percent of the population has access to enrollment and educational attainment are the indicators for electricity in Solomon Islands, which is significantly lower education; and housing, drinking water, sanitation, electricity, than nearly 90 percent on average among countries in the cooking fuel, and asset ownership are the indicators for living structural peer group, such as Comoros, Kiribati, Federal standards.2 The report compares the multidimensional States of Micronesia, Samoa, Timor-Leste, and Vanuatu. See World Bank (2017). 1 See UNDP and OPHI (2023) for the global approach to measuring multidimensional poverty. 2 8 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Figure ES1.  Multidimensional poverty rates Multidimensional poverty declined only moderately in Solomon Islands and other countries between 2009 and 2019, with the catch up of initially poorer wards. The incidence of multidimensional poverty— 40 or the percentage of the population in multidimensional Multidimensional poverty rate (%) 35 poverty—decreased from 43.5 percent in 2009 to 35.4 30 percent in 2019 (Figure ES2). Multidimensional poverty in 25 rural areas declined relatively rapidly, from 50 percent in 20 2009 to 43 percent in 2019. The multidimensional poverty 15 level remained low in urban areas, both in Honiara and other 10 urban areas. However, due to rapid urban population growth, 5 the urban share of the multidimensionally poor population 0 increased from 8 to 13 percent during the period. Several provinces, such as Malaita, Central, and Choiseul, experienced ) age ) ) ) ) 19 19 20 21 19 relatively robust multidimensional poverty reduction over the (20 r /20 /20 (20 (20 ave 18 19 nds Fiji ga decade. In contrast, the multidimensional poverty level in IC 20 (20 Ton LM sla ti ( rural Guadalcanal remains high, after only a moderate nI oa iba Sam mo reduction from 57 to 53 percent. At the ward level, 131 out of Kir o Sol 183 wards experienced multidimensional poverty reduction, Note: The multidimensional poverty rate indicates the percentage of the with a greater degree of reduction achieved among initially population in multidimensional poverty in a country. poorer wards, which reduced spatial inequality to some extent. Source: Staff calculations based on Population and Housing Census 2019 for Solomon Islands and UNDP and OPHI (2023) for other countries. With almost all multidimensionally poor Solomon Islanders being deprived in living standards, Multidimensional poverty is more prevalent in rural improvements in housing quality and access to basic areas, particularly in Guadalcanal Province. There is a services, primarily drinking water, drove a reduction in stark urban-rural contrast in non-monetary well-being, as multidimensional poverty. Nearly 86 percent of the poor nearly 43 percent of the rural population is multidimensionally are deprived in more than two dimensions—mainly education poor in 2019, in contrast to only 16 percent in urban areas and living standard dimensions (Panel A in Figure ES3). (Figure ES2). The most populous capital island, Guadalcanal, Between 2009 and 2019, access to safe drinking water contains both areas with the lowest multidimensional poverty improved in most wards, with the proportion of the population rate (Honiara City Council, 12 percent) and the highest in water deprivation declining from 29 to 17 percent (Panel poverty rate (rural Guadalcanal, 53 percent). Makira-Ulawa B). Approximately 77 percent of the rural population and 97 (50 percent) and Malaita (43 percent) provinces have percent of the urban population now enjoy improved access relatively high poverty rates as well. Of the approximately to water. By contrast, sanitation deprivation remains high, 249,100 people who were living in multidimensional poverty with certain wards, particularly in Honiara and Guadalcanal, as of 2019, about 87 percent were living in rural areas and experiencing a significant influx of migrants and worsening nearly two-thirds of them were residing in either Honiara, conditions. The improvement in housing deprivation was Guadalcanal, or Malaita provinces. relatively strong, with a nearly 20 percentage point reduction. Figure ES2.  Multidimensional poverty rates by location, 2009 and 2019 60 50 40 30 20 10 0 National Urban Rural Choiseul Western Isabel Central Rennell and Bellona Guadalcanal Malaita Tamotu Honiara Makira-Ulawa Urban/rural Province 2019 2009 Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 9 Education outcomes also lag behind other countries—for Figure ES3.  Changes in multidimensional poverty example, only 10 percent of working-age adults have dimensions and indicators, 2009 and 2019 completed secondary education—and there was no apparent improvement during the decade. (A) Three multidimensional poverty dimensions Health The diagnostic indicates that Solomon Islands needs to (7.4) accelerate reductions in multidimensional poverty and spatial inequality, but the country’s economic geography 0.3 presents significant challenges that necessitate a 0.2 strategic approach. The small and dispersed population 0.02 hinders the realization of economies of scale and increases 4.5 the cost for the government to provide quality public services and infrastructure. Additionally, the small domestic economy, 2.5 characterized by internal dispersion and remoteness from large export markets, limits private sector development and Education (26.1) international trade. These economic geography challenges are further exacerbated by limited transport and digital 23.4 connectivity, which constrain economic activity and people’s 4.6 access to markets and basic services. Meanwhile, rapid Living urbanization, driven primarily by large-scale migration to standards Honiara, has led to the urban population increasing by nearly (34.9) 6 percent annually from 110,000 in 2009 to nearly 200,000 in 2019. This rapid growth places significant pressure on the demand for urban services and infrastructure. Given these (B) Proportion of deprived population in each multidimensional poverty indicator geographic challenges, the government, with its limited 93 resources, must make strategic investments to accelerate 87 83 84 86 83 2009 2019 the reduction of rural and urban poverty and improve spatial 66 70 61 equality. 48 The persistent concentration of poverty in rural areas 29 19 21 14 12 17 highlights the need for more efficient and effective 10 8 investments. Substantial government spending has been directed towards components of multidimensional poverty Child mortality School enrollment Educational attainment Housing Drinking water Sanitation Electricity Cooking fuel Asset ownership that have since shown the most improvement. In 2016, Constituency Development Funds (CDFs) constituted nearly 9 percent of total government expenditure, with a significant portion allocated to housing improvements. Additionally, donor-funded government projects made significant Health Education Living standards investments in household water access. The total investment Note: The numbers in the figure add up to 35.4 percent, corresponding to in water over the course of the Rural Development Program the proportion of the multidimensional poor (not precisely proportional to and the Community Resilience to Climate and Disaster Risk the areas) (Panel A). Project (SBD 250 million) was roughly equivalent to the Source: Staff calculations based on Population and Housing Censuses annual CDF budget. However, the large expenditures 2009 and 2019. on housing and water access, which potentially contributed to reductions in multidimensional poverty, have strained as Temotu and Western to a lesser extent. This analysis could government budgets. Without a mechanism to align CDF be used to geographically target areas with high poverty. spending with development strategies, the sizable allocation Finally, the government must maintain service delivery of CDFs to housing could limit government spending in other amidst rapid urban population growth to harness critical areas.3 The diagnostic of this report also highlights urbanization for poverty reduction. The population in that multidimensional poverty is particularly severe and Honiara City Council and adjacent urban wards in Guadalcanal concentrated in specific wards and provinces. Among the Province surged to 170,000 in 2019, accounting for 23.5 183 administrative wards constituting the country, those percent of the national population. Given the significant with significantly high multidimensional poverty rates disparity in access to economic opportunities and basic (exceeding 50 percent) are predominantly concentrated services and infrastructure, migrants from rural areas in Guadalcanal, Malaita, and Makira-Ulawa provinces, as well 3  ccording to the World Bank Public Expenditure Review (2022), projects funded under the CDF mechanism did not follow a systematic A or standardized appraisal or selection process, at least until the recent reform. The recently approved CDF Act 2023 introduced several reforms on the governance system. 10 and Spatial Disparities Multidimensional Poverty  inSolomon Islands experience multidimensional poverty less frequently in urban population with improved access dropping from 81 percent in areas. However, despite the low incidence of multidimensional 2009 to 77 percent in 2019 in Honiara and further poverty, the number of individuals living in multidimensional deterioration observed in surrounding urban Guadalcanal poverty has rapidly increased in Honiara. The growing urban area. This poses potential public health risks in crowded population is straining service delivery and causing other areas, especially considering the intensified natural hazard potential problems in the city. For example, urban expansion risks due to climate change. If this situation remains over the past three decades has been mostly unplanned and unchanged and more people migrate to Honiara, congestion informal. Existing plans are outdated, not risk-informed, and costs will escalate, outweighing the benefits of urbanization. enforcement capacity is weak. In tandem, access to sanitation has already begun to deteriorate, with the proportion of the 11 Chapter 1. Introduction 1.1 Context Compounding these disadvantages are limited state capacity, institutional and social fragility, and political Solomon Islands is a lower-middle-income country with economy dynamics, which constrain the implementation low human development. The country’s per capita GDP of effective public policies. The quality of public sector was US$2,211 (purchasing power parity [PPP]) in 2022, well management and institutions in Solomon Islands is lower below the average of Pacific Island Countries (PICs). Low than most structural and aspirational peers. In part, this incomes are also reflected in the country’s low Human reflects the limited technical capacity of the bureaucracy, Development Index, which ranked 156th out of 193 countries though it also reflects the prevailing political economy in 2022. This compares to a ranking of 104th for Fiji, an dynamics that elevate interest groups over bureaucratic aspirational peer of Solomon Islands. Notwithstanding some strengthening. The combination of limited state reach, improvement since the turn of the century, the Human fragmented, localized identities, and the value of resource Capital Index (HCI) for Solomon Islands remains in the rents complicate the functioning of political markets. These bottom 20 percent of countries worldwide: a child born today interconnected development challenges drive the country’s is only expected to achieve 42 percent of their full potential. institutional and social fragility, contributing to violent events According to the most recently collected data in the 2012/13 and episodes of civil unrest—notably the Tensions, a Household Income and Expenditure Survey (HIES), 61 low-scale civil war between 1999 and 2003. The root causes percent of the population was considered poor based on the of the conflict (land disputes triggered by inter-island lower-middle-income poverty line of US$3.65 (2017 PPP migration, friction between traditional and non-traditional USD per person per day). authority, and access to resources and economic A challenging economic geography plays a crucial role in opportunities) persist today. shaping these outcomes. The country is the fifth most Economic growth in Solomon Islands has been slowing remote country in the world from large markets, being more down during the last decade or so. Solomon Islands remote than the average of the PICs.4 The population of achieved only moderate economic growth between 2009 721,000 is dispersed over an estimated 90 inhabited islands, and 2019, with fluctuations. The annual GDP growth rate stretching across 1.6 million squared kilometers of ocean. averaged around 4 percent during the decade (Figure 1). The Around 75 percent of the population lives in rural areas, economic growth was influenced by various factors such as mainly in smaller villages. The small and dispersed population the opening and closing of the Gold Ridge mine, El Nino and makes it challenging to realize economies of scale and makes tropical cyclones, and a peak in log production.5 Once it expensive for the government to provide quality public population growth is factored in, the average annual growth services and infrastructure. Additionally, a small domestic rate of GDP per capita was only 1.5 percent between 2009 economy with internal dispersion and remoteness from large and 2019, with no growth in several years. More recently, the export markets limits private sector development and COVID-19 pandemic halted economic growth in 2020 and international trade. Economic geography challenges are 2021. amplified by limited transport and digital connectivity, which further constrains economic activity. World Bank (2021). 4 World Bank (2024). 5 12 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Figure 1.  Solomon Islands achieved only moderate economic growth from 2009 to 2019 Annual GDP growth rate (%), 2009-19 Period for analysis 10 8 Annual growth rate (%) 6 4 2 0 -2 -4 -6 -8 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Year Annual GDP growth rate (%) Annual GDP per capita growth rate (%) Source: World Development Indicators. It is unknown to what extent such economic growth has development can be expected to intensify over time. With been translated into poverty reduction due to the lack of growth opportunities largely confined to natural resource- sufficient data. According to the most recent poverty based industries and to the urban services economy, growth assessment conducted by the Solomon Islands National will inevitably be concentrated around locations of the Statistics Office (SINSO) and the World Bank in 2015, it was underlying resources and in urban areas. At the same time, found that 12.7 percent of the population was poor based on the lower unit costs of some utilities and infrastructure in the 2012/13 HIES.6 This monetary poverty, measured based urban areas lead to pronounced differences in service on per adult-equivalent consumption expenditures and the coverage and quality along urban/rural lines. However, inequities national poverty line, is already outdated, creating serious could remain due to limits on mobility posed by inadequate knowledge gaps. 7 connective infrastructure, land systems, and island-scale ethnic divisions—and these are a source of fragility. Strengthening the foundation of well-being is a key pathway to development in Solomon Islands. The World The key questions are where and how multidimensional Bank’s Systematic Country Diagnostic identifies several poverty prevails in 2019; how much multidimensional priorities to improve the well-being of Solomon Islanders, poverty declined since 2009; what contributed to such such as improving people’s access to water, sanitation, reductions; and what needs to be done to accelerate electricity, education, and health services.8 These are poverty reduction going forward? To address these indispensable components of people’s well-being as they questions, this report aims to provide up-to-date and spatially directly determine current living conditions. However, access to granular diagnostics and analyses of the well-being of the these essential services also influences monetary poverty. This people, specifically focusing on multidimensional poverty, by is because they affect the productive capacity of workers—both utilizing the latest series of population censuses. The two current and future. For example, poor access to water and rounds of Population and Housing Censuses collected by sanitation services, not to mention education and health SINSO in 2009 and 2019 present an excellent opportunity services, will lower the future productive capacity of children. to analyze non-monetary poverty over time and across different regions. Compared to establishing comparability of Managing uneven development to reduce spatial income/consumption and monetary poverty measures, which inequality is also crucial. The Systematic Country require consistency in survey design, inflation adjustment, Diagnostic stressed the need to manage uneven development and other factors, analyzing changes in non-monetary as one of the pillars to achieve poverty reduction and shared poverty is relatively less demanding. Additionally, the prosperity.9 It argues that spatial patterns of uneven population census allows for spatial analysis by disaggregating 6 SINSO and World Bank (2015). The poverty lines vary across provinces due to their cost-of-living differences, ranging from SBD 3,569 in Temotu to SBD 10,334 in Honiara in 2012/13 prices. 7 SINSO is currently conducting the 2024/25 HIES, and this upcoming survey will provide long-awaited, updated insights into the well-being of Solomon Islanders. 8 World Bank (2017) formulates the following three pillars: strengthening the foundations of well-being, achieving inclusive and sustainable growth, and managing uneven development, with economic geography and state fragility as the two cross-cutting themes. 9 World Bank (2017). 13 the data geographically into smaller units such as wards.10 experiences situational poverty.12 Multidimensional poverty This report will diagnose the prevailing multidimensional poverty takes a broader approach, considering multiple factors of and spatial inequality, examine their trends, and, to the extent household well-being. Alongside monetary poverty, possible, identify the underlying drivers and future risks.11 multidimensional poverty serves as an indicator for the Sustainable Development Goals (SDGs).13 The remainder of this introductory chapter briefly introduces the methodology of measuring multidimensional poverty and This report follows a global methodology and measures the three key potential drivers for poverty reduction. multidimensional poverty in three dimensions: health, education, and living standards.14 Each dimension is 1.2 Measuring multidimensional poverty equally weighted at 33.3 percent or one-third (Figure 2), and they are further divided into nine indicators that emphasize The multidimensional poverty approach complements the attainment of a basic quality of life. Each person is monetary measures to assess household well-being. assigned a deprivation score based on their household’s level While income or expenditure indicators are important for of deprivation, with the health dimension having one understanding living standards, monetary poverty does not indicator, the education dimension comprising two indicators always reflect the full extent of deprivation, as there can be weighted as one-sixth each, and six indicators capturing various reasons why a household earns less income or the standard of living, each weighted one-eighteenth. The Figure 2.  Multidimensional poverty consists of health, education, and living standard dimensions Source: Authors.  10 Disaggregating monetary poverty to the ward level requires the application of a statistical method (small area estimation) to HIES and population census data. Ward-level poverty maps were published by SINSO and World Bank (2017). ’Poverty’ refers to multidimensional poverty in this report unless otherwise noted. 11 World Bank (2018). 12 “Proportion of men, women, and children of all ages living in poverty in all its dimensions according to national definitions” (SDG 1.2.2). 13 UNDP and OPHI (2023). 14 14 and Spatial Disparities Multidimensional Poverty  inSolomon Islands household deprivation score is calculated by aggregating the multidimensional poverty rates. Conversely, wards in Malaita scores for each indicator. A threshold of one-third is used to Province tend to have higher multidimensional poverty rates distinguish between the multidimensionally poor and the than monetary poverty rates. Both monetary and non-poor. Therefore, if the household deprivation score is multidimensional poverty rates are very high in Guadalcanal one-third (33 percent) or higher, everyone in that household Province. is categorized as multidimensionally poor. Despite some modifications made due to the limited availability of information in Solomon Islands’ census data, the Box 1. Multidimensional poverty measures multidimensional poverty numbers in this report are, with Individuals in a household are considered some caveats, broadly comparable to the global index.15 multidimensionally poor if their deprivation scores are This report examines multidimensional poverty in terms lower than one-third. The deprivation score is of incidence, headcount, and intensity (Box 1). Poverty calculated for each household based on the weighted incidence indicates the proportion of the population that is share of multidimensional poverty indicators they are multidimensionally poor. Poverty headcount refers to the deprived of (Figure 2). For example, the deprivation total number of individuals living in poverty. The score for a household that is deprived in child mortality Multidimensional Poverty Index (MPI) is a combination of (0.333), school enrollment (0.166), and housing (0.055) poverty incidence and the intensity of poverty, which reflects is calculated to be 0.55. Therefore, individuals in this the extent of deprivation calculated as the average household are considered multidimensionally poor. deprivation scores among the poor population. The calculated The multidimensional poverty measures calculated in multidimensional poverty numbers—in incidence, headcount, this report include the following: and MPI—are comparable across subnational areas over time and, with some caveats, between countries. Incidence (headcount ratio): The percentage of the •  population in multidimensional poverty, ranging This report disaggregates multidimensional poverty at from 0 to 100. several geographic levels to better analyze its spatial patterns. In addition to the national level, multidimensional Headcount: The population in multidimensional •  poverty indicators are reported at the province level, which poverty, ranging from 0 to the total population. includes the Honiara City Council among nine other provinces. Intensity: The average deprivation scores among •  Multidimensional poverty is also disaggregated into urban the multidimensionally poor population, ranging and rural areas. This report further distinguishes urban areas from one-third to one. into Honiara Urban Area (which consists of Honiara City Council and urban areas in adjacent wards in Guadalcanal MPI: The product of multidimensional poverty •  Province) and other urban areas (see Annex A2 for more incidence and intensity, ranging from zero to one. explanations). The finest geographic level to measure Multidimensional poverty with a deprivation score of multidimensional poverty is 183 wards, which remained one-third as the threshold is referred to as moderate unchanged between 2009 and 2019.16 multidimensional poverty, as opposed to severe Multidimensional poverty does not substitute monetary multidimensional poverty with 0.5 as the threshold. poverty, although there is a reasonable correlation The multidimensional poverty methodology used in between them at the ward level. Many of the this report broadly follows the global methodology. multidimensional poverty components are largely influenced Due to limited data availability, some modifications by public investment; for example, household access to water were made, including the increased weights to child and sanitation depends on, among others, the existence of mortality because of a missing nutrition indicator and infrastructure networks. On the other hand, housing and the use of the improved access criteria for drinking asset ownership are more related to household income levels, water because of missing information about the as they are bought in the market. Nevertheless, monetary proximity to water sources. and multidimensional poverty are overall correlated; wards with higher multidimensional poverty incidence tend to have See Annex A1 for more detailed explanations of the higher monetary poverty incidence (Figure 3). Meanwhile, methodology. there are some important patterns to note. Wards in Honiara tend to have higher monetary poverty rates than 15 For more detailed explanations, please refer to Annex A1. 16  he number of wards in each province has been determined based on population size. Elected representatives from each ward constitute T Provincial Assemblies and Honiara City Council (CLFG 2018). In 2019, the ward population ranged from 214 to 33,056, with the average being 3,939. 15 Figure 3.  Multidimensional and monetary poverty are the national population growth rate (2.6 percent per year), positively correlated at the ward level resulting in an increase in the urban population’s share from Multidimensional poverty rates (2009) and 19.8 percent in 2009 to 27.6 percent in 2019. monetary poverty rates (2009) at 183 wards In particular, the population in Honiara Urban Area, which includes the Honiara City Council and the 60 surrounding urban areas in Guadalcanal Province, has grown fast during the period. Its population accounts for about 85 percent of the country’s urban population.17 The 40 population growth in Honiara Urban Area represents approximately 60 percent of the overall national population increase between 2009 and 2019, indicating a significant 20 migration to Guadalcanal Island. It is worth noting that the population growth in all other provinces has been slower compared to the national average. 0 0 20 40 60 80 100 Box 2. Potential population undercount in the Multidimensional poverty (2009) 2009 Population Census Note: X-axis indicates multidimensional Honiara Guadalcanal poverty rate based Malaita on the 2009 Others The 2019 Population Census report (SINSO 2023) Population Census; y-axis indicates monetary poverty rate based on the 2009 Population Census and the 2013 HIES. R-squared = 0.11. Correlation points to a potential population undercount in the coefficient = 0.34. 2009 Population Census. Source: Staff calculations based on Population Census 2009 and SINSO & WB (2017). For example, the 2009 Census recorded 76,200 children aged 0 to 4 at the time of the census. Since no 1.3 Potential key drivers for poverty reduction children of this age would have been added or subtracted during the intervening 10 years, a certain This report examines three factors that have potentially percentage would have died and so the resulting 10-to contributed to reductions in multidimensional poverty 14-year-olds in 2019 should have been about 75,000. and inequality: internal migration and urbanization; Instead, the 2019 Census recorded 84,400. Some economic geography and jobs; and government 8,200 appeared during the decade that could have investments. Solomon Islands experienced unprecedented been missed in the 2009 Census (8.3 percent rapid migration toward Honiara, which is one of the fastest undercount) or misreported. growing cities in the world. Urbanization is often considered as a potential key driver for economic growth and poverty To deal with this potential population undercount, the reduction, mainly by facilitating agglomeration economies SINSO report applies a sub-national level adjustment and the shift of economic structure toward more productive to the 2009 population in Guadalcanal, Malaita, and sectors. It is critically important to examine whether that is Honiara. This report uses the adjusted population the case in a small island country like Solomon Islands. when looking at population trends, as in Table 1. Assessing the contribution of public sector investment, However, when comparing population/household which constitutes a large share of investment in the country, characteristics between 2009 and 2019, this report to poverty reduction is also crucial. This section briefly relies on the unadjusted 2009 population. This is introduces the context of these factors. because the characteristics of the undercounted Internal migration and urbanization population in 2009 are unknown, and the SINSO’s adjustment focuses on only the population differences. During the past ten years there has been a significant Whenever appropriate, this report compares the movement of people within Solomon Islands towards percentage of certain population/household the capital island. According to the two sets of Population characteristics (including poverty) between 2009 and Censuses, the overall population of the country has grown by 2019. It remains cautious when analyzing the changes 29 percent, from approximately 559,000 to 721,000, in the absolute number of the population in certain between 2009 and 2019 (as shown in Table 1). While this characteristics (for example, the changes in the population growth might be overestimated given the number of multidimensionally poor from 2009 to potential population undercount in the 2009 Census (Box 2), 2019). urban growth must have been outstanding. The growth rate of the urban population (5.9 percent per year) has exceeded  17 Honiara Urban Area is a concept used in SINSO (2023), though it is not an officially designated area. It includes Honiara City Council and urban areas in Tandai and Malango wards in Guadalcanal Province (see Annex A2). 16 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Table 1.  Solomon Islands experienced rapid urbanization, particularly around Honiara, between 2009 and 2019 Population by Province, 2009-2019 (thousands) 2009 2019 Difference Annual growth rate National 558.5 721.0 29.1% 2.6% Urban 110.5 199.1 80.3% 5.9%   Honiara Urban Area 91.3 169.7 86.0% 6.2%   Other urban 19.2 29.4 53.2% 4.3% Rural 448.0 521.8 16.5% 1.5%   Rural Guadalcanal 89.7 113.9 26.9% 2.4%   Other rural 358.3 407.9 13.9% 1.3% Choiseul 26.4 30.8 16.7% 1.5% Western 76.6 94.1 22.7% 2.0% Isabel 26.2 31.4 20.1% 1.8% Central 26.1 30.3 16.4% 1.5% Rennell and Bellona 3.0 4.1 34.8% 3.0% Guadalcanal 107.1 154.0 43.8% 3.6% Malaita 157.4 172.7 9.7% 0.9% Makira-Ulawa 40.4 51.6 27.6% 2.4% Temotu 21.4 22.3 4.5% 0.4% Honiara 73.9 129.6 75.3% 5.6% Note: Honiara Urban Area includes Honiara City Council and the urban areas of Tandai and Malango wards in Guadalcanal Province (Annex A2). The population in 2009 is adjusted to account for the potential population undercount (Box 2). Source: SINSO (2023). The movement of rural individuals and households to and amenities in urban areas may also be easier and more urban areas can directly influence poverty levels in both affordable given the low unit costs of service provisions due rural and urban areas through several channels. People to higher density and network effects. However, these direct migrate to urban areas for different reasons, such as expected effects may not exist if migration is highly selective, meaning higher earnings, family reunion and separation, and drought that those who leave rural areas are not poor even before and other shocks.18 One obvious scenario is when rural migration, or they would have escaped poverty even without poverty decreases simply because poor populations migrate migrating, or migrants remain poor in urban areas.20,21 On the from rural to urban areas. If these migrants successfully other hand, other migrants who are forced to leave rural escape poverty in urban areas, it will also lead to a decrease in villages due to conflicts and natural disasters often struggle national poverty levels.19 Obtaining better access to services in urban areas.22  18 A classic economic theory posits that people migrate to urban areas as long as their expected urban income, taking account of the likelihood of unemployment, exceeds rural income (Harris & Todaro 1970).  19 Examples of empirical evidence about the direct effect of migration are Beegle, de Weerdt, and Dercon (2011) for Tanzania and Bryan, Chowdhury, and Mobarak (2014) for Bangladesh.  20 For example, Hicks et al. (2017) find small productivity gaps between urban and rural workers after accounting for the differences in their characteris- tics in Indonesia and Kenya. See Lagakos (2020) for a review.  21 It is common for households with relatively high financial, human, and social capital to migrate from rural to urban areas. See a review by Lucas (2016).  22 This kind of forced migration is often referred to as ’pushed’ migration, as opposed to ’pulled’ migration. 17 Indirect effects of rural-to-urban migration can also consumption as employed, the share of workers in the play a crucial role in poverty reduction. Rural households agriculture sector slightly increased from 42 percent in 2009 can escape poverty through transfers from migrants in urban to 50 percent in 2019. Even when considering those who areas. These transfers can benefit non-recipient rural work for their own consumption as employed, nearly seven households as well.23 Additionally, the development of towns out of ten workers are still engaged in agriculture, while a resulting from migration can have spillover effects on quarter are in the service sector. In either case, the share of surrounding rural households.24 Migration and urbanization workers in the industry sector remains small. Stable sectoral can also lead to productivity gains through agglomeration shares in employment coincide with the lack of a sectoral effects and the reallocation of workers to more productive shift in GDP over the same period.26 sectors. However, migration can have adverse indirect effects Low-productivity jobs prevail across economic sectors. under certain conditions. Excessive migration to urban areas Most agricultural production is small-scale and largely can lead to negative spatial externalities, such as congestion, uncommercialized, providing the rural population with food air pollution, crime, and the proliferation of informal security and livelihoods. A relatively small, commercialized settlements. Insufficient job creation due to regulatory agriculture sector sits alongside subsistence production, barriers or skills mismatch can also keep migrants in poverty which is dominated by an export-oriented forestry industry, in cities and towns.25 but it also includes the harvesting of plantation crops as well Economic geography and jobs as a well-established oceanic fishing sector. The industry sector is fairly limited, with mining, food processing, and Jobs are another potential driving factor for poverty construction as the main industrial activities. Urban-based reduction; however, neither a structural shift toward government workers support the service-based economy— non-agricultural sectors nor strong job growth occurred comprising 57.1 percent of GDP—mainly in urban areas like at the national level. There is no sign of an employment Honiara. Retail, transport, financial intermediation, public shift from agriculture to industry and service sectors between administration, and real estate are the largest service sub- 2009 and 2019, according to the Population Census data in sectors. Tourism is currently limited despite an abundance of 2009 and 2019 (Table 2). Based on the ILO definition, which exquisite natural capital. does not consider individuals who work for their own Table 2.  Neither strong job growth nor a structural shift took place between 2009 and 2019 Employment trends, 2009-19 (percent) ILO definition SINSO approach 2009 2019 Diff. (pt) 2009 2019 Diff. (pt) Economic sectors (% of working-age workers) Agriculture 41.4 49.8 8.4 70.5 64.0 -6.5 Industry 13.9 10.0 -3.9 6.9 6.8 -0.1 Services 44.7 40.2 -4.5 22.7 29.2 6.5 Labor force status (% of working-age population) Active 28.9 36.5 7.6 57.1 54.8 -2.3  Employed 27.4 31.5 4.1 55.6 49.8 -5.8  Unemployed 1.5 5.0 3.6 1.5 5.0 3.5 Inactive 71.1 63.5 -7.6 42.9 45.2 2.3 Note: Only the working-age population is counted. See Annex A3 for ILO and SINSO approaches. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. For example, the spillover effects of cash transfers in rural Kenya are shown by Egger et al. (2022). 23 The role of secondary towns in poverty reduction has been illustrated in, for example, Gibson et al. (2017) and Christiaensen and Todo (2014). 24 See a review of empirical studies by Selod and Shilpi (2021). 25 World Bank (2024). 26 18 and Spatial Disparities Multidimensional Poverty  inSolomon Islands The persistent predominance of agricultural workers in The share of agricultural workers in Solomon Islands (68 Solomon Islands is not surprising, considering the percent in 2019) is similar to other countries with comparable country’s low per capita GDP and urbanization levels. per capita GDP levels. Other Pacific Island countries such as While the lack of a structural shift may be due to the small Fiji, Samoa, and Tonga have a lower share of workers in the population size, the island economy structure, and low human agriculture sector. Similarly, more urbanized countries tend capital, it is also true that the income level of Solomon to have less agricultural workers (Panel B). The agricultural Islands is simply too low. The worker share of the agricultural worker share of Solomon Islands is at the predictable level sector is negatively correlated with the GDP per capita level based on its urban population share. across low- and middle-income countries (Panel A in Figure 4). Figure 4.  Low levels of GDP per capita and urbanization are associated with a high share of agricultural workers Share of workers in agriculture in low- and middle-income countries, 2019 (A) GDP per capita (B) Urban population share 95% CI 80 95% CI 80 Fitted values % Employment in Agriculture, 2019 Fitted values Other countries Other countries Pacific countries 60 Pacific countries 60 Solomon Islands Vanuatu Solomon Islands Vanuatu 40 Solomon Solomon Tonga Fiji 40 Tonga Fiji 20 Samoa Samoa 0 20 -20 0 0 10000 20000 30000 20 40 60 80 100 GDP per capita, 2019 % Urban population share, 2019 Note: 134 low- and middle-income countries are analyzed. Employment is defined as persons of working age who were engaged in any activity to produce goods or provide services for pay or profit. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019 and World Development Indicators. The recent World Bank Country Economic Memorandum Given the lack of job growth or structural shift at the highlights that a fundamental underlying driver of national level, this report investigates the associations economic challenges in Solomon Islands is unfavorable between the changes in subnational employment economic geography.27 It arises from a three-dimensional patterns and multidimensional poverty reduction. Using challenge of remoteness from large export markets, internal the Population Censuses 2009 and 2019, this report dispersion of the population, and a lack of population density. examines whether labor force participation rates increased Solomon Islands is the sixth most remote country in the and unemployment declined during the period and how these world from large markets, as measured by GDP-weighted changes, if any, were related to poverty reduction. Poverty distance—being more remote than the average of PICs and reduction can be facilitated by a structural shift from low- lower-middle-income countries. In addition, the country productivity jobs (namely farm employment) to high- exhibits a high degree of internal dispersion and division. This productivity jobs (namely off-farm employment) and unfavorable economic geography raises production costs, enhanced productivity within each economic sector.28 This constrains the reach of infrastructure, and inhibits public report investigates these changes based on the information sector service delivery as well as private sector development. available in the two series of the Population Censuses. Private sector growth opportunities are largely confined to Considering the geographic constraints faced by the country, industries that can generate rents sufficient to outweigh the spatial aspects in the linkage between jobs and poverty are higher costs of production that result from the small size of particularly examined with the ward-level data. the domestic market, high costs of transport for all traded items, and susceptibility to natural disasters. These are predominantly natural resource-based industries or industries catering to niche markets.  27 World Bank (2024).  28 However, it is important to note that the economic growth was primarily driven by consumption rather than investment, and by the growth of the labor force rather than productivity growth (World Bank 2024). 19 Government policies and investments establishment in the early 1990s, there has been relentless growth in the public resources allocated to CDFs and in their It is also possible that the government played a key role importance. By providing public resources to national in multidimensional poverty reduction. This report takes government Members of Parliaments (MPs) to spend at their stock of the Solomon Islands Government (SIG) project discretion, CDFs bypass the provincial government system. financed by the World Bank and other development partners The size of budget allocation to CDF has substantially grown that involved the provision of household- or community-level over time, and their budget now exceeds that for provincial basic services—namely, water, sanitation, and electricity grants by a factor of six (Figure 5). In 2016, CDFs accounted between 2009 and 2019. As these are included as for 8.8 percent of total expenditure. Though equal in value for multidimensional poverty indicators, their improvements can each constituency, since the population varies so much directly result in reductions in multidimensional poverty. across constituencies, there are large variations in CDFs per After screening all the major donor-funded investment capita. Though there is no systematic evidence of their use, projects, this report identifies the following projects allocating CDFs fund anything from setting up constituency businesses the most significant amount of investments toward improving to providing households with solar panels or paying their household-level access to services across the country school fees, to distributing bags of rice at election time. While between 2009 and 2019: the Rural Development Program they serve as a mechanism to redistribute public resources (RDP) I and II and the Community Resilience to Climate and across the archipelago, including directly providing some Disaster Risk in Solomon Islands Project (CRISP). essential services to constituents, their operation is widely Constituency Development Funds (CDFs) represent a recognized as mainly patronage-based. The recently approved major means for the redistribution of public resources CDF Act 2023 introduced several reforms on the governance on a geographic basis. This report assesses the CDF’s system. Using the information available in the 2019 Census potential contribution to poverty reduction given its large data—albeit limited—this report explores how CDFs budgetary scale and nation-wide coverage. Following their contributed to multidimensional poverty reduction. Figure 5.  CDF Budget has increased significantly since 2009 CDF budget allocated to Members of Parliament (nominal SBD millions and percent of GDP), 2000-2022 Source: World Bank (2022a). 20 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Chapter 2. Poverty and Spatial Inequality Measuring multidimensional poverty using two series of Population Censuses at geographically granular levels, this chapter investigates various questions about poverty and spatial inequalities, such as: How many people faced multidimensional poverty in 2019? How much did poverty decline between 2009 and 2019? Where did poverty decline faster and where does it still persist? Was there a convergence of poverty across provinces and wards? And in what dimensions and indicators did multidimensional poverty improve? Key findings With a little more than one-third of the population facing multidimensional poverty, Solomon Islands is more severely deprived than other countries in comparison. About 35 percent of the population in Solomon Islands is estimated to face multidimensional poverty in 2019; in other words, one-in-three Solomon Islanders are multidimensionally poor. This multidimensional poverty level is far higher than other PICs—such as Kiribati (19.8), Samoa (6.3), Fiji (1.5), and Tonga (0.9)—and the average of lower-middle-income countries where data is available (24.3). People are more likely to be multidimensionally poor in rural areas, particularly in provinces such as Guadalcanal and Malaita. About 43 percent of the rural population is considered multidimensionally poor in 2019, as opposed to only 13 percent among the urban population. Provinces with relatively high multidimensional poverty rates include Guadalcanal (47 percent), Malaita (43 percent), and Makira-Ulawa (50 percent). Around 249,100 people are multidimensionally poor as of 2019, and nearly two-thirds of them reside in Honiara, Guadalcanal, or Malaita provinces Solomon Islands has made only moderate progress in reducing multidimensional poverty from 2009 to 2019. The incidence of multidimensional poverty decreased from 43.5 percent in 2009 to 35.4 percent in 2019. Multidimensional poverty incidence in rural areas declined relatively rapidly, from 50 percent in 2009 to 43 percent in 2019. Poverty has become more concentrated in urban areas, with the urban share rising from 8 percent to 13 percent. Several provinces, such as Malaita, Central, and Choiseul experienced relatively robust poverty reduction over the decade. In contrast, the poverty level in rural Guadalcanal remains high. The substantial decrease in multidimensional poverty in poorer wards contributed to a convergence of poverty across the entire country. Among the multidimensional poverty components, housing quality and access to water have robustly improved, particularly in rural areas. Between 2009 and 2019, access to safe drinking water improved in most wards. Approximately 77 percent of the rural population now enjoys improved access to water, and 97 percent in urban areas. While sanitation access has improved at the national level, certain wards, particularly in Honiara and Guadalcanal, which experienced a significant influx of migrants, face worsening conditions. 2.1 Solomon Islands’ multidimensional considered multidimensionally poor in 2019. Is this poverty level is higher than the average of multidimensional poverty level in Solomon Islands too high? lower-middle-income countries Has the progress in poverty reduction been too slow? When considering these questions, it is useful to put Solomon In Solomon Islands, a little more than one-third of the Islands in the regional and international contexts. This population is experiencing multidimensional poverty. At chapter begins with comparisons of multidimensional poverty the national level, an estimated 35 percent of people are between Solomon Islands and other countries. The global 21 multidimensional poverty is reported by UNDP and OPHI water and sanitation (Panel A in Figure 7), access to clean (2023) for a few PICs, including Kiribati, Samoa, Fiji, and cooking fuel and electricity (Panel B), and school enrollment Tonga. While this report measures multidimensional poverty and adults with secondary education completed (Panel C).32 for Solomon Islands by broadly following the same global measurement methodology, some modifications were made Figure 7.  Solomon Islands is lagging in many due to data limitations.29 As such, their comparison presented non-monetary outcomes in this section should be interpreted with caution. (A) Access to water and sanitation Multidimensional poverty in Solomon Islands is more prevalent than in other Pacific Island countries. The 100 comparison of the multidimensional poverty rates in Figure 6 clearly indicates the severe prevalence of poverty in Solomon Islands. The multidimensional poverty rate of Solomon Islands (35 percent) is at a higher level compared to other PICs—such as Kiribati (20 percent), Samoa (6.3 percent), and Fiji (1.5 percent)—and the average MPI scores among lower-middle- 0 SI Structural Aspirational PIC LMIC income countries where data is available (24 percent).30 peer peer People using at least basic drinking water services (% of population) Figure 6.  Solomon Islands’ multidimensional poverty People using at least basic sanitation services (% of population) rate is relatively high Comparison of multidimensional poverty rates (B) Access to energy between Solomon Islands and other countries 100 40 Multidimensional poverty rate (%) 35 30 25 20 0 15 SI Structural Aspirational PIC LMIC peer peer 10 5 Access to clean fuels and technologies for cooking (% of population) 0 Access to electricity (% of population) ) ) ) e ) ) 19 19 20 21 19 rag (C) Education (20 /20 /20 (20 (20 ave 100 18 19 nds Fiji ga IC 20 (20 Ton LM Isla ti ( oa iba on Sam om Kir Sol Note: The multidimensional poverty rate indicates the percentage of the population in multidimensional poverty in a country. Source: Staff calculations based on Population and Housing Census 2019 for Solomon Islands and UNDP and OPHI (2023) for other countries. 0 SI Structural Aspirational PIC LMIC Deprivation in living standards and education is peer peer particularly severe in Solomon Islands. This is not only Educational attainment, at least completed upper secondary, population 25+, based on the comparison with PICs but with lower-middle- total (%) (cumulative) income small state countries (considered as structural peers, Adjusted net enrollment rate, primary (% of primary school age children) such as Comoros, Kiribati, Federated States of Micronesia (FSM), Samoa, Timor-Leste, and Vanuatu) and upper-middle- Note: Structural peers include Comoros, Kiribati, FSM, Samoa, Timor-Leste, and Vanuatu. Aspirational peers include Fiji, Grenada, and Maldives. income small state countries (aspirational peers, such as Fiji, Grenada, and Maldives).31 Solomon Islands is lagging behind Source: Staff calculations based on Population Census 2019 and WDI. these countries in various outcomes, such as access to basic 29 Multidimensional poverty for many countries was measured based on 10 indicators from middle-income countries, as opposed to 9 indicators from Population Censuses in Solomon Islands. See Annex A for more details. 30 Lower-middle-income countries include Algeria, Bangladesh, Benin, Cambodia, Cameroon, Ghana, Guinea, Haiti, Honduras, India, Jordan, Kyrgyzstan, Lao People’s Democratic Republic, Lesotho, Mauritania, Mongolia, Morocco, Nepal, Nigeria, Pakistan, Papua New Guinea, Philippines, Sao Tome and Principe, Senegal, Tajikistan, Tunisia, Uzbekistan, Viet Nam, Zambia, and Zimbabwe. 31 The selection of structural and aspirational peers follows World Bank (2024). 32 Women remain disadvantaged across multiple dimensions, with poor health outcomes, including high maternal mortality and a higher likelihood of working in vulnerable environments. High rates of gender-based violence are also a significant impediment to (mainly women’s) well-being and that of their families, negatively impacting not only health but also education and economic opportunity (WHO 2021). The impacts of poor infrastructure disproportionately affect women’s time use because they are primarily responsible for unpaid domestic work. 22 and Spatial Disparities Multidimensional Poverty  inSolomon Islands 2.2 Multidimensional poverty is more Table 3.  Multidimensional poverty is more prevalent prevalent and concentrated in rural areas and concentrated in rural areas Multidimensional poverty rates, poor populations, In Solomon Islands, multidimensional poverty incidence and their shares among the national total, 2019 is higher in rural areas—particularly within Guadalcanal and Malaita—than in urban areas. At the national level, an estimated 35 percent of people—or about 249,000 people— Poor Poverty Share are considered multidimensionally poor in 2019 (Table 3). population rate (%) (%) However, the incidence of multidimensional poverty is (thousand) significantly lower in urban areas, with only 16 percent of the National 35.4 249.1 100.0 population in poverty, compared to 43 percent in rural areas. Notably, the poverty incidence in Honiara City Council is Urban 16.2 31.8 12.8 particularly low at only 12 percent. On the other hand, Guadalcanal, Malaita, Makira-Ulawa, and Temotu provinces  Honiara have relatively high poverty rates (more than 40 percent). 16.2 27.1 10.9 Urban Area The multidimensional poor population is concentrated  Other in rural areas, particularly in populous provinces. About 16.6 4.7 1.9 urban 87 percent of the country’s multidimensional poor population lives in rural areas in 2019. Moreover, nearly two-thirds of the Rural 42.7 217.3 87.2 poor population can be found in Honiara, Guadalcanal, or Malaita. These three provinces account for 6.3 percent  Rural 53.1 59.5 23.9 (15,700 individuals), 28.4 percent (70,900 individuals), and Guadalcanal 29.4 percent (73,300 individuals) of the country’s poor   Other rural 39.8 157.8 63.3 population, respectively. The concentration of multidimensional poverty in these areas reflects their large population. Choiseul 25.4 7.6 3.0 Interestingly, Honiara, despite having the lowest multi­ dimensional poverty incidence, accommodates more of the Western 32.0 28.7 11.5 poor population than many other provinces, including Temotu, Rennell and Bellona, Central, Isabel, and Choiseul. Isabel 25.3 7.2 2.9 Multidimensional poverty rates vary within provinces, while wards with significantly high poverty levels are Central 37.6 11.2 4.5 concentrated in a few provinces. While provinces like Rennell and Guadalcanal and Malaita have high overall multidimensional 19.2 0.7 0.3 Bellona poverty levels, it is important to note that each province consists of wards with both low and high multidimensional Guadalcanal 46.6 70.9 28.5 poverty rates (Figure 8A). Wards with significantly high poverty rates, exceeding 50 percent, exist even in low- Malaita 42.7 73.3 29.4 poverty provinces like Choiseul and Isabel. Nevertheless, high-poverty wards are predominantly concentrated in Makira-Ulawa 49.6 24.8 10.0 Guadalcanal, Malaita, and Makira-Ulawa provinces, as well as Temotu and Western to a lesser extent. In terms of the Temotu 41.1 9.0 3.6 number of multidimensionally poor people, Guadalcanal and Honiara 12.4 15.7 6.3 Malaita Islands stand out due to their large population sizes (Figure 8B). A further look at 183 wards in Figure 9 make clear the areas with high multidimensional poverty incidence Note: Numbers indicate for each location the proportion of multidimen- sionally poor population (column 1), the number of poor people (column and populations. 2), and the share of poor population among the national total (column 3). Honiara Urban Area includes the urban population in Honiara and Guadal- canal Province (see Annex A2). Source: Staff calculations based on Population and Housing Censuses 2019. 23 Figure 8.  Wards with high multidimensional poverty are concentrated in Guadalcanal and Malaita (A) Multidimensional poverty incidence at the ward level, 2019 24 and Spatial Disparities Multidimensional Poverty  inSolomon Islands (B) Number of the multidimensionally poor at the ward level, 2019 Source: Staff calculations based on Population and Housing Census 2019. Figure 9.  A group of wards are characterized by high poverty incidence and populations 100 80 Poverty incidence 2019 60 40 20 0 0 10000 20000 30000 40000 Population 2019 Honiara Guadalcanal Malaita Others Note: Markers represent 183 wards. Source: Staff calculations based on Population and Housing Census 2019. 25 One-third of the multidimensionally poor population are 2.3 Multidimensionally poverty incidence children under age 10, while there is no clear gender declined across the board between 2009 difference in the composition of the poor. As households and 2019 with younger children tend to be poorer, the multidimensionally poor population tends to be younger than the non-poor Over the years, there has been a moderate decline population (Panel A in Figure 10). Those who are older than 60 in multidimensional poverty incidence since 2009, account only for 5 percent of the multidimensionally poor primarily due to the progress made in rural areas and population, reflecting the country’s population composition poorer provinces. Nationally, the multidimensional poverty toward younger age cohorts. Males and females are almost incidence has decreased from 43.5 percent in 2009 to 35.4 equally split in each age cohort of the multidimensionally poor percent in 2019 (Table 4).33 This decline is even more population (Panel B). significant when considering the urban-rural gap in multidimensional poverty, which has shrunk during the past decade. This positive change can be attributed to the Figure 10.  One-third of the multidimensional poor are relatively strong multidimensional poverty reduction in rural children under age 10, while there is no areas, dropping from 50 percent in 2009 to 43 percent in gender difference 2019. Noteworthy reductions in multidimensional poverty Age and gender composition of the incidence have been observed in Choiseul (a 15 percent multidimensionally poor population (%) (cont.) point reduction), Malaita (a 14-point reduction), and Central (an 11-point reduction). On the other hand, the (B) By age and gender multidimensional poverty level in rural Guadalcanal remains very high at 53 percent. 100% Despite a slightly larger share of the multidimensional poor population living in urban areas than previously, the majority still lives in rural areas. Because of the rapid urbanization, the urban share of the multidimensionally poor population increased from about 8 percent in 2009 to 13 50% percent in 2019 (Panel B in Figure 11). Nonetheless, since poverty incidence in rural areas (43 percent) remains substantially higher than in urban areas (16 percent), nearly 90 percent of the country’s multidimensionally poor population (217,400 out of 249,100) still resides in rural 0% areas. The share in rural Guadalcanal increased from 22 0 to 10 11 to 20 21 to 30 31 to 40 41 to 50 51 to 60 61 to 70 71+ percent in 2009 to 28.4 percent in 2019 (Panel A). Age group Male Female (A) By age 100 5 5 16 14 25 30 22 21 33 28 0 All individuals Poor individuals 0-10 11 to 20 21-40 40-60 60+ Source: Staff calculations based on Population and Housing Census 2019.  33 To estimate multidimensional poverty for this report, households that have missing values in any MPI indicator are excluded from the Census data. Therefore, the population is slightly lower than the official report (see Table B1 in Annex B). 26 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Table 4.  Multidimensional poverty incidence moderately Figure 11.  Nearly two-thirds of the multidimensionally fell between 2009 and 2019 poor population live in Honiara, Guadalcanal, Incidence of multidimensional poverty, and Malaita. 2009-2019 (%) Spatial distributions of the multidimensional poor, 2009 and 2019 (%) 2009 2019 Difference (pt) (A) By province National 43.5 35.4 -8.1 28.1 24.3 Urban 17.9 16.2 -1.6 11.5 10.4  Honiara 17.4 16.2 -1.2 Urban Area 29.4  Other 35.5 19.5 16.6 -2.9 urban Rural 49.7 42.7 -7.0 28.4 22.0  Rural 4.0 6.3 56.8 53.1 -3.6 Guadalcanal 2009 2019   Other rural 48.1 39.8 -8.3 Honiara Guadacanal Malaita Western Other provinces Choiseul 40.0 25.4 -14.6 (B) By urban/rural Western 31.2 32.0 0.9 Isabel 31.6 25.3 -6.4 Central 48.2 37.6 -10.6 63.4 72.1 Rennell and 22.0 19.2 -2.9 Bellona Guadalcanal 52.6 46.6 -6.0 23.9 19.9 1.9 Malaita 57.0 42.7 -14.2 1.9 6.2 10.9 Makira-Ulawa 50.9 49.6 -1.3 2009 2019 Temotu 47.8 41.1 -6.7 Honiara Urban Area Other Urban Rural Guadalcanal Other rural Honiara 14.1 12.4 -1.8 Note: Numbers indicate the shares of multidimensionally poor populations by province. Honiara Urban Area includes the urban population in Honiara Note: Numbers indicate the proportion of multidimensionally poor popu- and Guadalcanal Province (see Annex A2). lation in each location. Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. The number of multidimensionally poor individuals is multidimensionally poor individuals by around 14.3 percent, or likely to have increased during the period from 2009 to 31,200 people (Table B2 in Annex B). However, once the 2019, but it is not entirely clear due to potential population undercount is adjusted, it is possible that the population undercount in the 2009 Census. Given the increase in the multidimensionally poor population was more potential population undercount in the 2009 Population and moderate—as low as 1 percent based on the assumption that Housing Census, it is difficult to provide precise estimates for the multidimensional poverty rate among the undercounted the trend in multidimensional poor populations. A larger population is similar to that of the rural population (50 percent) number of multidimensionally poor individuals in 2009 would (Box 3). Similarly, the number of multidimensionally poor make the decreases (or the increases) in the multidimensional individuals have likely increased in Honiara Urban Area, even poor population between 2009 and 2019 steeper (or less based on the most conservative scenario assuming the steep). Based on unadjusted population counts, the country’s multidimensional poverty rate among the undercounted rapid population growth has—despite the moderate decline in population is at a similar level to those in rural areas. multidimensional poverty incidence—increased the number of 27 Box 3. Estimating changes in the number of multidimensionally poor individuals between 2009 and 2019 SINSO (2023) mentions a potential population undercount in the 2009 Population and Housing Census. This makes it a challenge to estimate the changes in the number of multidimensionally poor individuals between 2009 and 2019. An assumption has to be made about the multidimensional poverty incidence among the undercounted population in 2009, and regardless of the assumption, the multidimensionally poor population is likely to have increased since 2009. When applying the same multidimensional poverty rate as in Honiara, Guadalcanal, and Malaita provinces (about 42 percent) or as in the national level (43.5 percent), the multidimensionally poor population is estimated to have increased by around 7,000 (or 3 percent). When the urban poverty rate (18 percent) is applied to the undercounted population, the multidimensionally poor population is estimated to have increased more (20,800 individuals or 9.1 percent). Table 5.  Number of multidimensionally poor individuals (thousands) 2009 2019 Diff. National 2009 population unadjusted 218.0 249.1 31.2 14.3% 2009 population adjusted   Poverty rate of the missing population:   1) 50% (same as in rural areas) 246.5 249.1 2.6 1.1%   2) 43.5% (same as the national level) 242.9 249.1 6.2 2.5% 42% (same as in Honiara, Guadalcanal, and Malaita)   3)  242.1 249.1 7.0 2.9%   4) 18% (same as in urban areas) 228.3 249.1 20.8 9.1% Honiara Urban Area 2009 population unadjusted 13.4 27.1 13.7 102% 2009 population adjusted   Poverty rate of the missing population:   1) 50% (same as in rural areas) 20.6 27.1 6.5 31.5%   2) 43.5% (same as the national level) 19.7 27.1 7.4 37.5% 42% (same as in Honiara, Guadalcanal, and Malaita)   3)  19.5 27.1 7.6 39.0%   4) 18% (same as in urban areas) 16.0 27.1 11.1 69.3% Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. The degree of deprivation among the multidimensionally multidimensional poverty, which represents the average poor population did not change between 2009 and 2019. share of MPI indicators that the multidimensionally poor are Despite the decrease in multidimensional poverty incidence, deprived of, remained at almost the same level, with 48.7 in there has been little improvement in the intensity of 2009 and 47.5 in 2019 (Table 6). Multidimensional poverty deprivation between 2009 and 2019. The multidimensional intensity remained at the same level across locations, poverty index (MPI) captures both multidimensional poverty including both in urban and rural areas. As a result, the MPI incidence and the intensity of deprivation among the changed in a similar way to the declining multidimensional multidimensionally poor population. The intensity of poverty incidence.34  34 Furthermore, reductions in severe poverty have been observed in terms of incidence, headcount, and intensity. Severe poverty is measured by raising the threshold of deprivation from 0.33 to 0.50. The incidence of severe poverty decreased from 19 percent in 2009 to 13 percent in 2019. Additionally, the number of people living in severe poverty decreased by 5 percent over the decade. The severe MPI also dropped from 0.116 in 2009 to 0.081 in 2019. It is worth noting that Guadalcanal and Malaita provinces, which had the highest severe MPIs in 2009, have made progress and are catching up with other province 28 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Table 6.  In contrast to the declining multidimensional poverty incidence, there was no change in the intensity between 2009 and 2019 Changes in multidimensional poverty intensity and MPI, 2009-2019 Intensity MPI 2009 2019 Diff. 2009 2019 Diff. National 48.7 47.5 -1.2 0.212 0.168 -0.044 Urban 45.8 44.3 -1.5 0.082 0.072 -0.010   Honiara Urban Area 45.4 43.9 -1.5 0.079 0.071 -0.008   Other urban 47.1 46.3 -0.7 0.092 0.077 -0.015 Rural 48.9 47.9 -1.0 0.243 0.205 -0.038   Rural Guadalcanal 49.1 47.6 -1.6 0.279 0.253 -0.026   Other rural 48.9 48.1 -0.8 0.235 0.191 -0.044 Note: Numbers indicate the MPI. Guadalcanal does not include Honiara. Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Figure 12.  MPI reduction in poorer wards resulted in a convergence between 2009 and 2019 Changes in MPI at Ward level, 2009-2019 A) Comparison of 2009 and 2019 (B) MPI distributions .6 15 2009 2019 .4 10 .2 5 0 0 .2 .4 .6 0 MPI (2009) 0 .2 .4 .6 Honiara Guadalcanal Malaita Others MPI Note: The wards below the 45-degree line experienced MPI reduction between 2009 and 2019 (Panel A). Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 29 There was a convergence of multidimensional poverty, Figure 13.  Almost all multidimensionally poor people particularly in poorer wards, between 2009 and 2019. are deprived in living standards, while being Out of 183 wards in the country, 131 wards experienced deprived in all three dimensions is rare reductions in MPI. The scale of MPI reductions has been Overlaps of three multidimensional poverty larger among wards that had relatively high MPI values in dimensions in 2019 (%) 2009, mainly in Guadalcanal and Malaita provinces (Panel A in Figure 12). As a result, the distribution of MPI at the ward Health level has shifted towards the lower end and become narrower (7.4) (Panel B). The reduced inequality is quantitively confirmed by 0.3 various statistics.35 0.2 2.4 Improvements in living standard 0.02 dimensions, particularly housing 4.5 and water, drove multidimensional 2.5 poverty reduction Education (26.1) Almost all Solomon Islanders in multidimensional poverty are deprived in the living standard dimension. The proportion of people deprived in the living standard 23.4 dimension is substantially high (34.9 percent), making almost 4.6 every poor person deprived in this dimension (Figure 13). Living While deprivations in health (7.4 percent) and education standards (26.4 percent) dimensions significantly overlap with the (34.9) deprivation in living standards, the health and education deprivations themselves do not overlap much, as only 2.7 percent of the population are deprived in both dimensions. Note: The numbers in the figure add up to 35.4 percent, corresponding to Consequently, only a tiny portion of the population (2.5 the proportion of the multidimensional poor (not precisely proportional to the areas). A household is considered deprived in each dimension if the depri- percent) are deprived in all three dimensions. vation score in each category is 1/9 or above, while a household is considered multidimensionally poor if the overall deprivation score is 1/3 or above. There were relatively strong improvements in housing and drinking water, which may have brought additional Source: Staff calculations based on Population and Housing Census 2019. benefits to women. Nationally, the proportion of the population deprived of these outcomes decreased by 18 Figure 14.  Deprivations in housing, water, and, to a lesser percentage points (from 66 to 48 percent) and 12 percentage extent, sanitation and cooking fuel improved points (from 29 to 17 percent), respectively (Figure 14). At The proportion of the population deprived the sub-national level, provinces such as Guadalcanal, of each MPI indicator, 2009 and 2019 Malaita, and Makira-Ulawa witnessed solid improvements in 93 access to improved water. Access to grid electricity has 2009 2019 87 83 84 86 83 70 improved only slightly between 2009 and 2019. As women 66 61 are primarily responsible for unpaid domestic work, including 48 29 food preparation and cooking and water collection and 19 21 14 12 17 10 8 purification, such improvements are likely to have benefited women’s time-use in particular. Child mortality School enrollment Educational attainment Housing Drinking water Sanitation Electricity Cooking fuel Asset ownership Health Education Living standards Source: Staff calculations based on Population and Housing Censuses 2009 and 2019.  35 The standard deviation of MPI at the ward level declined from 0.10 in 2009 to 0.09 in 2019. Moran’s I statistics for MPI, which measures the extent of concentration/dispersion of poverty across wards, declined from 0.717 in 2009 to 0.567 in 2019. 30 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Education Furthermore, there has been a slightly higher share of households that were deprived in terms of school Despite some improvements, educational levels in the enrollment in 2019 compared to 2009. The proportion of working-age population remain low, with relative gender households with at least one school-age child not currently parity up until the tertiary level. The proportion of households enrolled in school slightly increased from 15.1 percent in with no adult having completed primary education decreased 2009 to 16.4 percent in 2019 (Table 7).36 While there have from 18 percent in 2009 to 15 percent in 2019 (Table 7), been improvements in some provinces, such as Malaita, indicating progress in this area. However, this improvement was many others have experienced worsening school enrollment. not observed in Honiara. Also, the percentage of the working- It is worth noting that households deprived of school age population with only primary education completed declined enrollment tend to be deprived in other dimensions as well. from 69 percent to 56 percent, while the percentage of those For example, nearly 90 percent of households deprived of with incomplete secondary education increased (Figure 15). school enrollment are multidimensionally poor. In contrast, Overall, the education level in the country remains at a low level. only 24 percent of households not deprived of school As of 2019, nearly 90 percent of the working-age population enrollment are multidimensionally poor. have not completed Year 7 secondary education. Furthermore, the current formal educational enrollment for those aged 6 to 11 and 12 to 18 was at parity for male and female in both urban and rural settings, though males aged 19 to 24 were 5 percentage points more likely to be enrolled in formal education. Table 7.  Other than educational attainment in rural areas, there was no significant improvement in education deprivation Changes in educational deprivation indicators, 2009-2019 (%) Deprivation in school enrollment Deprivation in educational attainment 2009 2019 Diff. 2009 2019 Diff. National 15.1 16.4 1.2 18.0 14.7 -3.3 Urban 12.6 14.9 2.4 4.1 4.4 0.3   Honiara Urban Area 14.0 16.0 2.0 3.8 4.4 0.6   Other urban 8.3 9.3 1.0 5.1 4.9 -0.3 Rural 15.6 16.8 1.2 20.8 18.2 -2.6   Rural Guadalcanal 17.0 24.3 7.3 24.8 20.7 -4.2   Other rural 15.3 14.7 -0.6 19.8 17.5 -2.4 Note: Numbers indicate the proportions of the households deprived for each education indicator. School enrollment: At least one school-age child up to the (equivalent) age of grade 8 is not enrolled in school. Educational attainment: No adult in the household (equivalent age of grade 9 or above) has completed primary education. Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Source: Staff calculations based on Population and Housing Censuses 2009 and 2019.  36 The percentage of children (ages 6 to 14) not in school also increased from 16.1 percent in 2009 to 18.4 percent in 2019. 31 Figure 15.  Nearly 90 percent of working-age adults did reduced between 2009 and 2019 in most provinces, including not complete secondary education both urban and rural areas (Figure B1). Even at the ward level, Changes in educational attainment, the proportion of the deprived population was reduced 2009-19 (%) (Figure B1 in Annex B). Nonetheless, 31 wards (6 in rural Guadalcanal) still have more than 80 percent of the 100% population that is deprived of housing. While there have been moderate improvements in access to basic services, particularly water, there are still significant challenges in rural areas and certain provinces. The Solomon Islands Government National WASH Policy had the vision that all Solomon Islanders will have easy access to sufficient quantity and quality of water, 0% appropriate sanitation, and will be living in a safe and hygienic 2009 2019 2009 2019 2009 2019 2009 2019 environment by 2024.39 It appears that progress was made towards this goal, with improved water access increasing National Honiara Other urban Rural Urban Area from 71 percent in 2009 to 83 percent in 2019 (Table 9). In urban areas like Honiara Urban Area, more than 95 percent of No education/primary incomplete Secondary complete the population now has improved water access. However, Primary incomplete Tertiary complete nearly a quarter of the rural population still lacks access to Secondary incomplete Others improved water. Provinces such as Guadalcanal, Malaita, Makira-Ulawa, and Temotu continue to face water access Note: Ages 15 to 64 only. Honiara Urban Area includes the urban popula- tion in Honiara and Guadalcanal Province (see Annex A2). challenges. Source: Staff calculations based on Population and Housing Censuses At the ward level, there have been varying degrees of 2009 and 2019. reduction in water deprivation since 2009. Out of the As highlighted by the World Bank Public Expenditure total 183 wards in the country, 149 wards have seen a Review, government spending on education is high, but decrease in the proportion of the population lacking access it has not translated to increased education outcomes.37 to improved water (Panel A in Figure B2 in Annex B). The Government expenditures on education are approximately number of wards with only less than 10 percent of the 10.1 percent of GDP, markedly higher than the OECD average population deprived of improved water access has increased of 3.4 percent and the global average of 4.7 percent. The high from 60 in 2009 to 92 in 2019. Water deprivation has been level of spending partly reflects structural factors, including a rapidly improved in the wards that had relatively high MPI high proportion of school-aged children in the population, a values in 2009 and have experienced high rural population high cost of providing education services in the outer islands, growth since 2009. However, it is important to note that and disproportional spending on specific programs. The some wards, mainly in Guadalcanal Province, still have a high expenditures, however, are not currently resulting in similar percentage of residents lacking improved water access, with gains, especially in upper secondary and tertiary education. 80 percent or more facing this deprivation. This has left the country with a skills gap that jeopardizes In addition to the transition from unimproved to long-term development. For instance, according to the World improved access to water, there has been a notable shift Bank’s Human Capital Index, a child born in Solomon Islands in rural households towards relying on private piped today will only be 42 percent as productive when they grow water. The improvement in water access in rural areas is up compared to what they would have been if they enjoyed evident as many households have moved away from relying complete education and full health. on unimproved water sources, particularly surface water. Living standards Furthermore, the prevalence of private tap water usage has become more widespread among rural households (Table 8). Housing deprivation is reduced across the board, both in For instance, in rural Guadalcanal, the proportion of urban and rural areas. Following the global MPI households with private taps has experienced a significant methodology, a household is considered deprived of housing increase from less than 1 percent to 19 percent, while the if the floor is made of natural materials (mud/clay/earth, sand, reliance on public taps has decreased from 27 percent to 17 or dung), if the dwelling has no roof or walls, or if either the percent. Similar transitions have been observed in other rural roof or walls are constructed using natural or rudimentary areas. materials.38 Housing deprivation, defined in this way, was  37 World Bank (2022a).  38 Those rudimentary materials include carton, plastic/polythene sheeting, bamboo with mud/stone with mud, loosely packed stones, uncovered adobe, raw/reused wood, plywood, cardboard, unburnt brick or canvas/tent. Ministry of Health and Medical Services (2014). 39 32 and Spatial Disparities Multidimensional Poverty  inSolomon Islands However, access to improved water needs to be World Bank Systematic Country Diagnostic, it is estimated accompanied by good quality infrastructure. One of the that half of the communal standpipes in rural areas are not major challenges in providing water supply in rural areas is fully operational.40 A nationally representative household the lack of maintenance (World Bank 2017). Many survey conducted between 2015 and 2017 showed that 54 communities in rural areas have been given multiple water percent of households (59 percent in urban and 51 percent in systems by Members of Parliament (MPs) through CDFs, rural) perceived water quality as good, whereas 9 percent of donors, or NGOs. While there has been a strong focus on households—mainly in rural areas—perceived water quality supplying small-scale infrastructure, emphasis on ensuring as poor, as the water was polluted, cloudy, or muddy.41 proper maintenance has been insufficient. According to Table 8.  Many rural households shifted to improved water sources for drinking, including private water taps Changes in access to water in rural areas, 2009-19 (%) Rural Guadalcanal Other rural 2009 2019 Diff (pt) 2009 2019 Diff (pt) Private pipe 0.8 19.2 18.4 0.2 19.7 19.5 Public tap 27.1 17.0 -10.1 43.6 24.4 -19.3 Borehole/protected well 7.3 10.1 2.8 1.2 1.0 -0.1 Tank/other improved 11.7 15.8 4.1 25.0 36.8 11.8 Unimproved 53.1 37.9 -15.2 30.0 18.2 -11.8 Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Compared to water access, the improvement in While 116 wards have seen a decline in the proportion of the sanitation access has been slower, with 61 percent of population lacking sanitation access, 62 wards have people still lacking access. The population with improved experienced an increase since 2009 (Panel B in Figure B2 in sanitation access increased from 30 percent in 2009 to 39 Annex B). These wards with worsening conditions are mainly percent in 2019 (Table 9). However, in rural areas, only 26 located in Honiara and Guadalcanal, which have seen a percent of the population has improved access to sanitation significant influx of migrants in the past decade. Stronger as of 2019. Given the country’s high exposure to flooding and improvements in sanitation deprivation were observed in the extreme weather events, having access to improved wards with relatively high MPI levels in 2009. The number of sanitation is critical.42 Improving sanitation access is clearly a highly deprived wards, where more than 90 percent of the key to further reducing poverty for the next decade.43 residents lack sanitation access, has decreased from 78 in 2009 to 52 in 2019. At the ward level, there are mixed changes in improved sanitation access, with some wards experiencing worsening deprivation, particularly in Honiara and Guadalcanal. Unlike water access, improvements in sanitation access have not been consistent across all wards. World Bank (2017). 40 Anthonj et al. (2020). 41 Fleming et al. (2019). 42 The important role of urban infrastructure, particularly sanitation, for economic growth in lower-income countries has been reported by 43 several studies, including Castells-Quintana (2017). 33 Table 9.  Changes in access to basic services by province, 2009-2019 Water Sanitation Electricity 2009 2019 Diff. 2009 2019 Diff. 2009 2019 Diff. National 70.6 82.8 12.2 30.2 38.7 8.5 13.3 17.1 3.8 Urban 90.5 96.9 6.4 73.8 71.3 -2.5 56.1 53.8 -2.3  Honiara 90.8 96.7 5.9 77.2 72.5 -4.7 58.6 54.9 -3.7 Urban Area   Other urban 89.2 97.8 8.5 61.2 64.1 2.9 46.6 47.3 0.6 Rural 65.8 77.4 11.6 19.7 26.2 6.6 2.9 3.0 0.1  Rural 46.6 62.2 15.6 27.7 26.9 -0.8 4.5 4.2 -0.2 Guadalcanal   Other rural 70.2 81.7 11.4 17.8 26.1 8.3 2.5 2.6 0.1 Note: Numbers indicate the proportions of the population with improved access. Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Box 4. Definition of deprivation in water, sanitation, and electricity As explained in Section 1.2 (and Annex A), this report measures multidimensional deprivations following the global approach. A household is considered deprived of drinking water if it does not have access to an improved standard of drinking water (according to SDG guidelines).44 Specifically, households with unprotected wells, surface water, river/stream, or others are considered deprived of an improved standard of drinking water. In contrast, metered-Solomon Islands Water Authority (SIWA), protected well/spring, rainwater collection, or bottled water are considered improved sources of water. A household is considered deprived of sanitation if it does not have access to an improved standard of sanitation facilities (according to SDG guidelines).45 A household with no or shared toilet facilities (flush, water-sealed, or pit latrine) is deemed deprived of an improved standard of sanitation. Households with private access to toilet facilities (flush, water-sealed, or pit latrine) are considered to have adequate access to improved standards of sanitation. A household is considered deprived of electricity if it does not have access to electrification from grid electricity. Therefore, in Solomon Islands, if a household’s main source of lighting is from a generator, solar power, gas, kerosene lamp, wood/coconut, Coleman lamp, other, or none, it is considered to be deprived of electrification. There has been limited improvement in grid electricity 9 percent in 2009 to 84 percent in 2019. Solar lanterns access. Nationally, the proportion of the population with rather than solar home systems were used, as indicated improved electricity access has increased only slightly from by the high-frequency phone surveys (HFPS) collected by the 13 percent to 17 percent between 2009 and 2019 (Table 9). World Bank in January and February 2024. The evaluation Deprivation in electricity access is particularly severe in wards of this almost universal access to solar lanterns for lighting in Guadalcanal Province (Panel C in Figure B2 in Annex B). sources depends on its reliability and affordability. In fact, solar lanterns seem moderately reliable and affordable, at least as a In contrast, there was a remarkable uptake of solar source for lighting. Most people reported in the phone surveys that power for lighting sources. The proportion of the population solar lanterns were available for 12 hours a day, and that they paid relying on solar power for lighting sources increased from 44 This deprivation criteria based on unimproved access is different from the OPHI’s multidimensionally poverty methodology, which uses ’limited’ access. This is due to the lack of information in the Solomon Islands Census data. 45 Same as above. 34 and Spatial Disparities Multidimensional Poverty  inSolomon Islands nothing for it during the previous 30 days—a contrast to the Finally, inequality in living standards is clearly associated expensive grid electricity option. The respondents of the phone with education. As education is a key factor determining the surveys on average paid SBD 300 during the previous 30 days type of jobs and, therefore, earnings, it is not surprising that for electricity. household education level is positively correlated with living standards. But what is striking in Solomon Islands is the Nevertheless, solar lanterns cannot substitute electricity degree of their correlation. As shown in Figure 17, households or other better energy sources for other uses. Solar with better educated heads are less likely to face deprivations lanterns are used exclusively for lighting and charging mobile in all the living standard outcomes. For example, households phones according to the HFPS. Households still need to rely with heads who completed only primary education are almost on other energy sources for charging electric appliances and twice as likely to face housing deprivation compared to cooking. About 84 percent of households still rely on wood or households with heads who completed secondary education. coconut shells as energy sources for cooking. This clear correlation between education and multidimensional Except for mobile phones, no significant increase in poverty ascertains that the latter is not purely driven by household asset ownership is observed. As shown in public sector provisions. Except for housing and water, the Figure 16, asset ownership is one of the most severe gaps in living standards between low-educated and high- deprivations, next to electricity and cooking fuel. Item-wise, educated households did not change much between 2009 mobile phone ownership increased from 21 percent in 2009 and 2019. to 45 percent in 2019 at the household level. The proportion of households owning a computer also increased to 11 Figure 17.  Living standards are clearly associated with percent. In contrast, ownership of radios and canoes household heads’ education levels decreased during the period. This lack of change in asset Percentage of population deprived in living ownership implies that households may not have experienced standard indicators by household head improved livelihoods during the decade. education level, 2009 and 2019 Figure 16.  Mobile phone ownership increased, 100 90 % of population deprived but asset ownership remains low 80 Percentage of households owning 70 60 each asset, 2009 and 2019 50 40 50% 30 20 10 40% % of households 0 2009 2019 2009 2019 2009 2019 2009 2019 2009 2019 2009 2019 30% Housing Water Sanitation Electricity Cooking fuel Asset 20% None Primary complete 10% Lower secondary Higher secondary Tertiary 0% Source: Staff calculations based on Population and Housing Censuses o n or n ne ne Co io er f ri V or e ike ck r Ca no T M o ec tio d ut rat rat Tr u ho ho 2009 and 2019. rb Ra Ca mp ne ge ep le p to n Ge Mo bil Te Re tc ne r te In 2009 2019 Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 35 36 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Chapter 3. Migration, Urbanization, and Jobs The previous chapter confirms the reductions in multidimensional poverty and spatial inequalities between 2009 and 2019. The subsequent chapters analyze their potential key drivers. This chapter first focuses on internal migration and urbanization. It then examines jobs, which are another potential driver for multidimensional poverty reduction. The questions explored include: Did the massive migration to the Honiara Urban Area drive or contribute to reductions in poverty and spatial inequalities? How did this urbanization affect the job landscape of Solomon Islands? And did job growth and a shift to high- productivity, non-agricultural jobs contribute to the reductions in poverty and spatial inequalities during the decade? Key findings Honiara Urban Area stands out as the primary destination for migrants, attracting individuals from all corners of the country. The population in Honiara Urban Area experienced a rapid increase of 86 percent, growing from 91,000 in 2009 to 170,000 in 2019. Although this urban growth might be overestimated due to the potential population undercount in the 2009 Population Census, Honiara has been the magnet for substantial internal migration. Migration to Honiara Urban Area accounted for 50 percent of inter-provincial migration between 2014 and 2019. Within the Honiara Urban Area, wards in outer areas, including urban Guadalcanal, have witnessed particularly rapid population growth. In contrast, towns in provinces other than Honiara have experienced modest urban growth and remain relatively small in terms of population size. While the majority of the multidimensional poverty reduction during the decade took place in rural areas, internal migration still accounted for a significant share. When multidimensional poverty changes are geographically decomposed, internal migration accounted for 30 percent of the poverty reduction achieved between 2009 and 2019. Rural-to-urban migrants are generally better off compared to households in their origin rural wards and tend to have characteristics similar to those of other urban households. This selective nature of migration has become lessened over time. Lack of economic opportunities in villages seems to have driven migration as well, as out-migration was more prevalent in wards with higher multidimensional poverty rates in 2009. While a large number of households that have migrated to Honiara have better access to basic services, the increasing strain on urban services needs to be well managed. Given the wide gap in access to economic opportunities and basic services and infrastructure, migrants from rural areas face multidimensional poverty significantly less in urban areas. However, a growing body of evidence is illustrating how the increase in the urban population is putting a strain on service delivery and causing other potential problems in the city. For example, access to sanitation appears to have already started deteriorating, which has serious implications for public health in crowded areas— particularly in view of natural hazard risks intensified by climate change. Thus, the potential benefit to migrants needs to be weighed against these other costs to better identify the evolving policy response. 37 3.1 Urbanization has been driven by migration have relatively small population sizes, mostly ranging up to to Honiara 2,000 residents. Apart from Honiara, the relatively large towns include Noro (7,200 residents in Western Province) The recent urbanization of Solomon Islands was and Auki (7,000 residents in Malaita Province). primarily concentrated in the Honiara Urban Area. The urban population in the country experienced a rapid increase The wards experiencing rapid population growth are from 110,000 in 2009 to 199,000 in 2019. Consequently, primarily located in Honiara, Guadalcanal, and Malatia the urban population share also rose from 19.8 percent to provinces. Out of the 183 wards, 43 experienced population 27.6 percent. Notably, nearly 90 percent of the country’s growth faster than the national-level growth of 40 percent urban population growth occurred in Honiara Urban Area between 2009 and 2019. Among these 43 wards, 31 are in (Table 10). Urban Guadalcanal witnessed particularly rapid Honiara (8 wards), Guadalcanal (12 wards), and Malaita (11 growth, with a staggering increase of 132 percent. As of wards). Notably, the population more than doubled in 10 2019, Honiara Urban Area accommodates approximately 85 specific wards, including Noro (Western), West Gaongau percent of the country’s urban population. With the exception (Rennel and Bellona), Tandai and Malango (Guadalcanal), of Western and Malaita provinces, towns in other regions West Baegu/Fataleka and Sulufou/Kwarande (Malaita), and Nggossi, Kola’a, Vura, and Panatina (Honiara). Table 10.  Nearly 90 percent of the urban population growth occurred in Honiara Urban Area Urbanization in Solomon Islands, 2009-19 Urban population (thousands) Urban population share (%) 2009 2019 Difference 2009 2019 Difference All urban 110.5 199.1 80.3% 19.8 27.6 7.8   Honiara Urban Area 91.3 169.7 86.0% 100 100 0   Other urban 19.2 29.4 53.2% 100 100 0 Choiseul 0.8 1.1 30.0% 3.1 3.4 0.3 Western 9.8 14.6 49.7% 12.7 15.5 2.8 Isabel 1.0 1.3 38.2% 3.7 4.3 0.6 Central 1.3 1.5 18.4% 4.8 4.9 0.1 Rennell and Bellona 0.0 0.0 - 0.0 0.0 0.0 Guadalcanal 17.3 40.2 131.5% 16.2 26.1 9.9 Malaita 2.4 7.0 197.8% 1.7 4.1 2.4 Makira-Ulawa 2.1 2.1 1.6% 5.1 4.1 -1.0 Temotu 2.0 1.8 -8.9% 9.3 8.1 -1.2 Honiara 73.9 129.6 75.4% 100 100 0 Note: Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 38 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Figure 18.  Population changes at the ward level, 2009-2019 Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Honiara and other towns have experienced a significant Province also attracted many migrants because of the influx of migrants, accounting for 50 percent of opening of mining and logging operations. In contrast, interprovincial migration. The proportion of migrants provinces like Malaita and Central had more people leaving recently arriving at their current residences varies across than entering.46 locations. When focusing on cross-province migrants, Between 2014 and 2019, more people left provincial approximately 7 percent of the population migrated to their capital towns than entered, indicating insufficient present province in the last five years (Table 11). Notably, the economic opportunities in those areas. Many individuals share of recent migrants is exceptionally high in Honiara moved to the wards where provincial capital towns are located Urban Area (15 percent) and other urban areas (10 percent). (Table 11). For instance, nearly a quarter of the residents in Migration to Honiara Urban Area accounted for 50 percent of Tulagi (Central) and Gizo (Western) arrived there during this people who migrated across provinces between 2014 and period. Despite attracting a large number of migrants, these 2019, while migration to other urban areas accounted for towns had negative net migration rates as more people left. only 6 percent. Positive net migration rates are observed in The main destinations for individuals leaving provincial capital only a few provinces, such as Honiara and Guadalcanal towns were Honiara and Guadalcanal, likely in search of better Province, where the number of incoming migrants exceeded economic opportunities and access to amenities.47, 48 the number of outgoing migrants. Rennel and Bellona While people migrate for different reasons, climate change and disaster-related migration have been common in Solomon Islands (World Bank 2023). 46 Since 2008, weather-related events have triggered around 26,000 displacements. Two single events—Cyclone Uli in 2010 and the 2014 flash flooding in Honiara—were responsible for displacing 15,000 people. Some migrations were conducted at the community level, arranged by either the government or themselves. Among those who left provincial capitals between 2014 and 2019, 16 percent of them moved to Honiara or Guadalcanal, followed by Western 47 Province (16 percent). People who left non-capital towns during the same period also moved to those destinations. There is no information available in the Population Census about the reasons for migration. 48 39 Table 11.  More people left than arrived in places except for Honiara and Guadalcanal Composition of migrants and net migration rates, 2019 Provinces and urban/rural Wards of provincial towns Recent Other Net Recent Other Net migrants residents migration migrants residents migration (%) (%) rate (%) (%) rate National 7.1 92.8 - - - - All urban 14.2 85.7 - - - -   Honiara Urban Area 14.9 85.1 - - - -   Other urban 10.4 89.6 - - - - Rural 4.3 95.6 - - - -   Rural Guadalcanal 4.6 95.4 - - - -   Other rural 4.2 95.8 - - - - Choiseul 4.6 95.3 -30.9 10.6 89.4 31.0 Western 6.8 93.1 14.9 22.6 77.4 -19.4 Isabel 7.5 92.4 -15.1 17.8 82.2 -141.3 Central 4.2 95.7 -30.5 25.8 74.2 -159.0 Rennell and Bellona 23.0 76.9 151.4 - - - Guadalcanal 8.9 91.0 72.4 20.7 79.3 186.5 Malaita 2.8 97.1 -27.8 11.3 88.7 -222.9 Makira-Ulawa 3.9 96.0 -9.8 17.3 82.7 26.3 Temotu 3.8 96.1 -51.6 22.4 77.6 -8.5 Honiara 13.0 86.9 39.2 - - - Note: Recent migrants are the individuals who arrived in the current province (ward) from the other province (ward) during the last five years. The net migration rate is calculated as the difference between the number of people entering and leaving the province (ward) per 1,000 individuals during the last five years (2014 to 2019). Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Wards for provincial capitals are Batava (Choiseul), Gizo (Western), Buala (Isabel), Tulagi (Central), Auki (Malaita), Bauro Central (Makira-Ulawa), and Luva Station (Temotu). Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Honiara Urban Area has been a magnet for migrants most migrants to provincial capital towns originated from from all other provinces, while migration to provincial other wards within the same provinces. For instance, 60 capitals has primarily occurred within the same percent of migrants who arrived in Gizo (Western Province) provinces. Between 2014 and 2019, the majority of migrants between 2014 and 2019 came from other wards within to Honiara Urban Area originated from Malaita Province (35 Western Province. percent), followed by Western Province (15 percent), Isabel Outside of Honiara and Guadalcanal, more people left Province (8 percent), and Makira-Ulawa Province (6 percent). from wards located in remote areas with high poverty A significant portion of these migrants also moved from the rates. Regression analysis indicates that wards with higher wards of provincial capital towns (16 percent). In contrast, values on the remoteness index, which measures the level of 40 and Spatial Disparities Multidimensional Poverty  inSolomon Islands connectivity between wards, have negative net migration The majority of recent migrants to Honiara Urban Area, rates. This suggests that people are more likely to leave almost three out of four, rely on wages and salaries as remote wards that are poorly connected to other areas. their main source of income. In 2019, two-thirds of non- Additionally, wards with high levels of multidimensional migrant households in Honiara Urban Area also reported wages poverty in 2009 were more likely to experience migration and salaries as the primary income source (Figure 20). Recent outflows. Interestingly, an increase in poverty rates between migrant households in Honiara Urban Area are even more likely 2009 and 2019 is associated with lower net migration rates, to depend on wage and salary incomes, with 74 percent relying indicating a higher number of migrants leaving than entering. on this source. Only 6 percent of recent migrant households reported the sale of crops and fish as their main source of 3.2 Migration to Honiara was selective, with income. In contrast, in their origin wards, 47 percent of migrants’ characteristics better than rural households rely on the sale of crops and fish as their primary stayers and similar to Honiara residents income source. This stark contrast highlights the potential role of migration in shifting employment and income structures. Migration to Honiara Urban Area was selective but has become less so. Various individual characteristics are Figure 20.  Unlike rural stayers, about 75 percent of the associated with higher or lower probabilities of migrating to recent migrant households in Honiara Urban the Honiara Urban Area, and these characteristics have Area rely on wages and salaries remained relatively unchanged between 2009 and 2019 Comparison of the main source of income among (Figure 19). For instance, younger people have a higher migrants and other households, Honiara Urban likelihood of migration compared to older individuals. Area, 2019 (%) Moreover, individuals with higher educational attainment, particularly those with secondary or tertiary education, have 1.2 1.5 3.0 tended to migrate to Honiara Urban Area. However, the 13.7 influence of education on migration propensity has become 16.9 15.7 less significant over the period from 2009 to 2019. This 6.1 4.8 8.0 means that lower-educated people have had more chances to migrate to Honiara. As previously confirmed, individuals 7.3 tend to leave wards with higher levels of poverty. Finally, 46.9 there is no substantial difference in migration tendency between men and women after statistically controlling for other characteristics. 74.1 66.3 4.8 Figure 19.  People with higher education and/or living in poor areas tended to migrate to Honiara 29.6 Correlates with migration probability to Honiara Urban Area, 2009 and 2019 Migrants Non-migrants at destination Non-migrants at origin Female Age 25-34 Wages/salary Sale of crops/fish No income Age 35-44 Age 45-54 Own business Other income Age 55-64 Married Separated HH size 2019 Note: Migrants are those who arrived in the current locations between Education=primary 2014 and 2019. Honiara Urban Area includes urban population in Honiara Education=secondary 2009 Education=tertiary and Guadalcanal Province (see Annex A2). Poverty incidence 2009 Western Source: Staff calculations based on Population and Housing Census 2019. Isabel Central RenBell Migrants and non-migrant households have similar Malaita Makira levels of multidimensional poverty. The multidimensional Temotu poverty rates of recent migrants to major destinations are -.1 0 .1 .2 .3 .4 comparable to other households (Panel A in Figure 21). For Marginal effects on migration probability (%) example, in Honiara Urban Area, approximately 15 percent of recent migrants were multidimensionally poor in 2019, Note: The numbers are the marginal effects (%) estimated from a probit which is similar to the poverty rate of other households in regression model on the working-age individuals (age 15-64) for 2009 and 2019 separately, indicating the probability of migration associated with Honiara Urban Area (16 percent). Migrants in rural each household characteristic. Reference categories are ’age 15-24’ for Guadalcanal and other urban areas also have similar or even the age cohort, ’single’ for marital status, ’less than primary’ for education, and Choiseul for the province (origin province for recent migrants and cur- lower poverty rates compared to other households. This rent province for other individuals; Honiara and Guadalcanal are excluded pattern is consistent across other dimensions, such as access for the analysis). Poverty incidence 2009 indicates the poverty rate (0-1) in 2009 in the origin ward for recent migrants and in the current ward for to water (Panel B) and asset ownership (Panel C). The others. difference in asset ownership particularly implies that the Source: Staff calculations based on Population and Housing Censuses benefit of migration is not purely driven by the fact that access 2009 and 2019. to services is more available and affordable in urban areas. 41 In contrast, migrants tend to be better off than accounted for a reduction of 5.9 points, indicating that 72 households in their origin wards, although this percent of multidimensional poverty reductions occurred difference cannot be solely attributed to the migration within rural areas (Figure 22). Although the decrease in effect. Multidimensional poverty incidence is significantly multidimensional poverty incidence was not as dramatic in higher in the origin areas of recent migrants compared to Honiara and other urban areas, the population shifts toward their current status. For instance, approximately 35 percent these areas contributed 28 percent of the overall reduction in of people in the origin areas of recent migrants to Honiara the country, equivalent to 2.4 out of the 8.1 points. Urban Area are poor, while only 15 percent of the migrants themselves are currently poor (Panel A in Figure 21). Migrants Figure 21.  In terms of multidimensional poverty status, also have better access to water (Panel B) and higher levels of recent migrants and non-migrants are similar asset ownership (Panel C) compared to households in their in Honiara Urban Area unless they still rely origin areas. However, it is not possible to determine whether on agriculture as the main income source migrants escaped poverty after migration or if they were not Characteristics of migrants and non-migrants poor even in their origin areas due to the lack of information at origin and destination, 2019 (%) about their poverty status prior to migration. This is particularly relevant given the selective nature of migration. (A) Multidimensional poverty However, when comparing only households with 42.7 agriculture as their main source of income, recent 35.1 36.8 32.3 migrant households in Honiara Urban Area fare worse than other households in their origin wards and other 15.2 16.4 households in Honiara Urban Area. Approximately 43 percent of recent migrants who rely on agriculture as their All income source Agriculture as the main income source primary income source are classified as multidimensionally poor, which is a higher percentage compared to other Migrants Non-migrants in destination Non-migrants in origin households in their origin wards (32 percent) (Figure 21). (B) Deprived in water access Similarly, recent migrant households in Honiara Urban Area face challenges in terms of water access and asset ownership 17.1 if their primary income source is agriculture. Although less 16.0 15.6 14.3 than 10 percent of recent migrant households depend on agricultural income in the Honiara Urban Area (Figure 20), this data suggests that migration does not necessarily 3.7 3.2 provide a means for every household to escape poverty. All income source Agriculture as the main income source 3.3 Internal migration likely accounted for Migrants Non-migrants in destination Non-migrants in origin a significant share of multidimensional (C) Deprived in assets poverty reduction achieved during the decade 83.0 79.9 80.2 76.5 While multidimensional poverty mainly fell in rural 51.3 50.2 areas, internal migration accounts for 30 percent of the multidimensional poverty reduction achieved. Between 2009 and 2019, there was a moderate decline of 8.1 percentage points in the incidence of multidimensional All income source Agriculture as the main income source poverty. This change in multidimensional poverty can be Migrants Non-migrants in destination Non-migrants in origin broken down mathematically into the reductions within specific geographic areas, including Honiara Urban Area, Note: Migrants are those who arrived in the current locations between other urban areas, rural Guadalcanal, and other rural areas, as 2014 and 2019. Honiara Urban Area includes the urban population in well as the reductions due to population shifts across these Honiara and Guadalcanal Province (see Annex A2). areas.49 The decomposition analysis reveals that rural areas Source: Staff calculations based on Population and Housing Census 2019. The changes in poverty incidence between 2009 and 2019 are decomposed into intra-sectoral effects (poverty reduction within Honiara, 49 other urban areas, and rural areas), inter-sectoral effects (population movement across the locations), and interaction effects (residuals) following Ravallion and Huppi (1991). 42 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Figure 22. Internal migration accounts for nearly information on their poverty status prior to migration, the 30 percent of multidimensional poverty poverty rates of recent migrants are at least as low as those reduction of non-migrants in their destination areas. However, there Decomposition of changes in poverty incidence, are several concerns to consider. Unemployment rates in 2009-2019 urban areas were higher than in rural areas in 2019, and recent migrants in Honiara Urban Area are more likely to be Total change -8.10 unemployed after controlling for their characteristics. This is Honiara Urban Area -0.20 significant because households with unemployed members Other urban -0.10 are more likely to experience poverty in Honiara Urban Area. Rural Guadalcanal -0.59 Other rural -5.32 Additionally, migrant workers are more likely to be engaged Migration -2.41 in agricultural jobs, even when statistically controlling for 0.00 -1.00 -2.00 -3.00 -4.00 -5.00 -6.00 -7.00 -8.00 -9.00 other characteristics. Migrants who remain in agricultural occupations in Honiara Urban Area may fare worse living Contribution to poverty change (ppt) conditions compared to other households in their original Note: Numbers indicate the contribution to the changes in poverty incidence wards. In addition, high levels of gender-based violence between 2009 and 2019 from intra-sectoral effects (poverty reduction within throughout the country inhibits women in all situations, so Honiara, other urban areas, and rural areas), inter-sectoral effects (internal migration), and interaction effects (residuals) following Ravallion and Huppi women migrating to urban areas without support networks (1991). Interaction effects are not shown for presentation purposes. Honiara may face additional safety concerns.50 Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Third, the indirect effects of urbanization on poverty Source: Staff calculations based on Population and Housing Censuses reduction are also not clearly evident, and costs to 2009 and 2019. families left behind can be difficult to measure. Domestic transfers, which involve financial support from migrants to While the decomposition result suggests a sizable their rural households, can serve as an important indirect contribution of internal migration, there are several mechanism for poverty reduction through migration. caveats required. First, the migration of individuals However, the amount of domestic transfers received by rural from rural to urban areas has a selective nature, which poor individuals (excluding Guadalcanal) is relatively limited, limits the direct impact on multidimensional poverty with only around 20 percent of them receiving such transfers reduction. In theory, the migration of poor individuals and in the past 12 months.51 Therefore, the impact of domestic households from rural areas can contribute to poverty transfers from migrants on poverty reduction is not clearly reduction in their origin rural communities, although it may evident, given the data available, unless transfers were be a mere transfer of poverty from rural to urban areas. directly channeled toward, for example, investment in home However, the extent of this impact depends on the proportion improvements. There are also implications for the women of migrants who were actually living in poverty. A comparison ’left behind’ by partners traveling for work, with a risk of of the post-migration poverty status of recent migrants in violence and stigma and increased responsibilities to provide the Honiara Urban Area and non-migrants in their origin for families. There is no apparent evidence that urbanization wards reveals that migrants are generally better off. However, has facilitated a shift toward non-agricultural sectors or the it is important to note that the difference in poverty status creation of productive jobs, as reviewed below. between these two groups cannot be solely attributed to the direct effects of migration, as there is no available information Multidimensional poverty and jobs on the pre-migration poverty status of migrants. In fact, those who chose to migrate from rural to urban areas may Multidimensional poverty incidence among households have had a lower poverty rate even before migration, as it is working in the agriculture sector remained high. The evident that migrants tend to have at least higher levels of agriculture sector is characterized by three types of systems: education compared to non-migrants in their origin wards. 1) smallholder subsistence agriculture with occasional excess sales, 2) semi-commercial smallholders with Second, although recent migrants generally experience subsistence food production and deliberate market positive outcomes in the Honiara Urban Area, those who production, and 3) a small commercial sector including are unable to transition to non-agricultural jobs are plantations. Except for a palm oil plantation and a few poultry more likely to live in multidimensional poverty. Whether and pig units near Honiara, all agricultural production is migrants have escaped poverty as a result of migration smallholder farming. While the poverty rates declined depends on whether they would have remained poor if they between 2009 and 2019 across the board, the economic had not migrated. Although there is a lack of counterfactual sector in which household heads work is still clearly Sixty-four percent of women reported experiencing physical or sexual abuse by a partner in their lifetime in Solomon Islands (Ming et al. 2016). 50 A significant number of migrants sent cash transfers, as nearly 20 percent of rural households, regardless of their poverty status, received domestic 51 transfers within the last 12 months. However, the average amount of these transfers tends to be limited. Among rural recipients, 44 percent received less than SBD 500, 27 percent received between SBD 500 and SBD 1,000, and only 29 percent received more than SBD 1,000. There is no discernible difference in the average amount of transfers received between poor and non-poor households in rural areas. In rural areas outside of Guadalcanal Province, approximately 15 percent of poor households received domestic transfers within the last 12 months. Nearly 70 percent of these transfers originated from Honiara and Guadalcanal, while the remaining transfers mostly came from within the same provinces. There is no clear distinction in the average amount of transfers between Honiara/Guadalcanal and other provinces. 43 associated with their poverty levels (Figure 23). Around 38 At the ward level, there is no clear evidence that an percent of the population with household heads engaged in increase in population density has led to the creation of agriculture for pay was poor in 2009, which was higher than non-agricultural jobs. More workers are generally expected industry (25 percent) and services (16 percent). The poverty in the service sector in areas with higher population and rate among unpaid workers, who were mostly subsistence population density. However, except for wards in Honiara, farmers, was particularly high, standing at 47 percent. there is no clear relationship between population density and non-agricultural jobs at the ward level (Figure B3 in Annex B). Figure 23.  Multidimensional poverty rates are higher The changes in the working-age population between 2009 among agricultural households and 2019 have not shown any correlation with the changes in Poverty rates by household head’s economic the proportion of non-agricultural jobs at the ward level. sectors (%), 2009 and 2019 Additionally, the changes in the population with higher levels of education, such as those who have completed secondary 54.2 or tertiary education, have not been associated with an 46.7 45.2 increase in non-agricultural job opportunities 37.7 34.0 25.3 Figure 24.  No structural shift took place at the national 21.1 16.4 level, with the agricultural worker share increasing in urban areas The sectoral composition of workers, Unpaid work Agriculture Industry Services 2009-2019 (% of workers) (subsistence farmer) 100% 2009 2019 30 24 45 40 Source: Staff calculations based on Population and Housing Censuses 48 62 7 2009 and 2019. 73 69 66 12 77 10 Outside Honiara Urban Area, male, better educated, and 14 15 urban workers tend to work in the non-agricultural 14 68 sectors. Controlling for other characteristics, male workers 50 13 18 58 41 17 are nearly 10 percent more likely to engage in non-agricultural 25 16 38 18 16 jobs. Also, not surprisingly, those who completed secondary 0% 11 8 or higher education are highly likely to work in the non- 2009 2019 2009 2019 2009 2019 2009 2019 2009 2019 agricultural sectors compared to other low-educated workers. National Urban Honiara Other urban Rural Urban Area Given the better availability, urban workers are about 20 percent more likely to engage in non-agricultural jobs than Agriculture (for pay) Industry Services rural workers. The high probability in Rennell and Bellona Province relative to other provinces reflects the increased Note: Honiara Urban Area includes the urban population in Honiara and off-farm employment opportunities driven by the mining and Guadalcanal Province (see Annex A2). Only the working-age population (ages 15-64) is counted. logging boom. Source: Staff calculations based on Population and Housing Censuses The share of non-agricultural workers decreased, rather 2009 and 2019. than increased, between 2009 and 2019. The share of agricultural workers—which includes only those who work for pay—increased in rural areas (Figure 24). This appears to be due to some farmers switching from subsistence-based to market-based agriculture, as reflected in the reduction in the share of subsistence farmers. Also, there has been a notable shift in urban employment towards agriculture, with the percentage increasing from 11 percent in 2009 to 25 percent in 2019. This shift is observed in both Honiara and other urban areas. Shifts to non-agricultural sectors are not observed at the ward level, either (Figure 25).52 Between 2009 and 2019, the share of workers in the non-agricultural sectors increased more than 10 percentage points only in 11 out of 183 wards. 52 These wards are located in Western Province (4 wards), Makira-Ulawa Province (3 wards), Isabel Province (2 wards), Guadalcanal Province, and Malaita Province. Service sector jobs increased in most of these cases. 44 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Figure 25.  Only 11 wards experienced more than 10 percentage point increases in non-agricultural worker shares Ward-level map of changes in non-agricultural worker shares, 2009-19 Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. More people converted from subsistence to commercial worker household heads are very low, 13 percent for public agriculture in areas with better market access, which wage workers and 22 percent for private wage workers (Figure reduced multidimensional poverty significantly. A ward- 27). In contrast, as observed in the previous section, the level regression analysis suggests that the share of workers poverty rates among subsistence farmers tend to be high, with subsistence agriculture decreased between 2009 and followed by household heads who were either unemployed or 2019 in areas closer to populous wards, where markets are not active in the labor force. Interestingly, the poverty rate likely to be located. In addition, education also played a key among households with self-employed heads remained role, as the share of subsistence agricultural workers is more relatively high (35 percent), with no meaningful poverty likely to have declined in areas where the educational reduction between 2009 and 2019. attainment of working-age adults improved. The proportion In tandem with the rise in labor force participation and of unpaid worker heads (mostly subsistence farmers) employment rates, unemployment rates increased both decreased from 46 percent in 2009 to 31 percent in 2019, in urban and rural areas. The labor force participation rates and poverty rates among them substantially decreased from increased by 7.7 percentage points at the national level 54 percent in 2009 to 47 percent in 2019, accounting for a between 2009 and 2019 (Table 12).54 The employment to large portion of the poverty reduction achieved in the country the working-age population ratio slightly rose during the during the decade (Figure 26).53 same period, mainly in rural areas. While these changes Households with wage-worker heads are less likely to be indicate certain job growth, both urban and rural poor. The poverty rates among the population with wage- unemployment rates also went up by 3 to 4 percentage points. Whether female agricultural heads benefited similarly is unclear, as a previous work (World Bank 2019) indicates that government programs aimed at 53 improving agricultural production and pest control skills found it difficult to reach women. The ILOSTAT, which is referred to by the World Development Indicators as well, indicates the labor force participation rate to be 85.4 percent in 2019 54 in Solomon Islands. This is different from the number in Table 10 as it is a model-based estimate using the 2012/13 HIES and includes subsistence farmers who do not work for pay as active labor forces. 45 Figure 26.  Multidimensional poverty reduction was Figure 27.  Multidimensional poverty incidence is lower faster among subsistence farmers among households whose heads are wage Decomposition of changes in poverty workers incidence, 2009-2019 Poverty rates by household head’s economic sectors (%), 2009 and 2019 Total change -8.1 Agriculture -1.3 2009 2019 54 Industry -0.5 49 45 47 37 38 41 37 Services -0.8 33 36 35 28 28 Unpaid work -3.6 22 17 13 Not working -0.4 Sectoral changes -2.3 Employer Self-employment Wage (public) Wage (private) Other work for sale Unemployment Unpaid work Other inactive 0.0 -1.0 -2.0 -3.0 -4.0 -5.0 -6.0 -7.0 -8.0 -9.0 Contribution to poverty change (ppt) Note: Numbers indicate the contribution to the changes in poverty incidence between 2009 and 2019 from intra-sectoral effects, inter-sectoral effects, and interaction effects following Ravallion and Huppi (1991). Source: Staff calculations based on Population and Housing Censuses Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 2009 and 2019. Table 12.  Labor force participation and employment rates increased, and so did unemployment rates Labor force trends (ILO approach), 2009-19 (%) Labor force Employment rate Unemployment rate participation rate 2009 2019 Diff. 2009 2019 Diff. 2009 2019 Diff. National 28.8 36.5 7.7 27.4 31.5 4.1 1.5 5.0 3.6 Urban 45.2 48.8 3.6 41.4 41.9 0.5 3.8 6.9 3.1  Honiara 44.9 48.6 3.7 40.9 41.3 0.5 4.0 7.3 3.2 Urban Area   Other urban 46.1 50.0 3.8 43.5 45.2 1.7 2.7 4.8 2.1 Rural 24.0 30.8 6.8 23.2 26.7 3.5 0.8 4.1 3.3  Rural 26.9 33.4 6.6 26.1 28.4 2.3 0.8 5.0 4.3 Guadalcanal   Other rural 23.3 30.1 6.8 22.5 26.2 3.7 0.8 3.9 3.1 Note: Only the working-age population is included. Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Source: Staff calculations based on Population and Housing Census 2019. In Honiara Urban Area, unemployment rates are unemployed, even after controlling for other characteristics. particularly high among youth, individuals with low Interestingly, except for tertiary education, having higher levels of education, and recent migrants. A comparison of levels of education, such as incomplete or complete the unemployment status of working-age individuals in the secondary education, does not significantly improve the Honiara Urban Area reveals that those in the age group of 25 employment prospects of migrants. to 34 are more likely to be unemployed, with approximately Unemployment and job status are strongly linked to 10 percent of both male and female individuals in this age multidimensional poverty in the Honiara Urban Area. group being unemployed in 2019. Furthermore, individuals Households with an unemployed head have 22 percent with lower levels of educational attainment are more likely to higher odds of experiencing poverty compared to other experience unemployment. No significant difference was households, even after accounting for other characteristics. observed between males and females in terms of Additionally, even if households have employed heads, those unemployment rates. Regression analysis also indicates that engaged in agriculture and industry jobs still have a higher recent migrants are approximately 7 percent more likely to be 46 and Spatial Disparities Multidimensional Poverty  inSolomon Islands likelihood of being poor. Recent migrants and female-headed Area. Access to sanitation has already started deteriorating, households are not clearly associated with poverty after which has serious implications for public health in crowded controlling for other characteristics. areas—particularly in view of natural hazard risks intensified by climate change. If this situation remains unchanged and The increasing share of agricultural workers in Honiara more people migrate to Honiara, it will exacerbate congestion implies that there is insufficient job creation. Comparing costs, outweighing the benefits of urbanization, leading to the characteristics of workers in the Honiara Urban Area, further increases in poverty. In the short term, it is critically there is a higher likelihood for female workers, low-skilled important for the government to address immediate workers, and/or recent migrants to be employed in the economic infrastructure and service needs in the fastest agricultural sector. Female workers have a 16 percent higher growing urban areas. Access to reliable and affordable probability of working in the agricultural sector compared to electricity, as well as prompting own generation renewable male workers (Figure 28). Similarly, older workers and energy, such as through rooftop solar that can also feed back individuals with lower levels of education tend to be engaged into the grid, can help mitigate urban poverty. in agricultural jobs in Honiara Urban Area. Recent migrants also have a higher likelihood of being employed in agricultural In the medium term, it is crucial to lay the foundations jobs, at 3.3 percent. Therefore, it appears that the limited for sustainable urban development and service delivery availability of economic opportunities forced disadvantaged through improved urban management capacity and workers to engage in agricultural work. tools for planning and enforcement, land administration, and revenue mobilization. Urban expansion over the past three decades has been mostly unplanned and informal. Figure 28.  Female, low-skilled, recent migrants are Existing plans are outdated, not risk-informed, and more likely to be agricultural workers in enforcement capacity is weak. Interjurisdictional coordination Honiara Urban Area is needed to manage spillover effects from Honiara to peri- Correlates with agricultural employment, urban areas in Guadalcanal Province. Also, the outdated land Honiara Urban Area, 2019 registry, manual cadastral mapping, and limited property registration hinder land administration, planning, revenue Female Age 25-34 collection, and the business environment. Local governments Age 35-44 face a “service delivery trap” where limited resources Age 45-54 Age 55-64 constrain service delivery, and poor service coverage hampers Married revenue collection. Land-based revenue has significant Separated Education=primary potential, with only 40 percent of Honiara properties currently Education=secondary on tax rolls. It is crucial to modernize land administration Education=tertiary Education=other and revenue systems for broader coverage, efficiency, Recent migrants transparency, and improved public engagement. Household head HH member working in agriculture Sustained poverty reduction cannot be achieved without HH member working in non-agriculture strong economic growth. The World Bank Country -.2 -.1 0 .1 .2 Economic Memorandum suggests that without reforms, Marginal effects on agricultural employment probability (%) Solomon Islands cannot obtain upper-middle-income status by 2050.55 To spur economic growth by overcoming economic Note: The numbers are the marginal effects estimated from a probit geography constraints, it emphasizes the importance of regression model on the working-age employed individuals (age 15-64) in Honiara Urban Area in 2019, indicating the probability of working in the digitalization, improved transport connectivity, better urban agricultural sector associated with each household characteristic. Refer- planning, an enabling business environment, and education. ence categories are ’age 15-24’ for age cohort, ’single’ for marital status, The findings of this report suggest that these are important and ’less than primary’ for education. for accelerating poverty reduction as well. Facilitating job Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. creation is particularly crucial to fully harness the potential of urbanization in Honiara. A significant proportion of workers, particularly females, those with lower levels of education, and 3.4 Strategic urban planning and service recent migrants, are engaged in agricultural jobs in the area. delivery is essential to harness Addressing both demand-side and supply-side constraints is urbanization for poverty reduction vital to facilitate job creation in the Honiara Urban Area. On The findings underscore the need for the government to the demand side, it is important to modify or remove ensure that service delivery keeps up with rapid urban regulations that impose excessive costs on starting new population increases in order to prevent a rise in urban businesses and hiring additional employees. On the supply poverty. Although multidimensional poverty incidence side, there is a need to address the lack of skills among the remains low, the number of individuals living in multidimen­ labor force. sional poverty has rapidly increased in the Honiara Urban 55 World Bank (2024). 47 Chapter 4. Government Policies and Investments Towards Rural Areas Finally, this chapter investigates the role of the government in achieving reductions in poverty and spatial disparities. To assess its contribution made between 2009 and 2019, the chapter focuses on the CDFs and three WB-funded projects that supported housing and water improvements in rural areas. Also, synthesizing all the chapters, the report discusses policy implications to accelerate reductions in poverty and spatial inequality. Key findings Government investments contributed to poverty reduction, though their efficiency and effectiveness are not necessarily clear. A massive amount of the government budget has been allocated to the CDF for rural development, and a large proportion of the CDF was spent on housing. It appears that housing deprivation was indeed reduced in areas where more people positively evaluated CDF support. The government also heavily invested in household-and community-level water infrastructure, though the data indicated only moderate improvement in water deprivation. Together, the three WB-funded projects improved water access for 220,000 rural people in 150 wards between 2007 and 2022. However, the proportion of the rural population in water deprivation declined only from 34 percent to 23 percent. Going forward, accelerating rural investments is a key to poverty reduction. Achieving poverty reduction will require further rural investments, as well as the implementation of service delivery mechanisms that ensure ongoing maintenance and support. In rural infrastructure development, geographic targeting should be carefully considered given the spatial concentration of high-poverty areas. Furthermore, supporting subsistence farmers would be an effective approach to poverty reduction. A large proportion of rural households are still subsistence farmers, and their poverty incidence is particularly high. Supporting them to work for pay, rather than their own consumption, could result in a significant impact on poverty reduction. Those supports may include, for example, improving access to markets and information. 4.1 CDF likely contributed to staff (from selection to day-to-day management) and how multidimensional poverty reduction effective and sustainable the expenditure is. One of the challenges is rooted in the way CDF expenditure is reported, As explained in the World Bank Public Expenditure with funds recorded as spent at the time of the transfer to Review, comprehensive analysis of CDFs is difficult MPs and not at the point of execution, thus obscuring the because of the persisting gaps in how these funds are utilization of funds. However, it is understood that a sizable managed.56 The CDF is by far the largest mechanism for share of CDF resources is used for local infrastructure. subnational public funding and investment—more than 3 percent of GDP between 2009 and 2019 (Figure 5 in Chapter The limited evidence that does exist suggests that a 1)—but mostly operates independently from the provincial majority of resources are oriented toward spending on administrations.57 There is little information as to how CDF housing. For the latest publicly available information on CDF resources are used. The lack of data contributes to a poor spending, in 2015 there was some variation in the pattern of understanding of how CDFs are managed by MPs and their World Bank (2022a). 56 In addition to the CDFs, the Provincial Capacity Development Fund (PCDF) provides another key subnational funding mechanism designed to respond 57 to the priorities of local communities. It is significantly smaller in size than CDFs but presents the largest source of finance for capital expenditure by provincial governments. The two funds operate largely independently of each other with no structures to ensure collaboration, resulting in little coordination between the two funding mechanisms. 48 and Spatial Disparities Multidimensional Poverty  inSolomon Islands spending across constituencies. However, housing projects Figure 29.   About 40 percent of CDFs was spent were consistently reported as a large share of spending, on housing making up over 40 percent of the funds spent, while only 4 Share of CDF spending, 2015 percent was spent on water and sanitation despite it being a priority and a key infrastructure gap (Figure 29). The General Assistance, 3 government explains that under the recently approved new Rural electrification, 3 CDF Act that introduces various reforms mainly on the WASH, 4 governance of the program, the allocations and utilizations of funds are to be used more toward economic sectors without Basic infrastructure, 5 neglecting essential services and other areas.58 Source: SIG and UNDP (2018). In the 2019 Population and Housing Census, only about Administration, 13 Housing scheme, 42 one-third of households reported a positive contribution of the CDFs to their lives. The 2019 Census included a question about the main area of development assistance in which the CDFs have positively contributed to households. Transportation, 14 Income Interpreting these answers requires caution as each generation, 16 household was asked to choose only one area, even if the household received benefits in multiple areas. Also, it is important to know that among the households that responded to the question, 64 percent answered “none”.59 The most common positive responses were related to housing materials (19 percent), supplies for solar energy (12 percent), and education fees and scholarships (1.4 percent) (Table 13). The relatively high proportion of households choosing housing is consistent with the CDF budget allocation (Figure 29). Table 13.  People reported that CDFs contributed to housing and solar energy Percentage of households answering “What is the main area of development assistance that the CDF has contributed positively to your household?” Solar Housing Transport Education Cash Water None energy National 18.8 12.2 1.1 1.4 0.9 0.8 64.2 Urban 12.2 4.2 0.4 2.2 1.1 0.8 78.7   Honiara Urban Area 11.7 3.5 0.4 2.3 1.2 0.6 79.8   Other urban 14.8 7.3 0.5 1.3 0.6 2.0 72.8 Rural 21.0 14.9 1.4 1.1 0.8 0.8 59.3   Rural Guadalcanal 24.2 14.1 0.8 0.9 0.9 0.3 58.4   Other rural 20.1 15.1 1.5 1.2 0.8 1.0 59.6 Choiseul 18.3 21.5 0.6 0.9 0.6 1.5 55.4 For example, 40 percent to productive and resources sectors, 20 percent to essential services, 20 percent to cross-sectional and inclusivity, and 20 58 percent to social and cultural obligations. See https://solomons.gov.sb/new-cdf-law-enforced/ While only 36 percent of households expressed positivity about the CDF’s contribution to their lives, rural households were twice as likely to do so 59 compared to urban households. 49 Western 20.1 14.3 2.2 1.0 0.6 1.6 59.6 Isabel 28.4 13.3 1.6 0.6 0.6 0.4 54.2 Central 11.4 8.5 3.5 0.6 0.5 1.4 73.0 Rennell and Bellona 20.6 18.3 2.9 3.3 0.0 2.1 53.7 Guadalcanal 21.1 11.6 0.7 1.0 1.0 0.3 64.0 Malaita 22.1 14.2 0.8 1.4 0.9 0.9 59.3 Makira-Ulawa 12.0 13.8 1.3 1.9 1.3 0.3 67.6 Temotu 16.8 20.2 1.6 0.8 0.4 1.9 57.4 Honiara 11.8 3.5 0.4 2.7 1.2 0.6 79.4 Note: Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province (see Annex A2). Source: Staff calculations based on Population and Housing Census 2019. Wards with a higher proportion of people expressing poorer wards, and people could not have installed solar power positivity about the CDFs have experienced stronger without CDF support there. improvements in certain outcomes. Regression results Nevertheless, allocating a vast amount of public suggest that the proportion of households in a ward that investment into housing through the CDFs may be reported positive answers to the CDF contributions, or the neither an efficient nor effective approach to poverty CDF positive rate, is associated with improvements in poverty, reduction. According to the World Bank Public Expenditure education, housing, sanitation, and electricity. For instance, a Review, CDF spending is not in line with the National 10 percent higher CDF positive rate was associated with a 0.5 Development Strategy and infrastructure plans.60 Projects percent lower poverty rate, though the interpretation of this funded under the CDF mechanism do not follow a systematic result requires caution, as it does not indicate any causal or standardized appraisal or selection process. Strategically relationship (Figure 30). However, no clear association was reallocating public investment could achieve greater poverty found between the CDF-positive rates and changes in water reduction, and it is crucial that the recently started CDF deprivation. reforms will ensure such functions. Housing deprivation declined in the wards where a higher proportion of households expressed benefits Figure 30.  Multidimensional poverty improved more from CDF. A 10 percent higher CDF positive rate was in places where more people positively associated with a 1.5-point lower housing deprivation rate evaluated the CDFs (Figure 30). This result appears to correspond to a large Association between CDF positive rates and amount of CDF allocation to housing development, as shown changes in poverty outcomes at the ward level in Figure 29. Given that housing was the MPI indicator with the most significant reduction between 2009 and 2019, it % change in the outcome associated with a 10% increase in CDF positive rate appears that CDF’s contribution to multidimensional poverty -2 -1.5 -1 -0.5 0 0.5 reduction was sizable. Changes in poverty Access to solar power improved in the wards with higher Changes in education deprivation Changes in housing deprivation CDF positive rates. Next to housing, the largest share of Changes in water deprivation households listed solar power as the area of development Changes in sanitation deprivation supported by the CDFs (Table 13). The regression results also Changes in electricity deprivation imply that household access to solar power improved in the wards with higher CDF positive rates. However, it is difficult to Note: Bars indicate the percent change in each outcome associated with a attribute the remarkable improvement in solar-powered 10 percent increase in the CDF positive rate based on a ward-level regres- energy access between 2009 and 2019 solely to the CDFs. sion analysis. Error bars indicate 90 percent confidence intervals. This is because solar-powered electricity access has become Source: Staff calculations based on Population and Housing Censuses almost universal in the country due to improvements 2009 and 2019. everywhere, regardless of CDF-positive rates. It is still possible that the CDFs were more likely to be allocated to World Bank (2022a). 60 50 and Spatial Disparities Multidimensional Poverty  inSolomon Islands 4.2 The government heavily invested in hazards and climate change risks. One of the project household water access, though water components was climate change adaptation and disaster risk deprivation was reduced only moderately reduction investments (US$5.5 million), which involved the development of water supply facilities at the community and In addition to rural development through the CDFs, the provincial levels in Central, Rennell and Bellona, Guadalcanal, government also implemented various donor-funded Malaita, and Temotu provinces. The number of beneficiaries projects that potentially drove poverty reduction. This was estimated to be nearly 50,000. section focuses on household access to improved water sources for drinking, which was the area contributing to The RDP (I and II) and the CRISP together covered most multidimensional poverty reduction next to housing. After rural wards (152 wards), improving water access among reviewing the inventory of the past projects funded by donors approximately 220,000 people (Figure 31). As this is the (WB, ADB, JICA, and DFAT) that were active at any point total number of beneficiaries supported by the projects between 2009 and 2019, three projects were identified as between 2007 and 2022, the number of beneficiaries who those invested in household-level water facilities.61 improved water access thanks to these projects between 2009 and 2019 must be smaller than 220,000. The Rural Development Program (RDP) improved household water access in 134 wards between 2007 and Nevertheless, water deprivation in rural areas declined 2022. The RDP I was the WB’s first major intervention in only moderately between 2009 and 2019. According to Solomon Islands following the Tensions of 1998-2003.62 The the 2009 and 2019 Population Censuses, the proportion of US$37.4 million project (with a US$9.5 million WB loan) was the rural population in water deprivation decreased from 34.2 implemented from 2007 to 2015. The objective of the RDP percent to 22.6 percent. Considering the rural population was to raise the living standards of rural households by increase during the period, approximately 61,000 escaped establishing improved mechanisms for the delivery of priority water deprivation.65 This number is far smaller than the economic and social infrastructure and services through: 1) number of beneficiaries from the water investment through increased, cost-effective, and sustained provision of local the RDP and CRISP. services and basic infrastructure determined by participatory There are several potential reasons why only moderate planning prioritized by villagers; 2) increased capacity of improvement in water access was evident in the agricultural institutions to provide demand-driven agricultural Population Censuses. First of all, as these projects spanned services at the local level; and 3) support for rural business from 2007 to 2022, some people had already received development. The RDP I invested US$21.6 million to support benefits prior to the 2009 Census, and some people had not the development of water facilities, such as gravity-fed received benefits by the time of the 2019 Census. Secondly, systems and rainwater catchment systems, in 112 villages in it is also possible that many beneficiaries already had 79 wards across provinces except for Rennel and Bellona, and improved water access, and they simply replaced it. A study Honiara.63 The number of beneficiaries was reported to be that visited various RDP-targeted communities found that more than 80,000.64 As a continuation, RDP II further water facilities had been developed to replace water sources invested SBD 33.9 million in water access between 2015 and contaminated by logging operations.66 Thirdly, there might be 2022, benefiting about 84,000 people in 160 villages in 103 cases in which water facilities installed through the projects wards across provinces except for Honiara. RDP I and II were not functioning due to lack of maintenance. There is a together covered 272 villages in 135 wards in the country. possibility that new water facilities became useless over time The CRISP also made a significant investment in as logging operations expanded further into the interior, improving household water access in 38 wards across making secondary catchments and water piping from more five provinces. This US$9 million project was implemented distant and unimpacted water sources unusable.67 Finally, it from 2014 to 2020, with the objective to increase the could be a case that neither the project documents nor the capacity of selected rural communities to manage natural Census—or both of them—measured the outcomes and number of people properly. There were no other projects that involved water access improvements of rural households between 2009 and 2019. For example, JICA’s project for 61 improvement of water supply system targeted only urban areas (Honiara and Auki). World Bank (2016). 62 The most popular water interventions were rain catchment systems constituting gutters, pipes, and polytanks connected to anywhere between 4 and 63 30 public tap stands per community. The average tank size for such interventions was 10,000 liters per tank. Other water systems chosen in RDP include wells/boreholes, and local piped water systems, either gravity-fed or powered by solar/diesel pumps (Anderson 2023). A 2013 evaluation survey of 80 communities from the first four provinces where the project was active found that 71 percent of households had 64 satisfactory access to water compared to 33.3 percent before it started in 2008. https://www.worldbank.org/en/news/feature/2015/04/14/solomon-is- lands-empowering-communities-to-access-clean-water Calculated as the difference between 179,000 (the 2009 water deprivation rate applied to the 2019 rural population) and 118,000 (the 2019 water 65 deprivation rate applied to the 2019 rural population). Anderson (2023). 66 ibid. 67 51 Figure 31.  RDP (I & II) and CRISP invested in household water access in the majority of rural wards Wards targeted by RDP (I & II) and CRISP Source: Author work based on WB project documents. 4.3 Reducing concentrated rural poverty re- Going forward, accelerating rural investments to quires rural investments to be more effi- improve living standards—such as sanitation and cient and effective energy—remains crucial to multidimensional poverty reduction. Multidimensional poverty declined between While this report focuses on a period from 2009 to 2019 2009 and 2019, mainly among rural subsistence farmers. due to data availability, multidimensional and monetary The CDFs and other government investments supported poverty must have experienced a surge due to multiple improvements in, among others, housing and water, which recent shocks. While recent estimates of monetary poverty appear to have contributed to poverty reduction. Despite the are unknown due to the lack of data, the COVID-19 pandemic convergence in non-monetary outcomes across the provinces and civil unrest in 2021 likely affected the well-being and and wards during the decade, rural poverty remains relatively livelihoods of wide swathes of the population, with worsening high. A large portion of the rural population is still deprived in income, food security, education, and health indicators. For the living standard dimension, particularly sanitation and example, in 2022—during COVID-19 lockdowns— energy access. Inadequate access to these services employment fell from around 60 percent to 40 percent constrains health and human capital development, thereby according to the World Bank’s high-frequency phone lowering the labor productivity and incomes of both the surveys.68 With limited coping mechanisms and social current and future generations. Therefore, it is imperative to protection systems, such economic shocks can easily accelerate rural investments to improve their living standards. translate to income reduction and food insecurity. While the CDFs may have contributed to improvements in World Bank (2022b). 68 52 and Spatial Disparities Multidimensional Poverty  inSolomon Islands these areas, achieving multidimensional poverty reduction Supporting subsistence farmers in engaging in markets will require further rural investments, as well as the would be an effective approach to poverty reduction. A implementation of service delivery mechanisms that ensure large proportion of rural households are still subsistence ongoing maintenance and support. Promoting renewable farmers, and their poverty incidence is particularly high. energy, particularly solar mini-grids and off-grid, through Although it is unrealistic to facilitate their conversion to establishing adequate regulatory and institutional industry and service sector jobs, supporting them to work for frameworks would also be effective, lowering energy costs pay rather than their own consumption can still have a and promoting household access to electricity. significant impact on poverty reduction. The difference in their poverty rates is nearly 10 percentage points as of 2019. It is crucial to make rural infrastructure development Those supports may include, for example, improving access more effective and efficient so as not to squeeze budget to markets and information through transport infrastructure for other critical components for multidimensional to facilitate their transformation process to semi-commercial poverty reduction. For example, the World Development and/or commercial. As discussed in the previous chapter, Report 2009 highlighted the importance of a spatially blinded wards with better market access were more likely to have approach at the nascent stage of urbanization, which includes experienced a shift from subsistence to commercial the establishment of institutional foundations (such as land agriculture since 2009. The World Bank Country Economic policies) and basic service provisions in rural areas (including Memorandum recommends, among others, mechanization of remote areas).69 The RDP followed such an approach, smallholder farming with appropriate technologies and tools, targeting vast swathes of rural areas. Given the promoting ICT tools to share and receive information, and multidimensional poverty prevailing across rural areas, reducing the cost of doing business to facilitate private- attempting to cover all rural areas may still appear to be valid. sector investment.71 The RDP (2009-2022) and the ongoing However, the efficiency and effectiveness of a whole-rural Solomon Islands Agriculture and Rural Transformation approach should be carefully assessed. As revealed in Chapter Project (SIART) support subsistence smallholder farmers by 2, two-thirds of poor Solomon Islanders are concentrated in facilitating their semi-commercial/commercial agriculture Honiara, Guadalcanal, and Malaita, and wards with high activities with an aim to improve their livelihoods and well- poverty incidence are concentrated in Makira-Ulawa and being. Unlike the development of basic services, agricultural Temotu as well (Figure 9). Targeting those wards with high investments should be spatially differentiated/targeted by poverty (and large populations) with public investment could considering local products, market access, etc. speed up poverty reduction. The Solomon Islands National Infrastructure Investment Plan combines holistic and sectoral plans to develop and prioritize a list of projects, but it is unlikely to lead to effective geographic targeting.70 World Bank (2009). 69 World Bank (2022a). Its multi-criteria analysis framework evaluates projects based on 14 criteria, one of which is poverty: “Will the scheme provide 70 affordable transport, energy, water supply, and sanitation infrastructure to support socio-economic development and income generation?” and “What is the total population served by the scheme?” These are evaluated qualitatively due to lack of data. World Bank (2024). 71 53 References Albert, J., J. Bogard, F. Siota, J. McCarter, S. Diatalau, J. Fleming, L., C. Anthonj, M. B. Thakkar, W. M. Tikiosuva, M. Maelaua, T. Brewer, and N. Andrew. 2020. “Malnutrition Manga, G. Howard, K. F. Shields, E. Kelly, M. Overmars, and J. in rural Solomon Islands: An analysis of the problem and its drivers.” Maternal & Child Nutrition 16, no. 2 (March): Bartram. 2019. “Urban and rural sanitation in the Solomon e12921. Islands: How resilient are these to extreme weather events?” Science of the Total Environment 683: 331–40. Anderson, B. 2023. Community Driven Development, Climate Change, and Resiliency: Lessons from Solomon Gibson, J., G. Datt, R. Murgai, and M. Ravallion. 2017. “For Islands. Asia Pacific Issues, no. 160. Honolulu, HI: East- India’s rural poor, growing towns matter more than growing West Center. cities.” World Development 98: 413–29. Anthonj, C., J. W. Tracy, L. Fleming, K. F. Shields, W. M. Harris, J., and M. Todaro. 1970. “Migration, unemployment, Tikoisuva, E. Kelly, M. B. Thakkar, R. Cronk, M. Overmars, and development: A two-sector analysis.” American and J. Bartram. 2020. “Geographical inequalities in drinking Economic Review 60: 126–42. water in the Solomon Islands.” Science of the Total Environment 712: 135241. Hicks, J. H., M. Kleemans, N. Y. Li, and E. Miguel. 2017. “Reevaluating agricultural productivity gaps with Beegle, K., J. de Weerdt, and S. Dercon. 2011. “Migration and longitudinal microdata.” NBER Working Paper. Cambridge, Economic Mobility in Tanzania: Evidence from a Tracking MA: National Bureau of Economic Research. Survey.” Review of Economics and Statistics 5 (8): 1010– 1033. Lagakos, D. 2020. “Urban-rural gaps in the developing world: Does internal migration offer opportunities?” Journal of Bryan, G., S. Chowdhury, and A. M. Mobarak. 2014. Economic Perspectives 34, no. 3 (Summer): 174–92. “Underinvestment in a Profitable Technology: The Case of Seasonal Migration in Bangladesh.” Econometrica 82 (5): Lucas, R. E. B. 2016. “Internal migration in developing 1671–48. economies: An overview of recent evidence.” Geopolitics, History, and International Relations 8, no. 2: 159–91. Castells-Quintana, D. 2017. “Malthus living in a slum: Urban concentration, infrastructure and economic growth.” Journal Ming, M. A., M. G. Stewart, R. E. Tiller, R. G. Rice, L. E. Crowley, of Urban Economics 98: 158–73. and N. J. Williams. 2016. “Domestic violence in the Solomon Islands.” Journal of Family Medicine and Primary Care 5, no. Christiaensen, L., and Y. Todo. 2014. “Poverty reduction 1 (January-March): 0–16. during the rural–urban transformation–the role of the missing middle.” World Development 63: 43–58. Ministry of Health and Medical Services. 2014. The Solomon Islands Rural Water Supply, Sanitation and Hygiene Policy. Commonwealth Local Government Forum (CLGF). 2018. Honiara, Solomon Islands: Ministry of Health and Medical Commonwealth Local Government Handbook 2017/18. Services. London: Commonwealth Local Government Forum. Ravallion, M., and M. Huppi. 1991. “Measuring changes in Combes, P. P., C. Gorin, S. Nakamura, M. Roberts, and B. poverty: A methodological case study of Indonesia during Stewart. 2023. “An anatomy of urbanization in Sub-Saharan the adjustment period.” The World Bank Economic Review 5, Africa.” World Bank Policy Research Working Paper 10621. no. 1: 57–92. Washington, DC: World Bank. Selod, H., and F. Shilpi. 2021. “Rural-urban migration in Egger, D., J. Haushofer, E. Miguel, P. Niehaus, and M. Walker. developing countries: Lessons from the literature.” Regional 2022. “General equilibrium effects of cash transfers: Science and Urban Economics 91: 103714. Experimental evidence from Kenya.” Econometrica 90, no. 6 (November): 2603–43. Solomon Islands Government (SIG) and United Nations Development Programme (UNDP). 2018. Solomon Islands Development Finance Assessment. 54 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Solomon Islands National Statistics Office (SINSO). 2021. World Bank. 2017. Solomon Islands Systematic Country 2009 Population and Housing Census: Report on Migration Diagnostic: Priorities for Supporting Poverty Reduction & and Urbanisation. Honiara, Solomon Islands: SINSO. https:// Promoting Shared Prosperity. Washington, DC: World Bank. solomonislands-data.sprep.org/dataset/report-migration- and-urbanisation. World Bank. 2018. Country Partnership Framework for Solomon Islands for the period FY2018-FY2023. Solomon Islands National Statistics Office (SINSO). 2023. Washington, DC: World Bank. 2019 Population and Housing Census: National Report (Volume 1). Honiara, Solomon Islands: SINSO. World Bank. 2019. Enhancing the Economic Participation of Vulnerable Young Women in Solomon Islands. Washington, Solomon Islands National Statistics Office (SINSO), DC: World Bank. Ministry of Health and Medical Services (MHMS), and Pacific Community (SPC). 2017. Solomon Islands Demographic and World Bank. 2021. Archipelagic economics: Spatial economic Health Survey 2015. development in the Pacific. Washington, DC: World Bank. Solomon Islands National Statistics Office (SINSO) and World Bank. 2022a. Solomon Islands Public Expenditure World Bank. 2015. Solomon Islands Poverty Profile based on Review: Fiscal Reform and the Path to Debt Sustainability. the 2012/13 Household Income and Expenditure Survey. Washington, DC: World Bank. Solomon Islands National Statistics Office (SINSO) and World Bank. 2022b. Socio-economic impacts of COVID-19 World Bank. 2017. Solomon Islands Poverty Maps based on in Solomon Islands: Insights from High Frequency Phone the 2012/13 Household Income and Expenditure Survey Surveys, September 2022. Washington, DC: World Bank. and the 2009 Population and Housing Census. https://openknowledge.worldbank.org/entities/publication/ a608c7e6-2096-4392-a9ba-027b011bbbac. UNDP (United Nations Development Programme) and OPHI (Oxford Poverty and Human Development Initiative). 2023. World Bank. 2023. Local Responses to Climate Change and Global Multidimensional Poverty Index 2023: Unstacking Disaster-Related Migration in Solomon Islands. Washington, global poverty: Data for high-impact action. New York: UNDP; DC: World Bank. Oxford: OPHI. World Bank. 2024. Solomon Islands Country Economic World Bank. 2009. World Development Report 2009: Memorandum: Unlocking New Sources of Economic Growth. Reshaping Economic Geography. Washington, DC: World Washington, DC: World Bank. Bank. World Health Organization. 2021. Global Database on the World Bank. 2016. Project Performance Assessment Report: Prevalence of Violence Against Women. Geneva: World Health Solomon Islands Rural Development Program. Washington, Organization. DC: World Bank. 55 Annex A. Technical note A1. Multidimensional poverty According to the nine indicators, each person is assigned a measurement deprivation score based on their household’s deprivation. The health dimension has one indicator. The education dimension Following the global multidimensional poverty approach in comprises two indicators, each weighted as one-sixth. Six UNDP and OPHI (2023), this report measures indicators capture the standard of living, each weighted as multidimensional poverty consisting of three dimensions: one-eighteenth. Household deprivation score is calculated by health, education, and living standards. The maximum aggregating the score for each indicator. A threshold of one- deprivation score is 100 percent, and each dimension is third is used to differentiate the multidimensionally poor weighted equally at 33.3 percent or (Table A1). These from the non-poor. Thus, if the household deprivation score is dimensions can be subdivided into nine indicators one-third (33 percent) or higher, everyone in that household emphasizing the ease of achieving an essential quality of life. is categorized as multidimensionally poor. Table A1.  MPI components in this report Indicator Dimension Dimension Indicator Deprived if… weight weight Child Health Any child has died in the household. 1/3 1/3 mortality Any school-aged child is not attending Educational school up to the age at which they would 1/6 enrollment Education complete class 8. 1/3 Educational No eligible household member has 1/6 attainment completed six years of schooling. At least one of the three dwelling elements— Housing walls, floor, or roof—is constructed using 1/18 inadequate materials. Drinking The household lacks access to improved- 1/18 water standard drinking water. The household lacks access to improved- Sanitation 1/18 standard sanitation. Standard of living Electricity The household has no electricity. 1/18 1/3 The household cooks using solid fuel, such Cooking fuel as dung, agricultural crops, shrubs, wood, 1/18 charcoal, or coal. The household does not own more than one of these assets: radio, TV, telephone, Assets 1/18 computer, animal cart, bicycle, motorbike, or refrigerator, and does not own a car or truck. 56 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Due to the limited information available in the Solomon in the household (age equivalent of grade 9 or above) have Islands Censuses, several modifications were made to the completed primary education. If a household does not have measurement of multidimensional poverty in this report. any children aged 14 or younger this indicator is not Those modifications include the dropping of a nutrition applicable and deprivation in education is determined solely indicator and the revisions in deprivation definitions for child using the adult schooling (educational attainment) indicator. mortality and drinking water. Detailed explanation about With respect to living standards, a household is considered each indicator is as follows. deprived if it is deficient in any one of the five indicators: Nutrition is the most significant modification to the Solomon housing, drinking water, sanitation, electricity, cooking fuel, Islands MPI. Due to the lack of nutrition information in the and assets. 2009 and 2019 Population and Housing Censuses, the report A household is considered deprived of housing if the floor is dropped the indicator.72 The weight for the child mortality made of natural materials (mud, clay, earth, sand, or dung), if indicator was then revised from one-sixth to one-third, the dwelling has no roof or walls, or if either the roof or walls keeping the weight for the health dimension unchanged are constructed using natural or rudimentary materials such (1/3). Similar arrangements were made in some countries in as carton, plastic/polythene sheeting, bamboo with mud/ OPHI’s global MPI.73 stone with mud, loosely packed stones, uncovered adobe, Accordingly, the health dimension is derived only from child raw/reused wood, plywood, cardboard, unburnt brick or mortality: if a child has died in a household, that household is canvas/tent. Natural materials like traditional materials, considered to be health deprived. Common practice for the makeshift or improvised materials, and other materials are measurement of global MPI uses child loss five years prior to considered to be inadequately constructed. the survey.74 While the Solomon Islands Censuses have A household is considered deprived of drinking water if it information about the number of child deaths, they do not does not have access to improved standard of drinking water specify any time reference. As such, women (ages 15 to 49) (according to SDG guidelines).75 Specifically, households with may have reported child deaths more than five years ago. unprotected wells, surface water, river/stream, or other are Lacking the 5-year reference could result in overestimating considered deprived of improved standard of drinking water. child mortality. In contrast, metered-SIWA, protected well/spring, rainwater Nevertheless, the lack of the 5-year reference in the child collection, or bottled water are considered improved sources mortality indicator is unlikely to change the MPI measures of water. This deprivation definition is different from the substantially. As shown in Figure 13 in Chapter 2, most of OPHI’s global MPI methodology that adopts the SDG’s ’basic’ those who are deprived of the health dimension are also access standard, which incorporates both ’improved’ water deprived of the living standard dimension. More specifically, access and 30-minute proximity criteria. The Solomon while 7.4 percent of the national population was deprived of Islands Population and Housing Census includes the the health dimension in 2019, only 0.3 percent of the national proximity information only in the 2019 data. To maintain population was deprived of only the health dimension. As longitudinal comparability, this report sticks to the improved such, even if the deprivation in the health dimension is water access standard. Removing the 30-minute criteria overestimated due to the removal of a 5-year reference in could result in an underestimation of water deprivation. child mortality, it is unlikely to affect the overall MPI scores To assess the magnitude of the underestimation, Table A2 meaningfully. compares the deprivation rate of drinking water, the Education dimension depends on two factors: educational multidimensional poverty rate, and MPI with and without the enrollment and attainment. A household that has either one 30-minute criteria. The results show that the deprivation rate or both of the indicators applied is considered education for drinking water is lower by 4.1 percentage point if the deprived. For educational enrollment, a household is 30-minute criteria is not considered. This makes the considered as deprived if at least one child aged 14 or below multidimensional poverty rate lower by 0.9 percentage point, is not enrolled in school. For educational attainment, a which is not a substantial difference. household is considered deprived if none of the of the adults See SINSO, MHMS, SPC (2017) and Albert et al. (2020), which reported the prevalence of stunting among infants and young children and overweight 72 and obesity among women in rural communities in Solomon Islands. For example, 21 countries miss one indicator, and one country misses two indicators in the 2023 global MPI (UNDP & OPHI 2023). 73 UNDP & OPHI (2023). 74 This deprivation criteria based on unimproved access is different from the OPHI’s multidimensionally poverty methodology, which uses ’limited’ 75 access. This is due to the lack of information in the Solomon Islands Census data. 57 Table A2.  Changes in water deprivation based on A household is considered deprived of electricity if it does not different definitions, 2019 have access to electrification from grid electricity. Therefore, in Solomon Islands, if a household’s main source of lighting is from own generator, solar power, gas, kerosene lamp, wood/ Without With coconut, Coleman lamp, other or none, it is considered to be 30-min 30-min Diff. deprived of electrification. criteria criteria A household is considered deprived of cooking fuel if the % of population main source of cooking energy is limited to rudimentary deprived in 17.2 21.3 4.1 materials like charcoal, wood/coconut shells, other, or none. drinking water Cooking based on grid electricity, kerosene, or gas is considered adequate. Multidimensional 35.4 36.3 0.9 The household is asset deprived if it does not own more than poverty rate (%) one of these assets—radio, TV, telephone, computer, animal MPI 0.168 0.172 0.004 cart, bicycle, motorbike, or refrigerator—and does not own a car or truck. A household is considered deprived of sanitation if it does not have access to improved standard sanitation facilities To ensure comparability a decade apart, the selected (according to SDG guidelines).76 A household with none or questions from the Census questionnaires are worded either shared toilet facilities (flush, water-sealed, or pit latrine) is identically or very similarly in 2009 and 2019 rounds (Table deemed deprived of improved standard of sanitation. A3). The three dimensions of MPI are calculated by six Households with private access to toilet facilities (flush, indicators using six questions. Four of the six questions were water-sealed, or pit latrine) is considered adequate access to worded identically between the two censuses. Questions improved standards of sanitation. pertaining to access to sanitation and drinking water were worded differently between the waves, where the latter (2019) census posed the questions more descriptively. Table A3.  Questions for MPI indicators in 2009 and 2019 Census questionnaires Dimension Indicator Number Census 2009 /Census 2019 Question Health Child How many children of each sex did this woman give birth F4 mortality to who have died? Educational P16 Is the person now attending a formal educational institution? enrollment Education Educational P19 What is the highest level of education this person has completed? attainment H2 Main material used for the construction of walls? Housing H3 Main material used for the construction of floor? H4 Main material used for the construction of roof? Standard of living Electricity H10 Main source of lighting? Cooking fuel H11 Main source of cooking energy? Assets H16 Household durables? Census 2019 Question Census 2009 Question Do you share this toilet facility Sanitation H8 Main toilet facility? with others who are not members of your household? What is the main source of Drinking H6 Main source of drinking water? drinking water for members of water your household? Same as above. 76 58 and Spatial Disparities Multidimensional Poverty  inSolomon Islands The headcount ratio, H, is the proportion of those who are Figure A1.  Poverty and population density at the ward level multidimensionally poor in the total population. It is a measure of the incidence of multidimensional poverty, and (A) Monetary poverty and density can be calculated as, where q is the number of people who are multidimensionally poor, and n is the total population. The intensity of poverty, A, reflects the extent of deprivation by calculating a weighted average proportion of multidimensionally poor people (those with a deprivation score greater than or equal to 33.3 percent). (B) Non-monetary poverty and density where Si is the deprivation score that the i th multidimensionally poor person experiences. Then, the MPI value is the product of two measures: the incidence and the intensity of multidimensional poverty: MPI = H . A The relationships between population density and poverty, both monetary and multidimensional, are not clearly evident. A common concern about the use of multidimensional Source: Staff calculations based on Population Census 2009 and SINSO & WB (2017). poverty for spatial analysis is that it is by definition associated with population density; however, this is not the case for Solomon Islands. While wards in Honiara and Guadalcanal A2. Definitions of urban area and Honiara show a negative association between population density and Urban Area monetary poverty, meaning that wards with higher population Urban areas defined in the Population Censuses include density tend to have lower monetary poverty rates, this Honiara City Council and all provincial administrative centers relationship is not observed in other areas (Panel A in Figure except Rennell and Bellona, as well as some enumeration A1). For example, poverty rates vary in Malaita Province areas in Tandai and Malango in Guadalcanal.77 This definition regardless of the population density. On the other hand, non- did not change between 2009 and 2019 Censuses. monetary poverty is relatively more clearly associated with population density (Panel B), although many wards have both Analysis of urbanization and poverty should extend beyond low population density and low poverty rates. distinguishing only Honiara City Council and Guadalcanal Province in Guadalcanal Island as urban areas.78 Urban See SINSO (2023) and Combes et al. (2023) for a discussion on urban definitions from a global perspective. 77 Honiara City Council is the only local government body in Solomon Islands, established under Honiara City Act 1999 (CLGF 2018). 78 59 population growth in the capital region has already spread multidimensional poverty. Mixing it up with urban Guadalcanal beyond the boundaries of Honiara City. The urban parts of could mask the striking contrast between the two areas. As Guadalcanal Province, adjacent to Honiara City, are, in fact, such, considering adjacent urban areas of Guadalcanal the areas with the fastest population growth in the country. Province (Tandai and Malango wards) as part of Honiara Migrants have settled in these areas and are seeking Urban Area helps better analyze the actual patterns of economic opportunities in Honiara. As shown by the report, urbanization in terms of population movement, economy, and rural Guadalcanal is the most deprived area in light of access to services. Classification of locations in Guadalcanal Island Honiara City Council Guadalcanal Honiara Urban Island Area Urban areas of Guadalcanal Province Guadalcanal Province Rural areas of Guadalcanal Province A3. Measuring labor force status in A person is unemployed if he or she was not •  Population and Housing Censuses 2009 employed and actively looking for work (’yes’ to and 2019 P25) and available to work (’yes’ to P27). This report measures labor force status following the ILO A person is an active labor force if he or she was •  definition unless otherwise noted. Both the 2009 and 2019 either employed or unemployed. Censuses have the following questions to identify labor force A person is a potential labor force if he or she was •  status: not employed and either actively looking for work •  P20. During the last week, did this person do any or available to work. work? A person is not active labor force if he or she was •  •  P21. During the last week, did this person have a not employed and neither looking for work nor job at which he/she did not work? available to work. •  P22. What type of work/activity does this Unlike these ILO-based classifications, the 2019 Census person usually do? report considers individuals who were producing goods or services for their own consumption as employed. This affects •  P25. Did this person actively look for work? employment rates, labor force participation rates, and •  P27. Was this person available to work? sectoral shares, as shown in Table 2. Either approach would provide reasonable measures to analyze employment By applying the ILO definition to these questions, the labor patterns across locations and over time. The 2009 and 2019 status of a person can be determined as follows: Censuses have identical employment questions. In addition, both Censuses were collected at the same time of the year A person is employed if he or she worked (’yes’ to •  (November 22, 2009, and November 24, 2019); thus, P20) or was temporarily absent (’yes’ to P21) employment statistics are unlikely to be biased by during the last week, and the work was for pay seasonality.79 (P21 is neither voluntary work, unpaid family work, nor producing good/services for own consumption). 79 Labor force surveys are generally better suited for measuring labor statistics, though SINSO has never collected nationally-representative labor force surveys. 60 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Annex B. Additional Figures and Tables  Figure B1.  Housing deprivation declined in most wards Figure B3.   Outside Honiara, density is not clearly associated The proportion of the population deprived of with increases in non-agricultural jobs housing across 183 wards in 2009 and 2019 Share of non-agriculture workers and population density at the ward level, 2019 1 100 % Non Agriculture HH 2019 .8 80 Housing deprivation (2019) 60 .6 40 20 .4 0 0 2 4 6 8 10 log (Population density 2019) .2 Honiara Guadalcanal Malaita Others Fitted line (without Honiara) Fitted line (with Honiara) 0 Note: Markers represent 183 wards. 0 .2 .4 .6 .8 1 Source: Staff calculations based on Population and Housing Census 2019. Housing deprivation (2009) Honiara Guadalcanal Malaita Others Note: Wards below the 45-degree line experienced MPI reduction between 2009 and 2019. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Figure B2.  Changes in access to basic services at ward level, 2009 and 2019 (A1) Water: 2009 vs 2019 (A2) Water: distributions 1 25 .8 20 Deprivation in water (2019) .6 Number of wards 15 .4 10 .2 5 0 0 0 .2 .4 .6 .8 1 Deprivation in water (2009) 0 .2 .4 .6 .8 1 Honiara Guadalcanal Malaita Others Deprivation in water 2009 2019 61 (B2) Sanitation: distributions (B1) Sanitation: 2009 vs 2019 30 1 .8 Deprivation in sanitation (2019) 20 Number of wards .6 .4 10 .2 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Deprivation in sanitation (2009) Deprivation in sanitation Honiara Guadalcanal Malaita Others 2009 2019 (C1) Electricity: 2009 vs 2019 (C2) Electricity: distributions 1 150 .8 Deprivation in electricity (2019) 100 .6 Number of wards .4 50 .2 0 0 0 .2 .4 .6 .8 1 Deprivation in electricity (2009) 0 .2 .4 .6 .8 1 Malaita Others Deprivation in electricity Honiara Guadalcanal 2009 2019 Note: Wards below the 45-degree line experienced MPI reduction between 2009 and 2019. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 62 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Table B1.  Populations based on Census micro data, 2009 and 2019 (thousands) 2009 2019 Difference Annual growth rate National 496.0 703.2 41.8% 3.5% Urban 96.8 195.2 101.6% 7.0%   Honiara Urban Area 76.8 167.7 105.8% 7.2%   Other urban 20.0 27.5 94.1% 6.6% Rural 399.1 507.9 27.2% 2.4%   Rural Guadalcanal 75.7 111.8 47.8% 3.9%   Other rural 323.5 396.1 22.4% 2.0% Choiseul 25.3 29.7 17.2% 1.6% Western 71.4 89.2 25.0% 2.2% Isabel 24.4 28.5 16.8% 1.6% Central 25.4 29.6 16.6% 1.5% Rennell and Bellona 2.7 3.6 31.6% 2.7% Guadalcanal 90.3 151.8 68.1% 5.2% Malaita 134.8 171.1 27.0% 2.4% Makira-Ulawa 38.8 50.0 28.8% 2.5% Temotu 20.7 21.9 5.8% 0.6% Honiara 61.8 127.2 105.8% 7.2% Note: The numbers above are different from those in SINSO (2023) because: 1) the population in 2009 is not adjusted for undercounting; and 2) households with missing values in any MPI indicator are excluded. Honiara Urban Area includes urban population in Honiara and Guadalcanal Province. Sources: Staff calculations based on Population and Housing Censuses 2009 and 2019. Table B2.  Number of the multidimensional poor, 2009-2019 (thousands) 2009 2019 Difference (pt) National 218.0 249.1 31.2 14.3% Urban 17.5 31.8 14.3 82.0%   Honiara Urban Area 13.4 27.1 13.7 102.0%   Other urban 4.0 4.7 0.6 15.5% Rural 200.5 217.4 16.9 8.4%   Rural Guadalcanal 43.4 59.5 16.2 37.3%   Other rural 157.1 157.8 0.7 0.4% 63 Choiseul 10.3 7.6 -2.7 -26.3% Western 22.6 28.7 6.1 27.0% Isabel 7.8 7.2 -0.6 -7.7% Central 12.4 11.2 -1.2 -9.9% Rennell and Bellona 0.6 0.7 0.1 9.3% Guadalcanal 48.0 70.9 22.9 47.8% Malaita 77.5 73.3 -4.2 -5.4% Makira-Ulawa 19.9 24.8 4.9 24.6% Temotu 10.1 9.0 -1.0 -10.0% Honiara 8.8 15.7 6.9 78.7% Note: Numbers indicate the multidimensionally poor populations in each location. Honiara Urban Area includes urban population in Honiara and Guadalcanal Province. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Table B3.  Changes in access to basic services by province, 2009-19 Water Sanitation Electricity 2009 2019 Diff. 2009 2019 Diff. 2009 2019 Diff. National 70.6 82.8 12.2 30.2 38.7 8.5 13.3 17.1 3.8 Choiseul 72.7 91.2 18.5 11.5 13.2 1.7 2.3 4.9 2.6 Western 82.5 88.4 5.9 25.5 30.8 5.3 11.2 12.5 1.3 Isabel 87.3 94.8 7.5 33.8 40.2 6.4 5.0 5.3 0.3 Central 79.5 88.3 8.9 8.5 7.7 -0.8 4.0 3.8 -0.2 Rennell 96.5 98.6 2.0 83.6 87.8 4.1 0.8 9.2 8.4 and Bellona Guadalcanal 52.7 69.9 17.1 33.4 34.8 1.4 7.4 8.3 0.9 Malaita 62.4 77.4 15.1 23.4 37.3 13.9 3.4 3.3 -0.1 Makira-Ulawa 61.7 76.2 14.5 8.7 13.8 5.1 4.1 3.3 -0.7 Temotu 75.3 78.5 3.1 5.8 9.2 3.4 3.0 3.8 0.8 Honiara 92.3 98.4 6.0 80.6 77.4 -3.2 67.2 65.9 -1.3 Note: Numbers indicate the proportions of population with improved access. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 64 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Table B4.  Changes in the incidence of MPI indicators, 2009-2019 (Percentage points) Health Education Living standards Educational attainment enrollment Sanitation Electricity mortality Housing Cooking School Assets Water Child National -2.46 2.06 -2.68 -18.27 -12.20 -8.53 -3.83 -8.70 -2.39 Urban -1.54 3.43 0.26 -11.56 -6.39 2.48 2.30 -17.61 3.24  Honiara -1.52 3.17 0.42 -10.37 -5.90 4.66 3.74 -16.68 3.98 Urban Area   Other urban -1.33 1.54 -0.17 -10.98 -8.53 -2.90 -0.63 -13.61 7.17 Rural -2.48 1.93 -2.16 -16.10 -11.61 -6.58 -0.11 -1.75 0.50  Rural -4.43 9.81 -3.46 -14.98 -15.56 0.78 0.25 -2.94 -2.55 Guadalcanal   Other rural -1.95 -0.37 -1.97 -16.58 -11.43 -8.27 -0.13 -1.37 1.46 Choiseul -1.83 -1.24 -0.20 -25.44 -18.53 -1.72 -2.63 -1.14 1.75 Western -0.53 8.38 -1.02 -21.75 -5.92 -5.32 -1.29 -5.46 5.21 Isabel -1.95 -4.10 0.26 -23.19 -7.51 -6.43 -0.29 -2.57 1.96 Central -2.65 -4.29 -2.49 -2.73 -8.87 0.84 0.19 -2.02 3.29 Rennell -1.02 3.64 -5.61 1.56 -2.01 -4.15 -8.38 -5.85 -7.80 and Bellona Guadalcanal -4.19 8.84 -4.53 -18.69 -17.15 -1.42 -0.92 -6.83 -4.38 Malaita -2.80 -8.44 -3.89 -14.06 -15.05 -13.92 0.14 -1.02 0.92 Makira-Ulawa -0.94 16.22 -0.04 -14.40 -14.46 -5.08 0.74 -1.73 -1.66 Temotu -2.61 -2.01 -0.54 -14.37 -3.10 -3.41 -0.83 -1.37 -0.75 Honiara -1.20 1.98 0.87 -9.39 -6.04 3.17 1.28 -19.59 3.84 Note: Numbers indicate the changes in the proportion of deprived individuals between 2009 and 2019 (in percentage points) in each location. Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 65 Table B5.  Changes in education levels by Province, 2009-2019 (%) No education/ Secondary Secondary Primary complete Tertiary complete Primary incomplete incomplete complete 2009 2019 2009 2019 2009 2019 2009 2019 2009 2019 Diff. Diff. Diff. Diff. Diff. National 35.2 30.8 -4.4 33.3 25.7 -7.7 21.9 31.6 9.7 0.4 1.2 0.9 1.0 2.4 1.4 Urban 18.6 17.9 -0.7 27.8 18.7 -9.0 35.0 41.2 6.2 1.1 2.7 1.7 3.1 5.8 2.6  Honiara 18.8 18.1 -0.7 25.8 17.8 -8.0 35.7 41.2 5.5 1.2 2.9 1.8 3.5 6.2 2.7 Urban Area   Other urban 18.0 16.9 -1.0 35.6 24.8 -10.8 32.1 41.0 8.9 0.6 1.5 1.0 1.7 3.0 1.2 Rural 40.1 36.7 -3.3 35.0 28.9 -6.1 18.1 27.2 9.2 0.2 0.6 0.4 0.3 0.8 0.5  Rural 43.8 39.4 -4.5 30.4 25.6 -4.8 18.5 27.3 8.8 0.2 0.7 0.5 0.4 1.1 0.7 Guadalcanal   Other rural 39.2 36.0 -3.2 36.1 29.8 -6.3 18.0 27.2 9.3 0.1 0.5 0.4 0.3 0.7 0.4 Choiseul 25.8 23.4 -2.4 48.4 38.5 -9.9 19.7 30.2 10.4 0.1 0.4 0.3 0.3 1.0 0.6 Western 21.4 20.8 -0.6 48.3 36.9 -11.4 22.0 32.7 10.7 0.2 0.7 0.4 0.7 1.2 0.5 Isabel 28.8 26.1 -2.7 34.2 27.6 -6.6 29.7 35.6 5.9 0.2 0.5 0.4 0.3 1.0 0.7 Central 44.9 38.9 -6.0 30.9 23.3 -7.6 19.4 30.1 10.7 0.2 0.6 0.5 0.3 1.0 0.8 Rennell 24.2 18.7 -5.5 42.8 26.6 -16.3 21.0 35.4 14.4 0.1 1.8 1.7 2.0 1.8 -0.2 and Bellona Guadalcanal 41.2 34.8 -6.4 30.2 24.4 -5.8 20.6 30.6 10.0 0.3 1.1 0.8 0.6 1.8 1.2 Malaita 51.3 46.3 -4.9 27.5 24.2 -3.2 14.6 23.6 9.1 0.2 0.6 0.4 0.3 0.7 0.4 Makira-Ulawa 31.4 29.1 -2.3 38.3 33.3 -5.0 20.2 29.0 8.7 0.2 0.5 0.3 0.4 0.8 0.4 Temotu 43.8 40.9 -2.9 32.4 26.5 -5.9 17.4 26.2 8.8 0.1 0.6 0.5 0.4 0.8 0.4 Honiara 16.4 16.5 0.1 25.0 16.6 -8.4 36.9 41.9 5.0 1.3 3.2 1.9 4.0 7.0 3.0 Note: Ages from 15 to 64 years only. Other educational categories such as vocational training are not shown. Honiara Urban Area includes urban popu- lation in Honiara and Guadalcanal Province. Source: Staff calculations based on Population and Housing Censuses 2009 and 2019 66 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Table B6.  Changes in sectoral worker distributions, 2009-2019 National Urban Rural 2009 2019 2009 2019 2009 2019 Number of workers (thousands) Agriculture 30.6 64.7 2.8 13.6 27.8 51.0 Non-agriculture 43.2 65.1 23.0 41.4 20.2 23.7 Total 73.8 129.8 25.8 55.0 48.0 74.7 Share of workers (percent) Agriculture 41.4 49.8 10.9 24.7 57.9 68.3 Non-agriculture 58.6 50.2 89.1 75.3 42.1 31.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. Table B7.  Unemployment in Honiara Urban Area by sex, 2009-19 (%) Male Female 2009 2019 Diff. 2009 2019 Diff. All 4.2 7.5 3.2 3.8 7.0 3.2 Age: 15-24 5.0 7.8 2.8 4.5 7.2 2.7 Age: 25-34 5.0 9.5 4.6 4.5 9.3 4.8 Age: 35-44 3.2 6.8 3.6 2.8 6.4 3.6 Age: 45-54 2.7 4.9 2.2 1.7 4.4 2.7 Age: 55-64 2.3 3.6 1.3 1.0 2.0 1.0 Primary incomplete 5.4 9.9 4.6 3.1 6.9 3.7 Primary complete 4.0 7.3 3.3 3.5 6.7 3.3 Secondary incomplete 5.1 7.9 2.8 5.2 8.0 2.8 Secondary complete 2.6 5.9 3.3 2.7 6.2 3.5 Tertiary complete 1.8 3.7 1.9 1.6 3.4 1.8 Other 2.4 7.7 5.3 3.0 7.5 4.5 Note: Only the working-age population is included. Honiara Urban Area includes the urban population in Honiara and Guadalcanal Province. Source: Staff calculations based on Population and Housing Census 2019. 67 Annex C. Maps Figure C1.  Ward-level poverty incidence Changes 2009-2019 (percentage point) Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 68 and Spatial Disparities Multidimensional Poverty  inSolomon Islands Figure C2.  Number of the multidimensional poor Changes between 2019 and 2009 (percentage) Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 69 Figure C3.  Ward-level population (2019) Source: Staff calculations based on Population and Housing Censuses 2009 and 2019. 70 and Spatial Disparities Multidimensional Poverty  inSolomon Islands 71 72