Modernizing Social Protection and Labor Policy for Inclusive Growth November 2019 POLICY NOTE CONTENTS Abbreviations and Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Executive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Financing and Institutional Overview and Challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Policy Options for Improving Risk Management Systems in Côte d’Ivoire . . . . . . . . . . xv Institutional and Fiscal Framework for Reform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Objectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Framework and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Growth and Poverty Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Part I: Financing and Institutional Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Overall Social Protection and Labor Spending and Key Institutions . . . . . . . . . . . . . . 11 Comparative Expenditures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Key Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Labor Market Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Labor Market Agencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Expenditure Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Implementation Arrangements, Monitoring and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . 27 Targeting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Social Insurance and Pensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Overview of Pension Funds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH iii Demographics and Membership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Key System Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Productive Social Safety Nets and Social Assistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Review of Existing Social Safety Net Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Reform Agenda Underway. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Generosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Targeting and Effectiveness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Part II: Policy Options for Improving Risk Management Systems in Côte d’Ivoire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Designing and Implementing Social Insurance Programs for All. . . . . . . . . . . . . . . . . . 51 General Principles to Inform Reforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Implications for Côte d’Ivoire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 Integrating Social Insurance and Social Assistance Programs . . . . . . . . . . . . . . . . . . . . 73 Financing an Integrated System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Delivery Implications for Côte d’Ivoire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Revamping Labor Programs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Promoting Labor Productivity and Job Creation for Vulnerable Workers . . . . . . . . . . . 79 Facilitating Labor Market Transitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Part III: Institutional and Fiscal Framework for Reform. . . . . . . . . . . . . . . . . . . . . . . . . . 89 Achieving Universal Social Protection and Better Jobs: Policies and Fiscal Impacts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Creating Fiscal Space to Enable Reforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Managing the Political Economy of Reform and Building State Capabilities. . . . . . . . . 98 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 iv CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Figues, Tables, and Boxes FIGURE ES-1  SPL and other public expenditures as percent of GDP, 2016–2018. . . . . . . . . . . xiv FIGURE ES-2  Integrating social protection and labor policies across household income categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviii FIGURE ES-3  Integrating social protection and labor policies across household income categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix FIGURE 1  Opportunity, equity and resilience: the role of social protection and labor systems along the life cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 FIGURE 2  Growth and poverty trends, 1967–2016. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 FIGURE 3  SPL and other public expenditures as percent of GDP, 2016–2018. . . . . . . . . . . . . . 13 FIGURE 4  SSN expenditures as percent of GDP in Côte d’Ivoire and other low- and middle-income countries, 2016–2018. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 FIGURE 5  SSN coverage as percent of total population in Côte d’Ivoire and other low- and middle-income countries, 2016–2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 FIGURE 6  SPL and other public expenditures as percent of total public expenditures, 2016–2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 FIGURE 7  Distribution of SPL expenditures by program type, 2016–2018. . . . . . . . . . . . . . . . . 15 FIGURE 8  SPL coverage by region, 2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 FIGURE 9  Coverage (percentage within each quintile) of SPL programs by program type, 2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 FIGURE 10  Institutional mapping of key SPL actors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 FIGURE 11  Geographic distribution of social centers by geographic zone, 2016–2018. . . . . . 22 FIGURE 12  Geographic distribution of social workers by regional directorates, 2016–2018. 23 FIGURE 13  Pension contribution rates in Sub-Saharan Africa, 2016–2018. . . . . . . . . . . . . . . . . 35 FIGURE 14  Estimated pension expenditures and membership, 2016–2018. . . . . . . . . . . . . . . . 37 FIGURE 15  Program effects on poverty indicators, by target group. . . . . . . . . . . . . . . . . . . . . . . 46 FIGURE 16  Actuarially fair adjustments to the pension as a function of the retirement age. . 61 FIGURE 17  Incentives to contribute under the existing pension guarantee. . . . . . . . . . . . . . . . 62 FIGURE 18  Illustrative welfare and fiscal impacts of a universal pension system . . . . . . . . . . . 64 FIGURE 19  Illustrative fiscal and welfare impacts of universal health insurance. . . . . . . . . . . . 68 FIGURE 20  Severance pay in Côte d’Ivoire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 FIGURE 21  Welfare and fiscal impacts of an unemployment benefit scheme. . . . . . . . . . . . . . . 71 FIGURE 22  Integrating cash-transfers and old-age pensions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 FIGURE 23  Projections of Social Protection System Components. . . . . . . . . . . . . . . . . . . . . . . . 94 FIGURE 24  Deficit and Public Debt with Changes in Tax Revenues and Transfers. . . . . . . . . . 99 FIGURE 25  Capabilities and Reform Paths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH v FIGURE 26  Delivery system tools to maximize the effectiveness of SPL expenditures. . . . . 104 FIGURE 27  Integrating social protection and labor policies across household income categories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 TABLE 1  Poverty and human capital indicators by region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 TABLE 2  SPL expenditures by program type (local currency and USD), 2016–2018 . . . . . . . . 15 TABLE 3  SPL coverage (percentage within each region) by program type, 2015 (detailed data) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 TABLE 4  SPL coverage (percentage within each quintile) by program type, 2015 (detailed data) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 TABLE 5  Overview of main SPL programs and policies in Côte d’Ivoire, 2016–2018. . . . . . . . 19 TABLE 6  Overview of main SPL programs in Côte d’Ivoire (2017) included in this assessment based on program aims and data availability. . . . . . . . . . . . . . . . . . . . . . . . . . . 20 TABLE 7  Local structures and operating budget of MESP and associated institutions, 2016–18. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 TABLE 8  National labor market insertion projects and programs. . . . . . . . . . . . . . . . . . . . . . . . . 29 TABLE 9  Distribution of job seekers and beneficiaries by type of program. . . . . . . . . . . . . . . . 29 TABLE 10  Main pension system performance indicators, 2016–2018. . . . . . . . . . . . . . . . . . . . . 33 TABLE 11  Main parameters of the private and public sector pension schemes. . . . . . . . . . . . . 34 TABLE 12  Indexation policy choices made by countries in Sub-Saharan Africa. . . . . . . . . . . . . 35 TABLE 13  Mapping of social assistance programs and key indicators in Côte d’Ivoire, 2016–2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 TABLE 14  Aims of key social assistance programs in Côte d’Ivoire, 2016–2018 . . . . . . . . . . . . 41 TABLE 15  Transfer size with respect to poverty line and extreme poverty line. . . . . . . . . . . . . 44 TABLE 16  Simulations of cash transfer program effects on poverty and extreme poverty indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 TABLE 17  Core administrative functions of a Health Insurance Fund. . . . . . . . . . . . . . . . . . . . . . 67 TABLE 18  Structure of the Public Budget in Côte d’Ivoire, 2015–2021 . . . . . . . . . . . . . . . . . . . . . 96 TABLE 19  Distribution of the Public Budget by Sector in Côte d’Ivoire, 2016–2021. . . . . . . . . 97 TABLE 20  Projection of the Fiscal Deficit and Public Debt in Côte d’Ivoire, 2019–2025. . . . . 97 TABLE 21  State Capability for Selected Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 BOX 1  Recent trends in pension reform dialogue in Côte d’Ivoire. . . . . . . . . . . . . . . . . . . . . . . . 37 vi CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH viii CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH ABBREVIATIONS AND ACRONYMS AFD French Development Agency / Agence Française de Développement ALMP/PAMT Active Labor Market Programs / Programmes Actifs du Marché du Travail BCEAO Central Bank of West African States / Banque Centrale des Etats de l’Afrique de l’Ouest CNPS National Fund for Social Provisio / Caisse Nationale de la Prévoyance Sociale EU/UE European Union / Union européenne EUR Euro FCFA Franc of the African Financial Community / Franc de la Communauté Financière Africaine GDP/PIB Gross Domestic Product / Produit intérieur brut ICT/TIC Information, Communication and Technology / Technologies de l’informa- tion et de la communication INS/INS National Institute of Statistics / Institut National de la Statistiques M&E/S&E Monitoring and Evaluation / Suivi &Evaluation MEF/MEF Ministry of Economy and Finance / Ministère de l’Economie et des Finances MESP/MEPS Ministry of Employment and Social Protection / Ministère de l’Emploi et de la Protection Sociale MIS/SIG Management Information System / Système d’information de gestion MSCLCP/MSCSLP Ministry of Solidarity and Poverty Reduction / Ministère de la solidarité, de la cohesion sociale et de la lutte contre la pauvreté NGO/ONG Non-governmental Organization / Organisation non-gouvernementale NSPS/SNPS National Social Protection Strategy / Stratégie nationale de protection sociale SPL/PST Social Protection and Labor / Protection Sociale et Travail SSN/FSS Social Safety Nets / Filets de Sécurité Sociale PEJEDEC/PEJEDEC Youth Employment and Skills Development Emergency Project / Projet d’Urgence de Création d’Emploi Jeunes et de Développement des Compétences UNICEF United Nations Children’s Fund / Fonds des Nations Unies pour l’enfance US$/USD United States Dollar / Dollar US WAEMU/UEMOA West African Economic and Monetary Union / Union monétaire et économique Ouest-africaine WB/BM World Bank / Banque mondiale CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH ix CURRENCY EQUIVALENTS (Exchange Rate Effective November 30, 2019) XOF 594.174 = 1 USD) USD 1 = EUR 0.905734 FISCAL YEAR January 1—December 31 x CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH ACKNOWLEDGEMENTS This report was prepared as part of the World Bank’s technical assistance program1 (2015– 2018) in Côte d’Ivoire on social security and labor policy (contributory and non-contributory), «Building Productive and Resilient Social Protection and Labor Systems» and «Strengthening Social Protection and Labor Delivery Systems». The program was implemented in partnership with the Intersectoral Social Protection Technical Committee under the Presidency, respon- sible for the operationalization of the National Social Protection Strategy and the National Employment Plan (2016–2020), as part of the National Development Plan 2016–2020. This work has accompanied investment projects on strengthening the social protection and labor system for vulnerable populations, funded with support by the World Bank/International Development Association (IDA), under the Productive Social Nets Project and the Social Protection and Economic Inclusion Project, managed by the Ministry of Employment and Social Protection and the Ministry of Solidarity, Social Cohesion and Poverty Reduction. The analytical framework and the main recommendations of the report were discussed and developed with cross-sectoral Ivorian authorities and stakeholders during a round table held in Abidjan during March 2018 on “Towards an Integrated Economic Inclusion Platform,” in addition to workshops and discussions organized with the World Bank throughout the program. The team appreci- ates the effort, time and views shared by public and civil society representatives as part of the preparation of this work. The World Bank team included Heba Elgazzar (Senior Economist, Social Protection and Jobs) and David Robalino (Lead Economist and Manager, Jobs Group), with contributions by Claudia Rodriguez (Social Protection Specialist), Miglena Abels (Social Protection Analyst), Raphaela Karlen (Labor Specialist, Jobs Group), Luc Christiansen (Lead Economist, Jobs Group), Clement Joubert (Economist, Human Development Research Group), Roy Katayama (Senior Economist, Poverty), Patrick Premand (Senior Economist, Social Protection and Jobs), Chris Tullis (Consultant, Digital Development Specialist), and Giulia Parzani (Consultant, Operations). The work benefited from inputs from and discussions with Meritxell Martinez (Operations Officer, International Finance Corporation), Marina Novikova (Consultant, Economist), Emma Monsalve Montiel, (Consultant, Economic Analyst), Bienvenue Tien (Consultant, Economist), Samba Ba (Senior Economist, Macroeconomics), Solene Rougeaux (Senior Social Protection Specialist), 1 This work has received generous financial support from the Rapid Social Response Trust Fund and these donors (Sweden, Norway, the United Kingdom, Australia and the Russian Federation), with the sincere appreciation of the authors. No part of this report may be reproduced or disseminated in any form or by any means, electronic or mechanical, photocopying or recording, storage, or retrieval system, without the prior written permission of the World Bank. The observations, interpretations and conclusions expressed here are those of the authors and do not necessarily reflect the views of the Executive Directors of the World Bank or the countries that they represent, or the financial donors to the work. The World Bank does not guarantee the accuracy of the data cited in this report. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH xi Michael Muhanna (consultant, social protection), Michel Welmond (Program Lead, Human Development), Jacques Morisset (Program Lead, Macroeconomics) and Ian Walker (Lead Economist and Manager, Jobs Group). The team thanks the peer reviewers of this work for their helpful comments and suggestions, including Giuseppe Zampaglione (Lead Specialist, Social Protection and Jobs), Silvia Baur- Yazbeck (Financial Analyst, International Finance Corporation) and John Van Dyck (Senior Specialist, Social Protection and Jobs). The team also thanks the broader Côte d’Ivoire country team, including Oumou Coulibaly (Program Assistant, Abidjan Office), Lydie Anne Billey (Program Assistant, Washington) and Dora Harris (Senior Program Assistant, Washington) for their helpful logistical and office assistance. The report was prepared with excellent graphic design support by Will Kemp (Visual Storyteller, Global Corporate Solutions Publications Unit), Jeffrey Lawrence (Manager, Global Corporate Solutions Publications Unit), and their associates. The technical assistance program was mobilized under the general direction of Jehan Arulpragasam (Practice Manager, Social Protection and Jobs, Africa Region) and Pierre Laporte (Country Director, Côte d’Ivoire). Finally, the team extends its special appreciation to the communities across different regions of Côte d’Ivoire with whom it met. It is by sharing their experiences, aspirations and innova- tions that they have allowed us to enrich the present work. xii CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH EXECUTIVE SUMMARY Côte d’Ivoire’s improving economic context offers new opportunities to accelerate poverty reduction. As the world’s top exporter of cocoa, raw cashews, an exporter of oil,2 and with a growing manufacturing sector, the country is the largest economy in the West African Economic and Monetary Union (WAEMU). Since the end of conflict in 2011, growth has aver- aged approximately 9 percent, higher than almost all WAEMU comparators.3 However, the benefits of growth are yet to reach vast swathes of rural, peri-urban, and remote popula- tions outside of Côte d’Ivoire’s capital, with poverty at 46 percent in 2015, using the national poverty line of FCFA 750 per day (approximately US$1.30).4 To boost inclusive growth, the Government’s vision is to increase pro-poor spending and coverage of social protection and labor programs. To this end, the main objective of this policy note is to assess key social protection and labor policies, program expenditures and institutional arrangements in Côte d’Ivoire. The main policy question addressed by the note is to what extent spending levels and institutional arrangements have been effective in terms of reaching and supporting the poorest households. Financing and Institutional Overview and Challenges Overall, using an economic classification of social protection and labor (SPL) programmatic expenditure in this analysis, total SPL spending is estimated at 1.66% of GDP as of 2016– 2018. The analysis is based on available data for 2016 to 2018, as most spending remains stable. Out of a total of nineteen programs and initiatives identified, nine had sizable budget envelopes, beneficiaries, and available data. These programs were included in this analysis. Programmatic expenditures on SPL largely go to pension schemes for the non-poor (Figure ES-1). Spending (excluding general administration) included: contributory pensions schemes at 1.56 percent of GDP (57 percent of which is devoted to civil servants and 43 percent devoted to private, formal workers), non-contributory social assistance at 0.02 percent, active labor market programs at 0.06 percent, in line with other countries in Sub-Saharan Africa (SSA) at a similar development stage of SPL systems. Overall SPL spending as a percent of total public expenditure was 6.3 percent, 94 percent of which went to pension schemes, which are not considered pro-poor spending by the Government. As a result, the distribution of SPL financ- ing in Côte d’Ivoire shows that most of the spending does not benefit the poorest households. 2 While Côte d’Ivoire exports crude oil and refined oil products, it imports crude oil for domestic production. 3 World Bank Group Concept Note for Côte d’Ivoire Country Private Sector Diagnostic, February 2019. World Bank. 2019. Cote d’Ivoire Economic Update. Washington, DC: World Bank Group. 4 National Institute of Statistics (Institut National de Statistique), Centrale des Bilans. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH xiii More specifically, Côte d’Ivoire’s national employment policy strategy for 2016–2020 lays out key priorities for enhancing access to and quality of jobs, particularly among youth. Given that over 80 percent of the labor force is employed in the informal sector, the role of labor regulations in protecting most workers is limited. Coverage of active labor market pro- grams and policies (ALMPs) is relatively limited in Côte d’Ivoire, which has been challenging to improve given little coordination among agencies. Most these programs (six out seven) were launched following the political stalemate of 2011, with one that dates to 2010. As of 2016–18, the total estimated budget of all labor programs was around FCFA 40,994 billion (or approx- imately USD 74 million), of which 5% was financed through government sources and 95% through external sources. However, available documentations indicate that the target group and the beneficiary eligibility criteria are not clearly defined in most programs, particularly pro- grams that target youth at risks and vulnerable population. There is no consensual definition of vulnerability across programs. In terms of cost-effectiveness, most of the programs lack suffi- cient data, but indicative trends suggest labor programs tend to be costly with limited impacts as currently designed. Moving forward, coverage and eligibility of most ALMPs in Côte d’Ivoire remains a key issue, in addition to inter-agency and inter-sectoral coordination. Next, the pensions system is the main social insurance program in Côte d’Ivoire, given the paucity of coverage and information on other social insurance schemes. Coverage of social insurance schemes is very limited, and data has not been sufficiently available on some programs (such as unemployment protections, as well as maternity benefits and other income support). Two pension funds operate which include the national pensions for private sector workers (CNPS) and the pensions fund for civil servants (CGRAE). Together these funds cover FIGURE ES-1  SPL and other public expenditures as percent of GDP, 2016–2018 6.00% Expenditures as % of GDP, 2015 5.38% 5.00% 4.00% 3.00% 2.61% 2.00% 1.66% 1.79% 1.56% 1.38% 1.14% 1.00% 0.34% 0.02% 0.06% 0.02% 0.00% an r y P es PL it s n lth se r t ns en rio fic er io LM s t to ic io en ce lS ea E d s) at de ns f nm te si bu rv ns lA ef ta H uc In Se bi ra ro Pe A s tri To D Al e rT vi al n al ry ci C o En ag t he ll ci to So So n - d bu ,w O an o nd lN it y nd tri er on Ki ric a Al at n- ct s lC W le ie lI (e b s i d Al Al Su Source: World Bank staff calculations using data from Ministry of Economy and Finance, Côte d’Ivoire. xiv CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH only 6–10 percent of the population as of 2018, benefiting only civil servants and workers in the formal private sector. The Government plans to continue reforms to expand social insur- ance coverage from 10 percent currently to 30 to 50 percent of the population by 2020. To date, however, little is known on the performance of the pension funds or feasibility of expand- ing coverage in the face of high poverty and informality. Average pension expenditure per pensioner in CNPS amounted to 1.3 times GDP per capita whereas average pension expen- diture per pensioner in CGRAE exceeded 2 times GDP per capita in the country. A pensions reform vision has been articulated by the Government as of 2016 for expanding coverage to vulnerable workers, building on initial measures taken in 2012 to support sustainability of the two main funds. Finally, Côte d’Ivoire recently began implementing a new social safety net reform agenda, seeking to consolidate limited, vertical programs. The new system focuses on a national cash transfer program, supported by key building blocks such as a household registry, a reliable and efficient payment system, and services to boost human capital, household earnings and employment. Such a robust, productive social safety net system aims to channel resources to the poorest households while providing pathways to greater earnings. Under the program, the provision a monthly transfer of 12,000 FCFA to all 608,201 extreme poor households in the entire country would likely cost around US$163 million annually (0.70 percent of GDP). Such an investment is expected to reduce the extreme poverty headcount by approximately 25 percent, which would not otherwise be necessarily achieved only by means of local pilot initiatives for fewer households or by public investment in infrastructure and basic services alone, which do not stimulate consumption or investment in human or physical capital. Overall, the new productive program has been designed to institute a national, sustainable system, but the question remains how to scale-up by integrating across other SPL schemes. Policy Options for Improving Risk Management Systems in Côte d’Ivoire Improving standards of living in Côte d’Ivoire and accelerating economic growth will require structural reforms in its risk management system. This system comprises, on one hand, social insurance and assistance programs to manage risks such as a disease, disability, death, poverty, or unemployment and, on the other hand, labor programs to improve jobs and earn- ings opportunities. Thinking in terms of a risk management system is important to guide the allocation of limited public resources among competing programs. Investing today in programs that improve jobs and earnings opportunities is likely to reduce to the future cost of social insurance/assistance programs. Based on international experiences and recent innovations, Côte d’Ivoire can improve the design of social insurance/assistance and labor programs in the context of an integrated risk management framework. The focus in on three types of initiatives: CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH xv • Initiatives to expand the coverage of social insurance programs to informal workers which rely on new ICTs to ID, profile, enroll workers, collect contributions/savings, and make pay- ments, as well as improvements in financial literacy and behavioral nudges. • Initiatives to rationalize and integrate redistributive arrangements between social assis- tance and social insurance programs. • Labor programs that rely on investment subsidies to promote wage and self-employment among vulnerable workers, particularly in rural areas. Specifically, first, a reformed pension system managed by the CNPS could become the scheme for all workers, including informal sector workers. This would imply redefining the mandate of the system and modifying benefits formulas and eligibility conditions along the lines discussed above. Simulations using a prototype low-income country illustrate how the value of the pension resulting from targeted subsidies (TSI) and from individual contributions would change across income deciles. The fact that the value of the subsidies declines gradu- ally with income reduces negative incentives on labor supply. Under the assumption of univer- sal coverage, the fiscal cost of the system would be in the order of half a percentage point of GDP per year or the equivalent of a 0.6% tax on consumption. This seems to be a manageable level of expenditures to guarantee adequate pensions for all workers. As a reference, total expenditures in the existing pension system for civil servants and private sector workers are each in the order of 0.6 percent of GDP with very limited coverage. Second, like in the case of pensions, the coverage of social health insurance (SHI) in Côte d’Ivoire is limited to formal sector workers and can be extended as part of the same framework. In 2015 the government started a national project to create a stand-alone health insurance fund (CNAM) for the poor and informal sector workers. Assuming that the institutions necessary for the functioning and expansion of the new health insurance system—CMU— are in place, a key policy issue will be to define the most appropriate financing mechanism. Although at this stage there are no country specific data, simulations with the same prototype model used for pensions suggest that targeting subsidies might be a necessary condition to be able to expand coverage. In the calculations it is assumed that the cost of the package of health services offered by the CMU is equal to USD 100 per capita per year or 6% of GDP per capita. Third, like the majority of countries Côte d’Ivoire does not have an unemployment insurance program but it regulates severance pay through the labor code. Workers who lose their job for economic or technical reasons receive a lump-sum benefit that is proportional to the number of years of service: 30% of the monthly salary for the first five years of service; 35% for years 6–10; and 40% for years 11 and beyond. However, severance pay is not the most effec- tive form of protection for workers and yet it can affect the demand for labor. An alternative system for Côte d’Ivoire would involve setting up a proper unemployment insurance scheme. The unemployment rate for a majority of workers in Côte d’Ivoire is actually quite low; the main problem is underemployment and not unemployment. Initially, the system would be open, however, to both formal and informal sector workers with earnings above a given threshold, whether in wage or self-employment. xvi CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Fourth, as Côte d’Ivoire expands the coverage of cash-transfers, there is scope to gradually phase-out in-kind transfers and integrate, in a transparent way, subsidies for social insur- ance with anti-poverty transfers. As shown in recent studies, there is little rationale for having separate social assistance and social insurance programs, and/or multiple forms of redistrib- utive arrangements. The redistributive part of the social protection system therefore would have two elements: a basic cash/income transfer that would be targeted/tapered (TCT) like the current transfers provided by the productive safety-net programs; and subsidized insur- ance that is also likely to be tapered (TSI). The cash-component is critical to ensure that all workers, regardless of activity and type of work, are always able to support a minimum level of consumption. Finally, labor programs would need to be integrated and expanded, focusing on two inter- related objectives: creating better jobs for vulnerable workers and facilitating labor market transitions. Part of the agenda is to increase the productivity of the jobs most prevalent in low-income and poor households, particularly self-employment in agriculture and non-farm activities, while improving opportunities for job creation. This would require internalizing jobs related externalities and incentivizing investments in selected economic activities and regions. The other part of the agenda is to facilitate labor market transitions from inactivity/unemploy- ment into a job, and from low to higher productivity jobs. There are three operational models that Côte d’Ivoire could consider: (1) self-employment programs imbedded in supply chain development initiatives; (2) self-employment programs imbedded in aggregator programs that facilitate access to markets; and 3) self-employment programs imbedded in initiatives to produce public goods. Institutional and Fiscal Framework for Reform A holistic institutional and fiscal framework can best support achieving universal social protection while improving job opportunities. The social protection system envisioned for Côte d’Ivoire would integrate social insurance/assistance programs and active labor market programs. This integration would allow for a better allocation of limited public resources among competing programs. Depending on coverage targets and whether the health insurance system is universal or targeted expenditures by year 2025 could range between, the reformed social protection system, including health insurance, could cost between 4% and 12% of GDP. The largest item is the health insurance system, followed by the basic income guarantee and active labor market programs (Figure ES-2). It is also assumed that the reform of the pension system would require making part of the implicit pension debt (the part that is not backed by financial assets and the pay-as-you-go asset) explicit. Repaying this debt as individuals retire could cost around 0.3% of GDP per year. In terms of creating fiscal space, there are essentially three mechanisms that can be applied in Côte d’Ivoire to accommodate higher expenditures: 1) increasing revenues from taxes; 2) reallocating public expenditures; and 3) improving debt management. Based on different CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH xvii simulations, the expansion of the social protection system cannot be achieved sustainably by issuing debt. Being able to mobilize sufficient tax revenues is a pre-condition. Such reforms will not necessarily be easy in Côte d’Ivoire but are nonetheless achievable. Countries that have been able to introduce successful reforms have a policy-making process where political actors and social partners cooperate and are able to reach and enforce agree- ments. In general, successful reforms have been defined by: 1) strong leadership; 2) the exis- tence of good “aggregators” that can reduce the number of actors who can directly influence policy; and 3) open dialog and continuous interactions among these actors within a long-term planning horizon. Successful countries also have a strong bureaucracy to which the analysis and implementation of policies can be delegated. This tends to be underestimated. There can be countries with a good policy-making process but with no state-capability and therefore unable to design and implement reforms. In all cases, the successful implementation of given reform program requires accountability for results and therefore having the capacity to monitor and measure these results. Achieving proper accountability within public institutions – both among managers and front- line staff – is not easy. It often involves changes in human resources policies (e.g., perfor- mance assessment and promotion/remuneration policies) within the public administration. One place to start, however, is to setup robust M&E systems to track the relevant performance indi- cators. In the case of the pension system, for instance, the share of the labor force covered, the collection rate, or the time it takes to process new pensions and pay benefits. In the case of active labor market programs where there is room to outsource the provision of services to private providers, these M&E systems also play a key role in the design of payment systems based on performance. FIGURE ES-2  Integrating social protection and labor policies across household income categories Source: World Bank staff. xviii CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Conclusion Within the framework of its poverty reduction and employment agenda, Côte d’Ivoire has started reforms to improve the effectiveness of public expenditures and boost fiscal space through efficiency gains. As an initial step, new tools are being developed as of 2016 to foster coordination on delivery arrangements, with a view to address benefit levels and targeting over the mid-term. As the delivery framework improves, new avenues are expected to open to better track, target and tailor benefit levels and the type of measures to different groups within the SPL system (Figure ES-3). Moving forward, a broader dialogue will be needed to assess options for improving fiscal space over different time horizons, depending on the detailed social and labor objectives sought. Depending on various growth and poverty-reduction scenarios, domestic and external resource needs can be examined. As part of mid-term expenditure framework preparation, the relative cost-effectiveness of other pro-poor instruments vis-à-vis targeted SPL interventions can be assessed. Ultimately, together with economic growth strategies, modernizing Côte d’Ivoire’s SPL system can help support this agenda directly by catalyzing economic inclusion for the majority of Ivorians, equitably and at scale. FIGURE ES-3  Integrating social protection and labor policies across household income categories • Strengthen sustainable contributory social protection Highest income >80% of population • Reinforce productivity, working conditions, formal sector • Strengthen informal/formal sector High income 60–80% productivity • Adjust profiling mechanisms • Support quasi-contributory programs Middle income 40–60% • Facilitate exit from non-contributory programs • Subsidize productive non-contributory Low income 20–40% programs • Support non-contributory programs, Lowest income <20% of population basic living conditions, productive employment entry Source: World Bank staff. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH xix xx CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 1 INTRODUCTION Côte d’Ivoire’s improving economic context offers new opportunities to accelerate poverty reduction. As the world’s top exporter of cocoa, raw cashews, an exporter of oil,1 and with a growing manufacturing sector, the country is the largest economy in the West African Economic and Monetary Union (WAEMU). Since the end of conflict in 2011, growth has been increasing. From 2012 to 2018, growth averaged approximately 9 percent and converged to 7.2 percent in 2019, higher than almost all WAEMU comparators as well as Ghana, Sri Lanka, and Ethiopia.2 3 About 40 percent of growth has been driven by investment through govern- ment spending, followed by services, agriculture, trade, transport, construction, banking, and the digital economy.4 However, the benefits of growth are yet to reach vast swathes of rural, peri-urban, and remote populations outside of Côte d’Ivoire’s capital, Yamoussoukro, and the main cities such as Abidjan, Bouaké, and San Pedro. Among its population of 24 million as of 2017, poverty has decreased from 51 percent in 2011 to 46 percent in 2015, using the national poverty line of FCFA 750 per day (approximately US$1.30)6. Gross national income (GNI) hov- ers at US$1,489 as of 2015. Poverty reaches over 70 percent in rural regions, although poverty continues to gradually decline.5 To boost inclusive growth, the Government’s vision is to increase pro-poor spending and coverage of social protection and labor programs. The Government’s 2016–2020 National Development Plan (NDP) aims to transform Côte d’Ivoire into a middle-income economy by 2020 and substantially cut poverty. The Government’s poverty reduction vision is enshrined in its National Social Protection Strategy [(NSPS), 2016–2020], National Employment Plan [(NEP) 2016–2020], and the recently launched inter-sectoral Governmental Social Program/PSGOUV 1 While Côte d’Ivoire exports crude oil and refined oil products, it imports crude oil for domestic production. 2 World Bank Group Concept Note for Côte d’Ivoire Country Private Sector Diagnostic, February 2019. 3 World Bank. 2019. Cote d’Ivoire Economic Update. Washington, DC: World Bank Group. 4 National Institute of Statistics (Institut National de Statistique), Centrale des Bilans. 5 Based on national poverty line, which measures relative poverty. As of 2015, the national poverty line is FCFA 269,075 per capita per year. Source: Institut National de Statistique. 2015. Enquête sur le niveau de vie des ménages en Côte d’Ivoire: Profil de pauvreté. République de Côte d’Ivoire. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 1 (2019–2020).6 The strategy endorsed by the NSPS and NEP, supported by implementation plans under the PSGOUV, aims to transform Côte d’Ivoire’s social protection and labor pro- grams, which have historically been limited in attenuating poverty. Objectives The main objective of this policy note is to assess key social protection and labor policies, program expenditures and institutional arrangements in Côte d’Ivoire. The note aims to iden- tify policy options for improving allocative and technical efficiency for key programs, including how to modernize and coordinate financing arrangements, delivery systems and benefit pack- ages across different populations. The main policy question addressed by the note is to what extent spending levels and institutional arrangements have been effective in terms of reaching and supporting the poorest households. The first part of the policy note focuses on providing an overview of current programs; the second part focuses on policy options moving forward. As an initial analysis, the first part of the note takes stock and reviews key aspects of the main SPL programs in terms of: (i) financing, (ii) access and coverage, and (iii) governance and administration in service delivery. The second part of the note evaluates options for strengthening risk management as a well-coordinated, integrated system across different programs, with a focus on (i) fiscal implications, (ii) distributional effects, and (iii) modernizing design features. Framework and Methods The framework for the assessment is based on promoting equity through efficient public sector management, taking into account political economy dimensions. The World Bank Social Protection and Labor Strategy (2012–2022) emphasizes that well-coordinated social protection and labor systems are central to realizing three main policy aims along the life cycle.7 These aims include: (i) strengthening opportunity for all; (ii) strengthening equity (par- ticularly in mitigating chronic poverty); and (iii) building resilience for households (in terms of protecting against unforeseen risks and shocks). This note proposes to improve policies by leveraging coordinated financing and service delivery of the broader system. The work draws upon new and recent analysis using quantitative and qualitative methods based primarily on a historical review from 2014 to 2018. The analysis employs administrative data, household survey data, recently-completed WB reports and key stakeholder interviews 6 République de la Côte d’Ivoire, Communiqué du Conseil des Ministres du mercredi 09 janvier 2019 sur la stratégie de la mise en œuvre du Programme Social du Gouvernement (PSGOUV 2019–2020). The PSGOUV identifies pro-poor, social program spending priorities across sectors for the 2019 and 2020 budget cycle. The PSGOUV priorities include accelerating coverage expansion of the productive social safety net program under implementation, promoting economic inclusion among youth and females, expanding social tariffs for electricity (subsidies) for the poor, expanding access to potable water in rural regions, and expanding social housing benefits. 7 World Bank (2012). World Bank Group Social Protection and Labor Strategy, 2012-2022. Washington, DC: World Bank. 2 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 1  Opportunity, equity and resilience: the role of social protection and labor systems along the life cycle Equity: Social pensions Opportunity: Nutrition/ECD, Resilience: Old-age pensions, CCTs for pre-school, health disability insurance Pregnancy, Equity: OVC programs, Early childhood child allowances Old age School age Opportunity: CCTs for (girls’) education Opportunity: Employment Equity: Child services, entrepreneurship, allowances, school training and skills feeding Equity: Cash and in-kind transfers, public works programs Working age Youth Resilience: Unemployment, Opportunity: disability insurance Youth employment programs, skills training Source: Source : World Bank (2012). World Bank Group Social Protection and Labor Strategy, 2012–2022. Washington, DC: World Bank. among Ivorian authorities, non-governmental organizations and international partners. The main household surveys used include: the 2008 and 2014 National Household Living Standards Survey and the World Bank 2014 FinDex Survey on financial inclusion. The analysis has been based on using the classification of the World Bank SPL database (“ASPIRE”) and the distributional analysis of programs at the household level (using the “ADEPT” analytic tool). The suggested policy recommendations discussed in this policy note have been adapted to the Ivorian context, based on international experience discussed throughout the note. Moving forward, the Government plans to implement its NSPS as part of its National Social Protection Platform, an inter-ministerial committee representing 17 ministries created in July 2015 to coordinate policies and program delivery. Over the next phase, the Platform plans to develop and roll out the building blocks of a harmonized social protection and labor strategy, including establishing the institutional architecture of the National Social Protection Platform, namely governance, management information system and legal frameworks, needed to better coordinate and deliver social protection and labor programs. It is hoped that this policy note supports Côte d’Ivoire as it continues to implement its NSPS and consolidates its approaches to poverty reduction for the coming era. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 3 Growth and Poverty Profile While high growth and low inflation is projected, falling world cocoa prices and increas- ing oil prices are impacting Côte d’Ivoire’s revenue generation. Inflation was estimated at 0.7 percent in 2016 and the current account deficit is approximately 1.1 percent of GDP. The deficit reflects increasing investment-related imports and a drought-induced decline in cocoa and cashew nuts exports. Revenues fell short by 0.6 percent of GDP because of cocoa- and oil-related losses, as well as lower-than-expected revenues on tobacco and alcohol excises, income tax, and VAT. Limited fiscal space is closely tied to Côte d’Ivoire’s small tax base and income tax collection capacity. Tax revenue collection was estimated at 14.9 percent of GDP in 2016 (versus the Government’s target of 15.8), and total expenditures were estimated at 23.4 percent of GDP (versus a projection of 24.8 percent of GDP), equivalent to 4,950 F CFA billions. As a result, the overall budget deficit was 3.9 percent of GDP (meeting the target of 4.0 percent of GDP), with the shortfall was mainly covered by funds from WAEMU financial markets and foreign loans. Côte d’Ivoire has yet to restore poverty levels and human development indicators to pre- 1999 levels. In 1999, it plunged into a long period of political instability following a coup d’état, violence and partition (1999–2007), as well as an election crisis (2010–2012). Living standards plummeted as growth significantly declined, with GDP per capita in 2011 at half its level a generation earlier. GDP per capita moved from nearly 1,124,000 F CFA (US$ 2,400) in 1978, a pre-crisis peak, to around 535,000 F CFA (US$ 1,100) in 2011, reaching nearly 735,000 F CFA (US$ 1,500) by 2016. Between 1985 and 20088, using estimates for the national poverty line, the poverty rate increased from around 10 percent to 43 percent,9 reaching 46 percent by 2015. Between 2008 to 2015, poverty is estimated to have increased due to the post-election crisis. Inequality is moderately high with a Gini index of 40. Poverty continues to be overwhelmingly rural.10 57 percent of the rural population is living in poverty, compared to 35 percent in the urban areas. As the population is greater in rural areas, 61 percent of the poor are in rural areas. Likewise, poverty is higher among agricultural house- holds than non-agricultural households. Poverty is also higher for families with many young children, the less educated, the elderly, and those living in the northern regions of the country. The North-South disparities can be explained in part by the country’s two distinct agricultural regions. The forest region in the South is home to cocoa, coffee (mostly robusta), and palm oil production, while the drier savannah region to the North is home to cotton and cashew production. 8 Most recent time frame for which data is available. 9 Estimates prior to 2008 are national and not necessarily comparable to those of 2008 onward. Pre-2008 estimates using the national poverty line of 75,000 F CFA in 1985 prices (or equivalent to $1.40/per day per person in 2008 prices and exchange rates) for the 1985–2008 data. Due to a lack of data following the crisis, newer estimates with updated consumption aggregates are only available from 2008 onward. 10 All figures relate to national and World Bank poverty estimates for 2015. The next living conditions survey for Côte d’Ivoire is scheduled for early 2018, as part of the WAEMU harmonized survey initiative. The eight-member states of WAEMU, as well as several other countries, will be implementing a modernized survey to obtain data that are comparable both over time and across countries. The improved welfare data will be used to facilitate regional economic integration, monitor key social indicators, and obtain high-quality comparable data on the informal sector and agriculture. 4 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 2  Growth and poverty trends, 1967–2016 3000 60 Poverty headcount ratio*, $1.90 a day or national 2500 50 GDP per capita (constant 2010 US$) poverty line, percent (%) 2000 40 1500 30 1000 20 500 10 0 0 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Poverty headcount ratio*, percent (%) GDP per capita (constant 2010 US$) Source: World Bank Indicators, 2017. Côte d’Ivoire has relatively low formal unemployment, and in addition, faces a major challenge in terms of employment quality. Only 17.4 percent of the employed population holds wage jobs, and many wage jobs are informal. Employment is strongly concentrated in low-productivity occupations, such as agricultural self-employment (66.9 percent of employed population) and non-agricultural self-employment (29.3 percent), particularly among the poor, women, and those living in rural areas. The fact that formal wage employment starts from a very low base means that even strong growth of that sector would only absorb a small share of the population, most likely starting with the more educated. Projections of future employ- ment composition suggest that at most 30 percent of the population is likely to be in wage jobs by 2025. The recent decline in cocoa prices is expected to hurt poor cocoa farmers. The guaranteed producer price for cocoa was cut by 36 percent in 2017 from 1,100 to 700 CFAF per kg, a level lower than in 2012 (725 CFAF per kg) when producer prices starting its steady increase. If lower farm gate cocoa prices persist, they will stress poor cocoa farmers with limited coping strategies. The creation of new job opportunities and the expansion of well targeted safety net programs will be needed to help offset these negative impacts. Demographic pressures contribute to elevated poverty and risk in Côte d’Ivoire due to a rapidly growing and young population, similar to neighboring countries. The population is young, with 41 percent younger than the age of 15. Nearly half of the population is urban, with Abidjan housing close to 39 percent of the total urban population. Côte d’Ivoire’s annual CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 5 population growth rate is estimated to be 2.5 percent per year according to the 2014 popula- tion census. This high growth rate is due to the combined effect of a declining infant mortality rate (68 per 1000 birth as of 2012), high fertility rate (an average of 5 children as of 2012) and a high level of international migration (24.2 percent of the population was not Ivorian based on the 2014 population census). Due to the young structure of the population, the dependency rate is high and estimated to be 78.6 percent. With one out of two females being of childbear- ing age, the population is expected to continue to grow and remain young over the coming decades, increasing the burden on Côte d’Ivoire’s social and economic infrastructure. For example, at least 1,000 new classrooms a year will need to be created to keep pace with the increase in the number of school-age children. Côte d’Ivoire will not be able to achieve most of the Millennium Development Goals (MDGs) by 2015, particularly in terms of education and health in rural regions and across gender. TABLE 1  Poverty and human capital indicators by region Malnutrition, Net primary enrolment rate (%) Share of Chronic height-for-age food-insecure poverty rate (% less than households Regions Zones (%) –2 SD*) Girls Boys (%) NORTH North (Korhogo) 77.2 39.3 49.9 50.8 24.0 Northwest 57.8 31.8 52.6 68.8 22.0 (Odienné) Northeast 54.7 39.3 66.4 64.9 30.0 (Bondoukou) WEST West (Man) 63.2 34.2 64.7 76.1 23.0 CENTER Center 56.0 30.2 62.3 65.9 15.0 (Yamoussoukro) Center-North 57.0 28.5 66.8 72.1 8.0 Center-West 62.9 29.7 63.1 68. 12.0 (Daloa) Center-East 53.7 24.6 65.2 69.1 4.2 (Abengourou) SOUTH South 44.6 29.0 72.3 81.5 21.0 Abidjan 21.0 17.9 73.1 80.4 4.9 Southwest 45.5 29.2 54.1 72.9 25.0 (San Pedro) NATIONAL Rural 62.5 34.9 61.1 68.2 Urban 29.5 20.5 69.3 77.8 National 48.9 29.8 61.0 66.1 18.2 Sources: Poverty data: INS Household Living Standards Survey, 2008; health and education data: UNICEF MICS, 2011–2012. Food insecurity: National Food Security Survey (EASA), 2009. *SD = standard deviations points; malnutrition defined as percentage of children under age 5 whose height for age (stunting) is more than two standard deviations below the median for the international reference population ages 0–59 months. 6 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Côte d’Ivoire will need to significantly improve access to and quality of education and health services including family planning and water and sanitation as well as gender parity. Malnutrition is also higher in rural regions; 35 percent of children in rural regions are stunted as compared to 21 percent in urban regions as of 2011. There are disparities in access to basic services, and gender disparities across wealth and urban-rural groups. High costs and distance to schools have a strong impact on school enroll- ment and dropout, with household spending on education reaching as high as 33 percent of total national expenditure.11 For maternal health services, recent survey data indicate that 67 percent of respondents cited cost as the major barrier.12 This figure rises to 83 percent in rural regions as compared to 64 percent in urban Abidjan. Differences in primary school enroll- ment exist, but the most pronounced disparities exist in secondary education to which rural residents and the poor have limited access. Given the correlation between education and wel- fare, these disparities can limit future outcomes. Gender gaps in attainment persist at all levels in both urban and rural areas and at different wealth levels. The Ivorian population’s access to improved sanitation is especially critical with only 22 percent of the population in 2015 having access to improved sanitation facilities, and as low as 10 percent for the rural population (Joint Monitoring Program 2015), producing a negative impact on public health and child mortality. A lack of adequate infrastructure and financing hamper the development of sanitation facilities in rural areas, while most urban cities lack sanitation master plans. Overall, poor households’ ability to escape poverty is hampered by low earnings, a lack of systemic risk-mitigation measures, and inadequate access to and inclusion within markets. Unless alleviated through both income support, increased public investment to improve the quality of basic services and robust mechanisms to promote their inclusion within markets, these disparities are likely to constrain intergenerational mobility across the welfare ladder. 11 UIS, IIEP, & Pôle de Dakar. (2016). Who pays for what in education? The real costs revealed through national education accounts. Montreal: UNESCO Institute for Statistics, International Institute for Educational Planning, and Pôle de Dakar. 12 UNICEF (2011). Demographic and Health Survey (DHS): Côte d’Ivoire. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 7 8 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH PART I: FINANCING AND INSTITUTIONAL REVIEW Prior to 2015, the main social protection programs in Côte d’Ivoire have been contributory social insurance programs, while labor programs have largely targeted urban, semi- and relatively highly-skilled workers. As of 2016, the Government has begun gradually introducing new programs and policy reform planning to expand coverage and effectiveness. Social insur- ance has namely been pensions covering civil and private sector workers in the formal sector, accounting for 10 percent of the population. The Government plans to continue reforms to expand social insurance coverage from 10 percent currently to 30 to 50 percent of the popula- tion by 2020. Other programs have included nutrition, school feeding initiatives, food-for-work schemes and active labor market programs. Most of these programs have been supported by financing from external resources including the World Food Programme (WFP), European Union (EU), UNICEF, the International Labor Organization (ILO) and the World Bank (WB). The first part of the policy note is structured as follows. Section 2 assesses the overall institutional arrangements and expenditure patterns for social protection and labor programs, followed by in-depth discussions in subsequent sections by thematic area. Section 3 discusses labor market policy arrangements and expenditures. Section 4 reviews pensions and social insurance schemes. Section 5 assesses productive social safety nets and social assistance, including service delivery arrangements for SPL programs, including identification, targeting, payment modalities and monitoring and evaluation arrangements. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 9 10 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 2 OVERALL SOCIAL PROTECTION AND LABOR SPENDING AND KEY INSTITUTIONS Comparative Expenditures The NDP calls for increasing access to social services for vulnerable groups, including women, youth, and persons with disabilities. The NSPS aims to improve social protection for the most vulnerable to mitigate against economic shocks and to improve their resilience in the long term. The NSPS is organized around four main strategic pillars: (a) improving living stan- dards for poor and vulnerable households; (b) improving access to basic social services and investing in human capital; (c) strengthening social action against violence, abuse, exploitation, discrimination and exclusion; and (iv) extending contributory, formal social insurance. Priority will continue to be given to pro-poor spending. The Government’s planned fiscal adjustment and public financial management program includes include, inter alia, a reprioriti- zation of current spending, the implementation of new revenue measures, and measures to streamline the public wage bill. The Government categorizes pro-poor spending as spending on eleven sectors ranging from agriculture and infrastructure to social expenditures (health, education, and social protection and labor). Between 2011 and 2014, pro-poor spending has been relatively stable at 40–44 percent of total public expenditure, or 9.1–9.5 percent of GDP.13 Despite the volume of pro-poor spending, few programs explicitly target the poorest households. Most expenditure has been on education, at 23 percent of total public expendi- ture and 59 percent of total pro-poor spending. Pro-poor “social” spending (primarily general administration, as well as in-kind social assistance programs) was on average 0.7 percent of total expenditure, while “other poverty-fighting spending” (youth and other programs) was 0.2 percent. 13 All figures based on updated estimates from Ivorian authorities as of September 2017. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 11 Using an economic classification of social protection and labor (SPL) programmatic expendi- ture in this analysis, total SPL spending is estimated at 1.66% of GDP as of 2016–2018. The analysis is based on available data for 2016 to 2018, as most spending remains stable. Out of a total of nineteen programs and initiatives identified, nine had sizable budget envelopes, beneficiaries, and available data. These programs were included in this analysis. Programmatic expenditures on SPL largely go to pension schemes for the non-poor (Figure 3). Spending (excluding general administration) included: contributory pensions schemes at 1.56 percent of GDP (57 percent of which is devoted to civil servants and 43 per- cent devoted to private, formal workers), non-contributory social assistance at 0.02 percent, active labor market programs at 0.06 percent, in line with other countries in Sub-Saharan Africa (SSA) at a similar development stage of SPL systems (Figure 4). Overall SPL spending as a percent of total public expenditure was 6.3 percent, 94 percent of which went to pension schemes, which are not considered pro-poor spending by the Government. The distribution of SPL financing in Côte d’Ivoire shows that most of the spending does not benefit the poorest households. Even when excluding spending on the private sector pension fund (co-financed by employers and employees) from the analysis, results are similar. 90 percent of SPL funding by the Government remains devoted to pensions largely benefiting the non-poor. By contrast, spending on subsidies and transfers (including fuel price subsidies to the national oil refinery, electricity and budget transfers to cover pension fund deficits) was double that of SPL spending, at 2.61 percent. Although Côte d’Ivoire spends nominally on fuel subsidies and has begun phasing them out, subsequent analysis on social safety nets shows how regressive this spending has been. Coverage of SPL programs in Côte d’Ivoire remains very limited, similar to other countries in Sub-Saharan Africa (SSA) (Figure 5 showing SSN comparisons; international comparisons of pensions indicators are discussed in later in this policy note; labor market coverage across countries will be added subsequently). Coverage has gradually increased between 2002 and 2014 (similar estimations for 2016–2018). Based on household survey data,14 in 2002, labor market programs covered an estimated 0.4 percent; all social insurance, 5.6 percent; and all social assistance covered 20 percent, including cash transfers15 (including cash-for-food programs) at 0.5 percent. By 2014 (consistent for 2016–2018), labor market programs cov- ered 0.3 percent; all social insurance, 7.9 percent; all social assistance covered 27.5 percent, including cash transfers at 0.2 percent. While this means that 33 percent of the population are covered in total, the bulk is in-kind food assistance, followed by pensions for formal sector workers. Given that 46 percent of the population is under the poverty line, determining the relative effectiveness of different SPL instruments can help maximize the value of spending for the poorest households. 14 Data on how beneficiaries are distributed across quintiles are available from the National Household Living Standards Survey, which may not accurately reflect the number of beneficiaries found in administrative databases. Data available from administrative databases is used in discussing each individual program in this policy note. 15 The type of programs included vary from period to period. For cash transfers, programs in 2002 (notably food-for- cash) and 2014 (notably unconditional, targeted cash transfers) are not necessarily directly comparable. 12 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 3  SPL and other public expenditures as percent of GDP, 2016–2018 6.00% Expenditures as % of GDP, 2015 5.38% 5.00% 4.00% 3.00% 2.61% 2.00% 1.66% 1.79% 1.56% 1.38% 1.14% 1.00% 0.34% 0.02% 0.06% 0.02% 0.00% ns an r y P es PL it s n lth se r t en rio fic er io LM s t to ic io en ce lS ea E d s) at de ns f nm te si bu rv ns lA ef ta H uc In Se bi ra ro Pe A s tri To D Al e rT vi al n al ry ci C o En ag t he ll ci to So So n - d bu ,w O an o nd lN it y nd tri er on Ki ric a Al at n- ct s lC W le ie lI (e b s i d Al Al Su Source: World Bank staff calculations using data from Ministry of Economy and Finance, Côte d’Ivoire. FIGURE 4  SSN expenditures as percent of GDP in Côte d’Ivoire and other low- and middle- income countries, 2016–2018 Spending in constant USD 2011 PPP 3.5 408 450 3.0 400 SSN spending, % of GDP 350 2.5 300 2.0 250 1.5 164 200 83 150 1.0 64 100 0.5 57 1 16 22 25 50 0 11 11 0.0 0 re n a ya i e ia l da na m nd sia sia a al ga ni ric oo qu op na oi M n ha an la ne ay za ne Ke Af bi Iv er ai et hi G Rw al n do d’ Se am m Th Et Vi h Ta M ut Ca In te oz So Cô M Per capita spending ($ PPP) SSN spending, % of GDP Source: World Bank staff calculations using most recently-available national data at the time of analysis. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 13 FIGURE 5  SSN coverage as percent of total population in Côte d’Ivoire and other low- and middle-income countries, 2016–2018 60% 55% 50% 47% 46% % of population 40% 38% 30% 28% 30% 24% 21% 20% 16% 10% 11% 10% 9% 8% 6% 4% 3% 0.10% 0.09% 0% ct al ra ild m r sf et ) P) m dy ) er nt s) an VC m fe du ion et SN ov me an N tg h am si ) ) e) ) e) ) ra ns ap ifs nt er ty N or (C O gr ub Tr ety ro at (P og tra t P er ar y or pp ica ro S N et ns w f a H as Sa rf pr sh al pi o ux e af su Afr ga fe ia m ga p ci o a lS l C al Em i P lc So m ns h h ar h i ia na oc Et P So ra ut na lu tra i oc ic d tio S ts og So Ke oo io as (S A h di ive c at le Pr as (B ih am n N Fi ( on ct el (C oo s re e l( gr ) C du iv qu ga i er a de Ivo (L ro ny N ro bi m ne (P na SS (P d' Ke am Ca Se sia ha (P nia te oz Co G ne a M nz do Ta In Benificiaries, share of population Poverty headcount at national poverty line (% of population) Source: World Bank staff calculations using most recently-available national data at the time of analysis. FIGURE 6  SPL and other public expenditures as percent of total public expenditures, 2016–2018 Water and Environment 1.1% Interior 3.6% Defense 4.4% Health 5.7% Education 17.0% Subsidies and Other Transfers 8.3% Total SPL 5.3% All In-Kind Social Services 0.1% All ALMP 0.2% All Non-Contributory Social 0.1% Assistance All Contributory Pensions 4.9% 0.0% 5.0% 10.0% 15.0% 20.0% Expenditures as % of total Public Exp, 2015 (%) Source: World Bank staff calculations using data from Ministry of Economy and Finance,. 14 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 7  Distribution of SPL expenditures by program type, 2016–2018 All ALMP All In-Kind Social 3.6% Services 1.0% All Non-Contributory Social Assistance 1.5% All Contributory Pensions 93.9% Source: World Bank staff calculations using 2015 and 2016 data from Ministry of Economy and Finance. TABLE 2  SPL expenditures by program type (local currency and USD), 2016–2018 Dépenses annuelles 2015– Sous-secteur 2016 (en milliards de FCFA) (Mio USD) Toutes pensions contributives 247,8 495,6 $ Toute assistance sociale non contributive 3,8 7,6 $ Tous PMTA 9,5 19,1 Tous services sociaux en nature 2,6 5,3 Total 263,8 527,6 USD Source: World Bank staff calculations using data from Ministry of Economy and Finance,. * Annual social assistance spending is expected to increase in 2018–20 when a new national system is introduced, estimated at US $ 15 million per year starting in 2018. Steps have already started to shift public expenditures with a more targeted, focused approach to poor and vulnerable households. Within the framework of the NSPS, the Ministry of Employment and Social Protection (MESP) plays an important role in coordinating key programs. Prior to the creation of the NSPS steering committee in 2015, coordination across programs has been limited since financing, implementation, program databases, policies, targeting and delivery mechanisms remain largely independent. New national programs and initiatives are providing the tools to better identify the poorest households, their needs and delivery of investments and services to them. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 15 16 Coverage rate, percent of population (%) Al ls Al ls Coverage rate, percent of oc oc total population (%) ia 0 5 10 15 20 25 30 35 40 45 lp ia Al lp FIGURE 9  FIGURE 8  ro ls ro 0 5 10 15 20 25 30 35 te Co oc te ct io nt ia ct Al n rib li io ls O ut ns n ur 33.3 oc th or ia Al er y an li Pa ll so pe ce ns 7.9 ur ss ab ci ns Co an iv or al io nt ce e m in ns Ca la ar su rib ut sh Ac bo ke ra 6.4 or t ra tiv e rm t nc y ns la ar pr og e 1.8 pe fe bo ke O r, t ra ns io a rm m th llo ar pr s 0.3 er ns w og so Co Al ke ra an t m Q1 (poorest) ci nd ls ce pr s 0 al iti s, oc og Al in on la ia ra ll su N al st la m ab ra on re ss s SPL coverage by region, 2015 ca is 0.3 or nc -c sh so ta e Q2 m on tra rt nc ar ke tri bu ns pr og e 27.5 t to fe ra Urban pr Fo ry rp m og od so ro s Q3 0.2 ra gr Al m an d ci am ls s al in pe s 0 oc ia Fe Pu -k ns la in io e d Rural Fo bl w Q4 ss ic ns 0 tra od is ta ai w Sc ns ve or ho an nc O rs ks fe Fe d e th & ol rs e in er an d fe 26.9 w -k in so fo ed ai ta od in Total ve d ci rg fo g 0 rs tra al et an ns as ed rw d fe si or st su k 0 ta rs Q5 (richest) rg an bs et D ce id ed om pr ie D su In es og s Al om bs te t ic lr rn em ram es id s Source: World Bank staff calculations based on ADEPT analysis using 2015 Household Budget Survey. Source: World Bank staff calculations based on ADEPT analysis using 2015 Household Budget Survey. ie at pr io i it t Total t ic 0.3 0.5 pr s na vat an iv lp et ce at e ri v ran s at sf tra ns e e tra r s fe 20.4 20.4 rs ns fe rs 0 Coverage (percentage within each quintile) of SPL programs by program type, 2015 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 3  SPL coverage (percentage within each region) by program type, 2015 (detailed data) Area of residence Program Total Urban Rural Direct and indirect beneficiaries All social protection 33.3 34.3 32.3 All social insurance 7.9 10.8 5.1 Contributory pensions 6.4 8.2 4.7 Other social insurance 1.8 3.1 0.5 All labor market programs 0.3 0.4 0.1 Passive labor market programs 0.0 0.0 0.0 Active labor market programs 0.3 0.4 0.1 All social assistance 27.5 26.3 28.7 Cash transfer, allowances, last resort programs 0.2 0.2 0.1 Conditional cash transfer programs 0.0 0.0 0.0 Non-contributory social pensions 0.0 0.0 0.0 Food and in-kind transfers 26.9 25.4 28.2 School feeding 0.0 0.0 0.0 Public works & food for work 0.0 0.0 0.0 Fee waivers and targeted subsidies 0.3 0.4 0.2 Other social assistance programs 0.5 0.6 0.4 All remittances 20.4 23.1 17.8 Domestic private transfers 20.4 23.1 17.8 International private transfers 0.0 0.0 0.0 Source: World Bank staff calculations based on ADEPT analysis using 2015 Household Budget Survey. Notes: Program coverage is the portion of population in each group that receives the transfer. Specifically, coverage is: (Number of individuals in the group who live in a household where at least one member receives the transfer)/(Number of individuals in the group). Program coverage is calculated setting as expansion factor the household expansion factor multiplied by the household size. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 17 TABLE 4  SPL coverage (percentage within each quintile) by program type, 2015 (detailed data) Quintiles of per capita consumption Percent reporting receiving benefits (%) Program Total Q1 Q2 Q3 Q4 Q5 Direct and indirect beneficiaries All social protection 33.3 39.8 38.5 36.1 29.7 22.4 All social insurance 7.9 5.4 6.4 7.2 8.5 12.0 Contributory pensions 6.4 5.3 5.8 6.4 6.7 7.7 Other social insurance 1.8 0.1 0.7 1.1 1.9 5.1 All labor market 0.3 0.2 0.3 0.3 0.2 0.3 programs Passive labor market 0.0 0.0 0.0 0.0 0.0 0.0 programs Active labor market 0.3 0.2 0.3 0.3 0.2 0.3 programs All social assistance 27.5 36.6 34.6 31.5 23.0 11.8 Cash transfer, allow- 0.2 0.2 0.4 0.0 0.1 0.1 ances, last resort programs Conditional cash transfer 0.0 0.0 0.0 0.0 0.0 0.0 programs Non-contributory social 0.0 0.0 0.0 0.0 0.0 0.0 pensions Food and in-kind 26.9 36.4 34.0 31.1 22.3 10.7 transfers School feeding 0.0 0.0 0.0 0.0 0.0 0.0 Public works & food for 0.0 0.0 0.0 0.0 0.0 0.0 work Fee waivers and targeted 0.3 0.1 0.4 0.1 0.4 0.5 subsidies Other social assistance 0.5 0.3 0.2 0.6 0.5 0.7 programs All remittances 20.4 20.1 18.2 19.6 22.6 21.5 Domestic private 20.4 20.1 18.2 19.6 22.6 21.5 transfers International private 0.0 0.0 0.0 0.0 0.0 0.0 transfers Source: World Bank staff calculations based on ADEPT analysis using 2015 Household Budget Survey. Notes: Program coverage is the portion of population in each group that receives the transfer. Specifically, coverage is: (Number of individuals in the group who live in a household where at least one member receives the transfer)/(Number of individuals in the group). 18 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 5  Overview of main SPL programs and policies in Côte d’Ivoire, 2016–2018 Sub-sector (number of programs) Main programs Social Assistance (3) • Programme Intégré de Pérennisation des Cantines Scolaires (PIPCS) • Programme National des Filets Sociaux Productifs • Programme National de prise en charge des Orphelins et autres Enfants rendus Vulnérables du fait du VIH/SIDA (PNOEV) In-Kind Social Services (3) • Regional Social Assistance Offices and Social Workers (739 SW) • Système d’Observation et de Suivi du Travail des Enfants en Côte d’Ivoire-SOSTECI • Plan d’Action National (PAN) 2012–2014 de lutte contre la traite, l’exploitation et le travail des enfants Labor Market Programs (8) • Projet d’Appui au Traitement Economique du Chômage (PATEC) • Programme Senior (PATEC-modifié) • Projet Emploi Jeune et Développement des Compétences (PEJEDEC) • C2D Emploi • C2D Emploi/SCAED • Programme Spécial de Requalification (Phase 2) • Projet d’insertion Socio-Economique (PRISE) • Programme d’Appui à l’Amélioration de l’Employabilité et de l’Insertion des Jeunes— PAAEIJ (THIMO BAD) Inter-sectoral SP Programs (3) • Programme du Plan National Multisectoriel pour la Nutrition 2016–2020 • Campagne «zéro grossesse à l’école» • Projet d’Appui et d’Accès des Personnes Handicapées à la Formation Professionnelle Pensions and Social Insurance • Caisse Nationale de Prévoyance Sociale (CNPS) (Contributory) (2) • Caisse Générale de Retraite des Agents de l’Etat (CGRAE) Total number of main programs: 19 Source: World Bank staff compilation. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 19 TABLE 6  Overview of main SPL programs in Côte d’Ivoire (2017) included in this assessment based on program aims and data availability Sub-sector Type of program Programs evaluated for this analysis Social Assistance (3) School feeding / take-home Programme Intégré de Pérennisation des Cantines Scolaires (PIPCS) Cash transfers Programme National des Filets Sociaux Productifs Other food/in-kind program/transfer Programme National de prise en charge des Orphelins et autres Enfants rendus Vulnérables du fait du VIH/SIDA (PNOEV) Pensions and Social insurance (2) Old age pension (private) Caisse Nationale de la Prévoyance Sociale (CNPS) Old age pension (public) Caisse Générale de Retraite des Agents d’Etat (CGRAE) Labor Market Programs (3) Training (vocational, life skills, cash for Projet Emploi Jeune et Développement training) des Compétences (PEJEDEC) (programme national) Entrepreneurship support / Startup Projet C2D-Emploi incentives (cash and in kind grant, loans, training) Projet d’Insertion Socio-Economique des populations vulnérables dans l’ouest ivoirien (PRISE) In-Kind Social Services (1) Regional Social Assistance Offices and Social Workers (739 SW) Total number of programs evaluated: 9 Source: World Bank staff compilation. Key Institutions The MESP overseas SPL policy and implementation of the NSPS and national employment strategy through a network of local structures at the department and community levels, in coordination with other ministries and agencies (Figure 10). The annual operating budget for regional and local structures amounted to approximately 830 million FCFA as of 2015–2018, or the equivalent of US$37 million (Table 7). At the central level, the MESP has four Directorates, which are the General Directorate of Employment, General Directorate of Labor, General Directorate of Social Affairs and General Directorate of Vocational Training. Implementation of the Government’s social policy is overseen by the General Directorate of Social Affairs. At the local (deconcentrated) level, the Ministry operates a network of regional directorates, social centers at the community level, and different categories of social workers (general social workers who comprise the largest group, followed by educational assistants, youth workers, 20 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 10  Institutional mapping of key SPL actors Prog. Inter- - ministeriel (i.e., CNN) Direction du Min. Budget Budget Dir. Cantines Min. Education Direction Mutualite Presidence Primature PATEC (1 &2) PEJEDEC Min. Youth and Agence Emploi civic Jeune BCP-Emploi C2D Emploi Direction Prot. Soc. PRISE Min. ESP DG Travail DG Emploi UGP/Filets Sociaux Min Solidarite Source: World Bank staff and MESP, Republic of Côte d’Ivoire. etc.). As of 2018, there were 19 regional directorate offices, with an estimated 10 new offices planned over 2016–2020. The MESP also comprises 14 regional directorate offices. The MESP regional directorates mainly manage and coordinate the work and activities of social centers and social workers at the community level. At the local community level, social centers and social workers provide the selected social services including surveillance of community health, psychosocial development of mothers and children less than five years of age, family planning and education targeting women and youth, case management for specific risks as they arise, and overall support to facilitating improvements in living conditions for vulnerable groups. Côte d’Ivoire has regional social centers, with nearly 100 located through- out the country’s 81 departments, in which a total of approximately 730 social workers serve as of 2014. In terms of geographic distribution, the southern zone accounts for the vast majority of social centers at 42 percent and social workers at 34 percent, leaving wide disparities in access elsewhere (Figure 11 and Figure 12). The Government has also recently launched a continuing education program to build technical capacity among social workers in advance of project implementation and the development of an integrated information system to collect and monitor welfare information provided by social workers. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 21 TABLE 7  Local structures and operating budget of MESP and associated institutions, 2016–18 Structure Budget (FCFA millions) 67 centres sociaux du mesp dotes pour l’exercice 2014 404 19 directions regionales des affaires sociales du MESP 223 Complexes socio-educatif Ministère de la Solidarité (20) 108 Directions regionales Ministère de la Solidarité (14) 93 TOTAL 827 Source: World Bank staff and MESP, Republic of Côte d’Ivoire. FIGURE 11  Geographic distribution of social centers by geographic zone, 2016–2018 Ouest 15% Nord Sud 10% 42% Est 18% Centre 15% Source: MESP, Republic of Côte d’Ivoire. 22 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 12  Geographic distribution of social workers by regional directorates, 2016–2018 Worodougou et Béré Agnéby Tiassa et Mé 4% 5% Sud-Comoé Tonkpi Bélier Bounkani-Gontougo 8% 2% 4% 3% Poro-Tchologo et Bagoué Denguélé-Folon et Bafing 3% 2% N’Zi I ou Gbeke-Hambol 4% 6% Marahoué 3% Gbokle-Nawa-San-Pedro 5% Lôh-Djiboua Gôh 4% 3% Guemon-Cavally 3% Haut-Sassandra 3% Indénié-Djouablin Lagune 2 4% 18% Lagune 1 16% Source: MESP (formerly MEMEASFP), Republic of Côte d’Ivoire. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 23 24 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 3 LABOR MARKET PROGRAMS The key employment challenge Côte d’Ivoire faces is the significant lagging productivity of jobs, with low earnings and precarious working conditions rampant. Particularly vulnerable jobs include agricultural and non-agricultural self-employment in services and commerce where market access is weak, particularly among the poor, women and those living in rural areas. In evaluating public policies to support employment promotion and insertion among youth and vulnerable groups, this section sheds light on the emphasis of these programs in terms of populations, areas of intervention, relative costs and, ultimately, whether the mix of interventions and investments that exists responds sufficiently to Côte d’Ivoire’s diverse employment challenges moving forward. This section assesses spending on active labor market programs (ALMPs) in Côte d’Ivoire, a component of a broader jobs strategy16 to tackle these challenges. The assessment focuses on mainly on the structure of program spending and coverage, with an overview of key aspects of targeting, cost-effectiveness, monitoring and evaluation, and institutional arrange- ments useful for examining the potential to maximize the effectiveness of ALMPs. The section starts with a description of all ALMPs in the country, and followed with a detailed analysis of selected characteristics, say of a selected program.17 While developing the modern wage sector and improving access to formal employment is necessary over the long-term, over the 16 Sources: BCP-Emploi. 2017. Rapport de Suivi Financier Nr 20 au 31 Décembre 2016. Bureau de Coordination Programme-Emploi (BCP-Emploi), Abidjan : Côte d’Ivoire. Heckman, J. J. and J. A. Smith. 2004. “The Determinants of Participation in a Social Program: Evidence from a Prototypical Job Training Program,” Journal of Labor Economics 22 (2): 243–298. Ngom, D. and B. M. Kouadio. 2016. Evaluation de la Politique Nationale de l’Emploi 2012–2015. Rapport Provisoire, Abidjan : Côte d’Ivoire. PNE. 2016. Politique Nationale de l’Emploi (PNE) 2016–2020. Ministère de l’Emploi et de la Protection Sociale (MEPS), Abidjan : Côte d’Ivoire. World Bank. 2015. Côte d’Ivoire: Productive Social Safety Net Project. Project PAD, the World Bank: Washington, DC. 17 Forthcoming analysis will provide an in-depth understanding of the use of different instruments for different policy objectives and target populations, based on ongoing discussions with the Government and updated information on programs managed by Agence Emploi Jeunes and Direction Générale de l’Emploi, and will be reflected to the extent possible in this note. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 25 short- to mid-term, it is unlikely that even strong growth of that sector would absorb more than a small share of the population, primarily more highly educated, urban and non-poor popula- tions. ALMPs, if designed and targeted well for an emerging economy, could play an important role in Côte d’Ivoire’s economic transformation. Labor Market Agencies Côte d’Ivoire’s national employment policy strategy for 2016–2020 lays out key priorities for enhancing access to and quality of jobs, particularly among youth. The strategy identifies five main areas for intervention, including: (i) improving working and living conditions of workers; (ii) increasing employment opportunities regionally and at the local levels; (iii) strengthening mechanisms to promote employability and entrepreneurship, particularly among youth; (iv) pro- moting decent work for youth, women, and people with disabilities; and (v) improving the labor market information system and job matching mechanisms. Côte d’Ivoire’s labor code lays out a host of regulations in line with international conventions for decent work and standard legisla- tion, yet enforcement is limited to some segments of the formal sector at best. Given that over 80 percent of the labor force is employed in the informal sector, the role of labor regulations in protecting most workers is limited. The minimum wage remains at FCFA 60,000 per month, having increased in 2014 from FCFA 36,607 per month. Coverage of ALMPs is relatively limited in Côte d’Ivoire, which has been challenging to improve given little coordination among agencies. Two ministries oversee national employ- ment policies and/or targeted ALMPs. At a policy level, the Ministry of Employment and Social Protection [Direction Générale de l’Emploi (DGE)] is responsible for labor legislation, social dialogue, and employment strategies. As part of the national employment strategy, in January 2016 the new Ministry of Youth and Youth Employment was created, whose mandate is to manage key ALMPs targeting youth falling under two different agencies, Agence Emploi Jeune and Bureau de Coordination des Programmes-Emploi (BCP-Emploi)]. Various units exist that focus on a range of areas such as entrepreneurship, apprenticeship, public employment services targeting university graduates, monitoring, and, recently, employment support to vulnerable and poor communities. While labor code regulations and labor support to vulnera- ble populations remain with the Ministry of Employment and Social Protection, the division of ALMP responsibilities across ministries is not yet fully clear. The institutional arrangements for labor policy and governance remain complicated, with multiple employment-related agencies and programs operating largely independently and are often duplicative. Expenditure Patterns ALMPs in Côte d’Ivoire are largely financed through external resources, although since 2011 the share of government spending has been increasing steadily (Table 8). Several programs are coordinated and implemented by public agencies housed under the Ministère de la Promotion de la Jeunesse, de l’Emploi et du Service Civique, having moved from the Ministry of Employment and Social Protection during a cabinet reshuffle in January 2016. Most these 26 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH programs (six out seven) were launched following the political stalemate of 2011, with one that dates to 2010. As of 2016–18, the total estimated budget of all labor programs was around FCFA 40,994 billion (or approximately USD 74 million), of which 5% was financed through gov- ernment sources and 95% through external sources. Most of the labor programs are designed to cover young (18–35 years old) unemployed/ under-employed, and vulnerable individuals. Only one program (PATEC) is designed for long-term senior unemployed individuals. However, available documentations indicate that the target group and the beneficiary eligibility criteria are not clearly defined in most programs, particularly programs that target youth at risks and vulnerable population. There is no consen- sual definition of vulnerability across programs. In terms of cost-effectiveness, most of the programs lack sufficient data on detailed pro- grammatic costs and impacts to robustly assess cost-effectiveness, but indicative trends can be gleaned nonetheless. Distinguishing administrative cost from total is often not avail- able for most programs. However, recent and ongoing projects implemented by the Bureau de Coordination Programme-Emploi (BCP-Emploi) have begun to systematically collect this information. A recent evaluation of government employment policy between 2012 and 2015, indicates that the cost per beneficiary of the labor market programs in Côte d’Ivoire ranges from 193,000 to 1,268,000 F CFA, or approximately USD 325 to 2,100.18 The program with the highest cost per beneficiary is the Programme Spécial de Requalification (Phase 2), while the Projet d’insertion Socio-Economique (PRISE) has the lowest cost per beneficiary. Implementation Arrangements, Monitoring and Evaluation Coordinating the multiplicity of actors and program implementation remains a challenge. Due to the multiplicity of actors across several ministries most programs on labor market operate in silos, and often leading to each ministry setting its own agendas, with little com- munication and synergy of actions. For instance, in addition to the Agence Emploi Jeune, in the Ministry of Youth, there is also the Direction Générale de l’Emploi (DGE), at the Ministry of Employment and Social Protection. For a program with multiple components such as PRISE, several ministries are involved; as specific components are addressed to specific target group, ranging from youth (18–35 years old) to vulnerable populations. The latter is a priori under the responsibility of the Ministry of Employment and Social Protection, while the former is under the responsibility of the Ministry of Youth. Only programs coordinated by the BCP-Emploi have been monitored on a regular and systematic basis since 2011. The BCP-Emploi provides implementation reports on a quarterly basis for PEJEDEC, PRISE, and C2D Emploi/AFD. Information on other ALMPs are limited, and these programs are often not evaluated since inception. Institutional arrangements have 18 Republic of Côte d’Ivoire and International Labor Organization (Ngom D and Kouadio B) (2016). Evaluation de la Politique nationale de l’emploi, 2012–2015. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 27 directly affected both implementation and the coordination required for effective monitoring and evaluation of program processes and impacts. Government reshuffling has contributed to delays in implementation and budget execution. For example, the Programme Spécial de Requalification (Phase 1 et 2) was initially coordinated by the AGEPE. However, as the agency has been replaced with the Agence Emploi Jeune, clear lines of responsibility for that program in the Ministry of Youth have not been established. Targeting Key challenges in implementing ALMPs in Côte d’Ivoire appear to be on targeting and cover- age. This section highlights these issues with the case of PEJEDEC as a case study, given the data available from this program. The PEJEDEC project, established in 2011, aims to improve access to temporary employment among lower-skilled young (18–30) men and women in urban or semi-urban areas in Côte d’Ivoire that are unemployed or underemployed, as well as to develop skills for the program participants through this work experience and, for some par- ticipants, complementary training. The program has two operational components (temporary employment opportunities; skills development and employment support), and one component on institutional support. Although the PEJEDEC targets a specific age group (18–30 years old), participation to that program is on self-selection basis. Since inception, there has been an increase in program uptake, driven in particular by the number of applicants for the internship component (Table 9). The increased participation to the program may be due the substantial effort of the project coordinating unit to increase awareness among the population.19 In the case of PEJEDEC, project communication expenditures have been around 5 to 7 percent of total institutional capacity building expenditures. In addition, the program also specifically targets 30 percent of female participation into the program. The PEJEDEC monitoring documents show that this target has been met since inception. For instance, as of December 2016, 35 percent of the participants were women (BCP-Emploi, 2017). Furthermore, two impact evaluations conducted on this program, together with a forthcoming impact evaluation of the PRISE program, highlight important lessons.20 The results collectively show the importance of targeting ALMP programs for improved effectiveness. 19 As indicated in Heckman and Smith (2004), awareness of eligibility has been identified as a major source of program participation elsewhere. 20 Source/detailed findings to be added based on World Bank report (Crépon, Premand). 28 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 8  National labor market insertion projects and programs Annual budget Program Target (F CFA) Financing source Enrolment PEJEDEC Jeunes de 15 à 30 25 milliards IDA/Banque Mondiale 27 000 ans PAEIJ (THIMO BAD) Jeunes descolarisés 1.81 milliards BAD 2 000 Programme THIMO Jeunes et femmes 532 million Etat/FSE-THIMO 465 gouvernemental (FSE-THIMO) C2D Emploi Jeunes de 15 à 35 9 milliards C2D/AFD 18 248 ans C2D Emploi/SCAED Jeunes 1.2 milliards C2D/AFD 1 000 Programme Spécial de Diplômés chômeurs 713 millions Etat/AGEPE 500 Requalification (Phase 1) de longue durée Programme Spécial de Diplômés chômeurs 528 millions Etat/AGEPE 650 Requalification (Phase 2) de longue durée Projet d’insertion Socio- Population 1.358 milliards Japon (JSDF) 7 000 Economique (PRISE) vulnérable Projet d’Appui au Traitement Chômeurs de 18 à 361 millions Budget de l’Etat 722 Economique du Chômage 55 ans (PATEC) TOTAL 40.494 milliards 58 085 Source : Rapport de mise en œuvre de la Stratégie de Relance de l’Emploi/année 2014 ; PNE 2016 TABLE 9  Distribution of job seekers and beneficiaries by type of program 2013 2014 2015 2016 Job Job Job Job seekers Beneficiaries seekers Beneficiaries seekers Beneficiaries seekers Beneficiaries THIMO 20,947 6,381 26,909 12,693 26,909 12,693 26,909 13,693 Stage 132,053 1,827 161,324 2,388 173,364 2,388 173,364 2,388 Apprentissage 10,284 2,652 11,632 4,295 12,741 4,405 12,741 4,907 Formations 3,500 970 3,517 987 3,517 987 3,517 987 professionnelles Entrepreneuriat 258 220 258 4,019a 258 220 258 220 AGR — — — — 7,568 6,870 c 16,068 10,528d Total 167,042 12,050 203,640 22,525 b 224,357 27,563 232,857 32,723 Note : a) Y compris les 1 857 bénéficiaires THIMO et 1 942 bénéficiaires du projet PRISE formés en AGR ; b) Ce chiffre ne prend pas en compte les 1 857 bénéficiaires THIMO formés en puisqu’ils sont déjà comptabilisé au niveau de la sous composante THIMO ; c) Y compris les bénéficiaires THIMO et bénéficiaires du projet PRISE formés en AGR ; d) Y compris les bénéficiaires THIMO et bénéficiaires du projet PRISE formés en AGR Source : BCP-Emploi, Rapport de Suivi Financier Nr8, Nr12, Nr16, Nr20 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 29 Conclusions Moving forward, coverage and eligibility of most ALMPs in Côte d’Ivoire remains a key issue, in addition to inter-agency and inter-sectoral coordination. Although most of the existing ALMPs are national per design, most them do not covered all necessary areas, in particular rural areas, where most of the poorest households are concentrated. Using the case of, the PEJEDEC is limited to urban, and semi-urban areas; excluding mostly un-educated young people in rural areas. Implementation capacity remains limited in terms of public employment agencies, but the scope for greater public-private partnerships in rural regions and those that maximize local value chains have begun to garner the interest of governmental and non-governmental stakeholders. 30 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 4 SOCIAL INSURANCE AND PENSIONS This section provides a comparative overview of Côte d’Ivoire’s pensions systems identi- fies areas for future analysis, given very limited coverage and information on other social insurance schemes. Coverage of social insurance schemes is very limited and sufficient data has not been readily available on key programs (such as unemployment protections, as well as maternity benefits and other income support). Pensions and old age savings programs in Côte d’Ivoire exist but remain limited to a small share of the population. Two pension funds operate which include the national pensions for private sector workers and the pensions fund for civil servants, both operating under the auspices of the Ministry of Employment and Social Protection and the Ministry of Economy and Finance. Together these funds cover only 6–10 percent of the population as of 2018, benefiting only civil servants and workers in the formal private sector. The Government plans to continue reforms to expand social insurance coverage from 10 percent currently to 30 to 50 percent of the population by 2020. To date, however, little is known on the performance of the pension funds or feasibility of expanding coverage in the face of high poverty and informality. Overview of Pension Funds The pension sector in Côte d’Ivoire is comprised of two separate pension schemes—one covers private formal sector workers and is managed by “Caisse Nationale de Prevoyance Sociale” (CNPS) and the other covers public sector workers and is managed by “Caisse Generale des Fonctionnaires et Agents de l’Etat” (CGRAE). Both schemes are Defined Benefit21 (DB) schemes and are financed on a pay-as-you-go22 (PAYG) basis. CNPS provides 21 Defined benefit scheme is a type of pension scheme in which an employer promises a specified periodic benefit on retirement that is predetermined by a formula based on the employee’s earnings history, tenure of service and age. It is ‘defined’ in the sense that the formula for computing the employer’s contribution is known in advance. In a defined benefit scheme, the employer bears both the investment risk and the longevity risk. 22 Pay-as-you-go (PAYG) based financing: contributions from current workers and their employers being used to finance benefits for current pensioners. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 31 four sets of benefits, namely family allowance, maternity, work injury and pensions (old-age, survivorships and invalidity). The scheme also manages a health fund that provides health ben- efits to its members and their families. CGRAE currently provides two types of social security benefits, namely pensions (old-age scheme, survivorship, and invalidity) and family allowances. The focus of this analysis falls on pension benefits. Having separate pension schemes for public and private sector workers is not uncommon in the Sub-Saharan Africa region; 24 out of the 44 countries where data are available have separate pension arrangements for public and private sector workers. For many reasons, fully or at least partially integrated pension systems are preferred over having a separate pension system specifically for employees of the public sector (only four countries of the OECD—Belgium, France, Germany, Korea—have entirely sep- arate institutions and benefits for civil servants). (Palacios & Whitehouse, 2006) Demographics and Membership Côte d’Ivoire is currently one of the youngest countries in the world. In 2017–18, population above age 60 accounts for less than 5% of the total population and 53% of the population is between the ages of 15 and 60—more than 70% of the population is below the age of 30. UN population projections show that while the working-age population is projected to increase from 53% to 60% between now and 2080, the share of 60+ in total population is projected to increase even more, reaching 11% by 2080. The two-fold increase means that there will be more than 9 million individuals over the age of 60 by 2080. A shift of this magnitude means spending on the elderly will likely also increase from its current levels. Coverage of the pension sector in Côte d’Ivoire is currently low, though it is in line with that observed in most of Sub-Saharan Africa (on average about 10% of the population contributes to private sector pension schemes, though there is considerable variation between countries). Currently, the CNPS pension system covers 9.5% of the population above retirement age (age 60) and only 5.4% of the population between the ages of 15 and 60. In 2016–2018, CNPS provided old age, disability and survivor pensions to only an estimated 104,636 individuals. This means that, for the clear majority of the population, the responsibility to ensure old age income security currently rests with individuals and their families. The extremely low coverage of the working-age population implies that this is unlikely to change significantly for future cohorts of retirees—presently around 5.4% of working-age individuals are accruing pension rights in the CNPS pension scheme. Similarly, active and elderly coverage is also very low in the public-sector scheme (See Table 10). Low coverage rates among the active population is reflective of the labor market characteristics in the econ- omy—both pension schemes extend coverage only to formal public and private sector workers with employment contracts whereas the clear majority of the population is employed in the informal economy, many of whom in rural agriculture. The rate of elderly coverage is more than double the current coverage rate of the working-age population. This could be explained by the relatively low retirement age and length of service requirement for an old age pension. With the retirement age set to 60 and the minimum service period set to 15 years, it is possible that workers may only be contributing for a third of their full career and still meet the eligibility 32 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 10  Main pension system performance indicators, 2016–2018 CNPS CGRAE Number of contributors 672,251 240,556 Numbers of beneficiaries 104,636 82,867 System dependency ratio 15.6% 34.4% Contributors, % 15–60 population 5.6% 2.0% Contributors, % 15–64 population 5.4% 1.9% Beneficiaries, % 60+ population 9.5% 7.5% Total contributions 228 157.1 Total benefits, billions 107 141.45 Total benefits, % GDP 0.6% 0.8% Source: World Bank staff compilation. Estimates based on analytic extrapolations. conditions for a pension. For example, coverage of the working age population of 33% could still be consistent with 100% coverage among the elderly. While both schemes provided pensions to a comparable number of beneficiaries in 2016– 2018, CGRAE had only 240,556 contributors, which is about a third of the number of indi- viduals contributing to CNPS, resulting in a very high system dependency rate of 34%. This means that each pensioner is supported by just 3 contributors (this ratio is 1 to 6 in CNPS). Due in part to these unfavorable system demographics, more than 90% of contribution revenue was immediately needed to pay current pension benefits in CGRAE, compared to only 50% in the private sector scheme. The relatively higher system dependency rate in the public-sec- tor scheme is due to the system being much more mature (when pension systems are first introduced, there are no pensioners and only contributors, overtime the number of individuals eligible for a pension grows resulting in an increasing system dependency rate and pension spending) than the private sector scheme—CGRAE was set up in 1961 whereas CNPS was only introduced in 2000. Key System Parameters Several of the parameters in both schemes suggest that the systems may be promising overly generous benefits. Table 11 presents a summary of the main design characteristics of the pension sector in Côte d’Ivoire. The CNPS benefit formula includes a linear accrual rate of 1.7% of pensionable salary per year—the accrual rate is 1.75% in the CGRAE scheme. This means that a worker can expect to receive 50% of previous earnings after a 30-year long career, compared to the 40% suggested by ILO. In comparison, the regional average accrual CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 33 TABLE 11  Main parameters of the private and public sector pension schemes Parameters CNPS CGRAE Category of covered population Formal private sector workers Civil servants and Military Contribution rate (% of covered wage) 14% (5.6% employee and 25% (8.3% civil servant, 8.4% employer) 16.67% government) Accrual benefit rate 1.70% 1.75% Required basic length of contribution 15 15 Retirement age 60 60 and 65 (A4 and above)* Maximum replacement rate 50% No ceiling Number of year for base wage calculation 15 5 Pension indexation Inflation Wage growth Survivors pensions 50% 50% Source: World Bank staff compilation. * Civil Servants from category D to A3 retire at 60 and those from Category A4 and above retire at 65. rate is 1.9% which results in close to 60% replacement rate. While lower than that in some neighboring countries, the CNPS accrual rate is still higher than the average of 1.2% per year observed in high income countries. Accrual rates in high income countries are considered international best practice, although even these might not be fiscally sustainable in the longer run. The wages on which the pension benefit is calculated—the pension’s assessment base or reference salary—include the last 15 years in CNPS and only the last 5 in CGRAE. Computing pensions based on the average salary of the last few years could prove expensive for the pension system as contributions paid earlier in the career on a lower salary are not accounted for in the calculation of the pension. Typically, civil servants see steady increases in their sal- aries, as salaries are often based on seniority. Private sector salaries, by contrast, often peak earlier in the career around the late 40’s or early 50’s and then either hold steady or decline slightly. Thus, the ratio of benefits received to contributions paid may be markedly higher for civil servants than it is for private sector workers—resulting in overpayment of benefits for civil servants compared to contributions paid. International best practices recommend basing pen- sions on lifetime average wages. The rationale for this is that if a worker has paid contributions over the lifetime on the wages earned, then aligning benefits with contributions would imply basing the pension benefit on the average lifetime wage, since the average contribution was based on the lifetime average wage. In addition, extending the reference salary to full career will also eliminate inequities arising due to public and private sector workers having different age-earnings profiles. Pensions post-retirement in the CNPS scheme are increased in line with cost of living increases whereas pensions in CGRAE are indexed to wage growth which is typically the more generous of the two. Figure 13 and Table 12 present the different choices made by countries in 34 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 13  Pension contribution rates in Sub-Saharan Africa, 2016–2018 Source: World Bank Africa Pensions Database TABLE 12  Indexation policy choices made by countries in Sub-Saharan Africa Ad-hoc Wage Growth Inflation Angola INSS Benin FNRB Côte d’Ivoire CNPS Zimbabwe NPS Senegal IPRES Seychelles SPF Zambia PSPF Côte d’Ivoire CGRAE Sao Tome National Institute of Social Security (INSS) Madagascar Caisse Nationale de Zimbabwe PSPF Prévoyance Sociale (CNaPS) Kenya Civil Service Pension Scheme Togo Pension of Civil and Military (CSPS) Madagascar CRCM Officials (CRT) Mauritius National Pensions Fund (NPF) Madagascar CPR Zambia NAPSA South Africa GEPF—at least 75% Tanzania NSSF Ghana SSNIT inflation Tanzania PSPF Sierra Leone National Social Security and Insurance Trust Cabo Verde (INPS) Uganda PSPF Cabo Verde (AP) Mauritius Civil Service Pension Scheme Cameroon CNPS (CSPS) Source: World Bank Africa Pensions Database Sub-Saharan Africa, which still have young demographics and generate surpluses in their pen- sion funds, with respect to indexation policy. The clear majority of countries in Sub-Saharan Africa adjust pensions to wage growth or on an ad-hoc basis, which approximates wage growth and is typically higher than inflation. International best practice suggests indexation CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 35 post-retirement by inflation only, with the logic that an individual’s purchasing power should be maintained from the first day of retirement throughout the retirement period. Higher income countries also often used to index to wage growth but have had to move to inflation index- ation as their systems matured. Pensions are financed from employee and employer contributions at the rate of 14% and 25% of wages in CNPS and CGRAE respectively. Across the entire region, contribution rates in public sector schemes tend to be higher than those in private sector schemes. The CNPS con- tribution rate equals the regional average private sector contribution rate, whereas the CGRAE contribution rate ranks among the highest across public sector pension systems in Sub-Sharan Africa. There are some important differences in key system parameters between the public and private pension schemes. The public scheme includes a more generous pension assessment base and indexation mechanism, a higher accrual rate and no ceiling on replacement rates as compared to the private sector pension scheme. As shown in Figure 14, the public-sector scheme provides pensions to fewer beneficiaries but has higher expenditures. To support the higher benefits, CGRAE also has a markedly higher contribution rate which in the context of a civil service scheme where the Government is the employer is simply another expenditure item for the Government. In addition, the Government shoulders a larger share of the total pension contribution rate (67% vs. 60% in the private sector scheme) in the CGRAE scheme. In 2016–2018, total pension expenditures (both schemes combined) amounted to 1.4% of GDP which is very high compared to the number of people drawing pension benefits from the system. Average pension expenditure per pensioner in CNPS amounted to 1.3 times GDP per capita whereas average pension expenditure per pensioner in CGRAE exceeded 2 times GDP per capita in the country. It is difficult to properly assess the adequacy and generosity of pension benefits paid by each of the schemes without having access to individual wages and distributions of pension amounts, however, this rough measure raises important questions regarding the level of the benefits paid by the public-sector pension scheme as it signals a potential inequity between the two systems. If public sector wages are higher than those paid in the private sector, then it is intuitive that public sector pensions would also be higher—since more is being paid in contributions. If public sector workers are paying more into the pension system (because of higher wages and contribution rates), then it is expected that they also receive higher pensions in return. It’s also the case that since private sector scheme only started in 2000, it’s still immature, with the lat- est retirees having only about 17 years of contribution maximum, whereas in civil servants, they could have 30–40 years, so naturally benefits after 30–40 years are going to be higher than after only 17, even if the schemes were identical. As such, it is difficult to fully assess the equity of pension benefits between the two schemes despite some obvious differences with respect to key system parameters and performance indicators. 36 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 14  Estimated pension expenditures and membership, 2016–2018 180,000 120,000 Frais de pension (millions de la FCFA) 160,000 100,000 140,000 Nombre de bénéficiaires 120,000 80,000 100,000 60,000 80,000 60,000 40,000 40,000 20,000 20,000 — — 2010 2011 2012 2013 2014 2015 Benefits CNPS Benefits CGRAE Beneficiaries CNPS Beneficiaries CGRAE Source: World Bank staff compilation based on data provided by CNPS and CGRAE. BOX 1  Recent trends in pension reform dialogue in Côte d’Ivoire A pensions reform vision has been articulated by the Government as of 2016 for expanding coverage to vulnerable workers, building on initial measures taken in 2012 to support sustain- ability of the two main funds. Prior to 2012, the CNPS had been running a structural deficit of US$630 million due to the high statutory payouts in the face of limited revenues. Since the implementation of largely parametric reform measures in 2012 (targeting retirement age and contribution rates), the CNPS is no longer running a deficit as of 2014 and has experienced a boost in its reserves. As of 2016, the Government has begun preparing a strategy to design and implement two additional measures, one targeting coverage of selected informal sector workers, and a second aiming to institute a complementary pension fund for public sector workers to improve benefit levels. Both reforms envisage a gradual roll-out starting in 2019– 2020, supported by a recently-institute e-declaration platform to facilitate firms’ contribution collection. Over the long term, the Government aims to capitalize on these measures to bridge the gap in coverage and improve coordination between non-contributory and contributory benefits. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 37 Conclusions Overall, the pension sector in Côte d’Ivoire is currently not accessible by the overwhelming majority of the population. It is characterized by low coverage and relatively high generosity of benefits; pension expenditures are high relative to the number of beneficiaries drawing pen- sions. The system is fragmented—two separate pension schemes provide potentially inequi- table benefits to public and formal private sector workers which could impede smooth labor mobility between the public and private sectors. Both the public and private sector pension schemes can benefit from modernization of system parameters to more closely mirror inter- national best practice. Moving forward, both systems would benefit from a more dynamic and multi-dimensional actuarial valuation and economic simulation to project their financing needs, adequacy of benefits, equity and welfare in terms of poverty and labor outcomes over the medium and long terms. 38 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 5 PRODUCTIVE SOCIAL SAFETY NETS AND SOCIAL ASSISTANCE As welfare-enhancing programs that target the poorest, social safety net programs serve as platforms for coordinating across social protection and labor systems. Welfare among poorer households has been more negatively affected by recent shocks than that of the general pop- ulation. In Côte d’Ivoire, average consumption per capita among the bottom 40 percent of the income distribution has been decreasing at an annualized rate of approximately –2.6 percent between 2002 and 2014, a rate almost three times greater than that of the general population (a rate of –0.9 percent). Furthermore, the growth incidence curve for this period indicates that the bottom decile has experienced the greatest losses to consumption. As a result, Côte d’Ivo- ire is a lagging performer in Africa with respect to the shared prosperity index.23 Modernization would address two main areas: (i) consolidating risk-protection mechanisms and financing models, and (iii) improving linkages to sustainable, productive livelihoods and jobs. This section assesses the current state of social safety nets, institutional arrangements, delivery mechanisms and the potential of productive social safety nets for improving earnings among the working poor. Review of Existing Social Safety Net Programs Non-contributory social assistance measures are limited in Côte d’Ivoire. Interventions have been generally prompted by temporal events, largely driven and funded by donors and tend to use a varied number (and often inconsistent) of administrative delivery instruments (e.g. identification and eligibility criteria and targeting methods). Furthermore, several institutions 23 See World Bank (2018). Poverty and Shared Prosperity 2018: Piecing Together the Poverty Puzzle. Washington, DC: World Bank. Figure 2.9. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 39 TABLE 13  Mapping of social assistance programs and key indicators in Côte d’Ivoire, 2016–2018 Average annual expen- ditures on Targeted Geographic Year Number of beneficiaries Provider/ Program population Coverage Launched beneficiaries (F CFA) Financier Programme Intégré School children National 2016/2017 1,086,721 1,754,072,640 Government de Pérennisation and donor des Cantines Scolaires (PIPCS) Programme National Poor National 2016 50,000 to date 6,257,330,000 Government de Filets Sociaux households (national roll and donor Productifs out underway gradually) Programme National Orphans National 2016 179,425 42,730,000 Government de prise en charge children due to (14%) and des Orphelins et HIV/AIDS and donor autres Enfants their families rendus Vulnérables du fait du VIH/SIDA (PNOEV) Programme du Multi-sectoral National 2016 Non-specified Planned Government Plan National program tar- 50,000,000 (15%) and Multisectoriel pour la geting children per year donor Nutrition 2016–2020 and poor households Campagne “zéro School girls National 2014/2015 Cases of preg- grossesse à l’école” nancy decreased from 5,076 in 2012/2013 to 3,871 Projet d’Appui People with National 2016 75 project/ 22,000,000 Government et d’Accès des disabilities individual (100%) Personnes Handicapées à la Formation Professionnelle Système d’Obser- Children (5–17 National 2014/2015 1,911 children 205,393,700 Government vation et de Suivi year olds) identified (18.3%) and du Travail des in labor and donor Enfants en Côte identified d’Ivoire-SOSTECI communities Plan d’Actions Children National 2014/2015 Various services 8,782,188,065 Government National (PAN) (5–17 year (hotline, school (37.23%) and 2012–2014 de lutte old) in labor kits, teachers’ donor contre la traite, and identified housing, class- l’exploitation et le communities rooms, etc.) travail des enfants Network of Public Households National n/a n/a TBC Government Social Centers and Social Workers Source: World Bank staff compilation based on government data and administrative information. 40 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 14  Aims of key social assistance programs in Côte d’Ivoire, 2016–2018 Program Detailed Aims Year Initiated End Date Integrated Program for Promote school enrollment and retention, providing meals to 1998 Continuous Sustainability of School students and assisting village communities to progressively take Canteens (pipcs) ownership of the functioning of local school canteens, both in terms of supply and management, making it a vector of local development. Program introduced following gradual withdrawal of WFP in 1996. Productive Social Support income and livelihoods through cash transfers to poor 2015 2020 Safety Net Program households and develop operational system for functioning (Phase 1) social safety net and economic inclusion system linked to employment. Specifically, the project will support the establishment of key operational tools to be used to provide regular and predictable cash transfers to poor households, economic inclusion platform and associated accompanying measures on human capital. The underlying system include an objective targeting system with a social household registry, a reliable payment mechanism, and a sound monitoring and evaluation system. Support National Ensure the social care of orphans and other children made 2003 Continuous Program for Orphans vulnerable by HIV / AIDS and their families. Includes coordina- and Other Children tion, training and mobilization of actors to deliver seven core made Vulnerable by services, including monitoring and database management. HIV / AIDS (PNOEV) National Multisectoral Improve the nutritional status of the population in terms 2016 2020 2016–2020 Nutrition of reducing of stunting among children under five, global Program acute malnutrition, underweight children, low birth weight, and anemia. Measures include awareness, monitoring and education through health and education sectors, including improving coverage of school canteens; providing access to universal health coverage through targeted subsidies (targeting 80 percent coverage); agricultural sector measures to improve yield; distribution and quality of food; water sector measures to improve access to clean water; and improved monitoring through social services to vulnerable populations (targeting 70 percent coverage). National Zero School Improve education and health outcomes through strengthening 2013/2014 2015/2016 Pregnancy Program health education. sexual and reproductive (SRH) within the education system. Specifically, reduce teenage pregnancy (tar- geting 50 percent reduction); expand awareness and education of SRH through increased student involvement; increase avail- ability and coverage of community-level services and outlets for information, monitoring and awareness. Vocational Training Promote access for persons with disabilities to technical and 2014 2024 Support for Persons vocational training. Measures include support to vocational with Disabilities Project training modules; support to self-employment training for micro and small enterprise creation; support to labor market integration. Observatory and Support to implementation of ILO Convention 182 on the Worst 2013 2015 Tracking System for Forms of Child Labor ratified. Includes technical assistance Child Labor in Côte and mobilization of local actors and community to monitor and d’Ivoire (SOSTECI) report on employment among children aged 5 to 17. Linked to National Action Plan Against Child Trafficking, Exploitation and Labor (2012–2014). Source: World Bank staff compilation based on government data and administrative information. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 41 dealing with social service delivery such as health and education use a different set of instru- ments to target the poor. This includes subsidized healthcare services and school feeding and education bursary, etc. With the exception of two emergency Bank funded labor-intensive public work programs and a diminishing World Food program (WFP) school feeding, there are no well-designed nationally known and functional non-contributory safety net programs per se. The Government runs a social assistance program, called “les Secours Sociaux”. Started in 1967, the program provides needy persons a one-time financial assistance. Procedures are highly centralized in Abidjan and the program covers a very limited number of persons (only 220 persons in 2011). The paucity of social risk-mitigating mechanisms in Côte d’Ivoire leaves most poor house- holds highly vulnerable to shocks. Though a new social safety net system is being rolled out, Côte d’Ivoire has historically had narrow social protection coverage primarily provided through contributory social insurance schemes (pensions and health insurance) for mainly upper income categories, covering less than 10 percent of the population. Poorer Ivorians rely on informal safety nets, such as “mutuelles”, charitable associations, private remittances and community networks, an approach that tends to collapse in times of covariant shocks. As a consequence, existing mechanisms have had a marginal impact on poverty reduction, do not substantively support shared prosperity, and can hardly be counted on in case of crises. Reform Agenda Underway Côte d’Ivoire recently began implementing a new social safety net reform agenda, seeking to consolidate limited, vertical programs. A new vision for social promotion and protection has been enshrined in its 2014 National Social Protection Strategy, which is being implemented under the oversight of the National Social Protection Platform. The new system focuses on a national cash transfer program, supported by key building blocks such as a household registry, a reliable and efficient payment system, and services to boost human capital, household earnings and employment. Such a robust, productive social safety net system aims to channel resources to the poorest households while providing pathways to greater earnings. International experience has demonstrated that targeted, well designed social safety net systems can mitigate poverty, reduce inequality, encourage investments in human capital and improve livelihood diversification and earnings. Cash transfer programs in countries as diverse as Brazil, Ethiopia, and Bangladesh have demonstrated positive impacts on poverty and human capital in the short-term, while also mitigating the intergenerational transmission of poverty.24 National social safety net systems can help proactively manage shocks in the short- to long-term, particularly in fragile contexts. Existing regional disparities in poverty have been exacerbated by a protracted period of successive crises in Côte d’Ivoire over 1999–2011. While growth-promoting measures are vital to reduce poverty, additional equity-promoting redistributive mechanisms are also necessary. Instituting a national system for managing and 24 For a review of evidence, see: Fiszbein and Schady (2009). Conditional Cash Transfers; Reducing Present and Future Poverty. Washington DC: World Bank, and IEG (2011). Evidence and Lessons Learned from Impact Evaluation of Social Safety Nets. Washington DC: World Bank. 42 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH coordinating basic public services and non-contributory cash to the poorest households will help to accelerate poverty reduction and promote stability over the long-term in fragile-prone contexts. In addition, enhancing household earnings and linkages to broader markets can be achieved through economic inclusion, livelihoods and jobs mechanisms in a more systematic way. Economic inclusion programs include: (i) supply-side measures that seek to enhance human and financial capital related to employment, such as technical knowledge, awareness and access to funding, (ii) demand-side interventions that seek to stimulate market inclusion and (iii) the combination of the two approaches. Within each of these approaches lies different avenues depending on local needs, context and broader growth objectives. When designed well and targeted effectively, economic inclusion packages further boost resilience and risk protection among the poorest households. In Ethiopia, public works beneficiaries who were offered complementary interventions were more likely to be food secure, to borrow for pro- ductive purposes, use improved agricultural technologies, and operate non-farm own business activities.25 In rural Nicaragua, cash transfer beneficiaries who received additional business grants or vocational training were much better protected against shocks 18 months after the end of the program.26 In Bangladesh, a program that targets the ultra-poor by providing a basic stipend complemented with asset transfers and training was effective in inducing entry into self-employment in Bangladesh.27 Given the high prevalence of employment in low-produc- tivity occupations in Côte d’Ivoire, reducing poverty will also hinge on the ability of improving rural households’ ability to improve their productivity. A previous Labor Force Training Support Project in Côte d’Ivoire (1994–2002),28 which provided training and skills development for 100,000 informal sector workers, showed revenue significantly increased among beneficiaries in the agricultural sector, including among women. Generosity Côte d’Ivoire’s new social safety net program has been designed with these reform parame- ters in mind, including cash transfers, delivered to beneficiaries in rural regions since 2016, and support to economic inclusion, whose implementation will begin subsequently. In terms of cash transfers delivered to the initial cohort of beneficiaries, the program is generous by international standards. Beneficiaries receive a monthly household cash transfer of 12,000 FCFA (delivered in quarterly transfers), which represents 9 percent of the national poverty line, or 15 percent of the extreme poverty line, based on analysis using the 2008–14 household survey at the time of the program’s design. This transfer size is the equivalent of approximately 25 Gilligan, D. O., J. Hoddinott, et al. (2009). “The Impact of Ethiopia’s Productive Safety Net Programme and its Linkages.” The Journal of Development Studies 45(10): 1684–1706. 26 Macours, K., P. Premand, et al. (2012). Transfers, Diversification and Household Risk Strategies: Experimental Evidence with implications for climate change adaptation. Policy Research Working Paper Series. Washington DC, The World Bank. 27 Bandiera, O., R. Burgess, et al. (2012). Can Entry-level entrepreneurship transform the economic lives of the poor? UCL, mimeo. 28 World Bank (2005). Economic impacts of professional training in the informal sector: the case of the Labor Force Training Program in Côte d’Ivoire. World Bank Policy Research Working Paper. Report No. WPS3668. Washington DC: World Bank. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 43 TABLE 15  Transfer size with respect to poverty line and extreme poverty line Estimated Estimated Average transfer size Average transfer size Generosity over Generosity over Monthly Household relative to extreme relative to poverty average extreme average poor Transfer Amount poverty line line poor households households 10,000 13% 8% 18% 12% 11,000 14% 9% 20% 13% 12,000 15% 9% 21% 15% 13,000 16% 10% 23% 16% 14,000 18% 11% 25% 17% 15,000 19% 12% 27% 18% Source: World Bank staff estimates based on Household Budget Survey (ENV), 2008–14. Most recent data at the time of analysis. 15 percent of average household consumption among poor households, or 21 percent among extreme poor households29 (Table 15). Targeting and Effectiveness The effectiveness of social safety net programs in reducing poverty, particularly cash trans- fers, depends on the population targeted. Côte d’Ivoire’s social safety net program currently targets extreme poor households. The same transfer size contributes more to boosting household expenditure the poorer the household, influencing the severity of poverty at the household level. With the same budget envelope, distributing the same transfers more broadly across the poor and extreme poor may result in greater reductions in the poverty headcount, since even those households near the threshold would receive a boost, but not necessarily influence severity (or the poverty gap).30 As a result, simulations show that the program’s potential effects on most poverty and extreme poverty indicators are maximized if the program targets the extreme poor, as opposed to the poor in general. Table 16 presents simulations showing the effects on poverty of targeting the initial cohort of 35,000 households for the first five years of the national program (2015–2020). Two scenarios are examined: (a) selecting beneficiaries among all poor households, and (b) select- ing beneficiary among extreme poor households only. Figure 15 illustrates the main results, 29 The poverty line is FCFA 241,145 per capita per annum. The average household size among poor households is 6.34 and the average consumption per capita per annum for poor households is FCFA 155,006. The extreme poverty line is FCFA 138,630 per capita per annum. The average household size among poor households is 6.83 and the average consumption per capita per annum for poor households is FCFA 98,510. 30 de Janvry, A. and E. Sadoulet (2006). “Making Conditional Cash Transfer Programs More Efficient: Designing for Maximum Effect of the Conditionality.” The World Bank Economic Review 20(1): 1–29. 44 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 16  Simulations of cash transfer program effects on poverty and extreme poverty indicators Monthly Poverty Indicators Extreme Poverty Indicators Transfer Scenario (FCFA) ΔFGT(0) ΔFGT(1) ΔFGT(2) ΔFGT(0) ΔFGT(1) ΔFGT(2) 35000 randomly 10000 0.25% 0.42% 0.56% 0.55% 0.77% 0.94% selected poor households 11000 0.27% 0.46% 0.61% 0.59% 0.84% 1.01% 12000 0.30% 0.49% 0.65% 0.67% 0.90% 1.07% 13000 0.31% 0.53% 0.70% 0.70% 0.95% 1.13% 14000 0.33% 0.57% 0.74% 0.75% 1.01% 1.19% 15000 0.35% 0.60% 0.79% 0.80% 1.06% 1.24% 35000 randomly 10000 0.00% 0.46% 0.99% 1.41% 2.05% 2.50% selected extreme poor households 11000 0.01% 0.50% 1.07% 1.59% 2.20% 2.65% 12000 0.01% 0.55% 1.17% 1.75% 2.38% 2.84% 13000 0.01% 0.59% 1.25% 1.86% 2.51% 2.97% 14000 0.01% 0.64% 1.34% 1.99% 2.66% 3.14% 15000 0.01% 0.68% 1.41% 2.12% 2.78% 3.26% Source: World Bank staff estimates based on Household Budget Survey (ENV), 2008–2014. Simulations based on 1000 replications for each scenario, for a nationwide program with perfect targeting, no behavioral response and excluding administrative cost. where the provision a monthly transfer of 12,000 FCFA to 35,000 households selected among the extreme poor leads to a 0.55 percent reduction of the poverty gap, a 1.17 percent reduc- tion in the severity of poverty, a 1.75 percent reduction in the incidence of extreme poverty, a 2.38 percent in the extreme poverty gap and a 2.84 percent reduction in the severity of extreme poverty. The only indicator for which selecting the beneficiaries among the poor compares favorably to selecting the beneficiaries among the extreme poor is the poverty headcount (FGT(0)), which decreases by 0.35 percent if beneficiaries are selected among the poor. However, in order to maximize impacts on the severity of poverty or any extreme poverty indicators, selecting beneficiaries among the extreme poor is preferable. Simulations of poverty effects of a program scale-up show differing levels of costs and effects. For instance, a program with 160,000 households could cover approximately a third of extreme poor households in rural regions for an annual budget of US$42.9 million, or 0.18 per- cent of Gross Domestic Product (GDP). Such a level of coverage would lead to an 8.7 percent decrease in extreme poverty headcount, 12 percent decrease in extreme poverty gap and 14.5 percent decrease in extreme poverty severity. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 45 FIGURE 15  Program effects on poverty indicators, by target group 3.00% 2.50% 2.00% 1.50% 1.00% 0.50% 0.00% Poverty Poverty Poverty Extreme Poverty Extreme Poverty Extreme Poverty Headcount Gap Severity Headcount Gap Severity Targeting All Poor Targeting Extreme poor Source: World Bank staff estimates based on Household Budget Survey (ENV), 2008–2014. A program covering all 507,477 extreme poor households in rural regions would cost US$136 million annually, or 0.58 percent of GDP. Such a program would reduce the extreme poverty headcount by 25.4 percent, the extreme poverty gap by 35 percent and the severity of extreme poverty by 42.2 percent. Finally, a program covering all 608,201 extreme poor households in the entire country would cost US$163 million annually (0.70 percent of GDP). This program would be expected to lead to a 30.5 percent reduction in the incidence of extreme poverty, a 41.1 percent reduction of the extreme poverty gap and 49.1 percent of the severity of extreme poverty. This level of spending is comparable to other countries’ spending on similar cash transfer programs such as in Brazil. Such an investment is expected to reduce the extreme poverty headcount by approximately 25 percent, which would not otherwise be necessarily achieved only by means of local pilot initiatives for fewer households or by public investment in infrastructure and basic services alone, which do not stimulate consumption or investment in human or physical capital. Conclusions Overall, the new productive program has been designed to institute a national, sustainable system, but the question remains how to scale-up by integrating across other SPL schemes. Vertical, donor-funded social assistance programs, such as school feeding or emergency pub- lic works projects, have been introduced in Côte d’Ivoire. However, limited, vertical programs will not achieve more ambitious poverty-reduction goals without significant public investment in scaling-up and consolidating social safety net programs at a national level. Existing regional disparities in poverty have been exacerbated by a protracted period of successive crises in 46 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Côte d’Ivoire over 1999–2011, building on two preceding decades of economic downturn, resulting in a high level of poverty that will only harm consumption and purchasing power, which will retard inclusive growth targets. Program effects on the severity of extreme poverty are maximized with targeting rural regions with the highest incidence of extreme poverty. At the same time, as growth and fiscal space improves in Côte d’Ivoire, an equitable approach would eventually target the poor with differentiated benefit levels in both urban and rural regions. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 47 48 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH PART II: POLICY OPTIONS FOR IMPROVING RISK MANAGEMENT SYSTEMS IN CÔTE D’IVOIRE Improving standards of living in Côte d’Ivoire and accelerating economic growth will require structural reforms in its risk management system. This system comprises, on one hand, social insurance and assistance programs to manage risks such as a disease, disability, death, pov- erty, or unemployment and, on the other hand, labor programs to improve jobs and earnings opportunities. The latter are needed to reduce/prevent risks and improve the capacity of individuals to self-insure. Better jobs will also drive economic growth. Indeed, given current participation and employment rates and the dynamics of the working age population, labor productivity growth will be the main determinant of GDP per capita growth.31 For average labor productivity to increase, the productivity of existing jobs needs to increase, and more workers need to move from low to higher productivity activities (not necessarily sectors). Thinking in terms of a risk management system is important to guide the allocation of limited public resources among competing programs. In particular, there are tradeoffs between social insurance/assistance programs and labor programs. Investing today in programs that improve jobs and earnings opportunities is likely to reduce to the future cost of social insurance/assis- tance programs. Governments, therefore, should not design and manage social insurance programs taking labor market outcomes as given. Since there are policies and programs that can influence these outcomes, these need to be coordinated/integrated with social insurance/ assistance policies. 31 Luc Christiansen and Patrick Premand. 2018. “Jobs Diagnostics for Cote d’Ivoire: Employment Productivity and Inclusion for Poverty Reduction.” Jobs Series. Issue No. 2. World Bank. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 49 In Côte d’Ivoire, like in many other countries, social insurance, assistance and labor pro- grams have problems in terms of design and coordination. First, there are problems specific to each type of program. Social insurance programs are unable to cover the majority of work- ers who are in informal jobs. They cover only a minority of formal sector workers and yet are accumulating large unfunded liabilities. Assistance programs are fragmented, limited in scope, and probably not sustainable. Labor programs are unlikely to be having an impact on labor market outcomes, particularly when it comes to improving the quality (productivity, earnings, working conditions) of informal jobs. Second, there are problems in terms of coordination/inte- gration. Redistributive arrangements within social insurance and social assistance programs are largely disconnected, which creates gaps in coverage, can lead to a regressive distribution of limited public resources, and potential distortions in labor markets. Labor policies are also not coordinated with social insurance and assistance policies. Thus, there are no effective initiatives to improve jobs and earnings opportunities to reduce reliance on safety nets, or to provide access to social insurance to informal workers. This part of the policy note discusses international experiences and recent innovations to improve the design of social insurance/assistance and labor programs in the context of an integrated risk management framework. The focus in on three types of initiatives: • Initiatives to expand the coverage of social insurance programs to informal workers which rely on new ICTs to ID, profile, enroll workers, collect contributions/savings, and make pay- ments, as well as improvements in financial literacy and behavioral nudges. • Initiatives to rationalize and integrate redistributive arrangements between social assis- tance and social insurance programs. • Labor programs that rely on investment subsidies to promote wage and self-employment among vulnerable workers, particularly in rural areas. This part is foremost a policy discussion and therefore it does not discuss, in detail, issues related to implementation, management and administration presented earlier in the report. This part is organized in three sections. Section 6 starts by presenting a general framework to improve the design of social insurance programs and expand its coverage to the entire population. The framework is used to identify reforms that the Government of Côte d’Ivoire will need to consider over the short and medium term in pensions, health insurance, and unem- ployment benefit systems. The section also discusses the potential fiscal impacts of these reforms and some of the new technologies and administrative systems that would be needed to make the reformed programs work. Cases studies are used to illustrate some of these innovations. Section 7 focuses on the integration of social assistance and social insurance programs. It uses the case of old-age pensions to show how an integrated system of subsi- dies (transfers) can be used to ensure that all citizens have a pension during old-age, while providing incentives to contribute. Finally, Section 8 discusses the type of interventions that the government can consider to accelerate the rate of job creation in the formal sector and improve the quality of existing informal jobs. 50 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 6 DESIGNING AND IMPLEMENTING SOCIAL INSURANCE PROGRAMS FOR ALL Workers and their families are exposed to natural and man-made risks for which they often lack insurance. The risk of illness, disability, death, and earnings losses related to the weather (natural), but also the risk of unemployment or earnings losses due to fluctuations in prices, technological change, and economic downturns (man-made). In principle, workers can insure themselves against these risks by purchasing insurance products (e.g., health insurance) or by saving (e.g., deposits in savings accounts or the accumulation of assets). In practice, given imperfections in insurance and capital markets, liquidity constraints, and psychological biases (“myopia”), they often fail to have enough insurance or savings. For example, private insurance companies might not offer insurance products to cover all risks for all people; reliable saving instruments might not exist; low income workers might not be able to save enough or afford the premiums of existing insurance programs; and, even if not poor, individuals might not have the foresight and discipline to buy insurance and/or save. In response to the problems with private arrangements, most countries, like Côte d’Ivoire, have setup social insurance programs to help manage natural and man-made risks. Many have pension systems that offer old-age, disability, and survivorship pensions; social health insurance funds; and unemployment insurance schemes or severance pay. The programs are supposed to be mandatory to reduce adverse selection and “myopia,” and have explicit or implicit redistributive arrangements to protect low income workers. For instance, pension systems often include a minimum pension guarantee for those who are not able to contribute (save) enough. Similarly, in health insurance programs workers often contribute a percentage of their salaries which is below the cost of the package of health services. Unfortunately, existing programs tend to have limited coverage, can be regressive, might reduce the demand for formal labor, and are running out of fiscal space. Most programs are financed through contributions from employers and employees; they were designed for CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 51 formal workers in medium and large firms (plus civil servants). The expectation was that, as countries develop, the majority of workers would become formal wage employees. But even in middle-high income countries like Mexico and Chile, informal wage or self-employment con- tinue to occupy more than half of the labor force. In Côte d’Ivoire, social insurance covers less than 10 percent of the labor force.32 In addition, redistributive arrangements within insurance programs tend to be regressive (focusing on middle and high-income workers) and impose taxes on labor that can reduce incentives to create and take formal jobs. This aggravates the problem of coverage and can negatively affect labor productivity growth. Finally, like in Côte d’Ivoire, many social insurance programs are financially unsustainable, in part, because legis- lated social security contributions are not linked to the expected cost of the benefits provided. Several countries have been able to expand coverage considerably by introducing non-contributory social insurance schemes. Over 100 developing countries, including four low-income countries (Mozambique, Nepal, Tanzania and Uganda), have introduced “social pensions,” which are transfers paid to individuals after a certain age.33 In Bolivia, for instance, the social pension covers 97 percent of those older than 60 at a cost of 1.3 percent of GDP. Some 24 countries have adopted non-contributory health insurance such as Seguro Popular in Mexico. They offer a basic health package to informal sector workers who, by definition, are not covered by contributory health insurance. Workers don’t have to pay the cost (premium) of the package or pay only a small fraction. In Ethiopia and Ghana, for example, these programs cover 68 and 32 percent of the population.34 National Health Services managed by Ministries of Health also exist in most countries and could be considered non-contributory schemes. Many workers and their families rely on the public hospitals and clinics which are part of the system to obtain health care, although out-of-pocket expenditures are often high. Non-contributory schemes, however, tend to have problems of their own. First, they create inequalities in access: higher income workers are in contributory systems where benefits are higher, whereas low income workers are covered by less generous non-contributory schemes. Second, non-contributory schemes can become an implicit tax on formal jobs while constrain- ing labor mobility. The schemes are designed under the assumption that there is little or no mobility between formal and informal labor markets; workers are always in formal or informal jobs. However, labor market statistics show that transitions between formal and informal jobs (wage and self-employment) are quite common.35 Thus, when workers change jobs, they also need to change systems. This imposes costs and benefits. Workers moving from the informal sector to the formal sector, for instance, can get access to higher wages and better social insurance benefits, but they also need to pay social security contributions while losing the subsidies they had while in the informal sector. For some workers net benefits can be negative and they might prefer to stay in informal jobs, particularly if they are young and care less about health insurance and pensions. Finally, non-contributory schemes can be financially unsustain- able. Providing adequate pensions, health care, and income support for all can be costly. Many countries simply do not have the fiscal space to do so, particularly as population ages and pensions and health expenditures increase. 32 See Part I of this note. 33 See World Bank’s Pensions Watch: http://www.pension-watch.net/social-pensions-database/ social-pensions-database--/ 34 See Cotlear et al., (2015). 35 See Cho et al., (2015) 52 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH General Principles to Inform Reforms Although there are still debates about the best way to design and implement social insur- ance programs and expand their coverage to the informal sector, the most recent studies on the topic point to some general principles that Côte d’Ivoire could consider:36 • Having integrated social insurance programs that treat all workers and their families equally. This implies that all workers have the same rights and obligations regardless of where they work. Informal workers can enroll in the same system as formal workers. • Defining explicitly the benefits offered to plan members and their costs. This is critical to ensure the financial sustainability of the system as coverage and units costs change over time, and to be able to identify the most efficient financing mechanism (see below). • Linking individual contributions to benefits and having explicit distributive arrangements. This is important to improve incentives to enroll and contribute, reduce labor taxes, expand coverage, and improve equity. Hence, workers (and employers when available) can contribute to finance part or all the cost of expected benefits depending on their level of income. The difference is financed by the government in a transparent way. Governments therefore know who is receiving subsidies, how much they cost, and how to finance them. • Relying on new ICTs to facilitate the identification of workers, enrollment, profiling/ means-testing, and the collection of premiums/savings. Having the right policies is a necessary but not sufficient condition to improve the performance of social insurance programs; the right administrative systems and implementation arrangements need to be in place. • Financial and non-financial incentives to promote savings and insurance. These are important to address issues related to cognitive and psychological biases. Purely voluntary systems, even those that provide enough information about the benefits of the program, are likely to fail. Integrating programs. The tendency observed in many countries has been to design specific programs for different groups of workers. The proposal here is the opposite: the same social insurance program can cover all workers regardless of what they do. It is the case that workers in different jobs often face very different working conditions including in terms, for example, of proximity to urban centers and the variability of their incomes over time. This, however, doesn’t require different social insurance programs. Where and how workers enroll or how they pay contributions might differ, but the rules of the system—benefits and obligations—can be the same for all workers. There is no need to have, for instance, separate pension systems for civil servants and private sector workers, or for wage employees, own account workers, and farmers. The implication is that countries can have a single pension system, a single health insurance fund, and a single income protection program managed by one or separate institutions. All workers are expected to enroll in these systems irrespective of where they work or whether 36 See Packard et al., (2019) CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 53 they are inactive or unemployed. The only thing that can change is how much and how often they pay. As discussed below, depending on their level of income, workers can pay more or less during a given year, with the government covering the difference. It is also possible, again depending on their level of income, that some workers and their families receive or accumu- late benefits without paying for a given period of time. Defining and costing benefits. The need to understand how much a given program costs would seem obvious but often these costs are only known ex-post; when closing books at the end of a given budget exercise. The idea, instead, is to be able to define costs ex-ante and therefore have the capacity to predict how program expenditures are likely to evolve over time, and what it the best way to finance them. To this end, it is necessary to define explicitly the benefits that the program offers and to assess the likelihood that these benefits will be used/received by different population groups over time. The relevant parameters differ by the type of insurance program: • Old-age Pensions. In the case of old-age pensions, benefits can be defined by the share of life-time37 earnings that the program offers to replace at retirement for each year of contri- bution (the accrual rate). For instance, the system might offer to replace 80 percent of life- time income for a full-career worker (40 years of contributions) who retires at age 65. This would imply an accrual rate of 2 percent per year of contributions. Workers who contribute less, say 30 years, would receive a replacement rate of only 60 percent at age 65.38 Once the accrual rate has been defined, it is possible to calculate the cost of the program; the share of earnings (the contribution rate) that plan members would need to pay to the pension fund to finance their pension. For individuals with limited savings capacity, part of this contribution could be financed by the government (see subsection on redistribution below). What is important at this stage, however, is to be able to define what the equilib- rium contribution rate should be regardless of who pays. Whether the pension system is defined-benefit/pay-as-you-go or defined-contribution/fully-funded the formula is the same: the equilibrium contribution rate should be equal to the accrual rate times discounted life-expectancy at retirement.39 Thus, in order to keep the contribution rate constant and reduce labor costs, when life expectancy increases the retirement age should increase. • Unemployment benefits. In this case, benefits can be defined by 1) the share of wages that is replaced when the plan member becomes unemployed; and 2) the duration of the 37 The concept of life-time earnings is critical. In many pensions systems, like Ivory Coast, only the last few years of salaries are used to calculate the pension. This compromises the financial sustainability of the system and reduces incentives to contribute or declare salaries in full in the early years (since they don’t count towards the final pension). In defined-benefits systems, all salaries should be included in the calculation of the pension (the same way that in defined-contribution systems all contributions count towards the pension). Salaries, however, should be revalorized/ indexed to calculate life-time income. The rate of return to revalorize salaries is the “interest rate” that the pension system can afford to pay on contributions while remaining solvent. Its value depends on the structure of assets of the pension system. In pure pay-as-you systems, the value of the sustainable rate of return is given by the growth rate of the present value of expected future contributions (see Holzmann et al., (2005). 38 For a given number of years of contributions, workers retiring before age 65 would also receive a lower pension because they are expected to receive this pension for longer. Similarly, workers who retire after age 65 would receive a higher pension 39 Discounted life expectancy at retirement gives the average length of time retirees would be receiving their pension. Some will do it for longer and others for a shorter period (there would be some redistribution from healthy to sick individuals) but what matters here is the average (see footnote 3). 54 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH benefit. For instance, the program might offer to replace 70 percent of the salary and pay benefits for a period of 3 months. Again, once benefits have been defined, it is possible to calculate the cost of the program; the share of earnings that individuals (or the gov- ernment) would need to pay to finance the system. Regardless of whether the system is based on risk-pooling (unemployment insurance) or savings (unemployment individuals savings accounts), this equilibrium contribution rate is the same. It is equal to the replace- ment rate times the ratio of the average unemployment and employment rates of plan members.40 In practice, individuals face different risks and different durations of unem- ployment. This implies that within a pure risk-pooling system, low-risk individuals would be subsidizing high-risk individuals. Below we discuss options to eliminate this implicit redis- tribution. For now, as in the case of pensions, we simply emphasize the need to be able to calculate the cost of the program. • Health insurance. Defining benefits and estimating costs in the case of health insurance programs is a more complicated undertaking but it is also important to do it. The first step is to define the content of the different health services that the health plan provides, for instance, reproductive health and child care; outpatient and inpatient care for severe conditions; general surgery; drug supply; laboratory services and X-ray services. Given the unit costs of these services and utilization rates (how often plan members use the services during a given period of time) it is possible to calculate the expected cost of the plan for different population groups. In this case, individuals (or the government) would not be paying a share of their earnings but an insurance premium that is equal to the expected cost of the plan. Here again, if the same average premium is imposed on all plan members younger and healthier individuals would be subsidizing older and seeker plan members. Below we discuss options to address this. Linking workers contributions to benefits and having explicit redistributive arrangements. If all plan members face the same distribution of risks and we know the expected cost of the benefits the insurance program offers, it is possible to set a single contribution rate or insur- ance premium. The contributions plan members make and the premiums they pay would be linked to the benefits they receive. It would also be possible to define a criterion (e.g., house- hold income) to decide what share of the contribution rate or premium is paid by the individual (and/or employer when available) and which share is financed by the government. Therefore, the system can still involve subsidies, but these subsidies would be explicit. Individuals would know, ex-ante, how much they are receiving in subsidies, and the government would know, ex-ante, who is receiving the subsidies, how much they cost, and how to finance them. The problem with this proposal is that the population of plan members of pensions, unemploy- ment benefits, and health insurance plans is quite heterogenous. Different individuals have different life expectancies, different unemployment risks, and a different predisposition to var- ious health shocks. The policy question then is how narrow and precise should the calculation of contributions rates/premiums for different plan members be? In principle, the approach can vary by type of program. 40 Robalino David and Michael Weber 2013. “Designing and implementing unemployment benefit systems in middle and low-income countries: beyond risk-pooling vs savings.” IZA Journal of Labor Policy20132:12 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 55 • Pensions. In the case of pensions, it might be unavoidable to accept a certain level of implicit redistribution and set a single contribution rate despite differences in life expec- tancy. Estimating differences in life expectancy would be difficult. A single contribution rate would imply that men subsidize somewhat women who, on average, live longer. Also, probably, low-income workers would be subsidizing higher income workers if the latter have a longer life expectancy. But at the same time, low income workers are more likely to receive government subsidies in the first place. • Unemployment benefits. It is also difficult, and probably not practical, to estimate ex-ante the unemployment risks of different individuals. These risks depend not only on their individual characteristics but also the sectors/firms where they work. The alternative is to simply keep track of negative or positive balances in individual “unemployment accounts.” Given the equilibrium contribution rate to the system, high-risk individuals would be accu- mulating negative balances (up to a limit),41 while low-risk individuals accumulate positive balances (savings). The government would then subsidize the negative accounts and finance these subsidies from alternative sources. Two important sources to consider are: 1) dismissal taxes that would replace standard severance pay; and 2) taxes on accounts with positive balances. Essentially, the government would be offering plan members the following: “if you become unemployed you are allowed to receive a given percentage of your covered earnings42 for a given period of time (even if your account doesn’t have a positive balance), but if you contribute to the system and don’t use unemployment bene- fits you can keep a given fraction of the balance in your account.” Countries like Chile and Brazil have tried to merge savings and risk-pooling in unemployment benefit schemes. The option discussed here does it in a more transparent way, while reducing the risks facing workers, and by expanding the options the government has to finance the program. • Health insurance. In this case, governments can consider differentiating premiums by age groups (e.g., children, adults, and the elderly) and gender. In fact, most actuarial valua- tions of public health insurance funds project costs taking into account different utilization rates (and sometimes unit costs) by age and gender. Thus, household heads who enroll themselves and their family in the health insurance plan would pay a total premium that depends on the size and demographic structure of the household. Families who are not able to afford the premium can receive government subsidies. This proposal doesn’t deal with differences in the distribution of pre-existing conditions or the prevalence of risky behaviors such as smoking o consuming alcohol. While it would be possible to introduce corrections to take into account risky behaviors, it seems desirable (and appropriate) to average the cost of pre-existing and catastrophic conditions within the cost of the age/ gender group. 41 Having a limit is necessary to controls abuse and preserve the financial sustainability of the system. In regular unemployment insurance schemes, there are implicit limits given by the combination of a maximum duration of benefits and a minimum number of months of contributions. Individuals who reach the limit would need a backup system to guarantee a minimum level of income (see Section VIII). It would also be possible to adjust over time the contribution rates of higher-risk individuals. Individuals accounts would provide information to estimate these equilibrium contribution rates. They are equal to the replacement rate, times the risk of unemployment, times the duration of unemployment (or benefits). 42 Covering earnings versus salaries is important. It implies that own-account workers could also participate in the system. 56 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Identifying, enrolling and profiling Individuals. If the right policies are in place—integrated programs, clearly defined benefits and costs, explicit redistributive arrangements—enrolling all workers and their families in social insurance programs is a relatively simple task. Part of the processes involved are purely administrative and, in principle, straightforward thanks to new information and communication technologies: identifying plans members and their depen- dents; registering them in the new system; determining the type of coverage they are eligible for based on individual and household data; and issuing an insurance card. • Locating and Identifying the uninsured. A key priority for any administration should be to rapidly locate and identify all the uninsured population. In all cases, having a proper ID (National ID or birth certificate) is necessary. This is a problem in many countries, including the United States. • Registering/enrolling beneficiaries. Registration can take place in specific locations where individuals go to enroll voluntarily (including schools, churches, or mobile units), or at the household level. International experiences show that proper communication campaigns at the local level can help increase the number of individuals who register voluntarily. Nonetheless, voluntary registration is usually not sufficient. A census-type of verification might be needed to ensure that the entire population is enrolled in the insurance sys- tem. This, however, cannot be done regularly. Thus, various countries are experimenting with conditionalities. For instance, in order to obtain drivers licenses or passports, mobile phones, or to receive health services or social assistance payments, individuals must show proof of enrollment in the social security. Part of the registration process involves obtaining from individuals and dependents the necessary information to inform eligibility for different benefits (see below). The registra- tion process should be able to take place on-line and off-line. When registration is done off-line, the data collected should be uploaded daily to a central system. The information about beneficiaries and their families should, ideally, include biometric data which is stored in the social insurance card. This card should be issued at the point of registration, subject to ex-post audits that take place at the household level. • Proxy Means Tests (PMTs). Côte d’Ivoire is already applying PMTs to select beneficiaries of the Productive Safety Net Program. These are, essentially, statistical models that use read- ily available information about the individual and the household (collected at the time of registration/enrollment) to predict income or consumption levels. Many countries are using them to identify the beneficiaries of safety-net programs such as conditional or uncondi- tional cash-transfers. Similar models can be used to determine eligibility for different levels of subsidy within social insurance programs. There have been concerns about the accuracy of PMTs, as there is evidence that the percentage of eligible individuals who are erroneously excluded from the programs (exclu- sion errors) is high.43 The performance of PMTs, however, improves in the middle of the income distribution. Thus, one way to address this concern, is to claw-back subsidies only for households above the 4th or 5th decile of the income distribution. This ensures that 43 See Ravallion (2018). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 57 all other households, where the likelihood of miss-classification is higher, would be fully subsidize. Collecting Payments. There is an ongoing policy debate about whether is possible to collect meaningful contribution amounts from non-poor informal workers. Some studies suggest that collection costs can be higher than the revenue from contributions. At the same time, many informal workers are already paying for health services out-of-pocket. In the Middle East and North Africa region it is estimated that these payments represent, on average, 6 percent of total household expenditures and can be as high as 10 percent. Thus, poverty rates can increase by up to 20 percentage points when health expenditures are taken into account.44 There are also positive experiences, such as that of Korea, where premiums from non-wage employees now represent over 30 percent of revenues. Recent studies suggest that, in general, informal workers are willing to pre-fund health insurance. There are also emerging best practices for the design and implementation of efficient collection systems that should be taken into account. • Information. Ensuring that citizens have adequate information about the importance of having health insurance coverage, the functioning of the system, and the benefits and costs to plan members. • Affordability. Having premiums that are affordable relative to the level of income of the plan member. • Low transaction costs. Having in place a collection system that is simple, secure, and fast. New communications and information technologies can have an important role to play. Individuals should be able to make payments on-line or through their mobile phones. More innovative options to consider involve bundling health insurance premiums or contribu- tions for pensions with mobile phone cards. For instance, the price of a minute of voice or a mega-byte of data can be topped up by a given amount that goes to pay the premium owed by the plan member. In practice, the monthly premium can have an equivalent in terms of number of minutes of voice and/or megabytes. • Financial and non-financial “nudges.” There is a growing literature from behavioral eco- nomics about how to “nudge” changes in behaviors. For instance, reminders through text messages can encourage informal workers to enroll in voluntary pension plans.45 The right types of nudges are context specific and can be designed through experimentation and rigorous evaluations. • Accountability. It is key that plan members trust the system, and this can only happen if policymakers, managers, and providers are accountable to them. There are three levels of accountability: i) providing regular information about the performance of the health insur- ance system and individual providers; ii) having a system to report and address grievances on a timely manner; and iii) guaranteeing minimum standards in terms of the quality of health services. 44 See Elgazzar (2009). 45 See Akbas et al., (2016). 58 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Implications for Côte d’Ivoire Social insurance programs in Côte d’Ivoire (pensions, health insurance, and unemployment benefits) do not meet most of the design principles discussed above. As a consequence, their financial sustainability is at risk, they cannot be reliably expanded to the informal sec- tor, and they are likely contributing to a regressive distribution of income and lower levels of formal employment. In this section we discuss the types of reforms that the government could consider over the medium-term. The goal is to define a vision for the future that can guide policy initiatives over the short term. Because existing systems are small, one possibility to consider is starting reforms from the “bottom-up.”46 Essentially, Côte d’Ivoire could design and implement new social insurance arrangements for the poor and informal sector workers that can become the platform for a national integrated system over the medium-term. Something along these lines is already hap- pening in the case of health insurance through the new program for informal sector workers and the poor (Programme de Couverture Maladie Universelle, CMU). The new program, how- ever, still has design problems that would need to be addressed before coverage expands. Pensions The scheme for private sector workers (CNPS) and public servants (CGRAE)—both defined-benefit/pay-as-you-go—have common problems with benefit formulas and financ- ing mechanisms. Because the CNPS is small and still young, there are still no issues with payments, but the system is likely to be accumulating unfunded liabilities. The CGRAE, on the other hand, is a more mature scheme that already costs 0.6 percent of GDP and is exerting pressure on the budget. Going forward, both systems should take the following recommenda- tions into consideration: • Pension formula. All salaries should be included in the calculation of the pension, each revalorized by the sustainable rate of return of the system. This rate should be the result of an actuarial assessment. It could be close to 3 percent real per year, which has been the growth rate of the labor force and labor productivity. As an example, using this rate, the salary received 20 years before the time of retirement, should be multiplied by a factor of 1.8 (1.03^20), the salary received 10 years before retirement by a factor of 1.34 (1.03^30). Today the CNPS includes the last 15 years of salaries in the calculation of the pension and the CGRAE only 5. The salaries included are not indexed/revalorized. This reduces, instead of increasing the level of the pension. 46 See Cotlear et al., (2015). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 59 • Contribution rate. As discussed above, the contribution rate should be automatically linked to the accrual rate and life expectancy at the retirement age.47 If life expectancy increases, the retirement age should increase automatically to avoid increasing the con- tribution rate. In both schemes this “actuarial link” does not exit. Given an accrual rate of 1.7%, life expectancy at age 60 in Côte d’Ivoire (~16 years) and the resulting G factor (9.97), the equilibrium contribution rate in CNPS should be around 17 percent (17% = 1.75%*9.97) for workers who do not reach the 50 percent ceiling (less than 30 years of contributions), higher than the current 14 percent. In the CGRAE the contribution rate should be close to the actual level but there is no guarantee that this will continue given that there are no automatic adjustment mechanisms. • Early and delayed retirement. Individuals should be allowed to retire before or after the legislated age with actuarially fair adjustments in benefits. The accrual rate used to calculate the pension should be equal to the contribution rate divided by discounted life expectancy at retirement. Thus, individuals retiring early receive a lower accrual rate because they are expected to receive their pension for longer. Similarly, individuals who delay retirement should receive a higher accrual rate because they expect to receive the pension for a shorter period of time. Today, early retirement penalties in the CNPS and CGRAE are ad-hoc (5 percent per year). They become very low for people retiring before age 58, and there are also no adjustments for delay retirement (see Figure 16), • Indexation of pensions. Whether pensions are indexed with inflation, wages, or a com- bination is at the end a policy choice, but the type of indexation should be reflected in the calculation of the pension. The formulas discussed above assume that pensions are automatically indexed with inflation, which is the case of the CNPS. If pensions were to be indexed by wages like in the case of the CGREA, contribution rates should be higher. The general recommendation is to index pensions with inflation to control the cost of the system. • Ceiling on replacement rates. There should not be ceilings on the replacement rate because these can create negative incentives to contribute. Indeed, once the ceiling is reached, additional contributions do not increase the pension. The CNPS enforces a ceiling of 50 percent, presumably to control costs. Thus, a person contributing for 40 years would receive an effective accrual rate of only 1.25% (50%/40) that would require a contribution rate of only 12.5% (12.5%=1.25%*9.97). The system is essentially imposing an implicit tax on individuals who contribute more than 30 years (the time at which the ceiling is reached) that is used to subsidize those who contribute less. The recommendation is to eliminate the ceiling and set an affordable accrual rate instead. If the accrual rate becomes 47 The equilibrium contribution of the DB-PAYG pension system, which equates the present value of pensions paid with the present value of contributions is given by: β = αG, (3) Where α is the accrual rate (1.75%) and G is the annuity factor or “G factor,” essentially discounted life expectancy at retirement: L G= Σ a=R S (R, a) (1 + r ) a – R S(R,a) is the probability of surviving to age an at age R and r is the sustainable rate of return of the system. The G factor at age 60 in Côte d’Ivoire is estimated at 9.97067102 with an real interest rate of 3% (r = 0.03). 60 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 16  Actuarially fair adjustments to the pension as a function of the retirement age 50% 40% 30% Reduction in Pension 20% 10% 0% –10% –20% –30% 50 55 60 65 70 Retirement Age Source: World Bank staff calculations. too low for individuals with short careers—a very common problem in developing coun- tries—other explicit redistributive arrangements can be considered (see below). • Ceiling on covered earnings. It is good practice to impose a ceiling on coverage earn- ings that can be around 2.5 times economy wide average earnings. The idea is to allow higher income individuals to diversity sources of savings for retirement. They contribute to the pension system and their pension is calculated on only part of the salaries. Thus, they replace part of their earnings through the public pension system and part through private arrangements. • Minimum pensions. It is important to rethink the design of the minimum pension guaran- tees to improve incentives to contribute. Instead of guaranteeing a top-up of benefits to reach a minimum pension, the government should top-up contribution rates for low income workers in order to offer higher accrual rates (see discussion on redistribute arrangements below). The CNPS today offers an ad-hoc minimum pension guarantee equal to 30,000 Franc CFA/mo (USD 52/mo). Individuals with earnings equal or below the average—around 113,654 Franc CFA (USD 197)—have weak incentives to contribute. Indeed, whether they contribute 15 years or 30 years they would still get the minimum pension (see Figure 17). The system also does not protect workers who are not able to contribute for 15 years and therefore are not eligible for any pension. The scheme for civil servants has more structural problems given its nature and it is import- ant to consider the question of whether it should be closed to new entrants. The system does not really operate as a pension fund that accumulates reserves from contributions or pays pension from contributions. In practice, the CGREA is an accounting scheme: the salaries paid by the government are reduced by the amount of employees’ contributions and, when CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 61 FIGURE 17  Incentives to contribute under the existing pension guarantee 600 500 Value of the Pensions (USD) 400 300 200 100 0 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Number of Years of Contributions Source: World Bank staff calculations. workers retire, the government replaces the salary by a pension that is still financed from general revenues. The relevant policy question in this type of scheme is whether the wage/ pension bill is affordable. If not, interventions are needed to control the growth of the wage bill in the civil service and/or the level of pensions. The latter can be done by reducing the accrual rate. Going forward, an option to consider is closing the scheme to new entrants, mandating new civil servants to enroll in a, reformed, national pension system that would be managed by the CNPS. As discussed in the previous section, there is no rationale for having a separate pension fund for public sector workers. Towards a Universal Pension System A reformed pension system managed by the CNPS could become the scheme for all work- ers, including informal sector workers. This would imply redefining the mandate of the system and modifying benefits formulas and eligibility conditions along the lines discussed above. Like in the case of other countries that have reformed their pension system, it is possible to honor accrued rights to date for all plan member and grandfather current benefits and eligi- bility conditions to those who are relatively close to retirement. Hence, the new system would apply to all workers who enter the labor market after a given date and to existing plan mem- bers who were not grandfathered after that date. The latter would keep the value of pension wealth accumulated prior to the reform. As an illustration, the new pension system could have the following characteristics: 62 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH • It would target a replacement rate of 50 percent for full career workers (40 years) who retire at age 62, while allowing individuals to save more if they decide to. This implies setting a baseline accrual rate equal to 1.25% and a contribution rate of 11.25% (11.25% = 1.25%*9.97). • The system would aim to guarantee a minimum pension equivalent to 15% of average earnings. To this end, it would offer higher accrual rates and subsidized contributions to workers in the lowest deciles of the income distribution. These subsidies would depend on average incomes and expected contribution densities within the decile. Simulations using a prototype low-income country illustrate how the value of the pension resulting from targeted subsidies (TSI) and from individual contributions would change across income deciles. Low income workers, often informal sector workers, would be able to contribute/save little on their own, thus the value of the pension generated by their contribu- tions would be small (gray area in top panels of Figure 18). For instance, in the first decile, the value of the pension would be fully subsidized. As income increase, the value of the subsi- dized pension declines and the value of the contributory pension increases. Workers in deciles six and above would no longer receive subsidies. The fact that the value of the subsidies declines gradually with income reduces negative incentives on labor supply. The value of the subsidy within each decile is the difference between the contribution rate paid by workers and the equilibrium contribution rate. In the first decile, for instance, the equilibrium contribution rate is close to 40 percent of earnings. It is high because, in the example, workers in the first decile only contribute, on average, 50 percent of their active life and their incomes are equal to 17 percent of average earnings. To finance a pension of 15% of average earnings (the minimum) with only 20 years of contributions they would need to save around 40 percent of their incomes. If these workers do not have the capacity to contribute or contribute very little, most of their pension would come from explicit government subsidies. As incomes increase, equilibrium contribution rates decline, and workers can contribute more and for longer. In the example, workers in the 6th decile and above pay the equilibrium contribu- tion rate (see bottom-right panel in Figure 18). Under the assumption of universal coverage, the fiscal cost of the system would be in the order of half a percentage point of GDP per year or the equivalent of a 0.6% tax on consumption. This seems to be a manageable level of expenditures to guarantee adequate pensions for all workers. Expenditures are not higher because subsidies are, essentially, tar- geted to the first four deciles of the income distribution. In addition, costs are likely to decline as earnings and contribution densities increase. As a reference, total expenditures in the existing pension system for civil servants and private sector workers are each in the order of 0.6 percent of GDP with very limited coverage. As designed, the pension system would offer strong incentives to enroll and contribute. Low income workers and the poor would know that in order to receive a pension upon retirement, they would need to register and declare activity/earnings regularly, even if they do not pay contributions. Those who can contribute would know that the government is, essentially, matching their contributions. Higher income workers would be receiving a competitive rate CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 63 FIGURE 18  Illustrative welfare and fiscal impacts of a universal pension system Model assumptions (based on data from Sub-Saharan Africa) Decile Labor Force Participation Earnings (Share of Average) Contribution Density 1 0.330 0.174 0.5 2 0.359 0.266 0.5 3 0.373 0.336 0.5 4 0.388 0.407 0.5 5 0.398 0.489 0.6 6 0.427 0.590 0.7 7 0.449 0.730 0.8 8 0.476 0.944 0.9 9 0.490 1.360 0.9 10 0.567 4.705 1 Note: Consumption/GDP per capita = 0.81 Source: World Bank staff calculations. 64 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH of return on their contributions (3 percent real in the example) and the value of their pension would increase with the amount of their contributions. Still, as discussed in the previous section, the viability of the proposed pension plan depends on the ability of the government to profile/means test workers and collect contributions savings. An effective/agile PMT would need to be in place that can be applied regularly (e.g., every two or three years). Workers should be able to make contributions using their mobile phones. There should be mechanisms to nudge savings that could include: 1) reg- ular text-messages; 2) automatic deposits in the pension account when phone cards are purchased; 3) matching contributions which are deposited ex-ante in pension accounts and debited if individuals do not pay their own contributions after a given period of time. The most efficient combination of these nudges can be identified through experimentation and careful evaluation. Health Insurance Like in the case of pensions, the coverage of social health insurance (SHI) in Côte d’Ivoire is limited to formal sector workers. Thus, only around 5% of the total population is covered by SHI (essentially private and public sector workers and their families) through one of the two national pension funds and a network of designated service providers. The national premium is set at 1,000 F CFA per person per month, split 50%-50% employer/employee, and paid auto- matically through the payroll. There is a 30 percent co-payment and a schedule of exemptions. Revenues mobilized through this risk-pooling scheme reach around 21 percent of total health expenditures, which represent only 1.2 percent of GDP. Expenditures by the government and donors account for 22 and 26 percent respectively. Out-of-pocket expenditures account for 36 percent of the total. In 2015 the government started a national project to create a stand-alone health insur- ance fund (CNAM) for the poor and informal sector workers. In the scheme the poor are fully subsidized and informal sector workers (who enroll voluntarily) are expected to receive some, limited, subsidies. The Programme de Couverture Maladie Universelle (CMU), as it is called, has started targeting and enrolling the poorest households in coordination with the productive Social Safety Nets program. Targeting is conducted using the same methodology as that developed by the SSN program (see next section). As of 2019, CMU has enrolled 6% of the poor. Implementation among the informal sector has not yet started. Further studies are needed to inform decisions on design, target groups and enrollment strategy. Indeed, other countries attempting to reach universal health coverage have succeeded or failed based on issues related to system design and implementation Going forward, as the CMU expands to informal sector workers, the following issues will need to receive attention. • Level of integration. As discussed above, aiming for a more integrated system is an important policy goal. For now, the CMU is a parallel program for informal/poor workers that is expanding coverage and reducing out-of-pocket expenditures. Over the medium and long-term, it would be important to harmonize or integrate CMU with the contributory CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 65 programs for formal workers. As discussed above, not doing so can create inequalities within the health sector and distort labor markets by imposing implicit taxes on formal jobs. • Benefit package and costs. It is unclear at this stage what is the content and cost of the health package offered by existing schemes, but actual premiums seem quite low. In the contributory programs for formal workers, the per capita premium of 1,000 F CFA—USD 1.7 per month or USD 21 per year—is low by international standards. Per-capita premiums in a low-income country like Côte d’Ivoire are likely to be above USD 50 per year. To guarantee the financial sustainability of the new health insurance system it is critical to clearly define the content of the package of health services and estimate expected cost by age (children, adults, the elderly) and gender. The government can then assess how total expenditures in the program are likely to evolve and identify the best financing mechanisms. • Voluntary enrollment. Enrollment in the new health insurance fund should be mandatory for all workers even if enforcement is a challenge. At present, the CMU is effectively a voluntary scheme. The problem of having voluntary health insurance systems is adverse selection. Essentially, workers with pre-existing conditions or who are more likely to require health services are more likely to enroll. This increases the cost of the health plan. The alternative is to make health insurance mandatory for all workers and then identify mechanisms to promote and enforce enrollment (see Section VI, “General principles”). • Contracting and payment systems with providers. The government needs to acceler- ate initiatives to introduce payment systems that provide incentives to control costs and improve the quality of care. Existing studies suggest that access to health care is low and quality is poor. Most health care in urban areas is provided through public health centers, public hospitals and, increasingly, private clinics that are financed through direct out-of-pocket payments. Rural areas tend to rely on informal care givers, a limited num- ber of public health centers and roaming public health or social workers mainly providing nutrition and maternal care. For CNAM, there is a list of a pre-designated network of public service providers (health centers and hospitals) contracted through the contributory social health insurance programs to deliver primary, secondary and tertiary care. Setting in place contracts to pay these providers based on results and performance is a pre-condition for the success of the CMU. • Governance and management. The government could consider outsourcing some of the administrative functions of the CNAM to specialized management companies (third party administrators—TPAs). Indeed, managing the many administrative functions of a health insurance fund is a highly complex undertaking (see Table 17). Particularly challenging are processes such as enrolling plan members, collecting premiums, and purchasing services. Third party administrators operate at scale and can reduce administrative costs. They would be funded by an administration fee that is proportional to total health expenditures. • Information systems. In all cases, it is critical that CMU managers improve existing infor- mation systems to accurately track beneficiaries, contributions, and claims from health centers, dispensaries, clinics, and hospitals. Information submitted by providers should be linked to beneficiaries so that it becomes possible to track utilization, heath expenditures, and health outcomes at the level of the individual. 66 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 17  Core administrative functions of a Health Insurance Fund Functions Administrative Activities Resource Mobilization • Providing information • Identifying members • Registering, enrolling members • Billing and collecting contributions • Marketing Purchasing • Selecting, negotiating with health providers • Processing claims, paying providers/reimbursing patients • Gate keeping Overall Management • Customer service and education • Appeal mechanisms • Human resource management including training • Monitoring and evaluation • Accounting, auditing, reporting • Actuarial analysis, financial management, budgeting Source: World Bank staff. Towards a Universal Health Insurance System Assuming that the institutions necessary for the functioning and expansion of the new health insurance system—CMU—are in place, a key policy issue will be to define the most appro- priate financing mechanism. As previously discussed, countries have followed one of two paths: fully subsidized health insurance or targeted subsidies. At the end, the choice should be driven by the costs and benefits of targeting. Having in place the right targeting system can be a more complex undertaking that is prone to errors of exclusion. These, however, can be min- imized if in the bottom deciles of the income distribution subsidies are universal and targeting starts only above a certain level of income, close to the median. At the same time, targeting has the potential to reduce costs substantially. With a fix budget, targeting would allow plan administrators to offer a more generous package of health services. Although at this stage there are no country specific data, simulations with the same pro- totype model used for pensions suggest that targeting subsidies might be a necessary condition to be able to expand coverage. In the calculations it is assumed that the cost of the package of health services offered by the CMU is equal to USD 100 per capita per year or 6% of GDP per capita. Two financing mechanisms are considered: 1) universal subsidies; and 2) targeted subsidies. When targeting, subsidies for the first three deciles remain at 100% per- cent but decline to 50% for deciles 4, 5 and 6; deciles 7 and above do not receive subsidies (see top panels in Figure 19). With universal subsidies, the CMU would cost around 5.7 percent of GDP an amount that would probably be fiscally unsustainable. Indeed, as health costs increase and the population CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 67 FIGURE 19  Illustrative fiscal and welfare impacts of universal health insurance Source: World Bank staff calculations. ages, health insurance expenditures typically grow faster than GDP. Targeting subsidies would reduce costs to 2.6 percent of GDP but also increase substantially households’ health expen- ditures. Thus, for deciles 4, 5, 6, 7 effective premiums (premiums paid after the subsidy) could represent between 5 and 9 percent of average per capita consumption within the decile (see bottom-right panel in Figure 19). These numbers are simply illustrative of the tradeoff that the government will have to address: affordability of health insurance and fiscal sustainability. Unemployment Benefits Like the majority of countries Côte d’Ivoire does not have an unemployment insurance pro- gram but it regulates severance pay through the labor code. Workers who lose their job for economic or technical reasons receive a lump-sum benefit that is proportional to the number of years of service: 30% of the monthly salary for the first five years of service; 35% for years 6–10; and 40% for years 11 and beyond. Therefore, a worker who has been working for 5 years receives 1.5 months of salary, somebody working fifteen years 5.25 months of salary, and somebody working twenty-five years 9.25 moths of salary (see Figure 20). 68 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 20  Severance pay in Côte d’Ivoire Source: World Bank staff calculations. Severance pay is not the most effective form of protection for workers and yet it can affect the demand for labor. First, the system relies on the existence of a labor contract and there- fore it is not apt for self-employed workers. Second, employers seldom provision to pay severance benefits. Thus, employers facing financial difficulties who need to dismiss workers are also likely to be employers who do not have the necessary funds to compensate them. In addition, there are often lengthy legal processes associated with the payment of severance benefits. A study for Mexico, for instance, found that only 10 percent of workers eligible to receive severance were paid within the first years after dismissal.48 At the same time, employers trying to restructure and adopt new technologies to boost com- petitiveness can be deterred by high severance costs. In Côte d’Ivoire, young workers receive modest payments that are probably not adequate to protect them from the risk of unemploy- ment, but older workers can receive benefits that are too high and could discourage restruc- turing. By discouraging restructuring they can also hamper growth. Firms, aware of the cost of dismissing long-term workers, might also be reluctant to hire long-term.49 Absent structural reforms, Côte d’Ivoire might consider reviewing the formula used to cal- culate severance while mandating employers to create reserves or reinsure themselves. A new formula would aim to increase severance for young workers while reducing benefits for long-term workers. To this end, it is important to note that the risk of unemployment does not 48 See Kaplan, 2010 49 See Kuddo, Robalino, and Weber. 2015. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 69 increase linearly with the number of years of work; the opposite can happen. Older workers in a firm might have more difficulties finding a new job after dismissal and might require more support, but in general there is no rationale for increasing benefits in proportion to the length of service. An alternative benefit schedule could have only two levels. For example, workers who have been with the firm for 10 years of less receive 4 months of salaries, and workers with 20 years or more receive 6 months of salaries. In terms of reinsurance, one option to consider is to ask employers to pool risks. This can be done by contributing to an insurance managed by a third party. Contributions would be proportional to the wage bill that is subject to the payment of severance pay. Based on infor- mation about severance payments, unemployment rates, and the duration of unemployment, insurance companies could price the risk and bid accordingly for the management of the system. Towards a Universal Unemployment Benefits Scheme A more radical reform in Côte d’Ivoire would involve setting up a proper unemployment insurance scheme. The UB system would be designed according to the principles outlined in Section VI, “General principles”. Initially, it would not be a system that covers the entire labor force. The unemployment rate for a majority of workers in Côte d’Ivoire is actually quite low; the main problem is underemployment and not unemployment. The system would be open, however, to both formal and informal sector workers with earnings above a given threshold, whether in wage or self-employment. The common concern about workers abusing the system and claiming benefits while working informally would be addressed by conditioning payments to participation in labor programs (see Section VIII). The scheme would be managed by the same institution managing the universal pension system and/or the universal health insurance fund. Since, in principle, all workers would be enrolled, adding unemployment accounts to their existing accounts would be a relatively simple task. Once a new UB system is in place, severance pay could be replaced by a dismissal tax that is used to fund labor programs and/or redistribution within the insurance scheme. The rationale for having a dismissal tax is internalizing the social cost of unemployment.50 This would be a small tax that is applied to the salary of workers who are dismissed at a given point in time. Payments made by employers, however, would no longer go to workers. Instead, payments would accumulate in a fund that is used to finance active and passive labor programs (see next sub-section). The low-income country prototype model suggests that the fiscal costs of an unemployment benefits system could be quite modest. As an example, a system that covers the top five deciles of the income distribution and offers a replacement rate of 70 percent for a duration of six months, could cost the government around 0.12% of GDP; less if the revenues from a dismissal tax or a tax on savings in unemployment accounts is considered (see Figure 21).51 50 See Blanchard 2008. 51 The model was calibrated using an average unemployment rate across deciles of 5% and a duration of six months, which gives an unemployment risk of 0.88%. It was also assumed that the risk of unemployment decreases with income so that unemployment risks of deciles 6 to 10 are respectively: 1.75%, 1.32%, 0.88%, 0.15%, 0.29%. With these assumptions the equilibrium contribution rate of the system would be around 3.7% of wages. 70 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 21  Welfare and fiscal impacts of an unemployment benefit scheme Source: World Bank staff calculations. Because the risk of unemployment varies by income decile, equilibrium contribution rates and therefore the amount of savings accumulated in individual accounts also vary (see bottom-right panel of Figure 21). Thus, individuals in the top deciles pay more than what they should whereas those in the lowest deciles pay less. In the example, the contributions of deciles 6 and 7 are not enough to cover their benefits. Deciles 6 and 7, for instance, receive benefits equivalent to 4.2 months of salaries but self-finance only 2.1 and 2.8 respectively (see top-left panel of Figure 21). Deciles 9 and 10 contribute much more than what they receive in benefits. To top up the accounts of deciles 6 and 7, the government has several alternatives. One is to use general revenues which would cost 0.12% of GDP (0.32% if one doubles the unemployment rate). It could also tax the savings accumulated in the accounts of deciles 8 and 9 or use part of the proceedings of a dismissal tax. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 71 72 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 7 INTEGRATING SOCIAL INSURANCE AND SOCIAL ASSISTANCE PROGRAMS Although Côte d’Ivoire has several initiatives, the coverage of social assistance programs is low and mainly involves in-kind transfers. It is estimated that around 27 percent of the popu- lation has access to some form of social assistance, with larger shares observed in the bottom quintiles of the income distribution (35–36 percent). The main program is PIPCS (Programme Intégré de Pérennisation des Cantines Scolaires) which offers free meals to children in school and has around 1.1 million beneficiaries. Cash-transfers offered through the Productive Safety Nets Program reach less than 1 percent of the population. Total expenditures on safety nets therefore account for only 0.2 percent of GDP.52 The challenge, therefore, remains how to expand risk protection to more people in an efficient and equitable way. Financing an Integrated System As Côte d’Ivoire expands the coverage of cash-transfers, there is scope to gradually phase- out in-kind transfers and integrate, in a transparent way, subsidies for social insurance with anti-poverty transfers. This will allow, over time, to expand insurance coverage to all workers, improve equity, and avoid labor market distortions. As shown in recent studies, there is little rationale for having separate social assistance and social insurance programs, and/or multiple forms of redistributive arrangements.53 52 Ibid. 53 See World Bank (2019). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 73 The proposal, therefore, is to have social insurance programs connected with explicit redis- tributive arrangements with two objectives: 1) guaranteeing a minimum level of income (con- sumption) in all states of nature; and 2) subsidizing premiums/savings for those who cannot afford them (as discussed in the previous section). The redistributive part of the social pro- tection system therefore would have two elements: a basic cash/income transfer that would be targeted/tapered (TCT) like the current transfers provided by the productive safety-net programs; and subsidized insurance that, as shown in Section VI, is also likely to be tapered (TSI). The cash- component is critical to ensure that all workers, regardless of activity and type of work, are always able to support a minimum level of consumption. Subsidized insurance is necessary to ensure that everyone is covered against core risks regardless of the type employment and sector of work. Wage employees, informal own account workers, and farmers would all have access to the same system according to the same rules. The prototype low-income country model can be used to illustrate this type of integration using the case of old-age pensions. As in Section VI, the assumption is that the universal pension system offers a 50 percent replacement rate and a minimum pension equivalent to 15 percent of average consumption. In addition, there is a cash-transfer that is equivalent to 10 percent of average consumption with a 20 percent taper (i.e., for each 10 F CFA increase in income/consumption, the transfer is reduced by 2 F CFA). The actual cash-transfer of 12,000 per month per household is equivalent to 15 percent of GDP per capita or close to a per capita transfer of 5 percent of average consumption. Hence, the cash-transfer used in the simulations is around twice the existing transfer. The results show that in the presence of the cash-transfer, the role of the minimum pension guarantee would decline. In the first decile of the income distribution the cash-transfer would represent 57 percent of the pension and the pension subsidy 28.7 percent (see blue and orange areas in top-panels in Figure 22). In the second and third deciles, the cash transfer would represent 17.6 and 10 percent respectively, and the subsidized pension 10 and 9 per- cent. Contributory pensions would become the main source of income during-old age starting in the fourth decile of the income distribution. All workers (and each member of their families) are eligible to receive the cash-transfer (per capita). The cash-transfer would be phased-out, however, by the fifth decile Of course, this is only one possible design of the cash-transfer. Higher and lower levels of both the baseline transfer and the taper can be considered. Once coverage becomes universal, the entire system could cost around 2 percent of GDP. Most of these expenditures would be on the cash-transfer; subsidies to the pension sys- tem would in fact decline and represent only 0.24 percent of GDP (see left-bottom panel in Figure 22). Regardless, expanding the coverage of the integrated system would involve a large increase in safety nets expenditures from current levels. The government, however, is already spending 2.61 percent of GDP on various forms of subsidies and transfers, which are not accounted in the budget of the social protection system. If coverage expands gradually, there could be room to reallocate subsidies to target poor households. Expenditures could also be controlled by offering a less generous transfer and/or increasing the taper. Finally, better labor market outcomes can reduce the level of cash-transfers, but this is something that is likely to take several years (see Section VIII). 74 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 22  Integrating cash-transfers and old-age pensions Source: World Bank staff calculations. Over the short and medium terms, the government also needs to consider alternative sources of financing by boosting revenues from indirect taxes on goods and services.54 Regardless of income level, several countries have been able to increase fiscal revenues by introducing or reforming value added taxes (VAT) and excise taxes. The latter include taxes on different consumption goods such as tobacco and alcohol and certain type of exports. Real estate taxes, including taxes on land, are also an efficient mechanism to mobilize reve- nues while minimizing distortions in the economy.55 An assessment of the tax-system in Côte d’Ivoire is outside the scope of this report. However, international experiences suggest that there is room to improve tax administration through better management, governance, and human resources; modern information systems; and better audits and verification programs. 54 See Akitoby(2018) 55 See Auerbach (2006). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 75 Delivery Implications for Côte d’Ivoire Côte d’Ivoire is in a good position to integrate across programs owing to recently-introduced delivery system tools. The integration of social assistance and social insurance programs would require harmonizing/integrating administrative systems, particularly PMTs and regis- tries. One of the main challenges is having a cost-effective, practical identification system. The country has a tested targeting system that combines geographic, proxy-means, and commu- nity validation. It has been working well in the case of the Productive Safety Nets Program. The government is also considering using the system in the case of non-contributory health insurance. With simple modifications, the existing proxy-means model could be expanded to allow the classification of households in the middle of the income distribution. Hence, it can be used not only to allocate subsidies among the poor but also non-poor informal sector workers. The registry for the Productive Safety Nets Program can also become a platform for a national registry that includes all workers and their families. As the CMU program is implemented and health insurance coverage expands to the informal sector, the same platform can be used for the universal pensions and unemployment benefit systems. 76 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 8 REVAMPING LABOR PROGRAMS Better jobs outcomes in Côte d’Ivoire will require, in part, sustained reforms to increase private investments, improve allocative efficiency, and promote economic growth. In general, when economies grow, they create jobs. When there are no distortions in the allocation of resources—labor and capital—those jobs are created in the sectors and economic activities where they generate the most value. Thus, over time, jobs tend to move from low to higher productivity sectors, from rural to urban areas, and from informal to formal activities. These structural transformations contribute to lifting people out of poverty and increasing standards of living. For this to occur the government needs to continue to work on five reform agendas: 1) macroeconomic policies to ensure prices stability and reduce uncertainty; 2) a business envi- ronment (including appropriate infrastructure) that reduces transactions costs and promotes competition and entrepreneurship; 3) labor regulations—labor taxation, minimum wage policy, dismissal procedures, and labor contracts—that protect workers without reducing incentives to create formal jobs; 4) governance and the rule of law to enforce contracts and the appropri- ability of returns on investments; and 5) human capital development. However, recent studies suggest that in the presence of jobs related externalities these standard policies might not be enough, particularly over the short-term. There are two sources of externalities.56 First, jobs social externalities that occur when the social value of a job is higher than that capture by workers and employers. For example, in Fragile, Conflict, and Violent (FCV) settings there can be social externalities linked to jobs for young men, which reduce the risks of criminality and radicalization and contribute to social stability. In addition, youth who are employed learn on the job, build their human capital, and make other workers more productive. Jobs for young women can also produce externalities by facilitating human capital accumulation in their children through reduced fecundity and more spending on early childhood development. Second, there can be labor externalities. These occur in the presence of high unemployment and/or underemployment, when the market price of labor deviates from 56 See Robalino, Romero, and Walker (2018). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 77 the opportunity cost of labor which can be very low. These two types of externalities create a gap between social and private rates of return on investment projects. Policies that increase investments and maximize the rate of return to capital and therefore output, do not necessarily generate the distribution of jobs needed to address youth unem- ployment, low female participation rates, inequality, or poverty. Even if policy makers succeed in tackling other factors that undermine firms’ private investment returns, the private sector still might not invest enough or might not generate the optimal portfolio of investments from the point of view of jobs outcomes. Private investments that would be socially efficient—in part because of the number and types of jobs they could create—would do not take place. Instead, an economy can see “too much” capital going into investments that are less efficient for soci- ety from a jobs perspective. Indeed, at the global level, there is a large variance in the patterns of investment, growth, and job creation. The same level of investments in different sectors not only generates a different net number of jobs (in some cases negative) but also a different composition of jobs in terms of age, gender, and skills level. There is also path dependence in the allocation of investments and job creation across economic regions. Many rural regions offer little in terms of good job opportunities, yet most poor or vulnerable workers live there. The implication for Côte d’Ivoire is that the government might need to take a more active role in improving jobs outcomes, which will require revamping existing labor programs57 with a particular focus on the poor and vulnerable. Today, most jobs in the country are in farms and own account work in small, low productivity, household enterprises. Many of these workers are poor, some of them enrolled in Productive Safety Net Program. Existing labor programs, however, are doing little to improve their situation. Most of programs focus on youth age 18–35 and offer services to connect them to self-58 or wage employment.59 But their coverage is very low; only 0.3 percent of the population (similar rates across income quintiles) and mainly in urban areas (0.4 percent vs. 0.1 percent in rural areas). Annual expenditures are only F CFA 9.5 billion (USD 19. million) or 0.06% of GDP, financed mostly by donors (the government cov- ers only 5% of expenditures in labor programs). There are also issues in terms of design and implementation. Although administrative data are lacking for most programs,60 case studies suggest that there is limited institutional capacity including in public employment offices. Only the Youth Employment and Skills Development Project has been evaluated and the results suggest that the program has had limited impacts on beneficiaries’ labor market outcomes.61 57 Two ministries oversee national employment policies and/or targeted ALMPs. At a policy level, the Ministry of Employment and Social Protection [Direction Générale de l’Emploi (DGE)] is responsible for labor legislation, social dialogue, and employment strategies. As part of the national employment strategy, in January 2016 the new Ministry of Youth and Youth Employment was created, whose mandate is to manage key ALMPs targeting youth falling under two different agencies, Agence Emploi Jeune and Bureau de Coordination des Programmes-Emploi (BCP-Emploi)]. 58 An example is the C2D Employment Project. It offers entrepreneurship support, startup incentives (cash and in kind grant, loans, training. 59 The largest program is the Youth Employment and Skills Development Project (PEJEDEC) which offers training, vocational, life skills, and cash for training.) 60 Only programs coordinated by the BCP-Emploi have been monitored on a regular and systematic basis since 2011. The BCP-Emploi provides implementation reports on a quarterly basis for PEJEDEC, PRISE, and C2D Emploi/AFD 61 Crepon Study 78 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Going forward, labor programs would need to be integrated and expanded, focusing on two interrelated objectives: creating better jobs for vulnerable workers and facilitating labor market transitions. Part of the agenda is to increase the productivity of the jobs most prevalent in low-income and poor households, particularly self-employment in agriculture and non-farm activities, while improving opportunities for job creation. This would require internalizing jobs related externalities and incentivizing investments in selected economic activities and regions. The other part of the agenda is to facilitate labor market transitions from inactivity/unemploy- ment into a job, and from low to higher productivity jobs. This involves removing constraints that preclude access to good jobs: lack of information about job opportunities; inability to move to the sectors and/or regions where the jobs are; lack of work experience; lack of experi- ence with job-search; or the lack of behavioral and technical skills. Promoting Labor Productivity and Job Creation for Vulnerable Workers There is an emerging consensus in that the creation of good jobs, particularly in the for- mal sector, takes place through start-ups, SMEs, and stablished firms; not micro-enter- prises. The evidence, in fact, shows that micro-enterprises rarely growth and create jobs. ¨Few large firms start as minnows. ¨62 Granted, the growth in sales and profitability of firms doesn’t always bring jobs. In some cases, firms do not follow a “growth-focused” strategy but instead an “efficient-growth”63 strategy where profitability is achieved by controlling costs and adopting production technologies that destroy jobs.64 But when it comes to the creation of wage employment, it is usually the case that young, fast-growing firms (“gazelles”) and more stablished firms are the ones responsible. Therefore, as discussed above, policies that create the conditions for entrepreneurship and investment will continue to be critical to promote job creation and labor productivity growth. However, when self-employment is the most prevalent form of work, programs that aim to support micro-entrepreneurship are also likely to have an important role to play within a country’s jobs strategy. In general, these programs offer a combination of services to improve knowledge and life skills; managerial practices; access to capital; and access to markets. In most cases, the main focus of these interventions is on increasing sales and profits; not on improving jobs outcomes. Nonetheless, there are a few lessons that can be considered to guide the design of new programs with three objectives in mind: 1) enabling the creation of new businesses for own-account workers; 2) increasing the productivity and earnings of exist- ing businesses; and 3) improving their survival and chances to grow. 62 Aterido et al (2019). The review presented in this section has been guided by their research. 63 See Anderson-MacDonald, Chandy and Zia (2014) for a discussion of these two models. 64 In a comparative study of Turkey, Vietnam, Morocco, and Moldova, Aterido and Hallward-Driemeier (2018) find that only a quarter to a third of the firms that experience an increase in turnover are also experiencing an increase in employment. In addition, positive net jobs outcomes related to growth episodes are sometimes accompanied by considerable job destruction and sluggish or negative productivity increase. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 79 The evaluations of traditional programs show that impacts on employment rates and earn- ings have been limited. For instance, a recent meta-analysis looked at some 40 programs supporting small scale entrepreneurship.65 These programs offered a combination of services such as technical training (including basic knowledge in business management), life skills train- ing, access to credit, and mentoring/networking. The analysis showed that only one fifth of the programs were able to have an impact on beneficiaries’ employment rates and/or earnings. On average, programs that combined training with access to finance—integrated programs—were more likely to have an impact on labor market outcomes. But, in general, the interventions were more likely to influence behaviors and to some extent business performance (level of debt and sales). Micro-credit initiatives, although popular, have also displayed disappointing results. Some studies find positive impacts on business creation and survival,66 but others are more skeptical. For instance, a study for Morocco suggests that access to microcredit does not necessarily encourage households to create their own business but rather supports business expansion in agriculture-related self-employment, mainly through the purchase of more animals.67 A review of other microcredit programs finds modest effects on ownership, startups and closures.68 The results are more promising when it comes to investments, business size and profits of existing micro-enterprises. The results suggest that micro-credit can work not to promote entrepre- neurship but as insurance schemes that could, in fact, encourage wage employment.69 More recent evaluations show that micro-enterprises are not necessarily constrained in terms of finance and knowledge (at least not only), but rather entrepreneurial capabilities. The evaluation of a programs providing small-size grants to entrepreneurs found that only a few firms could grow and that capital was not enough to explain why these microenterprises were not growing and hiring workers.70 Other reviews of the literature also suggest that techni- cal and business training have limited impacts on sales and profits and none on job creation.71 Instead, the main constraint to the growth of businesses seems to be the lack of motivation and entrepreneurial spirit. Thus, programs that transfer cognitive skills and motivate partici- pants to solve entrepreneurial challenges seem more likely to influence jobs.72 There is some evidence that comprehensive packages that focus on the very poor and operate at the household level can increase consumption but there are concerns about their capacity to operate at scale. These so called “graduation” programs, initially developed by BRAC, offer cash-transfers, training, mentoring, and some form of productive asset (often animals). Engagements with beneficiaries are more continuous and last longer than in the traditional entrepreneurship programs. Most evaluations show that the programs are able to increase per capita consumption per month by an average of USD 5 dollars.73 There are no 65 Cho and Honorati (2014). 66 Attanasio et al., (2011) and Augsburg et al., (2012). 67 Crepon et al.,(2015) 68 Banerjee et al., (2015). 69 Bruhn and Love., (2014). 70 De Mel et al (2008. 71 McKenzie and Woodruff (2013). 72 Glaub et al., (2012) and Campos et al. (2017) 73 Banerjee et al. (2015) 80 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH impacts on job creation, but many households have been able to escape poverty as a result of the interventions. The main challenge facing these programs relates to their potential to expand. Indeed, unit costs are relatively high, and production goes to consumption or is sold in small local-markets where there is little potential to grow. There are also examples of group level interventions for youth that have had positive impacts although scale can be an issue as well. In an entrepreneurship program in Uganda, young adults were invited to form a group (21 people per group on average) and submit a grant proposal for a business start-up and vocational training. This program also provided non-financial support through mentoring and guidance in the development and submission of the business plan. Successful proposals received a grant of about USD 380 per individ- ual. Groups invested mostly in tools and materials, rather than skills training. After four years, results show an increase in business assets by 57 percent, work hours by 17 percent, earnings by 38 percent, number of paid workers by 36 percent.74 In summary, the various evaluations point to some program features that might influence the likelihood of success. First, since the programs are dealing with disadvantage, heterogenous, groups likely facing multiple constraints, integrated programs tend to work better than single interventions.75 Second, group interventions and in-kind transfers might generate economies of scale (including in procurement) and produce better results than programs that operate at the individual level and provide in-cash transfers.76 Third, intensive and specialized programs are more likely to have some long-lasting effects on entrepreneurs’ business practices than “light touch” ones.77 Related to this, follow-up visits and/or personalized coaching after training seem to increase the impacts of the programs.78 In terms of the delivery of technical assis- tance, relying on local role models who can share their experience is more efficient than in class or personalized technical assistance onsite.79 Nonetheless, the main challenge remains how go bring programs to scale when dealing with subsistence entrepreneurs and businesses that, in most cases, have little potential to grow. The beneficiaries of the programs discussed here do not enter self-employment/entre- preneurship by choice or vocation. They do it out of necessity given the absence of other income generating activities. For the most part, they lack key non-cognitive skills that are not easily transmitted: perseverance, discipline, risk-taking, leadership and communications. Even if some of these skills could be transferred, as discussed above, there is a question about the economic viability of their enterprises. These are not selected based on an understand- ing of market demands and the potential to generate profits and growth. Usually, businesses are chosen because others—family members and friends—are involved in them. They tend to produce non-tradable goods which are sold in very narrow and shallow local markets. There is a limited amount of these goods and services that can be consumed; at some point, 74 Blattman, Fiala and Martinez, (2014). 75 Berge et al., (2014); De Mel et al., (2014); Gine et Mansuri, (2011) 2014?; Karlan et al., (2015); Klinger and Schundeln, (2011); Marcours et al., (2012); Fafchamps et al., (2011). 76 Fafchamps et al., 2014). 77 Anderson-MacDonald et al., (2016) 78 Blattman et al., (2013); Higuchi, (2016); and Valdivia, (2012). 79 Lafortune et al., (2017) CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 81 having more entrepreneurs and/or higher labor productivity can, in fact, depress prices and incomes.80 Jobs Subsidies, Self-employment, and Larger-Scale Private Investments One way to deal with the problems discussed above is to have self-employment (SE) pro- grams that treat beneficiaries not as real entrepreneurs, but more as wage employees, or contract workers, who are paid for the provision of certain goods and services. Part of the design of the programs would involve profiling applicants. Those who have the potential to engage in real entrepreneurial activities would be connected to traditional entrepreneurship programs. Beneficiaries who stay in the SE program would still need to receive training and support to purchase inputs, but they would not be trying to conceive and setup a business on their own. Their business, in a way, would be part of larger private/public investments projects to develop parts of or entire supply chains, or produce public goods. These would be invest- ments that the private sector is unlikely to undertake alone given low expected-rates of return. However, once jobs social externalities are taken into account, social rates of return on invest- ments can be high enough to justify government subsidies. There are three operational models that Côte d’Ivoire could consider: (1) SE programs imbed- ded in supply chain development initiatives; (2) SE programs imbedded in aggregator pro- grams that facilitate access to markets; and 3) SE programs imbedded in initiatives to produce public goods. In the first two cases the demand for goods and services is market driven. In the last case, the demand is guaranteed by the government as part of its efforts to deliver social services. Some of the main elements in terms of design and implementation are discussed below. It is important to keep in mind that while several countries have been implementing this type of projects, there are no yet evaluations or evidence about potential impacts. Any new program would have to have well designed monitoring and evaluation systems. Supply Chains. Supply chains of relatively simple products, like chocolates, bring together three different types of activities: 1) the production of raw materials (e.g., cacao); 2) transforma- tion activities (e.g., drying cacao and producing chocolate powder, or transforming chocolate powder into chocolates); and 3) support services (e.g., distributing fertilizers and seeds, plant- ing and harvesting, renting machinery, or marketing final products). In developing countries, for several reasons, many of these supply chains are underdeveloped. Local producers might lack the initiative or knowledge, transaction and coordination costs can be high, expected private rates of return can be low. Yet, within each of these three sets of economic activities there are opportunities for job creation and for increasing the productivity existing jobs. Investment projects that aim to develop supply chains while maximizing job opportunities through SE programs would need to focus on four core activities: 80 Yoonyoung Cho & David Robalino & Samantha Watson, 2016. “Supporting self-employment and small-scale entrepreneurship: potential programs to improve livelihoods for vulnerable workers,” IZA Journal of Labor Policy, Springer;Forschungsinstitut zur Zukunft der Arbeit GmbH (IZA), vol. 5(1), pages 1–26, December. 82 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH • Preparing a business plan for the development of the supply chain in a given region. This is, naturally, the cornerstone of the project. The analysis and business proposal need to be objective, technically sounds, and able to attract private investors (see below). This is not an activity that can be conducted by government institutions or staff and consultants in international organizations. There are, however, several international foundations and social enterprises that, depending on the sector/product, would have the expertise to con- duct this type of analysis. Contracting and payment systems would need to ensure that, to some extent, those involved in the preparation of the business proposal share the benefits of success and the costs of failure. An important criterion to assess the viability of the proj- ect is its social rate of return, which takes into account jobs related externalities.81 • Mobilizing private investments to co-finance the project. As discussed above, jobs exter- nalities can justify the use of government subsidies to facilitate private investments. The costs of “discovery” and preparing an initial business plan are already part of the subsidy the government would be providing. In addition, there can be room to subsidize part of total investment costs; a form of matching grant. At the end, however, private investors need to participate, signaling that the project is worthwhile. Essentially, prospective inves- tors would be buying equity in the project at a subsidized price. They would be agreeing to fund, part, of the production process in exchange for a share of sales of the final product. Production costs would include, for instance, payments to local cacao producers, pay- ments to growers of cacao plants, equipment required to dry cacao (these costs would depreciate over time), wages of project managers and other wage employees, and water and electricity. Given that investors are assuming a risk, they should be able to influence the final business plan and decisions regarding project management/implementation, including the criteria to select and deliver the managers of SE programs. There are differ- ent types of contractual arrangements that could be considered. • Identifying the project management team. This team brings together the real managers and entrepreneurs behind the business “start-up.” They are highly qualified and moti- vated individuals who are hired on a competitive basis and who share part of the profits generated by the investment. They are not government officials or consultants paid by international organizations. More likely, they are young professionals interested in making a difference in terms of economic and social development and building field experience; they have the skills to compete for jobs in large consulting companies and investment banks. Some of them could have been involved in the preparation of the business plan. They would be in charge of all aspect of the business, including record keeping, account- ing, and financial management. • Designing and implementing SE programs. The main objective of these programs is to develop local income generating opportunities in different segments of the supply chain. The business model, by design, does not allow for imported inputs or support services that could be produced locally, or subcontracting the production of final products with exist- ing companies (if available). The project management team therefore has to subcontract with for profit or non-profit organizations the design and implementation of programs that would recruit workers who engage, as self-employed workers or micro-entrepreneurs, in the different economic activities identified within the supply chain. The programs are 81 See Robalino, Romero and Walker (2018) for a discussion about how to calculate social rates of return in the presence of jobs externalities. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 83 similar in many ways to traditional entrepreneurship programs. They need to register, profile, and select applicants and, eventually, deliver a given package of services including training and grants. They can also enforce standards not only in terms of production but also working conditions. The main difference is that the set of viable activities/businesses has been defined ex-ante. Applicants are not asked about their business ideas, but instead about their preferences about supply chain activities where they would like to engage. Organizations designing and implementing these programs would be selected competi- tively under contracting and payment systems that take into account performance and the achievement of specific labor market outcomes. Aggregators. These are simpler and cheaper models, relevant in particular for farmers, but likely to create fewer job opportunities and generate less value added. The government in this case subsidizes private “aggregators” who bid to manage the distribution and commer- cialization of a given product or products (produced locally) under certain labor standards. For instance, private investors can bid to distribute cacao of a certain quality to regional or inter- national markets. They become responsible for training local producers, providing machinery and inputs, and other activities related to the business, but they are only assessed based on outcomes. In principle, they can also engage in transformative activities controlling their own production processes, including decisions about how and where to purchase indirect inputs (e.g., fertilizers). Projects of this nature would require a simpler setup. Three activities need to receive special attention: • Assessing business opportunities. It is still necessary to have a good idea of the potential for commercializing various products in the region. This, again, would require subcon- tracting studies to specialized institutions. Contracting and payments systems based on performance are likely to have a less important role to play. • Selecting aggregators. Different criteria can be used, with different weights, to select “aggregators” who bid for specific projects. Recent analytical work suggests that it is important to include as the main criteria the level of jobs externalities generated per unit of subsidy.82 • Monitoring and evaluation. Whereas in the case of supply chain development projects there are strong built-in incentives to achieve project objectives, aggregator programs require allocating more resources to monitoring, evaluation, and enforcement. Participants in the program are not direct beneficiaries but agents to achieve the social objectives of the project in terms of improving jobs outcomes. Although they are taking an investment risk, they have weak incentives, for instance, to achieve objectives in terms of job cre- ation or labor productivity grow as per their business plans. This implies that contracting arrangement to disburse matching grants need to be conditioned on the achievement of specific outcomes. Enforcing these contracts requires having in place appropriate monitor- ing and evaluation systems. Public Goods. In this case, the government identifies social services that are needed by local communities. There is a broad range: child or elderly care; education and training; health 82 See Robalino, Romero, and Walker (2019) for a discussion of how to setup a competitive bidding mechanism for allocating matching grants in the presence of jobs externalities. 84 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH care; environmental protection; energy and water supply; basic infrastructure (e.g., second- ary roads, waste disposal); even the design and implementation of SE programs to support supply chain development (see above). Like in the case of aggregator programs, the govern- ment subsidizes, on a competitive basis, private investments in businesses that deliver these services. Projects are evaluated based on jobs outcomes with a focus on jobs for the local community. Entrepreneurs who apply to the grants, however, do not have to live in the region. In fact, the goal of this type of initiative is to mobilize social entrepreneurship talent from around the world. One difference with aggregators or supply chains development programs, is that the level of government subsidies can be higher. Indeed, the projects not only internal- ize jobs externalities but also the social externalities of the goods and services they deliver. Nonetheless, similar activities need to receive attention: 1) defining priorities in terms of social goods and services; 2) selecting social entrepreneurs and allocating matching grants; and 3) monitoring, evaluation, and enforcement (see above). Facilitating Labor Market Transitions Labor programs, often managed by public employment offices (PEOs), can have an important role to play in helping individuals find jobs or move between jobs. Like Côte d’Ivoire, most countries have them. They all offer similar services such as counseling, intermediation, job- search assistance, wage subsidies, technical and life skills training, and mobility grants. Some also include the type of entrepreneurship programs discussed in the previous section. Unfortunately, existing evaluations suggest that in most cases labor programs are not achieving the objectives for which they were designed. The most recent meta-analysis of around 80 programs implemented worldwide finds that only 30 percent of programs had a positive impact on beneficiaries’ employment rates and/or earnings.83 More troublesome, among those programs which had an impact the impact was relatively small. The report sug- gests that the main reason behind these findins are problems in terms of design and imple- mentation. Beyond limited institutional capacity and scarce resources, program managers do not always have the incentives to respond to the needs of employers and job seekers. In addi- tion, when the programs focus on wage employment and the economy is not creating enough jobs there is little room to improve labor outcomes. This is why it is important to accompany these programs with interventions to create and improve jobs opportunities (see previous section). The review of the evaluations offers a few recommendations about how to improve the design and implementation of labor programs to improve their chances of success. Four elements are emphasized: 1) profiling systems; 2) integration of services; 3) contracting and payments systems with providers; and 4) monitoring and evaluation systems. In the case of Côte d’Ivoire, the following reforms would need to be considered. • Introducing modern profiling systems (PSs). The main function of employment offices should be to register and profile jobs seekers. This is similar, in terms of administrative 83 See Kluve et al., (2018). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 85 processes, to the registration and means-testing of beneficiaries in the Productive Safety Net Program. The difference is that, in addition to means-testing, employment offices (or the managers of the labor programs) should also profile job seekers to assess the main constraints they face to find or switch jobs. Like the PMT, the Profiling System (PS)84 is a statistical model that, based on a short survey at the time of registration, groups job seek- ers into categories. From the “hard-to-serve,” individuals with the most binding constraints (e.g., psychological disorders), to those requiring minimal assistance. This is an important exercise because it will affect not only the types of services that individuals in each cate- gory receive, but also the cost to the package of services (see below). • Integrating and pricing services. The population of jobs-seekers is highly heterogenous; not all individuals face the same problems when searching for a job. This implies that labor programs should be able to offer a menu of services; an integrated package of service, that is adapted to individual needs (e.g., counseling, different types of training, coaching for interviews, wage subsidies). This is similar to the case of health insurance where there is a basic package of health services that all individuals can use. What they utilize at the end, however, depends on the health problems they have. Also, like in the case of health insurance, the package of labor services should be priced. The expected cost of the pack- age for hard to serve individuals, for instance, will be higher than that of individuals who require minimal assistance from the labor program.85 • Outsourcing the provision of services and paying based on results. Employment offices seldom have the resources and expertise to follow-up job seekers, coordinate with employers, and provide them with the set of services they need. An alternative model is to outsource the provision of services to specialized institutions, in the same way a health insurance fund outsources service provision to doctors, clinics, and hospitals. To this end, it is important to have the right contracting and payments systems in place. One option is to pay providers by services delivered (e.g., hours of counseling or training); the equivalent of fee-for-services for doctors. The problem with this method is that it does not give providers incentives to improve the quality of services, control costs, and ultimately improve labor market outcomes for beneficiaries. An alternative is to pay providers based on perfor- mance. There are different contracts that can be considered.86 Providers can receive a lump-sum payment for each program beneficiary they manage (the expected cost of the package for that particular beneficiary), but the renewal of contracts depends on having achieved certain targets with performance indicators (e.g., placement rates, employment rates, or changes in earnings). Or, they can bill by services provided at prices negotiated ex-ante. These prices would have two components: a fixed rate and a rate based on having achieved specific outcomes with the beneficiary in question (e.g., the beneficiary obtained an internship, a part time job, a full-time job). In both cases, average performance indicators by provider should be made public. • Introducing modern M&E systems. Program managers need to be able to generate admin- istrative data to regularly monitor costs and performance, and introduce corrective mea- sures as needed. This information is also critical to identify individuals who risk dropping 84 See Datta et al., (2018) for a discussion about the design and implementation of profiling systems. 85 See World Bank (2019) for a discussion about the similarities between well designed labor programs and health insurance programs. 86 See Datta et al., (2018) for a review of alternative types of contracts. 86 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH out of the system or cases where existing services (treatments) are not yielding the desired results. In addition, the information is needed to be able to enforce contracts and pay pro- viders. Today there are standardized instruments that can be used to collect the necessary data. These are essentially short surveys that are used to follow-up program beneficiaries, obtaining information about services used and different types of labor market indicators.87 87 See Krishnan et al., (2017). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 87 88 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH PART III: INSTITUTIONAL AND FISCAL FRAMEWORK FOR REFORM The government of Côte d’Ivoire is assessing options to improve the design and expand the coverage of its social protection and labor system from an inclusive growth perspective. Key questions include how to reform pensions, unemployment benefits, and health insurance systems to improve their financial sustainability and extend their coverage to informal sector workers; how to integrate redistributive arrangements within social assistance and social insurance programs to improve equity; and how to restructure active labor market programs to maximize their impacts on jobs. As the final part of the policy note, this section develops a holistic institutional and fiscal framework to support achieving universal social protection while improving job opportu- nities. This section starts by summarizing, briefly, the main types of reforms that would be required and their potential costs over the next, for example, fifteen years under different scenarios about coverage expansion. It then discusses the structure and expected evolution of the public budget in Côte d’Ivoire based on publicly available data and compares different options to create fiscal space such as: 1) higher revenues; 2) a reallocation of public expendi- tures; and/or 3) improvements in the management of the public debt. The discussion con- cludes with considerations for addressing institutional and political economy issues, and the types of administrative systems needed for successful implementation. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 89 90 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 9 ACHIEVING UNIVERSAL SOCIAL PROTECTION AND BETTER JOBS: POLICIES AND FISCAL IMPACTS The policy framework proposed to improve the design and expand the coverage of the social protection systems is based on three fundamental principles: • Having integrated social insurance programs that treat all workers and their families equally. There is no economic rationale for having separate programs for different types of workers. Doing this not only creates inequalities between workers and can fragment the labor market, it also increases complexity and administrative costs. Instead, all workers should have the same rights and obligations regardless of where they work. Civil servants, informal workers, farmers, and wage employees in the formal private sector can all enroll in the same system. The only difference would be in terms of how much and how they contribute. • Defining explicitly the benefits offered to plan members and their costs. This is critical to ensure the financial sustainability of the system as coverage expands and unit costs change over time, and to be able to identify the most efficient financing mechanism. In the case of pensions, this implies linking the contribution rate to the benefits accrual rate and life expectancy at retirement. For health insurance it implies defining the content of package of health services offered to plan members and estimating its average costs by age (children, adults, and the elderly) and gender. For unemployment insurance it implies linking the contribution rate to the replacement rate and the risk and duration of unemploy- ment for different population groups. • Having explicit and efficient distributive arrangements. The expansion of social protec- tion systems requires having in place the right redistributive arrangements to help those CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 91 who do not have the means to pay in full the cost of different insurance programs, or who are at risk of falling into poverty. Improving the design of these redistributive arrange- ments implies: 1) allocating subsidies based on means and not based on where individuals work; 2) integrating subsidies within social insurance and social assistance programs; and 3) financing subsidies though general revenues and not taxes on labor that can reduce incentives to create formal jobs. In practice, there would be two types of subsidies that can be universal or targeted: ― Subsidies to guarantee a minimum level of income for all workers regardless of employment status and whether they contribute or not to social insurance programs. ― Subsidies to fund all or part of the cost (contributions to) of social insurance programs. The social protection system envisioned for Côte d’Ivoire would integrate social insurance/ assistance programs and active labor market programs. This integration would allow for a better allocation of limited public resources among competing programs. In particular, there are tradeoffs between social insurance/assistance programs and labor programs. For instance, investing today in programs that improve jobs and earnings opportunities can reduce the future cost of social insurance/assistance programs. The integrated social protection system would comprise five programs: • Basic income guarantee (1.62% of GDP, covering the population). This is an anti-poverty transfer that is allocated at the individual level, not the household level. In the calculations, the transfer is equal to 10 percent of average consumption with a 20 percent taper (i.e., for each 10 F CFA increase in income/consumption, the transfer is reduced by 2 F CFA). The actual cash-transfer of 12,000 per month per household is equivalent to 15 percent of GDP per capita or close to a per capita transfer of 5 percent of average consumption. Hence, the cash-transfer used in the simulations is around twice the existing transfer. • Pensions (0.24% of GDP, covering the working age population). A reformed, defined- ben- efit pay-as-you-go (DB-PAYG), pension system managed by the CNPS would become the scheme for all workers, including informal sector workers. This would imply redefining the mandate of the system and modifying benefits formulas and eligibility conditions to ensure the financial sustainability of the system. It would also require financing the unfunded liabil- ities of the system. The new system would apply to all workers who enter the labor market after a given date and to the new contributions of existing plan members after that date. The main features of the new system are the following: ― It would target a replacement rate of 50 percent for full career workers (40 years) who retire at age 62, while allowing individuals to save more if they decide to. This implies setting a baseline accrual rate equal to 1.25% and a contribution rate of 11.25%. ― The system would aim to guarantee a minimum pension equivalent to 15% of average earnings. To this end, it would offer higher accrual rates and subsidized contributions to workers in the lowest deciles of the income distribution. These subsidies would depend on average incomes and expected contribution densities within the decile. 92 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH • Health insurance (universal at 5.67% of GDP, or with targeted subsidies at 2.55% GDP, covering the population). The assumption here is that the government continues with the implementation of the stand-alone health insurance fund (CNAM) for the poor and informal sector workers. The Programme de Couverture Maladie Universelle (CMU), would eventu- ally enroll all workers. The main features of the program are the following: ― The cost of the package of health services offered by the CMU is equal to USD 100 per capita per year or 6% of GDP per capita. ― Two financing mechanisms are considered: 1) universal subsidies; and 2) targeted sub- sidies. When targeting, subsidies for the first three deciles remain at 100% percent but decline to 50% for deciles 4, 5 and 6; deciles 7 and above do not receive subsidies. • Unemployment insurance (0.12% of GDP, at 50% of the working age population). The assumption is that Côte d’Ivoire introduces an unemployment insurance program that covers both formal and informal workers in the top 5 deciles of the income distribution. The common concern about workers abusing the system and claiming benefits while working informally would be addressed by conditioning payments to participation in labor programs (see next). The scheme would be managed by the same institution managing the universal pension system and/or the universal health insurance fund. The main features of the new system are the following: ― The program offers a benefit equal to 70% of earnings for a duration of 3 months. ― The contribution rate is set at 3.6% and the government covers deficits resulting from the participation of higher-risk workers. • Active labor market programs (1% of GDP—working age). Labor programs would need to be integrated and expanded, focusing on two interrelated objectives: creating better jobs for vulnerable workers and facilitating labor market transitions. The focus of the reform would be on the following: ― Adopting a modern profiling system ― Integrating and pricing services (e.g., training, counseling, job search assistance) ― Outsourcing the provision of these services to private providers through performance base contracts ― Introducing a modern monitoring and evaluation system The projections of fiscal costs presented below depend mainly on demographics. The assumption is that per capita costs in the base year are indexed with the growth of GDP per capita. In turn, GDP per capita depends on the growth rate of GDP which is given by the growth rate of the working age population and labor productivity growth. We assume that the annual real growth rate of labor productivity is 1%. Changes in the fiscal cost of each program are given by changes in the age composition of the population. We look at two scenarios: 1) low expansion of coverage where universal coverage is attained in year 2030; and 2) fast expansion where universal coverage is reached in year 2025. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 93 FIGURE 23  Projections of Social Protection System Components Source: World Bank staff calculations. 94 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Depending on coverage targets and whether the health insurance system is universal or targeted expenditures by year 2025 could range between, the reformed social protection system, including health insurance, could cost between 4% and 12% of GDP. The largest item is the health insurance system, followed by the basic income guarantee and active labor market programs (see Figure 23). It is also assumed that the reform of the pension system would require making part of the implicit pension debt (the part that is not backed by financial assets and the pay-as-you-go asset) explicit. Repaying this debt as individuals retire could cost around 0.3% of GDP per year. Creating Fiscal Space to Enable Reforms One of the main functions of economic policy is to mobilize revenues from taxes to address social externalities and fund the production of different types of public goods. This taxation necessarily creates distortions in products and labor markets that reduce output and jobs; a social cost. But there are also important social benefits. For instance, if taxes are used to deal with negative externalities such as pollution, carbon emissions, the loss of human capital, or positive externalities such as better education, health, and jobs for all. Taxes also finance the production of public goods such as national security, public health, and basic infrastruc- ture; goods that the private sector would not produce enough off. But the demands can be infinite and the resources that can be mobilized limited. This implies that governments need to have well defined priorities and strict rules to allocate public expenditures considering the social costs and benefits of taxes, transfers (subsidies), and public goods. The structure of the budget of a given country reflects, implicitly, the priorities in terms of externalities and public goods as well as fiscal space constraints; in Côte d’Ivoire health and social protection receive only a small share of the budget. In Côte d’Ivoire, the government spends close to 23.5% of GDP, which ; in line with expenditure levels for countries at the same level of economic development. About half of these expenditures are allocated to the salaries of civil servants (6% of GDP) and government operations (4%), and another 6% to investments (essentially the production of public goods). The rest of expenditures go to interest payments on the public debt (2%), pension expenditures (1.3%), and subsidies/transfers (1.7%). Usually, these transfers aim to address problems such as poverty, access to education, or access to risk management instruments (pensions, and health and unemployment insurance). In Côte d’Ivoire, however, the transfers also go to state owned enterprises which are not necessarily involved in the production of public goods; the private sector could produce the same goods and services (see Table 18). In terms of the sectoral allocation, the majority of the budget goes to fund the operations of the central government, housing, and education. Only 7 percent of expenditures to go health sector and only 1.2% to social protection (see Table 19). With the current structure of the budget, the integration and expansion of the social protec- tion system would severely increase the deficit and the level of public debt. In the scenario with a gradual increase in coverage and targeted subsidies in the health insurance system, the government deficit could reach 3.2% by year 2025 (excluding interest payments). As a result, the public debt would increase to 53% of GDP. This assumes that revenues and expenditures unrelated to the social protection system remain constant as a share of GDP (see Table 20). CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 95 TABLE 18  Structure of the Public Budget in Côte d’Ivoire, 2015–2021 En milliards de FCFA 2015 2016 2017 2018 2019 2020 2021 Agrégats Réalisation Réalisation Réalisation Estimation Estimation Estimation Estimation Recettes et dons 19.99% 19.77% 20.36% 20.37% 20.51% 20.72% 20.53% Recettes fiscales 15.08% 16.02% 16.47% 16.48% 16.98% 17.32% 17.41% Impots intérieurs 7.55% 8.58% 8.93% 9.14% 9.87% 9.95% 10.20% dont Impots sur revenus et salaires (Hors FDFP) 1.81% 1.93% 1.94% 2.16% 2.17% 2.01% 1.99% dont Impots indirects 3.03% 3.26% 3.32% 3.31% 3.35% 3.25% 3.37% dont TVA (hors part secteur électricité) 1.38% 1.42% 1.57% 1.78% 1.86% 1.89% 1.96% Droits et taxes à l'importation 5.55% 5.56% 5.53% 5.65% 5.56% 5.85% 5.80% Droits et taxes à l'exportation 1.98% 1.87% 2.01% 1.69% 1.55% 1.51% 1.41% Recettes non fiscales 3.47% 2.54% 2.69% 2.63% 2.36% 2.28% 2.11% Cotisations de sécurité sociale 1.96% 2.12% 2.06% 1.98% 1.93% 1.98% 1.83% Autres 1.51% 0.42% 0.64% 0.65% 0.44% 0.30% 0.28% Dons 1.44% 1.21% 1.20% 1.26% 1.16% 1.12% 1.01% Projet 0.66% 0.51% 0.52% 0.60% 0.55% 0.55% 0.49% Programme 0.78% 0.71% 0.68% 0.66% 0.61% 0.56% 0.52% Dépenses et prêts nets 22.81% 23.96% 24.87% 24.35% 23.55% 23.62% 23.39% Dépenses Courantes 14.17% 13.81% 14.98% 14.49% 13.72% 13.34% 12.88% Personnel 6.80% 6.69% 6.83% 6.83% 6.62% 6.30% 5.96% Prestations sociales 1.30% 1.25% 1.19% 1.26% 1.27% 1.33% 1.23% Subvention et transferts 2.12% 1.84% 1.94% 1.69% 1.50% 1.46% 1.42% Subvention aux écoles privées 0.50% 0.50% 0.50% Subvention aux EPN 0.35% 0.32% 0.30% Subvention coton 0.00% 0.00% 0.00% Subvention au secteur électricité 0.00% 0.00% 0.00% Fonds de réserves café cacao 0.00% 0.00% 0.00% Collectivités décentralisées 0.09% 0.08% 0.07% Bourse, kits scolaires et transports 0.27% 0.25% 0.24% Autres subventions 0.30% 0.30% 0.30% Fonctionnement (yc sociales ciblées) 3.96% 4.03% 5.02% 4.71% 4.33% 4.24% 4.27% Investissement 6.36% 6.73% 6.89% 6.93% 6.62% 6.80% 7.35% Financement intérieur 4.04% 5.04% 4.19% 3.96% 3.93% 3.88% 4.08% Financement extérieur 2.33% 1.69% 2.71% 2.97% 2.69% 2.92% 3.27% Emprunts 1.67% 1.21% 2.19% 2.37% 2.14% 2.36% 2.79% Dons 0.66% 0.49% 0.52% 0.60% 0.55% 0.55% 0.49% Autres (dépenses en capital, sortie de crise, déchets toxiques) 0.75% 1.70% 1.29% 1.16% 1.23% 1.41% 1.15% Prêts nets 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% Intérêts 1.52% 1.72% 1.71% 1.78% 1.98% 2.07% 2.00% Intérieur 0.74% 0.88% 0.92% 0.91% 0.81% 0.99% 1.04% Extérieur 0.78% 0.85% 0.79% 0.87% 1.18% 1.07% 0.97% Solde base ordonnancement -2.72% -4.00% -4.51% -3.98% -3.04% -2.90% -2.86% Solde budgétaire de base (Définition UEMOA) -1.15% -3.00% -2.32% -1.62% -0.91% -0.53% -0.07% Solde primaire de base -0.42% -1.89% -1.27% -0.50% 0.47% 0.97% 1.41% Financement 2.75% 3.98% 4.49% 3.02% 2.19% 2.97% 2.95% Extérieur 3.98% 2.14% 5.05% 5.64% 3.40% 2.67% 3.15% Tirages (yc appui budgétaire et Eurobonds) 5.14% 1.68% 7.85% 7.02% 4.06% 2.36% 2.79% Amortissement -1.18% -1.29% -3.21% -1.21% -1.40% -1.30% -0.68% Autre financement extérieur 0.01% 1.75% 0.42% -0.17% 0.75% 1.60% 1.04% Intérieur -1.22% 1.83% -0.57% -2.62% -1.21% 0.30% -0.20% Bancaire -0.77% 1.68% 0.30% -2.27% -1.19% 0.43% -0.01% Autre -0.45% 0.15% -0.87% -0.35% -0.02% -0.12% -0.19% Ecart (- sur financement) 0.07% -0.03% -0.02% -0.96% -0.85% 0.07% 0.10% PIB 100.00% 100.00% 100.00% 100.00% 100.00% 100.00% 100.00% Source: World Bank staff compilation based on available data from Ministry of Economy and Finance. 96 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 19  Distribution of the Public Budget by Sector in Côte d’Ivoire, 2016–2021 En millier FCFA Exécution Exécution Budget Budget Projection Moyenne 2019-2021 2016 2017 2018 2019 2020 2021 Service généraux des administrations 1,195,850 1,163,310 1,093,020 1,265,978 23.0% 1,174,012 20.3% 1,197,492 19.6% 1,212,494 20.9% publiques Défense 276,941 327,580 336,568 334,434 6.1% 577,206 10.0% 667,462 10.9% 526,367 9.1% Ordre et sécurité publics 212,431 209,970 234,679 256,770 4.7% 276,690 4.8% 276,360 4.5% 269,940 4.7% A aires économiques 367,785 332,755 402,994 417,151 7.6% 410,002 7.1% 582,468 9.5% 469,874 8.1% Protection de l'environnement 69,957 72,072 79,431 133,680 2.4% 119,927 2.1% 115,445 1.9% 123,017 2.1% Logements et équipements collectifs 545,344 925,272 1,217,513 1,190,533 21.6% 1,253,600 21.7% 1,192,768 19.6% 1,212,300 20.9% Santé 300,288 329,811 345,192 372,358 6.8% 343,993 6.0% 401,592 6.6% 372,648 6.4% Loisirs, culture et culte 115,326 143,506 162,485 136,515 2.5% 162,895 2.8% 173,644 2.8% 157,685 2.7% Enseignement 1,201,610 1,176,221 1,283,885 1,341,272 24.3% 1,398,189 24.2% 1,428,886 23.4% 1,389,449 24.0% Protection sociale 66,882 52,086 53,214 64,211 1.2% 58,027 1.0% 63,204 1.0% 61,814 1.1% Reserve de programmation - - - - 0.0% - 0.0% 0.0% - 0.0% TOTAL 4,352,414 4,732,583 5,208,981 5,512,902 100.0% 5,774,541 100.0% 6,099,321 100.0% 5,795,588 100.0% Dette Publique 1,520,437 1,973,417 1,547,278 1,821,400 2,119,711 2,011,326 1,984,146 TOTAL (y compris dette publique) 5,872,851 6,706,000 6,756,259 7,334,302 7,894,252 8,110,647 7,779,734 Source: World Bank staff compilation based on available data from Ministry of Economy and Finance. TABLE 20  Projection of the Fiscal Deficit and Public Debt in Côte d’Ivoire, 2019–2025 2019 2020 2021 2022 2023 2024 2025 Revenues 20.51% 20.72% 20.53% 21.92% 21.59% 21.28% 20.99% Total Taxes 9.87% 9.95% 10.20% 10.20% 10.20% 10.20% 10.20% Others 10.64% 10.76% 10.33% 11.72% 11.39% 11.08% 10.79% Expenditures 23.55% 23.62% 23.39% 23.16% 23.49% 23.82% 24.15% Subsidues and transfers 1.50% 1.46% 1.42% 1.42% 1.42% 1.42% 1.42% Health 1.43% 1.22% 1.32% 1.32% 1.32% 1.32% 1.32% Social Protection 0.25% 0.21% 0.21% 0.21% 0.21% 0.21% 0.21% Others 20.36% 20.73% 20.45% 18.45% 18.45% 18.45% 18.45% Integrated Social Protection System 0.00% 0.00% 0.00% 1.78% 2.10% 2.43% 2.76% Deficit -3.04% -2.90% -2.86% -1.24% -1.90% -2.54% -3.16% Public Debt 44.35% 41.01% 37.84% 45.48% 47.43% 50.07% 53.37% Source: World Bank staff compilation based on available data from Ministry of Economy and Finance. Notes: Expenditures excludes interests payments starting in 2022. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 97 There are essentially three mechanisms to accommodate higher expenditures, discussed below in the case of Côte d’Ivoire: 1) increasing revenues from taxes; 2) reallocating public expenditures; and 3) improving debt management. Higher revenues from taxes. Côte d’Ivoire mobilizes around 10 percent of GDP in taxes, but only 2 percent from income taxes and another 2 percent from VAT taxes. The government is already engaged in a series of reforms to increase revenues from taxes such as expanding the tax base, eliminating exemptions, strengthening enforcement and improving administrative efficiency.88 It seems reasonable to expect that revenues from income and consumption taxes could increase to around 10 percent of GDP in the next 10 years. This measure alone could generate sufficient fiscal space to accommodate the expansion of an integrated social pro- tection system without a major increase in the public debt. Other things being equal, by year 2030 the government budget could generate a primary surplus and the public debt would stabilize at around 40% of GDP. Reallocating public expenditures. There are different initiatives that the government is taking to control or reduce public expenditures, including by improving planning and the budget process, implementing a program to control the wage bill, and having a more efficient pro- curement system. There is also room, however, to rationalize subsidies and transfers and other social protection expenditures. Transfer to SOE would need to be gradually phase-out and as the new integrated system expands it would be able to replace the other programs. Thus, assuming that transfers and social protection expenditures converge to zero by 2030, the long-term deficit could be reduced by another 1% of GDP. Improving debt management. There are two interrelated aspects to public debt management: having the capacity to borrow and therefore the ability to run higher deficits; and reducing the cost of servicing the debt. The government has defined clear targets for managing the public debt in the context of the Stratégie pour la Gestion de la Dette à Moyen Term (SGDM). It is likely that this strategy could accommodate somewhat higher deficits and higher level of public debt in order to support the expansion of the social protection systems, but only at the margin. A seen in Figure 24, the expansion of the social protection system cannot be achieved sustainably by issuing debt. Being able to mobilize sufficient tax revenues is a pre-condition. Managing the Political Economy of Reform and Building State Capabilities An emerging consensus on reform internationally highlights that achievement depends on the policy-making process, which is the process of discussing, approving, and implementing public policies. Why are some countries able to implement complex social protection reforms and others are not? It is difficult to come up with a set of factors that would predict the prob- ability of success. GDP per capita, for instance, is not a key determinant; low income coun- tries like Bolivia, India, and Pakistan have been able to introduce important reforms in their 88 Loi des Finances and data from Ministry of Economy and Finance. 98 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 24  Deficit and Public Debt with Changes in Tax Revenues and Transfers Source: World Bank staff calculations based on available data from Ministry of Economy and Finance. pensions, health insurance, and/or safety nets programs. There are, nonetheless, some useful insights from the literature on the political economy of reform and the more recent literature on state capabilities. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 99 Countries that have been able to introduce successful reforms have a policy-making process where political actors and social partners cooperate and are able to reach and enforce agreements. In general, successful reforms have been defined by: 1) strong leadership; 2) the existence of good “aggregators” that can reduce the number of actors who can directly influ- ence policy; and 3) open dialog and continuous interactions among these actors within a long- term planning horizon.89 Successful countries also have a strong bureaucracy to which the analysis and implementation of policies can be delegated. This tends to be underestimated. There can be countries with a good policy-making process but with no state-capability and therefore unable to design and, in particular, implement reforms. Leadership is usually provided by the relevant line ministries or, in the case of reforms that involve multiple sectors, a higher office such as the office of the prime ministry or presi- dency. The responsible minister needs to make choices about priority reforms and invest sufficient time understanding the issues, communicating, and negotiating with different stakeholders. Having in a place a full-time team dedicated to coordinate the reform process and conduct part of the analytical work is also a must. In some cases, the management of the reform process and policy analysis can be delegated to an external, independent, commission managed by a recognized expert on the reform subject (see below). Different stakeholders and social partners need to be involved in the reform process from the beginning. Policy reforms that are planned behind “close-doors” have very little chance of surviving the political process and, if they do, very little chance of been implemented.90 The reform process will take longer as a result, but it will be more sustainable if the rationale for reforms, expected outcomes, fiscal costs, and implementation arrangements are widely discussed. An independent technical commission in charge of the designing of the reform program and preparing legislation can be an effective way to achieve political independence and fill-the gap when institutional capacity in the line ministries is weak. Chile is a country that has relied on technical commissions to prepare and implement various reforms, including the last reform to the minimum pension and the unemployment insurance system. The technical commission needs to be presided by somebody with the technical credentials, seniority, and communica- tion skills needed to lead the reform process, and who enjoys political independence. While small, the technical commission needs to be staffed to execute different functions including: 1) research and policy analysis; 2) technical consultations; 2) communications; 3) legal work; and 4) social dialog and political negotiations. 89 Schmidt-Hebbel, K (2009). Chile’s growth and development: leadership, policy-making process, policies, and results. Commission on growth and development working paper; no. 52. Washington, DC: World Bank; and Scartascini, CG and Stein, EH and Tommasi, M (2008). Political Institutions, State Capabilities and Public Policy: International Evidence. Inter-American Development Bank Research Department Working Paper No. 661. Washington DC: Inter-American Development Bank. 90 In 2005, for instance, the Minister of Finance of Egypt was able to develop and get approved by Parliament, in record time, a law to reform the pension system for private sector workers. There were not consultations with social partners or implementing agencies; only the Minister owned the reform. The Law was never implemented and, when the regime changed, it was forgotten. Organizations like the International Labor Organization (ILO) can facilitate social dialog around the reform of social protection systems. 100 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH TABLE 21  State Capability for Selected Countries Index of State Capability Examples of Countries Strong: SC > 6.5 Chile, Singapore, Korea, Qatar Middle: 4 < SC < 6.5 Colombia, Egypt, Lebanon, Malaysia, Peru, Tunisia, Vietnam, Weak: 2.5 < SC < 4 Bolivia, Ecuador, Kenya, Madagascar, Senegal, Uganda, Insufficient: SC < 2.5 Côte d’Ivoire, Iraq, Yemen, Zimbabwe Source: Based Adapted fromon Andrews et al., (2017) However, beyond policy analysis and social dialogue, it is important to recognize that reforming the social protection systems of any given country is a highly-complex, multi-year effort, that requires strong institutional capacity. Some of the activities involved include: drafting and approving new legislation; deploying public information campaigns; setting up or reforming the institutions in charge of managing the new programs; registering new plan members; setting up new management and information systems to process payments and collect contributions; defining and implementing new procurement and payments systems to outsource the provision of services based on performance; and setting up robust monitoring and evaluation systems. Indicators of state-capability suggest that Côte d’Ivoire would face considerable challenges when designing and implementing reforms of the social protection system, which it can over- come by building on steps it has taken to strengthen institutions since the end of its conflict in 2011. A recent study uses three indicators to measure state-capability.91 First, the Quality of Government (QOG) index that is a simple average of three indicators from them International Country Risk Guide: 1) corruption; law and order; and bureaucratic quality (0–4). Second, the public services indicator from the Failed State Index. This indicator rates countries on carry- ing out core state functions like infrastructure, sanitation, education, and health. Finally, three indicators from the World Governance Indicators: 1) government effectiveness; 2) control of corruption; and 3) rule of law. Re-scaling all these indicators in a 0 to 10 range and taking an average gives a measure of state capability. The average for Côte d’Ivoire is below 2.5 which is among the lowest levels of state capability observed in the world (see Table 21). A low, average, level of state-capability does not imply that a country should forgo structural reforms; on the contrary, probably in these environments these reforms are of even more importance. State capability, in fact, is built by solving problems. And, there can be important variations in capability across institutions within a country. Reforms that take place through certain organizations and specific sectors can help build overall state capability. What is important, however, is to align the pace of reform with existing capabilities and to gradually fill existing gaps. A mistake often made is to overload implementing agencies with new functions 91 Andrews M, Pritchett L and Woolcock M (2017). Building State Capability: Evidence, Analysis, Action. Oxford: Oxford University Press, 2017. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 101 FIGURE 25  Capabilities and Reform Paths Source: World Bank staff compilation. and tasks (“best practices”) that they are not ready to perform. For a reform process to gain normative traction within the institution it is necessary that de jure and de facto practices con- verge over time. This implies having realistic expectations about what can be accomplished within a given period of time. As illustrated in Figure 25, forcing the implementation of the jure “best practices” without the necessary capabilities can create stress and lead to failure. In terms of filling capability gaps, it is useful to understand what are the tasks/functions that need to be accomplished to design and implement a successful reform. Different tasks/ functions require different types of capabilities (see Table 22). In the case of the social protec- tion system, many of the functions involve policy analysis/policy making (e.g., designing the parameters of the new pension system or costing the health package), logistics (e.g., enrolling plan members and paying benefits), and the enforcement of obligations (e.g., collecting con- tributions). These are tasks that require analytical skills, that are transaction intensive, and that require the mastery of existing technologies (e.g., payment systems, biometric identification, or proxy means test). In Côte d’Ivoire, top managers in the relevant line ministries have the neces- sary skills for policy analysis and policy making and, if resources are available, can strengthen these skills through new recruitments and training. As discussed above, setting up an indepen- dent technical commission with local and foreign experts can also help. Other challenges exist regarding the capabilities of implementing agencies (e.g., pension funds) to carry out reforms, enforce regulations and manage the necessary logistics. Indeed, human resources are poorly trained, existing information systems are outdated, and there is limited capacity to manage change. At the same time, the status quo is not an option. The reforms of the social insurance programs can be an opportunity to adopt new information and communication technologies, upgrade management systems, and build human capital. In fact, 102 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH some of the new technologies can simplify existing processes such as the collection of contri- butions and the calculation and payment of benefits and reduce demands in terms of human capital (see next section). Côte d’Ivoire has also built capacity designing new administrative and information systems through the implementation of new cash-transfer program. The biggest challenges likely relate to broader service delivery, in particular, health services and active labor market programs. Realistically, even if the reforms discussed in the previous section are adopted, the expansion of coverage will need to proceed only gradually so that the “supply-side” has the time to adapt. Given fiscal space, the expansion of the basic income and pension programs could proceed at a faster pace because the capabilities needed are mainly transactional and new ICTs can help. For health insurance, and active labor market pro- grams (and therefore unemployment insurance) things are more complex because the neces- sary infrastructure needs to be in place, and because the agents in charge of service delivery need to have the right technical and socio-emotional skills, and the right institutional incentives to respond to the needs of plan members. In the case of active labor market programs, some of the reforms are about simplifying existing systems and outsourcing functions that the gov- ernment cannot implement fully, but this also takes time. In all cases, the successful implementation of given reform program requires accountability for results and therefore having the capacity to monitor and measure these results. Achieving proper accountability within public institutions—both among managers and front- line staff—is not easy. It often involves changes in human resources policies (e.g., performance assessment and promotion/remuneration policies) within the public administration. One place to start, however, is to setup robust M&E systems to track the relevant performance indica- tors. In the case of the pension system, for instance, the share of the labor force covered, the collection rate, or the time it takes to process new pensions and pay benefits. In the case of active labor market programs where there is room to outsource the provision of services to private providers, these M&E systems also play a key role in the design of payment systems based on performance. Conclusion Within the framework of the NSPS, Côte d’Ivoire has started reforms to strengthen governance, financing and delivery of its SPL system. New delivery tools have been under- way since 2016 to foster coordination. Key delivery tools being used that can be generalized across programs include a robust, unified identification system, harmonized targeting methods (with supplemental criteria tailored to different programs where needed), a national social reg- istry of households, e-payment systems and benefit tracking, and integrated program manage- ment information systems (Figure 26). Under the national productive social safety net program, inter-agency harmonization of targeting methods has also started. Targeting methods adopted by the national universal health coverage scheme as of 2017 will help to maximize the effectiveness of health insurance subsidies. In terms of identification, under the universal health care coverage initiative and national identification initiatives, the Government plans to develop a national identification CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 103 FIGURE 26  Delivery system tools to maximize the effectiveness of SPL expenditures Social Registry Unique Identification Exit and Follow-Up Systems Monitoring and Targeting Evaluation Benefits and Services Registration Delivery Source: World Bank staff. scheme for social protection and labor programs, an initial step towards harmonization. The main programs to be harmonized in terms of information systems, identification of beneficia- ries, targeting and financing strategy include social safety nets, pensions, universal health care coverage (in coordination with the Ministry of Health), educational grants (with the Ministry of Education), productive inclusion and labor programs for vulnerable groups and youth. As the delivery platform improves, new avenues are expected to open to better track, target and tailor benefit levels within the SPL system (Figure 27). While contributory social security regimes currently account for over 90 percent of public expenditure, this system only sup- ports up to 10 percent of the population, although a small fraction of this population remains under the poverty line as well. How to gradually improve revenue and earnings through non-contributory income support and livelihoods-enhancing spending for the vast majority of Ivorians, the “90 percent”, who are the working poor and in the informal sector also features within a long-term plan for strengthening coverage of social security schemes. Over the short- to mid-term, the need to create the fiscal space for SPL to effectively support this 90 percent remains a challenge, which this policy note has addressed by identifying options for reform. Côte d’Ivoire benefits from a long-standing public administration at the central and regional levels, although regional and local spending appears to be limited in half of the regions based on the distribution of social centers and nascent employment outreach. 104 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH FIGURE 27  Integrating social protection and labor policies across household income categories • Strengthen sustainable contributory social protection Highest income >80% of population • Reinforce productivity, working conditions, formal sector • Strengthen informal/formal sector High income 60–80% productivity • Adjust profiling mechanisms • Support quasi-contributory programs Middle income 40–60% • Facilitate exit from non-contributory programs • Subsidize productive non-contributory Low income 20–40% programs • Support non-contributory programs, Lowest income <20% of population basic living conditions, productive employment entry Source: World Bank staff. At the same time, extending and improving upon this foundation of de-concentrated services bears promise for future investments to expand coverage and deliver services, including livelihood-enhancing measures and linkages with value chains at the local level. Moving forward, broad dialogue will be needed to assess options for improving fiscal space over different time horizons. The choice of reforms depends on the specific social and labor objectives sought, and the different benefits packages and costs envisaged for different popu- lations. Depending on various growth and poverty-reduction scenarios, domestic and external resource needs can be examined. As part of mid-term expenditure framework preparation, the relative cost-effectiveness of other pro-poor instruments vis-à-vis targeted SPL interven- tions can be assessed. An assessment of the mix of different SPL instruments will also help to determine the evolution of Côte d’Ivoire’s SPL system over the long-term to promote inclusive growth. Ultimately, together with economic growth strategies, modernizing Côte d’Ivoire’s SPL system can help support this agenda directly by catalyzing economic inclusion for the majority of Ivorians, equitably and at scale. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 105 REFERENCES Akbaş, M., Ariely, D., Robalino, D. A., & Weber, M. (2016). How to help the poor to save a bit: Evidence from a field experiment in Kenya. Anderson-Macdonald, S.; Chandy, R. & Zia, B., The impact of marketing (versus finance) skills on firm performance: evidence from a randomized controlled trial in south Africa Mimeo. London Business School, 2014 Andrews M, Pritchett L and Woolcock M (2017). Building State Capability: Evidence, Analysis, Action. Oxford: Oxford University Press, 2017. Aterido, R. and M. Hallward-Driemeier (2018) “The Elusive Search for Gazelles: The definition for success affects the firms chosen but not the reality that very few firms maintain high performance for long” (Forthcoming) Attanasio, O., Barr, A., Cardenas, J. C., Genicot, G., & Meghir, C. (2012). Risk pooling, risk preferences, and social networks. American Economic Journal: Applied Economics, 4(2), 134–67 Augsburg B, De Haas R, Harmgart H, Meghir C. 2012. Banerjee, A.; Duflo, E.; Goldberg, N.; Karlan, D.; Osei, R.; Parienté, W.; Shapiro, J.; Thuysbaert, B. & Udry, C. A multifaceted program causes lasting progress for the very poor: Evidence from six countries Science, American Association for the Advancement of Science, 2015, 348, 1260799 Berge, L. I. O.; Bjorvatn, K. & Tungodden, B. Human and financial capital for microenterprise development: Evidence from a field and lab experiment Management Science, INFORMS, 2014, 61, 707–722 Bertrand, M., Crépon, B., Marguerie, A., & Premand, P. (2017). Contemporaneous and Post- Program Impacts of a Public Works Program: Evidence from Côte d’Ivoire. Blanchard, O. J., & Tirole, J. (2008). The joint design of unemployment insurance and employment protection: A first pass. Journal of the European Economic Association, 6(1), 45–77. Blattman, C., Fiala, N., & Martinez, S. (2013). Generating skilled self-employment in developing countries: Experimental evidence from Uganda. The Quarterly Journal of Economics, 129(2), 697–752. Bruhn, M., & Love, I. (2014). The real impact of improved access to finance: Evidence from Mexico. The Journal of Finance, 69(3), 1347–1376. Campos, F.; Frese, M.; Goldstein, M.; Iacovone, L.; Johnson, H. C.; McKenzie, D. & Mensmann, M. Teaching personal initiative beats traditional training in boosting small business in CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 107 West Africa Science, American Association for the Advancement of Science, 2017, 357, 1287–1290 Chemouni, B. (2018). The political path to universal health coverage: power, ideas and community-based health insurance in Rwanda. World Development, 106, 87–98. Cho, Y. & Honorati, M. Entrepreneurship programs in developing countries: A meta regression analysis, Labour Economics, Elsevier, 2014, 28, 110–130 Collins D, Saya U and Kunda T. The Impact of Community-Based Health Insurance on Access to Care and Equity in Rwanda. January, 2016. Management Sciences for Health, Medford, MA, USA. Cotlear, D., Nagpal, S., Smith, O., Tandon, A., & Cortez, R. (2015). Going universal: how 24 developing countries are implementing universal health coverage from the bottom up. The World Bank Crépon, B., Devoto, F., Duflo, E., & Parienté, W. (2015). Estimating the impact of microcredit on those who take it up: Evidence from a randomized experiment in Morocco. American Economic Journal: Applied Economics, 7(1), 123–5 De Mel, S., McKenzie, D., & Woodruff, C. (2008). Returns to capital in microenterprises: evidence from a field experiment. The quarterly journal of Economics, 123(4), 1329–1372. De Mel, S.; McKenzie, D. & Woodruff, C. Business training and female enterprise start-up, growth, and dynamics: Experimental evidence from Sri Lanka Journal of Development Economics, Elsevier, 2014, 106, 199–210 Elgazzar, H. (2009). Income and the use of health care: an empirical study of Egypt and Lebanon. Health Economics, Policy and Law, 4(4), 445–478. Fafchamps, M. & Minten, B. Impact of SMS-based agricultural information on Indian farmers The World Bank Economic Review, Oxford University Press, 2012, 26, 383–414 Fafchamps, M.; McKenzie, D.; Quinn, S. & Woodruff, C., Microenterprise growth and the flypaper effect: Evidence from a randomized experiment in Ghana, Journal of development Economics, Elsevier, 2014, 106, 211–226 Giné, X. & Mansuri, G. Money or ideas? A field experiment on constraints to entrepreneurship in rural Pakistan World Bank Policy Research Working Paper, 2014 Glaub, M.; Frese, M.; Fischer, S. & Hoppe, M. A psychological personal Initiative training enhances business success of African business owners Mimeo. National University of Singapore Business School, Mimeo. National University of Singapore Business School, 2012 Goswami, A. G., Medvedev, D., & Olafsen, E. (2019). High-Growth Firms: Facts, Fiction, and Policy Options for Emerging Economies. World Bank Publications. Gupta, N. D., Larsen, M., & Thomsen, L. S. (2015). Do wage subsidies for disabled workers reduce their non-employment?-evidence from the Danish Flexjob scheme. IZA Journal of Labor Policy, 4(1), 10. Hakeenah, Njenga, (2019). NSSF partners with M-Pesa in cashless drive. The Exchange [online] Available at: https://www.exchange.co.tz/nssf-safaricom-contributions-pension-efficiency- m-pesa/ [Accessed 7 Mar. 2019]. 108 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH Higuchi, Y.; Mhede, E. & Sonobe, T., Short-and Longer-Run Impacts of Management Training: An Experiment in Tanzania, Mimeo. National Graduate Institute for Policy Studies, Tokyo, Mimeo. National Graduate Institute for Policy Studies, Tokyo, 2016 Hughes N., Lonie S. “M-PESA: Mobile Money for the ‘Unbanked’ Turning Cellphones into 24-Hour Tellers in Kenya”. mitpressjournals.org. MIT Press. PDF. (2007). Jack W., Suri T. “Mobile money: The Economics of M-PESA”. microlinks.org. The National Bureau of Economic Research. PDF. (2011). Kalisa I, Musange S, Collins D, Saya U and Kunda T. The Development of Community-Based Health Insurance in Rwanda - Experiences and Lessons. July, 2015. University of Rwanda College of Medicine and Health Sciences - School of Public Health, Kigali, Rwanda and Management Sciences for Health, Medford, MA, USA . Kaplan, D. S., Sadka, J., & Silva-Mendez, J. L. (2007). Litigation and settlement: New evidence from labor courts in Mexico. The World Bank. Klinger, B., & Schündeln, M. (2011). Can entrepreneurial activity be taught? Quasi-experimental evidence from Central America. World Development, 39(9), 1592–1610. Kluve, J., Puerto, S., Robalino, D., Romero, J. M., Rother, F., Stöterau, J., ... & Witte, M. (2019). Do youth employment programs improve labor market outcomes? A quantitative review. World Development, 114, 237–253. Krishnan, Sudha Bala; Karlen, Raphaela Beatrice; Peterburs, Teresa Anna Maria; Tokle, Siv Elin. 2017. Monitoring and evaluation of jobs operations (English). Jobs M&E Toolkit; no. 1. Washington, D.C. : World Bank Group. Kuddo, A., Robalino, D., & Weber, M. (2015). Balancing regulations to promote jobs. The World Bank. Lafortune, J.; Riutort, J. & Tessada, J., Role models or individual consulting: The impact of personalizing micro-entrepreneurship training, American Economic Journal: Applied Economics, 2017 Christiansen L and Premand P. 2018. “Jobs Diagnostic for CÔTE D’IVOIRE: Employment Productivity and Inclusion for Poverty Reducton.” Jobs Series. Issue No. 2. World Bank. Macours, K., Premand, P., & Vakis, R. (2012). Transfers, diversification and household risk strategies: experimental evidence with lessons for climate change adaptation. The World Bank. Microfinance at the margin: experimental evidence from Bosnia and Herzegovina. Work. Pap. 146, Off. Chief Econ., Eur. Bank Reconstr. Dev., London Miriri, Duncan, (2018). Kenya’s Safaricom takes M-Pesa global with Western Union. Reuters Wire [online] Available at: https://af.reuters.com/article/investingNews/idAFKCN1NB1GA- OZABS?feedType=RSS&feedName=investingNews [Accessed 7 Mar. 2019]. Morawczynski O. “Exploring the Usage and Impact of ‘Transformational’ Mobile Financial Services: The Case of M-PESA in Kenya”. tandfonline.com. Journal of Eastern African Studies. (2009). Packard et al., Forthcoming 2019. “Protecting All - Risk-Sharing for a Diverse and Diversifying World of Work”. Social Protection and Jobs Global Practice White Paper. The World Bank. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 109 Ravallion, M., Jolliffe, D., & Margitic, J. (2018). Social Protection and Economic Development: Are the Poorest Being Lifted-Up or Left-Behind? (No. w24665). National Bureau of Economic Research. Republic of Côte d’Ivoire (2014). National Social Protection Strategy, 2016-2020. Republic of Côte d’Ivoire (2014). National Employment Plan, 2016–2020. Republic of Côte d’Ivoire (2015). National Poverty Profile. Republic of Côte d’Ivoire (2016). Bref Aperçu des Enjeux du Système de la Retraite. Republic of Côte d’Ivoire and International Labor Organization (Ngom D and Kouadio B) (2016). Evaluation de la Politique nationale de l’emploi, 2012–2015. Robalino D, Walker I, and Romero M. Maximizing the jobs development pay-offs of business plan competitions: results from an economic simulation. Forthcoming. Robalino, D. A., & Weber, M. (2013). Designing and implementing unemployment benefit systems in middle- and low-income countries: beyond risk-pooling vs savings. IZA Journal of Labor Policy, 2(1), 12. Safaricom Info Center. “M-PESA Rates”. [online] Available at: safaricom.co.ke. Safaricom. (2017). Scartascini, CG and Stein, EH and Tommasi, M (2008). Political Institutions, State Capabilities and Public Policy: International Evidence. Inter-American Development Bank Research Department Working Paper No. 661. Washington DC: Inter-American Development Bank. Schmidt-Hebbel, K (2009). Chile’s growth and development: leadership, policy-making process, policies, and results. Commission on growth and development working paper; no. 52. Washington, DC: World Bank. Solange, H. (August 2017). Community-Based Health Insurance (CBHI) Presentation. Rwanda Social Security Board. Unicef (2012). Côte d’Ivoire: Cash Transfer Feasibility Study for the National Social Protection Strategy. Unicef (2012) Côte d’Ivoire: Situational Analysis for the National Social Protection Strategy. Valdivia, M.,Business training plus for female entrepreneurship? Evidence from a field experiment in Peru Unpublished manuscript GRADE, Lima, December 2013 Vodafone. (2019). What is MPESA. [online] Available at: https://www.vodafone.com/content/ index/what/m-pesa.html [Accessed 7 Mar. 2019]. World Bank (2014). Impact of Crises on Human Development in Côte d’Ivoire. Internal Note unpublished. World Bank (2015). Côte d’Ivoire: Systematic Country Diagnostic. Washington DC: World Bank. World Bank (2015). Côte d’Ivoire Poverty and Targeting Diagnostic Report. Internal Note unpublished. World Bank (2015). Côte d’Ivoire Productive Social Safety Net Project Appraisal Document. Washington DC: World Bank. World Bank (2019). Côte d’Ivoire Social Protection and Economic Inclusion Project Appraisal Document. Washington DC: World Bank. 110 CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH World Bank (2019). Pensions Watch, Social Pensions Database. www.pension-watch.net/social- pensions-database/social-pensions-database--/. Accessed 3 Mar 2019 Yoonyoung Cho & David Robalino & Samantha Watson, 2016. “Supporting self-employment and small-scale entrepreneurship: potential programs to improve livelihoods for vulnerable workers,” IZA Journal of Labor Policy, Springer; Forschungs institut zur Zukunft der Arbeit GmbH (IZA), vol. 5(1), pages 1-26, December. CÔTE D’IVOIRE  |  MODERNIZING SOCIAL PROTECTION AND LABOR POLICY FOR INCLUSIVE GROWTH 111 Photo Credits: Universal Images Group North America LLC / DeAgostini / Alamy Stock Photo (cover); Eric Lafforgue / Alamy Stock Photo (p. viii); Godong / Alamy Stock Photo (p. xx); Mauritius Images GmbH / Alamy Stock Photo (p. 1); Media Minds / Alamy Stock Photo (p. 8); Charles O. Cecil / Alamy Stock Photo (p. 10); Jake Lyell / Alamy Stock Photo (p. 11); Godong / Alamy Stock Photo (p. 24); André Quillien / Alamy Stock Photo (p. 25); Film Filmryan / Alamy Stock Photo (p. 30); Mauritius Images GmbH / Alamy Stock Photo (p. 31); Eric Lafforgue / Alamy Stock Photo (p. 38); Olivier Asselin / Alamy Stock Photo (p. 39); Media Minds / Alamy Stock Photo (p. 48); iStock.com / aroundtheworld.photography (p. 51); Godong / Alamy Stock Photo (p. 72); Jake Lyell / Alamy Stock Photo (p. 73); Tommy Trenchard / Alamy Stock Photo (p. 76); robertharding / Alamy Stock Photo (p. 77); Jake Lyell / Alamy Stock Photo (p. 88); Michael Dwyer / Alamy Stock Photo (p. 90); Eric Lafforgue / Alamy Stock Photo (p. 91).