Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Stronger Social Protection and Labor Systems in Central Ameri- ca for a Resilient and Inclusive Recovery 2 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Acknowledgments 3 Acknowledgments This report was prepared by a core team led by Aylin Isik-Dikmelik (Senior Economist, Social Protec- tion and Jobs Global Practice), and comprising Ana Sofia Martinez, Ines Rodriguez, Nancy Banegas, Asta Zviniene, and Polly Jones, with excellent micro-data analysis and other contributions by Yanina Eliana Domenella, and research assistance by Roberto Lagos on social protection responses. Veronica Michel kindly helped to update analysis for the Honduras Jobs Diagnostics. The report was carried out under the guidance of Michel Kerf and Pablo Gottret. The team received valuable advice and guidance from Rita Almeida, as well as from the peer reviewers at different stag- es, Margaret Grosh, Pablo Acosta, Melis Guven, and Cornelius Fleischhaker. The team is also grate- ful to Social Protection and Jobs country teams in Central America: Clemente Avila, Gaston Blanco, Hugo Brousset, Ursula Martinez, Miriam Montenegro, Luz Stella Rodriguez, Maria Concepcion Steta, and Julieta Trias, for their helpful feedback and contributions. In addition, the team received valuable feedback from Ana Veronica Lopez and William Wiseman. Administrative support was provided by Lily Franchini and Alejandro Espinosa (Sonideas) has designed the cover and typeset the document. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 4 Contents Contents Acknowledgments   3 Executive Summary   8 I Why focus on Social Protection and Labor Systems   15 II Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19   19 A1. Despite progress, the effectiveness of SP spending and programs was limited before COVID-19   20 A2. Limited reach and potential mitigation impact of existing SP programs against the unprecedented COVID-19 pandemic   25 B1. Limited adaptiveness of SP and delivery systems   32 B2. Obstacles to Leverage SP Delivery Systems to Respond to COVID-19   36 C1. Structural issues Limiting Good Employment Outcomes for workers and the economy   38 C2. The COVID-19 pandemic exacerbated LM challenges   46 D1. Response of CA governments in SPL to COVID-19 emergency   55 III The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy   63 A1. Improve efficiency and effectiveness of SP spending and ensure right mix of programs to protect and build resilience of the poor and (newly) vulnerable   65 B1. Improve adaptiveness of the SPL system and further invest in delivery systems for increased preparedness in light of more frequent shocks   68 C1. Strengthen Labor Market Institutions and Services to Support People Back to (Productive) Work   71 Annexes 77 References 84 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Contents 5 Figures Figure ES1. Overview of Policy Directions    12 Figure 1. Total Social Assistance Expenditure (% GDP), Latin America and the Caribbean    21 Figure 2. Evolution of Composition of Social Assistance Expenditure in El Salvador, share of GDP, 2014-2019    21 Figure 3. Coverage of Social Assistance, Central America and LCR, by quintiles    22 Figure 4. Distribution of beneficiaries (beneficiary incidence), all SA, Central America and LCR, by quintiles    22 Figure 5. Distribution of beneficiaries (beneficiary incidence), El Salvador, 2019, by type of SA program, by quintiles    23 Figure 6. Distribution of beneficiaries (beneficiary incidence/progressivity) CCTs, Central America and LCR, by quintiles   24 Figure 7. Coverage Conditional Cash Transfer Programs, Central America and LCR, by quintiles    24 Figure 8. International poverty rate ($1.9 in 2011 PPP) from 2018 to 2023 (estimates) in Central America    25 Figure 9. Fiscal balance deficit as % of GDP for Central America and LCR average    26 Figure Box 1. Percentage of establishments with lower monthly sales compared to one year before the interview   27 Figure Box 2. GDP growth from 2019 to 2024 (estimates) in Central America and LCR    28 Figure Box 3. Percentage of households that ran out of food due to lack of money or other resources (last 30 days)    28 Figure Box 4. Education activities engaged in by households with children    29 Figure Box 5. Households with children that reported lack of internet as main deterrent to participate in school (among those not participating)    30 Figure Box 6. Percentage of household members who needed but were not able to access medical treatment    30 Figure Box 7. Reasons households could not receive medical attention when needed.    31 Figure Box 8. Contributors as percent of working age population (ages 15-64)    33 Figure Box 9. Poverty rates at $5.5/day, overall population vs. population over the age of 65.    34 Figure Box 10. Pension contribution rates assessed on insured wage income in 2019.    34 Figure 10. Coverage of social registries in selected countries, share of population    35 Figure 11. Adults who own an account in the financial system    36 Figure 12. Use of payment digital payment methods    36 Figure 13. Structure of Employment in Central America and Dominican Republic, by main sectors and by informal vs. formal    39 Figure 14. Share of Informal vs Formal Employment in Central America and DR    39 Figure 15. Overall LM Rigidity Index (as compared to global average)    40 Figure 16. Employment, informality, and unemployment in Central American countries for bottom 40% and top 60% of the population    41 Figure 17. Share of wage and unpaid workers in bottom 40 and top 60 percent of the population for Central American countries   41 Figure 18. Employment rate by education level in Central American countries    42 Figure Box 11. Labor Force Participation in Central America, by gender    44 Figure Box 12. Labor force participation in Central America, by education level and by gender    45 Figure Box 13. Unemployment rate in Central America, by gender    45 Figure Box 14. Youth NEET rate (rhs) and gender composition, Central America    46 Figure 20. Share of people who stopped working across the sub-region, total    47 Figure 19. Percentage of households that report a reduction of total income during quarantine    47 Figure 21. Share of people who stopped working across the sub-region, by sector    48 Figure 22. Share of people who stopped working across the sub-region, by employment type    48 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 6 Contents Figure 23. Percentage of people who permanently lost their employment during COVID-19 quarantine    49 Figure 24. GDP reduction and percentage of people who lost their jobs    49 Figure 25. Percentage of People (18+) that report losing their employment, total and by gender    50 Figure Box 15. Employment loss, by sector, compared to Q1 2020    51 Figure Box 16. Employment loss at detailed sector (1 digit ISIC) level, compared to Q1 2020    53 Figure 26. Structure of employment pre-COVID-19, by potential vulnerability to COVID-19 restrictions, by main sector   55 Figure Box 17. Share of employment women and men in service sector activities affected with mobility restrictions    57 Figure Box 18. Informality rate by type of activity    58 Figure Box 19. Coverage of main programs by quintiles    59 Figure Box 20. Beneficiaries distribution of main programs by quintiles    59 Figure 27. Overview of Policy Directions    64 Figure A.1. Percentage of households that ran out of food due to lack of money or other resources (last 30 days)    77 Figure A.2. Composition of SPL Responses by components and by regions    77 Figure A.3. Per capita social spending by sector Central America and LCR    78 Figure A.4. Social Spending by sector as a share of GDP, Central America    78 Figure A.5. Share of CCT transfers in beneficiaries’ welfare (adequacy), Central America and LCR, in Q1    78 Figure A.6. Share of establishments in or expect to fall in arrears in next 6 months    79 Figure A.7. Share of establishments that fired workers in the last 30 days    79 Figure A.8. Contribution of change in productivity (value added per worker), by country and major sector    79 Figure A.9. Difficulty of Hiring Index (as compared to global average)    80 Figure A.10. Rigidity of Hours Index (as compared to global average)    80 Figure A.11. Difficulty of Dismissal Index (as compared to global average)    80 Figure A.12. Labor force participation by age group and by Central American countries    81 Figure A.13. Informality by age group and by Central American countries    81 Figure A.14. Labor Force Participation Rate in Costa Rica, by youngest child in the household and by gender    81 Figure A.15. Probability of Employment in El Salvador, by youngest child in the household and by gender    82 Figure A.16. GDP growth in the agriculture sector from 2019 to 2024 (forecast)    82 Figure A.17. GDP growth in the industry sector from 2019 to 2024 (forecast)    82 Figure A.18. GDP growth in the service sector from 2019 to 2024 (forecast)    83 Boxes Box A. Broad Impacts of COVID-19 on households and human capital 27 Box B. Brief Panorama of Pension systems in Central America 33 Box C. Overview of Women’s LM Outcomes in Central America 44 Box D. Unequal LM Impacts of COVID and early recovery in Costa Rica 51 Box E. Zooming in on Panama: Main Elements of SP Response 57 Tables Table Box 1. Employment loss at detailed sector (1 digit ISIC) level, compared to Q1 2020    52 Table 1. Summary of SPL instruments utilized in response to COVID-19    55 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 7 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 8 Executive Summary Executive Summary Social Protection and Labor (SPL) Systems help individuals and societies manage risk and volatility and protect them from poverty through instruments that address the challenges of resilience, equity and opportunity. SPL systems include social safety nets, social insurance, and labor market programs. As recent events have shown, the relative emphasis among goals – resilience, equity and opportunity - can change over time, with demands put on SPL program design and delivery systems differing in each context. In relatively stable times, programs are likely to focus on human capital formation, equality of opportunity, poverty reduction, and redistribution. This was the case in Central America prior to the COVID-19 emergency, albeit with some shortcomings. The goal of shock-responsiveness (resilience) dramatically came to the fore during the pandemic, even if recognized earlier during natural disasters and now more broadly with climate change. Globally and in Central America, SPL systems had a criti- cal role in the response to the COVID-19 emergency. The nature and intensity of the unprecedented COVID-19 pandemic stretched SPL systems in Central America (and globally) and brought to the surface some gaps and new priorities. While no country could be fully prepared for the emergency, Central American SPL systems faced persistent challenges, to varying degrees, even in the relatively stable times that pre-date the pandemic (for example, underfunded poverty-targeted programs, lack of dynamic social registries and limited use of digital payments). These issues became obstacles to effectively leverage SPL systems during the pandemic. Furthermore, the nature of the crisis underscored the coverage gap for the newly vulnerable, which normally fall outside both safety nets and formal systems, what could be called the “missed middle”. This both compelled an expanded definition of vulnerability and validated the importance of improving the adaptiveness of SPL systems. Strengthening SPL systems is a priority investment both to lay a foundation for stronger inclusive growth in the sub-region and to build resilience, in light of increased shocks, including those related to climate change. Adaptive and effective SPL systems, which preserve, restore, and aim to augment human capital, are as important for inclusive recovery as structural policy reforms. SPL systems can also be leveraged for climate change adaption and mitigation. This report presents an overview of the challenges of SPL systems before and in the aftermath of COVID-19, with the main objective to provide directions for policy reforms.1 The focus is on social assistance and labor programs, while highlighting some links to pension systems. While recognizing that most countries face fiscal constraints (wors- ened by COVID-19), a clear finding is that investments are needed in the sub-region to improve SPL delivery systems and increase program resources in order to strengthen sector coverage and perfor- mance, including adaptability to crisis. 1 Since an in-depth analysis of pensions systems in Central America is planned as part of future analytical work, it falls outside of the scope of this report. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Executive Summary 9 The basic SPL pillars reflect varying levels of development in the sub-region. Social insurance (both contributory and non-contributory) exists in most countries but is underdeveloped: with the exception of Costa Rica and Panama, coverage of the working age population is below the LCR average. Unem- ployment insurance is absent. A range of social assistance programs constitutes the bulk of programs in most cases, though not necessarily the spending. Labor market programs and services are a very small part of the SPL system in most countries. While the level of development and challenges vary widely across and within Central America depending on which part of the SPL system is considered, there are three common themes. These are: (i) limited effectiveness of SP spending and programs; (ii) limited adaptiveness of SP and delivery systems; and (iii) structural issues limiting good employment outcomes for workers and the economy. The report analyzes these challenges both before and during the response to the COVID-19 pandemic. Effectiveness of SP spending and programs pre-COVID-19 Countries in Central America have made significant progress towards a more comprehensive social assistance system. Social safety nets covering different segments of the population are present in all Central American countries. These include Conditional Cash Transfer (CCT) Programs in most coun- tries (except Nicaragua) which aim to promoting human capital development and break the intergen- erational cycle of poverty. Most countries in the sub-region spend little on social assistance and mainly on broad categorical rather than poverty-targeted programs, leading to high coverage of both rich and poor. Most Central American countries spend less than the LCR average on social assistance. El Salvador and Costa Rica are exceptions; Honduras and Guatemala spend the least. In spite of limited resources, broad categor- ical programs outweigh other social assistance spending, in particular under-funded poverty-targeted programs. Coverage of social assistance in the sub-region of both the poorest and richest quintile is generally higher than the LCR average driven mainly by categorical programs. This spending pattern means that most countries are facing a real dilemma among the goals of fiscal balance, poverty relief coupled with human capital accumulation for the poor, and the wide distribution of benefits. Social insurance issues (including an expensive contributory system with high transition costs) only exacer- bate the dilemma. Overall social assistance in Central America is generally less progressive compared to the LCR aver- age; in most countries conditional cash transfer (CCT) programs are an exception, but their potential is handicapped by low coverage and benefits. Within the sub-region, even in countries where overall social assistance is most progressive (Costa Rica and Honduras), slightly less than one third of social assistance beneficiaries belong to the poorest quintile, compared to the LCR average of slightly less than half. There are significant differences across programs; for instance, CCT programs are well-tar- geted. With the exception of Nicaragua (without a CCT), the Dominican Republic, and Guatemala, CCT programs in the subregion are well-targeted (measured as share of total CCT beneficiaries that are from the first quintile), matching or exceeding the LCR average of 54 percent. However, CCT coverage (i.e., proportion of individuals in the bottom quintile that benefit from CCT program) is low. Only the Dominican Republic and Costa Rica come close to the LCR average. Similarly, with the exception of Honduras, CCT benefit levels are less adequate than the LCR average. Thus, despite good targeting, the low coverage along with inadequate benefits of CCTs limit their potential impact to reduce pov- erty; and improve human capital outcomes which are relatively low in all countries except Costa Rica. Limited reach of SP programs against the unprecedented COVID-19 pandemic The COVID-19 pandemic threatens to erode hard-won past gains in human capital and poverty re- duction. Estimates for 2020 suggest that poverty worsened in all Central American countries with the highest increases expected for Panama (50 percent) and Costa Rica (40 percent). Initial evidence Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 10 Executive Summary points to a disproportionate impact on the poor and vulnerable, further widening gaps in human capi- tal. The economic consequences of the pandemic caused widespread job and income losses, declines in food security, and constrained access to education and healthcare. Simulations of the impact of COVID-19 on human capital point to decreases in HCI (Human Capital Index) in the sub-region, with upper middle-income Costa Rica and Panama the most affected. Most countries in the sub-region acted swiftly to contain the impacts of COVID-19, however, the associated costs added to fiscal pressure. The size of emergency packages ranged from around 1.5 percent of GDP (Costa Rica) to 3.4 percent of GDP (Guatemala), except El Salvador where the cost reached 15.5% of GDP, putting most countries in a vulnerable fiscal situation. The sharp decline in employment rates as well as an increase in new retirees also put pressure on pension schemes in the sub-region. Even countries, such as Costa Rica, which were not confronting fiscal sustainability issues, have suddenly been caught in unexpected problems. Limited adaptiveness of SP and delivery systems Central American countries invested in their SP delivery systems over the past decade. All countries have made progress to integrate functions across programs, including through the development of social information systems and common payment platforms. Social registries support the processes of intake, registration and assessment, with coverage ranging from 70 percent or higher in Costa Rica and the Dominican Republic to 9 percent in El Salvador. A social registry is being piloted in Guatemala. However, social registry intake and registration have largely relied on infrequent en masse waves (with the exception of Panama and Costa Rica), so data is often not up to date. Payments in the sub-region include non-account based over-the-counter mechanisms in some places, although some countries, for example Panama, had progressed towards integration and digitization. The limited use of digital payments in most countries was a constraint. Obstacles to Leveraging SP Delivery Systems to Respond to COVID-19 Limited adaptiveness of delivery systems in Central America adversely affected the quality and time- liness of the social protection response to the pandemic, although there was an impetus for pay- ment digitization. Obstacles prevented the leveraging of social safety net programs to protect existing beneficiaries and expand coverage to the newly vulnerable: some governments had to create parallel systems to respond swiftly. Depending on the status of the social registry, countries in the sub-region applied many of the same approaches used globally to determine eligibility for emergency transfers: leveraging data from the social registry (Dominican Republic); on-demand applications (Costa Rica); and new data sources such as utility consumption (El Salvador and Guatemala). Cross-checks with administrative databases supplemented the first two approaches. COVID-19 provided an impetus for payment innovations, including increased digitization. In Guatemala, payments of the emergency transfer Bono Familia were digitized through the creation of virtual accounts. In Honduras, the govern- ment launched a comprehensive digital payments mechanism in 2020. Structural Issues Limit Good Employment Outcomes for Workers and the Economy Pre-COVID -19 most jobs in the sub-region were low quality. Most jobs across Central America are in services. While not all of these jobs are low quality, with the exception of Costa Rica and Panama, the majority of service jobs are informal, likely concentrated in low productivity elementary occupa- tions. In nearly all countries a significant share of employment is informal, ranging from 49 percent in Panama to 82 percent in Honduras, leaving large swaths of the employed vulnerable. Costa Rica is the exception with about one third informal employment. Informality has persisted in most countries, driven by multiple barriers for formalization, LM rigidities on a few dimensions, and minimum wage policies. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Executive Summary 11 Access to quality jobs is particularly limited for those in the bottom of the income distribution and those in rural areas. While most jobs are low quality across the sub-region, the incidence is much higher among the bottom 40 percent of the income distribution compared to the top 60 percent in most countries, for example, reaching a difference of 40 percentage points in Panama. In addition, the incidence of unpaid work is higher among the bottom of the income distribution while the re- verse is true for wage employment in all countries, with poor (and likely rural) households stuck in subsistence agriculture or household enterprises, while other (likely urban) households enjoy access to better jobs. Similarly, low-skilled workers, youth and women face worse labor market outcomes in most coun- tries. The low skilled are more likely to have an informal job or be unemployed, although the gap varies widely across countries, with the smallest differences between the high and low educated in the Dominican Republic and Panama. Youth have a lower rate of labor force participation, and even when employed are more likely to be in informal jobs and in lower productivity sectors, which are linked to lower earnings. The share of youth who are neither employed, nor in education or train- ing (NEETs) ranges from slightly under 20 percent in Panama and Costa Rica to nearly 30 percent in Honduras and Guatemala. Women constitute the majority of NEETs. While the gender gaps in LM outcomes have narrowed in almost all countries, persisting disparities are particularly large in some countries. For example, women in Guatemala participate in the labor market at less than half the rate as men. A cross-country analysis suggests that if women were to participate in the labor market at the same rate as men, per capita GDP would range from 16 percent higher in El Salvador (par with the LCR average) to 25 percent higher in Guatemala. Even when women participate in the labor market, they are more likely to be unemployed or informally employed and earn less, despite being more educated in most cases. Care responsibilities appear to be the most binding barrier for women. Despite gains in educational attainment, generally the supply of skills in the labor force in most coun- tries remains low and out of step with jobs of the future and capacity for employment support varies. Although the education level of the labor force in most countries has increased steadily, it remains low. In addition, while higher educational attainment generally leads to higher earnings, returns to education have declined in some countries (El Salvador and Guatemala). All countries in the sub-re- gion have some of the basic elements for employment support, but capacity varies widely. At one end is Costa Rica, which has modernized its main training agency to better fit the changing needs of the labor market. At the other end is Honduras, whose TVET system has the foundational elements, but is out of date with difficulties to provide training for skills in demand and to support youth and first-time job seekers. Other countries face similar issues. With the exception of Costa Rica, coverage of labor market programs is low. The COVID-19 pandemic exacerbated labor market challenges The impact of the pandemic exacerbated the sub-regional issues in the labor market. The COVID-19 shock worked through two main channels in the labor market: (i) confinement measures shuttered large parts of the economy; and (ii) consumer activities decreased to minimize the risk of infection. In general, countries with higher GDP contractions also lost more jobs. The negative impacts were con- centrated on high-contact employment, which tends to be informal, lower skilled with lower earnings, exacerbating existing divisions in the labor market. Many households experienced significant income losses: workers in services and industry were the most affected, as were generally the self-employed. Emerging data confirms the larger impact of the pandemic on youth, lower skilled, and women, all of whom sustained higher job losses, further increasing their vulnerability. Given the limited availability of automatic stabilizers, mechanisms did not exist to reach either formal or informal workers affected by the pandemic. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 12 Executive Summary SPL response of Governments to COVID-19 emergency Governments in the sub-region leveraged existing SPL instruments to varying degrees through the provision of social assistance, social insurance and labor market services to respond to the crisis. All countries (except Nicaragua) deployed cash transfers mainly to reach new beneficiaries, includ- ing informal workers. The scale and duration of benefits varied. The countries with the largest scale cash transfers included Panama eventually reaching almost 80 percent of the population. Likewise, in Guatemala, the coverage of cash transfers expanded from only 5 percent of households to 80 percent following the establishment of the Bono Familia temporary cash transfer program. Social assistance measures in nearly all countries included in-kind support to assist food security and utility fee waivers. Social insurance measures were limited due to narrow coverage and the lack of unemployment in- surance. Labor market measures included access to credits for micro, small and medium enterprises (formal and informal) (El Salvador and Panama) as well as credit guarantees (Costa Rica), wage sub- sidies or similar support for formal workers, (Guatemala, Dominican Republic, Honduras and El Sal- vador) and training (Costa Rica and Panama). In the case of the latter, some initiatives were coupled with social assistance support (Costa Rica). Across the sub-region labor regulations were adapted for remote work. Figure ES1. Overview of Policy Directions Improve e ciency and e ectiveness of social protection spending and Improve e ciency of social Carefully manage the Continue to build and Ensure the right mix of Rebalance the pension ensure right mix assistance spending with a transition away from protect human capital social protection schemes to boost fiscal of programs to shi away from less emergency programs to through adaptive and instruments, in particular sustainability protect and build progressive programs minimize any negative nutrition-sensitive social establishment of automatic resilience of the impacts on the poorest assitance programs, stabilizers, and enhace poor and (newly) incorporating behavioral the adaptiveness of vulnerable elements targeted programs Improve adaptiveness of the Social Protection and improve the Continue Develop social Increase the Strengthen Increase the Intensify the Leverage Improve Labor system and adaptiveness investing in the registries and adoption of grievance integration of use of administrative targeting further invest in of development rely on other digital payment redress social technology data through accuracy by delivery systems the SPL and strengthen- administrative methods and mechanisms information beyond interoperabili- using big data for increased system ing of social databases for account systems payments and ty/integration and machine preparedness in and increase protection deliv- crosschecks to ownership across the with social learning (with light of more preparedness ery systems, assess needs di erent protection caveats) frequent shocks for external to make them and conditions phases of the information shocks more inclusive, and determine delivery chain systems dynamic and eligibillity adaptive (Intro- ducing open and continuous registration) Strengthening labor market institutions and services to support people back to Be er understanding and Promote private sector Integrated labor demand Institutional strenghthening Promote labor market monitoring of labor demand, led/provide training to and supply programs that to improve labor reforms that support the (productive) improve employability and contribute to be er intermediation, supported creation of be er jobs including through use of work real-time information, for quality of matches, employment outcomes for by technology (digitization e ective (re)emplyment increasing the change of vulnerable groups. of employment services, policies long-term success adequate budgets) Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Executive Summary 13 Path forward to strengthen adaptive SPL systems for a more inclusive and resilient recovery – Directions for Policy Pre-existing and emerging challenges, as well as global knowledge, point to 3 policy directions for SPL systems in Central America to support resilient and inclusive recovery and beyond. The policy directions for SPL systems are: (i) improve efficiency and effectiveness of SP spending and ensure the right mix of programs; (ii) improve adaptiveness and invest in delivery systems; and (iii) strengthen employment services and institutions to support a quick return to (productive) jobs. Improve efficiency and effectiveness of SP spending and ensure the right program mix Measures to increase the efficiency of SP spending will differ among countries, but broadly imply a shift from less progressive programs to better targeted programs and/or improving the targeting of large categorical programs, for example by: • Improving the distributional impact of energy subsidies by moving to eligibility based on so- cioeconomic indicators as a first step, and later by streamlining them into the social assis- tance system to take advantage of mechanisms to identify the poor; • In the transition out of emergency programs, balancing the continued protection of the poor- est with reduced fiscal space; and • Rebalancing pension schemes to boost fiscal sustainability. Given the pandemic’s disproportionate impact on the poorest, it will be imperative to continue to build and protect human capital through more effective social assistance programs, namely: • Continue to invest in and ideally expand well-targeted CCTs; and • Incorporate nutrition-sensitive approaches, including behavioral elements, to enhance the proven positive results of progressive social assistance programs, such as CCTs. Successful models exist. A pilot intervention using home visits and group sessions with parents in Gua- temala had positive effects on children’s language and cognitive development. Brazil’s Feliz Crianza combines similar elements with a social assistance program. The mix of social protection instruments in the sub-region needs to reflect the importance of auto- matic stabilizers and the ability of programs to increase resilience and preparedness against future shocks. • Potential automatic stabilizers include progressive social programs with adequate coverage and flexibility to expand as well as unemployment insurance in countries where formality rates are higher even at the bottom of the income distribution (Costa Rica and the Dominican Republic). • The adaptiveness of programs can be improved by incorporating productive inclusion ele- ments as well as triggers and emergency protocols allowing for rapid scalability. Improve adaptiveness of the SPL system and invest in delivery systems for better shock preparedness. The COVID-19 shock reinforced the need to improve the adaptiveness of SPL systems, more so look- ing forward given Central America’s exposure to likely more frequent shocks. It is critical to continue investing in the development and strengthening of delivery systems to make them more inclusive, dynamic and adaptive. Drawing from lessons learned globally and in Central America during the emer- gency response, priorities include: Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 14 Executive Summary • Introduce open and continuous registration  on-demand as well as different registration mo- dalities (in-person at government offices, on-line, and by phone); • Develop dynamic social registries with better coverage  and rely on other administrative databases for cross-checks for validation, among other purposes, through investments in interoperability; ncorporating data from social regis- • Move towards integrated social information systems, i tries and beneficiary registries, among others, ideally with the capacity to monitor not only the demand but also the supply of programs;  ith benefits for government (ef- • Promote digital payment methods and account ownership, w ficiency gains) and recipients (convenience and financial inclusion); and  cross the delivery chain to improve effi- • Intensify the use of technology beyond payments a ciency and reduce processing times, for example, in intake and registration, notification, en- rollment, beneficiary data updates, and grievance redress. Strengthen Labor Market Institutions and Services to Support People Back to (Productive) Work A multi-pronged approach is needed to strengthen employment policies and services, aligned with a shift in the role of government from protecting workers from change to protecting them for change. The approach entails:  ncluding by using real- time data to understand • Improving labor market information systems, i and monitor labor demand, in order to guide re-employment support, particularly training;  combining demand and supply) interventions to sup- • Deploying well-designed and integral ( port in particular young and other vulnerable to get back to work, drawing on the evidence of positive and more sustained impacts, for example, by coupling wage subsidies with training and/or employment services; • Promoting private sector led and provided training to ensure relevant skills and better job matches, implying a change in the traditional TVET approach to training delivery; • Strengthening institutional capacity for intermediation,supported by technology to develop Job Matching Portals, the introduction of profiling to help prioritize resources for jobseekers and better match different participant groups with the appropriate ALMPs, and digitization of services;  struggle for countries in the sub-region; • Ensuring adequate budget to support employment, a and  o support the creation of better jobs, including modernizing • Promoting labor market reforms t LM regulations, aligning the minimum wage with labor productivity growth, and setting mini- mum standards across all types of contracts. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Why focus on Social Protection and Labor Systems 15 I Why focus on Social Protection and Labor Systems Social Protection and Labor (SPL) Systems around the world help individuals and societies manage risk and volatility and protect them from poverty and destitution-through instruments that address the challenges of resilience, equity and opportunity. SPL systems Improve resilience through insuring against drops in well-being from a range of shocks. Equity is enhanced through protecting against destitution and promoting equality of opportunity. Better opportunities for people come about by pro- moting better health, nutrition, education and skills development, along with helping men and women access productive work. The main SPL programs include social safety nets, social insurance, and labor market programs. As recent events have shown, the relative emphasis among these goals can change over time. Rightly, in relatively stable times, SPL programs are likely to focus on the priorities of human capital formation, equality of opportunity, poverty reduction and redistribution. This was the case for social assistance and labor programs in Central America prior to the COVID-19 pandemic, albeit with some shortcom- ings. The goal of shock-responsiveness (resilience) dramatically came to the fore during the COVID-19 pandemic, but within the sub-region and elsewhere it had been recognized earlier during natural di- sasters, and now more broadly with climate change. In analyzing SPL systems globally, and more spe- cifically, in Central America, the focus of this report, it is important to keep in mind these two different contexts as well as the different demands put on program design and delivery systems for each. Aligned in particular with the resilience goal, SPL Systems in Central America and globally, played a critical role in the response to the COVID-19 emergency. Countries across Central America2 deployed various SPL instruments to mitigate the impacts of the crisis, with all countries leveraging social assis- tance measures (in particular cash transfers, except Nicaragua).3 Across the globe, a little over a year into the pandemic more than 3000 SPL measures were planned or implemented in 222 countries, with the majority (~55%) being social assistance measures (Figure A. 1). This is not surprising, since one of the main goals of a SPL system is to help the vulnerable population cope with the negative effects of such a crisis: to help avoid irreversible losses to human capital, assets, and livelihoods and support them to return better prepared to the job market, as it includes a range of instruments (that can be deployed) aiming to protect against idiosyncratic and systemic shocks.4 Indeed, social protection and labor systems were employed to respond to the 2008-2009 global crisis across LCR (Grosh, Bussolo, 2 For the purposes of this report, Central America comprises the following countries: Costa Rica, Dominican Republic, El Salvador, Guatemala, Hondu- ras, Nicaragua, and Panama. 3 See sub-section II.D further details on the use of SPL instruments in response to COVID-19 pandemic across the sub-region. 4 ocial Protection and Labor instruments range from poverty targeted cash transfers and unemployment insurance programs, designed to act as automatic stabilizers, to labor market measures to help keep workers connected to their jobs or to facilitate other employment and support employability. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 16 Why focus on Social Protection and Labor Systems And Freije, 2014). Similarly, social benefits played a critical role in Europe and Central Asia acting as automatic stabilizers in some cases, and through concerted policy action relying on the strength of existing SPL systems in others (Isik-Dikmelik, 2012). Likewise, the response of countries in Central America built on recent progress made in many to establish progressive safety nets and social regis- tries, among other efforts. The nature and intensity of the downturn stretched SLP systems in the sub-region and laid bare some gaps. COVID-19 pandemic was an unprecedented shock that hit not only Central America, but all countries very hard…Central America was hit hard with contractions in GDP in 2020 ranging from -1.5% in Guatemala to -17.9% in Panama5 and with wide-ranging impacts on human development outcomes (as briefly outlined below in Box A. Given the speed, unexpectedness of the shock, as well as its wide ramifications in stretching supply chains, etc. it was impossible for any country to be fully prepared. Moreover, Central American SPL systems faced persistent challenges to varying degrees even in the relatively stable times that pre-date the pandemic. Challenges included: • limited efficiency and effectiveness of SP programs (e.g., limited progressivity of social assis- tance overall, low coverage of well-targeted programs limiting their impact etc.); • shortfalls in delivery systems (e.g., out of date, non-dynamic social registries with inadequate coverage, limited digital payments etc.); and • segmented labor markets, a labor force with inadequate skills, and limited institutional ca- pacity for labor market insertion, all obstacles to good employment outcomes. The COVID-19 crisis shone a light to and further deepened these structural issues, which presented obstacles to leveraging effectively existing SPL systems. Furthermore, the nature of this crisis under- scored the coverage gap for the newly vulnerable that normally fall outside safety net programs (e.g. transient poor, precariously employed, furloughed low income workers, etc.) and do not have access to the formal systems (e.g. severance payments in case of job loss, employment subsidy, unemploy- ment insurance if/when present, access to credit etc.), what could be called the “missed middle”.6 This in turn compelled a new and expanded definition of vulnerability and validated the importance of improving the adaptiveness of SP systems underscoring the need to build digital SP systems to better address shocks. Strengthening SPL systems is a priority investment not only to lay the foundation for stronger inclu- sive growth in the sub-region, but also to build resilience and mitigate further losses for the vulnera- ble, especially in light of increased shocks, including those related to climate change. Adaptive and effective SPL systems, which preserve, restore, and aim to augment human capital, are as important for inclusive recovery as structural policy reforms.7 At the same time, the poor and vulnerable in Cen- tral America (and the Caribbean) are among the most exposed in the world to natural disasters.8 The sub-region suffers from hurricanes, storms, flooding, and landslides, which have been increasing in frequency and intensity with disproportionate impacts on the poor and vulnerable and significant eco- nomic losses. 9 SPL systems can be leveraged for climate change adaptation and mitigation, building poor and vulnerable households’ and communities’ resilience (by diversifying livelihoods, encouraging savings, preventing poor coping decisions, etc.) and helping to prevent non-poor households from falling into poverty in case of shocks. They can also help manage the post-carbon transition through 5 Macro Poverty outlook, World Bank. April 2022. 6 Emerging evidence suggests that the newly vulnerable that are not beneficiaries of the safety nets nor the formal social security are the most affect- ed, while others could cushion the impacts a bit better due to receipt of such support (Lustig et al. 2020). 7 An effective and adaptive/resilient SPL system helps mitigate the impacts to the most vulnerable in the face of the crisis, helping to protect people’s human capital and increasing resiliency of households to shocks, while helping to prepare affected/vulnerable workers to return to employment/path to self-reliance during recovery 8 LCR Climate Change Action Plan PPT (2021). 9 Williams and Gonzales. (2020). Towards Adaptive Social Protection Systems in Latin America and the Caribbean: A Synthesis Note on using Social Protection to Mitigate and Respond to Disasters and Climate-Related Risks. World Bank, Washington, DC. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Why focus on Social Protection and Labor Systems 17 various SPL instruments (e.g., cash transfers that enable reform of oil, carbon, or utility subsidies; em- ployment programs to manage job transitions in the decarbonization process etc.) Considering the persistent and deepening challenges of SPL systems across the sub-region and their critical role in building resilience and protecting the poor and vulnerable before, during and after shocks (to mitigate further human capital and welfare losses), strengthening SPL systems is a priority and smart invest- ment for the sub-region. In this context, this report presents an overview of the challenges of SPL systems before and in the aftermath of COVID-19, with the main objective to lay the base for directions for policy reforms. The next section presents an overview of the long-standing challenges of the SPL systems in the sub-re- gion, how these challenges helped or constrained its capacity to mitigate the impacts of the COVID-19 pandemic, and the emerging challenges for SPL systems in the aftermath. It also documents the broader impacts of the COVID-19 pandemic on households and workers and summarizes the SPL responses deployed by countries across the sub-region. The third and final section presents directions for policy reform drawing on global knowledge. While the response to the pandemic was very partic- ular to that situation – programs were rolled-out fast, in some countries were large-scale, and tended toward flat benefits given the lack of time to develop a more differentiated design – future efforts to strengthen SP through sustainable programs can build on the experience gained. While recognizing the fiscal constraints faced by most countries, there is a clear need to invest in national SPL systems in the sub-region in order to strengthen the coverage and performance of the sector including adapt- ability to crisis, coupling needed improvements in delivery systems with more program resources. The report builds on and complements previous analysis (e.g., recently completed Jobs Diagnostics in several countries in the sub-region) and brings in new evidence through analysis of recent household surveys, Labor Force surveys, High Frequency Phone Surveys, and administrative data, focusing on before COVID-19 (most recent available: 2019 in most cases with exceptions). 18 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 19 II Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Most SPL systems in the sub-region include the basic pillars, though at varying levels of develop- ment. Social insurance (i.e., contributory) schemes exist but are underdeveloped in most countries. Unemployment insurance is absent across the sub-region. Non-contributory (social) pensions are present in all countries, though to a limited degree, to provide coverage to those not included in con- tributory schemes. A range of social assistance programs in various stages of evolution, constitute the bulk of the offer in most SPL systems, though not necessarily the spending. Finally, labor market programs and services, which promote employability and link workers to job opportunities, consti- tute a very small part of the SPL system in most cases.10 The level of development and challenges vary widely across and within Central American countries depending on which element of the SPL system is considered, although there are common themes... This section presents a snapshot of the challenges in SPL systems that pre-date COVID-19 as well as those that deepened or emerged during the pandemic, distilling common themes while pointing out differences. This is crucial to shed light on how to strengthen SPL systems to support a more resilient and inclusive recovery, as these struc- tural issues will remain even after the crisis has passed. The focus is on social assistance and labor programs, and their delivery, while highlighting broad links to pension systems.11 The challenges are broadly categorized under three themes: (i) limited effectiveness of SP spending and programs; (ii) limited adaptiveness of SP and delivery systems; and (iii) structural issues limiting good employment outcomes for workers and the economy. The rest of the section presents these challenges and their evolution during COVID-19 in turn. 10 Williams and Gonzales. (2020). Towards Adaptive Social Protection Systems in Latin America and the Caribbean: A Synthesis Note on using Social Protection to Mitigate and Respond to Disasters and Climate-Related Risks. World Bank, Washington, DC. 11 An in-depth treatment of the challenges in pension systems in the sub-region is planned as part of separate analytical work that is at early stages of development, and thus falls outside of the scope of this report. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 20 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 A1. Despite progress, the effectiveness of SP spending and programs was limited before COVID-19 Countries in the sub-region have made significant progress in the past decade towards a more com- prehensive social assistance system. Social safety net (SSN)/social assistance programs12 covering different segments of the population are present in all Central American countries. They include Conditional Cash Transfer (CCT) Programs in most countries (except Nicaragua), some of which were established more than a decade ago in an effort to promote human capital development and break the intergenerational cycle of poverty; and reflected a move towards more comprehensive assistance systems. 13 Despite progress on social assistance, structural issues persist. Structural issues include low to me- dium spending; limited progressivity largely due to the prevalence of categorical programs; and lower coverage and inadequate transfer amounts of the better-targeted programs, in particular CCTs, lim- iting their impact on human capital accumulation. Taken together with some of the fiscal and social sustainability issues faced in the social insurance system (including an expensive contributory system with high transition costs, as summarized later in the section in Box B), these challenges point to limited effectiveness and efficiency of SP programs and spending and indicate that there is room to improve the quality of spending. The rest of the sub-section provides more details on these challenges in turn. Spending on social assistance is below the LCR average for most Central American countries, with some exceptions, while its composition reflects an overall focus on categorical programs. While ac- cess to program level spending data varies, the most recent available data shows that spending on social assistance programs remains lower than the LCR average in most countries in the sub-region, with some exceptions, such as El Salvador and Costa Rica (Figure 1).14 Honduras and Guatemala are the lowest spenders on social assistance, especially when per capita spending in considered (Figure A. 2 and Figure A. 3), while the Dominican Republic, Panama and El Salvador’s spending are slightly below or roughly comparable with the LCR average, respectively. At the same time, spending on the traditional poverty targeted programs is outweighed by spending on categorical programs, highlight- ing the dilemma among the goals of fiscal balance, poverty relief/human capital promotion and wider sharing. For instance, in El Salvador, expenditure on electricity, gas and water subsidies represented 0.81% of GDP in 2019, while the expenditure on conditional cash transfer program represented 0.02% and social pensions, 0.08%. This is despite a steady reduction in spending on subsidies since 2014 when they represented almost 2% of GDP (Figure 2). In a similar vein, categorical programs comprise a bigger share of overall spending in other countries, such as Costa Rica where expenditures on cat- egorical programs is almost triple that of poverty-targeted cash transfer programs. This composition of spending may be deliberate reflecting the need to balance multiple social policy objectives. At the same time, considering that poverty reduction is one of the main objectives of social policy, this can lead to inefficiencies and/or coverage gaps, crowding out spending on more efficient approaches. The large, above LCR average, coverage of social assistance across the sub-region includes the poor and the rich. Coverage of social assistance programs in the poorest quintile of the population in al- most all countries in the sub-region exceeds the LCR average of 65% (Figure 3), ranging from 91% in 12 Social safety net programs are noncontributory interventions designed to help individuals and households cope with chronic poverty and vulnerabili- ty, targeted to poor and vulnerable households. Examples include unconditional and conditional cash transfers, noncontributory social pensions, food and in-kind transfers, school feeding programs, public works, and fee waivers. They are alternately known as social assistance programs. (World Bank, 2012 SPL Strategy 2012-2022) 13 Acosta, P., Almeida, R., Gindling, T., and C., Pena. (2017). Toward more efficient and effective public social spending in Central America. Washington, D.C.: World Bank Group. 14 Most recent pre-COVID-19 program level data for Costa Rica (from 2019 obtained as part of the DPF dialogue/support) indicate an overall social assistance spending of 1.86% of GDP, which reflects the positive evolution from 2014 (at 1.46% of GDP). The incorporation of the most recent program level data into the ASPIRE database is ongoing and thus not reflected in above Figure. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 21 Figure 1. Total Social Assistance Expenditure (% GDP), Latin America and the Caribbean 4.5 4.0 3.5 3.0 % of GDP 2.5 2.0 LAC,1.75 1.5 CA, 1.3 1.0 0.5 0.0 Trinidad and Tobago 2018 Chile 2018 Colombia 2018 Bolivia 2015 Argentina 2017 El Salvador 2019 Dominican Republic 2018 Jamaica 2018 Brazil 2018 Panama 2015 Paraguay 2017 Peru 2018 Mexico 2015 Uruguay 2015 Honduras 2018 Source: World Bank (2021). Atlas of Social Protection Indicators of Resilience and Equity (ASPIRE). Data is latest year available: Argentina (2017); Bolivia (2015); Brazil (2018); Chile (2018); Colombia (2018); Dom. Rep. (2018); El Salvador (2019); Honduras (2018); Jamaica (2018); Mexico (2015); Panama (2015); Paraguay (2017); Peru (2018); Trinidad y Tobago (2018); Uruguay (2015). Figure 2. Evolution of Composition of Social Assistance Expenditure in El Salvador, share of GDP, 2014-2019 3.5% Unconditional cash transfers Non-contributory health services 3.0% Other social care services Conditional cash transfers 2.5% Social pensions (non-contributory) School feeding 2.0% Other social assistance General Subsidies 1.5% Food and in-kind transfers Fee waivers and targeted subsidies 1.0% 0.5% 0.0% 2014 2017 2019 Source: World Bank (2021). Atlas of Social Protection Indicators of Resilience and Equity (ASPIRE). Panama to 62% in the Dominican Republic. This is largely driven by the prevalence of categorical pro- grams. For instance, school feeding programs are present in all countries, with coverage of the poorest quintile ranging from 23.9% in Guatemala to 71.2% in Costa Rica, while gas and electricity subsidies in El Salvador and Dominican Republic cover 39% and 65% of the poorest quintile, respectively.15 At the same time, the prevalence of categorical programs also translates into higher coverage of the richest quintile (top 20% of the income distribution) by social assistance programs. Indeed, this ranges from 67% of the richest quintile receiving at least one social assistance program in El Salvador to 32% in Panama, and 14% in Costa Rica, higher than the LCR average of 10%.16 15 Source, World Bank (2022) Atlas of Social Protection Indicators of Resilience and Equity (ASPIRE). In addition, scholarships programs have important coverage of the poorest population in Panama and El Salvador 16 In fact, El Salvador has the highest coverage of the richest quintile by SA programs in LCR by far, (ASPIRE). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 22 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Figure 3. Coverage of Social Assistance, Central America and LCR, by quintiles 100.0 90.0 80.0 70.0 % beneficiary household 60.0 50.0 40.0 30.0 20.0 10.0 0.0 CRI DOM SLV GTM HND NIC PAN LCR 1st quintile 2nd quintile 3rd quintile 4th quintile 5th quintile Source: ASPIRE performance indicators. Note: Latest available year before onset of COVID-19 pandemic: for Costa Rica (CRI) 2019, Dominican Republic (DOM) 2019, Honduras (HND) 2018, Panama (PAN) 2019 and El Salvador (SLV) 2019. Most recent data available for Guatemala (GTM) and Nicaragua (NIC) are from 2014 and included here only for completeness. For Latin America and the Caribbean 2010- 2019 weighted average using latest available survey year by country. Coverage may be underestimated because household surveys do not necessarily include the universe of social assistance programs. 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). Figure 4. Distribution of beneficiaries (beneficiary incidence), all SA, Central America and LCR, by quintiles 50.0 45.0 40.0 35.0 % beneficiary households 30.0 25.0 20.0 15.0 10.0 5.0 0.0 CRI DOM SLV GTM HND NIC PAN LCR 1st quintile 2nd quintile 3rd quintile 4th quintile 5th quintile Source: ASPIRE performance indicators. Note: Latest available year before onset of COVID-19 pandemic: for Costa Rica (CRI) 2019, Dominican Republic (DOM) 2019, Honduras (HND) 2018, Panama (PAN) 2019 and El Salvador (SLV) 2019. Most recent data available for Guatemala (GTM) and Nicaragua (NIC) are from 2014 and included here only for completeness. For Latin America and the Caribbean 2010-2019 weighted average using latest available survey year by country. Beneficiary incidence is: (Number of individuals in the group who live in a household where at least one member receives the transfer)/(Total number of direct and indirect beneficiaries). SA in most countries is less progressive compared to LCR, though there are significant differences across programs. In general, social assistance programs are less progressive in the sub-region than the LCR average (Figure 4). While on average in LCR a little less than half of SA beneficiaries (46%) belong to the poorest quintile, the share is considerably lower in the sub-region pointing to lower progressivity, though to varying degrees. For Costa Rica and Honduras on the higher end, around one third of SA beneficiaries belong to the poorest quintile, while in El Salvador only 22% come from the lowest quintile (Figure 4), indicating an almost uniform distribution. Nonetheless, there are significant differences across programs. For instance, in El Salvador electricity, gas and water subsidies, have an almost uniform distribution with a slightly higher share of beneficiaries belonging to the third and fourth richest quintiles of the population, while other programs (e.g., CCT, social pensions) are much Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 23 Figure 5. Distribution of beneficiaries (beneficiary incidence), El Salvador, 2019, by type of SA program, by quintiles 70.0 60.0 50.0 %Beneficiary distribution 40.0 30.0 20.0 10.0 0.0 Q1 Q2 Q3 Q4 Q5 CCT Social pensions In kind School feeding OSA Source: ASPIRE performance indicators; El Salvador (SLV) 2019. Beneficiary incidence is: (Number of individuals in the group who live in a household where at least one member receives the transfer)/(Total number of direct and indirect beneficiaries). more progressive (Figure 5), though they are smaller in coverage. Electricity, gas, and water subsidies, cover a much larger share (almost 7 out of 10 Salvadoreans), thus dominating overall SA performance. Indeed, SA incidence of SA does not contribute significantly to redistribution in El Salvador due to poor progressivity (targeting), with most resources going to energy subsidies and categorical programs. Indeed, conditional cash transfer (CCT) programs are well-targeted in most countries across the sub-region. Progressivity in CCT programs is mainly evident in Panama and El Salvador with a higher share of beneficiaries belonging to the first quintile than the LCR average of 54% (Figure 6). CCT pro- grams in Costa Rica and Honduras are more similar to the LCR average. Costa Rica’s “Avancemos” is the biggest CCT program in the region with 206,000 beneficiary households, half of which belong to the first quintile. CCT programs in the Dominican Republic17 and Guatemala18 are less progressive (Figure 6).19 However, CCTs across the sub-region have both low coverage and low adequacy of benefits in most countries, likely limiting their impact. The coverage of the first quintile for these programs is generally below the LCR average (49.1%). Particularly, “Comunidades Solidarias (urbanas y rurales)” in El Salvador only covers 3.7% of the poorest population. Other CCT programs cover roughly 3 out of 10 individuals in the bottom quintile. “Red de oportunidades” and “Senapan” in Panama reach 26% of the poorest population and “Bono Vida Mejor” in Honduras reaches 33%. The Dominican Republic has several CCT programs which reach in total 35% of the bottom quintile, namely: “Comer es Primero”, “Incentivo a la Asistencia Escolar”, “Bono Escolar Estudiando Progreso”, “Incentivo a la Educacion Superior”. In Guatemala coverage has dropped in recent years given the sharp reduction in the number of families covered by the CCT program, called Mi Bono Seguro, from roughly 700,000 in 2014 to slightly less than 100,000 in 2021 for Bono Social. Within the sub-region, Costa Rica’s CCT program, “Avancemos”, has the highest coverage of the poorest quintile (44.3%) (Figure 7). Similarly, with the exception of Honduras (where the average benefit level of the CCT represents 29.3% of its poorer beneficiaries’ welfare), the adequacy of other CCTs is below the LCR average of 17.2% of beneficiaries’ welfare20: the benefit level 17 The relatively poor performance in the Dominican Republic is due to several factors: a significant number of poor households are not included in the social registry SIUBEN; the targeting instrument has not been updated in the last decade; and although some beneficiaries have improved their situation and are now above the cut-off for the CCT, they have not been transitioned out of the program. 18 Latest data available for Guatemala corresponds to 2014. It should be noted that in 2017 the program had a recertification process, which likely result- ed in improved progressivity however lack of recent data limits assessment to confirm such improvement. 19 While distribution of benefits would be a better indicator of targeting, lack of benefit amount data in some countries prevents its calculation. 20 Welfare is measured based on income or consumption depending on the country. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 24 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Figure 6. Distribution of beneficiaries (beneficiary incidence/progressivity) CCTs, Central America and LCR, by quintiles 90 80 70 % of beneficiary households 60 50 40 30 20 10 0 CRI DOM SLV GTM HND NIC PAN LCR 1st quintile 2nd quintile 3rd quintile 4th quintile 5th quintile Source: ASPIRE performance indicators. Note: Latest available year before the onset of the COVID-19 pandemic: for Costa Rica (CRI) 2019, Dominican Republic (DOM) 2019, Honduras (HND) 2018, Panama (PAN) 2019 and El Salvador (SLV) 2019. Most recent data available for Guatemala (GTM) and Nicara- gua (NIC) are from 2014 and included here only for completeness. For Latin America and the Caribbean 2010-2019 weighted average using latest available survey year by country. Beneficiary incidence is: (Number of individuals in the group who live in a household where at least one member receives the transfer)/(Total number of direct and indirect beneficiaries). Figure 7. Coverage Conditional Cash Transfer Programs, Central America and LCR, by quintiles 60 50 % beneficiary households 40 30 20 10 0 CRI DOM SLV GTM HND NIC PAN LCR 1st quintile 2nd quintile 3rd quintile 4th quintile 5th quintile Source: ASPIRE performance indicators. Note: Latest available year before the onset of the COVID-19 pandemic: for Costa Rica (CRI) 2019, Dominican Republic (DOM) 2019, Honduras (HND) 2018, Panama (PAN) 2019 and El Salvador (SLV) 2019. Most recent data available for Guatemala (GTM) and Nicara- gua (NIC) are from 2014 and included here only for completeness. For Latin America and the Caribbean 2010-2019 weighted average using latest available survey year by country. Beneficiary incidence is: (Number of individuals in the group who live in a household where at least one member receives the transfer)/(Total number of direct and indirect beneficiaries). 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). for poor beneficiaries of Costa Rica’s “Avancemos” represents 13.3% of welfare, for Panama’s CCTs it is 10.5% and for the Dominican Republic it is 5.9%. (Figure A. 4). Low coverage along with inadequate benefit levels limit CCTs impact on poverty reduction, despite good targeting. The prevalence of categorical programs with limited progressivity crowds out the expenditure and reach of better targeted SA programs; and implies less focus on addressing the barriers to human capital formation and disparities across the socioeconomic distribution in the sub-region. Social as- sistance programs, in particular those that are targeted to the poor and vulnerable, are designed to directly or indirectly reduce poverty and develop human capital. The low coverage of the poorest by Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 25 targeted social assistance programs (such as CCTs) indicates limited access to such programs and the other services they provide, reducing the focus on human capital formation and on addressing the disparities across the socioeconomic quintile. Most countries in the sub-region have relatively poor human capital outcomes. Indeed, the Human Capital Index (HCI)21 of the countries in the sub-region, before the pandemic hit, was lower than the LCR average except for Costa Rica which reached 0.63 in 2020, showing an increase of 0.03 from 2010.22 Disparities abound across the different components considered for HCI, with Costa Rica having high- er expected years of school, learning-adjusted years of school and adult survival rate, than the other countries. On the other hand, Guatemala has the highest stunting rate of the sub-region, whereas in the Dominican Republic children have a 93% probability of not being stunted. In addition, stark dispar- ities exist across socioeconomic groups: in Guatemala, the HCI of a child born today in a household belonging to the top quintile is 0.63 while it is 0.40 for a child born in the poorest quintile, a gap of 22 percentage points, much higher than the average gap of 15 percentage points across 50 countries.23 A2. Limited reach and potential mitigation impact of existing SP programs against the unprecedented COVID-19 pandemic These challenges did not bode well for the arrival of the COVID-19 pandemic, which threatens to erode the hard-won human capital and deepen the disparities across socioeconomic groups with impacts to be felt for years to come. The effects of the COVID-19 crisis represent for Latin-Ameri- can countries the loss of a decade in a single shock24 hindering the living standards and wellbeing of its population. Before COVID-19 poverty had declined in the sub-region with the exception of Costa Rica with stagnant poverty rates and Nicaragua with higher poverty rates (Figure 8). Estimates for 2020 suggest that poverty rates will rise in all Central American countries with the highest increases Figure 8. International poverty rate ($1.9 in 2011 PPP) from 2018 to 2023 (estimates) in Central America 20 18 International poverty rate ($1.9 in 2011 PPP) 16 14 12 10 8 6 4 2 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020e 2021f 2022f 2023f 2024f Costa Rica Dominican Republic El Salvador Guatemala Honduras Nicaragua Panama Source: The World Bank, The Macro Poverty Outlook, April 2022 21 The HCI “measures the amount of human capital that a child born today can expect to attain by age 18, given the risks to poor health and poor educa- tion that prevail in the country where she lives”. It includes: (i) probability of survival to age 5, (ii) expected years of school, (iii) harmonized test scores, (iv) learning-adjusted years of school, (v) adult survival rate, and (vi) under 5 stunting rate Human Capital Project, October 2020. 22 Human Capital Project (2020). 23 Human capital Project, (2019). Guatemala-Insights from Disaggregating the Human Capital Index. 24 The World Bank, Renewing with Growth. 2021 Semiannual report of the Latin America and the Caribbean region. March 2021 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 26 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 expected for Panama (+58%) and Costa Rica (+110%) (Figure 8). In addition, despite lower overall levels (except Costa Rica), in the last decade there were gains in human capital accumulation, as evidenced by improvements in HCI in most countries,25 although rural vs. urban disparities persist. However, the significant impact of the pandemic on the poor and vulnerable in multiple dimensions risks reversing these hard-won gains in poverty and human capital. Indeed, the initial evidence points to disproportionate impact on the poor and vulnerable, further widening gaps in human capital. The economic consequences of the pandemic posed challenges on several dimensions, with widespread job losses and income reductions, declines in food security, constrained access to medical care, education etc., all of which has implications for protecting and building human capital. Inequalities persist in the countries of the sub-region and the impact of the pandemic also had unequal effects on different groups of the population. People living in poverty, informal workers, the rural population, women, children, youth, and the elderly were more prone to be affected by COVID-19 and the restriction measures implemented to control it. Simulations of the im- pact of COVID-19 on human capital point to decreases in HCI in all countries in Central America, with upper middle-income countries like Costa Rica and Panama being most affected.26 Box A presents an overview of the initial impacts of COVID-19 across several dimensions with implications for human capital. Labor market impacts are covered in sub-section II.C. Most countries in the sub-region acted swiftly to contain the impacts COVID-19, however, the asso- ciated costs put pressure on the already vulnerable fiscal situation. The size of the emergency pack- ages ranged from around 1.5% of GDP (in Costa Rica) to 3.4% (Guatemala), with the exception of El Salvador where the cost of the emergency response reached 15.5% of GDP,27 putting most countries in a vulnerable fiscal situation (in particular Costa Rica, El Salvador and Panama, Figure 9). Indeed, most countries of the region borrowed heavily in 2020 and 2021 in order to fund COVID-19 emergency response measures and cover tax revenue shortfalls from the related economic crisis. This increase in indebtedness will have to be addressed in the medium and long term and, among other measures, will likely bring forward pension reforms in order to increase their fiscal sustainability. (See section D for an overview of the government’s responses in SPL to the COVID-19 pandemic). Figure 9. Fiscal balance deficit as % of GDP for Central America and LCR average 12 10 Fiscal balance deficit (% GDP) 8 6 4 2 0 2019 2020e 2021f 2022f 2023f 2024f Costa Rica Dominican Republic El Salvador Guatemala Honduras Nicaragua Panama LCR average Source: Own elaboration based on The Macro Poverty Outlook, The World Bank, April 2022.. LCR average estimated based on population weights. 25 Between 2010 and 2020, 26 Based on COVID_HCI_SIMS tool produced by the Global HCI team. 27 The World Bank, Macro Poverty Outlooks 2021. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 27 Box A. Broad Impacts of COVID-19 on households and human capital The unprecedented nature of the COVID-19 crisis, with simultaneous supply and demand shocks and both regional and global impacts, had wide-ranging effects, and presented unique challenges for SPL systems. The COVID-19 crisis started as a supply shock with declining mobility resulting from fear of contagion and furthermore, from mobility restrictions imposed by government authorities.a This affected several economic sectors, mainly tourism, hotels, restaurants, and personal services. World demand for goods and services changed with the new living conditions brought by the pandemic. Bio-security supplies were highly demanded as well as appliances to adapt homes for work and entertainment.b On the supply side, firms reduced sales compared to the same period in 2019. According to the Business Pulse Survey and the World Bank Enterprise Survey, on average, 86% of firms in Latin America and the Caribbean (LCR) reported a drop in sales (data for 6 countries, see Figure Box 1). Sales reductions were higher for small firms in most countries and for firms in the other services sector. Of concern, 1 out of 2 firms said they were in arrears or expected to be in the next 6 months, worse for medium and small firms. 2 out of 10 firms reported that they had fired workers in the last 30 days, almost 3 out of 10 in large and medium firms. Firings were most common for establishments in the other services sector. Figure Box 1. Percentage of establishments with lower monthly sales compared to one year before the interview 120 100 Percentage of establishments 80 60 40 20 0 All Large Medium Small Micro Manufacturing Retail Other (100+) (20-99) (5-19) (0-4) Services Brazil Guatemala Honduras Mexico Nicaragua El Salvador Source: Business Pulse Survey and the World Bank Enterprise Survey. Data for Wave 1 for Brazil (June 19 to July 31, 2020), Guatemala (June 24 to August 7, 2020), Honduras (June 24 to August 7, 2020), Nicaragua (June 17 to August 7, 2020) and El Salvador (June 10 to August 7, 2020) The estimated GDP contraction for 2020 in the world is 4.3%, but worse in LCR, reaching 6.7%.c Within LCR, Central American countries faced GDP reductions from 1.8% in Guatemala to 17.9% in Panama, the latter one of the most affected countries in the region.d The Panamanian econ- omy depends on tourism, construction and retail, sectors that were heavily affected by the mobility restrictions during the pandemic. For instance, hotels and restaurants faced a 55.8% decrease followed by 51.8% in construction and 30.8% in real estate and business activities. Honduras and El Salvador, aside from the COVID-19 effects on their economy, were also heavily impacted by the Iota and Eta hurricanes. In Honduras the hurricanes affected around 3.9 mil- lion people and caused an economic loss of almost 7.2% of 2019 GDP (US$1.8 billion), severely damaging infrastructure and crops.e GDP contracted around 8-9% in both countries. Guatema- la, Costa Rica, and Nicaragua, report GDP decreases lower than the LCR average (see Figure Box 2). In general, the service sector has been most hit in the region as tourism was heavily affected by pandemic measures. In Panama and Honduras, the industry sector faced the high- est decreases. Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 28 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box A. Broad Impacts of COVID-19 on households and human capital (continued) Figure Box 2. GDP growth from 2019 to 2024 (estimates) in Central America and LCR 20.0 15.0 10.0 GDP growth (%) 5.0 0.0 2019 2020e 2021f 2022f 2023f 2024f -5.0 -10.0 -15.0 -20.0 Costa Rica Dominican Republic El Salvador Guatemala Honduras Nicaragua Panama LCR average Source: Own elaboration based on The Macro Poverty Outlook, April 2022, The World Bank. Note: The Macro Poverty Outlook. LCR average estimated based on population weights. Income reduction and jobs loss hindered the ability of households to maintain or increase their human capital accumulation. The pre-COVID-19 Human Capital Index (HCI) for the coun- tries in the sub-region (except Costa Rica) remain low compared to the LCR average.f The neg- ative effects of COVID-19 caused a worsening in all the areas that contribute to human capital accumulation. First, income reduction affected the access to food and nutrition. About 53% of households in Honduras and 47% in Guatemala reported running out of food because of a lack of money or other resources in the early stages of the pandemic. The situation is worst in households with children, worsening their malnutrition problems (see Figure Box 3). In Gua- temala, for instance, chronic malnutrition is a determining factor for a lower HCI. Before the pandemic, Guatemala’s share of children not suffering from stunting was the lowest in Central America at 0.53, while the sub-regional average was 0.78.g The impact of COVID-19 on children’s malnutrition is expected to decrease the HCI even further with higher effects in departments where chronic malnutrition was higher pre-COVID-19 (e.g., Totonicapan, Solola). Although the situation has improved after the first year of the pandemic, still almost one third of the house- holds in the region, mainly in Dominican Republic and Honduras, have reported to run out of food due to lack of money in June-July 2021 (see Annex A.1). Figure Box 3. Percentage of households that ran out of food due to lack of money or other resources (last 30 days) 70% 60% % of households 50% 40% 30% 20% 10% 0% Costa Rica Dominican Republic El Salvador Guatemala Honduras Total Presence of 65+ Presence of Kids Source: World Bank, High Frequency Phone Surveys, Washington DC. Data for Wave 1 (between May 21st and June 1st) Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 29 Box A. Broad Impacts of COVID-19 on households and human capital (continued) Figure Box 4. Education activities engaged in by households with children 100 90 80 70 % of HHs with children 60 50 40 30 20 10 0 Engaged Used Session/ Watched Listened to Other in any learning/ mobile learning meeting with educational educational education apps Lesson Teacher TV programs programs activities (tutor) on radio Costa Rica El Salvador Dominican Republic Guatemala Honduras Source: World Bank, High Frequency Phone Surveys, Washington DC. Data for Wave 2 (June 19th -June 28th) Second, access to education for children and youth has represented an enormous challenge since COVID-19 hit. Even though most households with children said that they were engaged in some educational activity, there were important differences remain across the region. For instance, in Costa Rica 96.4% of households with children said they were attending some sort of learning activity, while in Honduras the share was 83%. From 6 to 7 out of 10 households said that their children have sessions with teachers (albeit virtual). Other remote learning channels were also used: for instance, in countries with higher internet penetration (Costa Rica 81.2% and Dominican Republic 75.8%)h children use mobile learning apps (63% and 67.9%, respec- tively). A small share listens to educational programs on the radio (between 3.7% in Costa Rica and 12.6% in El Salvador); and others watch educational TV programs (from 27% in Costa Rica to 74.3% in El Salvador). See Figure Box 4. The quality of the remote engagement generates disparity among households. For example, online classes with teachers are more prevalent in wealthier (70%) than in highly vulnerable households (53%). For a significant share of households with children who were not engaged in any learning ac- tivity, lack of internet access was the main deterrent for participation. In Honduras more than 30% reported this to be the main deterrent, while in other countries of the sub-region the share is smaller. Differences exist among households depending on the level of education. In Hondu- ras, Dominican Republic and Guatemala, the lack of internet as a factor limiting participating in education activities, is lower for households with high education. This is different for El Salvador and Costa Rica (see Figure Box 5). Moreover, limited access to internet and mobile devices is more prevalent in poor and vulnera- ble households. For instance, in Panama, only 8% of students living in poverty have access to a computer and only 4% have connection to internet. 63% of these students have a mobile device but only 22% of them have internet connectivity. Disparities also exist in the access of schools to internet; for instance, 97% of schools in Panama Centro have access compared to 13% in the indigenous comarca Ngäbe-Buglé.i Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 30 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box A. Broad Impacts of COVID-19 on households and human capital (continued) These disparities are evident in the gaps that exist in the HCI among the Panamanian provinc- es. Before COVID-19, the indigenous comarcas had the lowest HCI. A World Bank’s simulation estimates that the effects of income reduction due to the pandemic in Panama will reduce learn- ing-adjusted years of school (LAYS) by between 0.68 to 1.05, which signifies a drop in the HCI of between 5.5% to 8.4%. HCI reductions are bigger in Coclé, Herrera, and Panamá. The indigenous comarcas have a lower, but still significant reduction placing them in a more vulnerable situation.j Figure Box 5. Households with children that reported lack of internet as main deterrent to participate in school (among those not participating) 60 50 % of HH with children 40 30 20 10 0 Honduras El Salvador Dominican Republic Guatemala Costa Rica Total High education Low education Source: World Bank, High Frequency Phone Surveys, Washington DC. Data for Wave 2 (June 19th -June 28th) Third, medical attention was also hindered by COVID-19. At the beginning of the pandemic not having access to medical treatment ranged from almost 10% of household members in Costa Rica to 26.2% in Dominican Republic. Even though this situation improved across time, still by the end of July 2020 between 2.5% and 24.8% of household members that needed medical treat- ment couldn’t get it, with worse effects in households with children and/or elderly (Figure Box 6). Figure Box 6. Percentage of household members who needed but were not able to access medical treatment 35 30 % of household members 25 20 15 10 5 0 Wave 2 Wave 2 Wave 2 Wave 2 Wave 2 Wave 3 Wave 3 Wave 3 Wave 3 Wave 3 Wave 1 Wave 1 Wave 1 Wave 1 Wave 1 Costa Rica Dominican Republic El Salvador Guatemala Honduras Total Presence of 65+ Presence of Kids Source: World Bank, High Frequency Survey, Washington DC. Wave 1 was between May 21st – June 1st, Wave 2 was June 19th -June 28Th, Wave 3 was July 18th – July 26Th Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 31 Box A. Broad Impacts of COVID-19 on households and human capital (continued) Not having access to medical treatment was driven by the over demand that the health sys- tem faced during the worst waves of the pandemic. At the early stage of COVID-19, at least 4 out of 10 household members in the Central American region said that they couldn’t access health treatment because of lack of personnel. Fear of contagion and lack of resources were other reasons not to get medical treatment when needed (see Figure Box 7). Figure Box 7. Reasons households could not receive medical attention when needed. 80 % reporting didn't go % reporting lack % reporting lack of 70 out of fear of contagion of personnel resources as main issue % of household members 60 50 40 30 20 10 0 Wave 2 Wave 2 Wave 2 Wave 2 Wave 2 Wave 3 Wave 3 Wave 3 Wave 3 Wave 3 Wave 1 Wave 1 Wave 1 Wave 1 Wave 1 Costa Rica Dominican El Salvador Guatemala Honduras Republic Source: World Bank, High Frequency Survey, Washington DC. Wave 1 was between May 21st – June 1st, Wave 2 was June 19th -June 28Th, Wave 3 was July 18th – July 26Th a The World Bank, The Cost of Staying Healthy. 2020 Semiannual report of the Latin America and the Caribbean region. October 2020 b The World Bank, Renewing with Growth. 2021 Semiannual report of the Latin America and the Caribbean region. March 2021 c Idem. d St Lucia is the LCR country with the higher GDP lose in 2020, Panama is the second fallowed by Barbados. e World Bank. WDI, Macro Poverty Outlook. f The HCI “measure the amount of human capital that a child born today can expect to attain by age 18, given the risks to poor health and poor education that prevail in the country where she lives”. It includes: (i) probability of survival to age 5, (ii) expected years of school, (iii) harmonized test scores, (iv) learning-adjusted years of school, (v) adult survival rate, and (vi) health growth. Human Capital Project, October 2020. g Simple average of not stunted rate based on the World Bank’s Human Capital Project, October 2020. Data for Guatemala, El Salvador, Nicara- gua, Honduras and Dominican Republic. Data not available for not stunted rate for Panama and Costa Rica. h WDI Indicators for individuals using internet as percentage of total population. For El Salvador 50.5% (2019), Guatemala 44.4% (2019), Panama 63.6% (2019), Nicaragua 27.9% (2017) and Honduras 32.1% (2017). i Taken from the World Bank’s ongoing study on human capital development in Panama. Data for students’ internet access from Encuesta de Hogares de Propósito Múltiple 2019. Data for schools’ internet access from COPEME (Septiembre 2020). Política Educativa: Educación a Distancia Propuesta. j Idem The COVID-19 pandemic also impacted pension schemes across the sub-region, putting further pres- sure on their fiscal and social sustainability, including in unexpected places like Costa Rica. While the impacts have not yet been fully documented it is already clear that sharp declines in employment rates have negatively affected the revenues of PAYG schemes and reduced contribution flows into individual pension saving accounts. The number of new retirees has also likely increased, as reduced employment opportunities for those who already have a right to a pension forced many into early retirement. Even countries that until very recently have felt unencumbered by short- and medium-term fiscal sustainabil- ity issues in their PAYG schemes have suddenly been caught in unexpected problems. For example, the still immature pension scheme of Costa Rica has over time accumulated sizable reserves to pre-fund its future liabilities, expecting to slowly divest them as their expenses rise above revenues in the decades ahead. However, a sudden drop in revenues caused by COVID-19 has made them dip into their reserves sooner than expected. Unfortunately, as they are trying to plan the sales of their reserves, mostly invested in government debt and mortgages, they are finding that the liquidity of these assets is highly restricted. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 32 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 In sum, the COVID-19 pandemic exposed and deepened the structural vulnerabilities in SP spending and programs, accelerating the need to improve efficiency and preparedness for any future shocks. The previously documented structural challenges in the SP spending and programs indicate that there is room to improve efficiency and effectiveness. The fiscal constraints that resulted from the COVID-19 pandemic further underscore the need to achieve better results and to use resources more efficiently. B1. Limited adaptiveness of SP and delivery systems Central American countries have invested in the development of their delivery systems over the past decades. While there are country-specific characteristics and variations, the sub-region shows com- monalities in the level of development of their delivery systems. In broad terms, all countries have tak- en steps to integrate functions across programs, and, among others, they have done this through the development of social information systems and investment in common payment platforms, among others. The following paragraphs will look in more detail at specific aspects of delivery systems in Central America pre-COVID-19, and then how these were leveraged, and the innovations introduced in order to respond to the pandemic. Development of social registries, but challenges In order to support the processes of intake, registration and assessment, social registries have been developed in the countries of Central America over the past decades, though their coverage is rela- tively low, with a few exceptions. Country strategies on social registries are quite varied with different possible paths to their establishment and use. Globally, social registries have been developed in at least 57 countries over the past decades and their coverage ranges from less than 5 percent of the popula- tion (e.g., Comoros, Cote d’Ivoire, Madagascar, Chad, Belize and Jamaica), to almost universal coverage (e.g., Argentina, Uruguay and Pakistan, see Figure 10). In the Latin America and Caribbean region, seven countries have social registries covering 70% or more of the population, including Argentina, Chile, Co- lombia, Costa Rica, Dominican Republic, Peru and Uruguay. A second group includes countries such as Brazil, Honduras and Ecuador, which cover between 30% and 69% of the population. A third group of countries whose social registries cover less than 30% of the population includes El Salvador, Haiti and the Plurinational State of Bolivia.28 In the Central America sub-region, coverage ranges from 85% in the Dominican Republic to 24 percent in El Salvador. In addition, Guatemala is currently working on a pilot for the social registry. Intake and registration approaches in the sub-region have largely relied on infrequent en masse reg- istration waves resulting in out-of-date data. For example, Honduras has registered approximately a million households through different registration waves that started in 2015 relying on home visits and registration desks.29 Similarly, the Dominican Republic carried out census sweeps in 2004, 2012 and 2018.30 Costa Rica has conducted census-sweeps every two years. In the case of Panama, after a census sweep in poor districts between 2006 and 2009, intake and registration has been performed on demand by each social program under the Ministry of Social Development (MIDES) through the Ficha Única de Proteccion Social (FUPS). 28 Social Protection and Jobs Responses to COVID-19: A Real-Time Review of Country Measures “Living paper” version 9 (May 15, 2020); H. Berner and T. Van Hemelryck, “Social information systems and registries of recipients of non-contributory social protection in Latin America in response to COVID-19”, Project Documents (LC/TS.2021/56), Santiago, Economic Commission for Latin America and the Caribbean (ECLAC), 2021. 29 Centro Nacional de Información del Sector Social, “Mejores Prácticas en la Implementación de la Estrategia Operativa para la Actualización de la Información Socioeconómica en el País”, Seminario Internacional sobre Actualización de Información De Sistemas De Información Social, Santo Domingo, October 2019; Centro Nacional de Información del Sector Social (https://ceniss.gob.hn/index.html) 30 “Social programs targeting system in the Dominican Republic: the experience of SIUBEN”, Knowledge Exchange Session Social Protection Dominican Republic - Saint Lucia, presentation, May 2021. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 33 Box B. Brief Panorama of Pension systems in Central America Pension systems in Central America display similar characteristics as those of the wider Latin America region. The PAYG based defined benefit schemes were introduced around the middle of the last century, starting with Costa Rica and Panama in 1941, and ending with Guatemala in 1969. However, with the exception of Costa Rica and Panama, the coverage of the working age population remains disappointing compared to the average rates achieved in Latin America, and especially in richer OECD countries, as shown in Figure Box 8. While low coverage rates are not surprising given a lower level of economic development and labor market formalization, the countries of the sub-region were feeling impatient with the low coverage and growing pension cost, which prompted the introduction of individual savings account based defined contribu- tion schemes in Costa Rica (1995), Honduras (2015), Panama (2005), and El Salvador (1998). The latter is aiming for the gradual but full elimination of the PAYG-based defined benefit pillar, while the rest of the countries foresee the defined contribution and defined benefit systems continu- ing to complement each other in the future. Figure Box 8. Contributors as percent of working age population (ages 15-64) 80% 70% 60% 50% 40% 30% 20% 10% 0% Honduras Guatemala Nicaragua El Salvador Panama Costa Rica LAC OECD Source: ILO 2017, WB Pensions Database, Inter-American Development Bank, and ILOSTAT. average average A hoped-for boost in coverage expected from the introduction of individual savings accounts did not materialize and, again following the trends of the wider region, Central American coun- tries have introduced non-contributory social old age pensions in Guatemala (2005), El Salva- dor (2009) and Panama (2010). Also, Costa Rica provides a means tested benefit for the elderly, disabled and widowed persons under the family allowances program and Honduras provides some support in-kind and for those with disabilities. Non-contributory benefits helped boost elderly pension coverage to 60-70 percent in Costa Rica and Panama, but the rest of the coun- tries still cover less than 20 percent of their elderly population. Further expansion of non-con- tributory pension benefits has run into budgetary and administrative constraints. However, the slow progress could also be at least partially justified by other, more urgent priorities. For exam- ple, poverty rates are higher among the population younger than 65, as shown in Figure Box 9, which might indicate a higher need for social assistance programs to be targeted to younger age groups. Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 34 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box B. Brief Panorama of Pension systems in Central America (continued) Figure Box 9. Poverty rates at $5.5/day, overall population vs. population over the age of 65. 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Panama Costa Rica Dominican El Salvador Nicaragua Guatemala Honduras Republic Total Over 65 Source: The World Bank. Poverty and Equity brief. October 2021. Note: Latest available year, for Panama 2019, Costa Rica 2019, Dominican Republic 2019, El Salvador 2019, Nicaragua 2014, Guatemala 2014, Honduras 2019. Rapid demographic aging and maturation of pension schemes also make the countries of the sub-region increasingly worried about the long-term fiscal and social sustainability of their pension programs. With the exception of Honduras and Guatemala, pension contribution rates in the rest of the region cannot be considered low, as shown in Figure Y, and so far, generally have been sufficient not to require additional budgetary financing. However, assuming a 15- year retirement period and real rate of return equaling real wage growth, a contribution rate of 15 percent can only self-finance a 20 percent replacement rate after 20 years of contribu- tions, while current pension promises for this short contribution period range from 40 percent in Costa Rica to 63 percent in Nicaragua. This clearly indicates that current pension promises are heavily subsidized, which can only result in an increased taxation burden on future genera- tions or politically difficult benefit cuts in the future. El Salvador, which has committed to fully transition to a purely savings-based system is already experiencing both – political pressure to maintain pension subsidies and budgetary costs associated with an unfunded minimum pen- sion increase of 50 percent in 2019. Figure Box 10. Pension contribution rates assessed on insured wage income in 2019. 20 18 16 14 12 10 8 6 4 2 0 Honduras Guatemala Costa Rica Panama El Salvador Nicaragua employee employer government Source: World Bank Socio-Economic Database for Latin America and the Caribbean Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 35 Figure 10. Coverage of social registries in selected countries, share of population 120 100 99 99 87 85 85 82 80 Share of population 75 75 73 60 60 50 47 45 40 40 37 36 30 29 25 24 23 20 19 14 11 10 10 8 8 6 3 3 2 1 0 0 0 Ur tina Do Pa uay ca n ep st ru Ph C a pi e lom s Jo ia Tu an Ec ey B r do l nd a Ge ras Se gia rth al al ac ne Sa ia au or ia iti M Iraq er s Le ijan n ali Gh ia Co omo a s ag e r ad Ja ize a In razi do ca Co nes Az ritiu te ro ic ilip hil Ho nesi C an ad ir aic ini ta no b El don an Ni P neg liv Ha M vad Co Pe rk M Ch nR rd M esti M d'Ivo l aR or u as Be m kis ua ug ba n ba Bo m rit au ge e l Ar Source: Own elaboration, based on data collected by the Delivery Systems GSG Limited development of modern payment mechanisms Before the COVID-19 pandemic, most social protection payment systems in Central America relied on non-account based over-the-counter payments, while a few had progressed towards integration and digitization. For example, in Guatemala the main cash transfer program (Bono Social) reached only 150,000 households and faced difficulties to process payments quickly.31 In Honduras, the over- the-counter payment system previously employed for the CCT Program (mobile ATMs in rural areas) prevented it from functioning as a shock-responsive safety net, though the government started work- ing on the transition to a digital account-based system in 2019.32 On the other hand, in Panama, a payment platform for MIDES programs (Plataforma Única de Pagos) was created in 201633, and ben- eficiaries receive quarterly payments through debit cards (Tarjeta Clave Social), in coordination with a national bank, (Banco Nacional de Panamá - BNP). In dispersed or isolated areas of the country, the BNP supports MIDES to deliver payments through mobile cash dispensers (which amount to 18 percent of total payments). In the case of Costa Rica, while transfers were made into bank accounts before the pandemic, the payment system was not centralized; rather each social program managed payments independently, leading to inefficiencies and lack of verification and control of the actual amounts paid to beneficiaries.34 In addition, in general before the COVID-19 pandemic, the widespread use of digital payments was limited to a few countries. In the case of Guatemala, El Salvador and Honduras, account ownership was 44%, 30% and 45% respectively (see Figure 11 ). Similarly, before the pandemic digital payment methods were only used by 33% of the population in Guatemala, 24% in El Salvador and 37% in Hondu- ras (see Figure 12 ).35 However, in Costa Rica, account ownership rates are higher overall, although the gap (14 percentage points) between men (75%) and women (61%) is still important.36 31 Activity Completion Summary, Guatemala - Strengthening safety net delivery systems (P172923) 32 Project Paper, Honduras, Additional Financing to the Social Protection Integration Project (P175718). 33 Government of Panama, law 54/2016. 34 IMF (2021), “Recomendaciones para el desarrollo del Sistema Unificado de Pago de Recursos Sociales (SUPRES) de la Tesorería Nacional”, technical report, May 2021. 35 World Bank 2020. Hacia un ecosistema de pagos digitales para impulsar la inclusión financiera. Diagnóstico en base a la publicación Payment aspects of financial inclusion (PAFI) del CPMI y Banco Mundial. 36 IMF (2021), “Recomendaciones para el desarrollo del Sistema Unificado de Pago de Recursos Sociales (SUPRES) de la Tesorería Nacional”, technical report, May 2021. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 36 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Figure 11. Adults who own an account in the financial system 60% 54% 55% 50% 44% 45% 40% 30% 31% 30% 20% 10% 0% Guatemala Bolivia El Salvador Honduras Nicaragua Latinamerica and the Caribbean Source: World Bank 2020. Hacia un ecosistema de pagos digitales para impulsar la inclusión financiera. Diagnóstico en base a la publicación Payment aspects of financial inclusion (PAFI) del CPMI y Banco Mundial. Figure 12. Use of payment digital payment methods 50% 46% 45% 40% 40% 37% 35% 33% 30% 24% 25% 25% 20% 15% 10% 10% 6% 7% 6% 4% 4% 4% 4% 5% 5% 3% 2% 2% 0% Made or received digital Used a cell phone or internet to access Have a digital account payments last year an account in a finacial institution Guatemala Bolivia El Salvador Honduras Nicaragua Latinamerica and the Caribbean Source: World Bank 2020. Hacia un ecosistema de pagos digitales para impulsar la inclusión financiera. Diagnóstico en base a la publicación Payment aspects of financial inclusion (PAFI) del CPMI y Banco Mundial. B2. Obstacles to Leverage SP Delivery Systems to Respond to COVID-19 In order to expand social protection programs vertically and horizontally as a response to the COVID-19 crisis, globally most countries leveraged the use of their delivery systems. The level of ma- turity of a country’s delivery systems as well as their dynamism and capacity to adapt in response to shocks were important factors when defining options for a social protection response to the crisis. Countries with better preexisting delivery systems tended to perform better in introducing and ex- panding social protection measures. For example, this included relying on existing social registries and other administrative databases and payment systems. The limited adaptiveness of the delivery systems in Central America affected the quality and timeli- ness of the response. The level of maturity, inclusiveness and dynamism of each country’s delivery sys- tems directly affected the social protection response. In some cases, challenges included out of date or the low coverage of social registries, limited digitalization of payments and overall lack of dynamism of delivery systems. These obstacles prevented leveraging of social safety net programs to protect existing beneficiaries and expand coverage to the newly vulnerable, with some governments having to create parallel systems to respond swiftly. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 37 Transitioning from census-sweep approaches to on-demand applications and interoperability of social information systems With the onset of the COVID-19 pandemic, one of the main challenges for governments across the globe was having access to data to assess households’ needs and conditions in order to determine eligibility for emergency transfers. Globally, three main approaches were used. The first approach was to leverage data from social registries, often cross-checking information with other administrative da- tabases. A second approach was on-demand applications, also often combined with cross-checks of self-reported data with administrative databases. The third approach used new sources of data. A key aspect was the existence of social registries, their level of coverage and whether data was up to date or not. In the Central America sub-region, the three main approaches seen at a global level and described above were also used. The Dominican Republic relied on SIUBEN, its social registry, to expand coverage of the PROSOLI program to an additional 690,000 non-beneficiary families registered in SIU- BEN (Quedate en Casa).37 In addition to relying on SIUBEN, the government conducted cross-checks with other administrative databases, including the Social Security Treasury and the General Directorate of Internal Taxes.38 In Costa Rica, the government introduced a platform to enable online applications for the emergency transfer Bono Proteger. Once applications were submitted, the government checked the applicant’s eligibility by pulling data from several administrative databases, including SINIRUBE (so- cial registry), SICERE (social security), Migration, Finance, electoral registry (TSE) and national electronic payment system (SINPE).39 Similarly, Panama relied on several databases to determine eligibility for the emergency transfer Panama Solidario, including the electoral registry, as well as databases on pensions, civil servants, formal workers, tax, and energy consumption.40 In the case of El Salvador, the social registry (Registro Único de Participantes- RUP) was not used, as data was not up to date and coverage was limit- ed. Instead, eligibility for emergency transfers was determined based on data on a gas subsidy, which was cross-checked with the RUP. In the case of Guatemala, given the absence of a social registry, eligibility for the emergency cash transfer Bono Familia was determined based on data on electricity consumption as a proxy of income. The government relied on administrative data from the energy utility companies along with other public databases, including RENAP (civil registry), social insurance and MIDES social pro- grams).41 In order to enroll beneficiaries, the government used text messages, call center, and a website. Moving from non-account based over-the-counter payments to integrated and digitized payments COVID-19 provided an impetus for innovations, leading to an increased digitization of payments. In Gua- temala, the emergency program Bono Familia introduced two digital solutions for its delivery. First, the program’s intake and registration could be done fully virtual via an online platform, SMS, or by phone. Sec- ond, payments administration and provision were digitized via temporary virtual accounts, which made it possible for beneficiaries to cash-out at a wider network of payment points and without the need of a card. Beneficiaries could access their payments via a one-time-password at any ATM or at participating stores with POS machines (in addition to in-person at bank branches).42 In Honduras, in 2020 the govern- ment launched a digital payment mechanism through the Honduran Bank for Production and Housing (Banco Hondureño para la Producción y la Vivienda, BANHPROVI), a public institution, directly contract- ed to make CCTs payments to beneficiaries, although due to other factors there were gaps in the regular cycle of payments during 2020-2021. Specifically, this service provides: (i) a Payment Network System 37 This included families in the lowest quality of life index categories (ICV1, ICV2 and ICV3). The emergency response also included a top-up to 811,000 beneficiary families that are part of the PROSOLI CCT program to purchase food and cover basic needs. 38 COVID-19 responses in Latin America and the Caribbean, brief. 39 Government of Costa Rica. 2021. “Noveno Informe Mensual de Seguimiento a la Ejecución del Bono Proteger”, (https://www.mtss.go.cr/elministerio/ despacho/covid-19%20mtss/plan_proteger/archivos/noveno_informe_proteger.pdf) 40 Panamanian Executive Decree 400, March 27, 2020, Executive Decree 833, December 28, 2020; and Executive Decree 432, July 1, 2021.,. 41 Activity Completion Summary, Guatemala - Strengthening safety net delivery systems (P172923). 42 Activity Completion Summary, Guatemala - Strengthening safety net delivery systems (P172923). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 38 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 with the competitive participation of multiple agencies and banking correspondents, which together of- fer broader coverage, with a standardized payment mechanism, offering multiple cash-out points; (ii) an information technology platform specializing in payments and e-commerce, the functionality and opera- tion of which was proven by a nationally funded payment pilot, which through e-wallets offers elements that ensure the traceability and audit of payment transactions; (iii) a simplified and standardized payment mechanism that allows the use of a one-time-password key through short messaging services (SMS) and the use of the national identity document to authenticate payments; and (iv) a foundation for advancing the financial inclusion of participants through the use of e-wallets and relations with financial institu- tions in the country.43 In Panama, payments of the Vale Digital component of the emergency program Panamá Solidario relied on the use of ID cards, which could be used by beneficiaries to make purchases in supermarkets and pharmacies. Costa Rica used digital payments for its emergency transfers and is working on implementing a centralized payment system for social benefits, called Sistema Unificado de Pago de Recursos Sociales (SUPRES).44 The Dominican Republic began to make digital payments and use electronic payments for the horizontal expansion of the PROSOLI CCT (Quedate en Casa)45. C1. Structural issues Limiting Good Employment Outcomes for workers and the economy Pre-COVID 19, most jobs in the sub-region were of poor quality, concentrated in lower productivity sectors, limiting good employment outcomes. Even in countries where there has been relatively robust job creation, such as the Dominican Republic and El Salvador,46 the majority of new jobs have been low quality.47 In some countries, such as Honduras and the Dominican Republic, the economy has moved steadily away from low productivity agriculture jobs, with some improvements in overall productivity due to this sectoral re-allocation.48 However, this structural transformation did not appear to translate into better jobs, as it was coupled with lower productivity growth in services in some cases (e.g. Hon- duras, Guatemala, Dominican Republic) and signs of premature deindustrialization (i.e. stagnant or a decreasing share of industrial employment, such as in the Dominican Republic and Honduras, Figure A. 7).49 Indeed, most jobs across the sub-region are in services, accounting for more than two thirds of the jobs in Costa Rica, Dominican Republic and Panama (Figure 13). While not all service sector jobs are low quality, declining within sector productivity and high levels of informality point to poor or deteriorating job quality. Indeed, with the exception of Costa Rica and Panama, the majority of service sector jobs are informal likely concentrated in elementary occupations with low productivity (Figure 13). In sum, the low quality of jobs was an overarching challenge (underscored by multiple structural challenges)50 for the sub-region limiting good employment outcomes for workers and the economy. Persistently high informality was prevalent across the sub-region, with some exceptions, highlighting the segmentation, driven by multiple barriers for formalization. In almost all countries in the sub-region 43 Project Paper, Honduras, Additional Financing to the Social Protection Integration Project (P175718). 44 IMF (2021), “Recomendaciones para el desarrollo del Sistema Unificado de Pago de Recursos Sociales (SUPRES) de la Tesorería Nacional”, technical report, May 2021. 45 World Bank (2021), “Respuesta de Protección Social a la Crisis de Covid-19 en la Republica Dominicana” 46 For instance, the employment to GDP growth elasticity in the Dominican Republic and El Salvador is comparable to other countries with a 1% increase in GDP growth associated with 0.55 and 0.9 percent increase in employment, which translated into creation of jobs (but not necessarily better jobs).(Winkler and Montenegro, 2021; Banegas and Winkler, 2020). 47 Broadly the following dimensions are taken into account to assess job quality: (i) the level and growth of earnings; (ii) type of employment (salaried, self-employed or unpaid family work); (iii) productive vs. unproductive sector growth; and (iv) formal vs. informal (legal definition, whether job offers social security affiliation). Lower earnings (growth), non-wage employment, low productive sector employment, and informality each individually (and collective- ly) point to lower quality. 48 For instance, in Honduras the main productivity growth between 2012-2016 was due to increased productivity in agriculture and the reallocation of labor/jobs from agriculture to higher productivity services, though within productivity in services has been declining pointing to deterioration of quality of jobs in the sector overall. HND JD. 49 Beylis et al. (2020). Going Viral: COVID-19 and the Accelerated Transformation of Jobs in Latin America and the Caribbean. World Bank Latin American and Caribbean Studies; Washington, DC: World Bank. 50 While some common threads most relevant for the labor market outcomes of vulnerable (such as informality, skills mismatches, etc.) will be touched upon in this report, a full treatment of the underlying structural issues, which vary according to country context, go beyond the scope of this report. Read- ers can refer to the Jobs Diagnostics for select Central American countries and Ulku and Zaourak (2021) for further details. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 39 Figure 13. Structure of Employment in Central America and Dominican Republic, by main sectors and by informal vs. formal Costa Rica Panama Dominican Republic El Salvador Guatemala Honduras 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Agriculture - formal Agriculture - informal Industry - formal Industry - informal Services - formal Services - informal Source: Own elaboration based on based on El Salvador 2019 (EHPM), Dominican Republic 2019 (ECNFT), Honduras 2019 (EPHPM), Guatemala 2018 (ENEI) using SEDLAC harmonization. Data for Panama 2019 (EML) and Costa Rica 2019 (ECE) provided by the NSO and standardized by the team. Data for Nicaragua not available. Note: Informality based on legal characteristics: not-salaried, employer/sala- ried/self-employed without social security Figure 14. Share of Informal vs Formal Employment in Central America and DR 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% HND 19 GTM 18 SLV 19 DOM 19 PAN 2019 CRI 19 Informal as % of employed Formal as % of employed Source: Own elaboration based on El Salvador 2019, Dominican Republic 2019, Honduras 2019, Guatemala 2018 using SEDLAC harmonization. Data for Panama 2019 and Costa Rica 2019 provided by the National Statistics Offices and standardized by the team. Note: Legal definition of informality is used. More specifically this refers to those employed (including non-salaried, employer, salaried, self-employed) without social security, as a share of total employed. at least half of employment is informal, ranging from 49% in Panama to 82% in Honduras (Figure 14). Costa Rica is the exception with around one third informal employment. As such, large swaths of the employed lack risk management mechanisms, leaving them vulnerable in the case of shocks. Informal- ity has persisted51 in most countries reflecting labor market segmentation, driven by multiple barriers for formalization.52 Indeed, the countries across the sub-region by and large did not adapt to a changing world, exacerbating labor market segmentation, making it difficult for firms to enter, survive, and create 51 For instance, in Dominican Republic the share of affiliation with the Social Security Institute has remained low over the last decade despite robust GDP and job growth (Winkler and Montenegro, 2021). 52 The potential barriers for formalization encompass a wide range of issues: limited competition and access to finance; lack of innovation, infrastruc- ture, and integration into GVCs; barriers to trade; and crime/insecurity etc. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 40 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Figure 15. Overall LM Rigidity Index (as compared to global average) 2.5 2 LM Reg. Rigidity Index (PCA) 2019 1.5 1 0.5 0 -0.5 -1 -1.5 . K Gr ua Sa ago st ) Ba ica Gu s a Ur iti Ni uay nd a Co evis ua Dom ia m d B ica n R da Ar blic Sa a da at r Br nd ala Ec na Ve Bo r zu ia B e a he Su hile e ico ru Ho ity) Pa as a C o lo ini Gu ado do o aic s a ad El tin li z i m m ,R b ne liv ha uc Ha Pe ur ad ca arbu g au ya ,T Do an in aR d a em m C na i na Be ua ug u C en b ra ela n N m Ba t. L nd lv as rb ep az To oP o ge ca Ja r l m S ex (M il ( i ini ico tig St ex An Tr M Source: Packard et al. (2019) formal jobs. Although, according to the overall LM rigidity index, the sub-region (except for Honduras and Panama) is more flexible compared to the global average (Figure 15),53 rigidities exist in some di- mensions, in particular for working hours (Annex Figure A. 8 – Figure A. 10. - show the dimensions of the index). This in turn, has resulted in worse labor market outcomes for women and youth, as documented for El Salvador (Banegas and Winkler 2020). Similarly, minimum wages may limit the growth of formal jobs in some countries through diverse channels. For instance, Honduras has the highest minimum wage in the sub-region, significantly above the average wage (and value added per worker).54 It appears to be binding in the formal sector, but due to its high level, it has limited ability to influence wages in the informal sector. Similarly, in El Salvador the average minimum wage is high, equaling roughly the average wage, but cannot be enforced due to large levels of informality (80% of informal workers earn less than the minimum wage), while presenting a barrier for formalization. While the average minimum wage is low by comparison in the Dominican Republic, the proliferation of minimum wages that apply to different industries, regions, etc., may impede efficient allocation and promote segmentation. Access to quality jobs is particularly limited for those in the bottom of the income distribution. While most jobs are low quality, the incidence is much higher among the bottom 40 compared to the top 60 across the sub-region (Figure 16). For instance, in Honduras and Guatemala practically all of the employed in the bottom 40 percent of the income distribution are informal (97% in both cases) while the incidence of informality is much lower among the top 60 percent (though still high at around two thirds). A large gap in informality levels across the income distribution characterizes most countries, reaching a difference of 40 percentage points in Panama. The incidence of unpaid work is also higher among the bottom of the income distribution while the reverse is true for wage employment for all countries (Figure 17). The divide is especially large in Panama, with 72 percent of the top 60 percent of the income distribution in wage employment compared to only 41 percent for poor workers. Similarly, there’s a sharp divide between the incidences of unpaid workers in the bottom 40 as compared to the top 60, with poor (and likely rural) households stuck in subsistence agriculture/household enterprises, while other (likely urban) households enjoy access to better jobs. The high incidence of unpaid work in Honduras and Guatemala reflects the continued prevalence of household enterprises though to a lesser degree for the better off compared to the poor population. There are myriad underlying reasons 53 In fact, based on their flexibility of labor regulation and spending on human capital and labor programs, Panama and Honduras can be classified as low flexibility and protection while GTM, DOM, SLV, NIC, CRI are considered high flexibility and low protection. Silva et al (2021). 54 The average minimum wage was 1.64 of the average value added per-worker, which is among the highest in the world. Walker and Michel (2018) Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 41 Figure 16. Employment, informality, and unemployment in Central American countries for bottom 40% and top 60% of the population Bottom 40% Top 60% 120 20 80 7 18 70 100 6 16 60 5 80 14 12 50 4 60 10 40 8 3 30 40 6 2 20 20 4 10 1 2 0 0 0 0 ca r ala as a a ca r ala as a a do do gu m gu m ur ur Ri Ri m m na na lva lva ra ra nd nd a a e e Pa Pa ca ca st st at at Sa Sa Ho Ho Co Co Ni Ni Gu Gu El El Total employment rate Share of informal workers Unemployment rate Source: Own elaboration based on LAC Equity Lab Tabulations of LABLAC and World Development indicators. Data for circa 2019. Note: Legal definition of informality is used. Those employed (including not-salaried, employer, salaried, self-employed) without social security as a share of total employed. There were inconsistencies for data for Dominican Republic and thus it is excluded from the graph. Figure 17. Share of wage and unpaid workers in bottom 40 and top 60 percent of the population for Central American countries Bottom 40 Top 60 80 90 70 80 60 70 60 50 50 40 40 30 30 20 20 10 10 0 0 ica . r ala as a a ica . r ala as a a ep ep do do gu m gu m ur ur aR aR em em nR nR na na lva lva ra ra nd nd Pa Pa ca ca st st at at Sa Sa ca ca Ho Ho Co Co Ni Ni Gu Gu ini ini El El m m Do Do Share of salaried workers Share of unsalaried workers Source: Own elaboration based on LAC Equity Lab Tabulations of LABLAC and World Development indicators. Data for circa 2019. Note: Legal definition of informality is used. Those employed (including not-salaried, employer, salaried, self-employed) without social security as a share of total employed. (including disparities in human capital across the income distribution that translate into worse em- ployment outcomes), however the fact remains that the poor, even when employed,55 are stuck in informal, labor-intensive, and low productivity jobs56 linked with low earnings, indicating their vulnera- bility against idiosyncratic and systemic shocks. Similarly, the low-skilled, youth, and women face worse labor market outcomes in most countries, with the resulting vulnerabilities for their welfare and economy at large. Overall, the labor market prospects of the skilled appear to be much better than those of their less-skilled peers (Figure 18). For instance, in the Dominican Republic low skilled workers are more likely to have an informal job, compared to their more educated peers.57 Likewise, in Costa Rica, while 79 percent of highly educated workers have a job, only 55 percent of the medium and 61 percent of the low-educated have jobs. The 55 Indeed, they are employed at much lower rates than their better off counterparts, with particularly wide gaps observed in Costa Rica (24 pp), Domini- can Republic (20 pps) and El Salvador (18 pp). (Figure 16) 56 Informal jobs are linked with low productivity, due to difficulties faced by small, informal enterprises in accessing capital. Walker and Michel (2018) and Banegas and Winkler (2020). 57 High school and college graduates are 8.5. and 16.5 percentage points less likely to have an informal job compared to their peers without primary education. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 42 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Figure 18. Employment rate by education level in Central American countries 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Guatemala El Salvador Costa Rica Honduras Dominican Republic Panama Low Medium High Source: Own elaboration based on El Salvador 2019 (EHPM), Dominican Republic 2019 (ECNFT), Honduras 2019 (EPHPM), Guatemala 2018 (ENEI) using SEDLAC harmonization. Data for Panama 2019 (EML) and Costa Rica 2019 (ECE) provided by the NSO and standardized by the team. Note: Employment as percentage of working age population (between 15 and 64 years old). Education level: low 0-8 years, medium 9-13 years, 14 or more years of education. employment gap varies widely across countries, however, with the smallest difference between high and low educated observed in the Dominican Republic and in Panama (7 and 6 percentage points re- spectively). Similarly, youth face worse labor market outcomes across the sub-region: they have lower labor force participation, are more likely to be employed in informal jobs (Figure A. 12) and in lower productivity sectors, which are linked to lower earnings.58 A large share of youth are neither employed, nor in education or training (NEETs) ranging from 18 and 19 percent in Panama and Costa Rica respec- tively to 29 percent in Honduras and Guatemala (Figure Box 14). High NEET rates signal challenges in the school to work transition and are associated with other social problems (e.g., poverty traps, teen pregnancy, crime, and violence etc.), with important scarring effects for labor market outcomes and lifetime earnings.59 While there has been progress in almost all countries in narrowing the gender gaps in LM outcomes, non-negligible disparities persist, and are particularly large in some countries, reflect- ing the multiple barriers women face in labor market inclusion. Box C presents a snapshot of women’s LM outcomes across the sub-region. Despite gains in educational attainment, generally the supply of skills in the labor force in most coun- tries remains low and out of step with jobs of the future. In most countries, the education level of the labor force has increased steadily, reflecting the gains in access to education of younger generations.60 At the same time, overall educational attainment remains low,61 partly due to the type of jobs available (labor-intensive, unproductive, low-skilled), which in turn is impacted by the skills available in the labor force, pointing to most countries being locked in a low equilibrium with polarization. Furthermore, high- er educational attainment does not necessarily translate into relevant skills, partly due to ineffective workforce development policies with weak linkages to market needs, resulting in challenges in work to school transition. At the same time, the changing world of work means increased demand for ad- vanced skills (such as problem solving, adaptability, digital skills, etc., in addition to foundational skills on literacy and numeracy), which appear to be in short supply in the sub-region. Indeed, evidence from El Salvador and the Dominican Republic (based on estimates of task content jobs following Lo Bello et 58 For instance, in DR salaried workers with informal jobs earn 33% less than their peers with formal jobs, while informal self-employed and employers earn about 42 percent less than their formal peers. In El Salvador, even after controlling for observable characteristics, women and youth are about 9 per- cent and 17 percent more likely to be in informal jobs than males and prime-age workers respectively. (Winkler and Montenegro, 2021; Banegas and Winkler, 2020) 59 De Hoyos Navarro, R., Popova, A. and Rogers, H. (2016). Out of school and out of work: a diagnostic of ninis in Latin America. Washington, D.C.: World Bank Group. 60 For instance, in Honduras as of 2019, 3 percent of youth had no education, less than one third the rate in the adult population (around 10%); and 9 percent had incomplete primary, compared to 23% for the adult population. (Updated analysis for Walker and Michel (2018) using 2019 data). Similarly, in Guatemala, the number of workers with at least advanced secondary education has increased between 2004-2018, and the number with incomplete primary has declined (GTM JD). Similar trends were present in El Salvador, Nicaragua, and Dominican Republic. 61 With only 58 % of the labor force having secondary education or less, as compared to 52 percent of LCR average. Costa Rica is the exception where the labor force is largely better educated on par with the region though lagging behind OECD countries. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 43 al. 2019), show that jobs that are intensive in the skills of the future62 (e.g., non-routine analytical, inter- personal, or non-routine manual) broadly have limited and declining prevalence, while jobs intensive in skills of the past (routine cognitive or routine manual) have higher prevalence.63 Not surprisingly, the jobs intensive in skills of the future are likely to be held by the more educated and are of higher quality (e.g. tend to be formal, in large firms, with higher earnings), while jobs of the past are of low quality and under threat of disappearance with technological change, not to mention more vulnerable to systemic shocks. This is yet another structural divide that is likely to be exacerbated by the COVID-19-pandemic. Furthermore, gains in educational attainment did not always translate into higher earnings, as was the case in El Salvador and Guatemala, likely leading to increased migrant outflows. While higher education- al attainment is in general linked with better employment outcomes, and in particular higher earnings, declining returns to education (i.e. wage skill premium) is observed in some countries such as El Salvador and Guatemala.64 This is particularly the case for college degrees: for instance, the average wage premi- um for college graduates (as compared to completed primary) in El Salvador declined from 75 percent in 2007 to 36 percent in 2017, resulting in an overall decline in real wages for college graduates.65 While the supply of college graduates increased over the same period in both cases, it appears that the growth in demand for such workers was insufficient, resulting in supply outpacing demand. The lack of oppor- tunities for more educated (mostly young) workers provides an impetus for international migration to maximize returns on skills. Selection into migration is not random and unsurprisingly, migrants from El Salvador and Guatemala, especially in recent years, tend to be more educated than their peers at home.66 At the same time, the coverage of labor market programs is low with varying degrees of institutional capacity for effective labor market inclusion, particularly for the vulnerable and unemployed. Effec- tive employment support requires robust institutional capacity, to be able to identify the needs of jobseekers, workers and employers, and to design and implement packages of programs tailored to needs. This involves a strong job training system (including TVET) linked with private sector needs as well as formal education and the effective delivery of employment services (e.g., job search assistance, labor market intermediation, hiring subsidies). While all countries have some of the basic elements required for employment support, capacity varies widely. Moreover, in general labor market programs in Central America, especially training, do not focus on priority groups (vulnerable and unemployed) and are generally perceived as taxes by contributing sectors, especially employers. On the one end is Costa Rica, which has been modernizing its main training agency, the National Institute of Learning (INA for its acronym in Spanish), to better fit the changing needs of the labor market. The moderniza- tion efforts are in-line with international evidence and include introducing the previously unavailable option of outsourcing to private accredited providers, prioritization of vulnerable students and work- ers (economically disadvantaged, women etc.), and the monitoring of labor demand among others. On the other end is the TVET system in Honduras, which has the foundational elements but is largely out of date, failing to provide training for skills in demand, and in need of significant reform to better support especially youth and first-time job seekers. In addition, the coverage of labor market programs is low, except for Costa Rica. Given the range of employability challenges especially among the poor, lower skilled, and youth, it is imperative to strengthen institutional capacity to support the transition to productive employment and expand the coverage of effective programs. 62 Following Acemoglu and Autor (2011), the task content of jobs can be measured in their intensity of the following 5 skill categories: Non-Routine Analytical, Non-Routine Interpersonal, Routine Cognitive, Routine Manual, and Non-Routine Manual. Jobs intensive in Routine (both cognitive and manual) skills are more likely to be automated and disappear with technological change while Jobs intensive in Non-Routine especially Analytical and interpersonal skills are likely to benefit from technological change and have been increasing in prevalence in the advanced economies across the globe (i.e. US, European Union etc.). 63 DR JD, SLV JD. There is a silver lining however, in that increased educational attainment in the labor force (in particular college education) can be a significant driver in the transition to jobs intensive in future of skills (especially in cases of unmet demand for a better educated workforce). This seems to have been the case in the Dominican Republic, where an increased share of college educated workers is driving the small increase in the share of the jobs that are intensive in non-routine cognitive and analytical skills. 64 A similar trend was observed in other countries in LAC, as documented by Aedo and Walker (2012). 65 Banegas and Winkler (2020). A similar trend is observed in Guatemala, though returns to tertiary education are much higher in absolute terms. More specifically, the returns to a college degree (as compared to incomplete primary) in Guatemala has decreased from 266 percent in 2004 to 184 percent in 2018 (Eberheard, forthcoming) 66 Sousa et. al., 2018. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 44 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box C. Overview of Women’s LM Outcomes in Central America Women’s labor force participation rate is very low relative to men in Central America, with potential important economic losses. Guatemala with a rate of labor market participation for women less than half the rate for men is the lowest in LCR. Although narrowing over the past decade in most countries, the participation gap between men and women in the sub-region ranges between 22 and 24 percentage points in Costa Rica and Panama up to 46 and 39 per- centage points in Guatemala and Honduras respectively (Figure Box 11), underscoring the barri- ers women face in accessing the labor market. A cross-country analysis indicates that if women were to participate in the labor market at the same rate as men, per capita GDP would be higher across the sub-region, ranging from 16% in El Salvador (on par with LCR average increase) to 25% in Guatemala (Cuberes y Teignier, 2016). While this analysis is a stylized model that does not take into account the demand side fully, it points to the potential size of the economic losses. Figure Box 11. Labor Force Participation in Central America, by gender Panama Costa Rica Dominican Republic El Salvador Honduras Guatemala 0 10 20 30 40 50 60 70 80 90 100 % of labor force participation Male force participation Female force participation Source: Own elaboration based on El Salvador 2019, Dominican Republic 2019, Honduras 2019, and Guatemala 2018 using SEDLAC harmoniza- tion. Data for Panama 2019 and Costa Rica 2019 are provided by the NSO and standardized by the team. Note: Labor force participation: Adult population that is in the labor force as share of working age population. Labor force includes employed workers and those who are actively seeking work The participation gap is largest for low-educated women while more educated women par- ticipate at much higher rates across the sub-region. The more educated women are the more likely they are to participate in the labor market (Figure Box 12) reflecting the likely higher re- turns to education, which may ease some of the most binding barriers (e.g., by being able to afford childcare). Indeed, the participation gap narrows drastically for highly educated women, for example, to only 6 percentage points in Honduras (Figure Box.12). Even when women do participate, they face worse labor market outcomes. Women in the la- bor force are more likely to be unemployed or informally employed (Figure Box 13). They also tend to have lower earnings, despite having higher education in most cases. For instance, the gender wage gap is about 23% in Dominican Republic, after controlling for the higher education levels of employed women (Winkler and Montenegro, 2021). Similarly, in El Salvador, the wage skill premium has been consistently higher for men compared to women, with skilled men mak- ing 17 percent more than women (Banegas and Winkler, 2020). Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 45 Box C. Overview of Women’s LM Outcomes in Central America (continued) Figure Box 12. Labor force participation in Central America, by education level and by gender 100 90 80 70 60 50 40 30 20 10 0 Female Male Female Male Female Male Female Male Female Male Female Male Guatemala Honduras El Salvador Costa Rica Panama Dominican Republic Low Medium High Source: Own elaboration based on El Salvador 2019, Dominican Republic 2019, Honduras 2019, and Guatemala 2018 using SEDLAC harmoniza- tion. Data for Panama 2019 and Costa Rica 2019 are provided by the NSO and standardized by the team. Note: Low education refers to: less than 9 years of education, medium: from 9 to 13 years of education and high: 14 or more years of education Figure Box 13. Unemployment rate in Central America, by gender 18 16 14 Unemployment rate % 12 10 8 6 4 2 0 Costa Rica Dominican Panama Honduras El Salvador Guatemala Republic Female unemployment rate Male unemployment rate Source: Own elaboration based on El Salvador 2019, Dominican Republic 2019, Honduras 2019, and Guatemala 2018 using SEDLAC harmoniza- tion. Data for Panama 2019 and Costa Rica 2019 are provided by the NSO and standardized by the team. While there is a wide range of barriers faced by women, care responsibilities emerge as a bind- ing constraint in most cases, especially when there is a young child in the household. In Costa Rica, women’s labor force participation registers a sharp increase once a child reaches 3 (in- dicating more potential outside childcare opportunities- Figure A. 13). Similarly, in El Salvador, the gap in the probability of employment between men and women narrows as the youngest child gets older, largely disappearing at school age. This is due to an increased probability of employment for women since men’s employment is not affected by the age of the youngest child (Figure A. 14). Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 46 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box C. Overview of Women’s LM Outcomes in Central America (continued) The school to work transition is particularly challenging for young women: the majority of NEETs are women, with long-term labor market impacts. Women constitute the majority of the NEET population, from slightly less than two-thirds in Costa Rica to around 80 percent in Honduras and Guatemala (Figure Box 14). There are numerous underlying reasons (high lev- els of adolescent pregnancy, which correlates with school dropout or delayed entrance to the labor market if at all, gender biased social norms, etc.) translating into lower levels of human capital accumulation (through formal education and on-the job training) and thus impacting their life-time earnings. Figure Box 14. Youth NEET rate (rhs) and gender composition, Central America 100% 35 90% 30 80% 70% 25 60% 20 50% 40% 15 30% 10 20% 5 10% 0% 0 Panama Costa Rica Dominican El Salvador Honduras Guatemala Republic NEET Female NEET Male NEET rate Source: Own elaboration based on El Salvador 2019, Dominican Republic 2019, Honduras 2019, Guatemala 2018 using SEDLAC harmonization. Data for Panama 2019 and Costa Rica 2019 provided by the NSO and standardized by the team. Note: NEET includes youth between 15-24 years old who are not in employment, education, or training (i.e., unemployed and/or inactive) C2. The COVID-19 pandemic exacerbated LM challenges Arrival of the COVID-19 pandemic, against this challenging backdrop, exacerbated these LM issues, which, if not addressed, threaten to have long term repercussions. The COVID-19 pandemic led to an unprecedented simultaneous supply and demand shock, which worked through two main channels in the labor market: (i) confinement measures shuttered large parts of the economy (that were con- sidered non-essential); and (ii) consumer activities, especially those requiring face to face contact, decreased to minimize risk of infection. The negative impacts were concentrated on high-contact em- ployment, which tends to be informal, low skilled, with low earnings, deepening the above document- ed divisions in multiple dimensions in most countries in the sub-region; and threatening to scar large swaths of workers in the long-term. Initial findings on the impacts of the COVID-19 crisis as they relate to structural issues in LMs in the sub-region are summarized below. At early stages of the pandemic, with stricter restrictions, households experienced significant in- come losses. In more informal economies such as Guatemala, El Salvador, and Honduras, almost 7 out of 10 households interviewed for the World Bank’s High Frequency Phone Survey (HFPS)67 reported decreases in income (Figure 19). This reduction was mostly a consequence of people stopping work when the pandemic hit. People working in services and industry were the most affected. In El Salva- dor, Honduras, and Dominican Republic more than 55% of respondents working in services reported 67 Nationally Representative High Frequency Phone Surveys undertaken by World Bank -First phase. Data collected May/June 2020 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 47 Figure 19. Percentage of households that report a reduction of total income during quarantine 90.0 81.4 80.0 64.5 71.7 73.8 68.1 68.8 69.9 70.0 70.0 66.3 66.1 63.1 61.7 63.9 60.0 59.2 60.0 60.0 53.9 50.3 50.0 40.1 47.7 45.1 43.0 44.2 38.8 40.7 38.9 40.0 30.0 20.0 10.0 0.0 i na i le lic ico ica y as or ala ia bia or ru ua liv ad ad Pe ub ur Ch aR em ex lom nt g Bo nd alv u ra ep ge M Ec st at Pa Co Ho nR S Ar Co Gu El ca ini m Wave 1 Wave 4 Do Source: World Bank, High Frequency Phone Surveys, Washington DC. Data for Wave 1 between May 21st and June 1st and Wave 4 between June and July 2021. stopping work. In these countries wholesale and retail activities account for 20% and hotels and restaurants for almost 10% of the economy (the latter only in El Salvador and Dominican Republic, in Honduras the share is smaller). These activities mostly require personal interaction, restricted espe- cially early on in the pandemic. More than 54% of people working in the industrial sector in Guatemala, El Salvador and Honduras also said they stopped working after the pandemic hit. Within this sector, manufacturing and construction activities are the most relevant representing 20% of the economy in these countries. People working in agricultural were the least affected as food production and its sup- ply are essential services (Figure 21) However, the impact was not low, as 2 to almost 4 people out of 10 in the agricultural sector said they stopped working due to COVID-19. In general, a higher percentage of self-employed people said they stopped working, relative to employees. This is most evident in El Salvador: 67% of self-employed workers paused working compared to 50% of employees (Figure 22). By mid-2021, income reduction was lower in the region with better improvements in Honduras, El Salvador and Guatemala where almost 5 out of 10 households reported decrease in income (Figure 19). Figure 20. Share of people who stopped working across the sub-region, total El Salvador 54 Honduras 52 Dominican Republic 51 Guatemala 42 Costa Rica 36 0 10 20 30 40 50 60 % of people who stop working Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 48 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Figure 21. Share of people who stopped working across the sub-region, by sector 59 Honduras 54 20 55 El Salvador 57 38 54 Dominican Republic 42 20 44 Guatemala 59 32 34 Costa Rica 29 31 0 10 20 30 40 50 60 70 Services Industry Agriculture Figure 22. Share of people who stopped working across the sub-region, by employment type 50 El Salvador 67 Dominican Republic 51 55 Costa Rica 29 49 Honduras 56 48 Guatemala 43 42 0 10 20 30 40 50 60 70 80 Employee Self-employed Source: Khamis et.al. “The Early Labor Market Impacts of COVID-19 in Developing Countries: Evidence from High-Frequency Phone Surveys”. The World Bank, 202168. Not only did the employed stop working when COVID-19 hit, but some lost their jobs permanently. Guatemala is among the countries with the highest percentage of people who lost their jobs per- manently in the Latin American and Caribbean region during the first months of the pandemic. By summer 2020, when the third round of the HFPS was applied, fewer people reported losing their jobs permanently. This was partially due to fiscal stimulus packages, which were implemented to mitigate the impact of the pandemic, and partially to its reliance on remittances and exports69 resulting in a lower GDP contraction compared to the other Central American countries. Honduras, on the other hand, reported a higher number of people losing their jobs permanently in the third round of the HFPS, along with the Dominican Republic (Figure 23). In Honduras the industrial sector was heavily affected and in the Dominican Republic the service sector was most hit, led by tourism. In general, countries with higher GDP contractions also have higher levels of jobs losses (Figure 24). Emerging data confirms the deeply unequal impact of the pandemic with youth, lower skilled, and wom- en sustaining large losses, increasing their vulnerability further. Reported job loss shows stark differenc- es between men and women: while between 18% (El Salvador) to 25% (Guatemala) of women reported having lost jobs, this proportion ranged from 10% (Dominican Republic) to 14.5% (El Salvador) for men (Figure 25). Even after the first year of the pandemic, women still reported more job loses than men, in some cases even bigger than the first months of COVID-19 as in Honduras where it doubled from 19% 68 People who stopped working are those who said they were working pre-pandemic but were no longer working the week preceding the interview. 69 World Bank. WDI, Macro Poverty Outlook. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 49 Figure 23. Percentage of people who permanently lost their employment during COVID-19 quarantine 35 30 25 20 15 10 5 0 i le i na ico y bli n as r ica ia or ala bia ru do ua pu ca li v ad Pe ur Ch c aR em ex lom nt lva g Re ni Bo nd u ra ge M i Ec st m at Sa Pa Co Ho Ar Co Do Gu El Wave 1 Wave 2 Wave 4 Source: World Bank, High Frequency Phone Surveys, Washington DC. Wave 1 between May 21st and June 1st, Wave2 between June 19th and June 28th, and Wave 4 between June and July 2021. Figure 24. GDP reduction and percentage of people who lost their jobs 20 HND DOM 18 permanently lost their employment 16 14 % of people who CRI 12 SLV GTM 10 8 6 4 2 0 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 % GDP reduction Source: For GDP contractions, The Macro Poverty Outlook, April 2022, The World Bank. For percentage of people who permanently lost their jobs, World Bank, High Frequency Phone Surveys, Washington DC. Data for Wave 3 (between July 18th and July 26th) to 40% in mid-2021 (Figure 25). The disproportionate impact on women can be explained by two main factors: (i) employment predominantly in client-facing sectors and occupations, which were among the hardest hit; and (ii) increased childcare and household responsibilities, largely due to school and child- care closures. While the recovery is at full speed, with ongoing school closures and threats of new waves, women are slower to return to the labor market and some may exit the labor force permanently. Similarly, job losses were disproportionately large for youth,70 who also were more likely to be employed in the most affected sectors. During recovery, new entrants to the labor market face an even more uphill battle, having to compete with more experienced workers for limited jobs. Indeed, the share of NEETs is likely to have increased, especially young women, underscoring the importance of having effective support for school-to-work transitions. Finally, the low-skilled have also been disproportionately impacted, which is not surprising as they are more likely to be employed in the most-affected sectors and held jobs that were not amenable to work from home.71 In contrast, workers with college degrees experienced smaller job losses. As regular survey data collected during 2020 (and beyond) became available in select countries (e.g., Costa Rica), it confirmed the initial findings from the HFPS on the disproportionate LM impacts of the pandemic on different groups and provided a deeper look into the uneven recovery across sectors 70 Based on HFPS, on average more than 25% of young workers (18 to 24 years old) in LCR were left without work in the first two months of the pandem- ic, compared to 14 percent of workers aged 25 to 64 years old (World Bank 2021, Jobs Interrupted). 71 For instance, in El Salvador, college graduates and formal workers with a salaried job were those that were more able to work from home. Banegas and Winkler (2020) Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 50 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 as well as these groups. Box D presents a deeper look into the LM impacts in Costa Rica. The pandemic, in turn, also exacerbated the aforementioned labor market rigidities on work hours, dismissal costs etc. Figure 25. Percentage of People (18+) that report losing their employment, total and by gender Costa Rica Dominican Republic El Salvador 35% 35% 20% 30% 18% 30% 16% 25% 25% 14% 20% 20% 12% 10% 15% 15% 8% 10% 10% 6% 5% 4% 5% 2% 0% 0% 0% Wave 1 Wave 2 Wave 3 Wave 4 Wave 1 Wave 2 Wave 3 Wave 4 Wave 1 Wave 2 Wave 3 Wave 4 Guatemala Honduras 30% 45% 40% 25% 35% 20% 30% Male 25% Female 15% 20% Total 10% 15% 10% 5% 5% 0% 0% Wave 1 Wave 2 Wave 3 Wave 4 Wave 1 Wave 2 Wave 3 Wave 4 Source: World Bank, LCR High Frequency Phone Surveys (HFPS). Wave 1 was between May 21st-June 1st, Wave 2 was June 19th -June 28th, Wave 3 was July 18th-July 26th, Wave 4 was between June and July 2021. Limited availability of UI or similar automatic stabilizers in the sub-region to respond to systemic shocks, further exposed, especially low earners, to these negative impacts. Unemployment insurance schemes play an important role in protecting formal workers against jobs loss, and act as automatic stabilizers during times of systemic shocks.72 Absence of these schemes, as well as similar automatic stabilizers (e.g., guaranteed minimum income programs, poverty targeted social safety nets with good coverage etc.) in the sub-region made government response imperative. Even among formal workers low earners were more susceptible to job losses due to their work not being amenable to do from home, so the lack of such systems disproportionately affected lower earners. For instance, Delaporte et al (2020) show that only 7% of workers were able to work from home in Guatemala, mainly those in formal, high paying jobs. Many (previously non-poor) informal workers found themselves newly vulnerable and in need of support in the face of job loss and prolonged unemployment. The high levels of informality across the sub-region and lack of mechanisms to reach workers (and businesses) in the informal sector, created additional challeng- es in reaching the most affected, underscoring the importance of suitable instruments and mechanisms to act swiftly. 72 UI systems played a first response and automatic stabilizer role in EU countries where they are well-established with adequate coverage, in the 2008 global financial crisis, as well as the covid-19 shock. Isik-Dikmelik (2012), European Commission (2011). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 51 Box D. Unequal LM Impacts of COVID and early recovery in Costa Rica The outbreak of the COVID-19 worsened labor market outcomes in general in Costa Rica, with the worst impact being felt in the second quarter (Q2) of 2020. During Q2 2020, labor force par- ticipation and employment rate were in their lowest point since Q1 2019 (57% and 44%, respec- tively, what represents decreases of 9% and 21% compared to the previous quarter). In Q2 2020, almost 21% jobs were lost compared to the previous quarter (see Figure 1). The already high unemployment rate increased by 92% (from 13% in Q1 2020 to 24% in Q2 2020), More jobs were lost in the informal than in the formal sector, with the former suffering 38% decrease in jobs in Q2 2020.a As a result, informality rate reached its lowest point since Q1 2019 in Q2 2020 (26%). The service sector was the one hit the hardest, resulting in (temporary) shifts in the struc- ture of the economy. Costa Rica’s economy is concentrated in the service sector, primarily in wholesale and retail, real estate, and business activities, representing around 69% of total em- ployment before the pandemic. Approximately one fourth of these service sector jobs (17 of the total 69% that is employed in this sector in Q1 of 2020) were activities that were vulnera- ble to the restrictions brought on by the COVID-19 pandemic.b Indeed, the restrictions hit this sector particularly hard with an employment loss of 23% in Q2 2020, as compared to previous (pre-pandemic) quarter. The manufacturing and agriculture sectors were affected less in com- parison, but still experienced 16% and 13% jobs loss respectively compared to Q1 2020 (Figure Box 15). As a result, the structure of the Costa Rican economy has turned slightly more to the agriculture sector (1.2 percentage points increase in its share) but is still concentrated in ser- vices (68% of employed).The concentration of the youthc and the informal workers in the vul- nerable activities was a worrying signal that pointed to their susceptibility to a potential crisis or to restrictions that were put in place as a response to the pandemic. Although there were signs of recovery by the first quarter (Q1) of 2021, they have not reached their pre-pandemic levels. Although there was a recovery in the subsequent quarters, one year into the pandemic the number of people employed is still lower than the pre-pandemic value (1.95 com- pared in Q1 2021 to 2.18 million in Q1 2020), with 38,559 additional people in the working age popu- lation (3,983,867 in Q1 2021). While the unemployment rate came down from its peak of 24% at the worst of the pandemic (Q2 2020) to 19% in Q1 2021, it is still far from its pre-pandemic level of 13%. The informal jobs also started to recover, however not fully yet with 13% fewer jobs in Q1 2021 com- pared to Q1 2020. One year into the pandemic, the number of employees decreased in all sectors except fishing. However, the changes were such that employees are slightly more concentrated in potentially affected sectors, increasing the vulnerability of the structure of the economy in 2021. Figure Box 15. Employment loss, by sector, compared to Q1 2020 0% -5% -10% -15% -20% -25% 2020 Q2 2020 Q3 2020 Q4 2021 Q1 Agriculture Manufacturing Services Total Source: SPJ team based on Costa Rica Labor Market Survey, Q1 2019 up to Q1 2021. Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 52 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box D. Unequal LM Impacts of COVID and early recovery in Costa Rica (continued) The service sector was the hardest hit in terms of job loss, however the degree of impact dif- fered across sub-sectors. For instance, hotels and restaurants, and other community, social and personal services experienced larger job contractions (losing 44.7% and 42.3% of the total jobs respectively compared to 2020 Q1), whereas education escaped relatively unscathed at a decline of 4.2% in its employment (Table Box 1). Similarly manufacturing overall was moderately affected, but some sub-sectors, were hit much harder, such as construction which experienced a 27% contraction of its total employment as compared to Q1 of 2020 (Table Box 1). Table Box 1. Employment loss at detailed sector (1 digit ISIC) level, compared to Q1 2020 Activity Sector(1 digit ISIC) Changes in # employees compared to 2020 Q1 (%) 2020 Q2 2020 Q3 2020 Q4 2021 Q1 Agriculture; hunting and forestry -12.7% -11.4% -8.4% -1.7% Fishing -39.6% -37.4% -12.8% 1.3% Mining and quarrying 12.4% -57.9% -36.6% -36.6% Manufacturing -10.3% -23.4% -13.7% -12.1% Electricity; gas and water supply -2.3% -8.8% 0.4% -6.4% Construction -27.1% -14.4% -5.9% -6.0% Wholesale and retail -21.4% -19.0% -12.9% -6.1% Hotels and restaurants -44.7% -21.5% -17.1% -18.5% Transport; storage and communications -13.2% -6.6% -5.2% -5.9% Financial intermediation -14.7% -9.7% -10.6% -28.1% Real estate; renting and business activities -22.0% -7.1% -14.7% -15.5% Public administration and defence -8.6% -21.3% -3.0% -18.7% Education -4.2% -3.4% 0.5% -10.3% Health and social work -17.7% -9.2% -21.3% -19.7% Other community; social and personal services -42.3% -25.5% -23.7% -8.3% Activities of private household as employers -34.6% -33.9% -16.5% -14.4% Extraterritorial organizations and bodies -100.0% -100.0% -100.0% -100.0% Source: Source: Own elaboration based on Costa Rica Labor Market Survey, Q1 2019 up to Q1 2021. Self-employed were also deeply impacted, with some adapting strategies to cope or innovate. Due to COVID-19, 26% of the self-employed operated with some restrictions, and 48% reported a decrease in their monthly profits. In order to cope with the pandemic, 32% of those who hire employees made at least one adjustment with them, being workday reduction the most com- mon. On the other side, 11% of paid employees reported they started working remotely. Of the total self-employed, only 6% reported adopting strategies to innovate their businesses in face of the changes introduced by COVID-19. Implementing or increasing internet sales was the most common strategy adopted. Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 53 Box D. Unequal LM Impacts of COVID and early recovery in Costa Rica (continued) Figure Box 16. Employment loss at detailed sector (1 digit ISIC) level, compared to Q1 2020 Panel A. By gender Male 0.0% 0.0% -5.0% -10.0% -5.0% -15.0% -10.0% -20.0% -25.0% -15.0% -30.0% -35.0% -20.0% -40.0% -45.0% -25.0% 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Total Male Low education Medium education High education Female 0.0% 0.0% -5.0% -10.0% -5.0% -15.0% -20.0% -10.0% -25.0% -15.0% -30.0% -35.0% -20.0% -40.0% -45.0% -25.0% 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Total Male Low education Medium education High education Panel B. By age group Adults 0.0% 0.0% -5.0% -10.0% -5.0% -15.0% -10.0% -20.0% -25.0% -15.0% -30.0% -35.0% -20.0% -40.0% -25.0% -45.0% -50.0% -30.0% 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Total Adults Low education Medium education High education Youths 0.0% 0.0% -5.0% -10.0% -5.0% -15.0% -10.0% -20.0% -25.0% -15.0% -30.0% -35.0% -20.0% -40.0% -25.0% -45.0% -50.0% -30.0% 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 2020 2020 2020 2021 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Total Youths Low education Medium education High education Employment rate Participation rate (rha) Source: Own elaboration based on Costa Rica Labor Market Survey, Q1 2019 up to Q1 2021. Note: We compare employment and participation rates to the levels of 2020 Q1. Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 54 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box D. Unequal LM Impacts of COVID and early recovery in Costa Rica (continued) Women and the youth, who were already suffering from worst employment outcomes, were  nem- hit harder by the crisis but with notable exceptions, such as highly educated women.d U ployment rate increased by 14 percentage points among the youth (from 31% in Q1 2020 to 45% in Q1 2021), compared to 5 percentage points among adults (from 9 to 14%). In addition, by Q1 2021, the percentage of youths who are not working nor in school or training (NEET) increased from 17% to 25% compared to Q1 2020, with a NEET rate especially high for young women (31%). Among those employed, the percentage in time-related underemployment increased from 13% to 17%. In the case of women, unemployment rate increased by 8 percentage points (to 26% in Q1 2021), compared to 5 percentage points among men (to 14% in Q1 2021). The evolution of employment and participation rates were also worst for women and youths (see total evolution in Figure Box 16). If we disaggregate by education, those with lower education suffered a larger decrease in participation and employment rates, as they were employed in the sectors hardest hit by the crisis and had to take on additional responsibilities at home due to school closures. On the ither hand, women with high education were virtually not affected by the crisis, likely due to then nature of their work (concentrated in sectors not directly impacted by the crisis) and the ability to work from home. a Informality is based on legal characteristics: not-salaried, employer/salaried/self-employed without social security. b Sectors non-affected or non-vulnerable to the pandemic are the following: (i) Electricity, gas, steam and air conditioning supply; (ii) Water supply, sewerage, waste management and remediation activities; (iii) Public administration and defense, compulsory social security; (iv) Human health and social work activities; and (v) Activities of extraterritorial organizations and bodies. In all the other sectors the workers were vulnera- ble to the pandemic. c Youth considers those aged 15 to 24 years old. d Low education refers to: less than 9 years of education, medium: from 9 to 13 years of education and high: 14 or more years of education The Covid-19 pandemic brought to the fore, the increasingly pressing need to transition to jobs of the future (including green jobs) and develop the required skills. The Covid-19 pandemic exposed and deepened the vulnerabilities and inequities in the LM (e.g., acceleration of structural transforma- tion towards deindustrialization observed in some countries, the low productivity equilibrium in most countries etc.), which present a huge challenge a but also a potential opportunity to transition to jobs of the future. Evidence points to such crises leading to potential creative destruction, presenting an opportunity to transition to better, more productive jobs, through concerted policy action. Indeed, some countries were more vulnerable to COVID restrictions than others due to their sectoral com- position (e.g., Dominican Republic and Costa Rica were among the most vulnerable, while Honduras appeared to be slightly less so partly due to its still large share of employment in agriculture, which overall was expected to be weather the pandemic a bit better (Figure 26)with uneven recovery fore- casted across the different sectors (Figure A. 15 to Figure A. 17). The still large prevalence of jobs of the past across the sub-region (with the exception of Costa Rica and Panama), further deepened the initial shock, with a large swath of the population not having jobs amenable to work from home (i.e., the physically intensive jobs and jobs that do not require use of information/communication technol- ogy: features of jobs of the past) and thus suffering severe impacts. Indeed, those with jobs amenable to work from home were less affected by the pandemic; they tend to live in urban areas with higher earnings, and more likely to be higher skilled and employed in jobs of the future.73 The displacement of large swath of workers due to the crisis and the limited availability of skills in the labor force for jobs of the future, present an opportunity to develop and augment the requisite skills to support a transition toward better jobs, including green jobs. 73 World Bank (2021). Jobs Interrupted.; Winkler and Montenegro (2021); Banegas and Winkler (2020) Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 55 Figure 26. Structure of employment pre-COVID-19, by potential vulnerability to COVID-19 restrictions, by main sector Honduras Guatemala El Salvador Dominican Republic Costa Rica 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Agriculture- not vulnerable Agriculture- vulnerable Industry- not vulnerable Industry- vulnerable Services- not vulnerable Services- vulnerable Source: The World Bank – SEDLAC standardization (CEDLAS and the World Bank). Data for Costa Rica 2019, Dominican Republic 2019, El Salvador 2019, Guatemala 2018 and Honduras 2019. Data for Nicaragua not available. Note: To determine the vulnerable sectors/activities, first the sectors that are con- sidered not vulnerable as they are essential were identified. The sectors are: (i) electricity, gas and water supply, (ii) public administration and defense, (iii) health and social work, (iv) extraterritorial organizations and bodies. Then, the activities within the non-essential sectors were identified; these activities are not vulnerable as they are essential or were not affected with the mobility restrictions imposed to mitigate COVID-19. D1. Response of CA governments in SPL to COVID-19 emergency The Central American and Dominican Republic governments have responded to the crisis to con- tain the negative impacts, leveraging existing SPL instruments to varying degrees. In March 2020, all countries declared a national state of emergency (State of National Alert in Nicaragua) and installed needed containment measures that affected the continuation of labor and education activities of most households in the region. These measures affected not only the poor but also the vulnerable population, which is overrepresented in the informal labor sector and lacks social protection. To coun- teract the effects of these measures, governments have provided a range of responses to mitigate in- come loss through the provision of social assistance, social insurance and labor markets support. Ta- ble 1 provides an overview of the type of SPL instruments deployed by countries across the sub-region. Amid social assistance responses, cash transfers have been deployed across all countries in the re- gion, except Nicaragua, to help not only the poor but also vulnerable households mitigate income loss due to the COVID-19 Pandemic restrictions. While all cash transfers programs shared a common aim of providing income support and aimed at reaching new beneficiaries such as the vulnerable Table 1. Summary of SPL instruments utilized in response to COVID-19   Social Assistance Social Insurance Labor Market Policies and Labor Regulations Cash In Kind School Utility and Paid sick Healthcare Pensions SS wage training labor shorter Transfer / food feeding financial leave insurance contributions subsidies / measures regulations work time voucher obligation support waiver/subsidy soft loans waivers Costa Rica √ √   √     √ √ √ √ √ √ El Salvador √ √ √ √ √ √     √ √ √   Guatemala √ √ √ √         √   √   Honduras √ √   √   √   √ √ √ √   Nicaragua   √     √           √ √ Panama √ √   √         √ √ √ √ Dominican √ √ √ √ √  √     √   √ √ Republic Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 56 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 population affected by the containment measures, including informal workers, the scale and dura- tion of the benefits varied across countries. All programs mainly expanded horizontally (reaching new beneficiaries). The countries with the cash transfers of largest scale included Panama with Plan Pan- ama Solidario, eventually reaching almost 80% of the population. In Guatemala the coverage of cash transfers was expanded from 5 percent of households to 80 percent (166,000 to 2.7 million) following the establishment of the Bono Familia temporary cash transfer program74. El Salvador’s, “Compensa- tion Bonus” reached 77.6 percent of the population (1.5 million households). In the Dominican Repub- lic, “Quedate en Casa” reached nearly half of the population. With the exception of Panama where monthly transfers started April 2020 and are expected to end in June 2022, the duration of the cash transfers has been mostly temporary for approximately an average of 3 months except for El Salvador and Honduras75 which provided a one-off cash transfer. In the Dominican Republic transfers started in April 2020 and the Government gradually wound down the program in early 2021. Given the over subscription of the program in Guatemala, the third payment was reformulated to a smaller amount than originally established.76 The Dominican Republic also had a program for self-employed workers (Programa de Asistencia al Trabajador Independiente – PA’TI) that provided a monthly benefit from May-December 2020 to nearly 190,00 beneficiaries. Box E presents broad impacts of the crisis and the main elements of the SP response in Panama (with particular focus on Plan Solidario). Other common social assistance responses in social assistance were in-kind support to assist food security and utility waivers to ensure continuity of priority services. All countries provided in kind support to assist the food security of affected households mainly through the delivery of food pack- ages or vouchers (Guatemala). Guatemala also distributed school meals and study guides in commu- nities where schools were closed. Almost all countries in the sub-region provided utility waivers in varied forms. For example, in Costa Rica the government waived suspension of water services due to non-payment for at least 60 days and requested restoring services canceled due to late payment as a measure to prevent the transmission of COVID-19. In Honduras, the government declared a moratori- um on suspensions for telecommunications (telephones and internet) services during the COVID-19 pandemic, and flexible payment plans or free basic packages to ensure continuity of services. Measures under social insurance to formal workers were limited due to their reach and the lack of unemployment benefits in the sub-region. Given the large proportion of informal workers in the sub-region, social insurance measures reached few workers. The predominant measures were paid leave during the quarantine period followed by ensuring continuity of health insurance. In El Salva- dor, Honduras and Nicaragua paid leave during quarantine periods was granted to workers vulnerable to COVID-19 due to health risks who were sent home due to social distancing measures. Access to health insurance was continued for displaced and/or suspended workers affected by the pandemic in Honduras and El Salvador. Guatemala made a one-time transfer of US$256 to public sector pension- ers. No country in the sub-region has unemployment benefits. 74 Wold Bank (2022) Guatemala SCD Update. 75 Emergency cash transfer program focused on self-employed and informal workers receiving government support under the COVID-19 crisis. (Gentilini et al., 2021) 76 Payments were made between May and December 2020 and that the third payment was reduced because there was over-subscription and no change in budget allocation. Analisis del Fondo de Familia, December 12, 2020, CIEN, Guatemala Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 57 Box E. Zooming in on Panama: Main Elements of SP Response Panama is the second most affected country from the COVID-19 pandemic and associated downturn within the Latin America and Caribbean region. In the LCR, Panama’s 17.9% contrac- tion in GDP was surpassed only by St. Lucia. Industry declined by 32.1%, followed by services with a reduction of 12.7%.a Tourism, transport, and communication services were affected by the closure of borders and downsizing of global commerce. Construction and mining activities closed when mobility restrictions were put in place. Women were disproportionately affected, as they tended to work in the hardest hit sectors. Due to the economic impact that Panama suffered from the pandemic, the unemployment rate in- creased to 18.5% in 2020, worsening the gender gap: the unemployment rate for men reached 12.5% while for women it rose to 24.7%.b This was largely due to the fact that a larger share of women was employed in services, which were more heavily affected by COVID-19. Indeed, before the pandem- ic, women’s participation in activities that require personal contact, such as hotels and restaurants, teaching, and financial and insurance activities, was significantly higher than men (Figure Box 17). Figure Box 17. Share of employment women and men in service sector activities affected with mobility restrictions 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Other service Financial and Hotels and Teaching Activities of households activities insurance restaurants as employers; activities undi erentiated activities of households Female Male Source: Own elaboration using Encuesta de Mercado Laboral 2019 Note: These activities represent more than 1/3 of the service sector. Informal workers were particularly vulnerable to COVID-19 impacts. Panama’s overall informal- ity rate, before COVID-19, was 49.1%. Activities heavily affected by COVID-19, such as mining, construction, manufacturing, industry, services, and transportation, employ unskilled workers and had higher informality rates ranging from 53 to 57% (Figure Box 18). In 2021, the labor market started to recover as half of the suspended employees returned to work and some economic activities started to resume. Still, the more vulnerable groups remained unemployed or sus- pended, deepening their vulnerability.c The Panamanian government implemented measures to mitigate the effects of COVID-19 on households leveraging its social protection system. Before COVID-19, Panama had five main social protection programs (Red de Oportunidades - RdO, SENAPAN, Angel Guardian - AG, 120 a los 65 – 120, and Pase-U).d The first four are managed through the National Social Registry. Pase-U (previously called Beca Universal) has high coverage in the poorest quintile, RdO and 120 only reach 20% of the poorest population. The other programs, SENAPAN and AG, are smaller reaching at most 5.3% of the poorest quintile. Still, most of the beneficiaries of the CCT programs, RdO and SENAPAN, belong to the first quintile. AG and 120, which are targeted to extreme poor and poor, show leakage in richer quintiles; and Pase-U, as a universal scholarship doesn’t target the poor population per se, but public and low-cost private schools are mainly attended by children from the poorest quintiles (Figure Box 19 and Figure Box 20). Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 58 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box E. Zooming in on Panama: Main Elements of SP Response (continued) Figure Box 18. Informality rate by type of activity Farmers and Agricultural, Forestry, Fishing and Hunting Workers 83.8 Cra smen and Workers in Mining, Construction, Manufacturing, Mechanics and Related Occupations 56.7 Unskilled Workers in Services, Mining, Construction, Manufacturing Industry, Transportation 53.2 and Other Elementary Occupations Operators of Fixed Installations and Machines; Assemblers,Drivers and Operators 47.0 of Mobile Machinery Total 41.9 Service Workers and Vendors in Shops and Markets 40.2 Technicians and Mid-level 14.3 Professionals Professionals, Scientists 14.0 and Intellectuals Directors and Managers of Public and 12.2 Private Sectors and Social Interest Organizations O ce Workers 7.7 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 Note: Informality based on legal characteristics: not-salaried, employer/salaried/self-employed without social security. Source: Own elaboration using Encuesta de Mercado Laboral 2019 In addition to protecting the poorest through existing programs, the government created a temporary program called Plan Panama Solidario (PPS) to cover the newly poor and vulnera- ble, with innovations in its implementation. The program was delivered in two different ways: (i) digital vouchers targeted to the poor and vulnerable population in urban areas, informal workers, formal workers with low salaries (less than $11,000 per year according to the 2019 income declaration) and formal workers suspended due to the pandemic countrywide; and (ii) food baskets and vouchers targeted to the poor and vulnerable population living in rural and indigenous areas where digital vouchers were not able to be implemented. Beneficiaries were identified through an innovative approach: matching different administrative databases to create the roster of beneficiaries. For this, the government of Panama had 3 different rosters with potential beneficiaries from which they validated and filtered information. The first list was salaried workers who were suspended because of the pandemic.e The second list included po- tential beneficiaries of affected economic activities provided to the National Authority for Gov- ernment Innovation (AIG, acronym in Spanish), by agencies that collect this information.f AIG validated the information with the Electoral Tribunal and excluded those included in rosters of government employees, private employees and retirees and pensioners. The third list consisted of potential beneficiaries based on their geographical area of residence. The eligible areas were determined by inter-ministerial coordination including different corregimientos. The AIG built the roster with information from the Electoral Tribunal and excluded individual from the same rosters used to filter the second list, plus rosters of suspended salaried workers, people with more than $11,000 in the 2019 income declaration and prisoners.g Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 59 Box E. Zooming in on Panama: Main Elements of SP Response (continued) Figure Box 19. Coverage of main programs by quintiles 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Red de SENAPAN 120 a los 65 Ángel Guardián Beca Universal Oportunidades Q1 Q2 Q3 Q4 Q5 Figure Box 20. Beneficiaries distribution of main programs by quintiles 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Red de SENAPAN 120 a los 65 Ángel Guardián Beca Universal Oportunidades Q1 Q2 Q3 Q4 Q5 Source: ASPIRE performance indicators. Note: Based on Encuesta de Mercado Laboral for 2019. Quintile distribution based on pre-transfer welfare. In 2020, PPS reached 7 out 10 Panamanians and 74.6 percent of those under the poverty line of 5.5 dollars (2011 PPP) a day. The middle and bottom 40 percent income groups had highest coverage (80.7 and 77.1 percent, respectively). The program covered more than 73 percent of those in informal employment, urban and indigenous populations. The program had a positive welfare effect in 2020 by reducing poverty by 4 pp (from 18.8 percent expected without mitiga- tion measures to 14.8 percent).h The government extended PPS through 2021 and introduced some reforms to reduce leakages and support economic recovery. In January 2021, the government introduced additional eligi- bility criteria, reducing the benefit to only one eligible member per household, excluding public sector employees, and individuals with properties worth more than $250,000 or those living in households with monthly energy consumption above 400KW. As a result, coverage of the rich- est quintile dropped from 49.3 percent to 28.8 percent, and from 70.4 percent to 45.9 percent in the second richest quintile. At the same time, 73.3 percent of the poor are still covered (relative to 74.6 percent in 2020), and 70.8 percent of the bottom 40 is covered (relative to 77.1 percent in 2020). In July 2021, the government introduced a conditionality for all beneficiaries of digital food transfers. Every month, starting in September 2021, each beneficiary should register and complete 24 hours of community worki or 10 hours of an on-line course in the ESTER learning platform, managed by the National Institute for Professional Training for Human Development (Instituto Nacional de Formacion Profesional y Capacitacion para el Desarrollo Humano INA- DEH).j In September 2021, 80 percent of beneficiaries opted for monthly on-line courses. Box continues on next page Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 60 Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 Box E. Zooming in on Panama: Main Elements of SP Response (continued) The size of the PPS transfer represented 10.3% and 10.5% of beneficiaries’ income in 2020 and 2021, respectively. This increase was a result of raising the transfer amount in late 2020 and 2021 and improving the targeting mechanism.k For the poor, the transfer amount would have represented 33.5 percent of their income in 2020 and 40.5 percent in 2021. For the bottom 40 percent Panama Solidario would have represented 23.6 percent in 2020 and 25.8 percent in 2021.Compared to Panama’s main social assistance programs, PPS represents almost as much as the share that the program 120 a los 65 represents of poor beneficiaries’ income (32.6 vs. 33.5%) while other programs have a lower benefit adequacy (ASPIRE). The total cost of PPS up to August 2021 was sizeable, reaching around 3.5% of GDP, re- flecting the severe impact of the pandemic in Panama. The total PPS cost up to August reached US$1.8M (without administrative and logistic costs). The annual fiscal cost (April 2020-March2021) totaled $1.3M which represented 2.5% of Panama’s GDP.l Compared to the region, the size of this intervention is higher than the emergency package measures in Hondu- ras (2.2% of GDP) and Costa Rica (1.4% of GDP), but lower than Guatemala (3.4% of GDP) and El Salvador (15.5% of GDP).m The government announced the extension of digital PPS vouchers until June 2022. a The World Bank, Renewing with Growth. 2021 Semiannual report of the Latin America and the Caribbean region. March 2021. b Before COVID-19 hit, unemployment rate was 7.1% in Panama, higher for women at 8.8% as compared to 5.8% for men. Banco Mundial. Expansion de la proteccion social durante la pandemia en Panama. Principales características del Plan Panamá Solidario y su impacto en mitigación de pobreza e inequidad. September 2021 (not published). c Banco Mundial. Expansión de la protección social durante la pandemia en Panamá. Principales características del Plan Panamá Solidario y su impacto en mitigación de pobreza e inequidad. September 2021 (not published). d Red de Oportunidades is a CCT targeted to extreme poor households. e This roster was provided by the Ministry of Labor and validated by the National Authority for Government Innovation (AIG, acronym in Spanish) using information from the Electoral Tribunal to verify their identity. f These potential beneficiaries included affected groups such as, lottery ticket sellers, artisans and artist, peddlers, people with disabilities, trainers, people engaged in tourist activities, athletes, people working at the airport. g Banco Mundial. Expansion de la proteccion social durante la pandemia en Panama. Principales características del Plan Panamá Solidario y su impacto en mitigación de pobreza e inequidad. September 2021 (not published). h Ibid. i The Ministry of Social Development (MIDES), in coordination with local councils, is managing the identification of existing community work in the areas of the where Vale Digital is implemented. AIG developed an instrument to assign each beneficiary to an identified community work, giving priority to the area selected by the beneficiary of the place of residence. j There are 10 on-line courses for Vale Digital beneficiaries in ESTER platform of INADEH, including personal and job soft skill training, introduc- tion to entrepreneurship, introduction to English, introduction to MS Excel, hospitality training, car preventive maintenance, and basic technics of food preparation. k Digital and physical vouchers increased from $100 to $120 in 2021, and the amount of the food baskets increased from $15 to $25 in 2020’s last quarter. l Banco Mundial. Expansion de la proteccion social durante la pandemia en Panama. Principales características del Plan Panamá Solidario y su impacto en mitigación de pobreza e inequidad. September 2021 (not published). m The World Bank, Macro Poverty Outlooks. April 2021 for Guatemala and El Salvador. Labor market reactivation measures included support to micro, small and medium enterprises (MSMEs) by providing access to credits with soft conditions not only to formal but also informal enterprises. Interventions to support businesses included providing soft credits for working capi- tal and/or capital and credit guarantees to encourage commercial lending. El Salvador established the FIREMPRESA program for US$600 million dollars of which US$360 million has been allocated to provide soft credits to formal MSMEs with credits up to US$ 125,000, and US$100 million to support informal businesses with credits up to US$ 15,000; the remaining funds are allocated to wage subsi- dies. Panama created the “Panama Agro Solidario” program for US$300 million to support small and medium agricultural and livestock producers with zero-interest loans up to US$ 100,000. Honduras and Costa Rica implemented a guarantee scheme that ranged between US$200 and US$300 mil- lion dollars to encourage commercial lending to SMEs by guaranteeing potential losses on new SME loans and complementary measures such as adapting the conditions of existing credits through debt restructuring. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Central American SPL systems faced persistent structural challenges that deepened in the aftermath of COVID-19 61 Wage subsidies were also widely used. Guatemala, Dominican Republic, Honduras and El Salvador approved wage subsidies to retain formal workers or provide income support to formal workers, whose contracts had to be suspended due to the COVID-19 pandemic. Guatemala approved a wage subsidy fund “Fondo para la Protección de Empleo” of US$ 260 million to support formal workers in the pri- vate sector whose contract has been suspended (75 quetzals or 9.7 dollars per day)77. The Dominican Republic through the “Fondo de Asistencia Solidaria a Empleados” (FASE) supported approximately 906,000 laid-off workers (expanded to include workers kept on the payroll by companies).. The benefit was temporary established on April 2, 2020 for 60 days and was extended until December of that year and subsidized part or all of the workers’ wages, depending on the salary level and other factors, al- though this cannot be less than RD$5,000 (US$92.00) nor greater than RD$8,5000 (US$157) per month. Total spending was US$353.2 million78. El Salvador within the FIREMPRESA fund provided a two-month wage subsidy of up to 50 percent of the total wage expenditures to encourage workers’ retention79. In Honduras, a one-off income tax (10 percent of salary expenses) was approved for businesses maintain- ing pre-crisis employment level. Training was also part of the reactivation measures and in some initiatives was coupled with social assistance support. Costa Rica launched the Protect Skills Plan, a scholarship program for online training promoted by the government to reach 50,000 beneficiaries of the Protect Bonus program. Panama partnered with the virtual learning platform Coursera to provide free access to training to more than 250,000 independent workers, entrepreneurs and students. Across the region, labor regulations were adapted and remote or tele work was introduced or pro- moted in most countries. The COVID-19 pandemic laid bare the need to establish more flexible work- ing arrangements to ensure continuity of work. Per available information, tele work was introduced or promoted in El Salvador, Guatemala, Nicaragua, Panama, and the Dominican Republic. 77 Gentilini et al. (2021). 78 World Bank (2021) “Respuesta de Proteccion Social a la Crisis de COVID-19 en la Republica Dominica” 79 https://www.asamblea.gob.sv/node/10277 62 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy 63 III The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy Strengthening SPL Systems is a crucial investment for “building back better,” with a double dividend of better protection during recovery and increased preparedness for future shocks. Social safety nets and labor market programs play a critical role to alleviate poverty; to manage against risk and build the resilience of the chronic poor and vulnerable (including the potentially vulnerable emerging “missed” middle); and to preserve and build human capital. They can improve income, productivity and self-re- liance by promoting access to better jobs. In the post-COVID-19 context, strengthening SPL systems will not only help the vulnerable population continue to cope with the negative effects of the crisis, protecting their gains on human capital accumulation and supporting them to return better prepared to the job market, it will also ensure that they are better prepared to buffer the impacts of any future shock in a more effective manner. Investment is especially important in the context of low to moderately adaptive SPL systems of the sub-region and increased exposure to shocks. SPL systems play a critical role in shock-responsive- ness, and the COVID-19 pandemic brought to the fore the need for improving their adaptiveness. SPL systems have mixed levels of readiness in the sub-region. For example, Costa Rica is better prepared: a moderate reach of targeted programs; solid delivery systems with an established social registry linked with other institutions; and improved capacity in adjusting labor market programs, especially training to labor market needs. Guatemala is less prepared: low coverage of targeted programs; a nascent but rapidly developing SP delivery system; and limited labor market programs and institutional capacity. In addition, as already mentioned, Central American is increasingly more exposed to shocks (including climate induced) and natural disasters with significant impacts for the poor and vulnerable population and the economy.80 Poor people are disproportionately impacted from shocks, as they tend to live in places that are more exposed to climate risk, and other shocks, and they do not have the means to self-insure. As such, experiencing a shock (idiosyncratic or systemic) not only means a loss of current 80 Two major hurricanes hit Central America in the midst of COVID-19 Pandemic, affecting the lives of more than 8 million people with economic losses reaching in tens of billions of dollars. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 64 The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy Figure 27. Overview of Policy Directions Improve e ciency and e ectiveness of social protection spending and Improve e ciency of social Carefully manage the Continue to build and Ensure the right mix of Rebalance the pension ensure right mix assistance spending with a transition away from protect human capital social protection schemes to boost fiscal of programs to shi away from less emergency programs to through adaptive and instruments, in particular sustainability protect and build progressive programs minimize any negative nutrition-sensitive social establishment of automatic resilience of the impacts on the poorest assitance programs, stabilizers, and enhace poor and (newly) incorporating behavioral the adaptiveness of vulnerable elements targeted programs Improve adaptiveness of the Social Protection and improve the Continue Develop social Increase the Strengthen Increase the Intensify the Leverage Improve Labor system and adaptiveness investing in the registries and adoption of grievance integration of use of administrative targeting further invest in of development rely on other digital payment redress social technology data through accuracy by delivery systems the SPL and strengthen- administrative methods and mechanisms information beyond interoperabili- using big data for increased system ing of social databases for account systems payments and ty/integration and machine preparedness in and increase protection deliv- crosschecks to ownership across the with social learning (with light of more preparedness ery systems, assess needs di erent protection caveats) frequent shocks for external to make them and conditions phases of the information shocks more inclusive, and determine delivery chain systems dynamic and eligibillity adaptive (Intro- ducing open and continuous registration) Strengthening labor market institutions and services to support people back to Be er understanding and Promote private sector Integrated labor demand Institutional strenghthening Promote labor market monitoring of labor demand, led/provide training to and supply programs that to improve labor reforms that support the (productive) improve employability and contribute to be er intermediation, supported creation of be er jobs including through use of work real-time information, for quality of matches, employment outcomes for by technology (digitization e ective (re)emplyment increasing the change of vulnerable groups. of employment services, policies long-term success adequate budgets) income/welfare for the poor, but also affects their human capital, with long-term implications for their productive capacity and well-being. In this context, investing in strengthening and improving the adaptiveness of SPL systems in the sub-region is a necessity, and a smart investment. Moreover, particular attention will need to be paid to more vulnerable and historically excluded groups, such as Afrodescendants, Indigenous People, women, people with disabilities, and in some countries, migrants, to make the recovery truly inclusive and resilient. Though data limitations do not permit an in-depth analysis of the situation in Central America, evidence from LCR indicate that peo- ple with disabilities are more likely to be poor, to live in slums, and have limited access to internet or accessible transport, which indicates they were among the most affected from Covid-19. In addition, living in rural areas or being an ethno-racial minority or women further compounds their probability of living in poverty.81 They are less likely to participate in the labor force, and when they do participate, they have lower quality jobs.82 Similar outcomes are observed for Afrodescendants.83 In Costa Rica, 81 World Bank (2021). Disability Inclusion in Latin America and the Caribbean. 82 Household head with disability almost twice as likely to be inactive as compared his/her peer without disability. Workers with disabilities face higher rate of informality and lower earnings. Ibid. 83 While there were significant gains in the past decades, Afro-descendants are still more likely to be chronically poor (2.5 times) and are the most underrepresented minority in decision-making positions. Freire et al. (2018). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy 65 poverty rates for Nicaraguan migrants exceeded 50 percent during the COVID-19 pandemic84. All of which point to the importance of thinking through inclusive social protection policies that take into account differentiated needs and barriers of vulnerable and excluded groups for inclusive growth. So- cial Protection programs and policies aim to promote inclusion by design, nevertheless data limita- tions may constrain identification and incorporation of the needs of some vulnerable groups, pointing to a requirement for disaggregated information. The pre-existing and emerging challenges, as well as global knowledge point to 3 policy directions for SPL Systems in CA countries towards a resilient and inclusive recovery and beyond. These are: (i) improve efficiency and effectiveness of SP spending and ensure right mix of programs to protect and build the resilience of the poor and (newly) vulnerable (e.g. shift away from weakly targeted categorical programs, such as subsidies which also would support greening efforts; expand coverage of well-tar- geted programs, in most cases with the double dividend of human capital accumulation; transition away from emergency programs while minimizing negative impacts on the poorest; and rebalance pension schemes to boost fiscal sustainability, etc.); (ii) improve adaptiveness of SPL systems and strengthen investments in delivery systems (e.g. including expansion of social registries, supporting dynamic inclusion, moving towards digital payments and promoting account ownership, etc.); and (iii) strengthening employment services and institutions to support people to return to (productive) jobs quickly. Figure 27 provides a summary of the specific policy directions under each broad theme, while the rest of the section presents further details drawing on global knowledge and experience. A1. Improve efficiency and effectiveness of SP spending and ensure right mix of programs to protect and build resilience of the poor and (newly) vulnerable In light of the weak fiscal situation across the sub-region, improving the efficiency of social assis- tance spending will be key, with a shift away from less progressive programs (including subsidies, which in turn can contribute to greening efforts). The additional fiscal constraints many countries face in the aftermath of COVID-19 underscore the importance of improving quality of spending in regular social assistance (i.e., beyond emergency measures), and there seems to be room to do so. While this will take different forms in different countries, it broadly implies a shift in the composition of programs, away from those less progressive (such as energy subsidies) into better targeted (such as poverty targeted programs with the double dividend of supporting human capital accumulation of the most vulnerable), and/or improving the targeting of large categorical programs (e.g., in-kind trans- fers, etc.). Considering the still non-negligible prevalence and regressivity of subsidies (in particular, electricity) in most countries in the sub-region (except Costa Rica)85, it will be important to improve their distributional impact, by shifting away from eligibility based on usage towards eligibility based on socio economic indicators as a first step, which would contribute to greening efforts (by eliminating incentives for excess usage and using a portion of savings to invest in greener technologies). Ideally these subsidies would be streamlined into the SA system to take advantage of existing mechanisms to identify the poor and vulnerable and limit leakage to the top quintile. For instance, in El Salvador, microsimulations indicate that if gas, electricity, and water subsidies used the same targeting meth- odology as the CCT programs (Comunidades Solidarias Rurales y Urbanas), this could potentially lead to a reduction of more than 1 percentage point in the poverty rate,86 even when using only 43 percent 84 World Bank. A Lost Decade for Poverty and Inequality Reduction: Costa Rica Poverty Assessment (forthcoming) 85 World Bank (2017) shows that the electricity subsidies in Panama are the most regressive in the region, followed by Honduras and Nicaragua. El Salvador’s subsidies are the least regressive in the region followed closely by those of Guatemala. The least regressive electricity of EL Salvador in fact has almost a uniform distribution, thus is far from being even mildly progressive. 86 2011 PPP $ 3.2 a day line Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 66 The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy of the fiscal resources.87 A portion of the fiscal savings can be applied to expand the very limited cov- erage of the CCT programs in El Salvador to protect and support the poorest and most vulnerable and their human capital accumulation through recovery and beyond, which would improve the overall progressivity and result in more efficient spending. While the political economy of such reforms can be complex, when social protection instruments are in place, they can enable reforms of oil, carbon or utility subsidies. Indeed, there are successful examples around the globe: Ukraine and Indonesia both managed such transitions while compensating the poorest and most vulnerable through the SA system.88 Such a reform of the electricity sector is being considered in the Dominican Republic. It will be important to carefully manage the transition away from emergency programs to minimize any negative impacts on the poorest. With the recovery underway in all countries (though to differing degrees), a shift in policies is required which also brings on new challenges. Moving away from emer- gency programs, that have largely served well to contain the negative impacts of the crisis on the most affected, is needed, as the expansion was temporary in nature and in most cases fiscally not sustain- able. Also, program design would be expected to evolve from the emergency context (“fast, large, and flat”). At the same time, such transition will require careful planning; the challenge will be to balance the continued protection of the poorest as fiscal space shrinks. Some countries in the sub-region have transitioned out of the bulk of emergency programs, as of summer in 2021, somewhat abruptly, as economic activity started to return to normal. For instance, Costa Rica accepted applications for the last payment for its emergency program Bono Proteger in December 2020. While as of summer 2021 an official announcement has not been made about the termination of the program, it is in effect eliminated. In contrast, given the depth of impact of COVID-19 in Panama (as outlined in Box E), the government extended its emergency program Panama Solidario through 2021, with a further extension until June 2022 for certain benefits, while introducing some reforms to reduce leakages and support economic recovery. In early 2021 the Dominican Republic gradually reduced the number of beneficia- ries and transfer amount for the emergency program (Quedate en Casa), before ending the program in April and transitioning 500,000 households categorized in the bottom range of the PMT index to the reformulated CCT program (SUPERATE). Similarly, continuing to build and protect human capital through adaptive and nutrition-sensitive social assistance programs, incorporating behavioral elements will be critical for the medium-term. The impact of social assistance programs broadly, and in particular CCTs, on human development outcomes is rigorously and widely documented across the globe.89 Their ability to prevent and mit- igate negative and long-lasting impact that shocks have on human capital formation (e.g. through prevention of bad coping mechanisms, such as pulling children out of school, decrease in dietary diversity/nutrition etc.) is also extensively documented. In addition the social assistance programs are increasingly incorporating nudges or co-responsibilities to provide incentives to use services provided by other agencies, to further support human capital accumulation and overall household welfare. Giv- en the disproportionate impact of the pandemic on human capital and welfare, especially of the poor- est and vulnerable, it will be imperative to continue to build and protect human capital through social assistance programs, incorporating nutrition-sensitive approaches or behavioral elements (nudges). This will imply, continuing to invest in and ideally expanding the well targeted CCTs across the sub-re- gion. In addition, incorporation of behavioral elements in particular for early childhood development have been shown to be successful and can be considered in most cases. For instance, a pilot interven- tion to improve development and nutrition of young children in poor rural areas in Guatemala, through home visits and monthly group sessions with parents, showed highly significant effects on children’s fine motor skills, language and cognitive development, improved parenting practices and increase in 87 Background analysis for SLV SCD Update by POV and SPJ teams. 88 Citation for Ukraine and Indonesia cases 89 For instance, there are more than a hundred of peer reviewed journal articles documenting the impacts of just one CCT program (Mexico’s Progresa which was among the first CCTs across the globe), and hundreds more that document CCTs impact on various human development outcomes extensively. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy 67 play activities. Similar behavioral elements are combined with social assistance programs in other parts of the world, with significant impacts (e.g., Brazil Feliz Crianza program), which can serve as a model for the countries in the sub-region. Ensuring the right mix of social protection instruments, and in particular establishment of automatic stabilizers, and enhancing the adaptiveness of targeted programs, further increase resilience and preparedness against a future shock (including climate shocks). Automatic stabilizers are the first line of response in times of systemic shocks across the globe. In addition to unemployment insurance schemes, social programs targeted based on economic need can act as automatic stabilizers, espe- cially when they reach a good share of the poor and vulnerable with the flexibility to expand. Having a carefully designed unemployment benefit system in place will allow income-smoothing for unemploy- ment risk at the household level for formal workers, in particular the low earners that were shown to be more vulnerable. While this would only cover a portion of the workers in most case due to high levels of informality, it may be appropriate for countries where share of formal employment is significant even among the bottom of the distribution, such as Costa Rica and Dominican Republic (insert ref to Figure 8, employment and informality bottom 40 etc.). Improving adaptiveness of the targeted social programs so that they have the ability to expand to protect the existing poor (i.e., vertical expansion) and cover the newly vulnerable (i.e., horizontal expansion) would also help improve preparedness and resilience. Improving adaptiveness could take the form of incorporating new design elements towards diversification of income sources and support return to employment, broadly economic inclusion (EI) approaches that are gaining increasing prominence in many countries around the world, including in the LCR. Recent evidence on EI programs points to promising and sustainable impacts, with benefi- ciaries saving, investing in productive assets, and earning more than they would have otherwise.90 In addition, triggers and emergency protocols that are embedded in the implementation rules of the pro- grams allow rapid scalability for potential expansions. For instance, social pensions in Mexico include emergency protocols in its program guidelines that allows for provision of one additional payment in case of emergencies (official declaration by competent state authority), which allows for flexible vertical expansion. Similar emergency protocols can be considered for the targeted programs in the sub-region, especially given the increased prevalence of climate related shocks/disasters and their disproportionate impact on the extreme poor which are more likely to reside in more areas exposed to such shocks. For example, Guatemala’s Bono Social provided top-up payments to CCT families in areas affected by several natural disasters. Honduras implemented a similar measure to respond to recent hurricanes. The recent reformulation in the Dominican Republic of the CCT (SUPERATE) includes in its foundation decree the delivery of emergency payments in response to shocks. Finally, an extensive agenda awaits in the sub-region to undertake the required rebalancing of the pension schemes to boost fiscal sustainability. Pension subsidies to the high- and middle-income populations will have to be reduced by applying maximum benefit ceilings, eliminating early retire- ment options and decreasing pension accrual rates. Expensive minimum pension guarantees also may need to be lowered in some countries, or at least offered at older ages. Contribution rates might also need to go up, or alternative streams of funding will need to be found. Finally, governments should make their debt issuance more diversified and less reliant on pension fund purchases. Building politi- cal support for such reforms is not easy but could be helped by clearly exposing intra- and intergener- ational unfairness of current system of hidden pension subsidies. A postponement of needed reforms would only make eventual changes more drastic and even less socially sustainable. Considering the vast needs, the challenges of the pension schemes in the sub-region as well as specific policy recom- mendations will be analyzed in upcoming work. 90 Andrews et al. (2021). The State of Economic Inclusion Report 2021: The Potential to Scale. Washington, DC: World Bank. Dominican Republic, Costa Rica, and Honduras incorporate productive inclusion approaches into their social assistance programs, with Costa Rica more advanced in its approach with an integrated package targeted to extreme poor, though with low coverage. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 68 The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy B1. Improve adaptiveness of the SPL system and further invest in delivery systems for increased preparedness in light of more frequent shocks The COVID-19 shock reinforced the need to improve the adaptiveness of SPL systems, which is par- ticularly imperative in Central America given its increased exposure to natural disasters and other shocks. An effective and adaptive/resilient SPL system helps mitigate the impacts on the most vul- nerable in the face of crisis, helping to protect people’s human capital and increasing the resiliency of households to shocks, while helping to prepare affected workers to return to employment and self-re- liance during the recovery. Considering the increased exposure of the Central American countries to natural disasters and other shocks, improving the adaptiveness of SPL systems is imperative for better preparedness and increased resilience of vulnerable in the sub-region. To this end, it is critical to continue investing in the development and strengthening of social pro- tection delivery systems, to make them more inclusive, dynamic and adaptive. The COVID-19 crisis underscored the importance of delivery systems, which enabled the fast delivery of emergency inter- ventions. Lessons on delivery systems adaptations and innovations that were introduced and tested globally as part of the emergency response are emerging and provide policy directions on future nec- essary reforms and investments. Introducing open and continuous registration will contribute to building more inclusive and dynamic delivery systems.91 During the COVID-19 pandemic, several countries introduced on-demand intake and registration approaches, introducing on-line or phone applications. Thailand received around 24 million applications from informal sector workers and farmers, more than half of the working age population. Within a few weeks, more than 6 million on-line applications were received in South Afri- ca. Argentina only used online registration for the Ingreso Familiar de Emergencia, and Brazil allowed families not registered in Cadastro Unico to submit an online application for the emergency cash transfers for informal sector workers and formal self-employed (Auxílio Emergencial). In Chile, people could apply for the Ingreso Familiar de Emergencia (IFE) online as well as in-person at ChileAtiende, the government’s multi-service centers.92 Central American countries could benefit from introducing on-demand registration as well as several registration modalities, including in-person at government offices, online and by phone. Countries that tested these approaches during the pandemic, such as Costa Rica and the Dominican Republic, could build on this experience and introduce on-demand registration for regular social protection programs, while other countries could gradually start incor- porating on-demand approaches. Developing dynamic social registries and relying on other administrative databases for cross-checks can be a powerful tool to assess needs and conditions and determine eligibility for social protection benefits and services. During the COVID-19 pandemic, several countries relied on existing social reg- istries, often cross-checking data by querying various databases. This approach was used in many Latin American countries, including Brazil, Chile, Colombia and Ecuador, as well as in Cambodia and Pakistan. Cross-checks with administrative databases were implemented in several countries, to com- plement data from social registries. This approach was often used to reach informal sector workers 91 Administrator-driven approaches, relying on infrequent census-sweeps, are generally chosen in contexts of capacity and financing constraints, and when a country first sets up social protection programs. This approach makes sense as a practical solution to challenges associated with a high degree of asymmetric information (or lack of data), weak administrative capacity (or low confidence in government institutions), or remote populations with little access to institutions or government services. While the administrator-driven approach tends to be the starting point for many countries, it is important to/systems tend to evolve to an on-demand or mixed approach as capacity and financing allows. With on-demand approaches, people proactively apply for one or more benefits and services, and the timing of their application is largely dictated by their own circumstances. Intake can take place at the house- hold or at a government office, it can rely on paper or digital forms (both offline and online), and the process can be carried out by the administrator (such as a social worker or caseworker) or it can be a self-service process. In practice, most countries use a combination of these modalities. For more on this topic, please refer to Lindert, Kathy; Karippacheril, Tina George; Rodriguez Caillava, Inés; Nishikawa Chávez, Kenichi. 2020. Sourcebook on the Foundations of Social Protection Delivery Systems. Washington, DC: World Bank, Chapter 4. 92 Rodríguez Caillava, Inés; Blomquist, John. “Delivery Systems Innovations in The Context Of The Response To Covid-19”. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy 69 by verifying that they were not formal sector workers found in tax and social insurance databases. In Turkey, where the social assistance information system used more than two dozen databases even before the pandemic, this exercise was relatively straightforward. In Pakistan, the NSER social registry was used in combination with other administrative databases, able to exclude the well-off, based on a list of wealth proxies that functioned as exclusion criteria. In Central America, Guatemala is currently working to develop a social registry, and could consider incorporating data validation through webser- vices. Countries with existing social registries, such as El Salvador and Honduras, could benefit from investing in increasing their coverage and performing cross-checks with other administrative databas- es through interoperability. The Dominican Republic could benefit from higher dynamism and efficien- cy by implementing the planned reforms that will lead to SIUBEN+, an integrated social information system that would be updated frequently through interoperability with other government databases, and would be able to monitor not only the demand but also the supply of social programs. In the case of Costa Rica, the government could leverage the lessons learned during the emergency response and introduce more dynamism to SINIRUBE through on-demand registration and interoperability with other government databases. Increasing the adoption of digital payment methods and account ownership can not only create ef- ficiency gains for the government and increase convenience for recipients, but also presents an un- met opportunity for improving financial inclusion. As a result of the extended use of digital payments during the pandemic, in Jordan the number of active e-wallets doubled in 4 months, reaching 1 million users. In Paraguay more that 1.5 million wallets were created to receive social protection payments. In Brazil, close to 70 million beneficiaries received aid through the Auxílio Emergencial program, for which the government set up a digital savings account with digital access channels to use the funds. It is estimated that 40% of beneficiaries did not have an account before the pandemic. In Colombia, the Ingreso Solidario program delivered assistance to more than 2 million beneficiaries by depositing funds in accounts held by e-money providers; close to 1 million beneficiaries did not have an account before the pandemic. In Argentina it is estimated that the recent opening of more than 1 million univer- sal free accounts was related to the requirement to have a bank account to receive Ingreso Familiar de Emergencia program payments.93 In Central America, countries that relied on digital payment methods for emergency transfers, such as Guatemala and Honduras, could build on this experience and intro- duce (or improve, in the case of the latter) these modalities for their non-emergency social protection interventions. Costa Rica would benefit from implementing the planned reforms towards the estab- lishment of SUPRES, an integrated and centralized payment system for social benefits. Panama could build on lessons learned from the emergency response, such as the payment method used for Vale Digital, and potentially increase payment methods as well as convenience and choice for recipients. Increasing the integration of social information systems can bring benefits to both governments and people. From the government perspective, integrated social information systems can improve efficien- cy, data accuracy and quality, help identify gaps in coverage, duplications and complementarities, as well as facilitate a coordinated response to crises. For people, they can help promote awareness of and access to numerous benefits and services, they can allow registration in several programs at once with simpler procedures, saving time and money. In Central America, a few countries are already working on the integration of their social information systems. An example is El Salvador, which is working on the development of a social information system that will include both social registry and integrated beneficiary registry functions, integrating information both on the demand for and the supply of social programs.94 Similarly, the Dominican Republic is currently working on the development of SIUBEN+, which will build on the SIUBEN social registry and will include both social registry and integrated ben- eficiary registry functions, as well as a geostatistical and geolocation information system which will 93 Gentilini, et al. (2021). 94 Government of El Salvador, Despacho Comisionada Presidencial para Operaciones y Gabinete de Gobierno, “Sistema Integrado de Información Social – SIIS”, presentation. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 70 The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy allow for the early identification of diverse risks, and contribute to increase the resilience and adapt- ability of social programs in the face of any shock.95 Intensifying the use of technology beyond payments and across the different phases of the delivery chain can create efficiencies and reduce processing times. Technology can be applied to several pro- cesses, including intake and registration, notification, enrollment, payments, beneficiary data update and grievance redress. During the pandemic, innovations relying on technology were introduced, in- cluding digital registration modalities, online platforms or SMS for notification, remote onboarding and online platforms for grievance redress. For example, in Colombia SMS was used to notify eligible households. Similarly, in Egypt, SMS was used to notify eligible households and inform on location and time to collect their payments and ATM cards. In Peru, the government launched an online platform through which citizens could check their eligibility for the transfer as well as find information about date, time and bank branch for their payment. Costa Rica, which introduced an online portal during the pandemic, could build on this experience and incorporate digital channels in its regular programs. The leveraging of administrative data through interoperability/integration with social protection information systems opens up possibilities not only for timely updating but also for its use for tar- geting. Countries in the region, such as Colombia, Brazil, etc., have used their integrated social pro- tection information system, integrating and cross-checking with other administrative databases, to mobilize their response and reach the newly vulnerable. Similarly, globally such approaches were em- ployed, including even when such systems were not contemplated previously. As part of its response to COVID-19, Togo used a voter database that covered 95 percent of adults and happened to include information on occupation to target informal sector workers. India used data on more than 200 million low-income women for whom bank accounts had been opened in a financial inclusion drive. Namibia relied on the civil registry checked against income tax records. Advances in technology and, in particular, big data and machine learning, offer the promise of poten- tial improvements in targeting accuracy, but are not a panacea; better data may matter more than greater sophistication in inference. Non-administrative big data, such as from satellite imagery, mo- bile phones and social media, and machine learning are expanding data and techniques96 at a dizzying pace, though they remain largely proxies for welfare rather than measurement per se. Big data are already being combined with traditional data to improve poverty maps and help predict which house- holds and areas are more at risk of natural disasters. Administrative data has long been used along- side traditional data to improve poverty maps; newer big data can similarly be incorporated. Moreover, historical data on localized natural disasters and drought combined with realized household poverty outcomes can be used to predict which households are at risk in the future. Such models can be used to prioritize the poor or vulnerable for covariate risk-mitigating social protection programs or public in- surance schemes, helping administrators to better manage covariate shocks. In a crisis or data-scarce environment such as post-conflict, using big data for determining eligibility may be the only option and an acceptable one. Big data can fill a gap when traditional data are either not available, as in many poor or fragile countries or in post-conflict settings, or not current, as in a crisis. In such circumstances, the ability to nonetheless conduct eligibility assessments for much needed assistance is vital. Several countries experimented with these approaches as part of their response to COVID-19. For example, Ecuador supplemented its incomplete social registry with geographic targeting using census data combined with mobile phone usage data. in In Nigeria, census data and high-resolution satellite im- agery using machine learning algorithms will be used to identify the location and sizes of dwellings, which allows for a mapping of the poorest urban areas. Togo 97 based on satellite and phone record 95 “Social programs targeting system in the Dominican Republic: the experience of SIUBEN”, Knowledge Exchange Session Social Protection Dominican Republic - Saint Lucia, presentation, May 2021. 96 Grosh, Margaret; Leite, Phillippe; Wai-Poi, Matthew; Tesliuc, Emil (forthcoming). A New Look at Old Dilemmas: Revisiting Targeting in Social Assistance. World Bank. 97 Gentilini et al. (2021) Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy 71 data using machine learning algorithms. The Democratic Republic of Congo also relied on Call Detail Records (CDR) combined with geographic targeting. Cote D’Ivoire, Liberia, Mozambique and Nigeria are also using some variant of these approaches. Whether big data will replace the need for traditional data for eligibility assessment depends upon whether challenges arising from their newness can be fully understood and solutions crafted. 98 A recent paper on the Togo experience shows that targeting based on CDR performs worse in simulations than a traditional PMT, suggesting that the use CDR might not be that accurate beyond emergency, post-conflict or data-scarce contexts.99 C1. Strengthen Labor Market Institutions and Services to Support People Back to (Productive) Work Initial signs point to a potential K-shaped recovery, with varying paths necessitating differentiated policies. Emerging data points to a swift K-shaped recovery (Fig XX, the estimated rebound in GDP). However, wide differences are expected across countries, localities, sectors and sub-groups of popula- tion. As detailed in II.C, low-skilled, largely informal workers in affected sectors were disproportionately impacted, along with youth and women. Evidence from previous crises indicates that they are likely to face deeper scarring and take longer to get back to employment.100 There are already signs that recovery in some service sub-sectors, especially hotels and tourism, across the sub-region is lagging. The differences across countries are also clear: Panama and Guatemala are rebounding relatively fast, driven by strong growth in industry and services, despite the severity of the initial shock, while Costa Rica which was hit relatively less hard, is recovering more evenly, estimated to reach its pre-COVID growth by 2023. This points to the need for differentiated policy responses, to ensure that vulnerable workers, in particular in the most affected sectors, can be supported back to productive employment. Strengthening (re) employment policies and services through a multi-pronged approach, while keep- ing the required customization in mind, will be critical for building back better. As economic activity ramps up, it will be crucial to support workers back to employment to minimize long-term impacts. This needs a multi-pronged approach, which should be aligned with the recommended shift in the role of government policy in light of the changing world of work, namely rather than protect workers from change, protect them for change by supporting job transitions and reemployment (Packard et al 2019).101 This includes, but is not limited to, improving labor market information systems to better understand the needs of employers; promoting private sector led training to ensure relevant skills and better job matches; employ integral interventions/ALMPs, such as wage subsidies with training, to support youth/new entrants to LM that are at risk of long-term scarring, strengthening institutional capacity for intermediation, including use of technology, while ensuring adequate budgets; and pro- moting labor market reforms (e.g. adjustments to regulations) that support the creation of better jobs. First, in the context of a rapidly changing labor market, a better understanding and monitoring of labor demand, including through use of real-time information, is a critical element for effective (re) employment policies. Re-employment support can take many forms, including hiring subsidies, skills training (technical and socioemotional), job search assistance and labor intermediation. The effective deployment of these measures hinges on having information on the labor market (e.g., what sectors and occupations are hiring, what type of skills are demanded, etc.), helping to identify skills needs and ways to improve job matches for the most vulnerable affected workers. This information will be 98 Gentilini et. al., 2021 99 Relative to methods that require a comprehensive social registry (a hypothetical exercise; no such registry exists in Togo), the machine learning ap- proach increases exclusion errors by 9-35%. For more detail, refer to Aiken, Emily; Bellue, Suzanne; Karlan Dean; Udry, Christopher R.; Blumenstock, Joshua. “Machine Learning and Mobile Phone Data Can Improve the Targeting of Humanitarian Assistance”. NBER Working Paper 29070, July 2021. 100 Silva, et al. (2021) 101 Packard et al. (2019) Protection for all: Social Protection and Labor White paper Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 72 The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy particularly useful now, considering the varying speed of recovery across localities and sectors as well as the potential sectoral realignment that may take place. While traditional needs assessments and employer surveys are key to have a sense of needs on the ground, real-time labor market data, such as vacancies posted online, is increasingly being utilized102 for a better and more up to date under- standing of LM shifts. Big data techniques have been key for the analysis of online data; while having limitations, there is potential to inform a wide range of actors and workforce development policies. Countries across the globe, such as Australia, Malaysia, and Turkey, use online data to monitor vacan- cies and skills shortages. Such data is incorporated into the Labor Market information eco-system in Korea (World Bank, 2020), which is used to inform policy priorities, to target training subsidies for scarce skills (e.g. Australia); informs migration policies (e.g. Korea, Malaysia), and is an input for work- force development (e.g. Korea). In Mexico, recent analysis and integration of real-time with traditional labor market information as well as consultations with employers, is providing a wealth of information to improve employment support programs and policies.103 Some countries in the sub-region are well on their way to monitor shifts in labor demand, while oth- ers may benefit from innovative approaches to that end. Costa Rica and El Salvador are building labor market information systems, with the capacity to provide data, in particular on occupations and skills in demand.104 Guatemala, Honduras, and Panama are at the early stages of building such systems and using the data to inform policy design. Concerted effort to build and systematize labor market informa- tion would not only help understand the needs of the market and thus ensure that training programs are in-line with the demands of the labor market, but also can help in workforce development and broader alignment of the education system (including TVET) with realities on the ground. For instance, a deeper understanding of labor demand by types of jobs/occupations/skills would allow the design and implementation of employment programs that can promote green jobs creation. Given that the jobs of the future are increasingly more reliant on non-routine cognitive skills, such as analytical (e.g., problem-solving) and interpersonal (e.g., proactivity, responsibility, client orientation),105 it will be im- portant for the education system to emphasize foundational skills with a view to fungibility, while the changes in demand for professional skills can be largely handled through the close monitoring and adjustment of re/upskilling policies and programs. Second, and relatedly, promoting private sector led and provided training, to the extent possible, will ensure relevant skills, improving the quality of matches and increasing chances of long-term suc- cess. Literature has clearly shown that private sector-led (where demands are identified based on their needs) and, in particular, private sector provided training has a higher chance of success in promoting quality matches. This is crucial for firm-specific skills, which improve productivity in the longer-run. This involves public private partnerships, including with multi-national firms, where firms provide training or work in close collaboration to define training content. For instance, the Entra21 Program in Argentina for vulnerable youth106 implemented in close collaboration with the private sector, resulted in higher formal employment (by 8 percentage points) and earnings (by 40%) for participants as com- pared to the control group (Alzúa et. al., 2016). Similar approaches have been used in other countries (Germany107, US, and Brazil108 ) with positive outcomes in particular for disadvantaged groups. 102 See Kurekova, Beblavy, and Thum (2014) on uses of online data for labor market analysis—as cited in CSC (Critical Skills Monitoring Committee) (2018), which provides further information on use of real-time data. Burning Glass analyzes and structures the job vacancy data in multiple countries, through machine learning and textual analysis, using the skills taxonomy they developed Burning Glass (2018). 103 The dashboard only shows select indicators to inform and facilitate the responsibilities of job counselors, while the tool itself includes a rich set of information that informs a broader set of stakeholders at different levels. 104 El Salvador has developed a LMIS, which was recently launched (at the beginning of December 2021). The recent modernization of the Instituto Na- cional de Aprendizaje en Costa Rica envisions undertaking relevant labor demand studies and monitoring as well as the establishment of an LMIS which includes/aggregates vacancy information across the country. 105 Autor et. al (2012), Bussolo et. al (2018). 106 Youth aged 18-25, with secondary degree from households in the lowest quintile. The program combined classroom training with on-the job training 107 Public Private Partnership, which provides low-skilled German youth (with less than secondary education) with a combination of counseling, training and work experience, improved the probability of employment in the short and medium term (Ehlert et al 2012) 108 The Lei do Jovem Aprendiz, apprenticeship program in Brazil resulted in higher probability of permanent employment and higher earnings, especially for less educated workers. DataFolha Instituto de Pesquisa, & CIEE (2019). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy 73 There are efforts across the sub-region to align training programs with the needs of the private sector, though to varying degrees. This necessitates a change in approach and delivery of training through TVET. For instance, in Costa Rica, the recent modernization of the main TVET system (INA) that allows for outsourcing of training not available at INA to private accredited providers, as well as an orientation towards vulnerable youth (through provision of scholarships/vouchers) with close mon- itoring of private sector demand and the links with the national employment system, is aligned with the international evidence. On the other hand, INFOP, the principal publicly funded TVET in Honduras, largely serves the already employed without much support for labor market insertion, and even then, with outdated programs not linked to employer demand. While INFOP has outsourced some training (around 20% of its budget) to private providers, which have better links to labor demand, it still fell short of serving the needs of the private sector.109 The government proposed ambitious reforms in 2019 in response to pressure from the private sector, however implementation has been slow. Other TVET agencies in Central America (El Salvador and the Dominican Republic) face similar issues. Third, in addition to training, other labor programs can contribute to better employment outcomes for vulnerable groups, especially when implemented through integrated programs addressing both la- bor demand and supply. For example, evaluations often find that wage subsidies have positive employ- ment effects and that combining them with training and/or employment services improved long-term employment for the participants. Impact can also be enhanced by using complementary measures such as covering childcare costs. Programs designed to simultaneously address several constraints at once including supply weaknesses (skills gaps and opportunity costs), low demand (training and hir- ing incentives) and information gaps (labor intermediation) have the potential for achieving sustained jobs for the participants. Other jobseekers may be best served by on-the-job training opportunities or entrepreneurship support. A meta-analysis of studies of ALMPs in LA and Caribbean, Escudero et al (2019) found that employment and formal employment are the outcomes most likely to be positively affected by ALMPs and that ALMPs are more likely to be effective for women and for young people. Fourth, institutional strengthening to improve labor intermediation, supported by technology, will further improve support to affected workers on the road to recovery. The capacity of Ministries of Labor, employment services and other actors, to provide job search assistance and intermediation, es- pecially for the most vulnerable, needs to be strengthened. While most countries have some network of employment services with local presence, the responsibility for labor intermediation is fragmented in most cases. Globally, the rapid development of artificial intelligence/machine learning techniques to analyze large swaths of textual information (CVs, online vacancies) led to rapid improvements in matching algorithms and development of Job Matching Portals. Job Matching Portals are increasingly becoming a core element of labor market information systems (e.g. WorkNet of Korea, UWV of the Netherlands, VDAB of the Flanders region of Belgium, Pol’e Emploi of France etc.). While private sector companies develop most job matching portals (e.g., WCC in Colombia) there are examples of in-house development (e.g. Korea) or public-academic partnerships (e.g. a collaboration of the National Employ- ment Service with ITAM in Mexico). In the region, Job Matching Portals exist in Costa Rica (under the National Employment Office in collaboration with the National Learning Institute), while efforts are ongoing in El Salvador. Job Matching Platforms can facilitate and lighten the work of job counselors, while further professionalization of counselors would likely be needed to provide effective support to the more vulnerable. On the other hand, more straightforward approaches, such as instituting a basic profiling system to better prioritize resources may be a low-hanging fruit. For instance, most countries in Europe employ a three-level prioritization system categorizing distance from the LM or risk of long- term unemployment (low, medium, high) based on counselor judgment or statistical profiling or both, which allows for differentiated levels of support (e.g. job search assistance and link with employers; some counseling and link with ALMPs; and intense counseling, link with ALMPs, and referrals to other 109 HND JD Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 74 The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy services to address barriers, respectively). In Ireland this prioritization determines who will have first access to group orientation workshops (i.e., the high group), to get them on the path to employment as quickly as possible to minimize the probability of long-term unemployment. Strengthening the insti- tutional capacity for employment services would also help to improve implementation of employment programs by matching different participant groups to specific programs (e.g., reskilling, employment subsidies, green public works, etc.) and to help vulnerable workers to manage job transition, for exam- ple, in the eventual necessary move towards green jobs (e.g., decarbonization). Digitization of employment services would round out the efforts and can result in efficiency gains. The Covid-19 pandemic pushed a rapid digitization of many services, including employment services. Countries where a good base existed were able to move to digital means more rapidly. For instance, the public employment services in Estonia adopted a multi-channel delivery strategy in 2018, with online applications, webinars, and chat rooms, telephone and email. Online applications in particular allowed them to cope with the large volume of unemployment insurance claims during the pandem- ic.110 Live video and other online communication tools for free calls have proven to be a good, low-cost way to provide information and advice to job seekers and employers (e.g., Uruguay). Adaptation of employment services to allow for continued social distancing, through digitization, will be key to be able to reach and serve more job seekers, potentially increasing cost-effectiveness and flexibility in implementation. At the same time, it will be important to have multiple delivery mechanisms to en- sure access to services for all job seekers, especially those without the access or ability to use digital channels. Providing effective employment support requires adequate budgets, a struggle for the countries in the sub-region. In the Dominican Republic, active labor market programs (labor market intermedia- tion, job search assistance, training, wage subsidies etc.) comprised only 1% of SP spending, or 0.07% of GDP (2018).111 In Honduras TVET spending (through INFOP) was 0.18% of GDP (2016). Costa Rica, on the other hand, spent about 0.22% of GDP on ALMPs (World Bank 2019). This compares poorly to the pre-pandemic OECD average of 0.52% of GDP,112 though more or less on par with LCR countries (with some exceptions such as Brazil and Argentina).113 Adequate budget is needed to provide effective em- ployment support that can bridge the transitions through the recovery and to green jobs. These policies can only go so far in the absence of structural reforms to support job creation. A rich set of recent analyses outline the wide range of structural barriers to job creation (e.g., limited com- petition and access to finance; lack of innovation, infrastructure, and integration into GVCs; barriers to trade; and crime/insecurity etc.).114 It is important to recognize that strengthening employment ser- vices and institutions is only one, albeit important, instrument to improve employment outcomes in the medium to long term. Indeed, scarcity of labor demand leading to limited job creation (especially of good jobs) is a binding constraint across certain localities in the countries of the sub-region and can only be addressed through broader structural policies. A detailed treatment of these falls outside the scope of this report. Finally, promoting labor market reforms to support the creation of better jobs will be critical for a sustained recovery. This includes modernizing LM regulations, as well as other incentives (e.g., em- ployment subsidies) to foster demand for workers (especially vulnerable workers) to help promote employment growth.115 In addition, it will be important to ensure that minimum wage is consistent 110 ILO (2020). COVID-19: Public employment services and labour market policy responses 111 Winkler and Montenegro (2021) 112 OECD Stats. Public expenditure on activation policies. 113 ASPIRE 114 Ulku, Hulya; Zaourak, Gabriel. 2021. Unleashing Central America’s Growth Potential: Synthesis Report (English).  Washington, D.C.: World Bank Group. Banegas, Nancy; Winkler, Hernan. 2020. El Salvador Job Diagnostic: Understanding Challenges for More and Better Jobs in El Salvador - An Integrated Approach. Job Series; No. 24. Washington, DC: World Bank. 115 Silva et al. (2021) and Ulku and Zaourak. (2021). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery The path forward for strengthened adaptive SPL systems for a more inclusive and resilient recovery- Directions for Policy 75 with labor productivity growth. Some modernization of the LM regulations already took place across the sub-region with the establishment of tele/flexible work. At the same time, COVID-19 deepened the divide on access to jobs of the future linked with the use of technology. Adapting the labor market regulations to the jobs of the future will be key to ensure better protection for all workers and a more even playing field. For instance, instead of differentiating based on permanent vs. temporary contracts, establishing a set of minimum standards that apply to all contracts, with simpler dismissal procedures could be considered (Packard, et. al). 76 Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Annexes 77 Annexes Figure A.1. Percentage of households that ran out of food due to lack of money or other resources (last 30 days) 60% 53% 50% 46% 44% 40% 40% 34% 35% % of households 30% 30% 27.9% 23% 24% 20% 10% 0% Costa Rica Dominican Republic El Salvador Guatemala Honduras Wave 1 Wave 4 Source: World Bank, High Frequency Phone Surveys, Washington DC. Data for Wave 1 (between May 21st and June 1st) and Wave 4 (between June and July 2021) Figure A.2. Composition of SPL Responses by components and by regions 100% 11% 90% 18% 16% 24% 21% 22% 24% 32% 14% 80% 70% 11% 18% 15% 23% 26% 21% 60% 29% 50% 40% 75% 65% 61% 63% 30% 59% 58% 55% 20% 39% 10% 0% AFR EAP ECA LAC MNA North America SAR World Social assistance Social insurance Labor market Source: Gentilini et al. (2021). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 78 Annexes Figure A.3. Per capita social spending by sector Central America and LCR 3500 3297 3000 Per Capita Social Spending, 2500 2448 2374 2010 Constant USD 2160 2000 1902 1898 1793 1705 1500 1324 1000 919 769 745 696 566 556 466 459 500 456 436 296 180 173 158 0 ica i le y os zil Ba o as ba i na bia ico a or n R ru li c Pa a ay r na Ho ia Gu ras ala a do ua m aic gu g li v ad Pe ra ub gu m Ch ad Cu ba ya aR em ex lom nt u na ua ug ra Bo B m ha nd av Gu ra rb ep To ge M Pa ca Ec st Ur Ja at Sl Co Ba Ar Co nd Ni El da ca ini da m ini Do Tr Education Health Social Protection Source: Guatemala Social Expenditure Review as background analysis for Guatemala SCD Update. Figure A.4. Social Spending by sector as a share of GDP, Central America 25% 21.1% 20% 15.9% 7% 15% 13.2% 13.4% 13.5% 8% 5% 4% 7% 10% 7.6% 6% 3% 3% 4% 5% 5% 4% 2% 7% 6% 3% 4% 3% 4% 0% Guatemala El Salvador Panama Nicaragua Honduras Costa Rica Education SP/GDP Health SP/GDP Social Protection SP/GDP Source: Guatemala Social Expenditure Review as background analysis for Guatemala SCD Update. Figure A.5. Share of CCT transfers in beneficiaries’ welfare (adequacy), Central America and LCR, in Q1 Honduras 20.2 Latin America & 17.2 the Caribbean Costa Rica 13.3 Panama 10.5 Dominican 5.9 Republic 0.0 5.0 10.0 15.0 20.0 25.0 Source: ASPIRE performance indicators. Note: Latest available year, for Costa Rica (CRI) 2019, Dominican Republic (DOM) 2019, Guatemala (GTM) 2014, Honduras (HND) 2018, Nicaragua (NIC) 2014, Panama (PAN) 2019 and El Salvador (SLV) 2019. For Latin America and the Caribbean 2010-2019 weighted average using latest available survey year by country. Adequacy is incidence is: (transfer amount received by a group)/(Total welfare aggregate of the beneficiaries in that group). Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Annexes 79 Figure A.6. Share of establishments in or expect to fall in arrears in next 6 months 80 70 Share of establishments 60 50 40 30 20 10 0 Guatemala Honduras Nicaragua El Salvador All Large (100+) Medium (20-99) Small (5-19) Micro (0-4) Manufacturing Retail Other Services Source: Business Pulse Survey and the World Bank Enterprise Survey. Data for Wave 1 for Brazil (June 19 to July 31, 2020), Guatemala (June 24 to August 7, 2020), Honduras (June 24 to August 7, 2020), Mexico, Nicaragua (June 17 to August 7, 2020) and El Salvador (June 10 to August 7, 2020) Figure A.7. Share of establishments that fired workers in the last 30 days 80 70 60 Share of establishments 50 40 30 20 10 0 Brazil Guatemala Honduras Mexico Nicaragua El Salvador All Large (100+) Medium (20-99) Small (5-19) Micro (0-4) Manufacturing Retail Other Services Source: Business Pulse Survey and the World Bank Enterprise Survey. Data for Wave 1 for Brazil (June 19 to July 31, 2020), Guatemala (June 24 to August 7, 2020), Honduras (June 24 to August 7, 2020), Mexico, Nicaragua (June 17 to August 7, 2020) and El Salvador (June 10 to August 7, 2020) Figure A.8. Contribution of change in productivity (value added per worker), by country and major sector Panama 2012-2019 Total=2.7% Honduras 2012-2019 Total=-0.7% Guatemala 2012-2019 Total=1.8% Dominican Republic 2012-2019 Total=3.5% Costa Rica 2012-2019 Total=2.4% El Salvador 2012-2019 Total=1.5% -3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 4.00 5.00 Within Agriculture Within Industry Within Services Inter-sectoral Agriculture Inter-sectoral Industry Inter-sectoral Services Source: Own elaboration using JobStructure Tool. Data for 2019 based on ILO estimates. The World Bank – Jobs Group. 80 Di culty of Dismissal Index (PCA) Rigidity of Hours Index (PCA) 2019 Di culty of Hiring index (PCA) 2019 2019 (normalized annual) (normalized annual) (normalized annual) -1 -0.5 0 0.5 1 1.5 2 2.5 3 -2 -1.5 -1 -0.5 0 0.5 1 1.5 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 Colombia Antigua and Barbuda Bolivia Grenada Belize Costa Rica Source: Packard et al. (2019) Source: Packard et al. (2019) Source: Packard et al. (2019) Haiti Bolivia Honduras Jamaica Chile Panama Uruguay Colombia Venezuela, RB Argentina Dominica Brazil (Sao Paulo) Brazil (Sao Paulo) Grenada Peru Costa Rica Haiti Argentina Dominican Republic Jamaica Chile Guatemala St. Lucia Belize Nicaragua Mexico (Mexico City) St. Lucia El Salvador Peru Grenada Antigua and Barbuda Suriname Mexico (Mexico City) Dominica Trinidad and Tobago Ecuador St. Ki s and Nevis Uruguay Dominican Republic Trinidad and Tobago Argentina Guatemala Barbados Bahamas, The El Salvador Bolivia Barbados Colombia Ecuador Antigua and Barbuda Figure A.10. Rigidity of Hours Index (as compared to global average) Figure A.9. Difficulty of Hiring Index (as compared to global average) Venezuela, RB Belize Brazil (Sao Paulo) Dominica Figure A.11. Difficulty of Dismissal Index (as compared to global average) Chile Costa Rica Haiti St. Lucia Honduras Jamaica Ecuador Panama Suriname Guyana Dominican Republic Trinidad and Tobago Bahamas, The St. Ki s and Nevis Uruguay Suriname Guatemala Bahamas, The Honduras Nicaragua Barbados Panama El Salvador St. Ki s and Nevis Peru Guyana Nicaragua Mexico (Mexico City) Venezuela, RB Guyana Annexes Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Annexes 81 Figure A.12. Labor force participation by age group and by Central American countries 90 80 % Labor force participation 70 60 50 40 30 20 10 0 El Salvador Costa Rica Dominican Republic Panama Guatemala Honduras Youth Adult Source: Own elaboration using SEDLAC harmonization for El Salvador 2019, Dominican Republic 2019, Honduras 2019, Guatemala 2018. Data for Panama 2019 and Costa Rica 2019 provided by the NSO and harmonized by the team. Note: Youth considers people from 15 to 24 years old and adult, from 25 to 64. Labor force participation: Share of population that is in the labor force as percentage of working age population for each age group. Figure A.13. Informality by age group and by Central American countries 100 90 80 % of informal workers 70 60 50 40 30 20 10 0 Costa Rica Dominican Republic Panama El Salvador Guatemala Honduras Youth Adult Source: Own elaboration based on El Salvador 2019, Dominican Republic 2019, Honduras 2019, Guatemala 2018 using SEDLAC harmonization. Data for Panama 2019 and Costa Rica 2019 provided by the NSO and standardized by the team. Note: Informality based on legal characteristics: not-salaried, employer/salaried/self-employed without social security. Youth considers people from 15 to 24 years old and adult, from 25 to 64. Figure A.14. Labor Force Participation Rate in Costa Rica, by youngest child in the household and by gender Participation Rate by Age of Youngest Kid in the HH .8 .9 Participation rate .6 .7 .5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Age of the youngest kid in the household (< = 14) Men Women Source: Own elaboration using Costa Rica Labor Force Survey 2020 Q1 (Pre- COVID-19) Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 82 Annexes Figure A.15. Probability of Employment in El Salvador, by youngest child in the household and by gender 0.1 0.05 0 1 or less 2 3 4 5 6 7 8 9 Marginal e ect -0.05 -0.1 -0.15 -0.2 -0.25 Age of youngest child Women Men Source: Based on EHPM 2019, updating analysis in Banegas and Winkler (2020). Note: Each line shows the difference in the probability of labor force partic- ipation between a person (male and female) with children and a person without children, according to the younger child’s age. They are estimated using a linear probability model, controlling for other individual characteristics such as age, education, and location. Figure A.16. GDP growth in the agriculture sector from 2019 to 2024 (forecast) 8.0 Real GDP growth at constant factor prices 6.0 4.0 2.0 0.0 2019 2020e 2021f 2022f 2023f 2024f -2.0 -4.0 -6.0 -8.0 Costa Rica Dominican Republic El Salvador Guatemala Honduras Nicaragua Panama Source: The World Bank, Macro Poverty Outlooks. April 2022. Own elaboration. Figure A.17. GDP growth in the industry sector from 2019 to 2024 (forecast) 40 Real GDP growth at constant factor prices 30 20 10 0 2019 2020e 2021f 2022f 2023f 2024f -10 -20 -30 -40 Costa Rica Dominican Republic El Salvador Guatemala Honduras Nicaragua Panama Source: The World Bank, Macro Poverty Outlooks. April 2022. Own elaboration. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery Annexes 83 Figure A.18. GDP growth in the service sector from 2019 to 2024 (forecast) 15 Real GDP growth at constant factor prices 10 5 0 2019 2020e 2021f 2022f 2023f 2024f -5 -10 -15 Costa Rica Dominican Republic El Salvador Guatemala Honduras Nicaragua Panama Source: Own elaboration based on Macro Poverty Outlooks, The World Bank. April 2022. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 84 References References Acosta, Pablo A., Rita Almeida, Thomas Gindling and Christine Lao Pena. 2017. “Toward More Efficient and Effective Public Social Spending in Central America.” Directions in Development—Human Development. World Bank, Washington, DC. Alzúa et al. 2016. “Long-run Effects of Youth Training Programs: Experimental Evidence from Argentina” Economic Inquiry, vol. 54(4), pages 1727-1991. doi: https://doi.org/10.1111/ecin.12348 Andrews et al. 2021. “The State of Economic Inclusion Report 2021: The Potential to Scale.” World Bank, Washington, DC. Autor, David and Daron Acemoglu. 2012. “What Does Human Capital Do? A Review of Goldin and Katz’s The Race between Education and Technology.” Journal of Economic Literature, 50:2, 426–463 Avitabile, Ciro D’Souza, Ritika Gatti, Roberta V. Chapman and Emily Weedon. 2019. “Insights from Dis- aggregating the Human Capital Index.” Washington, DC. Banegas, Nancy and Hernan Winkler. 2020. “El Salvador Job Diagnostic: Understanding Challenges for More and Better Jobs in El Salvador - An Integrated Approach.” Job Series; No. 24. Washington, DC. Berner, Heidi and Tamara Van Hemelryck. 2021. “Social information systems and registries of recipients of non-contributory social protection in Latin America in response to COVID-19.” Project Documents (LC/TS.2021/56). Economic Commission for Latin America and the Caribbean (ECLAC), Santiago. Beylis et al. 2020. “Going Viral: COVID-19 and the Accelerated Transformation of Jobs in Latin America and the Caribbean.” World Bank Latin American and Caribbean Studies. World Bank, Washington, DC. Bussolo, Maurizio, Ivan Torre and Hernan Winkler. 2018. “Does Job Polarization Explain the Rise in Earn- ings Inequality? Evidence from Europe.” Policy Research Working Paper; No. 8652. World Bank, Washington, DC. Centro de Investigaciones Económicas Nacionales. 2020. “Análisis del Fondo Bono Familia.” CIEN, Guatemala. https://cien.org.gt/wp-content/uploads/2020/12/Documento-Analisis-Fondo-Bo- no-Familia.pdf Centro Nacional de Información del Sector Social. 2019. “Mejores Prácticas en la Implementación de la Estrategia Operativa para la Actualización de la Información Socioeconómica en el País.” Semi- nario Internacional sobre Actualización de Información De Sistemas De Información Social, Santo Domingo. https://ceniss.gob.hn/index.html Cuberes, David and Marc Teignier. 2016. “Aggregate Effects of Gender Gaps in the Labor Market: A Quantitative Estimate.” Journal of Human Capital, University of Chicago Press, vol. 10(1), pages 1-32. DataFolha Instituto de Pesquisa, & CIEE Centro de Integração Empresa Escola. (2019). Aprendiz CIEE, o que nos disseram os jovens que concluiram o programa. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery References 85 De Hoyos Navarro, R., Halsey Rogers and Migue Székely. 2016. “Out of school and out of work: a diag- nostic of ninis in Latin America.” World Bank, Washington, DC. Delaporte, Isaure and Werner Peña. 2020. “Working From Home Under COVID-19: Who Is Affected? Evidence From Latin American and Caribbean Countries.” GLO Discussion Paper Series 528. Glob- al Labor Organization (GLO). Eberhard-Ruiz, Andreas. 2021. Guatemala Jobs Diagnostic. Washington, D.C.: World Bank Group. Ehlert, M. 2012. “Buffering income loss due to unemployment: Family and welfare state influences on income after job loss in the United States and western Germany.” Soc Sci Res. 41(4):843-60. doi: 10.1016/j.ssresearch.2012.02.002. Escudero et al. 2019. “Active Labour Market Programmes in Latin America and the Caribbean: Evi- dence from a Meta-Analysis.” Journal of Development Studies, Taylor & Francis Journals, vol. 55(12), pages 2644-2661. Freire et al. 2018. “Afrodescendientes en Latinoamérica: Hacia un marco de inclusión.” World Bank, Washington, DC. Garcia Mora, Maria Elena, Steven Schwartz and German Freire. 2021. “Disability Inclusion in Latin America and the Caribbean: A Path To Sustainable Development.” World Bank, Washington, DC. Gentilini, Ugo, et al. Social Protection and Jobs Responses to COVID-19 : A Real-Time Review of Country Measures (May 14, 2021) (English). COVID-19 Living Paper Washington, D.C. : World Bank Group. Retrieve from: http://documents.worldbank.org/curated/en/281531621024684216/Social- Protection-and-Jobs-Responses-to-COVID-19-A-Real-Time-Review-of-Country-Measures- May-14-2021 Government of Costa Rica. 2021. “Noveno Informe Mensual de Seguimiento a la Ejecución del Bono Proteger.” https://www.mtss.go.cr/elministerio/despacho/covid-19%20mtss/plan_proteger/archi- vos/noveno_informe_proteger.pdf. Government of El Salvador. Despacho Comisionada Presidencial para Operaciones y Gabinete de Go- bierno. “Sistema Integrado de Información Social – SIIS.” Presentation. Grosh et al. Forthcoming. “A New Look at Old Dilemmas: Revisiting Targeting in Social Assistance.” Grosh, Margaret, Maurizio Bussolo and Samuel Freije. 2014. “Understanding the Poverty Impact of the Global Financial Crisis in Latin America and the Caribbean.” Directions in Development--Human Development. World Bank, Washington, DC. International Labour Organization. 2020. “COVID-19: Public employment services and labour market policy responses.” International Monetary Fund. 2021. “Recomendaciones para el desarrollo del Sistema Unificado de Pago de Recursos Sociales (SUPRES) de la Tesorería Nacional.” Technical report, May 2021. Isik-Dikmelik, Aylin. 2012. “Do Social Benefits Respond to Crises? Evidence from Europe and Central Asia during the Global Crisis.” Social Protection and Labor Discussion Paper; No. 1219. World Bank, Washington, DC. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery 86 References Lo Bello et al. 2019.  “From Ghana to America: The Skill Content of Jobs and Economic Develop- ment.” Policy Research Working Paper; No. 8758. World Bank, Washington, DC Lustig, Nora et al. 2020. “The Impact of COVID-19 Lockdowns and Expanded Social Assistance on In- equality, Poverty and Mobility in Argentina, Brazil, Colombia and Mexico.” World Bank, Washington, DC. Mejia-Mantilla et al. 2021. “Jobs Interrupted: The Effects of COVID-19 in the LAC Labor Markets.” World Bank, Washington, DC. Michel, Veronica., and Ian Walker. 2019. “Honduras: Jobs Diagnostic”. World Bank, Washington, DC. Packard et al. 2019. “Protecting All: Risk Sharing for a Diverse and Diversifying World of Work.” World Bank, Washington, DC. Silva et al. 2021.  “Employment in Crisis: The Path to Better Jobs in a Post-COVID-19 Latin Ameri- ca.” World Bank Latin America and Caribbean Studies. World Bank, Washington, DC. Sousa, Liliana D.; Garcia-Suaza, Andres. 2018. Remittances and Labor Supply in the Northern Triangle. Policy Research Working Paper; No. 8597. World Bank, Washington, DC. © World Bank. https:// openknowledge.worldbank.org/handle/10986/30446 License: CC BY 3.0 IGO. Ulku, Hulya and Gabriel Zaourak. 2021. “Unleashing Central America’s Growth Potential: Panama.” World Bank, Washington, DC. Ulku, Hulya and Gabriel Zaourak. 2021. “Unleashing Central America’s Growth Potential: Synthesis Re- port.”  World Bank, Washington, DC. Williams, Asha and Sarah Berger Gonzalez. 2020. “Towards Adaptive Social Protection Systems in Latin America and the Caribbean: A Synthesis Note on using Social Protection to Mitigate and Respond to Disasters and Climate-Related Risks.” World Bank, Washington, DC Winkler, Hernan and Miriam Montenegro. 2021. “Dominican Republic Jobs Diagnostic.” Job Series; No. 28. World Bank, Washington, DC. World Bank. 2012. “Resilience, equity, and opportunity: the World Bank’s social protection and labor strategy 2012-2022.” World Bank, Washington, DC. The World Bank. 2020. COVID-19 responses in Latin America and the Caribbean (internal mimeo). World Bank. 2020. “Hacia un ecosistema de pagos digitales para impulsar la inclusión financiera. Diag- nóstico en base a la publicación Payment aspects of financial inclusion (PAFI) del CPMI y Banco Mundial.” World Bank, Washington DC. World Bank. 2020. “Respuesta de Protección Social a la Crisis de Covid-19 en la Republica Domini- cana.” World Bank, Washington, DC. World Bank. 2020. “The Human Capital Index 2020 Update: Human Capital in the Time of COVID-19.” World Bank, Washington, DC. World Bank. 2020. “The Cost of Staying Healthy.” LAC Semiannual Report. World Bank, Washington, DC. Stronger Social Protection and Labor Systems in Central America for a Resilient and Inclusive Recovery References 87 World Bank. 2021. “Expansión de la protección social durante la pandemia en Panamá. Principales características del Plan Panamá Solidario y su impacto en mitigación de pobreza e inequidad.” Unpublished. World Bank. 2021. “Macro Poverty Outlook, Spring Meetings 2021: Country-by-country Analysis and Projections for the Developing World.” World Bank, Washington, DC. World Bank. 2021. “Renewing with Growth: 2021 Semiannual report of the Latin America and the Carib- bean region.” Washington, DC. World Bank. 2022. “Guatemala - Systematic Country Diagnostic: Building a Stronger Social Contract through Productive, Inclusive and Sustainable Growth.” World Bank, Washington, DC. World Bank. 2022. “Macro Poverty Outlook, Spring Meetings 2022: Country-by-country Analysis and Projections for the Developing World.” World Bank, Washington, DC. World Bank. Forthcoming. “A Lost Decade for Poverty and Inequality Reduction.”