UNCOVERED: THE ROLE OF THE IMF IN SHRINKING THE SOCIAL PROTECTION Case Studies from Tunisia, Jordan and Morocco

UNCOVERED: THE ROLE OF THE IMF IN SHRINKING THE SOCIAL PROTECTION Case Studies from Tunisia, Jordan and Morocco Jihen Chandoul and Chafik Ben Rouine Laith Al-Ajlouni Boutaina Falsy and Jamal Azouaoui Editor Salma Hussein Language Editing Hazel Haddon Coordination Imene Cherif

Table of Contents Foreword 6 Introduction 8 Salma Hussein How international financial institutions affect social protection in Tunisia 21 Chafik Ben Rouine and Jihen Chandoul I. Introduction 22 II. International financial institutions and social protection 25 III. The impact of IFIs on social protection in Tunisia 32 IV. A reading of IMF documents: The subsidy reform central to IMF structural adjustment in Tunisia 43 V. Conclusion 57 VI. Bibliography 59 Appendices 62 IMF conditionalities and social protection in Jordan: Regression or improvement? 78 Laith Alajlouni I. Introduction 79 II. Defining social protection systems, globally and in the Jordanian context 81 III. The history of social protection in Jordan 83 4 Uncovered: The role of the IMF in shrinking the social protection

IV. Challenges to social protection spending in Jordan 88 V. The IMF and Jordan 92 VI. Data gaps 110 VII. Conclusion 111 VIII. Policy recommendations 112 Social spending in Morocco: The influence of IMF conditionality 115 Boutaina Falsy and Jamal Azouaoui I. Introduction 116 II. The failings of a neoliberal socioeconomic model 118 III. Social protection in Morocco: Between universality and econometric targeting 127 IV. The COVID-19 crisis: Opportunities and obstacles to social protection 143 V. Recommendations 151 Appendices 152 Case Studies from Tunisia, Jordan and Morocco 5

Foreword When the Covid-19 pandemic hit the MENA region with lockdowns and closures, the social impacts were felt almost immediately. The social impact was so outsized because social protection had been severely weakened over the past decades, mostly in the name of austerity and smaller state budgets. Also, by trying to rely more on direct cash transfers and to become more“targeted”, new protection programs have excluded some of those most in need. Other important components of social protection, such as healthcare and education, have seen waves of de-funding and privatizations in recent years, creating larger barriers to their access. At the same time, inequalities have sharply risen, as the middle class has come under increasing pressure. Social protection should not only be a tool to protect the poorest and most vulnerable, but should also protect the middle classes from sliding into poverty in times of crisis. For FES, an organization rooted in the values of social democracy and social justice, social protection has always been a major priority. On the global level, programs have focused on social protection floors and the creation of a global social protection fund. In the MENA region, the debate is just starting, but gaining significantly in relevance. This is due to the efforts of many international and local organizations, some of which are FES’ long-time partners. With this publication with a regional and comparative outlook, we hope to support these concerted efforts. In Tunisia, where at the time of publishing discussions are under way for a third major IMF program since 2011, painful economic reforms and inflation have almost crushed the middle class, which was so important for the democratic spring following the revolution. The expansion of social protection and to cast the net as wide as possible, must be important policy priorities. Jihen Chandoul and Chafik Ben Rouine have shown impressively how cuts in subsidies have hurt the population, without an adequate program for cash assistance in place. In Jordan, high unemployment and rising taxes have created social discontent. Some public programs are in place to protect the poorest, but large and significant gaps remain. Latih Al-Ajlouni did a great job to research the policies in Jordan and their impact on social protection and social spending. In Morocco, the Covid-19 pandemic exposed the weakness of the existing social protection architecture, which following the IFIs recommendations also relied mostly on targeting. Boutaina Falsy and Jamal Azouaoui show how the lessons from the crisis might help create a more universal protection system in the country. This publication would not have been possible without the countless support and dedication of many collaborators. First and foremost, Salma Hussein did a great job at editing the publication, giving it 6 Uncovered: The role of the IMF in shrinking the social protection

its regional comparative outlook and supporting the authors with ideas and literature. My colleague Imene Cherif coordinated the publication project from the beginning, from the selection of authors to the organization of the national workshops and online meetings with the authors. Her tireless efforts made this publication possible. The publication was also greatly supported by a small but very active Scientific Committee, comprised of Ferdinand Eibl(King’s College London), Nidhal Ben Cheikh, Bessma Momani(University of Waterloo) and Noureddine El Aoufi who was later replaced by Mohamed Said Saadi. I wish to thank them all for their invaluable contributions and support for this project. The publication is set to appear simultaneously in three languages(English, Arabic, and French), which came with its own challenges. A large number of translators were involved: Albert Cherif, Nancy du Plessis, Amel Lafif-Jedidi, Ferid Rahali as well as the teams from Industry Arabic and Docstream. Finally, all languages were proofread and language-edited, by Hazel Haddon(English), Yahia Wagdi (Arabic), and Amel Lafif-Jedidi(French). I would like to thank them all for their great contributions to this project. I am certain that readers will find these studies a useful contribution to the debate around social protection and social spending, both in the MENA region as well as globally. For us, these studies are the beginning of a much larger effort to support the education and public debates on the nexus of social protection, austerity and the role of international financial institutions. Out of this, we hope to shape a future where social protection for all and the building of social protection floors are central aspects of policy making to assure basic welfare for all. Coming social and economic challenges for the MENA region and the world, such as the coming climate emergency, can make a large number of people suddenly dependent on social protection and support. If these systems are then unprepared and underfinanced, the stability and continuity of societies can be endangered very fast. We need to discuss now, with the lessons of the pandemic still in fresh memory, how social protection can be made more accessible and sustainable, without leaving anybody behind. Thomas Claes Project Director Tunis, September 2022 Case Studies from Tunisia, Jordan and Morocco 7

Introduction Salma Hussein This book aims to evaluate some of the main approaches that the International Monetary Fund (IMF) has pursued in Arab countries in the name of furthering social protection. 1 The past 10 years (2011-2021) have constituted a unique decade for a number of reasons. This book comes at a crucial moment due to the combination of three factors in particular. Firstly, this was of course a time when popular demands for greater equality, dignity, social justice, and ending corruption came to the forefront. The region had achieved moderate to high growth rates during the decade leading up to what was at first called the Arab Spring, while rates of unemployment, especially among youth, remained the highest in the world. Meanwhile, rates of poverty continued to rise, or at the very least, remained constant in most countries in the region, especially non-oil-exporting countries. Secondly, public debt accelerated during this decade, especially foreign debt. In the wake of these wide-reaching uprisings, and perhaps in an effort to preclude further unrest, many countries in the Arab world were quick to take on foreign loans. They hoped this would provide relatively more fiscal space to meet demands for social justice. The period from 2011 to 2021, which is the focus of this study, was characterized by dependence on foreign debt. The main driver behind this trend was the IMF: the Fund distributed hard-currency loans, and its experts could approve or reject policies adopted by the countries receiving its loans. The Arab world made a historic number of loan agreements with the IMF during this period. Eighteen IMF agreements were signed with Arab countries between 2010 and 2021. 2 The IMF offered many different kinds of loans, which had different conditionalities attached. Table 1 shows the number of agreements associated with IMF loans in various Arab countries during the period under study. 1 The IMF works with countries from the Arab world to Central Asia, an area it refers to as the Middle East and Central Asia region(MCD). The Arab world falls within the IMF’s MENAP(Middle East, North Africa, Afghanistan and Pakistan) region, which is further divided into oil-importing and oil-exporting countries. The case studies in this book focus on the former, i.e., oil-importing countries in the Arab world. The IMF classifies the three countries that are the subject of this study as “Arab countries in transition.” 2 There were further agreements signed with other Arab countries(Somalia, Djibouti and Mauritania), which signed loan agreements with the IMF in 2020. This number excludes annual reviews carried out by IMF experts in a larger number of Arab countries. These reviews took the form of Article IV consultations, which also include neoliberal demands with regard to fiscal and monetary policies. For example, the IMF pressured Lebanon to lift fuel subsidies, despite the ramifications of this decision for macroeconomics, poverty, and the economic recession. The IMF’s demands also included curtailing inflation through monetary policy alone and reducing budget deficits, especially through freezing hiring and wages for government positions, and cutting back comprehensive or universal social safety nets. 8 Uncovered: The role of the IMF in shrinking the social protection

Table 1. IMF agreements with Arab countries, 2010-2021 Country Tunisia Morocco Jordan Egypt Sudan Yemen Iraq Number of agreements, 2010-2021 Year 3 2013 2016 2020 4 2012 2014 2016 2018 3 2012 2016 2020 3 2016 2020 2020 1 2021 2 2010 2014 2 2012 2019 Type of loan Stand-By Arrangement Stand-By Arrangement Rapid Financing Instrument Precautionary and Liquidity Line Stand-By Arrangement Extended Fund Facility Extended Fund Facility Extended Fund Facility Stand-By Arrangement Rapid Financing Instrument Extended Credit Facility Extended Credit Facility Extended Credit Facility Stand-By Arrangement Stand-By Arrangment Source: Compiled by author from MONA(Monitoring of Fund Arrangements) Database, International Monetary Fund. Thirdly, this decade during which Arab populations sought social justice, decent work, and an end to cronyism came three years after the major global crisis of 2008, and perhaps as a result of this crisis. This crisis, and its repercussions for Arab countries, changed economic thought entirely. In the wake of 2008, neoliberal institutions began to produce a new body of literature that undermined the assumptions of the Washington Consensus, which had shaped the IMF and World Bank(WB) agendas for decades. Many of the new studies revised the assumption that inequality was good and instead linked growing inequality with unsustainable growth, and social safety nets with political stability. The IMF embraced these ideas, while the Arab Spring and fluctuating food and energy costs led the G20 to adopt the concept of social protection floors and to recognize the need for inclusive growth. 3 3 Independent Evaluation Office of the IMF(2017), The IMF and Social Protection Evaluation Report, IMF, Washington DC., p. 1. Case Studies from Tunisia, Jordan and Morocco 9

This decade of intensive Arab cooperation with the IMF provides excellent material for addressing the following questions: How has the IMF’s changing attitude towards social protection shaped the new generation of IMF programs? Has the fiscal space created by IMF programs resulted in the expansion of social protection? Has unemployment risen or fallen? What have been the effects on poverty and inequality? In order to fully address these questions, this book takes as case studies three middleincome, non-oil-exporting countries: Morocco, Tunisia and Jordan. The governments of these three countries worked closely with IMF experts throughout the period under study in order to address their peoples’ aspirations for justice and progress. It would certainly be impossible to publish a book on social protection in 2022 without examining the impact of the COVID-19 pandemic. IMF documents indicate that the financial institution attributed the socioeconomic woes of the countries it was helping during the pandemic to the spread of COVID-19, as if poverty, unemployment and inequality had not existed prior to 2020. It is also worth noting that despite acknowledging the prevalence of these social problems and the importance of government spending to remedy them, the IMF still required the three countries in question, as well as Egypt and Iraq, to reimpose austerity measures and reduce budget deficits as early as 2021. This was the case even though these countries were unable to distribute vaccines to their populations and lacked social safety nets to cover the majority of their citizens. The details of this are explored in greater depth at the end of each of the three country case studies. Defining social protection 4 What is social protection? This might seem like a simple question, but answering it is a complex matter and is the subject of contention among international development and financial institutions. Of these institutions, the IMF uses the most ambiguous definition. In the IMF documents pertinent to the case studies, the IMF’s definition was the most theoretically narrow, and the most detached from practice. Generally speaking, there are two contrasting conceptual approaches to social protection. The first is the approach that has been adopted by international financial institutions, particularly the IMF and the World Bank. 4 The entire research team contributed to the case study analyses in this section, while most of the text was written by Chafik Ben Rouine and Jihen Chandoul. 10 Uncovered: The role of the IMF in shrinking the social protection

The second aproach, which we hope will gain a greater foothold and develop the policies necessary to establish itself, is grounded in the Universal Declaration of Human Rights and espoused by the International Labor Organization(ILO). Social protection, as defined by the ILO, is a human right that must be honored through the pursuit of social policies and programs, as per Articles 22 and 25 of the Universal Declaration of Human Rights. 5 The declaration includes a number of human rights mechanisms that have been adopted by the United Nations regarding the right to social security. This right is also included in the International Covenant on Economic, Social and Cultural Rights. It has also recently been added as one of the 2030 Sustainable Development Goals: to“implement nationally appropriate social protection systems and measures for all, including floors, and by 2030 achieve substantial coverage of the poor and the vulnerable.” The ILO has encouraged countries to adopt social protection floors, 6 which are defined as the ability to access healthcare, basic income security for the elderly and disabled, support for dependent children, and basic income security for the unemployed and the working poor, as well as government programs guaranteeing job opportunities. According to the ILO, social protection floors include four guarantees: 1. Access to essential health care, including maternity care; 2. Basic income security for children, providing access to nutrition, education, care and any other necessary goods and services; 3. Basic income security for persons in active age who are unable to earn sufficient income, in particular in cases of sickness, unemployment, maternity and disability; 4. Basic income security for older persons. The ILO adds that these floors should also provide universal access to basic social services such as healthcare, water, sanitation, education, food security, housing, and any other services that governments consider to be a national priority. The ILO sees social protection floors as at the heart of its policies for poverty eradication, overcoming inequality, and redistributing resources to achieve social justice. 5 Specifically, Article 22 of the Universal Declaration of Human Rights establishes that:“Everyone, as a member of society, has the right to social security and is entitled to realization, through national effort and international co-operation and in accordance with the organization and resources of each State, of the economic, social and cultural rights indispensable for his dignity and the free development of his personality.” Article 25 declares:“Everyone has the right to a standard of living adequate for the health and well-being of himself and of his family, including food, clothing, housing and medical care and necessary social services, and the right to security in the event of unemployment, sickness, disability, widowhood, old age or other lack of livelihood in circumstances beyond his control. Motherhood and childhood are entitled to special care and assistance. All children, whether born in or out of wedlock, shall enjoy the same social protection.” 6 The concept of social protection floors was adopted in 2009 at the International Labour Conference as part of the ILO’s Global Jobs Pact. In 2012, the ILO unanimously approved the Social Protection Floors Recommendation for the national level(Recommendation No. 202) during that year’s International Labour Conference. Case Studies from Tunisia, Jordan and Morocco 11

In 2016, the World Bank adopted the concept of social protection floors, including the above four points. The concept of social protection floors should therefore be considered a basic tenet of citizens’ rights, and separate from aid during crises or charitable aid. In practice, however, the World Bank’s programs and conditions have shrunk universal social protection to a more basic form, without social protection floors. According to the World Bank: Universal social protection coverage includes: providing social assistance through cash transfers to those who need them, especially children; benefits and support for people of working age in case of maternity, disability, work injury or for those without jobs; and pension coverage for the elderly. Assistance is provided through social insurance, tax-funded social benefits, social assistance services, public works programs and other schemes guaranteeing basic income security. 7 However, this World Bank definition overlooks the need to establish universal social protection programs, especially with regard to education, health, decent housing, water and sanitation. As per the vision put forward in this book, social protection in its fullest sense must include a mix of labor policies and various forms of support and security, whether funded through public sources or through societal contributions towards a wide-reaching social safety net that will protect all citizens. The COVID-19 pandemic has reaffirmed the importance of this principle, which the UN highlighted as a priority in its Framework for the Immediate Socio-Economic Response to COVID-19. The programs and policies adopted by international funding institutions have been at odds with this approach of providing universal and comprehensive protection to all citizens at different stages of life. These institutions have instead limited their interpretation of social support to helping the poorest sectors of society. 8 First of all, it is clear that the IMF has not adopted an official definition of social protection. 9 The IMF uses various other terms instead, such as“social safety nets,” or more frequently,“social spending” and“social safeguards,” in discussing broader issues such as education and healthcare. 10 7 World Bank, Social Protection& Labor Overview, https://www.worldbank.org/en/topic/socialprotection/overview. 8 The chapter on Tunisia includes a table that summarizes the approaches of different regional and international funding institutions, in comparison with the rights-oriented vision for social protection espoused by the ILO. 9 Independent Evaluation Office of the IMF(2017), The IMF and Social Protection Evaluation Report, p. 1. 10 Independent Evaluation Office of the IMF(2017), p. 1. 12 Uncovered: The role of the IMF in shrinking the social protection

The IMF does not see social protection as one of its responsibilities, 11 nor as a right in general. However, it has recently begun to approach social protection from the standpoint of macroeconomic stability, since maintaining political and social approval for its macroeconomic policies could depend entirely on preventing growing unrest among the most vulnerable sectors of society. This is what the IMF refers to as a“macro-critical” objective, i.e., an issue that would affect macroeconomic stability. 12 Even within this narrow understanding of social protection, IMF experts have ignored critical issues related to macroeconomic questions in all three case studies examined here. They have insisted on increasing energy prices, for example, or currency devaluations, despite growing pressures on the most vulnerable members of society and widespread protests. We also find that in all three case studies, the IMF depended to a significant extent on the World Bank to establish monetary aid programs targeting the poor. However, the World Bank did not measure any ex post facto impacts of the stringent social measures that the IMF had pursued. Neither institution offered the social protection floors to which the World Bank had committed in 2016, 13 and for which the former director of the IMF had also expressed support in 2009. 14 How do the IMF and the World Bank evaluate social protection in the Middle East? Both the IMF and World Bank see social protection in the Middle East and North Africa as sparse and lacking sufficient government spending. They hold that the social safety net does not cover all citizens and fails to encompass all categories of vulnerability and social risk. According to the World Bank, government spending on social protection in the Middle East and North Africa averages 1% of the national GDP, which is less than the global median and the median for developing countries and of sub-Subharan Africa(1.5%), and less than half the median in Europe and Central Asia. Social protection programs for the elderly, for example, cover less than 25% of the population in the region. 15 The World Bank report discussed in the previous paragraph did not include spending on education and healthcare. However, an IMF report on social spending(including spending on education and 11 Ibid., p. 2. 12 Ibid., p. 4. 13 World Bank(2016), World Bank, ILO announce New Push for Universal Social Protection(press release), World Bank, Washington. https://www.worldbank.org/en/news/press-release/2016/09/21/world-bank-ilo-announce-new-push-for-universal-socialprotection 14 Independent Evaluation Office of the IMF(2017), previously mentioned reference, p. 1. 15 World Bank(2018), The State of Social Safety Nets 2018, The World Bank, Washington, DC. https://openknowledge. worldbank.org/handle/10986/29115 Case Studies from Tunisia, Jordan and Morocco 13

healthcare) observed that“social spending in the region is generally lower than in other parts of the world…Governments in the region devote 10.4% of GDP on average to social spending, compared to an EM[emerging markets] average of 14.2%.” 16 These disparities are even more pronounced with regard to purchasing power parity(PPP) per capita spending. Emerging market countries in the region, including our three case studies, spend an average of USD 1,220 on social outlays, which is less than the average of USD 1,978 spent by emerging market countries globally. 17 The significant additional spending that will be necessary for the region to reach the Sustainable Development Goals attests to the challenge that the region faces in order to provide social protection floors. The median country in the Middle East needs“to spend an additional 5.3 percent of GDP per year by 2030 to achieve five critical SDGs[Sustainable Development Goals].” Some countries might need to spend even more since “the level of public health care expenditure is generally lower than in global comparators.” 18 However, a close examination of the IMF documents pertaining to each of the three case studies— Tunisia, Jordan and Morocco—suggests that there has been an absence of or insufficient set of measures adopted to create the fiscal space necessary to expand the social protection base. The IMF recommended social protection measures that aimed merely to address the social side effects of the measures that the Fund had itself imposed on these countries. The findings from these case studies align with a report issued by the IMF’s Independent Evaluation Office that found that most governments did not have any dialogue with the IMF about social protection. In cases in which dialogue had occurred, the IMF’s recommendations were often“generic based on a mechanical application of an empirical methodology,” and governments usually did not find them relevant to their country’s circumstances, or“sufficiently respectful of their country’s social and cultural characteristics and values.” 19 Approximately 93% of the programs pursued by the IMF between 2010 and 2015, a total of 57 programs, made reference to social spending in low-income countries. However,“only 19 percent of these arrangements…(in 7 countries) contained indicative targets(ITs) defined to focus primarily and specifically on social protection.” Such targets had the least stringent conditionality of IMF programs, and failure to meet these targets do not result in suspension of the IMF agreement or freezing of the IMF payments. Furthermore, the“definitions of the indicative targets in the rest of the arrangements either did not include expenditures on social protection or were too broad or insufficiently specific to determine if social protection expenditures were a meaningful component.” 20 16 IMF staff(2020), Social Spending for Inclusive Growth in the MENA and Central Asia Region, IMF, Washington DC., p. 7. [original hyperlink broken] 17 IMF staff(2020), Social Spending for Inclusive Growth in the MENA and Central Asia Region, p. 7. 18 Ibid., p. 8. 19 IMF Independent Evaluation Office(2017), p. 20. 20 Ibid., p. 21. 14 Uncovered: The role of the IMF in shrinking the social protection

Negative repercussions: The far-reaching impacts of IMF programs IMF programs have exacerbated poverty and inequality in ways that extend far beyond the scope of in-cash or in-kind subsidies. In order to understand the net macro-level socioeconomic impacts of these programs, the social ramifications of these fiscal and monetary policies must also be examined, as well as their effects on growth and its structure, and the extent to which they create decent jobs. Both the World Bank and the IMF lack appropriate models to examine the full effects of these programs. The limited statistical capacities of developing countries and lack of transparency also makes it difficult to access complete data for this kind of independent research. For example, the monetary policy approach that sought to devalue local currencies causes inflation, which pushes wide swathes of the population below the poverty line. Rising energy costs and valuedadded taxes also make inflation worse. IMF experts’ demands for interest rates raises have not only slowed investing but also produced various other negative social repercussions. One is the rising cost of servicing the public debt, which means further shrinking of the fiscal space for social protection, known as the“crowding-out effect.” In addition to reducing the resources available for social protection, and because of growing public debt, public money in the form of interest payments on government debt is moved out of the treasury and into the hands of the wealthiest subset of society, who loan their surpluses to the government. This arrangement exacerbates inequality. Ultimately, this means that public money is redistributed from the poorest sectors of society to the wealthiest individuals at the top of the socioeconomic ladder. This could be named “selective austerity.” Despite efforts to reduce public debt, debt has risen: In the three countries that are the focus of this book, the IMF required cutting public debt relative to GDP. However, the three countries ultimately ended up either with a higher level of debt, or failed to reduce debt to the expected level, even before the COVID-19 pandemic began. Public debt as a percentage of GDP was supposed to fall in oil-importing Arab countries(it was estimated at 100% of GDP in 2021). Rising debt has led to high financing needs, reaching about 50% of GDP(USD 390 billion) during 2021-2022, compared to 2018-2019 levels. 21 Countries such as Tunisia will continue to have the same elevated rates of public debt after 2023, while Jordan might see a downward trend after that time. 22 21 Regional Economic Outlook MENA(April 2021), International Monetary Fund, Washington DC., p. 4. 22 Regional Economic Outlook MENA(April 2021), p. 6. Case Studies from Tunisia, Jordan and Morocco 15

IMF programs have once again countributed to eroding fiscal space, particularly given the lack of focus on reforming tax structures so that the greatest tax burden falls on those with the highest income, whether individuals or corporations. In light of the conditions that IMF has imposed on these three countries and the recommendations it has proposed, the common threads among the case studies with regard to social protection include: • Reducing budget deficits through cutting social spending and wage bills while also increasing indirect taxes, even when the budget deficit was around 4%, as was the case in Morocco in 2012, according to the agreement on emergency liquidity assistance. • Reducing universal subsidies in all forms and replacing them with cash subsidies targeting only the poorest sectors of society. In the three case studies, energy subsidies were removed, one of the IMF’s most severe demands in terms of social impact. This condition was usually imposed as a prior action, i.e., a measure that had to be implemented before the signing of the agreement with the IMF. • The IMF has overlooked education and healthcare spending, which are the most important pillars of universal social protection, and also constitute the third and fourth Sustainable Development Goals. It has also neglected to develop governance of public institutions dealing with education and health, even when these are at the heart of the program’s goals, as was the case in Morocco. The three case studies examine public spending on healthcare and education in order to evaluate the IMF’s actual interest in these areas, particularly since the most recent report of the IMF’s Independent Evaluation Office overlooked education and healthcare as categories of social spending after examining these factors in a special report in 2013. • In all three case studies, the IMF engaged in bias by omission with regard to social protection and labor market policies. None of its programs in the region have dealt with unemployment benefits or the need for a plan to distribute such aid, 23 nor have IMF reform programs included pension funds for workers in the informal economy, funding for active labor market policies, or similar programs to address structural problems raised in the reports of the IMF experts. The Independent Evaluation Office found that less than 10% of agreements between the IMF and EM countries(including the three case studies) included social spending floors, and less than 15% of agreements included conditionality to strengthen social safety nets. In general, the Independent 23 Ibid., p. 16. 16 Uncovered: The role of the IMF in shrinking the social protection

Evaluation Office report found that developed countries received more recommendations and conditionalities than any other region with regard to social protection. The three case studies focus on the key common recommendation among IMF programs with regard to effects on social protections, namely cash transfer programs that the IMF implements in exchange for lifting universal subsidies(on food or energy). The Independent Evaluation Office has acknowledged that these measures are implemented without regard for social protection; rather, they see the energy subsidy removal as linked to eliminating budget deficits. This policy has affected the Middle East more than any other region in the world, according to the Independent Evaluation Office report. 24 It is worth noting that this has happened as part of all types of loans and agreements between these three countries and the IMF(see Figure 1). In all three case studies, it was clear that despite all the new programs that the World Bank had helped implement, social protection had been eroded in one way or another over the course of the decade under study. These countries cut most of their energy and food subsidies, which meant increased economic pressure on much of the population due to reduced disposable income. It also exacerbated inflation, as did other IMF-recommended policies, which led to the erosion of actual income, particularly at the bottom of the socioeconomic ladder. Meanwhile, these programs were unable to reach the poorest sectors of society that they had been trying to target. They failed to reduce poverty or to protect growing numbers of vulnerable persons whose livelihoods were affected such that they were unable to provide for their children’s health and education, based on government spending in these areas. The savings achieved through lifting universal subsidies did not result in higher sustainable allocations for education and healthcare spending, nor a shift towards subsidy programs for the lowest-income groups. At the same time, the IMF’s Independent Evaluation Office report indicated that the recommendation to cut universal subsidies, especially for energy, aimed to reduce budget deficits rather than to increase social protection. Research methodology The book focuses in particular on the erosion of universal subsidy programs and their replacement with targeted subsidy programs. It draws upon the three case studies to generate a systematic analysis in an international context. The three case studies were all subject to peer and specialist review. 24 Ibid., p.15. Case Studies from Tunisia, Jordan and Morocco 17

The case studies evaluate the targeted monetary aid programs with regard to their various impacts on poverty and social justice, as per the specific circumstances of the programs in each of the three countries. We also examine government spending on healthcare and education in these three countries as a percentage of GDP, as an indicator of the IMF’s engagement with the most important elements of social protection and the third and fourth Sustainable Development Goals. The case studies aim to shed light on complementary aspects of IMF programs with regard to the question of social protection: • The Tunisian case study engages in an in-depth reading of documents published by the IMF evaluating the economic situation in the country. These documents entirely overlook what the IMF calls macroeconomic vulnerabilities, such as the degree of popular and political support for the conditions attached to IMF loans, even though the IMF itself has committed to engaging with these principles. This is the first time that such an in-depth study on this topic has been carried out by French- and Arabic-speaking researchers with access to a wealth of documents in these languages. • The Jordanian case study focuses on the government’s success in ensuring that teachers’ salaries are excluded from the conditions of its IMF agreements, unlike most similar agreements with countries in the region, which stipulate the need to reduce public wages as a percentage of GDP. The study also analyzes changes in income tax structure in the IMF agreement. • The Moroccan case study provides an in-depth analysis of the social protections that the kingdom has provided for its citizens. It also lay outs the most important social and economic factors prior to the pandemic, following a decade of social demands and four agreements with the IMF. Finally, the book offers the first analysis of the COVID-19 pandemic’s effects on the economic policies that the IMF has recommended to the three governments in question. For example, the IMF has promoted a systematic process for restructuring debt with all developing countries. In practice, however, debts have been frozen for very few low-income countries. The IMF has used its lending instruments to create fiscal space for middle-income countries, including our three case studies. Figure 1 depicts the scope of IMF lending in different regions of the world in 2020-2021, and shows that the Middle East was the IMF’s third biggest client during the COVID-19 pandemic. 18 Uncovered: The role of the IMF in shrinking the social protection

Figure 1. The IMF lending map, 2020-2021(in nms of SDR*) 90,000 80,000 70,000 2020 2021 82,086 60,000 50,000 40,000 47,335 30,000 20,000 10,000 0 1,345 1,570 Asia& Pacific 4,419 3,661 14,977 7,469 19,014 7,990 Europe Middle East & Central Asia Sub-Saharan Africa Western Hemisphere April 30, 2020: SDR1= USD 1.36640 September 15, 2020: SDR1= USD 1.41664 * The Western Hemisphere refers to North and South America. Argentina has taken on the lion’s share of credit, constituting half of the region’s IMF lending in 2020. Source: IMF annual reports(2020-2021). 25 Furthermore, the three case studies indicate that, according to the IMF, the COVID-19 pandemic and its impacts on the most vulnerable sectors of society should not hold back progress in lifting universal subsidies, devaluing local currency, or freezing austerity measures. All of the temporary public spending packages were to stop after the first year of the pandemic, despite slow vaccine distribution, the vulnerability of informal workers, and slowing or declining economic growth. The IMF’s sole consideration is reducing public budget deficits. The IMF expects oil-importing Arab countries to see deficits fall after 2021, as a result of recovery, the end of fiscal aid packages, and efforts to rein in public spending. 26 25 https://www.imf.org/external/pubs/ft/ar/2020/eng/downloads/imf-annual-report-2020.pdf https://www.imf.org/external/pubs/ft/ar/2021/eng/downloads/imf-annual-report-2021.pdf 26 Regional Economic Outlook MENA(April 2021), previously mentioned reference, p. 6. Case Studies from Tunisia, Jordan and Morocco 19

In sum, IMF programs and recommendations require ongoing borrowing and paying of debts, even at the expense of exacerbating the prevalence of disease or hunger, or of pushing the population below the poverty line. Three workshops were held to discuss the findings from the three case studies and to develop and discuss recommendations . 27 This book is being published at a historical moment similar to that in Latin America at the beginning of the foreign debt crisis in the 1980s, when most developing and low-income countries became tied up in foreign loans that initially seemed relatively inexpensive and necessary for development and growth. A decade later, these debts proved a heavy burden for Latin American countries, as the disadvantages of working with the IMF and the World Bank became clearer. The global crisis of 2008 then had the effect of pushing the Bretton Woods institutions to conduct important analyses of, and ultimately make revisions to, the Washington Consensus. We find ourselves today at a similar juncture, with foreign debt service burdens increasing in developing and low-income countries. The IMF and the World Bank have adopted a different discourse than that which they espoused in the 1980s and 1990s. This contemporary discourse is grounded in awareness of the ills of inequality of income and of opportunity, and supports imposing international taxes on giant corporations and progressive taxes on the rich. This discourse also ostensibly supports the Sustainable Development Goals and social protection. We offer here the first book to conduct a systematic study of the major disparities between this discourse from the IMF and the IMF’s practice on the ground, and we hope it will be useful to the reader. 27 Case studies also include sections, rich in information, on the historical background of the country’s cooperation with the IMF, on the country’s social protection systems, and on the conditionality attached to different types of IMF programs. 20 Uncovered: The role of the IMF in shrinking the social protection

HOW INTERNATIONAL FINANCIAL INSTITUTIONS AFFECT SOCIAL PROTECTION IN TUNISIA by Chafik Ben Rouine and Jihen Chandoul Translation from French Nancy Du Plessis

Table of contents I. Introduction 22 II. International financial institutions and social protection 25 III. The impact of IFIs on social protection in Tunisia 32 IV. A reading of IMF documents: The subsidy reform central to IMF structural adjustment in Tunisia 43 V. Conclusion 57 VI. Bibliography 59 Appendices 62 I. Introduction From 2013 until today,“post-revolution” Tunisia has been marked by the forceful intervention of international financial institutions(IFIs)—especially the International Monetary Fund(IMF) and the World Bank(WB)—with respect to its economic and social reforms. Since 2013, the IMF has granted three loans to Tunisia: a Stand-By Arrangement(SBA) in 2013 and 2015, 1 an Extended Fund Facility (EFF) between 2016 and 2020, 2 and an emergency assistance loan to address the outbreak of COVID19 3 under the Rapid Financing Instrument(RFI) in 2020.The IMF became involved in Tunisia as part of the Deauville Partnership, launched in 2011 following uprisings in many Arab countries. It is an alliance and cooperation of G8 countries, Turkey, the Arab Gulf states and IFIs to help transitioning countries access conditional financing in return for implementing neoliberal structural reforms. 4 1 https://www.imf.org/external/np/loi/2013/tun/052413.pdf 2 https://www.imf.org/external/np/loi/2016/tun/050216.pdf; https://www.imf.org/external/np/loi/2018/tun/071118.pdf 3 https://www.imf.org/en/Publications/CR/Issues/2020/04/14/Tunisia-Request-for-Purchase-Under-the-Rapid-FinancingInstrument-Press-Release-Staff-Report-49327 4 Chandoul, J.(2015).“Le Partenariat de Deauville, à l’origine des politiques économiques en Tunisie”. Observatoire tunisien de l’économie. 22 Uncovered: The role of the IMF in shrinking the social protection

The IMF intervened during an economic and social crisis exacerbated by the outbreak of COVID-19 in 2020. Since 2013, Tunisia has witnessed massive protests and sit-ins by new social movements demanding economic and social reforms to reduce inequality, distribute wealth more equitably, and ensure decent living conditions. In January 2018, there were protests against the austerity measures and the IMF’s involvement. The emergency brought about by the COVID-19 pandemic has exacerbated Tunisia’s economic and social crisis: household consumption has dropped by 8%, overall investment is down by 4.9%, exports have plummeted by 9.6%, and GDP has declined by 4.4%. A study by the United Nations Development Programme(UNDP) and the Tunisian Ministry of Development, Investment and International Cooperation estimates that, as a direct consequence of the lockdown and the ensuing drop in production, unemployment increased from 15% to 21.6% in the first quarter of 2020, with about 475,000 people falling below the poverty line. 5 The global crisis caused by the COVID-19 pandemic has made it all the more important to strengthen social protection schemes, including for the informal sector,­and to increase social spending. According to the most recent study published by the National Office of Statistics(INS) in September 2020, 6 the informal sector accounts for 44.8% of jobs in Tunisia, 81.2% of which are held by men and 18.8% by women. Agriculture and fisheries remains the sector most affected by informal labor, with 85.6% of employees undeclared, followed by building and public works(69.2%) and trade(64.7%). Almost one Tunisian in six continues to live in poverty, with strong regional inequalities and poverty rates especially high in the depressed areas of the Centre-West and North-West. 7 The COVID-19 crisis has put social protection back on the global agenda, particularly Target 1.3 of the first Sustainable Development Goal(SDG) on instituting universal social protection floors(SPFs). This describes a basic set of social rights, including access to essential services. 8 In 2017, as the International Labour Organization(ILO) recommends, Tunisia’s Ministry of Social Affairs began conducting studies and examining the feasibility of setting up a social protection floor to remedy the obvious shortcomings of the social security system. 9 This study primarily aims to analyze IMF involvement in Tunisian social protection, particularly its non-contributory schemes, between 2013 and 2021. It does not include the contributory social protection system, the national social security funds. 5 UNDP and the Tunisian Development, Investment and International Cooperation Ministry(June 2020).“l’Impact de la pandémie du COVID-19 sur l’économie tunisienne pour l’année 2020”. https://www.tn.undp.org/content/tunisia/fr/home/ library/environment_energy/etude-sur-l-impact-economique-du-covid-19-en-tunisie-.html 6 INS(September 2020).“Les indicateurs de l’emploi informel 2019”. 7 INS(September 2020). Tunisia Poverty Map. 8 Indicator 1.3.1:“Proportion of population covered by social protection floors/systems, by sex, distinguishing children, unemployed persons, older persons, persons with disabilities, pregnant women, newborns, work-injury victims and the poor and the vulnerable.” 9 The reform process is being carried out with the support of the ILO and UNICEF. The Tunisian government plans to create four pillars guaranteeing universal health care, a minimum income for poor people, a minimum income for the elderly and persons with disabilities, and a minimum income for children: despite Tunisia’s progress in extending social protection coverage, several population groups are still not covered by insurance or social assistance schemes. How international financial institutions affect social protection in Tunisia 23

“Non-contributory social protection” refers to: • targeted transfers through social safety nets and social assistance programs; and • non-targeted transfers through public subsidies(food and energy). Several population groups remain beyond the reach of contributory social protection schemes because of their precariousness and their informal work. Social protection for these groups, who have limited contributory capacity due to inactivity, unemployment, illness, old age and disability, is covered by non-contributory social assistance programs. Since 2012, IMF intervention in Tunisia has focused on targeted and non-targeted transfers by conditioning financial support—first by reducing energy subsidies to direct some of the resource savings toward targeted social safety nets, and then eliminating food subsidies and replacing them with cash transfers targeting the poorest(when the necessary database is in place). The IMF maintained and confirmed its approach in its latest report on Tunisia in February 2021. The report is part of the Article IV consultation that considers the effects of the health crisis. The IMF stated that, once the new database of the poorest households(Amen Social) is developed, the authorities should make cash transfers to them to offset the eliminated energy subsidies. 10 It“calls on the authorities to strictly prioritize spending in favor of health and social protection, while exerting strict control over the civil service bill, ill-targeted energy subsidies, and transfers to inefficient stateowned enterprises.” 11 The IMF report emphasizes that reforming targeted and non-targeted transfers, namely energy and food subsidies, must be some of the first measures undertaken to alleviate the social consequences of the COVID-19 pandemic—as part of“a social compact.” 12 In this study we have also included the World Bank, which is central to implementing the IMF’s approach to social protection and has supported the IMF in Tunisia since 2012 with several loans and technical assistance programs, including loans during the COVID-19 emergency . 13 The IMF report of 2021 states,“The authorities are working with the World Bank to expand the Amen Social program to facilitate cash transfers and other services.” 14 We seek to answer the following questions: How do the IFIs approach social protection? Do the IMF and World Bank take different approaches? How do their interventions impact non-contributory social protection? 10 IMF Country Report: Tunisia, 2021 Article IV Consultation, 26 February 2021. 11 IMF Country Report: Tunisia, 2021 Article IV Consultation, 26 February 2021, p. 23. 12 Ibid.“Given past reform failures and resistance, staff urges the authorities to consult and communicate with the broader public a medium-term reform program(…). To be credible and gain buy-in, the reform program(…) would need to be supported by a social compact, with the main stakeholders committing to support reforms within their remit.” 13 World Bank(2013). Towards greater equity: energy subsidies, targeting and social protection in Tunisia. Washington, DC; World Bank(2013). Tunisia Social Protection Reforms Support Project, Project Appraisal Document(P144674). Washington DC. In 2020, the World Bank approved USD 35 million for Tunisia to fight the COVID-19 pandemic; in 2020, it granted USD 175 million to mitigate the economic impact of COVID-19. 14 IMF Country Report: Tunisia, 2021 Article IV Consultation, 26 February 2021, p. 19, n. 8. 24 Uncovered: The role of the IMF in shrinking the social protection

The study is based primarily on a review of the literature, including loan and policy documents and research studies on the IMF and World Bank positions on social protection in general, and their interventions in Tunisia in particular . It also analyzes laws and decrees adopted in exchange for conditionality, including the latest decrees adopted during the COVID-19 pandemic, and documents on reforming Tunisia’s social safety net. A budgetary analysis based on Tunisian finance laws since 2012 was carried out in order to shed light on how energy and food subsidies in budgets have evolved over this period. The study also used qualitative interviews with Nidhal Ben Cheikh, a social protection expert and former research director at the Center for Research and Social Studies(CRES), and others . The study has two parts. In the first, we present our analysis of how the IFIs define and approach social protection; in the second we analyze the impacts and implications of its approach to reforming social protection in Tunisia since 2012, including the evolution of the socioeconomic crisis linked to the COVID-19 pandemic. II. International financial institutions and social protection Defining social protection There is no international consensus on how to define“social protection,” and different actors have adopted different definitions and approaches. Here we analyze the IMF and World Bank definitions and how they differ from United Nations(UN) definitions, in this case the ILO’s human-rights-based definition. Within the IMF, the notion of social protection has evolved over many decades. In June 2019, the IMF published a policy paper on its commitment to social spending 15 in which social protection includes social insurance and social assistance programs. Social insurance aims to protect households against shocks that may negatively affect their income and welfare and is usually financed by contributions or taxes on wages(contributory social protection). Social assistance/social safety nets aim to protect households against poverty and are financed by government revenues(non-contributory social protection). Social spending is defined as social protection, as defined above, with country-specific expenditures on basic education and healthcare. 16 15 IMF Policy Paper, A strategy for IMF engagement on social spending, June 2019. 16 Basic education can include primary and secondary education; basic health often refers to a nationally defined basic healthcare package. How international financial institutions affect social protection in Tunisia 25

In its intervention, the IMF includes social spending, the need to address inequalities in IMF surveillance activities and the important role of social spending in promoting sustained and inclusive growth, as well as the need to provide resources to achieve the SDGs on education and health. The IMF also includes the need to use social protection floors in its programs. We note that social protection has never been seen as one of the IMF’s responsibilities. As its policy paper on social spending confirms, the IMF has always relied on the relevant expertise of other institutions, mainly the World Bank. The World Bank is the main institution promoting and implementing the IMF definition and approach to social protection. It defines social protection as“help[ing] individuals and societies manage risk and volatility and protect them from poverty and destitution—through instruments that improve resilience, equity, and opportunity” 17 including social assistance programs(social safety nets) and social insurance(mainly old age and disability pensions and unemployment insurance). In the same document, the World Bank strategy on social protection systems for the Middle East and North Africa(MENA) region shows that its definition is quite similar to that of the IMF. In fact, its main objective there is to move away from fragmented social protection and employment programs toward:(i) effective social safety nets(e.g., targeting cash assistance and mandatory employment programs), and(ii) reforming non-targeted transfers(food and energy subsidies) and improving targeting to increase effectiveness. 18 Although the World Bank is one of the main actors thinking about social protection, other conceptions and definitions promoted by UN organizations are increasingly influencing the debate. These include human rights mechanisms such as the United Nations Economic and Social Council(ECOSOC), as well as the ILO and UNICEF(the United Nations International Children’s Emergency Fund). There are the two main—and contradictory—views of social protection: the view promoted by IFIs, led by the World Bank; and the ILO’s human rights approach. The rights-based approach does not define social protection as an urgent and temporary response to a crisis situation, nor as charity, but as a set of permanent guarantees stipulated in law that define beneficiaries as rights holders(see Appendix F for a comparison of the definitions). These definitional differences reflect two totally different approaches to social protection, which have important implications. 17 World Bank,“The World Bank’s Social Protection& Labor Strategy 2012–2022: Resilience, Equity and Opportunity”, p. 1. 18“The World Bank’s Social Protection& Labor Strategy 2012–2022” notes that subsidy reform will invariably harm the middle class and be very difficult to implement in the MENA region(p. 70). 26 Uncovered: The role of the IMF in shrinking the social protection

The IFIs’ residual approach This section uses these definitions to analyze IFIs’ approach to social assistance and the extent to which it is fundamentally different to the rights-based approach. The economic crisis in East Asia triggered a shift in how the IFIs approach social protection, in light of growing criticism of the impacts of the structural adjustments of the 1980s on people. As a result, the World Bank, which had been working with the IMF, advocated a new, social risk management approach. 19 This consists of adopting upstream social risk mitigation measures during a crisis, trying to address extreme poverty by using social safety nets for the poorest populations. Social safety nets are minimum cash transfers to the poorest that aim to repair the trauma of the economic crisis and structural adjustments. Viewed as compensation to the poorest for austerity measures, they have become the World Bank’s main social protection programs. The World Bank defends a residual and minimalist approach to social protection which translates into a policy of“social management risk”— which complements that of the IMF. The IMF’s social spending strategy paper 20 emphasizes the principles of fiscal sustainability and expenditure efficiency in assessing social protection and social spending. In other words, social spending is considered critical to fiscal and macroeconomic balances if it is unsustainable, inadequate and inefficient under current policies. These principles are IMF priorities that justify policies of optimizing and rationalizing spending to reduce the budget deficit in austerity programs. The IMF uses conditionality to encourage the reduction and/or elimination of non-targeted transfers—food and energy subsidies—and the reallocation of part of the savings to welltargeted and well-designed pro-poor social safety net programs. Furthermore, although the IMF now includes social spending, including for education and healthcare, this is more of a public relations exercise than a real step forward. Indeed, in IMF programs, social protection floors/social safety nets and public spending on education and healthcare are defined as “indicative targets”—non-binding guidelines—whereas the main macroeconomic objectives of IMF loans are binding and must be implemented by states. In Table 1, we summarize the differences between the two approaches and their implications for social protection. 19 World Bank(1990). World Development Report 1990: Poverty; World Bank(2001). World Development Report 2000/2001: Attacking poverty; World Bank(2001). Social Protection Sector Strategy: From Safety Net to Springboard; World Bank (2006). World Development Report 2006: Equity and Development. 20 IMF Policy Paper: A Strategy for IMF Engagement on Social Spending: Background Papers, June 2019. How international financial institutions affect social protection in Tunisia 27

Table 1. IFI approach to social protection vs. human-rights-based approach IMF/World Bank’s residual approach to social protection Transformative/human-rights- based approach(ILO/UN) Concept Social protection is a plaster to offset the negative effects of structural reforms or minimize social risks Social protection is a human and justiciable right that the state is obligated to ensure through investments/policies and by reducing inequality and poverty Beneficiaries The poorest 21 Universal—even if implemented gradually on priority populations Central principle Effective budgeting and the“macrocriticality” of social protection Human rights/social justice Efficient public spending to maintain the budget deficit—which explains why non-targeted transfers are abolished to rationalize spending Tools Social safety nets defined as a minimum A social protection floor to guarantee a (poverty line) decent standard of living Positioning vis-à-vis the budgetary space Reduce fiscal space under austerity measures by streamlining social protection spending Challenge austerity measures and their impact on fiscal space, social spending and social policies: increase fiscal space and allocations for social spending and protection Objectives and disparities Approach Short-term measures to mitigate impacts on the most vulnerable groups without addressing inequalities and structural poverty Residual approach to mitigate social risks Short-, medium- and long-term social policies to ensure social security for all and reduce inequalities Transformative, human-rights-based approach to reduce inequalities upstream of interventions. This raises questions about the feasibility and effectiveness of targeting to fight disparities, as advocated by the IFIs, as opposed to broader social protection programs that promote universal social protection schemes. 21 The World Bank’s approach to social safety nets has evolved somewhat by integrating targeted transfer programs with conditions through participation in programs to boost human capital(such as free healthcare, along with cash transfers). However, this type of conditionality only has value if the problem is related to the demand for social services. If the social services are lacking or low quality, a conditional program would be of no great interest(Merrien, F.-X., 2013, La protection sociale comme politique de développement: un nouveau programme d’action international, International Development Policy, Revue internationale de politique de développement, pp. 68–88.) and would not address structural inequalities and vulnerabilities. 28 Uncovered: The role of the IMF in shrinking the social protection

Targeting methodologies: Different approaches and their limitations The World Bank’s preferred targeting method is the heavily criticized proxy means test(PMT) 22 that attempts to predict a household’s level of well-being using a statistical model. It was developed in response to the difficulties of making real assessments of household incomes in developing countries with large informal economies. PMT limitations in targeting the poorest The PMT has limits in identifying the poorest households. Targeting is costly and application requires skill and capacity. 23 The method suffers from inclusion and exclusion errors due to significant information asymmetry resulting from the minimal availability of disaggregated data and lack of regular surveys of households. A PMT cannot track the dynamic nature of household income. 24 Given the predominantly informal and fluctuating nature of developing country economies and their weak capacity for conducting censuses and household surveys, in most such countries, many of the poorest households are not targeted. An Australian study 25 evaluated the PMT using data from Bangladesh, Indonesia, Rwanda and Sri Lanka and found that errors of exclusion(under-coverage) and inclusion(leakage) ranged from 44% to 55% when 20% of the population was covered, and from 57% to 71% when just 10% was covered.“At these coverage levels – which are realistic for many programs in developing countries with limited financial resources – eligible households have no better than a one-in-two chance of being selected, and in some cases even view proxy means testing as a lottery.” 26 Other studies have also revealed exclusion errors in the method. In Indonesia, 93% of the poorest 5% of households in the Keluarga Harapan (PKH) conditional cash transfer program were not included. 27 An Oxford Policy Management study found that with 62% of the poorest population in northern Kenya targeted, 26% were excluded from the 22 This methodology estimates household income by linking indicators to household spending or consumption and is based on national household surveys. Such costly surveys are not regularly conducted in developing countries, so the PMT relies on a set of proxies(usually between 10 and 30) that best reflect a household’s situation. Indicators include demographic characteristics(e.g., age of household members and household size), human capital characteristics(the head of household’s educational level and the children’s schooling), physical characteristics of the home(type of roof or dwelling), durable goods(refrigerators, televisions, cars, etc.) and productive assets(land or animals). To calculate a score for each household, each indicator is weighted according to its estimated impact on household spending. Households below the defined threshold are eligible for the social assistance program. 23 Kidd, S., Gelders, B., and Bailey-Athias, D.(2017).“Exclusion by design: An assessment of the effectiveness of the proxy means test poverty targeting mechanism”. ILO Social Protection Department. 24 Ibid. 25 Australian Aid(2011). Targeting the poorest: An assessment of the proxy means test methodology, p. 31. 26 Ibid. 27 Alatas, V., Banerjee, A., Hanna R., Olken, B. A. and Tobias, J.(2012). Targeting the Poor: Evidence from a Field Experiment in Indonesia. How international financial institutions affect social protection in Tunisia 29

Hunger Safety Net Programme. 28 In Mexico, the Oportunidades program, which covers about 20% of the population, was estimated to have an exclusion error of around 70%. 29 In Cambodia, approximately 56% of households living in poverty were excluded by the ID-Poor PMT targeting mechanism. 30 Similarly, according to an ILO study assessing PMTs, 31 governments that choose a restrictive approach to poverty targeting can make them somewhat more effective over time. But improving targeting implies significantly higher administrative costs to collect more information on the target population’s income or welfare indicators for the authorities to triangulate the PMT score and confirm and adjust. They can also invest in grievance mechanisms for citizens to appeal PMT results by reporting their actual incomes, which can reduce exclusion or inclusion errors but not eliminate them. For example, despite significantly improving its PMT, Georgia’s targeted social assistance program still has exclusion errors of 50%. 32 CRES’s 2017 evaluation of social assistance programs in Tunisia 33 linked the challenges and difficulties of targeting to“institutional weaknesses, poor coordination between various state services, and the large informal sector which prevent the existence of a tax administration capable of observing the income level of all families.” The report also affirms that the social protection floor“is the most adequate and appropriate solution for freeing a large part of the population from the yoke of chronic poverty and for empowering economically vulnerable populations or those with specific needs(the elderly, the handicapped etc.).” Price subsidies like those for food or energy, which are considered in broader social protection schemes, are often deeply embedded in the political and institutional structures of developing countries. Often, the targeted social assistance that replaces them, such as cash transfer programs, are quite or completely new, complex, and costly to implement, which makes them subject to cuts that leave many vulnerable households without any coverage, risking further impoverishment given the rise in prices caused by cutting subsidies during a recession. PMTs belong to a fiscal and restrictive approach that favors low taxation and limited social spending and targets extremely poor people to reduce costs. They are completely different from the rights28 Silva-Leander, S. and Merttens, F.(2016). Hunger Safety Net Programme: Assessment of programme targeting – Oxford Policy Management. 29 Veras, F., Peres, R., and Guerreiro, R.(2007). Evaluating the Impact of Brazil’s Bolsa Família: Cash Transfer Programmes in Comparative Perspective, IPC evaluation note No. 1, International Poverty Centre, Brasilia, Brazil. 30 World Bank(2011). Assessment of the Identification of Poor Households(IDPoor) Program’s Procedures and Implementation. Draft Technical Report for the Human Development – Social Protection Unit East Asia and Pacific Region. Washington, DC. 31 Kidd, S., Gelders, B., and Bailey-Athias, D.(2017). Exclusion by design: An assessment of the effectiveness of the proxy means test poverty targeting mechanism. International Labour Office Development Pathways. 32 Baum, T., Mshvidobadze, A., and Posedas, J.(2016). Continuous Improvement: Strengthening Georgia’s Targeted Social Assistance Program. Washington, DC: World Bank Group; Gelders, B. and Kidd, S.(2016). Child Wellbeing and Social Security in Georgia: The case for moving to a more inclusive national social security system. Georgia: UNICEF. 33 CRES(May 2017). Évaluation de la performance des programmes d’assistance sociale en Tunisie: Pour optimiser le ciblage des pauvres et freiner l’avancée de l’informalité, p. 33. 30 Uncovered: The role of the IMF in shrinking the social protection

based approach that favors using social protection floor to progressively achieve universal social protection. A PMT can be improved by combining it with other means of verification to increase coverage of the poorest. With respect to targeting, greater coverage of programs reduces the exclusion of poor people. The progressive achievement of universal social protection is only possible if the authorities recognize that increasing social spending is key to reducing inequalities: that makes it possible to expand coverage so programs can become much more effective, either by using universal categories or a combination of universal and targeted schemes, some based on income verification and some not. Alternative targeting method for universal approaches Universal social protection schemes, which are much more likely to benefit all poor households and are less costly in terms of counting capacity, do not rule out targeting. Indeed, in 1991, American sociologist and political scientist Theda Skocpol suggested the potentially progressive method of “targeting within universalism,” which is less likely to lead to errors than simply targeting cash transfers, as Slater and Farrington wrote in 2009. Other targeting methods involve singling out particularly vulnerable population groups on the basis of gender, age or ethnicity(women, children, older people, ethnic minorities), as the ILO, UNICEF and other organizations advocate. While that might result in individuals or households that are not especially poor benefiting from transfers, it is less costly and complex to implement than systems that only select by income.“Categorical” targeting does not stigmatize individuals and has greater public support. A growing number of experts and organizations advocate programs like universal child benefits and non-contributory oldage pensions that are based on simple criteria. How international financial institutions affect social protection in Tunisia 31

III. The impact of IFIs on social protection in Tunisia Subsidies in Tunisia Who benefits? One of the first questions about promoting a new policy of targeting subsidies in Tunisia concerns their redistributive effect: to justify a structural reform of food and energy subsidies, we must know who benefits. We also examine the targeted social transfer programs already in place in Tunisia. The redistributive effect of food subsidies In 2013, coinciding with the start of the first IMF program in Tunisia, CRES and INS in partnership with the African Development Bank(AfDB) studied the redistributive effect of food subsidies(CRES/INS 2013). 34 Among the objectives was assessing how food subsidies impact inequality. As the authors point out, the General Compensation Fund(CGC) did not introduce food subsidies to fight inequality. The main objective of food subsidies was to control inflation and preserve purchasing power, to strengthen Tunisia’s competitiveness through ensuring low wages by limiting pay increases, and to improve nutrition for all Tunisians. The CRES/INS study describes food subsidies, which count as price subsidies, as indirect transfers, as opposed to direct social transfers. The authors used two different approaches to assess the redistributive effect of food subsidies. The first, by quintile, makes it possible to evaluate the distribution of food subsidies by population group, from the poorest(the first quintile) to the richest(the last quintile). The study shows that“real indirect transfers…account for 7.7% of the real value of per capita household consumption in the first quintile, compared with only 1.5% of the real value of per capita consumption in the richest quintile. The share of subsidies in the real value of consumption thus decreases as per capita consumption increases.... Thus, the food subsidy system in Tunisia is relatively progressive.” 35 34 CRES and INS(2013). Analyse de l’impact des subventions alimentaires et des programmes l’assistance sociale sur la population pauvre et vulnérable. 35 CRES/INS 2013, previously mentioned reference, p. 15. 32 Uncovered: The role of the IMF in shrinking the social protection

The quintile approach shows that in relative terms, food subsidies benefit the poorest more than the richest and that the relative benefit decreases as wealth increases. This finding is all the more important as discussion about the study has created some bias around the debate on food subsidies. In fact, to justify reforming them, public attention was focused on subsidy distribution in absolute terms, rather than on the consumption level of each class. Referring to a report in the official press agency, TAP, the online newspaper Kapitalis published an article headlined:“Tunisia: Food subsidies actually benefit the rich.” 36 Most analysts have focused on the dichotomy between poor and non-poor, without analyzing the subsidy system’s progressive nature or judging its unfairness. For example, the World Bank summarizes its report,“The Unfinished Revolution”, which has been very influential in framing the debate, by stating,“The current model relies mostly on untargeted food and fuel subsidies, which are expensive and inequitable—because they largely benefit the rich.” 37 A footnote accompanying this assertion maintains,“While food subsidies in Tunisia are significantly less inequitable, yet they also benefit the rich the most.” 38 The graph from the 2013 CRES/INS study summarizes the distribution of food subsidies in Tunisia by social class, showing that it is not true that food subsidies benefit the rich more. They actually more generally benefit the middle class. Figure 1. Distribution of food subsidies in Tunisia 9.2% Poor 22.8% Household leaks 7.5% Upper class Source: CRES/INS(2013). 60.5% Middle class 36 http://www.kapitalis.com/kapital/23692-tunisie-les-subventions-alimentaires-profitent-plutot-aux-riches.html 37 World Bank(2014). The Unfinished Revolution: Bringing Opportunity, Good Jobs and Greater Wealth To All Tunisians, p. 30. 38 Ibid., p. 32. How international financial institutions affect social protection in Tunisia 33

The authors’ second“continuous approach” using the Lorenz curve shows that the food subsidy system is indeed progressive. 39 Not only is Tunisia’s food subsidy system progressive, but its contribution to reducing inequalities also comes very close to perfect equity. Yet while the CRES/INS study supports the conclusion that the food subsidy system is progressive and equitable, the World Bank’s 2014 report asserts the opposite, without citing the CRES/INS study, which must be referred to when judging the fairness of the subsidies. The World Bank concedes that food subsidies are“less inequitable” than energy subsidies but refuses to call them progressive, much less equitable. Since the World Bank has always promoted a targeted approach to subsidies, this shows how important it is to analyze the policy bias of IFIs with respect to subsidy reforms. This is particularly important because the only study on the redistributive effect of energy subsidies was conducted by the World Bank, as we will see in the next section. It is also important to assess the role of food subsidies on the agricultural sector and not just as a reductionist analysis of consumers. These subsidies are mainly regulated by World Trade Organization (WTO) agreements and complement the way international institutions(the IMF, the World Bank or the WTO) treat food subsidies. The redistributive effect of energy subsidies The only French-language study on the redistributive effect of energy subsidies in Tunisia was carried out by the World Bank in 2013: Vers une meilleure équité: les subventions énergétiques, le ciblage, et la protection sociale en Tunisie(Towards greater equity: energy subsidies, targeting and social protection in Tunisia). 40 Its main objective was to determine who benefits from energy subsidies which, unlike food subsidies, do not exclusively target households. Most energy subsidies—especially for fuel—do not go directly to households. Energy subsidies generally include subsidies for electricity (which mainly concern the Tunisian energy company, STEG) and subsidies for fuel(mainly the Tunisian petroleum refining company, STIR). The World Bank study analyzed only the direct effect of these various energy subsidies(see Figure 2). To convey a sense of the complexity of energy subsidies, we describe them in detail in Appendix C. 39 CRES/INS 2013, previously mentioned reference, p. 26. 40 Banque Mondiale(2013). Vers Une Meilleure Equité: Les Subventions Energétiques, Le Ciblage, et La Protection Sociale En Tunisie.(Towards greater equity: energy subsidies, targeting and social protection in Tunisia). 34 Uncovered: The role of the IMF in shrinking the social protection

Figure 2. Distribution by wealth quintile of the direct benefits of energy and food subsidies in Tunisia(%) Energy subsidies 70 Distribution of subsidy benefits (%) 60 Q1 Lower income Q2 50 Q3 Q4 40 Q5 Higher income 30 20 10 0 Electricity and gas Petrol LPG gas Diesel Petroleum Distribution of subsidy benefits (%) 70 60 Q1 Lower income Q2 50 Q3 Q4 40 Q5 Higher income 30 Food subsidies 20 10 0 Cereals Sugar Vegetable oil All Source: World Bank(2013). World Bank staff calculations. Values have been adjusted based on gas prices as of April 2013.“Q” stands for quintile. How international financial institutions affect social protection in Tunisia 35

We can differentiate between two types of subsidized products: those with very unequal profiles (such as petrol and diesel) and those with unequal profiles that are similar to food subsidies(such as electricity and liquefied petroleum gas, LPG). If we cross-reference these profiles with their share of household consumption(see Appendix B), we note that diesel accounts for only 2% and that petrol is almost exclusively used in the transport sector. More conventional products account for 18%(LPG) and 15%(electricity and gas) of household consumption. This shows why the progressivity or equity of food and energy subsidies should not be analyzed in absolute terms, as the World Bank has done, but rather in relation to real household consumption. The CRES study on food subsidies distinguishes between distribution in absolute terms and in relation to consumption(transfer rate). Figure 3 shows that the poorest quintile’s share of subsidies is the highest when compared to consumption. Figure 3. Share of subsidies in Tunisia by wealth quintile compared to consumption(%) Share of the total subsidy received Rate of indirect transfers(consumption) 20,1% 20,9% 21,5% 21,3% 16,3% 8,4% 5,9% 4,4% 3,1% 1,5% Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Source: CRES/INS(2013). We can only judge whether a subsidy is progressive or not by measuring it against actual consumption (transfer rates). But the World Bank study simply concludes that“the distribution of energy subsidies is more unequal than that of food subsidies.” However, only after we correlate subsidies to actual consumption can we judge whether energy subsidies really are more unequal than food subsidies. Even if the World Bank’s statement were true, it says nothing about how progressive energy subsidies are in absolute terms, rather than just in comparison to food subsidies. Energy subsidies may well be progressive in absolute terms yet more unequal than food subsidies, which are also progressive. 36 Uncovered: The role of the IMF in shrinking the social protection

The World Bank authors did not use the Lorenz curve to assess the progressivity of energy subsidies as the CRES/INS study does. Furthermore, they do not mention the indirect redistributive effect of energy subsidies, especially for products like petrol and diesel that affect transport and agriculture. Only in the chapter on the impact of economic reforms do the authors conclude that“the magnitude of the indirect effects almost equals that of the direct effects. According to the World Bank staff’s calculations, the impact on consumption of a 10% increase in prices over a year is huge for the poorest quintile. It is equivalent to a reduction of more than 50% in diesel subsidies and 60% in electricity subsidies. 41 Comparing the World Bank study with the CRES/INS study on energy subsidies for households (both from 2013), we note that the World Bank did not address the complexity of the direct and indirect effect of subsidies on inequalities. Thus, we can draw no conclusions regarding the fairness of energy subsidies. Instead of the radical approach promoted by the World Bank and the IMF, it would be better to place reforming energy subsidies within the framework of a fair and gradual energy transition to reduce the consumption of polluting fuels, for example by replacing public transport fleets with more environmentally friendly electric buses. This would kick off the transport sector’s transition to cleaner energy by tackling the crux of the problem and reduce consumption(fewer polluting fuels) whereas fiddling with the price punishes the weaker classes without offering them any substitute. More evidence of the positive impact of food and energy subsidies is explained in the next section. The PNAFN and subsidy reform The World Bank and CRES/INS studies of 2013 also attempted to evaluate Tunisia’s National Programme of Support to Needy Families(PNAFN). We examined how, in 2017, Bibi and Ben Cheikh evaluated the CRES/INS study on the PNAFN and its connection to subsidy reform in a comprehensive report of more than 300 pages. 42 In partnership with the AfDB, Bibi and Ben Cheikh reviewed the PNAFN and the free medical assistance program(AMG II). We focus on what they wrote about the PNAFN, Tunisia’s only cash transfer program. 43 41 World Bank, 2013, previously mentioned reference. 42 Bibi, S. and Ben Cheikh, N.,«Evaluation de la performance des programmes d’assistance sociale en Tunisie», 2017, Tunis. 43 The CRES authors explain that the PNAFN was created in 1986 within the IMF’s structural adjustment program, which aimed to rationalize Tunisia’s General Compensation Fund(food subsidies).“Fearing that the gradual dismantling of food subsidies would undermine the purchasing power of poor and vulnerable populations, the Tunisian authorities identified 78,000 families identified for receiving compensatory transfers on the basis of the 1985 INS ECBNVM[Enquête Consommation Budget Niveau de Vie des Ménages].” How international financial institutions affect social protection in Tunisia 37

Like today, the IFIs planned to rationalize food subsidies and replace them with a more targeted direct transfer program. Between 1986 and 2010, the number of families covered by the PNAFN increased from 78,000 to 118,000. After the revolution, the number rapidly swelled—from 176,000 in 2011 to 202,000 in 2012 and 235,000 in 2014, the last year covered by the CRES/INS report. The CRES/INS report evaluated many aspects of the PNAFN that are difficult to summarize. Below, we concentrate on the PNAFN’s targeting process and how it compares with that of food and energy subsidies. We find these two aspects important because they offer very different approaches that have been carried out simultaneously in Tunisia: the universal approach of subsidizing prices for food and energy, and the targeted approach of supporting the poorest populations. Table 2. Accuracy and errors in the current targeting in Tunisia The measurement unit and the household Ideal solution 0.1%- 8.3% Implementing with perfect information PNAFN AMGII Non-targeted 8.3% 0 0 Total 8.3% Current implementation 0.1%- 8.3% Current results PNAFN 3.9 AMGII 4.1 Non-targeted 0.5 Total 8.3 Target groups 8.31%- 30% 30.1%- 100% 0 0 21.7 0 0 70 21.7 70 Target groups 8.31%- 30% 30.1%- 100% 3.1 1.4 10.9 6.7 7.7 61.9 21.7 70 Total 8.3% 21.7 70 100 Total 8.3 21.7 70 100 Note: Households are ranked in ascending order according to the level of consumption per person that would have been observed in the absence of the PNAFN and AMG II. Source: CRES(2017). 38 Uncovered: The role of the IMF in shrinking the social protection

From Table 2 above, we note the following: • Targeted cash transfers exclude more than half of the poor. Only 3.9% of the poorest 8.3%, who should be targeted by the PNAFN, are actually covered. The authors point out that that its targeting accuracy is 46.9% and that 53.1% of the needy families that should be covered are not. • Food and energy subsidies do a lot to reduce poverty in Tunisia. In Table 3 below(especially the figures we have highlighted in green), the CRES(2017) authors calculate the effect of the PNAFN, the AMG II, food subsidies and energy subsidies on poverty(its incidence, 44 the income deficit or poverty gap, 45 and its severity 46 ). These figures show that food subsidies contribute most to declining incidences of poverty, followed by energy subsidies, the PNAFN and the AMG II. The same holds for the poverty deficit; only the PNAFN helps reduce the severity of poverty more than food and energy subsidies. Table 3. Effects of social transfers in Tunisia on poverty Standard of living ρi ρi% P 0 (Z) π 0 (T i ; Z) π 0 (T i ; Z)% E 0 (T i ; Z) P 1 (Z) π 1 (T i ; Z) π 1 (T i ; Z)% Before transfers 211,865 35,8 16,2 Individual is the unit of measurement PNAFN AMGII Food subsidies -- 3,000 1,339 8,789 14,38 6,40 42,01 Effect on the incidence of poverty(P 0 (Z)) -- 1,56 0,8 2,82 21,73 11,1 39,21 1,51 1,73 0,93 Effect on the poverty gap(P 1 (Z)) -- 1,93 0,8 2,82 21,73 11,1 39,21 Energy subsidies 7,785 37,21 2,01 27,96 0,75 2,01 27,96 After transfers 232,787 28,6 28,6 44 Proportion of people below the poverty line. 45 Average relative gap between the poverty line and poor households’ average spending. 46 Average square of the gap between the poverty line and poor households’ average spending. How international financial institutions affect social protection in Tunisia 39

E 1 (T i ; Z) - 1,51 1,73 0,93 0,75 Effect on the severity of poverty(P 2 (Z)) P 2 (Z) 9,9---- 4,8 π 2 (T i ; Z) - 1,85 0,66 1,50 1,09 π 2 (T i ; Z)% - 36,20 12,89 29,46 21,45 E 2 (T i ; Z) - 2,52 2,01 0,70 0,58 Note: π α (T i ;z) is the absolute contribution of the T i transfer to the reduction of P α (z) as defined by equation (4-12). π α (T i ;z)% is the absolute contribution. pi is the average per capita household gain from the T i program. pi% is the share of Ti resources in all resources deployed to reduce poverty. E α (T i ; z) is the benefit ( π α (T i ; z)%)- cost(pi%) ratio described by the equation(4-13). Source: CRES(2017). We summarize the findings of interest to us in Table 4 below. Because the SDG indicators, especially those of SDG 1 on poverty eradication(Indicators 1.1.1 and 1.2.1), only address the incidence of poverty, it seems reasonable to us to put more importance on how subsidies impact the incidence of poverty than on how they impact its severity. The food and energy subsidy schemes are much more in line with the 2030 Agenda for Sustainable Development and its SDGs than the PNAFN targeting system, because they contribute more to achieving SDG 1(ending poverty everywhere) and respect the central principle of“leaving no one behind” with universal coverage by design(100% targeting). Table 4. Comparing transfer systems in Tunisia Accuracy of targeting (%) Effect on the incidence of poverty (%) Effect on the poverty gap (%) Effect on the severity of poverty (%) Source: CRES(2017). PNAFN 46.9 21.73 21.73 36.20 Food subsidies 100 39.21 39.21 29.46 Energy subsidies 100 27.96 27.96 21.45 40 Uncovered: The role of the IMF in shrinking the social protection

Conclusions on the redistributive effect of food and energy subsidies Who benefits from subsidies? Are food and energy subsidies equitable? Our analysis leads us to conclude that food subsidies are progressive and equitable and contribute the most to eradicating poverty. Energy subsidies do more for poverty eradication than PNAFN targeted transfers. We therefore find that the universal approach of energy and food subsidies is more likely to achieve SDG 1(eliminating extreme poverty and hunger), and to leave no one behind, whereas the PNAFN excludes more than half of needy families. However, PNAFN targeting might be worth maintaining with regard to extreme poverty and the severity of poverty. Replacing food and energy subsidies is not in line with the 2030 Agenda and the challenges of this century. We should organize social protection instruments more intelligently, rather than be compelled to choose between them. What is more, the results we have summarized and analyzed on the redistributive effect of social transfers are just some of the tools that states can use to reduce inequalities. As we have shown in another study, fiscal redistribution(through a progressive tax) mainly acts on inequalities“from above” by reducing the incomes of the richest, while social redistribution(through social transfers) mainly acts“from below” by transferring benefits to the poorest. The combination of generous social redistribution and progressive income taxes has meant that Tunisia, like Algeria, has a structure of inequality“close to that of Europe and China where the middle class holds the largest share of national income”. 47 The IMF and public spending on education and healthcare Good education for all citizens is a good example of a universal subsidy. SDG 4 calls for ensuring “that children universally—including both boys and girls—will be able to complete a full course of primary education by 2015.” The same holds for health, with SDG 3 requiring“ensur[ing] healthy lives and promot[ing] well-being for all at all ages.” Both are crucial to sustainable development. IMF programs usually impact both sectors indirectly, with fiscal consolidation, local currency devaluation and debt servicing payments causing public spending on healthcare and education to decline. The IMF’s pre-pandemic programs in Tunisia did not mention spending on either education or healthcare. 47 Ben Rouine, C.(2020). Inégalités et Fiscalité en Afrique du Nord. Observatoire Tunisien de l’Economie/Tax and Inequality in North Africa, Tunisian Observatory of Economy 14/01/2020. How international financial institutions affect social protection in Tunisia 41

Figure 4. No noticeable evolution of health expenditure(before COVID) 3.20% 3.00% 3.07% % of GDP 2.80% 2.60% 2.51% 2.59% 2.70% 2.64% 2.66% 2.71% 2.68% 2.55% 2.40% 2.49% 2.20% 2.00% 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Tunisian Ministry of Finance. Figure 5. Inconsistency of expenditure on education 6.00% 5.92% 5.80% 5.60% 5.40% 5.20% 5.40% 5.24% 5.24% % of GDP 5.00% 4.79% 4.80% 4.60% 4.65% 4.56% 4.76% 4.59% 4.40% 2011 2012 2013 Source: Tunisian Ministry of Finance. 2014 2015 2016 4.67% 2017 2018 2019 2020 42 Uncovered: The role of the IMF in shrinking the social protection

Figures 4 and 5 show how four main phases of budget spending on healthcare and education evolved as a percentage of GDP between 2011 and 2020. Healthcare spending increased between 2011 and 2013, following the revolution. Between 2013 and 2016, spending more or less stagnated. From 2016 until 2019, under IMF-imposed austerity, healthcare spending declined as a percentage of GDP. Finally, in 2020, after the outbreak of COVID-19, healthcare spending increased both in gross terms due to the centrality of the crisis, and also as a percentage of GDP, due to the sharp decline in GDP. Spending on education did not follow the same trajectory as healthcare spending. From 2011 to 2013, it fell as a percentage of GDP, and began increasing somewhat steadily from 2014 to 2017. In 2018 there was a significant drop in spending on salaries in primary and secondary education. Expenditures increased as percentage of GDP in 2019(mainly because of slow growth), yet they remained below the 2017 figure. Spending rose again in 2020 during the pandemic with a base effect due to the sharp decline in GDP in 2020. IV. A reading of IMF documents: The subsidy reform central to IMF structural adjustment in Tunisia Subsidy reform has been key to structural reforms since the first IMF structural adjustment program. In this section we focus on subsidy reform in IMF arrangements and its counterpart, direct transfers to vulnerable populations based on the PNAFN. It should be noted here that almost all IMF review reports have been published in English, with just a few in French. Not a single report has been published in Arabic, Tunisia’s official language. To more broadly disseminate this information, which has been at the heart of social struggles since the revolution, we reproduce IMF reports verbatim. Our analysis is divided into three phases that correspond to the principal stages of the IMF structural adjustment. This allows us to analyze how the IFIs’ positions and relationships with the Tunisian authorities have evolved in parallel with the national and international situation. It highlights the periods when the IMF and the Tunisian authorities were more or less aligned, as well as when they were not(notably in 2019). The first phase corresponds to the first Stand-By Arrangement, which ran from June 2013 to April 2016. The second phase corresponds to the second Extended Fund Facility arrangement, from June 2016 to February 2020. The third corresponds to the COVID-19 emergency, from March 2020 to February 2021. How international financial institutions affect social protection in Tunisia 43

Phase 1: June 2013 to April 2016 The Stand-By Arrangement was signed in 2013, around the time that two assessments of food and energy subsidies were published. The first arrangement of mid-2013 indicated that subsidies would be central to the structural reforms. The IMF regularly conducts reviews to assess how reforms are progressing with respect to the timetable set out in the agreement and the disbursement of loan installments. The reviews are essential for understanding the framework and evolution of subsidy reforms in Tunisia. When an arrangement is signed, the first review report is published(we call it “R0”), in which the IMF presents its own analysis of the situation and shows how aligned the Tunisian authorities’ views are with the IMF. IMF conditions constitute the core of the structural adjustment program, and we explain the various types in Appendix C. Here we focus on the structural reforms linked to subsidy reforms. In the R0 that accompanied the signing of the Stand-By Arrangement in June 2013, the IMF specified that energy subsidies were central to the structural reforms. The IMF shares the World Bank’s view that energy subsidies must be reduced by regularly increasing prices and introducing an automatic fuel price formula. It also involves improving and extending the PNAFN to poor and vulnerable populations. While the World Bank’s assessment of energy subsidies had not been made public at the time of the IMF’s R0, the IMF used it to justify the reform:“The existing PNAFN system provides cash transfers and free health cards to 9 percent of the population(an estimated 235,000 households), but the latest World Bank estimates(using the 2005 household survey) indicate leakage to non-poor of around 60 percent.” 48 The IMF relies on the World Bank’s dramatic figures that estimate 60% of PNAFN transfers do not go to the poor. In addition, the redistributive effect of energy subsidies was analyzed with technical assistance from the World Bank and the IMF as part of a conditionality – a“structural benchmark” (SB) – of the structural adjustment program. 49 We note that the IMF’s approach to food subsidies is different from their approach to energy subsidies. 50 When the R0 report was published in June 2013, the CRES/INS evaluation of food 48 IMF, 2013, p. 22. 49“The authorities’ focus on improving the targeting of the social safety net is critical as subsidy reforms for energy products are under preparation(Box 5). Both the authorities and staff recognize that any future ad-hoc increases in energy prices need to be accompanied by compensating measures for the poor and vulnerable, even though the new registry or its accompanying targeting system may not yet be in place. The authorities—with World Bank and Fund TA—are currently analyzing the welfare and social impact of reforms for each energy product, and are working on designing a strategy to support households during subsidy reform(structural benchmark for August 2013, see Box 5).” IMF Country Report No. 13/161, June 2013, p. 22. 50“As for food subsidy reform, given the importance of food subsidies for poorer households and despite existing leakages (e.g., to hotels or industry), the government will assess at a later stage how to reform these, and may be able to use lessons from the energy subsidy reform to design compensatory measures.” Ibid., point 26, p. 25. 44 Uncovered: The role of the IMF in shrinking the social protection

subsidies was not yet public. But the study proved that food subsidies are progressive—discrediting the IMF’s obsession with them. 51 In the R0, the structural reforms linked to the energy subsidy reform were formulated as structural benchmarks: • adopting a Ministry of Industry decree approving a new automatic fuel pricing formula by August 2013; and • submitting a new targeted household support program, to accompany the reform of generalized energy subsidies, to the Cabinet. In its first and second reviews(R1 and R2), which were published together in February 2014, the IMF welcomed energy price increases for the cement industry. 52 As for the automatic fuel pricing formula, on p. 14 the IMF noted:“[A] new automatic fuel price formula has been designed, allowing convergence to international prices over time, but without a smoothing mechanism for increases higher than USD 6 per barrel, and with lags for smaller price increases. Staff urged the authorities to continue working on improving the formula to ensure that the price formula is not abandoned and remains sustainable in the event of large increases in international prices.” The IMF considered this structural reform to have been implemented but was concerned that it could have an upward effect on international prices. It even urged Tunisian authorities to not abandon the automatic formula. On the other hand, IMF staff was dissatisfied with the delayed implementation of the structural benchmark on targeting beneficiaries and“urged that fuel price increases be accompanied by the implementation of a household compensation mechanism. In parallel, staff has urged the authorities to step up their communication campaign ahead of any fuel price increase.” 53 In its third review(R3) of July 2014, IMF staff continued to express dissatisfaction with the delay in implementing the targeted program strategy, notably introducing the unique social identifier(IMF, 2018). 51 The fact that, unlike energy subsidies, the sentence on food subsidies does not include“the authorities and IMF staff” but only“the government” clearly demonstrates that the IMF resisted including food subsidies, which staff do not agree with, although both sides have agreed to show that energy subsidy reform is effective before attacking food subsidies. 52 This involved a 35% increase in electricity tariffs in January 2014 and the elimination of electricity subsidies for cement companies in June 2014. Gas prices were to be raised by 47% in January 2014 and eliminated, with another 47% increase in June 2014. 53 P. 24. How international financial institutions affect social protection in Tunisia 45

Only during the IMF’s fourth review(R4) of September 2014 did the IMF finally approve the program’s implementation. 54 The IMF also noted that the authorities were in the process of finalizing the unique social identifier, which was already available to 7.5 million citizens. The fourth and penultimate review concluded with general satisfaction regarding the energy subsidy reform, with the two structural benchmarks on fuel price adjustments considered as met. Finally, according to the fifth review(R5) published in late December 2014,“The evaluation of existing programs(vulnerable households, school allowances), recently completed, shows less leakage toward the non-poor than previously estimated. Based on this information, and the introduction of a unique social identification number for 8.5 million Tunisians, a number of measures are being developed to enable the establishment of a new social safety net that should accompany the reduction of energy subsidies”(translated from French). From the in-depth evaluation that CRES carried out in 2014 but only published in 2017, we eventually learned that the World Bank’s estimate that 60% did not go to the poor was exaggerated. The fact that the IMF postponed its food subsidy reform based on the first CRES/INS assessment of 2013 and attached greater importance to the assessment of leakages by CRES than by the World Bank should have prompted the Tunisian authorities to re-evaluate the redistributive effect of energy subsidies before proceeding with the reform. The sixth and final review of the structural adjustment program under the Stand-By Arrangement (R6), published in October 2015, expressed the IMF staff’s general satisfaction with the subsidy reforms. However, on the automatic fuel price formula, the two sides part ways.“Staff urged the quick revision of the existing asymmetric automatic fuel price adjustment formula—which is currently triggered only if there is a cumulative USD 6 a barrel increase within a quarter—into an automatic symmetric formula that will smooth large fluctuations in international prices and ensure full cost recovery and appropriate tax collection. The authorities agreed with this approach, but argued that such a revision would take time and can only be implemented in 2016.” 55 In R6, the IMF regretted that the electricity tariff increase was suspended(see Appendix D). 56 In this new program, the IMF demanded two conditions that negatively impacted the well-being of the majority of the population: 54 This included a 10% increase in cash transfers to vulnerable populations; an increase in the number of PNAFN beneficiaries from 220,000 in 2013 to 235,000 in July 2014, with the aim of reaching 250,000 by the end of the year; and the “temporary” transfer of 80 Tunisian dinars(TND) to about 285,000 PNAFN families to coincide with a rate increase(which corresponded to the Ramadan transfer). 55 P. 23. 56“Staff regrets the suspension of programmed increases in electricity tariffs, which are still heavily subsidized—20 percent below cost recovery even after last year’s efforts to adjust tariffs for medium and low voltage industries and eliminate subsidies for cement companies. Staff urged further increases in electricity tariffs towards cost-recovery levels to continue reducing untargeted and regressive energy subsidies. The authorities argued that gains in overall energy subsidies from lower international oil prices do not justify increasing tariffs at this juncture,”(R6, p. 23). 46 Uncovered: The role of the IMF in shrinking the social protection

• the devaluation of the Tunisian dinar(TND); and • the ending of energy subsidies in adverse conditions using an automatic formula to review petrol prices each quarter when international oil prices were rising. The IMF first made elimination of energy subsidies a structural benchmark, then, faced with the recalcitrant Tunisian government, made it a prior action(PA) – a condition for validating payment of a new debt tranche. The Tunisian authorities were forced to raise prices in December 2017 and again in 2019. The IMF noted the lack of adequate social protection to shield the Tunisian population from the negative impacts of its measures. The World Bank recommended not lifting fuel price controls before cash transfer systems were in place to compensate for the resulting deterioration in Tunisians’ real incomes. At the time, by installing a database on vulnerable households before March 2017 as a structural benchmark, the IMF’s insistence on abolishing energy subsidies became the focus of political and social confrontations. Phase 2: June 2016 to February 2020 The second phase of structural adjustment began with the signing of a new Extended Fund Facility arrangement whose initial report(R0) was published in June 2016. With fairly low international oil prices, the program under the Extended Fund Facility arrangement began by gradually reforming energy subsidies. The only obligatory structural reform was the database on vulnerable households to be created by March 2017. With regard to the automatic fuel price formula,“The authorities argued that the symmetric mechanism and associated smoothing formula should initially be implemented on a quarterly basis, 57 but agreed with staff that monthly adjustments starting in January 2017 are necessary to ensure that it[the mechanism] can be sustained in case of large fluctuations in international oil prices.” 58 The government allowed an automatic and symmetric formula to be introduced but maintained a quarterly fuel price adjustment until at least January 2017, although the IMF preferred it to be monthly. The first Extended Fund Facility review(R1), published in July 2017, did not report much progress. The automatic adjustment mechanism was to be applied to the three main fuels starting that month. To enforce implementation of the reform, the IMF added a new structural benchmark with the aim of reapplying the automatic fuel price adjustments by September 2017:“We will reestablish the quarterly adjustment mechanism for gasoline and diesel prices(regular and 50), which will be adjusted monthly starting in January 2018. In addition, to reduce subsidies, the 2017 budget includes an increase in the prices of electricity(5 percent) and gas(7 percent), while keeping the social tariffs unchanged.” 59 57 MEFP[Memorandum of Economic and Financial Policies], p. 22. 58 IMF Tunisia Country Report No. 16/138, June 2016, p. 13. 59 IMF Middle East and Central Asia Department, Tunisia: First Review Under the Extended Fund Facility, Request for Waivers of Non-Observance of Performance Criteria and Rephasing of Access, 2017, issue no. 203, IMF, Washington. p. 60 https://www.elibrary.imf.org/view/journals/002/2017/203/article-A001-en.xml How international financial institutions affect social protection in Tunisia 47

This new structural benchmark must be analyzed against the evolution of the price of oil. As Figure 6 shows, oil prices began to rise in early July 2017. We will see how the increase in international oil prices played a central role in the structural adjustment program. Figure 6. The pressure of the increase in oil international prices in July 2017 Europe Brent spot price FOB(USD per barrel) 90 85 80 75 70 65 60 55 50 45 40 Jun 01, 2016 Aug 01, 2016 Oct 01, 2016 Dec 01, 2016 Feb 01, 2017 Apr 01,2017 Jun 01, 2017 Aug 01, 2017 Oct 01, 2017 Dec 01, 2017 Feb 01, 2018 Apr 01, 2018 Jun 01, 2018 Aug 01, 2018 Oct 01, 2018 Dec 01, 2018 Feb 01, 2019 Apr 01, 2019 Jun 01, 2019 Aug 01, 2019 Oct 01, 2019 Dec 01, 2019 Feb 01, 2020 Source: US Energy Information Administration. July 2017 was a turning point in energy subsidy reform in more ways than one. The rise in oil prices ended what the IMF characterized as their“depoliticization.” As part of the Extended Fund Facility agreement, the IMF also liberalized the Tunisian dinar, causing it to decline precipitously between April 2016 and February 2019. International oil prices were rising during the same period, thereby multiplying the effect on fuel prices and energy subsidies. The second half of 2017 saw the World Bank return to energy subsidy reform as well. 48 Uncovered: The role of the IMF in shrinking the social protection

Although the IMF publicly exposed the World Bank’s exaggerated estimates regarding PNAFN leakages at the end of the Stand-By Arrangement, in the second Extended Fund Facility review, completed in March 2018, the IMF devoted an entire paragraph to the World Bank’s assessment, 60 without providing any figures to show how food subsidies only benefit the rich. With the rise in international prices and a likely renewed“politicization” of fuel prices, the IMF/World Bank pairing closed ranks and divided their roles along two main issues. The first was the removal of energy subsidies in a hostile environment(rising international oil prices), driven mainly by the IMF. The second issue was establishing a targeting policy by: • The IMF imposing a condition to force completion of the database on vulnerable families; and • The World Bank overseeing the establishment of a legal framework merging all state social assistance programs(PNAFN, AMG I and AMG II)(see Appendix E). THE IMF’s tougher stance on energy subsidy reform The IMF hardened its tone in R2, and for the first time since June 2013, deployed its most powerful weapon to reform energy subsidies—prior actions. The IMF insisted on a“significant ad hoc increase in fuel prices” before it would make its review. 61 This meant that the authorities didn’t just owe the prices in December 2017, but that they also had to pledge to increase them in each quarter of 2018. 62 The IMF then further tightened its grip on Tunisia by increasing the number of annual reviews from two to four. 63 As former Finance Minister Hakim Ben Hammouda pointed out:“The change announced by the IMF and the addition of two more reviews signals the Fund’s dissatisfaction with the progress of our reforms and commitments.” 64 60“The social safety net remains poorly targeted, largely regressive, and fragmented. The World Bank has estimated that nearly a quarter of Tunisians are net beneficiaries of rather generous social transfers(that represent up to one-fifth of total income). However, only 2 in 5 of these beneficiaries live below the national poverty line. Sizeable subsidies on food and energy(together amounting to 3.1 percent of GDP in 2017), which mainly accrue to the better-off, largely explain this poor targeting. For example, energy subsidies benefit rich households up to 30 times more than those with lower income; and cash and health transfer programs for low-income families cover only about 12 percent of Tunisia’s poor.” IMF Tunisia, Second EFF Review, issue no. 120, June 2018, p. 21. https://www.elibrary.imf.org/view/journals/002/2018/120/article-A004-en.xml?rskey=ekEd3A&result=1 61 Ibid. 62 Réalités, La Tunisie et le FMI, les Non-Dits du Dernier Accord. https://www.realites.com.tn/2018/03/la-tunisie-et-le-fmi-lesnon-dits-du-dernier-accord/ 63 Ibid. 64 Ibid. How international financial institutions affect social protection in Tunisia 49

A few months later, in the R3(third review) published in July 2018, the IMF multiplied its orders, again brandishing prior actions. 65 For the IMF, it was no longer a matter of discussing and exchanging views with the Tunisian authorities, but of imposing orders so that the program and disbursements could continue. It should be noted that the IMF’s manner and approach became even harsher as international oil prices rose and the Tunisian dinar plummeted. The R4 published in October 2018 coincided with the peak in the price of Brent crude oil. The IMF continued to use prior actions to impose orders on the Tunisian authorities:“Energy price hikes. Signature of ministerial orders(“arrêts”) stipulating the increase in(i) fuel prices in September and(ii) electricity/gas tariffs…; and implementation of the September 1 price hikes for fuels, electricity, and gas.” 66 Below we reproduce the relevant“Text Table 1” of the IMF prior actions and structural benchmarks(Table 5). Table 5. IMF-required energy price adjustments in Tunisia, 2018 TOTAL Fuel products(STIR) Essence premium Gasoil premium Super lead free Unleaded Regular Heavy Fuel Inductriel fuel Tariffs(STEG) Electricity High tension Middle tension Low tension Adjust. (millimes/%) September 1 2018 Impact (TD mio.) 252 60 mil.... 60 mil.... 60 mil. 17 60 mil. 11 60 mil. 51 60 mil. 5 60 mil. 1 167 132 23% 6 23% 106 12% 25 65 IMF Middle East and Central Asia Department, Tunisia: Third Review Under the Extended Fund Facility Arrangement and Request for Modification of Performance Criteria, Volume 2018, Issue 218, IMF. https://www.elibrary.imf.org/view/journals/002/2018/218/002.2018.issue-218-en.xml 66 IMF Middle East and Central Asia Department, Tunisia: Fourth Review Under the Extended Fund Facility Arrangement and Request for Modification of Performance Criteria(R4), Volume 2018, Issue 291, IMF. p. 69. https://www.elibrary.imf.org/ view/journals/002/2018/291/002.2018.issue-291-en.xml 50 Uncovered: The role of the IMF in shrinking the social protection

Gas High pressure Middle pressure Low pressure 35 23% 11 23% 19 12% 5 Source: IMF EFF Review 4. Table 5 shows that the IMF ordered very high energy price increases, yet stated in its R4 review,“Fuel prices remained about 25 percent below market prices in early September.” 67 , 68 In the midst of an avalanche of orders from the IMF, the Tunisian authorities refused to make a monthly adjustment as the international price of Brent crude oil peaked. They maintained their preference for a quarterly price adjustment by invoking“popular expectations” for the first time. The authorities had good reason to do so. The year 2018 was marked by record-breaking inflation; rising interest rates, fuel prices and taxes; increases in the trade deficit and public debt; and falling international reserves. As a result, the year had seen heated protests. The reference to popular expectations reflected the mounting social tensions due to the structural adjustment program that the Extended Fund Facility had accelerated. The devaluation of the Tunisian dinar created a surge in popular sentiment that destabilized society. The authorities’ refusal to impose a monthly price adjustment affected the pace of the reviews. While the IMF had conducted and published three reviews between March and October 2018, the fifth review was not published until July 2019, a sign of deep disagreements between the two parties. Meanwhile, as Figure 7 shows, the international price of Brent crude oil passed the October 2018 peak and continued to fall until the end of December 2018. From January 2019, Brent crude prices started rising again, until April 2019. It was at this point that the IMF chose to make their field visit to complete the fifth review. Taking advantage of a rise in the price of Brent crude oil, IMF teams came to reimpose prior actions with fuel price increases. However, things did not go as planned. During the IMF’s visit from March 27 to April 9, 2019, the authorities were forced to increase fuel prices surreptitiously during the Arab 67 The IMF analyzed energy prices adjustments as follows:“The authorities committed to remedy the slippages on energy price adjustments accumulated over the summer, supported by prior action(MEFP¶14 and Table 2). First, to replace the originally agreed monthly fuel price increases, they implemented an average 5.5 percent hike in fuel prices on September 1, to be followed by a similar increase in November. Staff would have preferred a monthly schedule of pre-set increases to depoliticize the adjustments, but the authorities argued that the quarterly schedule with ad hoc-adjustments would be better aligned with popular expectations. Second, more than two-thirds of the electricity and gas tariff adjustments committed under the Third Review were implemented in early September; and an administrative order stipulates further hikes in October and November.” Ibid., p. 9. 68 Ibid., p. 4. How international financial institutions affect social protection in Tunisia 51

League summit, despite having assured the media that reports of a price increase were just rumors. 69 That provoked a social uproar of rare intensity, and this occurred during the IMF visit. From April 1, roads were blocked across the country to protest the fuel price increases 70 and civil servants staged a general strike. 71 In a historic development, the Tunisian General Labour Union(UGTT), the Tunisian Confederation of Industry, Trade and Handicrafts(UTICA) and the Tunisian Union of Agriculture and Fisheries(UTAP) issued a joint press release on April 6 rejecting the latest fuel price increases. 72 The widespread refusal sent a strong message to the IMF about the“popular expectations” of energy subsidy reform. The IMF was unable to conclude the fifth review after its April 2019 visit, and needed another, from July 11 to 17 2019, 73 to complete it. This did not prevent the IMF continuing to impose orders via prior actions in its fifth review(R5), published in July 2019. Nevertheless, for the first time, the IMF was forced to concede:“At the same time, there has been strong opposition among Tunisians to measures that may affect their purchasing power such as energy subsidy reduction. Finding a set of policies and reforms that strike a workable balance between the economically necessary and socially acceptable is therefore crucial.” 74 This was the last concession, as the IMF did not conduct any other visits or reviews under the Extended Fund Facility. The IMF and Tunisia resumed relations only with respect to COVID-19, but in a different framework, which we consider below. Phase 3: From the COVID-19 outbreak until today First, we must explain what had transpired in Tunisia politically before the outbreak of COVID-19. • After the fifth IMF review in July 2019, during the second half of the year, all attention was focused on the legislative and presidential elections. • After the elections, it took several months for the political actors to agree on a government. The Assembly of the Representatives of the People(ARP) announced its confidence in Elyes Fakhfakh’s government on February 27, 2020, just a few days before the first cases of COVID-19 in Tunisia. 69 Direct info, Tunisie-Augmentation du prix du carburant: Blocage de l’autoroute de Sfax. https://directinfo. webmanagercenter.com/2019/04/01/tunisie-augmentation-du-prix-du-carburant-blocage-de-lautoroute-de-sfax/ 70 African Manager, Les louages bloquent tout à cause des prix du carburant, avril 2019. https://africanmanager.com/les-louages-bloquent-tout-a-cause-du-prix-des-carburants/ 71 UGTT, Communiqué de presse commun, avril 2019, UGTT. http://www.ugtt.org.tn/2019/04/5323-2/ 72 Ibid. 73 IMF, Tunisia: IMF Staff Concludes Visit, press release, July 2019, IMF Washington. https://www.imf.org/fr/News/Articles/2019/07/17/pr19286-tunisia-imf-staff-concludes-visit 74 R5, pp. 8–9. 52 Uncovered: The role of the IMF in shrinking the social protection

• Crucially, the crisis in Tunisia facilitated the temporary granting of full powers to the government under Article 70 of the Tunisian constitution. 75 While the Tunisian authorities’ priority was saving lives, the IMF and the World Bank seized the opportunity to mount a fresh attack. First, the IMF tackled energy subsidies, insisting that its support during COVID-19 would not be linked to any conditions. In the first few months at least, COVID-19 resulted in a sharp fall in Brent crude oil prices, as shown below(Figure 7). Figure 7. The COVID-19 pandemic lowers oil prices Europe Brent spot price FOB(USD per barrel) 70 60 50 40 30 20 10 0 Mar 02, 2020 Apr 02, 2020 May 02, 2020 Jun 02, 2020 Jul 02, 2020 Aug 02, 2020 Sep 02, 2020 Oct 02, 2020 Nov 02, 2020 Dec 02, 2020 Jan 02, 2021 Feb 02, 2021 Source: US Energy Information Administration. The price of Brent crude oil fell from USD 52 in early March 2020 to around USD 9 in late April 2020. Thereafter, the price almost regained its pre-COVID-19 level, then began to rise further in November 2020. The IMF granted Tunisia a loan under a Rapid Financing Instrument program created specifically to respond to COVID-19. The signing of the new agreement masked the abrupt 75 Parliament adopted Law No. 2020-19 of 12 April 2020 empowering the head of government to issue decree-laws to combat COVID-19. The government was granted legislative power for a period not to exceed two months in four main fields. With respect to fiscal, financial and social fields, Article 1 includes“measures aimed at adjusting the procedures and terms of health and social coverage of social security beneficiaries affected by the coronavirus.” https://legislation-securite. tn/law/104786 How international financial institutions affect social protection in Tunisia 53

termination of the Extended Fund Facility, which was supposed to end in May 2020. This was the first time in the history of Tunisia’s relations with the IMF that a program was not completed. A single Rapid Financing Instrument report was published in April 2020; the government’s letter of intent was sent on April 2, 2020. We believe that the IMF pressured the Tunisian authorities behind the scenes by imposing prior actions for the Rapid Financing Instrument agreement. Indeed, Tunisia’s letter of intent was sent on April 2, 2020, and the IMF Executive Board approved the agreement on April 10, but the Tunisian authorities issued a decree that shows it must have been a secret. That is to say, on March 31, 2020, the minister of energy, mines and the energy transition and the minister of finance set out the composition and functioning of the Technical Commission charged with fixing and monitoring the selling prices of imported finished petroleum products and locally refined petroleum. Although the Extended Fund Facility agreement was terminated after the Tunisian authorities refused to apply a monthly adjustment and the IMF conceded that its reform was unpopular, Article 5 of the decree states:“The prices of unleaded petrol, ordinary diesel and sulfur-free diesel are set by the Commission in accordance with the rules relating to the monthly automatic adjustment mechanism, after observation of the average import prices in Tunisian dinars for each of these products during a period of one month ending before the first day of the month affected by the adjustment. The value of the monthly adjustment of the selling price may not exceed the rate of one point five per cent (1.5%) of the selling price in effect since the last adjustment, either up or down. The rate mentioned in the previous paragraph shall be increased to two per cent(2%) as of 1 January 2021.” The fact that this decree was issued three days before the letter of intent was sent seems to be fairly convincing evidence that the IMF demanded its publication as a strong condition(a prior action) before it would grant the loan under a supposedly non-conditional Rapid Financing Instrument program. 76 Taking advantage of the low prices—at least in late March and early April—the IMF re-imposed a monthly adjustment with a maximum monthly variation of 1.5% upwards or downwards, to be increased to 2% from January 2021, which is supposed to equal at most 6% per quarter. Based on the data published in the monthly energy report, we have been able to reproduce the evolution of fuel price subsidies during the COVID-19 pandemic emergency. The fuel subsidy per unit is defined as the difference between the transfer price per unit at which STIR sells(selling price) and the unit import price at which STIR buys(import price). 76 The IMF report on Tunisia’s RFI request states:“Seizing the opportunity of low international oil prices, the authorities adopted legislation to move to an automatic monthly price adjustment mechanism that will gradually eliminate subsidies on the three main fuels and thus remove a major structural risk to the budget going forward. The mechanism will be applied for the first time by April 8, resulting in a 1.5 percent decline in pump prices across all covered fuel categories.” IMF Tunisia EFI Request, April 2020, p. 5. 54 Uncovered: The role of the IMF in shrinking the social protection

Table 6. Evolution of fuel subsidies in Tunisia Fuel SSP (super unleaded) (millime/L) Apr. 16 Dec. 16 Dec. 17 Dec. 18 Dec. 19 Jan. 20 Feb. 20 Mar.20 Apr. 20 May 20 Jun. 20 Jul. 20 Aug. 20 Sep. 20 Oct. 20 Nov. 20 Dec. 20 Jan. 21 Feb. 21 Mar. 21 Apr. 21 May 21 Jun. 21 Import price 658 768 1048 1381 1416 1346 1293 994 926 961 938 936 929 921 913 916 1074 1140 1239 1299 1319 1344 Source: Tunisian Ministry of Energy. Transfer price 822 817 934 1077 1138 1138 1138 1101 1080 1054 1029 1004 1004 1004 1004 1004 1004 1033 1067 1149 1149 1149 Subsidy 164 49 -114 -304 -278 -208 -155 107 154 93 91 68 75 83 91 88 -70 -107 -172 -150 -170 -195 Selling price 1650 1650 1800 1985 2065 2065 2065 2035 2005 1975 1945 1915 1915 1915 1915 1915 1915 1955 1995 2095 2095 2095 How international financial institutions affect social protection in Tunisia 55

We can see that, starting in April 2020, the import price fell below the STIR transfer price, canceling out the gasoline subsidy for all of 2020. Thus, according to the new automatic monthly formula, the price at the pump fell from 2,065 millimes per liter in January 2020 to 1,915 millimes per liter in August 2020. 77 We observe that the rise in the price of Brent crude began to accelerate beginning in late January 2021, due in particular to strict agreements restricting Organization of the Petroleum Exporting Countries(OPEC) production and some countries resuming activity. As a result, in February 2021, the authorities began to raise the price at the pump. 78 Other fuels followed more or less the same trend as petrol(see Appendix D). Only the 13 kg gas cylinder(LPG) was subsidized throughout the COVID-19 emergency. At the end of February 2021, the IMF published its annual Tunisia report under the Article IV consultation. That was part of the beginning of negotiations between Tunisian authorities and the IMF on a new structural adjustment program, which constitutes the fourth phase. In preparation for that, the IMF published a reform scenario on energy subsidies. 79 Regarding the IMF’s proposed new approach to a fourth phase of structural adjustment, the Article IV report added:“Given past failures and resistance, the authorities should consult and communicate with the broader public a medium-term reform program that will take the country in a new direction. To ensure its success, the reform program and associated medium-term fiscal framework would need to be supported by a social compact, with the main stakeholders committing to support the reforms within their remit.” 80 The IMF clearly stipulated that the monthly adjustment proposed by the Tunisian authorities was insufficient and that they should increase the maximum monthly threshold because of rising international fuel prices. Seeing that an increase has a greater psychological impact on the public than a decrease, combined with the harsh socioeconomic effects of the COVID-19 emergency on vulnerable populations, could we be at the dawn of new social tensions? Will a social compact suffice to justify likely monthly energy price increases? 77 The pump price generally follows the STIR transfer price, with the difference due to a more or less fixed margin and various government charges and taxes. 78 In late February 2021, Director-General of Hydrocarbons Rachid Ben Dali indicated that there would also be increases at the beginning of March 2021, adding:“[T]he last revision according to the self-regulation mechanism was 2% when it should have been 18%.” 79“The 2021 budget foresees a decline in energy subsidies to 0.3 percent of GDP, with monthly price changes limited to 2 percent under the automatic fuel price adjustment mechanism. However, oil prices are currently projected to exceed the US$45 per barrel envisaged in the 2021 budget. Staff advised reforming the adjustment mechanism for the three main fuels by letting prices at the pump move more closely with import prices and widening the band for monthly price adjustments.” IMF Tunisia: 2021 Article IV Consultation, February 2021, p. 16. 80 IMF(2021). Tunisia 2021, Article IV Consultation, 26 February 2021. 56 Uncovered: The role of the IMF in shrinking the social protection

V. Conclusion Since 2012, supported by the World Bank, IMF involvement in social protection in Tunisia has focused on targeted transfers(social assistance) and non-targeted transfers(energy and food subsidies) by conditioning its financial support on: i) the reduction of energy subsidies to lower the budget deficit and direct some of these resources as targeted transfers to the“poorest” and ii) the elimination of food subsidies when a database for targeting“the poorest” is in place. This approach consists of adopting upstream measures to mitigate social risks through social safety nets for the poorest populations. Aiming to target extreme poverty and mitigate the trauma of structural adjustments, it is understood as compensating the poorest for austerity measures. The World Bank champions a residual and minimalist approach to social protection which translates into a policy of“social risk management.” This approach is fundamentally opposed to the humanrights-based approach that views social protection as a fundamental universal right and advocates establishing—even gradually—a social protection floor for all to fight against inequalities. There are many limits to targeting“the poorest” using the PMT, which has difficulties regarding inclusion and exclusion due to the significant information asymmetry that results from the low availability of disaggregated data and the irregularity of household surveys to monitor the situation, in Tunisia as well as other developing countries with large informal economies. A growing number of experts and organizations now advocate broader programs, such as universal child benefits or universal non-contributory old-age pensions, that are based on simple categorical criteria. Tunisia is also slowly implementing a social protection floor to ensure social security for all, as guaranteed by the constitution. Price subsidies that are part of generous social protection schemes, including those for food and energy, are often rooted in the political and institutional structure of developing countries, while the targeted social assistance programs(such as cash transfer programs) that replace them are often new, complex and costly to set up. That makes them subject to cuts, which leave many vulnerable households without coverage and at risk of further impoverishment when prices rise, because subsidies are eliminated during recessions. By analyzing the impact of IFI interventions on non-targeted transfers in Tunisia, we have shown that the“scientific” arguments the IMF and World Bank used to replace its universal approach to food and energy subsidies with their targeted approach were fallacious: food subsidies are progressive, equitable and more effective for eradicating poverty than energy subsidies or the PNAFN social assistance program. As for energy subsidies, the World Bank study on this topic is not comprehensive enough for us to be able to judge their redistributive effect. However, energy subsidies do have a positive effect on poverty eradication that surpasses PNAFN targeted transfers. In the first phase of structural adjustment, the IMF seemed to adopt a consultative approach with the Tunisian authorities and even chose the CRES analysis over the World Bank How international financial institutions affect social protection in Tunisia 57

regarding the PNAFN leakages. But in the structural adjustment’s second phase, the IMF got back on track and promoted the World Bank’s analysis, despite its flaws. Gradually, the two IFIs abandoned all exchanges with the Tunisian authorities and almost exclusively ordered prior actions. This so increased social tensions that the IMF was forced to abandon the Extended Fund Facility program in July 2019. After the first COVID-19 shock, the IMF encouraged countries to continue expansionary fiscal policies by increasing public spending to mitigate disparities exaggerated by the socioeconomic crisis. 81 However, despite its more“progressive” rhetoric—including the goal of reducing inequality— the IMF has maintained its approach to social spending and protection. Indeed, in a September 2020“How To” note on social spending during and after the COVID-19 crisis, 82 the IMF continued its usual austerity programs(fiscal consolidation). Debt sustainability and channels for efficient spending justify the implementation of austerity programs, including reductions in social spending that“will most likely be deemed macro-critical through the fiscal sustainability and spending efficiency channels.[...] Country teams could also assess the effectiveness and efficiency of those measures in the context of social objectives and discuss the appropriate timeline for phasing out those that will not drive better social outcomes during the recovery.” During a press conference on January 19, 2021, the IMF emphasized that exceptional support measures that increase fiscal spending should be temporary, but not removed too soon in the name of fiscal consolidation. The global COVID-19 crisis offered the IMF and the World Bank an unhoped-for opportunity to apply the famous shock treatment during the third phase of structural adjustment. Taking advantage of the COVID-19 emergency, the two IFIs ordered the implementation of all the reforms they had failed to force through during the second phase(i.e., monthly adjustment, application texts of the “Amen Social” law, and unique identifiers). On March 1, 2021, in an interview with Express FM radio, Minister of Economy, Finance and Investment Support Ali Kooli indicated that a Tunisian delegation was going to negotiate a new agreement with the IMF in early April 2021. With Brent crude oil prices increasing in tandem with global economic activity, this move is certain to open a fourth phase of structural adjustment. It is hard to anticipate its outcome, but social unrest is a certainty. 81 IMF Managing Director Kristalina Georgieva highlighted the growing inequalities and the importance of avoiding a premature withdrawal from expansionary fiscal policies, as well as the need to step up social spending to support the most vulnerable households. 82 IMF(2020). How to Operationalize IMF Engagement on Social Spending during and in the Aftermath of the COVID-19 Crisis, p. 9. 58 Uncovered: The role of the IMF in shrinking the social protection

VI. Bibliography Alatas, V., Banerjee, A., Hanna R., Olken, B. A. and Tobias, J.(2012). Targeting the Poor: Evidence from a Field Experiment in Indonesia. Australian Aid(2011). Targeting the poorest: An assessment of the proxy means Test Methodology. Banque Mondiale(2013). Vers Une Meilleure Equité: Les Subventions Energétiques, Le Ciblage, et La Protection Sociale En Tunisie.(Towards greater equity: energy subsidies, targeting and social protection in Tunisia). Baum, T., Mshvidobadze, A. and Posedas, J.,(2016). Continuous Improvement: Strengthening Georgia’s Targeted Social Assistance Program. The World Bank. Ben Rouine, C.(2020). Inégalités et Fiscalité en Afrique du Nord. Observatoire Tunisien de l’Economie/ Tax and Inequality in North Africa, Tunisian Observatory of Economy 14/01/2020. Bibi, S. and Ben Cheikh, N.(2017).«Evaluation de la performance des programmes d’assistance sociale en Tunisie». Chandoul, J.(2015). Le Partenariat de Deauville, à l’origine des politiques économiques en Tunisie. Observatoire Tunisien de l’Economie. CRES(May 2017). Évaluation de la performance des programmes d’assistance sociale en Tunisie: Pour optimiser le ciblage des pauvres et freiner l’avancée de l’informalité. CRES(2019). Etude de faisabilité des garanties du Socle National de Protection Sociale. CRES and INS(2013). Analyse de l’impact des subventions alimentaires et des programmes l’assistance sociale sur la population pauvre et vulnérable. CRES, UNICEF and Ministère des Affaires Sociales.(2019).«La garantie d’une allocation pour enfants en Tunisie, Faisabilité, coût, impact et financement». Devereux, S. and Sabates-Wheeler, R.(2004). Transformative Social Protection, IDS Working Papers No. 232. ILO(2009). Recovering from the crisis: A Global Jobs Pact adopted by the International Labour Conference at its Ninety-eighth Session, Geneva, 19 June 2009. ILO(2012). Social protection floors for social justice and a fair globalization. Report IV(1), International Labour Conference, 101 st Session. IMF(2013). Tunisia: Letter of Intent, Memorandum of Economic and Financial Policies, and Technical Memorandum of Understanding May 24, 2013. How international financial institutions affect social protection in Tunisia 59

IMF(2013). Tunisia: Request for a Stand-By Arrangement—Staff Report; Press Release on the Executive Board Discussion; and Statement by the Executive Director for Tunisia. IMF(2016). Tunisia: Letter of Intent, Memorandum on Economic Financial Policies, and Technical Memorandum of Understanding May 2, 2016. IMF(2018). Tunisia: Letter of Intent, Memorandum of Economic Financial Policies, and Technical Memorandum of Understanding June 22, 2018. IMF(2018). Tunisia: Third Review under the Extended Fund Facility, and Request for Waiver of Applicability and Modification of Performance Criteria. IMF(2019). Policy Paper, A strategy for IMF engagement on social spending, June 2019. IMF(2020). Tunisia: Letter of Intent. April 2020. IMF(2020). Tunisia Request for Purchase under the Rapid Financing Instrument, IMF Country Report No. 20/103, April 2020. IMF(2020). How to Operationalize IMF Engagement on Social Spending during and in the Aftermath of the COVID-19 Crisis, 2020. IMF(2021). Tunisia 2021, Article IV Consultation, 26 February 2021. INS(2020). Les indicateurs de l’emploi informel 2019. INS(2020). Carte de la pauvreté en Tunisie. Kidd, S. and Gelders, B.(2016). Child wellbeing and social security in Georgia: The case for moving to a more inclusive national social security system. UNICEF: Georgia. Kidd, S., Gelders, B. and Bailey-Athias, D.(2017). Exclusion by design: An assessment of the effectiveness of the proxy means test poverty targeting mechanism, International Labour Office, Social Protection Department. Merrien, F.-X.,(2013). La protection sociale comme politique de développement: un nouveau programme d’action international, International Development Policy, Revue internationale de politique de développement, pp. 68–88. PNUD, and Ministère du Développement, de l’Investissement et de la Coopération Internationale. (2020). L’Impact de la pandémie du COVID-19 sur l’économie tunisienne pour l’année 2020. Silva-Leander, S. and Merttens, F.(2016). HSNP: Assessment of programme targeting. Oxford Policy Management. Veras, F., Peres, R. and Guerreiro, R.(2007). Evaluating the Impact of Brazil’s Bolsa Família: Cash Transfer Programmes. In Comparative Perspective, IPC Evaluation- Note No. 1, International Poverty Centre: Brasilia, Brazil. 60 Uncovered: The role of the IMF in shrinking the social protection

World Bank(1990). World Development Report 1990: Poverty. Washington, DC. World Bank(2001). World Development Report 2000/2001: Attacking Poverty World Bank(2006). World Bank Development Report 2006: Equity and Development. World Bank(2011). Assessment of the Identification of Poor Households(IDPoor) Program’s Procedures and Implementation. Draft Technical Report for the Human Development – Social Protection Unit East Asia and Pacific Region. World Bank(2013). Tunisia Social Protection Reforms Support Project, Project Appraisal Document (P144674). World Bank(2014). The Unfinished Revolution. Bringing Opportunity, Good Jobs and Greater Wealth To All Tunisians. World Bank(2018). Loan Agreement for Development Policy Loan 8885-TN(Closing Package). Washington, DC. World Bank(2021). Resilience, equity and opportunity – The World Bank’s Social Protection and Labor Strategy. 2012–2022 How international financial institutions affect social protection in Tunisia 61

Appendices Appendix A: Fuel subsidies in Tunisia Table A.1 shows that regular diesel was heavily subsidized until the COVID-19 crisis(except in 2016). In 2020, falling prices pushed the transfer price above the import price, eliminating subsidies on ordinary diesel for the rest of the year. Table A.1 Evolution of the ordinary diesel subsidy Ordinary diesel(millime/L) Apr. 16 Dec. 16 Dec. 17 Dec. 18 Dec. 19 Jan. 20 Feb. 20 Mar. 20 Apr. 20 May 20 Jun. 20 Jul. 20 Aug. 20 Sep.20 Oct. 20 Nov. 20 Dec. 20 Jan. 21 Feb. 21 Import price 549 716 1023 1414 1484 1381 1343 1184 1112 1060 1024 1005 996 977 964 963 Source: Tunisian Ministry of Energy. Transfer price 828 769 884 1051 1124 1124 1124 1095 1080 1062 1045 1027 1027 1027 1027 1027 Subsidy 279 53 -139 -363 -360 -257 -219 -89 -32 2 21 22 31 50 63 64 Selling price 1200 1140 1280 1480 1570 1570 1570 1550 1530 1510 1490 1470 1470 1470 1470 1470 1470 1500 62 Uncovered: The role of the IMF in shrinking the social protection

Table A.2 shows that the prices for diesel 50 has developed similarly to regular diesel. Table A.2 Evolution of the subsidy for diesel 50 Diesel 50/SS Apr. 16 Dec. 16 Dec. 17 Dec. 18 Dec. 19 Jan. 20 Feb. 20 Mar. 20 Apr. 20 May 20 Jun. 20 Jul. 20 Aug. 20 Sep. 20 Oct. 20 Nov. 20 Dec. 20 Jan. 21 Feb.21 Import price 607 763 1045 1464 1522 1391 1406 1036 1028 983 972 965 954 941 934 939 Source: Tunisian Ministry of Energy. Transfer price 870 838 953 1104 1169 1169 1169 1135 1116 1094 1072 1027 1049 1027 1005 1005 Subsidy 263 75 -92 -360 -353 -222 -237 99 88 111 100 62 95 86 71 66 Selling price 1450 1420 1560 1745 1825 1825 1825 1800 1775 1750 1725 1700 1700 1675 1650 1650 1650 1685 Table A.3 shows that the heavy fuel oil subsidy has evolved differently from the other three main fuels. Heavy fuel oil subsidies were totally removed for just three months in 2020(transfer price equals import price), then began to be subsidized again starting in October 2020. How international financial institutions affect social protection in Tunisia 63

Table A.3 Evolution of the heavy fuel oil subsidy Heavy fuel oil(DT/T) Apr. 16 Dec. 16 Dec. 17 Dec. 18 Dec. 19 Jan. 20 Feb. 20 Mar. 20 Apr. 20 May 20 Jun. 20 Jul. 20 Aug. 20 Sep. 20 Oct. 20 Nov. 20 Dec. 20 Jan. 21 Feb. 21 Import price 336 500 748 1062 1053 804 Transfer price 401 400 444 573 634 634 634 645 634 586 634 602 634 602 637 637 637 637 637 644 637 658 637 658 637 Subsidy 65 -100 -304 -489 -419 -170 -11 48 32 35 0 0 -7 -21 -21 Selling price 510 510 560 710 780 780 780 780 780 780 780 780 780 780 780 780 780 780 Source: Tunisian Ministry of Energy. Table A.4 shows that although international prices have fallen, gas cylinders are still subsidized to the same extent, allowing the subsidy to be somewhat reduced. Unlike other products, the selling price of gas cylinders has not changed for years. 64 Uncovered: The role of the IMF in shrinking the social protection

Table A.4 Evolution of the gas cylinder subsidy LPG(13 kg/cylinder)(DT/ cylinder) Apr. 16 Dec. 16 Dec. 17 Dec. 18 Dec. 19 Jan. 20 Deb. 20 Mar. 20 Apr. 20 May 20 Jun. 20 Jul. 20 Aug. 20 Sep. 20 Oct. 20 Nov. 20 Dec. 20 Jan. 21 Feb. 21 Import price 10.3 11.559 15.899 18.027 17.871 19.59 20.869 16.126 15.38 15.11 15.1 14.82 14.73 14.95 15.04 15.46 Source: Tunisian Ministry of Energy. Transfer price 3.032 2.913 3,06 3.001 2.882 2,882 2.882 2.782 2.782 2.782 2.782 2.782 2.782 2.782 2.782 2.782 Subsidy -7.268 -8.646 -12.839 -15.026 -14.989 -16.708 -17.987 -13.344 -12.598 -12.328 -12.318 -12.038 -11.948 -12.168 -12.258 -12.678 Selling price 7.4 7.4 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 7.7 How international financial institutions affect social protection in Tunisia 65

Appendix B: The complex reform of energy subsidies This appendix indicates the complexity of energy subsidy reform compared with food subsidy reform. Subsidies for fuel have historically been larger than those for electricity. Table B.1 shows that the most consumed fuels are diesel, followed by petrol, LPG, heavy fuel oil and paraffin. Table B.1 Total expenditure in Tunisia(explicit and implicit) on energy subsidies, 2013 Indicator LPG Petrol Diesel 50 Diesel Heavy Paraffin Electricity TOTAL ppm 0.2% fuel oil Subsidy rate, April 68 15 16 26 62 37 27/ 50 p.a. 2013(%) Total consumption 225 884 230 1,739 103 36 2,169 5,385 at selling price, April 2013(MD) Estimated rate of 214 23 22 40 165 66 30 p.a. price increase to remove the subsidy, April 2013(%) Expenditure Amount in April 2013(MD) 483 199 50 693 170 23 1,671 3,290 As percentage 0.7 0.3 0.07 1 0.2 0.03 2.4 4.7 of GDP(%) Projected amount end – 2013(MD) 749 321 75 1,071 214 32 2,569 5,032 As percentage of GDP(%) 1.0 0.4 0.1 1.4 0.3 0.0 3.4 6.6 Source: World Bank(2013). Sources: World Bank staff calculations; Ministry of Finance and Ministry of Industry data. Note: GDP in 2012: 70,400 MD. Expected GDP in 2013: 76, 240 MD. P.a= not applicable. Projected amounts at end-2013 as a percentage of GDP: explicit subsidies: 4.7%; implicit subsidies: 1.9%; total: 6.6%. 66 Uncovered: The role of the IMF in shrinking the social protection

In Figure B.1, the World Bank presents the fuel consumption for each economic sector. Diesel is mainly used in the agriculture and transport sectors; industry consumes mostly gas, either directly or in the form of electricity; households consume very little diesel, but a fair amount of LPG. Figure B.1 Fuel consumption in Tunisia by sector, 2012 5% Fuel oil Electricity 15% Electricity 2% Diesel 4% Kerosene Households 18% LPG 42% Solid biofuels 89% Natural gas 2% LPG 30% Petrol 10% Natural gas 15% Electricity 9% Charcoal 1% Natural gas 6% Fuel oil Transport Agriculture 68% Diesel 27% Electricity 5% Fuel oil Industry 38% Electricity 54% Natural gas 78% Diesel 16% Natural gas Commercial and public services 11% LPG 1% Kerosene 0% 3% Kerosene LPG Source: International Energy Agency(IEA), 2012. World Bank staff calculations. Source: World Bank(2013). 34% Diesel How international financial institutions affect social protection in Tunisia 67

In order to give a better idea of the destination of each subsidized energy product, the World Bank has summarized this information in Table B.2, which indicates the primary recipient of each fuel. Table B.2 Main energy users in Tunisia(by product) Product LPG Petrol Kerosene Diesel Fuel Source: World Bank(2013). Main user Households Transport Households Transport Electricity Share of consumption(%) 68 100 85 57 73 The most notable finding is that the transport sector is the largest user of fuels such as petrol and diesel and thus receives the largest proportion of energy subsidies. Households not connected to the gas network tend to use only LPG(generally in cylinders). At first glance, energy subsidies appear to be more important economically than socially. However, transport, the most heavily subsidized sector, also plays a key economic and social role in Tunisia. Given the transport sector’s preponderance, it is useful to try to estimate how it indirectly impacts household costs. Diesel appears to be significant to both transport and agriculture, which indirectly impacts food prices. Focusing solely on subsidized products’ direct impact on households does not allow us to analyze the dynamics of the indirect impacts of diesel, petrol, fuel oil, gas and electricity(Table B.2 does not show electricity use, which is mainly industrial). Both Figures B.2 and B.3(below) confirm transport’s central role, especially public transport, in household consumption by quintile. The richest quintile uses much more private transport and is much better connected to the natural gas network. Although no figures distinguish between public and private transport, energy subsidies for public transport appear to play a greater role for households that are not wealthy than those that are. The same is true of the LPG cylinder subsidy, which mainly and proportionately benefits the poorest households. 68 Uncovered: The role of the IMF in shrinking the social protection

Figure B.2 Share of transport expenditure in Tunisia by standard of living(%) Share of transport expenditure by standard of living (%) 10 Public transport 8 Private transport 6 4 2 0 1 (794) 2 (1368) 3 (1905) Quintile(DTN/person) 4 (2670) Figure B.3 Share of gas expenditure in Tunisia by standard of living(%) 50.0 Rate of connection to natural gas (% of households) 40.0 Natural gas connection 30.0 20.0 10.0 0.0 1 (794) 2 (1368) 3 (1905) Quintile(DTN/person) Source: EBCM 2010, NSI. World Bank staff calculations. Source: World Bank(2013). 4 (2670) 5 (5064) 5 (5064) How international financial institutions affect social protection in Tunisia 69

Appendix C: The main types of IMF conditionality The main types of IMF conditionality specifically applied to Tunisia are: • Quantitative conditions : Referred to as QPCs(quantitative performance criteria), they are conditions that the borrowing state must fulfil to ensure that the IMF Executive Board disburses scheduled tranches. They are the IMF’s basic austerity measures: a minimum threshold for primary fiscal deficit, maximum threshold for credit expansion, minimum threshold for foreign exchange reserves, and maximum threshold for current expenditure. The IMF sometimes adds country-specific quantitative performance criteria and can introduce indicative targets to its reviews, and in those cases, non-compliance does not delay disbursements. • Structural reforms : A negotiated program in which the IMF has the final word regarding a loan agreement or the disbursement of a tranche. The reforms are usually accompanied by an implementation plan. The IMF has two tools to enforce reforms, which often depend on its satisfaction or the extent to which the state has implemented the reforms: Structural benchmarks : These are structural reforms that must be implemented according to a precise timetable which may change without causing a breach of contract or stalling the disbursement of a tranche. Prior actions : As their name suggests, a government must complete these reforms before any review or agreement is signed. IMF staff will refuse to visit the country to sign a new agreement or conduct a field visit to review a program until the government has satisfactorily implemented prior actions. The IMF’s most powerful tool, prior actions amount to extortion and are often used to express the IMF’s displeasure with the government’s slow pace and/or resistance to implementing the measures. For example, at the end of the Stand-By Arrangement in April 2016, the IMF insisted that Tunisia pass the Central Bank Independence Law before signing a new Extended Fund Facility. 70 Uncovered: The role of the IMF in shrinking the social protection

Appendix D: How energy subsidy expenditures evolved during the IMF program in Tunisia As Figure D.1(below) shows, the price of a barrel of oil had been very high since December 2010, when the revolution began, before it fell sharply—from USD 110 in July 2014 to USD 26 in January 2016. Figure D.1 International oil prices, 2010-2016 Europe Brent spot price FOB(USD per barrel) 120 100 80 60 40 20 Source: US Energy Information Administration. R6 stated,“Staff and the authorities agreed that lower oil prices provide an important opportunity to depoliticize fuel prices and reduce energy subsidies in a sustainable manner. The 56 percent decrease in international oil prices since the beginning of 2014 has not yet translated in decreases in domestic retail fuel prices, which are, for some products, 30-50 percent above international levels[…], thus providing additional savings to the budget.”(IMF Tunisia Article IV Consultation, Sixth Review, October 2015, p. 23). Dec 17, 2010 Mar 17, 2011 Jun 17, 2011 Spe 17, 2011 Dec 17, 2011 Mar 17, 2012 Jun 17, 2012 Spe 17, 2012 Dec 17, 2012 Mar 17, 2013 Jun 17, 2013 Spe 17, 2013 Dec 17, 2013 Mar 17, 2014 Jun 17, 2014 Spe 17, 2014 Dec 17, 2014 Mar 17, 2015 Jun 17, 2015 Spe 17, 2015 Dec 17, 2015 Mar 17, 2016 How international financial institutions affect social protection in Tunisia 71

Table D.1 Budgetary expenditure on energy subsidies in Tunisia, 2012-2015 mTND SBA- R1 SBA- R2 SBA- R3 SBA- R4 SBA- R5 SBA- R6 2012 2111 2111 2111 2111 2111 2111 2013 2854 2854 2854 2854 2854 2854 2014 2500 2500 2460 2353 2353 2353 Source: Authors’ compilation from IMF Review 6. Note: Black is realized; grey is projected by the IMF. 2015 1489 1489 1441 1394 1826 392 Energy subsidies peaked in 2013 and decreased slightly in 2014 following the drop in international prices. The IMF’s final review(R6) projected a significant reduction in the energy subsidies budget, from TND 2.3 billion in 2014 to TND 392 million in 2015, due to the drop in international prices. Table D.2 Budgetary expenditure on energy subsidies in Tunisia, 2014-2020 Review in mTND EFF- R0 EFF- R1 EFF- R2 EFF- R3 EFF- R4 EFF- R5 – energy subsidy EFF- R5 – fuel subsidy(STIR) EFF- R5 – electricity& gas subsidy (STEG) 2014 3112 2015 918 918 918 918 918 Source: IMF reviews. Note: Black is realized; grey is projected by the IMF. 2016 166 197 197 197 197 197 2017 194 1427 1550 1550 1550 1550 2018 209 1276 2089 2706 2983 2700 1500 1200 2019 878 2142 2608 1625 983 2020 2190 1432 758 72 Uncovered: The role of the IMF in shrinking the social protection

IMF data on energy subsidies, particularly its first projections(in grey) during the initial review(R0), show how it became less optimistic that the government would eliminate energy subsidies regardless of international prices. During the Extended Fund Facility’s R0, the IMF planned to virtually abolish subsidies, beginning in 2016 with a budget of around TND 166 million and no further increases planned. This was because, in 2016, prices were still very low. However, projections for the year 2018 show a major readjustment as the IMF realized the resistance it faced. Starting with R5, the IMF began to record the energy subsidy spending managed by STIR(fuel imports) and those managed by STEG(gas imports and electricity subsidies). In its final review, the IMF was more optimistic about a decline in the subsidy for STEG than that for STIR. How international financial institutions affect social protection in Tunisia 73

Appendix E: The return of the World Bank and the implementation of a targeting policy As we have noted, the IMF reinstated the World Bank when high international oil prices were indicating that it would be hard to eliminate energy subsidies. On June 29, 2018, the World Bank signed a loan agreement with Tunisia for an Investment, Competitiveness and Inclusion Development Policy Financing(ICI) program. However, most of the measures concerned past energy subsidies 83 and instead of presenting the usual matrix of conditions linked to the disbursement of tranches, the loan was to be disbursed in one go after the reforms were put in place. This indicates that in coordination with the IMF, the World Bank had also chosen to use prior actions to force compliance by the Tunisian authorities. These included: • #6: To contain the electricity and gas subsidies, the Borrower has approved an electricity and gas tariff adjustment in line with its Energy Subsidy Reduction Policy Note, pursuant to the Borrower’s Minister of Energy, Mines and Renewable Energy’s Letter addressed to STEG, dated May 10, 2018. • #9: To establish the foundation for a well targeted and fiscally sound expansion of the coverage of social safety nets, the Borrower’s Council of Ministers:(a) has approved and transmitted to Parliament the Social Protection Draft Organic Law(Loi Organique Amen Social) which, in particular, establishes the principle of targeting of safety net programs; and(b) has approved the new targeting model, including key implementation procedures, all pursuant to the Borrower’s:(i) Council of Ministers’ Decision dated September 13, 2017 and transmittal letter to Parliament dated December 26, 2017, and(ii) Council of Ministers’ Decision dated May 7, 2018, respectively. 84 Prior Action 6 shows that the World Bank was coordinating energy price rises with the IMF, particularly those managed by STEG. As for the implementation of a targeting policy, we observe that the World Bank did not exchange with the Tunisian authorities and not only used a prior action to force them to pass a law, but also to adopt the World Bank’s social protection targeting principle, as opposed to a universal principle. Prior Action 9 resulted in the Tunisian parliament adopting Law 2019-10 of January 30, 2019 on creating the“Amen Social” program. It refers the introduction of a number of implementing acts and the social affairs minister issuing a decree to establish a scoring model to judge the multidimensional deprivation of the categories, a decree establishing a Higher Council for Social Development, another to create a National Agency for Social Inclusion and Development to manage the database of vulnerable populations, and a fourth decree defining the conditions and procedures to be eligible for the Amen Social program, as well as others. 83 World Bank(2018). Loan Agreement for Development Policy Loan 8885-TN(Closing Package). Washington, DC. 84 ICI Development Policy Loan Agreement between Tunisia and the IBRD, p. 5. 74 Uncovered: The role of the IMF in shrinking the social protection

With regard to the database of vulnerable populations, it should be noted that during the R5, the IMF finally judged that it had been put into place in March 2019, a few weeks after the Amen Social law had been adopted. As for the World Bank ICI program, as of December 23, 2019, 85 only one progress report had been published and the project was deemed unsatisfactory because no implementing legislation had been published. Generally, when IFIs like the World Bank and the IMF force Tunisian authorities to pass a law(as was the case with the law on public-private partnerships), the Tunisian authorities express their displeasure by delaying publication of implementing legislation. That usually leads to the IFIs waiting for the Tunisian authorities to return to them when they need funding, at which time they can be forced to publish the implementing texts. Even if that takes years, in the end, the IFIs prevail because of the Tunisian authorities’ financial dependence, as we shall see below. During the COVID-19 crisis, the IMF imposed an automatic monthly formula and the World Bank acted in tandem. On June 15, 2020, the World Bank signed a new loan with the Tunisian authorities called First Resilience and Recovery Emergency Development Policy Financing or FRRE – DPF 86 and imposed Priority Actions as, deploring the lack of Amen Social law implementation texts, it did in the late December 2019 ICI loan progress report. The World Bank’s new PAs included: • 1. To enhance the capacity of the social protection system to respond to the COVID-19 crisis and future shocks, and improve household resilience:(a) the Borrower has expanded cash transfers as part of the immediate COVID-19 response pursuant to Law-Decrees No. 2020-3 and 2020-4, dated April 14, 2020 published in the Official Gazette No. 32, dated April 14, 2020;(b) the Borrower has approved Decree(Décret Gouvernemental) No. 2020-317,dated May 19, 2020, and published in the Official Gazette No. 45, dated May 20, 2020, setting the eligibility criteria and delivery procedures of different services under the AMEN Social Program; (c) the Minister of Social Affairs has issued Minister Order(Arrêté) dated May 19, 2020 and published in the Official Gazette No. 45, dated May 20, 2020, adopting the use of a new targeting model; and(d) the Borrower’s Council of Ministers has issued a decision on May 7, 2020, to integrate additional social services in the Amen Social registry. • 2. To establish the legal framework for citizen digital identification and interoperability for better and more efficient service delivery(of social programs in particular), the Borrower:(a) has enacted Law-Decree No. 2020-17, dated May 12, 2020, on Unique Citizen ID(“Identifiant Unique des Citoyens” or“IUC”) published in the Official Gazette No. 41 dated May 12, 2020; and(b) has adopted Decree No. 2020-312, dated May 15, 2020 and published in the Official Gazette No. 43, dated May 15, 2020, setting the objectives of the IUC, its content and technical 85 https://documents.banquemondiale.org/fr/publication/documents-reports/documentdetail/902641577116160470/ disclosable-version-of-the-isr-tunisia-investment-competitiveness-and-inclusion-p161483-sequence-no-01 86 https://documents.banquemondiale.org/fr/publication/documents-reports/documentdetail/384481592239471040/ official-documents-loan-agreement-for-loan-9137-tn-closing-package How international financial institutions affect social protection in Tunisia 75

specifications to define authorized user agencies and the purposes for their use, and to define the protocol and process interoperability related to the IUC.(First Resilience and Recovery Emergency Development Policy Loan Agreement between Tunisia and the IBRD) Action 1 of the loan agreement informs us that the World Bank was not happy with the nonpublication of the decrees in the Amen Social law that it had imposed through the ICI program and insisted on publication of some of the implementation acts. In fact, the World Bank had not demanded that decrees related to the creation of new institutions be published, but rather that the application acts related to the criteria and scoring model for targeting vulnerable populations. Thus, Article 2 of the minister of social affairs’ May 19 decree on determining the scoring model stipulated that the scoring model would adopt the PMT. In the middle of the COVID-19 crisis, the World Bank went so far as to impose its targeting model on the Tunisian authorities. With regard to the citizens’ identification at the heart of the targeting reform, the World Bank also took advantage of the temporary concentration of powers in the hands of the government to force passage of Decree-Law 2020-17 of May 12, 2020. The bill should have been discussed in parliament, allowing for a public debate on the conditions for creating and using the unique social identifier, the IUC. The fact that this law was passed as a prior action during a period of“democratic suspension” during one of the world’s biggest health emergencies indicates how far the World Bank was prepared to go to impose its targeting reform. During a hearing of the ARP Finance Committee on December 14, 2020, the World Bank“called for the adoption of a social identification mechanism in order to direct subsidies to those who need them” 87 —a sign that the World Bank was still not satisfied with the implementation of the reform. In response, the parliamentary finance committee chair, Haykel Mekki,“recalled that before independence, Tunisia and the WB had cooperated on financing the public education system and constructing infrastructure without Tunisia being subjected to excessive conditions” and declared, “We do not want relations between the colonizer and the colonized, and would prefer to preserve our traditional partnership on a friendly win-win basis.” 88 The World Bank Country Manager for Tunisia, Tony Verheijen, recalled the World Bank’s respect for Tunisia’s sovereignty. 89 However, the remarks by the head of the Finance Committee to the ARP show that the Tunisian authorities are increasingly reluctant to take orders(via priority actions) from the World Bank and the IMF. The fourth phase of structural adjustment will show if the two parties can interact on a new footing, or if the IFIs will dig in their heels. 87 https://www.webmanagercenter.com/2020/12/14/460771/la-bm-appelle-la-tunisie-a-accelerer-les-reformes-convenuesdepuis-2018/ 88 Ibid. 89 Ibid. 76 Uncovered: The role of the IMF in shrinking the social protection

Appendix F: Definitions of social protection Table F.1 summarizes the different definitions of social protection used by the IMF, the World Bank, and United Nations agencies, particularly the ILO, which is a key international actor influencing definitions and approaches to social protection. Table F.1 Various definitions of social protection IMF Social protection is a policy instrument that provides benefits in cash or in kind to the most vulnerable individuals or households, including: 1. Social assistance(social safety nets/targeting the most vulnerable) 2. Social insurance(pensions) World Bank Human rights mechanisms/ UN/ILO Social and labor protection help individuals and societies manage risk and volatility and protect against poverty through instruments that enhance resilience, equity and opportunity: 1. Social assistance(social safety nets) such as cash transfers and targeted food aid 2. Social insurance: old-age and disability pensions, unemployment insurance Social protection(or social security) is a basic universal right. It protects everyone through a defined set of basic social security guarantees—the national social protection floor. The national social protection floor is a combination of policy measures that a society provides to protect its members against economic and social distress caused by the absence or substantial reduction of employment income due to various contingencies(sickness, maternity, industrial accidents, unemployment, disability, old age or the death of the breadwinner), healthcare and benefits for families with children, the elderly and persons with disabilities. How international financial institutions affect social protection in Tunisia 77

IMF conditionalities and social protection in Jordan: Regression or improvement? by Laith Alajlouni

Table of contents I. Introduction 79 II. Defining social protection systems, globally and in the Jordanian context 81 III. The history of social protection in Jordan 83 IV. Challenges to social protection spending in Jordan 88 V. The IMF and Jordan 92 VI. Data gaps 110 VII. Conclusion 111 VIII. Policy recommendations 112 I. Introduction The history of the IMF’s active engagement with Jordan goes back to 1989, when the Kingdom signed its first Stand-By Arrangement(SBA) with the Fund. 1 Jordan subsequently signed a series of agreements with IMF until its first“graduation” from its programs in 2004. These agreements entailed a series of monetary, fiscal, and business reforms that included the elimination of subsidies on necessary goods, privatization, and the liberalization of the financial sector. After these agreements, the government of Jordan followed expansionary policies from 2004 to 2010. However, this period of high spending and moderate taxation was challenged by the global financial crisis, and more importantly, by the Arab Spring and the resulting regional turmoil from 2011 onward. The consequent economic hardship pushed Jordan to request the IMF’s help again in 2011, thus finding itself in a new series of commitments to the Fund’s programs. Table 1 details the history of IMF programs in Jordan. 1 https://bit.ly/3tUPQUT IMF conditionalities and social protection in Jordan: Regression or improvement? 79

Table 1. History of IMF lending commitments with Jordan Facility Extended Fund Facility Extended Fund Facility Stand-By Arrangement Stand-By Arrangement Extended Fund Facility Extended Fund Facility Extended Fund Facility Stand-By Arrangement Stand-By Arrangement Source: https://bit.ly/3nQdVLP Date of arrangement March 26, 2020 Aug. 24, 2016 Aug. 03, 2012 Jul. 03, 2002 Apr. 15, 1999 Feb. 09, 1996 May 25, 1994 Feb. 26, 1992 Jul. 14, 1989 Expiration date Ongoing Aug. 23, 2019 Aug. 02, 2015 Jul. 02, 2004 May. 31, 2002 Feb. 08, 1999 Feb. 09, 1996 Feb. 25, 1994 Jan. 13, 1991 In this respect, this study will focus on the impact of the IMF reform programs in Jordan during the era following the Arab Spring, primarily from 2011 to 2021, during which Jordan signed four subsequent agreements. These agreements included fiscal consolidation programs that involved the elimination of electricity and bread subsidies, as well as raising income tax rates. By contrast, this period also witnessed a change in the IMF’s approach in Jordan, as it further encouraged the Jordanian government to expand social protection schemes. This study will therefore analyze this puzzling situation and examine the impact of these programs in order to assess whether they resulted in improvement or regression in the quality and scope of social protection in Jordan. We find that the IMF reforms were associated with inflation hikes leading to a reduction in Jordanians’ disposable income throughout the implementation of fiscal consolidation programs. The following sections will define and conceptualize social protection systems in the Jordanian context, then present the history and development of the Jordanian social protection system. Subsequently, the study will discuss the root causes of the weak social protection system in Jordan, therefore laying the foundation to discuss the impact of IMF policies on the quality and scope of these programs. Finally, a conclusion, comments on data gaps, and recommendations are provided. 80 Uncovered: The role of the IMF in shrinking the social protection

II. Defining social protection systems, globally and in the Jordanian context Social protection systems are defined in various ways by different institutions. Aside from some minor differences, what all these definitions have in common is that social protection systems should be designed for the purpose of helping everyone who is vulnerable, in need, at risk of falling into poverty, or is failing to attain education, healthcare, and basic daily needs. Additionally, the benefits of social protection systems go beyond protecting vulnerable groups, as they have been proved to effectively promote increased and sustained economic growth and productivity. 2 Similarly, social protection reduces inequalities that could hamper economic growth and social mobility. As such, social protection has always been at the center of the global community’s interests. For instance, the Sustainable Development Goals(SDGs) announced by the United Nations General Assembly in 2015 emphasize the importance of social protection systems. One of SDG 1’s targets (Target 1.3) stresses the necessity to“implement nationally-appropriate social protection systems and measures for all” in order to eradicate poverty and promote equality. Moreover, SDG 3(in Target 3.8) and SDG 8(in Target 8.b) highlight the need to achieve universal healthcare coverage and to build a global strategy for youth employment by 2020, respectively. To better understand how social protection systems are defined, this study will reference definitions from the International Labour Organization(ILO), World Bank(WB), the United Nations International Children’s Emergency Fund(UNICEF), and the Asian Development Bank(ADB). These will inform the process of conceptualizing social protection in the Jordanian context. The ILO defines social protection systems as“the set of policies and programs designed to reduce and prevent poverty and vulnerability throughout the life cycle. Social protection includes benefits for children and families, maternity, unemployment, employment injury, sickness, old age, disability, survivors, as well as health protection. Social protection systems address all these policy areas by a mix of contributory schemes(social insurance) and non-contributory tax-financed benefits, including social assistance.” 3 The World Bank defines social protection similarly, as the set of“policies and programs that help individuals and societies manage risk and volatility, and that protect them from poverty and destitution through instruments that improve resilience, equity, and opportunity.” 4 UNICEF refers to social protection as“the range of policies and programs needed to reduce the lifelong consequences of poverty and exclusion.” 2 Ostry, Jonathan David, Andrew Berg, and Charalambos G. Tsangarides. Redistribution, inequality, and growth. International Monetary Fund, 2014. 3 ILO.“World social protection report 2017–19: Universal social protection to achieve the Sustainable Development Goals.” (2017). 4 Costella, Cecilia, and Adea Kryeziu. Resilience, equity, and opportunity: The World Bank’s social protection and labor strategy 2012-2022-consultations report. No. 82582. The World Bank, 2012. IMF conditionalities and social protection in Jordan: Regression or improvement? 81

The ADB provides a more comprehensive definition, referring to social protection as“the set of policies and programs designed to reduce poverty and vulnerability by promoting efficient labor markets, diminishing people’s exposure to risks, and enhancing their capacity to protect themselves against hazards and interruption/loss of income. Social protection consists of five major elements:(i) labor markets,(ii) social insurance,(iii) social assistance,(iv) micro and area-based schemes to protect communities and(v) child protection.” 5 Relying on the above, the study at hand conceptualizes social protection in the Jordanian context based on four main elements:(i) the government’s social spending,(ii) access to quality health services,(iii) access to quality education, and(iv) cash transfer social assistance. Article 6 of the Jordanian constitution stipulates that“the state guarantees jobs and education within the range of its capabilities” and that“the law protects motherhood, childhood, and the elderly” as well as sponsoring“people with disabilities.” Although the Jordanian constitution clearly states that the law should provide social protection for vulnerable groups, there is no common, referential, or specific definition of social protection in Jordan. In other words, there is no clear and obligatory definition of social protection for Jordanian policy-makers to refer to when they design social policies. For example, the National Strategy for Social Protection 2019-2025 6 lacks a clear-cut definition of social protection, although it outlines three main areas of focus to improve the social protection regime in Jordan: first, increasing the available opportunities for Jordanian families to achieve self-sufficiency; second, enabling citizens through healthcare, education, and social services; and third, focusing on targeted social assistance that enables the poorer segments of society to meet their basic needs. Given this, it could be argued that the social protection system in Jordan is characterized by two main aspects: it is designed in a way that takes into account limited state capacity 7 and the lack of a welldefined state commitment for social protection that informs policies. These elements are reflected in the efficiency, effectiveness, and scope of Jordanian social protection systems, which will be covered in the following sections. 5 Asian Bank for Development Social Protection Framework, 2003. 6 Jordan did not have a clearly articulated vision for social protection before this strategy. 7 As mentioned above, the Jordanian constitution makes it clear that social protection policies should be balanced with the state’s“capabilities.” 82 Uncovered: The role of the IMF in shrinking the social protection

III. The history of social protection in Jordan The legal and institutional framework Since its establishment in 1946, state–citizen relations in the Hashemite Kingdom of Jordan have been characterized by a rentier social contract. 8 This implies that citizens receive social benefits in exchange for being loyal to the governing regime, so the state is always responsible for providing social services, including education, healthcare, public sector employment, and other services, in order to guarantee the political loyalty of its subjects. However, these social benefits were not the result of a deliberate process to design and build a universal social protection system. Were the institutional design of the Jordanian state to be examined, it would be clear that it only has one official institution mandated by law to build a“comprehensive social security system,” 9 namely the Ministry of Social Development. Overall, the ministry works on two main areas: organizing social initiatives and the work of social associations operating in the country, and providing social assistance to the poorer segments of society. However, the ministry has not been successful in building a universal social protection system in Jordan. In addition to the Ministry of Social Development, a number of other state institutions can be considered part of the social protection system in Jordan. These include the Ministry of Labor, the Ministry of Health, the Ministry of Education, and the Social Security Corporation. Labor policies are under the remit of the Ministry of Labor, which is legally required to oversee and organize the labor market with a view to protecting labor rights and creating jobs. The ministry, however, has not been effective in delivering on this mandate. According to the global Sustainable Development Report, Jordan’s 2020 score on protecting“fundamental labor rights” was 0.5/1. 10 Moreover, the Jordanian labor market is highly informal. For example, Jordan’s Strategy Forum estimates that 41.4% of workers are informal workers. 11 This indicates that 41.4% of the labor force lack basic labor rights protections like health insurance and social security; hence the ministry’s failure to deliver on its primary mandate. 8 Al-Ajlouni, Laith Fakhri, and Allison Spencer Hartnett.“Making the Economy Political in Jordan’s Tax Revolts.” Middle East Report Online(MERIP). https://merip. org/2o19/o2/making-the-economy-political-in-jordans-tax-revolts 24(2019): 2o19. 9 Ministry of Social Affairs law and its amendments No. 14 of 1956. 10 https://bit.ly/3zguoeV 11 https://bit.ly/3y8jxCl IMF conditionalities and social protection in Jordan: Regression or improvement? 83

On healthcare, around 6.5% of the government’s budget in 2019 was allocated to the Ministry of Health. However, only 56% of Jordanians benefited from the ministry’s health insurance. 12 Overall, health insurance policies in Jordan—public and private—cover only 72% of the population. 13 Nonetheless, Jordanians still seem more satisfied with their health system than citizens of other countries in the Middle East and North Africa(MENA). According to the Arab Barometer data from 2019, 63% of Jordanians said that they were completely satisfied or satisfied with the healthcare system, a higher percentage than in Tunisia(39%), Algeria,(32%), or Morocco(18%). The Jordanian Ministry of Education offers free access to primary and secondary school education, but only 73.5% of Jordan’s population go to public schools(schools run by the Ministry of Education or by UNRWA, t he UN Relief and Works Agency for Palestine Refugees in the Near East), while the remaining 26.5% attend private schools. 14 The socioeconomic situation in Jordan For the past 10 years, Jordan has been experiencing a challenging socioeconomic situation that has revealed the fragility of the country’s social policies, including social protection. Currently, the overall unemployment rate in Jordan is 25%. 15 Although 62% of the population is aged between 15 and 65(i.e., working age), Jordan has an extremely low overall economic participation rate, at just 34% by the end of 2020, as shown in Figure 1. According to the ILO labor force data, Jordan had the lowest male labor market participation rate in the MENA region in 2019, at 63.8%. This was lower than in Tunisia(68.9%), Morocco(70.1%), Egypt(71.1%), and Saudi Arabia(78.4%). More importantly, Jordan had the third lowest female labor participation rate in the world in the same year(14.6%), only ranking above Yemen and Iraq. 12 Department of Statistics data. 13 Department of Statistics data. 14 Ministry of Education statistical report 2020. 15 Q1/2020 Employment and unemployment data, the Department of Statistics. 84 Uncovered: The role of the IMF in shrinking the social protection

Figure 1. Low labor market participation rates in Jordan 100.0% Economic participation rates(2010-2020) 80.0% 60.0% 40.0% 39.5% 39.0% 38.0% 37.1% 36.4% 36.7% 36.0% 39.2% 36.2% 34.3% 34.0% 20.0% 0.0% 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Jordan’s Department of Statistics, Employment and Unemployment Survey 2020. These labor market figures indicate structural challenges that hinder Jordanians from joining the labor market. In fact, these challenges are explained by the failures of the social protection system and the absence of the welfare state, and would have been less severe had there been a functioning social protection policy. Significant informality, poor public transportation, and the absence of decent work environments and labor rights protections are the main reasons why many Jordanians are reluctant to join the labor force. The high unemployment rates are also explained by the poor formal education system in Jordan. The poverty rate in Jordan reached around 15.7% in 2018, 16 and, according to the World Bank, 17 is expected to increase by 11% due to the COVID-19 pandemic. Within the same context, 75.8% of formal workers in Jordan earn less than JOD 600 a month, which is equivalent to USD 846.3. 18 This level of income, coupled with the overall social and economic environment, does not enable Jordanians to be more economically active, and therefore limits their chances of social mobility and makes it increasingly challenging for them to afford a decent living in light of the absence of public goods and services. It is undeniable therefore that Jordan faces a complex socioeconomic situation, given the low effectiveness of the social protection system, which, were it strengthened, could have acted as a buffer to protect the most vulnerable from falling into poverty after the COVID-19 pandemic. 16 https://bit.ly/3yb4toE 17 https://bit.ly/370mREM 18 Social Security Corporation annual report, 2019. IMF conditionalities and social protection in Jordan: Regression or improvement? 85

Social spending in Jordan As mentioned above, this study focuses on social spending in terms of health, education, and social assistance. Between 2008 and 2020, the government of Jordan spent an average of 19.4% of its annual budget in the form of social spending(excluding spending on subsidies), per Figure 2 below. 19 Figure 2. Social spending in Jordan in 2020 is relatively lower than 10 years ago 40.00% Social spending percentage of total public spending(2008-2020) 35.00% Ministry of Social Development share Ministry of Education share 30.00% Ministry of Health share Total social spending as share of public spending 25.00% 20.00% 18.25% 18.52% 20.73% 19.09% 17.64% 15.00% 9.67% 10.00% 10.33% 11.36% 10.60% 10.77% 5.00% 6.96% 6.41% 7.81% 6.96% 6.55% 0.00% 1.62% 2008 2009 1.78% 2010 2011 2012 1.57% 2013 2014 2015 2016 1.52% 2017 2018 0.33% 2019 2020 Source: Social Security Corporation annual report, 2019. The government of Jordan spent around 10.7% of the government’s annual budget on education, the sector that receives the largest amount of public spending, with a total allocation accounting of around 3.35% of GDP. 20 Despite the education sector’s receipt of the most government funds as a percentage of GDP, it still receives less money than the education sector in Morocco(5.3%), Tunisia (6.6%), or Egypt(3.8%), and education outcomes in Jordan remain moderate or weak. As Table 2 shows, Jordanian school students performed below the global average on the three sections of the 2018 Programme for International Student Assessment(PISA) test. 21 19 Spending on subsidies is not clearly outlined in the published governmental fiscal data. 20 According to 2020 figures. 21 https://bit.ly/37jxXVk 86 Uncovered: The role of the IMF in shrinking the social protection

Table 2. Jordanian students’ scores on the 2018 PISA test Reading Math Science Jordan’s score 419 400 429 Global median 466 468 458 Source: https://bit.ly/37jxXVk Ranking second in terms of public spending is health, per state fiscal data. From 2008 to 2020, the Ministry of Health received around 7.1% of the government’s annual budget, which accounts for 2.04% of 2020 GDP. This spending is reflected in the relatively high satisfaction rates with the health system in Jordan compared to those in other countries of the MENA region(63% of Jordanians are completely satisfied or satisfied with the healthcare system). To support its program of cash assistance, the Ministry of Social Development 22 receives approximately 1.6% of annual public spending in Jordan, as shown in Figure 2. This includes both the current and capital expenditures of the ministry’s activities, thus demonstrating that it is potentially underbudgeted compared to the other ministries. Available data on social assistance shows direct social assistance costs the Jordanian government an average of JOD 88.4 million per year(2013-2019), which constitutes around 1.1% of the total government spending—see Figure 3 below. Figure 3. Different types of financial assistance in Jordan 100.0 National Aid Fund annual social assistance by category(JOD million) 90.0 84.9 84.9 80.0 70.0 79.1 60.0 50.0 40.0 30.0 Regular financial aid Immediate aid Physical rehabilitation aid 71.6 Temporary financial aid Regular financial aid Total 20.0 12.3 10.0 4.7 0.0 2013 2014 2015 2016 Source: National Aid Fund annual reports(2015-2019). 2017 2018 92.0 80.1 10.4 2019 22 The ministry is responsible for distributing social assistance through the National Aid Fund. IMF conditionalities and social protection in Jordan: Regression or improvement? 87

Direct social assistance in Jordan falls under five main categories, defined as follows: 1. Regular financial aid: monthly financial aid provided to needy households struggling with permanent poverty and need. These monthly transfers range from JOD 50 to JOD 200, depending on the social and economic situation of the recipient household. 2. Temporary financial aid: monthly financial aid provided to needy households struggling with temporary poverty and need. These monthly transfers range from JOD 50 to JOD 200, depending on the social and economic situation of the recipient household. 3. Emergency financial aid: financial assistance provided to households that need urgent help due to exceptional circumstances such as forced displacement, natural disasters, and other unexpected and urgent events. 4. Immediate aid: a one-time financial transfer provided to households that need a certain sum of money to solve an emergent financial challenge. 5. Physical rehabilitation aid: a one-time financial transfer limited to JOD 600 provided to ill people and people with special needs to assist them in buying medical equipment. IV. Challenges to social protection spending in Jordan The Jordanian social protection system faces challenges from several political and economic factors. Issues such as clientelism, tribalism, and the mismanagement of local development are considered the main political obstacles facing the establishment of an effective, efficient, and comprehensive social protection system in the country. Economic factors such as the state’s poor fiscal capacity and the need for fiscal consolidation reforms represent the main economic challenges. This section will explore these issues and their impact on the Jordanian social protection system. Political challenges to the social protection system The economic and fiscal challenges that prohibit the establishment of a universal and effective social protection system in Jordan are closely related to, and ultimately caused by, the political challenges. As discussed above,“from the early decades of Jordanian statehood, the Hashemite regime cultivated a broad social contract that insulated the ruling elite from demands for political inclusion by providing government jobs and public services to key constituencies. The government used foreign aid inflows 88 Uncovered: The role of the IMF in shrinking the social protection

to subsidize basic commodities and expand the public sector to bolster national employment. Under this arrangement, citizens tolerated minimal political rights in exchange for state jobs and economic patronage.” 23 Similarly, Tell shows that, since the colonial period in Jordan, the government of Jordan has subsidized tribal leaders to cooperate in maintaining the security of the monarchy. 24 Moreover, this clientelistic approach was later applied to the Jordanian industrialist community also, and this group“became the chief beneficiary of industrial subsidies and government procurement, developing a vested interest in the continuous inflow of aid and close political ties with the West.” 25 Contemporarily, Rodríguez argued in 2018 that tribal organizations played a role as a client in a patron-client setting in Jordan. She also argued that the tribes themselves played the role of a subpatrons for their own clients. The same argument was adopted by Kao in 2015. 26 This clientelistic model has shaped development policies in Jordan for years, diverting state resources from being allocated to universal social protection schemes and priority development projects toward elites who gain their legitimacy through their influence and power inside their constituencies, through tribal politics. Moreover, the patron–client model has diverted social protection institutions goals from protecting the vulnerable toward being tools for“the reproduction of existing power relations,” which made the“formal system of social protection to be both extremely resilient and inefficient in tackling poverty and income inequality.” 27 Jordanian social and developmental public policies were therefore not designed to eliminate existing problems or to achieve optimal social and economic results, but rather have been mainly driven by the need to satisfy tribal leaders and other influential people who can influence public policy-making and are able to control the public through their clientelistic networks. Consequently, the prioritization of public spending in Jordan was mainly dependent on satisfying the ruling elites and their subjects in the different Jordanian constituencies. This inefficient spending and clientelistic policy-making have negatively influenced redistribution policies in Jordan and led to the weakening of state capacity, which in turn has hindered the establishment of a universal social protection system. Economic challenges to the social protection system The Jordanian state’s fiscal capacity has had major effects on the social protection system, as Jordan has been struggling with a chronic budget deficit for over 20 years. In 2020, for example, the 23 Laith Fakhri Al-Ajlouni, Allison Spencer Hartnett“Making the Economy Political in Jordan’s Tax Revolts,” Middle East Report Online, February 24, 2019. 24 Tell, Tariq. The Social and Economic Origins of Monarchy in Jordan. Springer, 2013. 25 Ibid. 26 Kao, Kristen Elaine. Electoral Institutions, Ethnicity, and Clientelism: Authoritarianism in Jordan. University of California, Los Angeles, 2015. 27 Cerami, Alfio.“Social Protection and the Politics of Anger in the Middle East and North Africa.” Social Policies 2, no. 1 (2015): 117-138. IMF conditionalities and social protection in Jordan: Regression or improvement? 89

Jordanian government budget deficit constituted 7.4% of GDP after receiving grants, and 10.5% of GDP excluding grants. This deficit led to a surge in public debt in the last few years, which increased from approximately 54.5% of GDP in 2010 28 to approximately 101.2% of GDP in 2020. 29 These rates were increasing despite of the implementation of the IMF-backed fiscal consolidation programs in Jordan, as will be elaborated further in this paper. We found that despite the implementation of these programs, the ratio of debt service to GDP increased from 3.8% in 2010 to 9.4% in 2020. The high public debt and chronic budget deficits have constrained the Jordanian government’s public spending, leaving little room for fiscal maneuvering. However, room can be created through revision of spending terms(like security spending) and by implementing political reforms that would eliminate losses due to clientelism and corruption. For instance, public debt loan interest represented 17.8% of total public revenues(including grants) in 2020, and 13.6% of the total public expenditures in the government’s budget in 2020. 30 Along the same lines, in 2020 civil servants’ wages and pensions represented nearly 23.8% of the government’s public revenues, and nearly 18.1% of the government’s total expenditures. Additionally, Jordan is spending significant amounts on security and military activities, with military spending in 2020 representing 37.5% of the government’s total revenues and 28.6% of the government’s total expenditures. 31 As a result, 83.6% of the government’s revenues and 60.3% of its expenditures are spent on rigid items, whether economic(wages and loan interest) or functional(military and security spending). 32 It is clear therefore that enhancing social spending in Jordan requires long-term structural fiscal reforms to reduce public debt, debt service, civil servants’ related costs, and military and security spending. These reforms should not only be spending-related, however, as the government of Jordan faces more complicated issues relevant to the mobilization of public revenues. For instance, Jordan’s tax revenue to GDP ratio(tax effort) in 2019 was roughly 14.8%. This is lower than many countries in the region, including Egypt(15.4%), Turkey(17.1%), Morocco(22.3%), and Tunisia(22.8%). 33 These figures indicate that neither the Jordanian government’s tax collection nor is its management of tax evasion are effective. The government has room for an increased tax mobilization, but this requires political will and improved tax justice by moving gradually from depending on regressive taxes(sales taxes or other indirect taxes) to more progressive and direct taxes. 28 https://bit.ly/3yqkNBL 29 https://bit.ly/3fysLlb 30 https://bit.ly/3lwhZQ7 31 https://bit.ly/3rVbCXw 32 Author’s own calculations based on official data. 33 https://bit.ly/3rZ74Qe 90 Uncovered: The role of the IMF in shrinking the social protection

For example, an IMF-backed income taxation reform led to protests in Jordan’s capital Amman in 2018, mainly from professional unions and the private sector, who viewed the reforms that dealt with tax evasion as very harsh. 34 In this case, had the government been able to enforce the legislation by persuading the general public that it would improve tax justice, it would have been easier politically to pass the reforms. This also indicates the need to ensure and guarantee tax justice in Jordan in order to increase people’s willingness to pay taxes. Examining the structure of Jordanian tax revenues reveals that the tax collection regime has several inefficiency concerns. For example, tax revenues in Jordan are concentrated around a single source, namely sales tax, an indirect form of taxation. In 2019, sales tax constituted 70.6% of the total tax revenues collected by the government(see Figure 4). Moreover, tax revenues collected from individual businesses(0.9%) were very low compared to the size of this segment of the Jordanian economy and the economic activity it generates. Small and mid-size enterprises(SMEs) constitute 95% of enterprises in Jordan, 35 yet they only account for 0.9% of Jordan’s tax revenues, versus 16.8% paid by the large corporations, which make up only 5% of economic enterprises. These variations in the sources of tax revenues indicate significant tax evasion by SMEs. According to the official estimates, the government loses roughly USD 1.128 billion annually in tax revenues due to tax evasion, 36 which amounts to 17% of the tax revenues collected per year. These distortions in the tax regime are due to the easy tax collection solutions adopted by the government. As a result, it is easier for the government to increase its tax revenues by increasing sales taxes than by collecting income taxes. IMF policies clearly contribute to this approach, as governments tend to use easy tax collection approaches in order to meet the IMF requirements to raise public revenues and reduce budget deficits and public debt levels. 37 This informality is one of the factors that prohibits building a more effective social protection system; not only does it deprive workers of their right to social security and health insurance, it also leads to reduced tax revenues and thus fewer resources to finance the social protection system in Jordan. The size of the informal sector therefore needs to be reduced. 34 https://bit.ly/3Ha8Kxf 35 https://bit.ly/3Cgt8ud 36 https://bit.ly/3Cgv4mt 37 https://bit.ly/3zrEjgW IMF conditionalities and social protection in Jordan: Regression or improvement? 91

Figure 4. Sales tax is two thirds of tax revenue in Jordan Tax revenue structure in Jordan(2019) Sales tax Corporate tax International trade tax(customs) Employee income tax Real estate tax Individual business income tax 5.9% 4.1% 1.7% 0.9% 16.8% 70.6% 0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% Source: Ministry of Finance, Monthly bulletin, February 2020. The above illustrates the Jordanian government’s structural challenges due to its fiscal capacity. These are twofold; first, the rigidity of government expenditure, and second, the inefficiency of tax collection. Until these two issues are resolved, the Jordanian government will not have the fiscal capacity to improve its social protection system. V. The IMF and Jordan Jordan’s relationship with the IMF dates back to 1989, when Jordan requested IMF assistance after a severe economic crisis had hit the country in the later part of that decade. The first agreement (a Stand-By Arrangement) was signed in 1989 with the goal of reducing the account deficit. This agreement was then followed by six other agreements, until Jordan“graduated” from these programs for the first time in 2014. 38 Harrigan et al. divided these reform agreements into three phases:“The first phase consisted of a brief period of liberalization of the financial sector, the capital account and the exchange rate regime, which culminated in a twin banking and currency crisis. The second phase, from 1992–1999, was characterized by stabilization of the macro-economy and a limited process of structural reform accompanied by a brief growth rate recovery in the mid-1990s. The third phase, 2000–2004, was characterized by a more concerted reform effort under the new regime of King Abdullah II and an annual average real GDP growth rate of 5.4%.” 39 38 https://bit.ly/3fFBi5K 39 Harrigan, Jane, Hamed El-Said, and Chengang Wang.“The IMF and the World Bank in Jordan: A case of over optimism and elusive growth.” The Review of International Organizations 1, no. 3(2006): 263-292. 92 Uncovered: The role of the IMF in shrinking the social protection

After graduating from these programs, the Jordanian economy improved, and the average economic growth rate during the period from 2004 to 2008 was around 7%. 40 However, the global financial crisis and the regional turmoil that followed the Arab Spring, combined with the mismanagement of state resources in Jordan, caused an economic and financial crisis in the country after 2011, and the average annual economic growth rate during the period from 2011 to 2018 was 2.2%. Unemployment rates and public debt also skyrocketed. As a result, the government of Jordan was obliged once again to request IMF help, and it signed a 36-month Stand-By Arrangement in 2012. This agreement aimed at supporting fiscal reforms and the electricity sector, and Jordan received USD 2 billion to implement these reforms. They included “fiscal adjustment underpinned by expenditure and tax reforms, and comprehensive reforms in the electricity sector to bring NEPCO[National Electric Power Company] back to cost recovery.” 41 To do so, the Jordanian government had to eliminate electricity subsidies for households that consumed more than 600 kilowatts a month, 42 in addition to liberalizing fuel prices. 43 In 2016, Jordan signed a 36-month Extended Fund Facility(EFF) with the IMF. This agreement aimed at supporting Jordan’s reform agenda, in accordance with its 2025 Vision. The reforms that were requested in order to implement this agreement were mainly related to advancing fiscal consolidation, improving the business environment, and promoting employment. 44 Under the terms of this agreement, the fiscal consolidation agreed upon would target the reduction of the“general sales tax and customs duty exemptions, and amend the income tax law.” IMF emphasized the“need to reach operational cost recovery” for electricity, and the consolidation of the“Water Authority of Jordan’s finances,” 45 which meant eliminating subsidies on water and electricity services. It is not yet clear how much the government of Jordan would save from this. Officials argue that the elimination of electricity subsidies would help in the reduction of the Jordanian National Electric Power Company’s debt. In January 2020, only two months before the COVID-19 pandemic hit Jordan, the authorities reached the final Extended Fund Facility agreement with the IMF, for four years. This agreement aimed primarily to improve the competitiveness of the business environment and to boost inclusive growth. Unlike the previous agreements, this facility did not focus on fiscal reforms, except for tackling tax evasion and reforming tax administration with a view to improving collection. 46 40 https://bit.ly/3kmPja4 41 https://bit.ly/3xFJ3ip 42 https://www.ammonnews.net/article/162776 43 https://bit.ly/3krmycq 44 https://bit.ly/3yFVqMm 45 https://bit.ly/2VOHQbi 46 https://bit.ly/2X7bmJJ IMF conditionalities and social protection in Jordan: Regression or improvement? 93

The IMF and taxation in Jordan As with any conventional IMF program, fiscal consolidation and tax administration reform were pivotal in the IMF strategy in Jordan after 2011. In general, the IMF welcomed the majority of the fiscal consolidation measures taken by the government of Jordan during the period 2011-2021. For example, in 2012, the IMF welcomed“the authorities’ revenue-raising efforts and supports their intentions to eliminate inefficient tax exemptions and privileges that narrow the tax base, as well as broadening the coverage of the tax system(including through revamped property taxes) to better target wealthy taxpayers.” 47 These measures in 2012 were expected to yield around 1.2% of GDP in additional revenues(0.9% from removing tax exemptions, 0.3% from taxes on luxury goods, and 0.2% from non-tax revenue). In fact, these measures were praised by the IMF Executive Board in the same year. Yet, the board argued that“further fiscal consolidation will be essential over the medium term to return fiscal and external balances to a sustainable level.” 48 Paradoxically, the IMF in the same year pointed to the need to decrease the“tax burden in product and labor markets” which“can help create jobs and lower unemployment.” 49 The IMF explains weak public revenues in Jordan as the result of“weak tax administration, declining non-tax revenue, and also a slow economy,” 50 in addition to arguing that the reduction in tariffs from trade agreements has led to a decline in revenues. 51 The IMF also argued that“concerns about social tensions” led to putting on hold“plans for consolidation” in the period between 2012 and 2014. However, despite the IMF’s awareness of the social tensions caused by fiscal consolidation measures, it argued that fiscal consolidation and reducing subsidies between 2012 and 2014“affected mostly banks, hotels, telecommunication, rich households, and non-Jordanians,” calling for more broadbased adjustments“to minimize distortions and maintain competitiveness.” This, in turn, had a formidable impact on the middle and lower classes in Jordan. During the period of the study(2011-2021), two IMF-backed income tax laws were advanced by the government and implemented. In 2014, the government passed an income tax law that was expected to raise the tax revenue to GDP ratio by 2.5%, if it was combined with reductions in tax exemptions, according to the IMF. 52 However, despite the IMF’s expectations and the implementation of these reforms, the tax revenue to GDP ratio increased by just 0.5% between 2014 and 2018, as shown in Figure 5 below. 47 IMF-Jordan Article IV consultation, May 2012(IMF Country Report No. 12/119). 48 IMF Executive Board conclusion of the 2012 Article IV Consultation with Jordan, April 2020. 49 IMF-Jordan Article IV consultation, May 2012(IMF Country Report No. 12/119). 50 IMF-Jordan Article IV consultation, June 2014(IMF Country Report No. 14/152). 51 This shows inconsistencies in IMF policies; the IMF welcomed trade liberalization earlier in the 2000s. 52 IMF-Jordan Article IV consultation, June 2014(IMF Country Report No. 14/152). 94 Uncovered: The role of the IMF in shrinking the social protection

Moreover, the available data shows that, in 2016, the government of Jordan increased the prices of fuel, cigarettes, clothes, and alcohol through a reduction in sales tax and customs exemptions. 53 The same approach regarding taxation in Jordan continued in 2018. The IMF Executive Board “supported the continued gradual and steady fiscal consolidation. They were encouraged by the authorities’ commitment to continue to remove exemptions on the general sales tax and customs duties.” 54 This was translated into removing more sales tax exemptions in the general budget of the Jordanian government for 2018. Moreover, it also resulted in a new income taxation law that primarily aimed to“reduce the income tax thresholds for individuals and families toward international standards.” 55 Again, these reforms did not lead to significant increases in the tax revenues for the government of Jordan, as shown in Figure 5 below. Figure 5. Jordan’s tax revenue to GDP ratio increased over the last decade 18.00% 16.00% 14.00% 12.00% 10.00% 8.00% 6.00% 4.00% 2.00% 0.00% Tax revenue to GDP(2008-2020) 14.90% 15.02% 15.39% 16.76% 12.21% 12.30% 13.12% 13.94% 14.75% 15.02% 15.58% 12.14% 12.28% 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Ministry of Finance data. Overall, the IMF approach to taxation in Jordan was centered around raising new taxes and removing tax exemptions. These policies were associated with increases in inflation rates leading to a reduction in the purchasing power of Jordanians and their disposable income. Figure 6 below shows inflation hikes in the periods that witnessed the introduction of new taxes and the removal of tax exemptions. 53 Jordan’s official gazette, issue 5411, 21/7/2016. 54 IMF Executive Board conclusion 2017 Article IV Consultation with Jordan. 55 IMF-Jordan Article IV consultation, July 2017(IMF Country Report No. 17/231). IMF conditionalities and social protection in Jordan: Regression or improvement? 95

Figure 6. Inflation rates fluctuate with the rhythm of tax reforms, 2011-2019 6.0% 5.0% 4.0% 3.0% 2.0% 1.00% 0.00% -1.00% -2.00% 1.91% 2011 4.51% 2012 Inflation rate in Jordan(2011-2019) 4.87% 2.87% 3.33% 2013 2014 -0.87% -0.79% 2015 2016 2017 4.50% 2018 0.26% 2019 Source: Department of Statistics data. The impact of IMF conditionality on the social protection system in Jordan Does Jordan have room to maneuver its economic policies? In general, the IMF agreements with Jordan entail two types of conditions. The first are quantitative performance criteria(QPC), which preset“specific, measurable conditions for IMF lending that always relate to macroeconomic variables under the control of the authorities. Such variables include monetary and credit aggregates, international reserves, fiscal balances, and external borrowing.” 56 The second are structural benchmarks(SBs), which include the“reform measures that often are non-quantifiable but are critical for achieving program goals and are intended as markers to assess program implementation.” 57 In the case of Jordan, quantitative performance criteria included debt reduction, fiscal deficit reduction, improving tax revenues, and maintaining monetary stability. Structural benchmarks for Jordan included more specific measures like the integration of“the Income and Sales Tax Departments in a function-based revenue department, and the establishment of a temporary directorate to oversee the large- and medium-taxpayers’ offices” in 2004. 58 56 https://bit.ly/3yI9x44 57 https://bit.ly/2VRpcQ0 58 https://bit.ly/3xGkZLZ 96 Uncovered: The role of the IMF in shrinking the social protection

Prior to 2016, Article IV consultations between the IMF and the Jordanian authorities show that the IMF was more involved in supporting deep structural reforms using structural benchmarks, in combination with the conventional quantitative performance criteria of debt and fiscal deficit reduction. However, after 2016, IMF conditions represented either a continuation of the reforms already required by the earlier programs, or reduced structural reforms that mainly related to the reform of public financial management by improving the transparency and efficiency of tax collection. Furthermore, the IMF became more supportive of building social safety nets after 2016. However, this change did not have a notable impact on the improvement of the social protection system in Jordan. Table 3 below lists the main structural reform measures(conditions) required by the IMF since 2004 and tracks their changes and impact. Table 3. Conditionalities required by the IMF since 2004 Article IV consultation (Year) 2004 Agreement type SBA Conditionality Measures required type QPC 1. Reduction of government debt and rebalancing its composition in favor of domestic debt. 2. Raising the price of petroleum products and increasing other taxes in April 2004 to achieve the fiscal deficit target of 3.9% of GDP. SB 1. Enrolling new military recruits in the Social Security Corporation and narrowing of eligibility criteria for military pensions. 2. By March 2004, integrating the Income and Sales Tax Departments in a function-based revenue department and establishing a temporary directorate to oversee the large- and mediumtaxpayers’ offices. 3. Developing a risk-based audit plan to audit general sales tax(GST) taxpayers. 4. Reducing the ratio of GST nonfliers to 10% of the total taxpayers. 5. Using a single treasury account at the Central Bank of Jordan(CBJ). 6. Hiring an externally recognized audit firm for the auditing of the 2003 CBJ. 7. Financial statements, with the provision of a review by an international partner. IMF conditionalities and social protection in Jordan: Regression or improvement? 97

2007 No agreement (post-program monitoring discussions) 2012 SBA SB 2016 EFF SB 1.“Savings on transfers and other current spending, a better targeted social safety net, and improved public financial management.” 2. Strengthening the income tax regime by widening the tax base. 3.“Phasing out food subsidies.” 1 4.“Legislating a new public debt ceiling of 60 percent of GDP in 2011.” 1. Implementing an income tax law yielding additional revenue of about 1% of GDP. 2. Introducing a commitment control system through the Government Financial Management Information Systems to register, report, and account for expenditure commitments against cash allocations issued by the Ministry of Finance. 3. Preparing draft amendments to the CBJ law to strengthen autonomy and oversight, in line with Fund advice. 4. Announce a medium-term electricity/energy strategy incorporating the inputs provided by the World Bank, including a timetable and measures for bringing NEPCO back to cost recovery. 5. Increasing energy tariffs. 6. Implementing a national unified registry for targeting subsidies. 1. Submitting a new income tax law to Parliament (consistent with program understandings, including on the PIT thresholds, the rates’ structure, and transfer-pricing). 2. Submitting to Parliament a draft budget law for 2017 in line with program understandings and projections for 2017-19, including the estimates of tax expenditures. 3. Submitting to Parliament a new tax exemptions’ framework to reduce GST exemptions on domestic and imported goods and services, and exemptions on customs duties. 4. Publishing the public debt management strategy. 5. Finalizing the reorganization of the Public Debt Directorate. 6. In the water and energy sector, the government should“publish studies on cross-subsidization and options for price adjustments in response to oil price changes.” 7. Adopting an automatic electricity tariff adjustment mechanism with effective implementation on January 1, 2020. 8. Submitting to Cabinet and publishing an updated action plan on how to reduce the water sector’s losses over the medium term. 98 Uncovered: The role of the IMF in shrinking the social protection

2020 EFF SB 1. Submit to Parliament an amended investment law that removes all articles related to tax incentives, in line with TA(technical assistance) recommendations. 2. Adopting and announcing a three-year plan, with an evenly phased implementation starting in 2021, that(a) redirects electricity subsidies only to those who most need them; and(b) uses part of the achieved savings to reduce electricity tariffs for key business sectors, while containing NEPCO losses. 3. Expanding targeted temporary reductions in socialsecurity contributions to all new startups. 4. Implementing the Digital Volume Verification System(track-and-trace system) to monitor tobacco production and reduce cigarette smuggling. Overall, the IMF was quite involved in encouraging the Jordanian government to implement structural reforms in the subsidy system, narrowing down social spending, and accelerating privatization and trade openness. Despite this, the IMF conditions left some leeway for the Jordanian government to maneuver its measures in order to achieve the IMF’s objectives. In fact, a review of Article IV consultations with Jordan shows that the IMF structural benchmarks and quantitative performance criteria left enough room for the Jordanian government to manage these conditions. Furthermore, the majority of these measures were achieved in agreement with the Jordanian government, and many of them were initiated and proposed by the Jordanian government. In terms of social spending, the IMF approach to social spending schemes in Jordan can be divided into two main periods: pre-2016 and post-2016. During both periods, the IMF supported decreasing the universality of social spending and the movement toward a more efficient allocation of social spending through targeted assistance. However, before 2016, the IMF was more direct about cutting down social spending, which was dealt with as a fiscal burden on the Jordanian government’s budget. By contrast, after 2016 the IMF encouraged better allocation only, without calling for major reductions in social spending, although it still required the elimination of water and electricity subsidies, despite both being essential for attaining the SDGs. This IMF approach notwithstanding, the aggregate numbers of public spending and social spending in Jordan do not show a major decline(see Figures 7 and 8 below). In other words, it could be argued that the IMF programs did not lead to a major change in the macroeconomic and fiscal policies adopted in Jordan. For instance, the Jordanian government’s fiscal deficit excluding grants remained around or even higher than its average in the period from 2008 to 2020, 59 despite the implementation of Stand-By Arrangement and Extended Fund Facility programs with the IMF. 59 The average fiscal deficit to GDP ratio between 2008 and 2020, excluding grants, was 6.98%. The annual deficit remained around this figure over the same period. IMF conditionalities and social protection in Jordan: Regression or improvement? 99

The ratio of public expenditure to GDP did not decline significantly despite the implementation of the Stand-By Arrangement agreement between 2012 and 2016. In fact, the public spending to GDP ratio increased from 26.93% in 2012 to 28.06% in 2016. Likewise, despite another Extended Fund Facility agreement in the period from 2016 to 2020, the public spending to GDP ratio increased to 31.14% in 2020, as shown in Figure 8. However, this increase in public spending can be explained by the increase in debt service. The debt service to GDP ratio increased from 3.8% in 2010 to 9.4% in 2020, as shown in Figure 9 below. More importantly, debt service was a major source for consuming public revenues, where it constituted around 41.3% of the total government revenues in 2020, compared to 9.3% in 2010(Figure 10). Figure 7. Jordan’s budget deficit grew despite fiscal consolidation 12.00% 10.00% 8.00% Fiscal budget deficit to GDP ratio(2020) 10.42% Deficit(excluding grants) to GDP ratio Deficit(including grants) to GDP ratio 10.05% 7.38% 6.00% 4.68% 4.00% 2.00% 0.00%1.50% 2008 2009 4.31% 2010 2011 2012 6.72% 2.15% 2013 2014 2015 5.04% 2.47% 2016 2017 2018 2019 2020 Source: Author’s own calculations based on data from the Ministry of Finance and the Central Bank of Jordan. Figure 8. A structural gap: Jordan’s public revenues and expenditures(to GDP) 35.00% 30.00% 24.05% 25.00% 20.00% 19.37% 15.00% Public revenues and expenditures to GDP(2008-2020) 27.26% 16.84% 29.98% 22.26% 28.27% 23.23% 31.14% 21.09% 10.00% Total expenditures to GDP ratio 5.00% Local revenues to GDP ratio 0.00% 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author’s own calculations based on data from the Ministry of Finance and the Central Bank of Jordan. 100 Uncovered: The role of the IMF in shrinking the social protection

Figure 9. Jordan’s debt service is at record highs 10.0% 9.0% 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0% 0.0% 3.8% 2010 3.9% 2011 4.5% 2012 Debt service to GDP Ratio(2010- 2020) 4.7% 6.0% 5.5% 6.1% 5.8% 5.3% 2013 2014 2015 2016 2017 2018 8.9% 2019 9.4% 2020 Source: Author’s own calculations based on data from the Ministry of Finance and the Central Bank of Jordan. Figure 10. Jordan’s debt service is shrinking the fiscal space 50.0% Debt service to domestic and total revenues ratio(2010-2020) 40.0% Debt service to total revenues ratio Debt service to domestic revenues ratio 33.8% 30.0% 20.0% 15.5% 29.4% 10.0% 46.5% 41.3% 0.0% 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Source: Author’s own calculations based on data from the Ministry of Finance and the Central Bank of Jordan. The IMF and Jordan during the COVID-19 crisis When the COVID-19 pandemic erupted in 2020, the small state of Jordan was already facing a complex socioeconomic situation. The unemployment rate had reached 19.1% by the end of 2019, before the arrival of the pandemic. This rate subsequently reached 25% by the end of the first quarter of 2021. Similarly, Jordan’s informal workers were the most affected by the pandemic. Moreover, the public debt to GDP ratio increased from 96.6% by the end of 2019 to 111.65% by the end of 2020. 60 60 https://bit.ly/3k3UiMB IMF conditionalities and social protection in Jordan: Regression or improvement? 101

Most importantly, the pandemic disrupted Jordan’s main sources of foreign capital, such as tourism, expats’ remittances, and exports. The COVID-19 pandemic has therefore exacerbated an already complex economic and social situation in Jordan. Three months before the pandemic arrived in Jordan, the IMF reached a staff-level agreement with Jordan on a four-year Extended Fund Facility. 61 The agreement was“centered on increasing growth and stimulating job creation, strengthening external and fiscal stability, increasing transparency, and improving social spending.” 62 To achieve these objectives, the agreement focused on Jordan’s five-year reform program, which was presented by the government of Jordan at the London initiative in 2019. 63 This reform program is centered“on a pro-growth reform agenda—which is based on measures to improve tax administration and reduce tax evasion, as well as more effective public-sector investment, reduced business costs, and measures to improve government transparency and the investment climate. Key reforms include reduced electricity prices for businesses to improve competitiveness through reforming energy tariffs in Jordan by‘reducing cross-subsidies contributed by businesses and received by households,’ together with development of a plan to reduce production costs and direct households’ subsidies only to those who need it.” 64 Accordingly, the IMF Executive Board approved providing Jordan with USD 1.3 billion to support the implementation of the program. Although the agreement to implement this program was concluded in the pre-COVID-19 period, the program represents a further advancement in the IMF’s changing approach toward Jordan after 2016. The program focused on improving employment and reducing poverty by improved public investments (through public-private partnership projects) and the reduction of business costs through the reduction of electricity prices. The IMF therefore started to adopt more of a Keynesian approach in dealing with the Jordanian economy. This was evident in the structural benchmarks set out by the IMF in this agreement. Under this agreement, the structural benchmarks were completely different from those associated with previous agreements. The majority of structural benchmarks in the new agreement were related to the improvement of the business environment and to promoting inclusive economic growth. Consequently, as the COVID-19 pandemic started to gain momentum in Jordan, the IMF declared the necessity of adjusting the program modalities to help Jordan adapt to and contain“the rapidly changing circumstances.” 65 The IMF therefore approved an additional USD 396 million in“emergency assistance to Jordan to address the COVID-19 pandemic.” 66 This emergency financial assistance was 61 https://bit.ly/3z1UBOk 62 https://bit.ly/3z1UBOk 63 The London initiative is a major international conference organized by the UK government to support investment, growth, and jobs for Jordan. The conference tried to support the Jordanian economy by mobilizing global political leaders, CEOs of international firms, and investors to invest in Jordan. 64 IMF Country Report No. 20/101 65 https://bit.ly/3k8R8qN 66 https://bit.ly/3y0Ookf 102 Uncovered: The role of the IMF in shrinking the social protection

provided to Jordan under the Rapid Financial Instrument(RFI) to“help address the country’s balance of payment needs and allow for higher spending on healthcare, containment, and assistance to households and companies most affected by COVID-19 crisis.” 67 With regard to social spending, the IMF continued to adopt the same narrative of improving the “effectiveness and efficiency” of social spending instead of supporting more universal social spending schemes. However, the IMF supported the expansion of the National Aid Fund’s cash transfer program over the years 2019-2021. Yet this did not lead to a significant increase in the national fund’s spending or the number of beneficiaries. The National Aid Fund’s expenses witnessed a limited increase from JOD 145.5 million in 2020 to a budgeted JOD 201 million in 2021. 68 The IMF also adopted a revised definition of social spending in the Jordanian context. Per the IMF, it includes:“(i) non-wage components of the education and health sectors’ current expenditure envelope; (ii) NAF’s social protection programs; and(iii) the school feeding program.” 69 However, these changes did not represent a paradigm shift or a major change in the IMF approach toward social spending in Jordan. Despite the pandemic and the new Keynesian approach, the IMF position remained static in terms of focusing on better targeting of social spending instead of increased scope or universality. In fact, over the past few years, there have been some gradual or“fragmented” institutional changes regarding the IMF’s discourse and requirements, which are related to the Washington consensus. 70 For example, with the global financial crisis, the IMF started to reduce its use of pro-market terms such as deregulation, trade liberalization, interest rate liberalization, and structural adjustment, among others. 71 However, despite that this attitude being applied to Jordan to some extent, the general practice did not lead to a reduction in neoliberal, pro-market policies, and did not lead to the strengthening of the Jordanian social protection system. The social impact of the IMF policies IMF programs have been important in maintaining Jordan’s macroeconomic stability, especially in the post-Arab Spring era. Nevertheless, the social impact of these policies has been negative. In addition, Jordan is still facing a challenging socioeconomic situation, despite the implementation of these programs. To better understand the socioeconomic situation in Jordan, the following section will look at a number of different indicators and assess the current challenges and inequalities in order to understand their origins and their relationship to the IMF’s policies. 67 https://bit.ly/3zeZzHL 68 https://bit.ly/3oC5nHd 69 https://bit.ly/3zfaWzr 70 Kaya, Ayse, and Mike Reay.“How did the Washington consensus move within the IMF? Fragmented change from the 1980s to the aftermath of the 2008 crisis.” Review of International Political Economy 26, no. 3(2019): 384-409. 71 Ibid. IMF conditionalities and social protection in Jordan: Regression or improvement? 103

In terms of human development, Jordan performs better than the oil-importing countries in the region, and slightly weaker than the oil-exporting countries, per the World Bank’s Human Capital Index 2020 (see Figure 11). 72 Despite the implementation of IMF programs throughout the last 10 years, and despite the limited state fiscal capacity, Jordan has maintained a relatively acceptable performance in terms of human capital development(i.e., health and education) compared to the region. Figure 11. Jordan is above most non-oil exporting countries in MENA MENA region countries' Human Capital Index scores(2020) United Arab Emirates Bahrain Qatar Oman Iran, Islamic Rep. West Bank and Gaza Saudi Arabia Kuwait Jordan Algeria Tunisia Lebanon Morocco Egypt, Arab Rep. Iraq Yemen, Rep. 0.67 0.65 0.64 0.61 0.59 0.58 0.58 0.56 0.55 0.53 0.52 0.52 0.50 0.49 0.41 0.37 0.0 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Source: The World Bank’s Human Capital Index. 72 The Human Capital Index(HCI) measures the amount of human capital that a child born today can expect to attain by age 18, given the risks of poor health and poor education that prevail in the country where she lives. It is designed to highlight how improvements in current health and education outcomes shape the productivity of the next generation of workers, assuming that children born today experience over the next 18 years the educational opportunities and health risks that children in this age range currently face. 104 Uncovered: The role of the IMF in shrinking the social protection

On the health aspect, the life expectancy for both women and men in Jordan is considered one of the lowest in the region according to World Bank data(see Figures 12 and 13). Moreover, a deeper look into data from Jordan’s Department of Statistics shows that Jordan is facing a challenging public health situation. For instance, 51% of Jordanians above the age of 50 have been diagnosed with diabetes. 73 Additionally, 43% of women between aged 15-49 have been diagnosed with anemia. One third of Jordanian children also suffer from anemia. Being diagnosed with anemia is directly related to social class:“38% of the poorest children are diagnosed with anemia compared to only 18% of children who belong to the richest segment of society.” 74 There is a clear relationship between socioeconomic inequalities and the health situation of children; for example, in 2018, Amman had the lowest anemia rate among children(25%) compared to the least developed cities of Ajloun(41%) and Maan(37%). Figure 12. Life expectancy for males is relatively low in Jordan Life expectancy at birth, male(years) Qatar United Arab Emirates Lebanon Bahrain Oman Algeria Kuwait Turkey Tunisia Saudi Arabia Jordan Middle East& North Africa Arab World Egypt, Arab Rep. Sudan 79.14 77.29 77.14 76.43 76.12 75.67 75.76 74.72 74.68 73.93 72.83 72.63 70.30 69.74 63.46 0 10 20 30 40 50 60 70 80 90 Source: World Development Indicators database, World Bank. 73 Population and household health census 2017-2018, the Department of Statistics: https://bit.ly/3syNnPe 74 Ibid. IMF conditionalities and social protection in Jordan: Regression or improvement? 105

Figure 13. Life expectancy for females is above the regional median Life expectancy at birth, female(years) Qatar Lebanon Turkey United Arab Emirates Tunisia Bahrain Saudi Arabia Kuwait Jordan Middle East& North Africa Libya West Bank and Gaza Egypt, Arab Rep. Arab world Sudan 82.01 80.86 80.56 79.31 78.72 78.42 76.75 76.58 76.30 76.07 75.95 75.75 74.35 73.83 67.18 0 10 20 30 40 50 60 70 80 90 Source: World Development Indicators database, World Bank. With regard to education, global quantitative indicators show that the Jordanian education system is suffering major quality issues. For example, Jordanian school students performed below the global average on the three sections of the PISA test 2018, as mentioned above. Moreover, the Human Capital Index 2020 report shows that 52% of 10 years old students in Jordan suffer from“learning poverty.” In other words,“52% of 10-year-olds cannot read and understand a simple text by the end of primary school. This is higher than both the average for its region(48%) and the average for Jordan’s income group(38%).” 75 75 https://bit.ly/3zb6fGy 106 Uncovered: The role of the IMF in shrinking the social protection

Consequently, the poor quality of the public education system in Jordan resulted in increased educational inequality in the country. For instance, Jordanians have become more likely to send their children to private schools in order to obtain a better quality of education, with around 30% of Jordanian students attending privately owned schools. 76,77 This leads to an increased financial burden for Jordanian households, in addition to more education inequality. Global indicators show that education inequality is evident in Jordan. For example,“Students from the richest 20 percent of households in Jordan score 448, while those from the poorest 20 percent score 384, a gap of 64 points on a scale that ranges from 300(minimal attainment) to 625(high attainment). This gap is larger than the typical gap across the 50 countries(55 points).” 78 Similarly,“a child in the richest 20 percent of households who starts school at age 6 can expect to complete 11.7 years of school by her 18th birthday while a child from the poorest 20 percent can expect to complete 10.7 years of school, a gap of 1 years of school.” 79 This poor situation in terms of the provision of both healthcare and education can be explained in terms of the failure of Jordan’s local development strategies and the failure of the national social protection system. Jordan could perform significantly better in terms of access to quality public health and education services. However, the poor efficiency of spending on these services and the randomness of prioritization of spending efforts may be hindering the improvement of the overall social protection system in the country. For example, although the Jordanian government maintained its level of spending on health and education throughout the implementation of the two fiscal consolidation programs with the IMF in the periods of 2012-2016 and 2016-2020, as shown in Figure 14 below, failures in both the educational and health systems still exist. 80 In fact, our content analysis of the IMF documents shows that the Fund itself did not request any significant reductions from the Jordanian government in spending on health and education. Even when the IMF does raise the subject of cutting the public sector wage bill, the Fund asked the government to exclude the health and education sectors from any intended cuts. Moreover, the IMF requests the government increase investments in these projects through the implementation of public-private partnership projects. However, the IMF did request some reforms that have had negative social impacts. In 2012, the IMF criticized the government additional spending that was caused by some social protection programs. Mainly,“a higher wage bill(by 0.6 percent of GDP as a result of a civil service reform expected to yield savings in the medium term); higher pensions and health outlays(0.3 percent of GDP); and spending on housing and medical assistance for Syrian refugees(0.5 percent of GDP).” 81 76 https://bit.ly/385xhn2 77 Other estimates shows that 26.5% of students in Jordan go to private schools. 78 https://bit.ly/3gnTNfv 79 https://bit.ly/3y8ax0a 80 The ratios of public spending on health and education to GDP kept growing, except for a slight decline during the period 2014-2017. 81 IMF Country Report No. 12/343. IMF conditionalities and social protection in Jordan: Regression or improvement? 107

Under the Stand-By Arrangement agreement of 2012-2016, the IMF was tracking government allocations to the health exemptions fund(Royal Court subsidies for needy patients),“focusing on targeting to avoid abuse.” 82 In 2015, the IMF criticized the government for exceeding budget allocation for the health fund by JOD 55 million. The Fund dealt with the health fund transfers as a“key driver of fiscal deviations,” estimating that it contributed around 0.9% of GDP in terms of additional public expenditure. 83 This approach continued under the Extended Fund Facility(20162020), during which the IMF continued to track government transfers to the health fund. 84 To control government transfers to this fund, new regulations were implemented in early 2018“whereby medical exemptions could only be granted by the Royal Court, on a time-bound basis, following strict eligibility criteria and predetermined prices.” 85 The available data shows that the IMF projects to reduce the transfers from around JOD 278 million in 2017(actual) to around JOD 162 million in 2024(projected). With regard to the Extended Fund Facility(2016-2020), and despite its emphasis on the need to increase social spending in Jordan, the IMF’s approach to social spending remained the same, but this time with a focus on education. In 2018, the IMF expressed its“concern” regarding the spending efficiency of the education sector, arguing that teachers’ wages were two times higher than GDP per capita. 86 In 2019, the IMF set a spending floor of JOD 1.75 billion for social spending in Jordan, or around 6% of GDP that year, 87 but the government did not meet this requirement, as shown in Figure 14 below. 82 Sixth Review Under The Stand-By Arrangement, May 2015, IMF Country Report No. 15/115. 83 IMF Country Report No. 16/295. 84 Staff Concluding Statement of the 2016 Article IV Mission and First-Review under the Extended Fund Facility, November 14, 2016. 85 Second Review Under The Extended Arrangement Under The Extended Fund Facility, May 2019, IMF Country Report No. 19/127. 86 https://www.mof.gov.jo/ebv4.0/root_storage/ar/eb_list_page/%D9%85%D8%B1%D8%A7%D8%AC%D8%B9%D8%A 9_%D8%A7%D9%84%D8%A5%D9%86%D9%81%D8%A7%D9%82_%D8%A7%D9%84%D8%B9%D8%A7%D 9%85_%D9%88%D8%AA%D8%B1%D8%B4%D9%8A%D8%AF%D9%87_%D8%A2%D8%B0%D8%A7%D8%B 1_2018_%D9%85%D9%84%D8%AE%D8%B5_%D8%AA%D9%86%D9%81%D9%8A%D8%B0%D9%8A-0.pdf 87 Second Review Under The Extended Arrangement Under The Extended Fund Facility, May 2019, IMF Country Report No. 19/127. 108 Uncovered: The role of the IMF in shrinking the social protection

Figure 14. Government spending on health and education in Jordan improved but remained low 4.00% 3.50% 3.00% 2.50% 2.33% 2.00% Government spending on health and education to GDP(2008-2020) 2.82% 1.75% 3.29% 2.26% 3.00% 1.97% 1.50% 1.67% 1.00% Education spending as% of GDP 0.50% 0.00% 2008 Health spending as% of GDP 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 3.35% 2.04% 2019 2020 Source: Author’s own calculations based on data from the Ministry of Finance and the Central Bank of Jordan. Consequently, these failures to invest properly in human capital in Jordan have led to structural distortions and inequalities in the labor market. For instance, the economic participation rate in Jordan was 34.5% by the end of the first quarter of 2021. The fact that 65.5% of the population is economically inactive shows that Jordan is suffering from a serious problem of encouraging and enabling its population to join the labor market. Moreover, when it comes to gender disparities, the available data shows that 54.8% of Jordanian men are economically active, compared to just 14% of Jordanian women. More importantly, unemployment figures in Jordan indicate major distortions in the educational system. For example, the unemployment rate for Jordanian women who obtained a bachelor’s degree or above is 73.7%, while the same figure for men is 19.7%. Therefore, we can argue that Jordanian development policy failures are twofold: first, Jordan was weak in building productive and healthy human capital, and second, Jordan failed to build an enabling ecosystem that allows Jordanians to be active in the economy. Finally, the country’s Sustainable Development Report of 2021 shows that Jordan is performing moderately in terms of achieving the SDGs. 88 According to the report, Jordan ranked 72nd globally. However, Jordan’s score of 70.1 was slightly higher than the regional average(67.1). In terms of poverty elimination(SDG 1), the poverty headcount ratio at USD 3.20 per day in Jordan was 4.75%, which is higher than Lebanon(0.9%), Tunisia(2.87%), and Algeria(2.89%). Yet, this measure of poverty in Jordan was lower than countries like Morocco(6.37%), Iraq(19.62%), and Egypt(24.77%). By contrast, Jordan’s performance in terms of its population’s access to electricity, clean water, and 88 https://2021.dashboards.sdgindex.org/profiles/jordan. IMF conditionalities and social protection in Jordan: Regression or improvement? 109

sanitation services was one of the best globally. On labor rights and decent work conditions, Jordan’s performance was around the regional average. Based on the above, we cannot claim that the IMF programs had a direct negative or positive impact on the social situation in Jordan. However, IMF programs could be more supportive in terms of building a Jordan that is more resilient to socioeconomic shocks. In other words, the IMF could encourage Jordan to reform its fiscal policy in order to better allocate the available funds to social protection schemes. This will be better explained in the section covering recommendations below. VI. Data gaps In general, it could be argued that the Jordanian government is transparent in terms of the data it publishes on public finances and other economic and social issues. However, gaps do exist, and there is still room for improvement. Detailed fiscal data, in addition to microdata on poverty, informality, labor market issues, social assistance, and other socioeconomic issues and policies, is still lacking. The following are the main areas where Jordan lacks data: • The government of Jordan has not published updated microdata on poverty rates since 2010. This data is necessary to understand how poverty and the social class structure have changed in Jordan over time. More importantly, this data is critical to assess how IMF policies have affected Jordanian society since 2011. • The government of Jordan does not publish detailed data on how much it spends subsidizing food, electricity, and water services. This data is important to assess how much the government would save by eliminating these subsidies. It is also important to understand how the government will reallocate this money after eliminating subsidies. • Data on informality in Jordan is lacking. Updated and detailed data about informal workers and informal institutions is important in order to understand these vulnerable groups and how they can be included in social protection programs. • As with economic informality, Jordan lacks data on tax evasion. This is necessary to understand the Jordanian government’s potential to improve data collection. This would result in improved state fiscal capacity and a greater ability to expand social protection systems and programs. • The National Aid Fund data lack details on the beneficiaries of the cash assistance programs and whether or not they graduate from its programs. 110 Uncovered: The role of the IMF in shrinking the social protection

• The data available on unemployment benefits provided by the Jordanian Social Security Corporation is not sufficient. • Survey data on Jordanians’ attitudes toward the IMF programs is lacking. Moreover, Jordan needs survey data to assess the population’s satisfaction with the current social protection system. • Survey data on Jordanian policy-makers’ perceptions of the IMF programs is necessary to better understand how they engage with the IMF and how they view the alternative solutions. VII. Conclusion This study sought to analyze the social protection system of Jordan, its legal framework, and public social spending, and how these aspects have been affected by IMF conditionality in the period from 2011 to 2021. The study found that the Jordanian social protection system was already weak before these agreements, as Jordan lacked a well-designed legal and institutional framework for social protection. In addition, the current institutions that are responsible for providing social protection services suffer from many functional and capacity issues. Jordan has high levels of poverty, unemployment and informal labor, and its health and education services are of moderate quality. These challenges were exacerbated by the weak fiscal capacity by the state, combined with the IMF conditions. Yet, despite these challenges, this study finds that the government of Jordan can create a room for an increase in social spending through reprioritizing its spending, enacting political reforms that reduce corruption and clientelism, and by enacting progressive fiscal reforms. Moreover, these reforms should be combined with reforms in the health and education sectors, in addition to reforming the business environment. Jordan signed four subsequent agreements with the IMF during the period from 2011 to 2021. These agreements entailed a wide variety of reforms, including the removal of subsidies on bread and electricity and reforms to the tax regime. However, the macroeconomic situation in Jordan remained almost the same, or worsened. Unemployment rates reached 25%, the debt to GDP ratio reached 111.65%, and fiscal deficits witnessed a limited decline. On the social protection side, Jordan’s performance in the health and education sectors regressed, and poverty rates increased. Overall, although we cannot infer a causal relationship between IMF conditionality and the deteriorating economic and social situation in Jordan, we argue that these programs did not lead to improvements in the Jordanian social protection system. IMF conditionalities and social protection in Jordan: Regression or improvement? 111

VIII. Policy recommendations Fiscal policy • The Jordanian government should be more effective in mobilizing tax revenues. This should not be done through raising new taxes. Instead, the Jordanian government should be more effective in tackling tax evasion, in addition to imposing more progressive taxation in order to improve tax justice. • In order to improve Jordanian’s willingness to pay taxes, the government of Jordan must take concrete steps toward increasing tax justice in Jordan. • The Jordanian government needs to develop a clear strategy to reduce its dependency on indirect taxation(sales taxes). This form of taxation constitutes a burden on the middle class and the poorer segments of society, leading to decreased purchasing power and reduced quality of life. • The Jordanian government needs to provide incentives and exemptions for informal businesses to encourage them to formalize. These can include exempting formalizing businesses from licensing fees or taxation for the first year, or reducing social security subscriptions. • The Jordanian government needs to review military and security spending in order to reallocate some funds to health, education, and social assistance spending. • The Jordanian government needs to be more transparent regarding its expenditures and revenues. For example, there are no available figures on how much the government would save by eliminating some public goods and services subsidies(e.g., electricity), and where the saved funds would be re-allocated. Education and health • Government spending on education consists mainly of wages, and this high expenditure is due to the high number of teachers. The student-to-teacher ratio in Jordan is around 15.9, which is close to the Organisation for Economic Co-operation and Development(OECD) countries’ ratio of 14.5. However, the salaries are low. The government therefore needs to look into ways to allocate more funds to the ministry with the purpose of increasing teachers’ wages to improve their performance, which means they will deliver a better quality of education. 112 Uncovered: The role of the IMF in shrinking the social protection

• The Jordanian government needs to increase spending on curriculum reform and teacher training. • With regard to health, the Jordanian government needs to look into ways to increase the rate of health insurance coverage among its population. Moreover, health spending efficiency must be increased by reducing waste and corruption. • The Jordanian government needs to increase health spending on protective measures like awareness campaigns in order to reduce the levels of chronic diseases such as diabetes. • The high levels of anemia among Jordanian children show that the government needs to increase spending on early childhood healthcare. • The Jordanian government needs to look into ways to utilize the current primary health care centers in order to reduce the pressure on Ministry of Health hospitals, improving services and accessibility at both levels. Labor market • Jordan needs to address the issue of informal workers. As mentioned above, informal workers constitute around 41% of the workforce in Jordan. Formalizing these workers is necessary to protect them from falling into poverty, as it allows them to receive social security benefits and health insurance. • The Ministry of Labor needs to improve the effectiveness of its monitoring and law enforcement efforts in order to guarantee decent work conditions for those who work in Jordan. • Concrete measures need to be implemented in order to increase the employability of Jordanian women and therefore their rate of economic participation. These measures include the need to subsidize transportation for working women, to improve the quality of transportation, and to reduce the pay gap between men and women. Moreover, the Jordanian government should support businesses that take gender balance into consideration. • Similarly, active labor market policies(ALMPs) in Jordan lack proper, evidence-based design and implementation. Proper research needs to be conducted in order to better implement these programs, leading to increased employability of Jordanians and improved economic participation. IMF conditionalities and social protection in Jordan: Regression or improvement? 113

Civil society organizations • Civil society organizations in Jordan need to transfer their advocacy efforts from qualitative and opinionated advocacy campaigns to more evidence-based approaches. In other words, while pushing for universal social protection systems, Jordanian civil society needs to make arguments based on data in order to influence policy-making in favor of evidence-based solutions. • The donor community needs to increase funding directed to Jordanian civil society to improve their social protection advocacy programs’ effectiveness and reach. • Jordanian civil society organizations lack proper economic policy training and therefore need capacity building programs on how to read, understand, and analyze economic data in order to allow them to generate rigorous policy papers that advocate better social protection systems. • Jordanian civil society organizations should initiate programs that aim at monitoring government expenditure and the delivery of social protection services. • Due to the increasingly pressing economic challenges in Jordan, civil society organizations may need to build a coalition that lobbies for more inclusive economic reform and a universal social protection system. A coalition of this nature would help unify efforts, raise more funds, and increase the effectiveness of civil society advocacy campaigns. • Civil society organizations in Jordan need to increase efforts in advocating for governmental transparency on issues like public spending, social assistance, poverty rates, and other socioeconomic issues. This requires intensive training for civil society actors in order to strengthen their technical capacities in overseeing these issues, as governments usually find ways to influence public opinion and civic actors on these topics. 114 Uncovered: The role of the IMF in shrinking the social protection

SOCIAL SPENDING IN MOROCCO: THE INFLUENCE OF IMF CONDITIONALITY By Boutaina Falsy and Jamal Azouaoui Translation from French Nancy Du Plessis

Table of contents I. Introduction 116 II. The failings of a neoliberal socioeconomic model 118 III. Social protection in Morocco: Between universality and econometric targeting 127 IV. The COVID-19 crisis: Opportunities and obstacles to social protection 143 V. Recommendations 151 Appendices 152 I. Introduction In recent decades, the traditional interpersonal solidarity that made Moroccan communities so cohesive has declined at an ever-greater pace, 1 leaving more and more Moroccans stranded on the roadside of development, helpless before the vagaries of life. This major phenomenon is threatening a society already eaten away by anomie and decaying values. As this deadly process intensified in the late 1990s, social and institutional actors and public opinion recognized the urgent need for a new type of solidarity, with universal and equitable protection for all. But there is no consensus regarding universal social protection. Many crucial questions remain: Is the choice between welfare and insurance, professional and/or national solidarity, or between the Beveridge Model, the Bismarck Model, or a hybrid system? Should the scheme be single-pillar, or should we maintain several pillars at the risk of weakening the foundations? This social question has not yet been resolved, because it is not easy to respond to social issues in a country with varying statutes, limited budgetary space for action, and social actors who are often marked by intense corporatism. 1 Ishan Guennoun,”Cohésion Sociale, institutions et politiques publiques”, OCP Policy Center, Policy Brief – 17/38(November 2017),[https://www.policycenter.ma/publications/coh%c3%a9sion-sociale-institutions-et-politiques-publiques-0]. 116 Uncovered: The role of the IMF in shrinking the social protection

Nevertheless, in the early 2000s, an exuberant political and economic situation made it possible to initiate many reforms, such as basic medical coverage through the Mandatory Health Insurance (AMO) and the Medical Assistance Plan(RAMED), as well as the National Human Development Initiative(INDH) and the Social Cohesion Support Fund(FACS). After a sharp slowdown during the 2010s when many subsidies were eliminated, a new generation of targeted subsidies is being developed under the auspices of the IMF and the World Bank(WB). They are supported by the Single Social Registry(RSU), 2 a national databank of social security beneficiaries that aims to serve the most vulnerable. The RSU will also make it possible to distribute other benefits, such as family allowances. Morocco’s government has also been striving to make up for lost time with respect to public education. Illiteracy remains very high, with 44% of the population unable to read or write. Public healthcare remains the poor relation in the state budget. Morocco is the black sheep of the Middle East and North Africa(MENA) region and gets failing grades from the Organization for Economic Co-operation and Development(OECD). Are these negative scores only due to budgetary constraints? Does the policy of the lowest social offer come from the fear of crippling debt? Or is it an accepted governance mode that reflects Morocco’s strategic orientations, encouraged by the IMF(and the World Bank)? This case study analyzes the Moroccan state’s decisions on social protection, exploring how it is connected with government debt to the big international financial institutions(IFIs), particularly the World Bank and the IMF, bearing in mind that these institutions have their own definitions of social protection. We analyze education and healthcare spending and programs supported by the IMF and the World Bank, and their impact on the development of social protection. In the first section of the study, we look at how the IMF gradually restructured the Moroccan economy and social protection by applying conditionalities. In the second part, we analyze Morocco’s social protection system. We also study IMF discourse and policy. Finally, we look at the effects of the worst health crisis in Morocco’s history, the COVID-19 pandemic. This emergency revealed the limits of the neoliberal model, with its negligible social dimension. We also look at how the COVID-19 pandemic might yet produce a universal social security system. The appendices to this study present the global context, with a wealth of information on the Moroccan social protection system: its various programs, its beginnings, its governance and its limits. We also outline the history of the IMF and the World Bank in the Kingdom. These many different elements combine to offer the reader an image of Morocco’s reality and strategic choices. 2 The Single Social Registry is an information system for coordinating social protection activities with a national database on beneficiaries of social protection. It includes only those who are not covered by contributive systems. Social spending in Morocco: The influence of IMF conditionality 117

II. The failings of a neoliberal socioeconomic model Structural socioeconomic weaknesses: 2011–2020 The intervention of the Bretton Woods institutions, and the IMF in particular, has profoundly reshaped the economy of the Kingdom of Morocco since independence in 1956. 3 This intervention, along with the massive injection of capital starting in the late 1990s, has strongly impacted the country’s economic growth and thus Moroccans’ purchasing power. Yet Morocco remained dependent on foreign capital because of its balance of payments deficit, which was caused by an undiversified economy and mediocre investment performance. By the late 2010s, Morocco’s low economic redistribution model was losing momentum. Although living standards continued to improve, they only increased by an average of 2.7% in constant Moroccan dirham(MAD) terms between 2013 and 2019, compared to 3.6% 4 between 2007 and 2014. Rural Moroccans were particularly affected by the slowdown: their rate of advancement was halved, dropping from 4.6% between 2007 and 2014 to 2.7% between 2013 and 2019. 5 It should be noted that these figures are averages and do not fully reflect the inequalities of Moroccan society. One of the Moroccan economy’s main structural weaknesses is the inability of its national savings to keep up with one of the highest investment rates in the world(34% in 2019). 6 However, instead of investment generating additional points of growth, which releases more savings and allows for more investment, the opposite occurred. The growth was not equal to the efforts 7 and the marginal coefficient of capital 8 remained too high for the economy’s savings capacity. Indeed, generating 1 point of GDP in Morocco requires injecting seven units of capital. 9 3 For more information on the role of the IMF and the World Bank in reconfiguring the Moroccan economy, see Appendix A. 4 High Commission for Planning(HCP, the national statistics office of Morocco), Voluntary National Review of SDG implementation, 2020, 107,[https://www.hcp.ma/file/218567/]. 5 HCP and the authors. 6 From 2000 to 2019, the Kingdom made unprecedented investments in an unevenly distributed effort that mostly benefited the tertiary sector, where the rate of investment exceeded 45%. In industry, the rate of investment remained constant, at around 29%. However, agriculture, a major provider of employment that is decisive for Morocco’s trade balance, saw investments fall to an average of 7%. See High Commission for Planning, Etude sur le Rendement du Capital physique au Maroc, 2016, 28-29,[https://www.hcp.ma/Rapport-relatif-a-l-etude-sur-le-Rendement-du-Capital-Physiqueau-Maroc_a1653.html]. 7 On average 3.7%. See World Bank Group, Economic Update October 2021, 2021, 2,[https://thedocs.worldbank.org/en/ doc/c8738e9cd55e692fe59a71c43cb2505f-0280012021/original/Morocco-Economic-Update-October-2021.pdf]. 8 The average number of units of additional investment required to generate one unit of GDP. 9 High Commission for Planning, Etude sur le Rendement du Capital physique au Maroc, 10. 118 Uncovered: The role of the IMF in shrinking the social protection

Morocco also has been unable to diversify its economy and has very little industry. Since the early 1980s, the sectoral distribution of value added has shifted slightly toward services, but not enough, as agriculture continues to be key to wealth creation. Investments are increasingly concentrated in the mechanical and electrical industries, as well as in construction and public works, 10 all major importing sectors. The trade deficit is further deepened by importing hydrocarbons and cereals. Moreover, the national economy remains oriented toward the domestic market, where growth is driven by demand. These factors significantly increase imports, particularly of consumer goods, which, given the lack of export dynamism, aggravates Morocco’s negative trade balance. Morocco’s continued structural trade deficit indicates a detrimental international specialization and further supports the impression that it has a consumption-oriented economy. Low redistribution means economic growth does not benefit everyone The increase in living standards in Morocco conceals deep disparities. Not all households have benefited equally from this economic growth, and standards of living differ greatly depending on a person’s social background and whether they live in a rural or urban area. When median expenditures are disaggregated by social category, the inequalities are much more evident. In 2019, one in two Moroccans had an annual expenditure of less than MAD 15,187(MAD 1,266 per month, or about USD 130). 11 Disparities between urban and rural areas are even more pronounced: in cities, the median standard of living is around MAD 18,040 per person(MAD 1,503 per month, or about USD 150), while in the countryside, it is around MAD 11,233(MAD 936 per month, or below USD 100). Also in 2019, the standard of living of the richest 10% of Moroccans was equivalent to above MAD 37,631. The wealthiest decile of the Moroccan population dominated total consumption expenditure at 30.9%, almost 10 times more than that of the poorest decile, whose standard of living is below MAD 7,402(around USD 750). In 2019, the richest 20% of Moroccans monopolized 46.1% of total household consumption(47% in 2014), compared with 7% for the 20% least well-off Moroccans(6.7% in 2014). According to the latest publication by the High Commission for Planning(HCP, the national statistics office of Morocco), the Gini coefficient, which was 39.5 in 2014, had dropped to 38.5 in 2019, indicating a slight reduction in income disparity. The HCP, which had stopped communicating any data regarding the Gini coefficient since 2014, abruptly changed its policy with the advent of 10“The economic sector covering all activities relating to the design and construction of public and private buildings, both industrial and not, and infrastructure such as roads or pipelines”(Translated from French). See“Construction”, Wikipedia, last accessed May 11, 2022,[https://fr.wikipedia.org/wiki/Construction]. 11 High Commission for Planning, Evolution du niveau de vie des ménages et impact de la pandémie COVID-19 sur les inégalités sociales, March 2021, 3,[https://www.hcp.ma/Evolution-du-niveau-de-vie-des-menages-et-impact-de-lapandemie-COVID-19-sur-les-inegalites-sociales_a2676.html]. Social spending in Morocco: The influence of IMF conditionality 119

COVID-19 and confirmed that the 2019 Gini coefficient was 38.5. 12 In any case, this modest progress was swept away by the economic turmoil caused by COVID-19, and the pandemic battered a country with glaring socioeconomic vulnerabilities. Socioeconomic weaknesses reinforce each other Poorly performing investments, a high level of investment, and a chronic trade deficit have caused disequilibrium in Morocco’s balance of payments. 13 Social disparities pose a political risk that can only be mitigated by social measures—which must be funded. All these factors increase the country’s dependence on creditors like the IMF. Morocco’s structural weaknesses make the IMF an indispensable partner which is solicited as soon as internal and external difficulties appear. The IMF returns in 2012 Morocco turned to the IMF four times between 2011 and 2020. Table 1 outlines the Kingdom’s current relationship with the IMF. Only in 2020, under the weight of the COVID-19 pandemic, did Morocco need the funds made available by its Precautionary and Liquidity Line(PLL) arrangement of 2018(almost USD 3 billion). Table 1. Morocco’s position in the IMF as of 30 November 2021(in USD) Article VIII SDR department Net cumulative allocation Holdings Outstanding purchases and loans Precaution and Liquidity Line(PLL) (SDR million) 1,418.66 1,463.09 (SDR million) 1,499.80 Allocation(%) 100.00 103.13 Quota(%) 167.69 12 Ibid., 4. 13 The IMF’s founding charter states that it can provide temporary financing to countries with payment imbalances if they furnish the“adequate safeguards” specified by the Fund. In return, the IMF supervises their economic policies through a system of conditions. Generally speaking, the“Washington Consensus” is the neoliberal concept that shapes the conditions. Policies should be market-friendly, outward-oriented and macro-economically stable. A typical IMF policy package includes 10 policy prescriptions: fiscal discipline, reduction of general subsidies, tax reform, financial liberalization and free movement of capital, unified and competitive exchange rates, trade liberalization, removal of barriers to foreign direct investment, privatization, deregulation and property rights. Any country that does not comply with the IMF’s conditions will automatically be deprived of loans, which reduces its access to development aid and international capital markets. 120 Uncovered: The role of the IMF in shrinking the social protection

Latest Financial Commitments Type Date of agreement PLL 17 Dec. 2018 PLL 22 Jul. 2016 PLL 28 Jul. 2014 Expiration date 7 Apr. 2020 21 Jul. 2018 21 Jul. 2016 Amount approved Amount drawn (SDR million)(SDR million) 2,150.80 2,150.80 2,504.00 0.00 3,235.10 0.00 Projected payments to Fund (SDR million; based on existing use of resources and present holdings) Forthcoming 2021 2022 2023 Principal Charges/Interest 15.77 15.77 Total 15.77 15.77 2024 962.10 11.71 973.81 2025 537.70 1.75 539.45 Source: IMF(2021). 14 During this period, the IMF asked the Moroccan government for two specific measures that, if not compensated by an increase in income, could harm the well-being of most of the population: a flexible exchange rate and the abolition of universal energy subsidies. IMF experts have consistently called for greater fiscal restraint, including in phases of demand contraction and recession, which call for counter-cyclical fiscal policies. 15 As this study shows, exaggerated austerity has tied the government’s hands by preventing it from increasing budgets for social programs, including education and healthcare. Unlike OECD countries, Morocco was forced to adopt austerity budget measures for two years, even during the COVID-19 pandemic. The IMF forecasts a 1% decline in the budget deficit in 2021, a figure that is expected to fall further in 2022. 16 14 International Monetary Fund, Morocco: Financial Position in the Fund as of November 2021, Last accessed in February 2022,[https://www.imf.org/external/np/fin/tad/exfin2.aspx?memberKey1=680&date1k creditors ey=2099-12-31]. 15 To revive the economy and boost growth. 16 International Monetary Fund, PRESS RELEASE NO. 21/370 – IMF Staff Completes 2021 Article IV Mission to Morocco, 10 December 2021,[https://www.imf.org/en/News/Articles/2021/12/10/pr21370-morocco-imf-staff-completes-2021-articleiv-mission-to-morocco]. Social spending in Morocco: The influence of IMF conditionality 121

Causes of the economic and political crisis In the period 2010-2011, faced with several closely linked exogenous and internal shocks, 17 the Moroccan government requested IMF help. The political crisis in Morocco that followed the Arab Spring coincided with economic instability: along with the demonstrations, a decline in external demand from the EU, rising energy costs, and the balance of payment imbalance 18 combined to create a serious crisis. The situation was urgent. The public was demanding fundamental political reforms, while the current account deficit widened considerably to 9.7% of GDP, and the budget deficit reached 7.4% of GDP. 19 Noting Morocco’s economic and political turmoil, on 3 August 2012 the IMF agreed to the Kingdom’s request by approving a first Precautionary and Liquidity Line(PLL) 20 of SDR USD 4.1 billion. 21 This first arrangement was followed by a second for SDR 3.2 billion, 22 approved on 28 July 2014. The third PLL arrangement, for SDR of USD 2.5 billion, 23 was approved on 22 July 2016, and the fourth, for USD 2.97 billion, on 17 December 2018. Conditionalities in this series of PLL arrangements have given the IMF greater leverage, especially since Morocco is once again over-indebted. 17 In the 2013 Article IV Consultation, Directors“Executive Directors commended the economy’s resilience in the face of significant external shocks and challenging domestic conditions.” See International Monetary Fund Middle East and Central Asia Dept., IMF Executive Board Completes Third Review Under the PLL Arrangement with Morocco and Concludes 2013 Article IV Consultation, 06 March 2022, 2,[https://www.elibrary.imf.org/view/journals/002/2014/066/article-A002en.xml]. 18 The current account deficit rose to 4.3% of GDP in 2010 and to about 8% in 2011, despite strong exports and a significant increase in remittances and tourism receipts. See International Monetary Fund, IMF Report No. 12/239 – Morocco: Request for an Arrangement Under the Precautionary and Liquidity Line: Staff Report; Press Release on the Executive Board Discussion; and Statement by the Executive Director for Morocco., August 2012, 5,[https://www.imf. org/en/Publications/CR/Issues/2016/12/31/Morocco-Request-for-an-Arrangement-Under-the-Precautionary-and-LiquidityLine-Staff-Report-26184]. 19 International Monetary Fund, IMF Report No. 12/239, 4. 20 This is a sum calculated from the recipient’s quota that is made available in exchange for an annual repayment agreed with the IMF. According to the fund’s website,“The PLL provides financing to meet actual or potential balance of payments needs of countries with sound policies… It is intended to serve as a backstop or help resolve crises under wide-ranging situations. The PLL combines a qualification process(similar to that for the FCL, but with a lower bar) with focused e-post conditionality aimed at addressing the remaining vulnerabilities identified during the assessment of qualification.” A country’s PLL eligibility indicates the creditworthiness of its economy and the economic policies pursued by the authorities, which contributes“to consolidation of market confidence in the country’s policy plans.” See International Monetary Fund, La ligne de précaution et de liquidité(LPL) du FMI, 30 March 2016,[https://www.imf.org/fr/About/Factsheets/ Sheets/2016/08/01/20/45/Precautionary-and-Liquidity-Line]. 21 Approximately USD 6.2 billion at the time it was approved. 22 About USD 5 billion when it was approved. 23 About USD 3.5 billion when it was approved. 122 Uncovered: The role of the IMF in shrinking the social protection

A new constitution and an IMF arrangement The Kingdom responded to the challenges of the moment by activating two strategic levers: historic political reform and IMF assistance. Unlike other countries in the MENA region, the Kingdom did not remain inflexible. Rejecting protectionism, it chose a peaceful political response: 24 a constitutional reform that the IMF welcomed. 25 The IMF stated that Morocco had performed strongly in three of the five areas considered for PLL eligibility:“financial sector and supervision, monetary policy,[and] data adequacy” 26 and had recorded “moderate underperformance in the other two areas, fiscal policy, external situation and market access.” However, although the IMF noted 27 Morocco’s“persistent social and political risks and growing inequalities” with satisfaction, it enjoined Morocco to pursue its efforts of“budgetary rebalancing,” which required better control of public expenditures, especially those related to compensation. 28 Morocco’s fiscal policy was focused on reducing public spending, which included raising the administered price of oil. 29 The government also committed itself to rapidly adopting a more flexible exchange rate regime. Both measures directly impact inflation—broadly seen as a tax on the incomes of the poorest. The IMF also demanded the reduction of“non-essential” current expenditures to compensate for the increased wage bill of about 0.25% of GDP 30 “linked to the strengthening of the workforce in the priority areas of health, education, security and defence, and the full-year effect of the May 2011 salary increases.” 31 24 Contrary to the strategic choices made by most nations facing the Arab revolutions, Morocco decided in favor of a political solution to the crisis on February 20, 2011. The head of state entrusted the drafting of a new constitution to an ad hoc commission whose members were drawn from all political trends within Moroccan society. The result was a constitutional text that eased tensions. From a long-term perspective, this historic decision indicates a real capacity within government to anticipate its future and to evaluate the advantages/disadvantages of major political decisions. This capacity that would be deployed once again a few years later by introducing universal social protection, a new major reform of the reign. 25 The IMF welcomed the political reforms and underlined that“pressures persist for further reforms to address, among other issues, still-high youth unemployment, income inequalities, and governance.” See International Monetary Fund, IMF Report No. 12/239, 5. 26 Ibid., 20. 27 The IMF announced that the Moroccan economy was robust enough to be eligible for IMF support(PLL). See ibid., 25-27. 28 An assertion which may not always be confirmed: curiously, the HCP, which had stopped communicating all data relating to the Gini coefficient since 2014, abruptly changed its policy after the COVID-19 crisis. A decision to not publish an indicator as important as the Gini coefficient for a five-year period does not really correspond to what the IMF calls“data quality.” 29 Higher energy prices reduced the budget deficit by 0.6% of GDP in 2012. See International Monetary Fund, IMF Report No. 12/239, 18. It should be noted that subsidies on petroleum products were definitively abolished at the end of 2015 without causing social unrest. 30 The government pledged to not fill positions vacated by retirees, and to improve its performance standard. 31 In contrast, in its 2014 report, the IMF no longer distinguished between different ministerial departments and supported “limiting the net creation of new positions, salary advances, and promotions to a well-defined budget envelope, in order to reach their medium-term goal.” See International Monetary Fund, IMF Report 14/241 – Request for an arrangement under the arrangement under the precautionary and liquidity line and cancellation of the current arrangement, 5 August 2014, 11,[https://www.imf.org/en/Publications/CR/Issues/2016/12/31/Morocco-Request-for-an-Arrangement-Under-thePrecautionary-and-Liquidity-Line-and-41818]. Social spending in Morocco: The influence of IMF conditionality 123

However, the main structural reform heralded the changes to come, namely dismantling the compensation fund’s main prerogatives 32 and eliminating most oil subsidies, 33 which were to fall from 5.5% of GDP in 2012 to 1.2% of GDP in 2015. In return, a more equal redistributive system 34 targeting vulnerable populations was to be developed. The Moroccan government made clear that it was developing this particular policy commitment in close collaboration with the World Bank. However, persistent disparities did not prevent the IMF from supporting(or pressuring) the government, which, for its part, made no secret of its willingness to abolish subsidies for energy 35 and food. To drive its point home, the IMF cited an HCP study stating that about 43% of Morocco’s total food and energy subsidies 36 were captured by the top quintile of the population. 37 The IMF urged the government to put an end to a system that not only did not benefit the population but had also become an unbearable burden on the budget. Unfortunately, this approach overlooked a number of facts by viewing social protection solely through the prism of budgetary cost 38 and not as protection against the vagaries of life. It also omits the inflationary impact of dismantling subsidies, which hits the most disadvantaged hardest. 39 At the time, the World Bank estimated that“the impact would be three times higher for the poorest than for the richest” and admitted that“compensatory measures would be inevitable” to avoid the poverty rate rising. 40 32 Unlike the IMF, Moroccan social legislation does not include compensation within social protection. In its August 2012 report, the IMF stated:“Morocco’s current social protection system relies primarily on a generalized subsidy system, but also includes several more targeted initiatives.” See International Monetary Fund, IMF Report No. 12/239, 16. 33 In 2011, compensating petroleum products cost the equivalent of 5.1% of Moroccan GDP, at MAD 41.4 billion(EUR 3.7 billion), compared with almost nothing in 2003. See Pierre-Olivier Rouaud,”Réforme des subventions aux carburants au Maroc: les bons conseils du FMI”, L’Usine Maroc, 14 June 2013,[Réforme des subventions aux carburants au Maroc: les bons conseils du FMI(usinenouvelle.com)]. 34 In its letter of intent, the Moroccan government announced that, in the future, it would implement a program to reduce “inequalities in incomes and opportunities, both broadly and across regions and genders, and strengthening social service delivery and social protection.” See International Monetary Fund, IMF Report No. 12/239, 60. 35 According to the IMF,“[T]he authorities are considerably shielding the budget from risks linked to the volatility of international oil prices.” See International Monetary Fund, IMF Report 14/241, 11. 36 In 2014, the IMF still encouraged“progressively reduc[ing] subsidies on wheat and butane, the latter in part through measures aimed at substituting butane consumption in agriculture by the use of renewable energy”. Fortunately, none of the subsequent governments abolished the subsidies on butane and soft wheat, understanding that they are key to basic food security. See ibid., 11. 37 International Monetary Fund, IMF Country Report 12/239, p. 16(Box 2). 38 In 2014, the IMF continued to encourage“the possible progressive reduction of the butane subsidy, by introducing incentives for using renewable energy sources, as well as measures to gradually reduce the subsidy on soft wheat flour.” See International Monetary Fund, IMF Report 14/241, 48. 39 The most disadvantaged are very sensitive to food and transport costs. 40 Maude Chardaire,“Le Maroc, le pays qui a(presque) supprimé les subventions aux produits pétroliers”, Novethic, 09 November 2016,[Le Maroc, le pays qui a(presque) supprimé les subventions aux produits pétroliers(novethic.fr)] 124 Uncovered: The role of the IMF in shrinking the social protection

But the government of the time(a coalition led by the Justice and Development Party, PJD) 41 did not heed the warning; what it took away with one hand it did not offer with the other. Although the IMF supported programs that“better target the poor,”, such as RAMED and TAYSSIR, these schemes offered limited support for healthcare and education. They could not compensate for the loss of purchasing power caused by the end of energy and food subsidies. What has been achieved after a decade of PLL arrangements Despite dismantling most of its compensation system, Morocco never reached the IMF’s“indicative targets.” Back in 2014, during the second PLL arrangement, the IMF had sounded the alarm. 42 The public account had improved slightly in the mid-2010s, but only temporarily. After the deficit declined to 3.5% of GDP in 2017, 43 the government’s budgetary position improved. But the following year, it rose to 3.9% of GDP, and in 2020, during the COVID-19 crisis, it reached the unprecedented level of 7.6%. For structural reasons, Morocco’s debt and chronic state budget deficit could never be reduced. The overall public debt rate, already very high in 2019 at 80.4% of GDP, 44 rose to 94% in 2020. 45 This failure did not result in any questioning of government or IMF policies. On the contrary, in its 2014 report the IMF encouraged Morocco to stay the course:“The 2014 budget targets a continued prudent reduction in the deficit through structural fiscal reforms. The further reduction to 4.9 percent of GDP is to be achieved mainly through a continued reduction in the cost of energy subsidies while social protection for the most vulnerable is being expanded through dedicated programs.” 46 41 On 29 November 2011, the leader of the conservative, right-wing Islamist party Abdelilah Benkiran was appointed to head the government by royal decree. 42 In 2014, the IMF declared that“risks to the balance of payments from external shocks remain, justifying access under a PLL arrangement. Notwithstanding the reduction of vulnerabilities achieved under the 2012 PLL arrangement, the external and fiscal positions remain vulnerable to external shocks.” See International Monetary Fund, IMF Report 14/241, 20. 43 In 2014, the IMF continued to encourage“the possible progressive reduction of the butane subsidy, by introducing incentives for using renewable energy sources, as well as measures to gradually reduce the subsidy on soft wheat flour.” See International Monetary Fund, IMF Report 14/241, 48. 44 According to the IMF,“Between 2015 and 2017, the cyclically adjusted primary deficit would be almost halved from 2.6 percent to 1.4 percent of GDP and close to zero percent by 2019. Such a primary balance path is consistent with putting the public debt-to-GDP ratio on a downward path after its expected peak in 2015 so it is gradually reduced closer to 60 percent of GDP by 2019.” See International Monetary Fund, IMF Report 14/241, 10. 45“Dette publique au Maroc: Que penser du niveau de dette au Maroc?”, Centre Marocain de Conjoncture, 16 March 2021,[Dette publique au Maroc: Que penser du niveau de dette du Maroc?(cmconjoncture.com)]. 46 International Monetary Fund, IMF Report 14/241, 18. Social spending in Morocco: The influence of IMF conditionality 125

In so doing, the IMF continued to advocate standard neoliberal tenets, describing subsidies as almost entirely responsible for the twin budget and current account deficits. 47 This contradicts its reports that explicitly or implicitly highlighted the poor performance of public investment and Morocco’s nondiversified economy, factors that contributed as much, if not more, to the problem of twin deficits, through public debt and deficiencies in the national current account. The IMF continued to support targeted assistance programs. However, these programs, by their very nature, did not fulfil Moroccan expectations regarding social justice and protection. Not until 2020 was the 2012 ambition to create a more equitable redistributive system finally placed on the agenda. In April 2020, events forced the Moroccan government to draw on funds available under the PLL. Plunged into an economic depression by the COVID-19 crisis, the state had no choice but to borrow nearly USD 3 billion, repayable over five years with a three-year grace period. To obtain this loan, the economy and finance minister urgently passed Decree-Law No. 2.20.320 on exceeding the external borrowing ceiling. Besides responding to the economy’s structural weaknesses, the current public debt overhang, and social inequalities, the IMF gets help from Moroccans to continue its control and surveillance. • The consolidation of accounts and mobilization of public savings have not put an end to IMF control. Morocco’s central bank(Bank Al-Maghrib) and the Ministry of Finance and Economy play leading roles in supervising the IMF’s structural and quantitative guidelines. With rare exceptions, these two bodies systematically adopt neoliberal economic, fiscal and monetary policies inspired by the Washington Consensus. Notably, in 2012, the request to the IMF for the PLL was signed by the governor of the Central Bank and the economy and finance minister of the day. • Morocco is subject to continuous surveillance in the form of annual consultations under Article IV of the IMF Articles of Agreement. Each year, IMF teams examine the economic policies the Moroccan government has implemented to“consolidate macroeconomic stability,” as well as ongoing systemic reforms. Bank Al-Maghrib and the Ministry of Finance and Economy are the IMF’s main interlocutors during its annual surveillance missions. In almost all their annual reports, they agree with the thrust of the IMF’s assessments. IMF reports are further scrutinized by creditors, rating agencies, multilateral organizations and the Kingdom’s main partners, giving the IMF immense leverage that very few countries could counter. 47 In 2016, the IMF“commended the authorities for their prudent macroeconomic management and the ongoing improvements to the policy framework which have helped yield positive outcomes, including strengthened fiscal and external buffers, a recovery of growth, and progress on unemployment and poverty reduction.” See International Monetary Fund, IMF Report No. 16/35- Morocco: 2015 Article IV Consultation-Press Release; Staff Report; and Statement by the Executive Director for Morocco, 08 February 2016, 2,[https://www.elibrary.imf.org/view/journals/002/2016/035/ article-A000-en.xml]. 126 Uncovered: The role of the IMF in shrinking the social protection

III. Social protection in Morocco: Between universality and econometric targeting Social protection in Morocco Article 31 of the Moroccan constitution of 2011 stipulates:“The State, the public establishments and the territorial collectivities work for the mobilization of all the means to facilitate the equal access of the citizens[female and male] to conditions that permit their enjoyment of the right: to healthcare, to social protection, to medical coverage and to the mutual or organized joint and several liability of the State.” 48 However, to arrive at a quantitative view of social protection in Morocco and a meaningful analysis of the systems involved, it is helpful to summarize the various sectoral strategies and budgets dedicated to social protection. This study argues that the definition of social protection should cover all institutional mechanisms, public or private, taking the form of a collective welfare system and/or implementing a principle of social solidarity, which cover the costs resulting for individuals or households from the existence of a certain number of identified social risks(illness, old age, unemployment, poverty, education, or occupational accidents). It implies paying benefits to individuals or households facing one or more of these risks to compensate for both the increase in the beneficiaries’ costs and the decrease in their resources. The Moroccan model was largely inspired by the French system of social protection schemes, particularly the intersecting nature of benefits(unemployment, old age, illness, education, occupational accidents, etc.) and the budgetary and accounting approach to social protection and national health accounts. However, Morocco’s statistical system and monitoring of data related to these sectors is less developed. The influence of targeting techniques can be seen in the system for accessing these benefits, which operates via a series of eligibility criteria that are regularly re-evaluated in light of Morocco’s international funding; see the section below on the systems for targeting and allocating benefits. 48 For more information on the creation, development and institutions of social protection in the Kingdom, see Appendix B. Social spending in Morocco: The influence of IMF conditionality 127

Nonetheless, receiving social benefits and access to social protection is regarded as a human right, in terms of ensuring income security, reducing poverty and inequality, and promoting social inclusion and dignity. 49 It is therefore the state’s responsibility to generalize, or universalize, social protection and make it effective through well-designed economic, social and fiscal policies. Social protection is also an essential component of the Sustainable Development Goals(SDGs), which the entire international community has committed to achieving by 2030. The Moroccan system has two main components(see Figure 1): 50 • A contributory component of social security including the various pension funds(the Moroccan Pension Fund, CMR; the Collective Pension Scheme, RCAR; the National Pension and Insurance Fund, CNRA; the National Social Security Fund, CNSS; and the Moroccan Health Insurance Fund, CMAM, formerly the National Fund of Social Welfare Organizations, CNOPS). • A non-contributory publicly funded component made up of a set of social assistance programs in the form of conditional cash transfers that link the social protection schemes to a number of conditionalities to be fulfilled by the target population(INDH, RAMED, TAYSSIR, school canteens, boarding schools, scholarships, vocational integration programs, social assistance institutions, etc.) Other commodity and fuel price support programs have been deployed in the name of poverty alleviation but fall outside the scope of social protection. Many of these programs were designed and implemented in the late 1980s when the Moroccan government adopted a more liberal economic model under the conditionalities imposed by the International Bank for Reconstruction and Development(World Bank, WB), the IMF, the Paris Club and others. 49 In 1944, the ILO’s Declaration of Philadelphia recognized social security as a fundamental human right. This was followed by the Universal Declaration of Human Rights in 1948 and the International Covenant on Economic, Social and Cultural Rights in 1966. 50 For more details on its history and governance, see Appendix B. 128 Uncovered: The role of the IMF in shrinking the social protection

Figure 1. The Moroccan social protection system Contributive systems Moroccan social protection system Transfer programs (noncontributive) AMO Pension schemes AT/MP Conventional scheme MRE RAMED Aid for education Aid for employment INDH Aid for countering violence against women Rural aid This diagram allows us to depict the structure of the Moroccan social protection system. It includes social protection schemes as well as social assistance programs in various fields, including education, poverty reduction and women’s inclusion(the list is not exhaustive). This heterogeneous and inefficient system covers just a small fraction of the population—millions of citizens have no protection(see Figure 2). Social spending in Morocco: The influence of IMF conditionality 129

Figure 2. Public spending on social assistance programs in Morocco(% of GDP) All social assistance Cash transfers Conditional cash transfers Fee waivers 2016 1.1 In-kind Other social assistance Public works School feeding Social pension 2016 0.7 2016 0.1 2016 0.1 2016 0.1 2016 0.0 Morocco Morocco Morocco Morocco Morocco Morocco Morocco Morocco Morocco Source: World Bank(2016). 51 According to the World Bank, Morocco’s system of social assistance(aside from compensation funds) lacks a coherent strategy, a simple institutional framework, or clear priorities for targeting a few core programs covering the main risks in each phase of life. A recent United Nations International Children’s Emergency Fund(UNICEF) study identified 140 different schemes, many of which are small and focused on niche target groups. Some 50 government agencies(ministries, departments, institutions and commissions) are responsible for program governance, implementation and funding. Yet despite this impressive number of programs and institutions, important gaps remain in the network of social protection, such as for out-of-school children and people with disabilities. Several program mandates overlap and duplicate similar measures, for example, by supporting the same population sub-groups. 52 51 World Bank Group, Aspire Data – Morocco, 2016,[https://www.worldbank.org/en/data/datatopics/aspire/country/ morocco]. 52 International Monetary Fund. Middle East and Central Asia Dept, Morocco: 2012 Article IV Consultation and First Review Under the Two-Year Precautionary and Liquidity Line, 2013, 23[https://www.elibrary.imf.org/view/journals/002/2013/096/ article-A001-en.xml]. 130 Uncovered: The role of the IMF in shrinking the social protection

The SDGs and ILO universal coverage Social protection ensures that individuals and households can access healthcare and have secure incomes, especially in the event of old age, sickness, disability, occupational injury, maternity or the loss of a breadwinner. 53 Since the late 19th century and the creation of welfare states, there has been a shift from individual responsibility to collective solidarity. Since the 1970s, all welfare states have been faced with very strong growth in healthcare, unemployment and pension spending—whether they follow the Bismarck 54 or the Beveridge 55 models—and have ratified the International Labour Organization(ILO) conventions that constitute the main references for international social security law. These include two particularly important normative references. • The Social Security(Minimum Standards) Convention, 1952(No. 102) specifies the minimum level of social security benefits and the conditions for their provision, as well as the nine main branches in which protection is guaranteed. • The Social Protection Floors Recommendation, 2012(No. 202), adopted 60 years later, provides general guidance for extending and complementing social security systems by establishing national social protection floors accessible to all persons in need. With this in mind, Morocco had ratified 65 ILO conventions as of November 14, 2020, 56 of which 49 conventions are in force, 11 were denounced and 3 repealed; 5 instruments were repealed; none had been ratified in the previous 12 months. Morocco has not yet ratified 41 ILO technical conventions relating to social protection. On June 14, 2019, Morocco began the ratification procedure for the Migration for Employment Convention, 1949(No. 97), the Social Security Convention, 1952(No. 102) and the International Agreement on the Promotional Framework for Occupational Safety and Health, 2006(No. 187), a treaty that encourages member states to establish a national occupational safety and health policy and program. Morocco’s behavior reflects its desire to implement the Kingdom’s international commitments respecting labor standards. Morocco’s commitment to ratifying ILO conventions related to social protection is summarized in the following table. 53 International Labour Organization,“A universal challenge: Social security for the world” World of Work, no. 49(01 December 2003): 11-13,[https://www.ilo.org/wcmsp5/groups/public/---dgreports/---dcomm/documents/publication/ dwcms_080606.pdf]. 54 Work-related insurance. 55 Protection based on collective solidarity, irrespective of any professional activity. 56“NORMLEX- Ratifications for Morocco, 2020”, International Labour Organization, Last accessed 14 November 2020, [https://www.ilo.org/dyn/normlex/fr/f?p=NORMLEXPUB:11200:0::NO::P11200_COUNTRY_ID:102993]. Social spending in Morocco: The influence of IMF conditionality 131

Table 2. Conventions on work ratified by Morocco Field All fields Medical care Health Unemployment Old age Disability Survivors’ benefits Employment injury Maternity Equality of treatment Safeguarding rights Source: Authors’ compilation. Convention 102 130 168 128 121 183 118 157 Ratified or not Yes No No No No Yes No No This paragraph also recalls that Morocco’s commitment to achieving the SDGs also includes strengthening social protection by building sustainable, inclusive and universal systems and in particular, meeting the objectives of SDG 1, SDG 3, SDG 4, SDG 5, SDG 8 and SDG 10. 57 Table 3. Themes of the SDGs relevant to social protection Sustainable Development Goal SDG 1 SDG 3 SDG 4 SDG 5 SDG 8 SDG 10 Source: Authors’ compilation. Aim End poverty everywhere Good health and well-being Quality education for all Gender equality Promote decent work Reduce inequality 57“Sustainable Development Goals- 17 Goals to Transform Our World”, United Nations, Last accessed May 2022,[https:// www.un.org/sustainabledevelopment/]. 132 Uncovered: The role of the IMF in shrinking the social protection

How IMF conditionality impacts education and healthcare spending In general, IMF conditions have created long-term effects that cause“scarring.” 58 To illustrate how conditions impact social spending, we have isolated two significant budget items from the state’s social policy: education and healthcare. We integrate public spending on education and health with the two pillars of the universal subsidy and argue that increasing it will help Morocco achieve SDG 3 and SDG 4. Such spending is crucial to compensating for income inequalities and facilitating social mobility. All the PLL arrangements between the IMF and Morocco have noted the relationship between unemployment and education. The IMF promotes better education to meet the needs of employers faced with poorly trained personnel. Yet under its supervision, there has been no improvement in Morocco’s fiscal space, which is a sine qua non for real improvement in public education. 1. Education In 2018, Thomas Stubbs, Bernhard Reinsberg, Alexander Kentikelenis and Lawrence King published a landmark study on IMF conditionality and its effects on the volume of social spending. 59 Their findings confirmed the conclusions of all previous studies on the subject. 60 That is, conditions negatively impact educational spending, 61 and each additional spending condition corresponds to a decrease of at least 0.47% of GDP in spending on education. We can extrapolate how Morocco’s education budget would have developed had it not been deprived of valuable funding: in Figure 3, we used budget figures from finance laws over the last decade 62 to plot a graph with two parallel lines, one showing real spending and the other spending without IMF (and World Bank) conditions. 58 Coined in early 2020 by three IMF researchers to refer to a permanent loss of productive capacity, the term has since caught on. 59 Thomas Stubbs et al.,“How to evaluate the effects of IMF conditionality”, The Review of International Organizations, no. 15(December 2018), 29–73,[https://link.springer.com/content/pdf/10.1007/s11558-018-9332-5.pdf]. 60 With the notable exception of those commissioned by the IMF. 61 IMF experts disagree with studies showing that IMF conditions have a negative effect–for example, in“What Happens to Social Spending in IMF-Supported Programs?” the authors write:“[E]ducation and health spending has risen during IMF-supported programs at a faster pace than in developing countries as a whole. Over a five-year period in low-income program countries, education spending as a share of GDP increases by about 0.8 percentage point of GDP, and for health, about 1 percentage point of GDP. For other developing countries, the effect is estimated to be neutral.“ See Benedict Clements, Sanjeev Gupta and Masahiro Nozaki, IMF Staff Discussion Note 11/15- What Happens to Social Spending in IMF-Supported Programs? International Monetary Fund, 31 August 2011, 16,[https://www.imf.org/en/Publications/StaffDiscussion-Notes/Issues/2016/12/31/What-Happens-to-Social-Spending-in-IMF-Supported-Programs-25190]. 62 From 2011 to 2020. Social spending in Morocco: The influence of IMF conditionality 133

Figure 3. The impact of structural and quantitative conditions on Morocco’s educational budget 8.0 7.5 7.0 6.5 98.3 6.0 6.3 5.9 5.5 5.0 2012 GDP in current prices(billions of USD) Public spending on education(% of GDP) GDP surplus in absence of conditions 110.1 103.3 5.8 5.6 5.5 5.3 2013 2014 2015 2016 2017 118.1 5.7 5.3 2018 2019 125 120 112.2 115 6.9 110 6.4 105 100 95 90 85 80 2020 Source: World Bank, Ministry of Economy and Finance of the Kingdom of Morocco. 63 Figure 3 illustrates the impact of the quantitative and structural conditionalities of the IMF(and the World Bank) on budget allocations to education. While in relative terms the figure of 0.47% may seem modest, in absolute terms it represents between MAD 4 billion and 5 billion, or from 7% to 9% of the annual budget for public education, vocational training and higher education. Moreover, the IMF’s quantitative conditions impose budgetary arbitrations correlated with the growth rate. Admittedly, this rule is not strictly applied in the case of education, but the growth of GDP and educational spending often converge with a lag of one or two years, as illustrated in Figure 4. 63 Government expenditure on education by budget year http://lof.finances.gov.ma/fr/budget/47/64; GDP in current USD, https://data.worldbank.org/country/MA. 134 Uncovered: The role of the IMF in shrinking the social protection

Figure 4. Morocco’s GDP growth compared with education expenditure Percent % 15 10 5 0 -3.1 -5 -4.8 -10 2012 2013 GDP growth rate Growth rate of pub. expend. on edu 5.8 2.1 3.0 0.6 2014 2015 2016 2017 10.8 8.1 2018 Source: Authors’ calculations from World Bank data. 64 2019 8.7 -6.3 2020 Although the two lines are not perfectly aligned, they nevertheless contradict the official discourse, which insists that the budget for public education, vocational training, higher education and scientific research is sacrosanct. However, the approach based on the budgetary variable, the effectiveness of which depends closely on the quality of public governance, needs to be qualified, as the emergency program of 2009-2012 shows. Although the state made a one-off effort, spending MAD 12 billion to upgrade the education system, 65 the results were disappointing. A report by the Court of Audits published in 2018 describes an unprecedented fiasco and waste of resources. 66 The report’s authors identified a multitude of shortcomings: • A persistently high rate of school dropouts, estimated at more than 270,000 students per year; • Overcrowded classrooms at all levels; • Suboptimized use of schools; • The use of schools without minimum facilities; • A persistent teacher surplus and shortages; • Limited effectiveness of social support programs; • An unreliable information system. 64 We have calculated the growth rates of the education budget and GDP using figures extracted from the morasses of the Finance Acts, and from World Bank data: https://data.worldbank.org/country/MA; http://lof.finances.gov.ma/fr/budget/47/64. 65 The Moroccan government called on no less than five donors to finance its emergency plan: the European Union, the European Investment Bank, the French Development Agency, the African Development Bank and the World Bank. The strategy’s resounding failure was coupled with an increase in debt equivalent to 1% of GDP. 66 Cour des Comptes du Royaume de Maroc, Rapport relatif à l’évaluation du programme d’urgence-Ministère de l’éducation nationale-, May 2018,[http://www.courdescomptes.ma/upload/MoDUle_3/File_3_613.pdf]. Social spending in Morocco: The influence of IMF conditionality 135

Even today the education system suffers from the same ills:“Cum hoc ergo propter hoc.” 67 Because the same causes produce the same effects, Morocco is far from achieving quality education, as advocated by the SDGs. 68 For example, the International Association for the Evaluation of Educational Achievement’s 2020 report ranks Morocco among the bottom five countries in the world in mathematics and science. 69 Moreover, the Moroccan education system is now one of the most unequal in the world, 70 with three types of schooling: • A generally poor public education system that favors quantity over quality; • Private education that is only slightly better than its public counterpart but capitalizes on middle class parents’ fears that their children will experience social regression; • Private education, often foreign, that is inaccessible to the lower and middle classes. 2. Healthcare Since independence, healthcare in Morocco has been the poor relation of public policy, and Health Department resources are well below the minimum required. Even if we add the special funds of SEGMA 71 and the INDH, 72 and spending by the various welfare and insurance schemes, as well as household grants, to the ministerial budget, the average share of GDP devoted to this crucial sector remains below 5%—see Figure 5 below. This level is close to the average of other lower-middleincome countries(5.6% of GDP), but lower than the average of upper-middle-income countries (6.7% of GDP), high-income countries(7.8% of GDP), or nearby countries like Tunisia(6.9% of GDP). 67“With this, therefore because of this.” 68 United Nations, Éducation de qualité: Pourquoi est-ce important?, 2016,[https://www.un.org/sustainabledevelopment/ fr/wp-content/uploads/sites/4/2016/10/Why_it_matters_Goal_4_French.pdf]. 69 TELQUEL, N°932, 18 December 2020, 14. 70 Jean-Pierre Chauffour, Country Economic Memorandum: Morocco2040- Emerging by Investing in Intangible Capital, World Bank Group, 2017, 230,[https://openknowledge.worldbank.org/bitstream/handle/10986/28442/9781464810664. pdf?sequence=2&isAllowed=y]. 71 Autonomously run state services. 72 Launched in 2005, the National Initiative for Human Development(INDH) is run by the Ministry of the Interior. Since 2018, it has funded maternal and child health programs. 136 Uncovered: The role of the IMF in shrinking the social protection

Figure 5. Healthcare is the poor relation of public policy in Morocco 7.0 6.0 5.7 5.7 5.3 5.1 5.2 5.2 5.3 5.0 Tax % 4.0 Pub. expend. on health in all sectors as% 3.0 Pub. expend. on health as% GDP in current USD 1.9 1.6 1.5 1.4 1.7 2.0 1.3 1.3 1.3 1.0 0.0 2012 2013 2014 2015 2016 2017 2018 2019 Sources: World Bank, Ministry of Economy and Finance of the Kingdom of Morocco. 73 , 74 2.0 2020 Moroccan healthcare financing is regressive and inequitable Direct household contributions are the primary source of financing for the healthcare sector. According to the most recent Ministry of Health statistics(from 2015), more than 51% of the health sector is financed by direct household payments, 24% by social security contributions, 22% by fiscal resources, less than 1% by other private household spending, and 1% by international cooperation. Compared to the average for middle-income countries, Morocco is much more dependent on direct payments by households for health services. 73 Regarding the data on private and public sector health expenditure as% of GDP, see World Bank Group, Current health expenditure(% of GDP) – Morocco, Last modified 30 January 2022,[https://data.worldbank.org/indicator/SH.XPD.CHEX. GD.ZS?locations=MA]. Regarding the data on government expenditure on health by fiscal year, see Ministry of Economy and Finance,[…][http://lof.finances.gov.ma/fr/budget/47/64]. Regarding the data on GDP in current USD, see World Bank Group, Morocco – Overview, 2022,[https://data.worldbank.org/country/MA]. It is important to specify that these are provisional budgets. It is indeed impossible to derive the actual achievements from the documents of the Ministry of Economy and Finance. The Settlement Acts only provide the reader with aggregated figures. To facilitate our work, we had no choice but to rely on the morasses. Despite extensive research, we were unable to access summary tables showing health expenditure over several years, and we had to extract the data document by document. The most recent figures are often approximate and drowned in a narrative inflation that complicates the analysis. We extracted the other data from the information cited above. 74 For information on the population of Morocco, see High Commission for Planning, Population du Maroc par année civile (en milliers et au milieu de l’année) par milieu de résidence: 1960 – 2050, Last accessed 16 May 2022,[https://www.hcp. ma/Population-du-Maroc-par-annee-civile-en-milliers-et-au-milieu-de-l-annee-par-milieu-de-residence-1960-2050_a677. html]. Data on per capita expenditure was calculated from the Ministry of Finance and Economy and the World Bank. Social spending in Morocco: The influence of IMF conditionality 137

SDG 3 To attain SDG 3 by 2030, the Moroccan government should increase healthcare spending by 2.5%(to 8.2% of GDP), of which 2.2% would be public funds. That would increase per capita healthcare expenditure from USD 178 in 2017 to USD 419 in 2030, allowing for a 2.6-fold increase in doctors and a 3.6-fold increase in medical staff(assuming human resource health policies remain unchanged). This amounts to MAD 107 billion(MAD 7.7 billion annually), including MAD 77 billion in public spending(MAD 5.5 billion per year). Source: OECD 2020.75 Targeting and social benefit allocation 75 In 2012, the IMF described Morocco’s“generalized” or universal subsidy system for flour, sugar and petroleum products such as gasoline, butane and diesel. This system cost the state budget the equivalent of 6.1% of GDP. Energy subsidies accounted for 84% of the total subsidy bill. 76 The chart depicting the Moroccan social protection system showed a diversified, underfunded system that has gradually been liberalized by switching from benefits linked to employment(contributory social insurance programs) or collective, cultural and religious solidarity(non-contributory social assistance programs) to various linked targeting systems that control access to benefits. This analysis brings us back to observing how the Moroccan state’s conception of social measures and their place in public policy has evolved alongside its relationship with international financial institutions, notably the IMF and the World Bank. Targeting in Morocco In the mid-1990s, the Bretton Woods institutions began to promote means-tested targeting of households. 77 Means-testing is presented as the only reliable and effective method for making conditional resource transfers in developing countries—countries that have been funded by these institutions, especially to help implement social policies. 75 OECD, Mobilisation des recettes fiscales pour le financement de la santé au Maroc, 2020, 13,[www.oecd.org/fr/fiscalite/ politiques-fiscales/mobilisation-des-recettes-fiscales-pour-lefinancement-de-la-sante-au-maroc.htm.] 76 International Monetary Fund. Middle East and Central Asia Dept. Morocco: 2012 Article IV Consultation and First Review Under the Two-Year Precautionary and Liquidity Line, 2013,[https://www.elibrary.imf.org/view/journals/002/2013/096/ article-A001-en.xml]. 77 It is important to remember that targeting is the whole process(methodology, tool, information system) that makes it possible to identify entities in a given population(individuals/households/areas) presenting certain common characteristics. If resources are limited, targeting aims to optimize the distribution/redistribution of resources by ensuring that aid reaches those who need it most first. 138 Uncovered: The role of the IMF in shrinking the social protection

Targeting is not new in Morocco; in fact, it has been the subject of controversy related to the most appropriate methods to reach targeted populations or priority areas for social policy and programs. The most telling examples of the consequences of targeting in Morocco are RAMED and TAYSSIR, which are part of a large program of targeted free health care, first and foremost education. RAMED RAMED, or the Medical Assistance Plan, is based on the principles of social assistance and national solidarity and aims to benefit poor and vulnerable persons who are not covered by compulsory health insurance. A conditional assistance program, it uses econometric targeting to provide the poorest people with basic health coverage. RAMED was officially launched by Morocco’s prime minister in November 2008 in a pilot project supported by the EU, the World Bank, the IMF and the African Development Bank(AfDB) in what was then known as the Tadla-Azilal region(including the provinces of Béni Mélal, Azilal and Fquih Bensalah). The program targeted an estimated 420,000 people in an area with a population of 1.5 million. Following the pilot phase, RAMED’s regulatory and procedural provisions for eligibility were adjusted in 2010 to target the population and their access to health services at public hospitals and health institutions, as well as government-run health services. On 13 March 2012 in Casablanca, His Majesty King Mohammed VI approved the scaling up of RAMED to all regions. RAMED is expected to reach about 8 million people, of whom 2.7 million are impoverished and 5.3 million are vulnerable, according to poverty and vulnerability rates established by the HCP in 2007. RAMED thus became the first public program in Morocco to use econometric targeting—the direct targeting of households on the basis of their socioeconomic conditions, rather than geographical or community criteria—to identify poor and vulnerable households. However, the method proved problematic during implementation. TAYSSIR With the TAYSSIR program, launched in 2008, the state’s methodology evolved from geographic targeting to individual household targeting. According to the Ministry of National Education, eligible children enrolled in TAYSSIR receive a monthly allowance from their parents to help them stay in school. The program is targeted by district. Rural communes must have a poverty rate above 30% and a school-dropout rate above 8%. Social spending in Morocco: The influence of IMF conditionality 139

In 2016, TAYSSIR switched from geographically targeting, by district, to targeting individual households. The change in method was implemented in two phases, as detailed below. • In the short term, the program budget was optimized by harmonizing its database with that of the school system, to improve overall functionality and guarantee reliable data. In addition, the beneficiary database had to be finetuned to eliminate those who were ineligible: salaried households with stable incomes, assets, or family allowances, etc. This required tallying and cross-checking the program data with the information systems of other partners and relevant bodies, in particular the Centre for Social Security(CNSS), the National Treatment Center(CNT), the Moroccan Interprofessional Retirement Fund(CIMR), the National Pension and Insurance Fund(CNRA) and the Directorate General of Taxes(DGI). • Plans for the near future include setting up an RSU based on a package of the RAMED and TAYSSIR programs as well as providing direct aid to widows. The social registry should provide a reliable, comprehensive and unified source for identifying each social program’s potential target population. Targeting methods in Morocco: Geographical targeting, especially with poverty maps that have been used to implement several aid programs. Income-level targeting(to locate vulnerability and poverty) used by RAMED. Geographical and categorical targeting adopted for TAYSSIR. In its first review in 2012 of the PLL arrangement, IMF experts wrote:“The government aims to introduce a more efficient system by gradually replacing universal subsidies with well-targeted assistance, building on the existing framework, including the National Initiative for Human Development, and the targeted pilot programs‘RAMED’ and‘Tayssir.’” 78 “The next steps[should] involve better identification of vulnerable groups and developing a consensus on[IMF] reform modalities and programs.” 79 In the medium term, the IMF recommends lowering the costs of the system of subsidies and social protection, while ensuring better targeting, in a four-pillared strategy: • Reducing the cost of energy subsidies; • Creating a solidarity fund financed by taxes on tobacco, alcohol and public land purchases, along with higher taxes on top incomes, including for luxury cars; 80 78 International Monetary Fund, IMF Report No. 12/239, 16. 79 Ibid., 16. 80 Ibid., 13-14. 140 Uncovered: The role of the IMF in shrinking the social protection

• Using a fully targeted social protection scheme(in consultation with the World Bank); 81 • Having a hedging mechanism against oil price volatility so price increases are borne by households – limiting the cost of the subsidy. According to the World Bank, by 2016, the cost of the remaining subsidies(on food and LPG) fell to about 1.4% of GDP, while spending on non-subsidy social safety net programs went up to about 0.4% of GDP. 82 With social spending around 1.8% of GDP, Morocco is close to the worldwide median level. 83 In 2018, vulnerable populations saw their purchasing power impacted by the elimination of food and energy subsidies, yet 75% of the poorest remained outside the scope of RAMED. That forced the government to improve its targeting.“[The authorities] also noted that the ongoing regionalization is an opportunity to better coordinate and adapt national policies and programs to local needs, address the still large urban–rural gap in poverty rates.” 84 The IMF welcomed the decline in public spending to 3% of GDP through canceling universal subsidies. This sum has never been re-allocated to targeted grants, however, which account for just 0.4% of GDP. 85 Econometric targeting The results of the various targeting methods used in Morocco and other countries reveal their wide margin of error and exclusion. This is particularly likely to happen when the targeting criteria are too inclusive in relation to available resources and are poorly weighted or not rigorously defined. Experience also shows that commissions frequently had to make discretionary decisions in the final identification of beneficiaries, which increases the risk that certain groups or potential beneficiaries will be excluded. Evaluations of schemes like RAMED and TAYSSIR have shown that it is difficult to identify each poor household on the basis of declaratory statements alone, even if they are legitimized by local authorities and administrations. The crisis caused by the COVID-19 pandemic has confirmed these conclusions. The government cast a wide net to compensate those severely impacted by the crisis, but its targeting was imprecise. As the interior minister admitted to Parliament last summer, many ineligible people benefited from aid, while those targeted did not receive it. 81 Ibid., 14. 82 Ibid., 24. 83 Ibid., 24. 84 International Monetary Fund. Middle East and Central Asia Dept., Morocco: 2017 Article IV Consultation-Press Release; Staff Report; and Statement by the Executive Director for Morocco, 2018, 20,[https://www.imf.org/en/Publications/CR/ Issues/2018/03/12/Morocco-2017-Article-IV-Consultation-Press-Release-Staff-Report-and-Statement-by-the-45713]. 85 Ibid., 24. Social spending in Morocco: The influence of IMF conditionality 141

On 10 July 2020, the National Human Rights Council(CNDH) submitted to both houses of Parliament recommendations and comments on Draft Law 72-18 regarding the system for targeting beneficiaries of social support programs and the establishment of the RSU. The document, copies of which were also sent to the prime minister and interior minister, included 12 recommendations and observations. The CNDH also stressed the need to ensure that targeting serves the challenge of building an inclusive national social protection scheme that achieves equality, rejects discrimination and limits the possibilities of excluding individuals and groups. The CNDH also emphasized the urgency of developing a targeting system that considers the new strategy development model and can adapt to changes that might affect national policy on social support, not only in terms of strengthening program convergence and integration, but also with respect to transforming support programs into an integrated system in sectoral public policies. The Court of Auditors’ Annual Report for the years 2016-2017 highlighted the fund’s weaknesses and poor governance in terms of supporting social cohesion, exposing its lack of a comprehensive strategy for implementing social support programs and incoherent programming of resources and expenditures. The report also criticized the governance of the TAYSSIR targeting method, support for people with disabilities, RAMED’s implementation and the“One Million School Bags” program. Acknowledging these failures in his Throne Day speech on July 29, 2018, His Majesty King Mohammed VI launched a universal targeting mechanism based on the RSU, which will use data from the new National Population Registry(RNP) and be administered by the National Registry Agency(ANR). From this point on, the programs would not define a given target using techniques that could vary from one program to another. Instead, a single population-targeting system would be developed to help social programs use their criteria to locate beneficiaries. The RSU being rolled out uses advanced identification technologies with a more accurate scoring system. The Moroccan government does not anticipate its implementation before 2023. The IMF saluted the reform:“The decision to harmonize all current social assistance programs into a single family-allowance scheme starting from 2023 is a step in the right direction. The transfers will be targeted based on the unified social registry(which is expected to become operational in 2022) and funded through the reallocation of current transfers and the gradual elimination of the gas subsidy.” 86 86 International Monetary Fund. Middle East and Central Asia Dept., Morocco: 2020 Article IV Consultation-Press Release; Staff Report; and Statement by the Executive Director for Morocco, 2021, 17,[https://www.imf.org/en/Publications/CR/ Issues/2021/01/04/Morocco-2020-Article-IV-Consultation-Press-Release-Staff-Report-and-Statement-by-the-49987]. 142 Uncovered: The role of the IMF in shrinking the social protection

Instead of developing and making two complementary systems more effective, under the aegis of the IMF and the World Bank, targeted subsidies will replace universal subsidies. IV. The COVID-19 crisis: Opportunities and obstacles to social protection Importance of strong social safety nets in combating insecurity The IMF stressed that Morocco was on the right track before the COVID-19 pandemic broke out and plunged the country into one of the most serious crises in its modern history. However, official figures contradict that assessment, showing that, on the eve of the pandemic, and after almost 10 years of IMF programs and measures, Morocco was already suffering both economically and socially. According to the independent constitutional institution created by His Majesty King Mohammed VI, “Two thirds of the working population(60%) are not covered by any pension scheme and almost half(46%) of the working population are not covered by health insurance.” 87 Almost no Moroccan workers—except for a minority of employees in the formal private sector—have occupational hazard insurance. Moreover, the country has no specific social protection scheme for children, the unemployed or persons with disabilities. Despite a multitude of schemes and assistance programs (see Figure 1 and Appendix B), Morocco’s social protection system faces multiple risks and challenges to both social security and social assistance. These include inefficient governance, demographic factors, uncertain funding and the importance attached to the design and generalization of public social protection policies. In coming decades, social protection will be influenced by a number of factors. These factors are explored below. By 2050, Morocco’s population pyramid will be reversed Demography and labor market dynamics have always affected social systems, with governments responding by making parametric adjustments or far-reaching systemic reforms. Changes in Morocco’s demographic structure have directly impacted social policies, particularly health insurance schemes (payments and consumption patterns) and pension schemes(increasing the retirement age, changing 87 Economic, Social and Environmental Council(CESE), La protection sociale au Maroc: Revue, bilan et renforcement des systèmes de sécurité et d’assistance sociales, n° 34/2018, 2018, 14,[https://www.almendron.com/tribuna/wp-content/ uploads/2020/04/av-as34-vf.pdf]. Social spending in Morocco: The influence of IMF conditionality 143

the method of calculating pensions, etc.). Demography is an essential reference for defining primary needs, such as education, health, employment and social security schemes. It is one of the major trends determining a country’s economic, social and cultural development. Policy guidelines, social protection and governance Public action on social assistance is marked by low equity due to the programs’ excessive fragmentation and lack of coordination. As a result, people who are not in need are overprotected, while needy people find themselves excluded from social assistance nets. The socioeconomic crisis unleashed by the COVID-19 pandemic illustrates the crucial role of social protection, and especially the impact that the lack of social safety nets has on economic dynamics and vulnerable populations who have slipped into precariousness. During the crisis, Moroccan decisionmakers concluded that a population that has been made precarious is a threat to the already fragile social peace. The Moroccan government clearly needs to develop public social protection systems and mechanisms, based on solidarity and proportionate to its resources, for persons and groups not yet covered: the informal sector and the self-employed; unsalaried workers; members of the liberal professions; cooperative workers; caregivers, especially women in rural areas; single mothers; children, including deprived children(“street children,” working children and children in prison); persons with disabilities; the elderly(Article 10 of the Covenant); persons living in poverty and social exclusion(Article 11 of the Covenant); victims of occupational accidents and diseases; and the unemployed. Developing a new social protection policy will henceforth be a pillar of public growth policy and an essential component of any new development efforts. But the success of this huge project to transform the Moroccan model will depend on the government coalition’s negotiating power vis-àvis powerful multilateral organizations. The COVID-19 pandemic’s impact on Morocco’s socioeconomic model In 2020, Morocco faced unprecedented economic and social upheaval amid the COVID-19 pandemic. Not only did the health emergency devastate whole economic sectors(tourism, the aeronautics industry, pharmaceutical laboratories, etc.), it was compounded by an environmental crisis, with a severe drought that slashed agricultural production. The dual crisis caused purchasing power to plummet and led to massive unemployment, a chain of bankruptcies and the first recession since 1995. According to the HCP and the World Bank, the growth rate was-7%, although this was also due to the collapse of exports in the wake of the disruptions to global value chains and the fall in 144 Uncovered: The role of the IMF in shrinking the social protection

tourism revenues. A World Bank report summarized the effects:“The crisis is having a severe impact on jobs and household incomes, generating a spike in unemployment and a deterioration of poverty and vulnerability indicators.” 88 The report estimated that, between September 2019 and September 2020, almost 581,000 people lost their jobs, nearly half of them in agriculture.(It is worth reiterating that the poverty rate is already higher in rural areas). Because of the lockdown, during the second quarter of the year, the number of hours worked per person per week dropped from 45 before the outbreak of COVID-19 to 22, and 62% of workers suffered income losses.“In this context, poverty and vulnerability rates are expected to spike,” noted the World Bank report. 89 Nonetheless, assessing the socioeconomic impact of the COVID-19 crisis in Morocco is a complex task, given the country’s still relatively weak statistical capacity. 90 A devastating crisis for public finances The first months of the COVID-19 crisis hit the government budget hard, with slippage in public accounts at least as drastic as after the 2008 crisis. At that time, however, the state was much less indebted. This time, already in dire straits, it was forced to accept more debt in order to support the economy, fight the pandemic, meet its expenses, provide aid and assistance to the population, and plug the loopholes in the tax system. The accumulation of crises culminated in a sharp decline in tax revenues for the fiscal year 2020, 91 with public debt jumping 12 points to 76% of GDP for the treasury alone, and overall debt at 92%. A high social cost The pandemic and lockdowns in Morocco and its main international partners plunged the Moroccan economy into an unprecedented recession. Internal and external demand and supply slumped dramatically, causing a domino effect that led thousands of businesses in the formal and informal sectors to cease operations. 92 In the third quarter of 2020, the economy hemorrhaged 344,000 rural 88 World Bank Group, Morocco Economic Monitor: From Relief to Recovery, September 2020, ix,[https://documents1. worldbank.org/curated/en/365441608752111119/pdf/Morocco-Economic-Monitor-From-Relief-to-Recovery.pdf]. 89 Ibid., 4-5. 90 Ibid., 5. 91 This was not true for corporate tax, which is based on the previous fiscal year. Revenues from VAT and the information and communications technology(ICT) sector that are closely linked to national consumption fell by more than 10%. Income tax revenue fell by 5% due to the numerous redundancy plans. Registration and tax stamps, which are linked to real estate, financial and land transactions, also experienced more than 20% slippage. In the end, the state lost MAD 17.4 billion in earnings, including a 7.7% drop(MAD 13.4 billion) in tax revenues. 92 The crisis has also shown the limitations of the economic model and led to the creation of the Mohammed VI Investment Fund. The equivalent of MAD 45 billion, or 4.5% of national GDP, will be made available to the fund, one third from the general budget and the remaining MAD 30 billion raised from public and private institutional investors. Multilateral organizations such as the IMF, the AfDB, the World Bank and the European Bank for Reconstruction and Development (EBRD) may be asked to contribute. The fund will operate as a legal entity in six areas: industrial restructuring, innovation, small and medium-sized enterprises(SMEs), public-private partnership(PPP) infrastructure, agriculture and tourism. Each sector will have its own thematic fund managed by specialized companies selected by tender. Social spending in Morocco: The influence of IMF conditionality 145

jobs and 237,000 urban jobs. Within a few months, the pandemic, combined with the drought, erased 20 years of job creation, returning the country to its 1999 figures. The unemployment rate jumped from 9.9% to 12.7%—up 3.3% compared with the same period of 2019. The social costs of the pandemic and the drought have been high: nearly 1.5 million Moroccans are unemployed, with precarious workers, women and young people the most exposed. This is a very high figure given the country’s low activity rates, 93 ineffective social safety nets for the unemployed, and the high percentage of non-monetized agricultural jobs. COVID-19 and the Moroccan socioeconomic model The effects of the pandemic have impacted the most vulnerable populations strongly, with millions of vulnerable Moroccans catapulted into absolute poverty. Whereas a year earlier, 1.7% of the population was in absolute poverty, 94 in the wake of the pandemic, 11.7% of the population, or 4.2 million Moroccans, 95 were made indigent by depriving them of resources to meet their basic food and non-food needs. At the same time, the vulnerability rate more than doubled, from 7.3% before the lockdown to 16.7% during. In urban areas, the vulnerability rate jumped from a pre-lockdown rate of 4.5% to 14.6% during the lockdown, and in rural areas, from 11.9% to 20.2%. 96 Overnight, almost a third of the population, or 11.2 million Moroccans, moved into precarity. Social inequalities worsened, surpassing the socially intolerable threshold: the Gini coefficient reached 44.4%, up from 38.5% before the pandemic. This major phenomenon— a direct consequence of the lack of robust social safety nets—has exposed the flaws of the Moroccan model. Its“social minimum” model is a legacy of the IMF, which favors economic growth at the expense of social justice. As a social crisis loomed, Morocco adopted emergency measures Faced with this social crisis and its imminent political risks, the Moroccan government has reacted by adopting urgent measures and resorting to more debt. For the first time, Morocco used the funds earmarked under the 2018 PLL arrangement, estimated at nearly USD 3 billion, or 3% of GDP, repayable from 2025. This is in addition to other external and domestic debt. 93 At the same time, the crisis has permanently excluded thousands of Moroccans from the labor market. They have abandoned all hope of finding a job, de facto ceasing to belong to the labor force and be counted as part of it. The activity rate has fallen from 44.9% to 43.5% as many people have stopped looking for work. 94 According to the HCP,“absolute monetary poverty” implies non-satisfaction of, or the satisfaction of only the minimum, basic requirements for the maintenance of the individual. 95 High Commission for Planning, Evolution du niveau de vie des ménages et impact de la pandémie COVID-19 sur les inégalités sociales, 2020, 5,[https://www.hcp.ma/file/221006/]. 96 Ibid., 5. 146 Uncovered: The role of the IMF in shrinking the social protection

The World Bank stresses that Morocco’s stock of external debt increased by 19% in 2020, and that the country is the second largest borrower in the MENA region after Egypt. 97 Morocco has also resorted to domestic debt through banks and public institutions, which together covered some 70% of Morocco’s financial needs in 2020. 98 CNSS staff on sick leave benefited from a MAD 2,000 allowance(around USD 220) from the IMF, 99 along with support from the national basic health scheme(AMO) and family allowances. A total of 800,000 employees received a total of MAD 2 billion from the Coronavirus Pandemic Fund. For recipients of RAMED, the Economic Watch Committee decided to support households. 100 Those with two or fewer persons received a lump sum of MAD 800. Households of four or more persons received a monthly amount of MAD 1,200. Another 4.3 million families whose members work in the informal sector but are not affiliated with RAMED received compensation similar to that received by RAMED beneficiaries. The state’s rapid and large-scale distribution of public aid helped mitigate the effects of the pandemic confinement on household living standards. The absolute poverty rate was reduced by 9% nationally, from 11.7% before the transfer of public aid to 2.5% afterwards. 101 This was a decline from 7.1% to 1.4% in urban areas and from 19.8% to 4.5% in rural areas. The Gini coefficient dropped from 44.4% without transfers to 38.4% after receiving public aid. 102 Morocco was the second most generous country in terms of aid distributed to these citizens, as shown in Figure 6. However, two figures illustrate the vulnerability of a workforce faced with very high unemployment and minimal unemployment benefits. Some 1.5 million people in the labor force are unemployed, with only a few more than 13,000 receiving or having received benefits at or below the minimum wage. 103 Knowing that employment slumped from 41.2% 104 to 37.9%, 105 it is obvious that many vulnerable families have fallen into absolute poverty. 97 World Bank, International Debt Statistics 2022, 2022, 115,[https://openknowledge.worldbank.org/bitstream/ handle/10986/36289/9781464818004.pdf]. 98 Jeta Menkulasi, Cesar Serra, and Suchanan Tambunlertchai“Four Questions About Debt and Financing Risks from Covid in the Middle East and North Africa”, IMF News, 28 April 2021,[https://www.imf.org/en/News/Articles/2021/04/27/ na042821-4-questions-about-debt-and-financing-risks-from-covid-19-in-mena]. 99 To help them avoid personal bankruptcy, indebted employees were also granted a four-month grace period in the form of an exceptional and temporary suspension of bank loan repayments until the end of June. This amounts to a rescheduling, since loan repayments have been postponed until after the original deadline. 100 Medical assistance plan. 101 High Commission for Planning, Evolution du niveau de vie des ménages, 6. 102 Ibid., 6. 103 The minimum wage is MAD 2,638(around EUR 240). 104 High Commission for Planning, Principaux indicateurs du marché de travail relatifs à l’activité, à l’emploi et au chômage au titre du quatrième trimestre 2019, 06 February 2020, 2,[https://www.hcp.ma/downloads/Activite-emploi-et-chomagePrincipaux-indicateurs_t21667.html]. 105 High Commission for Planning, Evolution des indicateurs du marché de travail au titre du troisième trimestre de 2020, November 2020, 3,[https://www.hcp.ma/Evolution-des-indicateurs-du-marche-de-travail-au-titre-du-troisieme-trimestrede-2020_a2621.html]. Social spending in Morocco: The influence of IMF conditionality 147

Figure 6. Segments of the workforce that benefited from government aid in Morocco and other countries(%) 70 60 56 50 40 Jordan February 2021 Morocco November 2020 56 Morocco February 2021 Tunisia November 2020 Tunisia February 2021 Egypt February 2021 42 41 34 34 30 24 22 23 24 20 20 20 15 12 10 8 5 17 12 645 2 12 10 0 14 11 9 54 0 Farmer Employer/ Public wage Regular formal Regular informal self employed priv. wage priv. wage i/s of est. 70 60 58 50 46 50 52 48 40 35 31 32 29 29 29 30 26 26 20 20 22 23 21 22 20 17 15 18 18 18 16 14 14 12 11 10 7 0 Regular informal Irregular wage Unemployed Out of Labor Force priv. wage o/s of est. Total Source: Caroline Krafft, Ragui Assaad, and Mohamed Ali Marouani, based on data from the Economic Research Forum(2021). 106 106 Caroline Krafft, Ragui Assaad, and Mohamed Ali Marouani,“The Impact of COVID-19 on Middle Eastern and North African Labor Markets: Glimmers of Progress but Persistent Problems for Vulnerable Workers a Year into the Pandemic”, ERF Policy Brief, No. 57(June 2021): 17,[https://erf.org.eg/publications/the-impact-of-covid-19-on-middle-eastern-and-north-african-labormarkets-glimmers-of-progress-but-persistent-problems-for-vulnerable-workers-a-year-into-the-pandemic/] 148 Uncovered: The role of the IMF in shrinking the social protection

The COVID-19 crisis may produce a new social protection system The fact that a large number of Moroccans very quickly fell into precarity shows the limitations of emergency measures, the temporary nature of which cannot provide a lasting response to their needs. The COVID-19 pandemic has revealed a deficient health system, victim to decades of neglect, looting and disinvestment. The same applies to social protection, which remains inadequate and reserved for a tiny minority of the population. We are currently witnessing a sharp and widespread rise in inequalities and an explosion in poverty rates. This situation demonstrates the need for sustainable social solidarity. His Majesty King Mohammed VI has ordered social protection be generalized within five years. Four projects have been launched: AMO, pensions, family allowances and job-loss compensation. Work on AMO and RAMED was slated to begin in 2021 and to take two years. The implementation of the two programs will be accelerated. Universal health coverage is to be extended to all Moroccans, at the same level and with the same criteria, over two years(2021–2023). The scheme is to be financed by taxes and social security contributions, with benefits no longer dependent on wage status. RAMED will be replaced by the expanded AMO, with costs for new entrants covered by taxes. In the second phase(2024–2025), to benefit the active population, old-age and unemployment risks will be covered by retirement pension benefits and job-loss compensation. To guarantee the success of this systemic reform and to prevent financial losses, the government has promulgated a law on a national population register, which covers the whole population, including minors and newborns, as well as foreigners who are settled in Morocco. The registry will facilitate the collection, registration and storage of identity, demographic and biometric data, as well as online and real-time authentication based on a Digital Civil and Social Identifier(DCSI). Even if we disregard the inequalities and misery created by the neoliberal order, a state that allocates so few resources to social spending and so many resources to servicing its debt 107 is torpedoing its economic competitiveness. Social spending significantly impacts production, with a 1% growth in social expenditure adding about 0.1% to GDP—representing a multiplier of about 0.6. During severe economic downturns, the effect is larger. The effect is comparable between countries with low and high debt-to-GDP ratios, and countries with high and low levels of trade openness. In addition, spending on education, healthcare and unemployment benefits have the greatest impact on activity. Investment is not attracted solely by variables such as wages or taxes; an entrepreneur needs welltrained, healthy personnel who are able to adapt to innovations. 108 107 According to the 2021 Finance Law, almost MAD 78 billion will be allocated to debt repayment—nearly double the national budget for education. 108 The entrepreneur also needs an independent and efficient judiciary, which is far from the case in Morocco. No liberal system can function without a functional and honest judicial regulator. Yet the IMF never mentions this institution, which is so crucial for a market economy. Social spending in Morocco: The influence of IMF conditionality 149

Apparently recognizing the limits of the IMF and World Bank model, the Moroccan authorities have launched a huge project to provide universal social protection(see below). However, financing this reform is a real challenge at a time when public accounts have deteriorated and state debt has reached new heights, aggravated by the current drought. The problem is structural: year after year, tax revenues cover only about 60% of total expenditure. With the COVID-19 emergency, the ratio has dropped to 53%. The plan to extend universal social protection(2021–2025) Framework Law 9.21 is the basis for generalizing social protection, creating a hybrid system combining insurance and assistance, contributory and non-contributory. Its objective is to protect Moroccans from a wide range of hazards that can occur over a lifetime. The law aims to extend healthcare to all Moroccans within two years(2021–2023), with the same level and criteria. Financed by both taxes and social security contributions, benefits will be disconnected from wage status. Medical assistance(RAMED) will be replaced by the extension of compulsory health insurance(AMO), with costs related to new entrants covered by taxes. In the second phase(2024–2025), the risks of old age and unemployment will be covered by pension benefits and compensation for loss of employment for the working population. To ensure the success of this systemic reform and prevent financial losses, the government promulgated a law on the National Population Register that covers the Moroccan population, including minors and newborns, as well as foreigners settled in Morocco. This registry will facilitate the collection, registration and storage of identity, demographic and biometric data, as well as online and real-time authentication based on a Digital Civil and Social Identifier(DCSI). In the medium term, this extension of social protection will make it possible to fulfil the conditions for creating a comprehensive, universal and solidarity-based national social security system, providing dignity and inclusion for all, as well as social cohesion and economic development. Starting in January 2021, it will gradually extend compulsory health insurance and family allowances to all households, as well as retirement pensions and job-loss compensation(IPE) for the labor force that is not yet covered. This project of societal and state transformation is hampered by a tight budget, largely caused by the Moroccan economy’s stop-and-go activity as a result of the COVID-19 pandemic. However, the state plans to allocate an unprecedented amount of public policy funds: 22 million additional beneficiaries in 2021 and 2022 7 million school-age children in 2023 and 2024 5 million working people in 2025 Access to job-loss compensation in 2025 MAD 14 billion for generalizing compulsory health insurance MAD 20 billion for family allocations MAD 16 billion for pensions MAD 1 billion for job-loss compensation 150 Uncovered: The role of the IMF in shrinking the social protection

V. Recommendations A – Develop the debate on the relationship between the financial management of public budgets and social policy design using a three-step process: • Initiate new research by connecting academics and practitioners; • Raise awareness among public authorities and other national institutions of the importance of updating databases(financial, sectoral, etc.) and making them available to the various research circles; • Structure the debate between research circles and the IMF and World Bank teams, and also include OECD and UN bodies, notably the ILO. B – Governance of social policies: • Should we move toward regionalizing public policies of a social nature? • Support trade unions to effectively intervene(upstream) in the public policy decisionmaking process, particularly social policy. C – Educate trade unions and other civil society actors about the economic requirements for managing the state budget: • Initiate targeted training for committed actors; • Make documentation available to all stakeholders(popularize key concepts); • Help these actors design their own advocacy. D – Encourage exchanges between civil society and academic circles within the MENA region. The implementation of these recommendations can be structured around three pillars. Pillar 1 Conceptualisation Pillar 2 Empowerment Pillar 3 Networking Social spending in Morocco: The influence of IMF conditionality 151

Appendices Appendix A. The IMF and the World Bank’s restructuring of the Moroccan economy The birth, development and deficiencies of a neoliberal socioeconomic model Morocco 1964: the first IMF loan Newly independent Morocco wanted to catch up with the countries of the North and escape the rut of underdevelopment. But massive capital flight and the departure of thousands of economic operators in the decade after independence forced the state to take the place of the faltering private sector. 109 The government viewed a voluntarist policy influenced by Keynes as its best option, and opted for a public deficit and its corollary, indebtedness, believing they could trigger the much hoped-for“take-off.” Michel Labonne described this budgetary voluntarism as accompanied by strong interventionism connected to a voluntarist view of development involving substantial public investment in infrastructure. 110 But difficulties arose in the first plan(1960-1964) due to excessively ambitious objectives 111 and the low level of national savings available, hindering agricultural and industrial development projects. 112 Morocco turned to the IMF and contracted its first loan of USD 1.3 million in July 1964. The same scenario was repeated with the three-year plan of 1965–1967 and the five-year plan of 1968–1972. None of the objectives were achieved. It was impossible to overcome the intrinsic weaknesses 113 of a national economy marked by growing unemployment, very insufficient rates of 109 In 1968, private investment only represented 29% of national investment. See Abdeljalil Agourram and Abdel-Aziz Belal, “Bilan de l’économie marocaine depuis l’indépendance“, Bulletin économique et social du Maroc 32, no. 116(1970): 150, [http://www.abhatoo.net.ma/maalama-textuelle/developpement-economique-et-social/developpement-economique/ reperes-du-developpement-economique/politique-economique/l-economie-marocaine-depuis-l-independance]. 110 Michel Labonne,“Ajustement structurel au Maroc: le secteur agricole en transition?”, In: Les agricultures maghrébines à l’aube de l’an 2000., ed. by Mahmoud Allaya,(Montpellier: CIHEAM Options Méditerranéennes: Série B. Etudes et Recherches n. 14, 1995), 297-305,[https://om.ciheam.org/om/pdf/b14/CI960057.pdf]. 111 Jamil Salmi, Planification sans développement. Evaluation de l’expérience marocaine de planification 1960-1979. (Casablanca: Les éditions maghrébines, 1979). 112 M. Saâdi, Le processus d’industrialisation par substitution d’importations au Maroc,(Mémoire DES, University of Casablanca, Faculté de droit, 1976). 113 Aside from 1965, from 1960 to 1966 Morocco experienced quasi-stagnation or very slow growth in overall production. From 1957 to 1966, the overall growth rate was about 2%, well below the population growth rate of 3%. See Abdeljalil Agourram and Abdel-Aziz Belal,“Bilan de l’économie marocaine depuis l’indépendance“, 149. 152 Uncovered: The role of the IMF in shrinking the social protection

saving 114 and investment, as well as a very tricky balance of payments. Following this double failure, with a still anemic private sector, the young state was forced to finance most of its investments through external debt. This was especially necessary because there were also many post-independence social emergencies in the realms of education and healthcare. The turning point in 1983: The IMF and the IBRD impose the neoliberal order In 1983, in the grip of falling phosphate prices, drought, the war in the Sahara and the oil crisis, and the Moroccan state was unable to honor its commitments and decided to appeal 115 to the IMF and the International Bank for Reconstruction and Development(IBRD). 116 It was granted breathing space in the form of debt rescheduling 117 but in return had to commit to very strict structural fiscal consolidation. The new approach was not limited to consolidating public accounts—it aimed to transform the economy, state and even Moroccan society. The IMF and IBRD hold sway over Morocco The IMF and IBRD’s shock therapy produced a new economic(and societal) neoliberal model coordinated by a state with reduced powers, whose main function would henceforth be regulating economic, social and political policies. The government began by cutting spending on healthcare, 118 education and support for basic consumer goods. At the same time, it promoted globalization and dismantled customs protections. It also developed the financial market and liberalized the foreign exchange market while depreciating the national currency. 119 The reform package produced fairly positive macroeconomic effects, with an average 5% GDP growth between 1985 and 1990 and reduced internal and external imbalances. However, the results of the austerity policy and its corollary—the shrinking of the scope of public services—were mixed. In Morocco, IMF conditionality has reshaped the social, institutional and market structure and still guides major budgetary decisions, with the IMF acting as the channel for implementation of the neoliberal ideas of the Washington Consensus in Morocco. Conditions helped the IMF and later 114 Public savings financed only a modest part of the country’s investments—one third in 1968. See Abdeljalil Agourram and Abdel-Aziz Belal,“Bilan de l’économie marocaine depuis l’indépendance“, 165. 115 Authors’ note: this turning point is still a painful moment in the collective memory of the generation of Moroccans who lived through and suffered the most serious crisis of the contemporary Moroccan state. 116 This is a very sensitive issue: the Treaty of Fez(1912) was imposed on Morocco because of the country’s indebtedness, and to implement fiscal and administrative reforms. 117 This was the first of a long series, with several reschedulings between 1983 and 1992. 118 The structural adjustment plan reduced public expenditure on health, except for the University Hospital Centres(CHUs) in Rabat and Casablanca. Public hospitals, which had provided 90% of care in 1980 only provided 66% in 1987. Worse still, hospitals were operating at 50% capacity due to the unavailability of medical equipment. See Jean-François Clément,“Les effets sociaux du programme d’ajustement structurel marocain“, In: Politique étrangère 60, no. 4, 1005. 119 The MAD was overvalued by 25%. Social spending in Morocco: The influence of IMF conditionality 153

the World Bank forge a new socioeconomic model for more than 36 million Moroccans. Over 38 years, systemic change has been made gradually, with the country reshaped through a series of adjustments, strung together by a string of quantitative and structural conditionalities. But the social balance sheet of the system’s backbone, the state, has been mixed, to say the least, as it has clearly failed to significantly reduce inequalities and unemployment. From external to domestic debt: Waning international donors Following painful structural adjustments, the issue of foreign debt became central, with the government realizing that external debt can threaten national sovereignty. To avoid that misfortune, the authorities opted for domestic debt—so much so that, in 1997, for the first time in modern Morocco, local savings were more in demand than foreign savings. Public debt management was facilitated by the emergence of a public debt market deep enough to meet the needs of the treasury. This approach allowed the state to switch from external financing making up 80% of the treasury’s total debt in 1984 to it representing only 21.6% of total external debt stock in 2019. External debt gradually gave way to domestic debt, which rose from 20% of the total outstanding debt in 1984 to 78% in 2019. 120 Switching to the domestic market eased the grip of international donors(bilateral and multilateral) who became much more discreet. Today, they only finance major projects. 121 This strategic choice was not random: in the eyes of public authorities, it offered(and still offers) real political and economic advantages: 1. The loss of power of bilateral and multilateral international donors after 1992 made it possible to avoid a possible“diktat” putting the country under IMF tutelage again. 2. Bank Al-Maghrib sets the key rates that constitute the reference for the domestic market. This Moroccan institution has kept the average cost of the debt reasonable, with domestic debt of 4.29% in 2019 against 2.6% of external debt. 122 3. The low cost of external debt should not obscure exchange rate risks, or sudden spikes in the spread in the event of creditor panic. The historically low levels are mainly due to the abundance of liquidity that the world’s leading central banks have injected into the international financial system since 2012. However, this cyclical situation can change very quickly should there be a sudden downturn in the economic situation. In any case, it depends on decision-makers that Moroccan authorities cannot influence. 120 For the first time since 2008, the treasury’s outstanding debt fell in late 2018 by 0.3% of GDP, from 65.2%(51.8% for domestic and 13.4% for foreign debt) to 64.9%(50.9% for domestic and 14.0% for foreign debt) in late 2019. 121 The brutal depression caused by the COVID-19 crisis forced the government to appeal to the IMF for the PLL mechanism. 122 Ministry of Economy and Finance of the Kingdom of Morocco, Projet de Loi de Finances pour l’année budgétaire 2022 – Rapport sur la dette publique, 2021, 40,[https://www.finances.gov.ma/Publication/db/2022/08-%20Rapport%20 Dette%20publique_Fr.pdf]. 154 Uncovered: The role of the IMF in shrinking the social protection

4. If there is a debt crisis, the public authorities could rekindle inflation and reduce the value of public debt. They could also block possible financial recovery programs by raising the level of the tax burden and/or cutting spending. 123 5. The cost of domestic debt remains relatively stable, and the government has access to abundant savings. Moroccan institutions are even finding it difficult to invest in international markets and are forced to acquire a certain volume of treasury bonds for prudential reasons. 6. Rating agencies, whose word often determines the cost of debt, do not accord as much importance to domestic debt as to foreign debt. Appendix B. The history of social protection in Morocco In Morocco, a large number of welfare institutions derive from tribal customs and are based on the precepts of Islam. Religion and the tribe establish the rules and procedures for collecting and redistributing community wealth. This is evident in post-independence legislation: institutions for administering social benefits are governed by financial and sustainability rules, but entitlement to benefits is intimately linked to socio-professional status, employability, and codes largely inspired by the Islamic model(the family code, rules of inheritance, etc.). However, the modern Moroccan social security system began during the protectorate, with the introduction of the first modern system offering social insurance, exclusively for French workers posted to Morocco. It included a system of mutual societies to provide compensation for accidents at work, and to cover medical care. In the under-developed private sector, similar initiatives were rather timidly developed, including the well-known Moroccan Interprofessional Retirement Fund(CIMR), created in 1949. This explains why Morocco’s social protection has evolved into a highly heterogeneous system with inadequate coverage rates and unequal access to coverage of social risks, both in terms of benefits (the system does not cover all social security benefits) and population(the schemes are based on socioprofessional status, covering private sector and semi-public employees and civil servants, but there is no operational coverage for self-employed workers, although there is one for the“vulnerable”). 123 To stabilize(or reduce) the public debt(D) as a percentage of GDP(Y), the primary balance(i.e., excluding interest) in points of GDP must be equal to(or greater than) a“stabilizing primary balance.” This(SPB) is equal to the product of the debt by the difference between its apparent interest rate i(ratio of the interest burden to the stock of debt at the end of the previous year) and the rate of growth in value of GDP g(see sheet on the stabilizing balance). Hence the formula: SPB / Y=(i – g) x D/ Y. See Fançois Ecalle,“Les notes d’analyse II) Déficit et dette publics, politique budgétaire 3) L’inflation est-elle une solution pour alléger la dette publique?“, FIPECO, 23 April 2022,[https://www.fipeco.fr/fiche/Linflation-estelle-une-solution-pour-r%C3%A9duire-le-co%C3%BBt-de-la-dette-publique-%3F]. Social spending in Morocco: The influence of IMF conditionality 155

It should also be pointed out that Moroccan social protection includes support for needy populations, along with other assistance mechanisms such as literacy training, national mutual aid, and above all compensation in the form of price subsidies for basic necessities, namely“national flour,” sugar and butane gas. Social protection in Morocco: Structure, governance and parameters Over time, Morocco has built up a large collection of compulsory and optional social programs and nets to support its economic development model. In addition to contributory schemes(health insurance, pensions, industrial accident and job-loss compensation, unemployment insurance, etc.), social protection also includes all other forms of social assistance as well as certain government antipoverty programs such as INDH, the human development initiative. The Moroccan system has two main components: i) contributory social security comprising the various pension funds(CMR, RCAR, CNRA), the national social security fund(CNSS) and national health insurance fund(CMAM); and ii) funded non-contributory social assistance programs in the form of conditional cash transfers that link social protection programs to numerous conditions that must be satisfied by the target population. These include INDH, RAMED, TAYSSIR, school canteens, boarding schools, scholarships, work integration programs, social assistance institutions, and others. All these actors provide their clients with the various social benefits offered by the various social assistance schemes and programs. All public and private actors administer some or all of the social benefits affecting healthcare, pensions, work accidents and occupational diseases, job loss, poverty, education, and so forth. These include: • Compulsory health insurance; • Retirement pensions; • Disability pensions; • Family allowances and other maternity benefits; • Conditional cash transfer programs(RAMED, TAYSSIR, etc.); • Job-loss compensation. The various social protection actors are summarized here. 156 Uncovered: The role of the IMF in shrinking the social protection

Figure B.1 Morocco’s social protection system in detail Organismes publiques et institutions nationales: - Autorité de contrôle des assurances et de la prévoyance sociale(ACAPS, ex DAPS). - Caisse Marocaine des Retraites (CMR). - Caisse Nationale de Sécurité Sociale(CNSS) - Caisse Marocaine d’Assurance Maladie(CMAM, ex CNOPS). - Agence Nationale de l’Assurance Maladie(ANAM). - Agence de Développement Social(ADS). - Entraide Nationale(EN). - Conseil Economique et Social et Environnemental(CESE). - Haut-Commissariat au Plan (HCP). - Conseil National des Droits de l’Homme(CNDH). - Caisse de Dépôt et de Gestion (CDG). - Fondation Hassan 2. - Régime Collectif d’Allocation de Retraites(RCAR). - Caisse Nationale de Retraites et d’Assurances(CNRA). - Initiative Nationale pour le Développement Humain(INDH). - Fondation Mohammed VI de Promotion des Œuvres Sociales de l’Education-Formation. Société Civile: - Association Marocaine des Droits Humains. - Organisation Marocaine des Droits Humains. - Association Marocaine de Solidarité Et de Développement (AMSD). - Centre d’Etudes en Droits Humains et Démocratie(CEDHD). - Forum des Alternatives Maroc (FAMS). - Espace Associatif. - Collectif marocain pour la promotion des droits des personnes en situation de handicap(CHM). - Association Démocratique des Femmes du Maroc(ADFM). - Ligue Marocaine pour la Protection de l’Enfance. - Association Nationale de Lutte contre le Sida(ALCS). PS Compagnies Privées d’Assurances: - AYA ASSURANCES. - SANAD. - RMA WATANIYA. - SAHAM ASSURANCES. - WAFA ASSURANCE. Parlement: - La Commission des secteurs sociaux. - La Commission des Finances et du Développement Economique. Syndicats et représentants des employeurs: - Union Marocaine du Travail (UMT). - Confédération Démocratique du Travail(CDT). - Union Nationale du Travail au Maroc(UNTM). - Union Général des travailleurs du Maroc(UGTM). - La Confédération Générale des Entreprises au Maroc(CGEM). Ministères: - Le ministère de l’Economie et des Finances. - Le Ministère de l’Education Nationale et de la Formation Professionnelle. - Ministère de l’Emploi et des Affaires Sociales. - Ministère de la Santé. - Ministère de la Solidarité, de la Femme, de la Famille et du Développement Social. - Ministère de l’Artisanat, de l’Economie Sociale et Solidarité. - Délégation Interministérielle aux Droits de l’Homme. Acteurs Universitaires et Centres de Recherche: - Chaire Prévoyance et Retraites, de l’Université Internationale de Rabat. - Association Marocaine des Sciences Economiques(AMSE). - Centre d’études et des recherches démographiques. Social spending in Morocco: The influence of IMF conditionality 157

There are other commodities and fuel price support programs for combating poverty that fall outside the scope of social protection. Many of these programs were conceived and launched in the late 1980s, when Morocco was transitioning to a new, more liberal economic model, following the terms of its external financing and indebtedness, especially to the IBRD and the IMF. Appendix C. IMF conditionality The IMF dominates global economics and plays a critical role in the long-term development of low- and middle-income countries. Its founding charter allows it to provide temporary financing to countries with imbalances in their balance of payments in return for“adequate guarantees” specified by the Fund. In exchange, the IMF monitors the countries’ economic policies through a system of conditions. Generally speaking, the Washington Consensus is the neoliberal conceptual framework that structures IMF conditions. 124 These are based on three assumptions: policies should be marketfriendly, outward-looking, and macroeconomically stable. A typical IMF policy package consists of 10 policy prescriptions: fiscal discipline, reduction of general subsidies for financing human resources and infrastructure, tax reform, financial liberalization, unified and competitive exchange rates, trade liberalization, the removal of barriers to foreign direct investment, privatization, deregulation, and property rights. Any country that fails to comply with conditionality will be automatically deprived of IMF loans and its access to development aid and international capital markets will be limited as a result. This is why conditions are such crucial instruments. The different types of conditions According to the IMF website, approval of an arrangement or a program review is based on a variety of economic policy commitments agreed with the national authorities. The commitments can take different forms: prior actions 125 are actions the country agrees to undertake before the IMF Executive Board approves financing or completes a program review, because they are viewed as essential to achieving the program’s objectives or monitoring its implementation. They ensure that the program is headed in the right direction or will be put back on track if it deviates from agreed policies. They may relate, for example, to eliminating price controls or adopting a budget law consistent with the program’s budgetary framework. 124 John Williamson,“What Washington Means by Policy Reform”, in: Latin American Adjustment: How Much Has Happened?, ed. John Williamson,(Washington DC: Institute for International Economics, 1990), 7-20. 125 Socially, these are the toughest conditions. The IMF is loath to announce an arrangement with a country until certain measures are implemented, such as the devaluation of its currency or higher energy prices. In the case of Morocco, the PLL arrangements between 2012 and 2020 were not linked to prior actions(at least they were not made public). 158 Uncovered: The role of the IMF in shrinking the social protection

Quantitative performance criteria are specific, measurable conditions, macroeconomic variables that the country can control like certain monetary and credit aggregates, international reserves, fiscal balances and external borrowing. For example, a program may foresee a minimum level of net international reserves, a cap on the central bank’s net domestic assets, or a ceiling on government borrowing. Indicative targets can be set in addition to quantitative performance criteria to assess a country’s progress toward a given program’s objectives. This is sometimes done when quantitative performance criteria cannot be established because of statistical uncertainties about economic trends(e.g., what will happen in the months to come). After the uncertainties have been removed, the targets normally become quantitative performance criteria with the necessary modifications. Structural benchmarks are reform measures, often non-quantifiable, that are essential for achieving program objectives. They serve as markers for assessing the program’s implementation during reviews. They vary from program to program and may, for example, take the form of initiatives to improve the functioning of the financial sector, or to strengthen the social security system or the management of public finances. If a quantitative performance criterion is not met, the IMF Executive Board may approve a formal waiver to allow the review to be completed, provided it is satisfied that the program will be properly implemented—either because the deviation is minor or temporary, or because the national authorities have taken or will take corrective action. Unmet structural benchmarks and indicative targets do not require waivers and are instead assessed as part of the program’s overall performance. The IMF’s Monitoring of Fund Arrangements(MONA) database is in the public domain and covers all aspects of program conditionality. Why IMF conditions are elastic The nature and number of conditions depend on the country’s economic situation, as well as on a much more prosaic factor—the number of countries the IMF is supporting at the time. The more countries, the scarcer the resources, and the more stringent the conditions. The opposite phenomenon also occurs. 126 When the IMF has plenty of liquidity, it is more generous and recruits more borrowers, expecting to earn interest on loans. 127 This strategy validates criticism of an institution that seeks to maximize revenues, protect future budgets and maintain a position of global power. 128 126 Valentin Lang,“The economics of the democratic deficit: The effect of IMF programs on inequality”, Discussion Paper Series, no. 617(September 2016),[https://www.econstor.eu/handle/10419/162960]. 127 Sarah Babb and Ariel Buira,“Mission creep, mission push and discretion: The case of IMF conditionality”, in: The IMF and World Bank at sixty, ed. Ariel Buira,(London: Anthem Press, 2005), 59-83. 128 Axel Dreher and Roland Vaubel,“The Causes and Consequences of IMF Conditionality”, Emerging markets finance& trade 40, No. 3(2004): 26–54,[https://doi.org/10.1080/1540496X.2004.11052571]. Social spending in Morocco: The influence of IMF conditionality 159

Conditionality is linked to political priorities: The case of Morocco 129 Another factor that explains IMF generosity is geopolitical in nature. 130 Countries aligned with the United States tend to be treated more favorably by the IMF and may benefit from fewer conditions. Similarly, major IMF shareholders may advocate a softer conditionality in exchange for political influence on the UN Security Council. In addition, significant countries have greater bargaining power and can benefit from fewer conditions than countries that are less geopolitically important. 131 Since independence, the Kingdom of Morocco has almost always aligned itself with the resolutions of two permanent members of the UN Security Council, the United States and France, and has been their unconditional ally. These two countries are also major contributors to the IMF. Is there a causal link? The Kingdom has got to look after its economic interests, of which debt is a major component. It is rational for a state to maintain a relationship of trust with its main public creditors, whether they intervene directly on a bilateral basis or indirectly in consultation with the IMF, 132 with other public creditors or via multilateral institutions. Although observers in 1983 praised the clever Moroccans who obtained certain reductions during negotiations with the IMF, 133 this cannot have been the only factor explaining the IMF’s(highly) relative magnanimity. After all, what influence do individuals, or even a state like Morocco, have on the IMF juggernaut? While we have glossed over the correlation between the geopolitical dimension and the measures adopted, it is obvious that the magnitude of the adjustments necessitated it, just as it was politically 129 Several studies find that domestic political conditions in the borrowing country and international policy factors influence conditionality. See Teri L. Caraway et al.,“International negotiations and domestic politics: The case of IMF labor market conditionality”, International Organization 66, No. 1(2012):27–61, 2012.; Axel Dreher and Nathan M. Jensen. “Independent Actor or Agent? An Empirical Analysis of the Impact of U.S. Interests on International Monetary Fund Conditions.” The Journal of Law& Economics 50, no. 1(2007): 105–24.[https://doi.org/10.1086/508311].; Dreher, Axel.“IMF Conditionality: Theory and Evidence.” Public Choice 141, no. 1/2(2009): 233–67,[http://www.jstor.org/ stable/40270954].; Dreher, Axel, Jan-Egbert Sturm, and James Raymond Vreeland.“Politics and IMF Conditionality.” Journal of Conflict Resolution 59, no. 1(February 2015): 120–48,[https://doi.org/10.1177/0022002713499723]. Erica R. Gould,“Money Talks: Supplementary Financiers and International Monetary Fund Conditionality”, International Organization 57, no. 3(2003): 551–86,[doi:10.1017/S0020818303573039].; Pierre 2008. Chwieroth and Nelson also show that conditionality varies according to the professional ties and shared beliefs of IMF staff and the borrowing country officials. See Jeffrey M. Chwieroth,“Professional ties that bind: how normative orientations shape IMF conditionality”, Review of International Political Economy 22, no. 4(2015), 757-787,[https://doi.org/10.1080/09692290.2014.898214] ; Stephen Nelson,“Playing Favorites: How Shared Beliefs Shape the IMF’s Lending Decisions”, International Organization 68, no. 2(2014): 297-328,[https://www.cambridge.org/core/journals/international-organization/article/abs/playingfavorites-how-shared-beliefs-shape-the-imfs-lending-decisions/9F2D7EE62C0AB7C3C2E21C0C2DD27CD7]. 130 The elasticity of conditionality is roughly constant with respect to geopolitical importance(Hainmueller et al., 2016). 131 In 2004, Dreher and Vaubel analyzed the underlying negotiations that determine IMF conditionalities. See Axel Dreher and Roland Vaubel,“The Causes and Consequences of IMF Conditionality”. 132 Ibid. 133 Michel Labonne,“Ajustement structurel au Maroc”. 160 Uncovered: The role of the IMF in shrinking the social protection

necessary to avoid at all costs a repetition of violent food riots, like those of 1981 and 1984 134 that shook the regime. However, if the Moroccan structural adjustment program was less dogmatic than that administered to African, Latin American and Eastern European countries, the measures imposed were no less severe 135 and very hard on the population. 136 This perception is not subjective: in 2006, Dreher 137 demonstrated that IMF intervention reduces growth rates, after Barro and Lee 138 had found in 2005 that higher participation rates in IMF programs negatively impact economic growth, democracy and the rule of law. 134 Ibid. 135 In the 1980s, Moroccans experienced them as a trauma. 136 One of the very first consequences of the structural adjustment program was the government’s decision to suspend recruitment, thereby abruptly depriving Moroccans of a vector of social ascension that the failing private sector was unable to provide. 137 Axel Dreher,“IMF and economic growth: The effects of programs, loans, and compliance with conditionality”, World Development 34, no. 5(2006), 769–788,[https://doi.org/10.1016/j.worlddev.2005.11.002]. 138 Robert J. Barro and Jong-Wha Lee,“IMF programs: Who is chosen and what are the effects?” Journal of Monetary Economics 52, no. 7(2005), 1245–1269,[https://doi.org/10.1016/j.jmoneco.2005.04.003]. Social spending in Morocco: The influence of IMF conditionality 161

Publisher: Friedrich-Ebert-Stiftung Regional Office MENA Economic Policies for Social Justice 4, rue Bachar Ibn Bord 2078 La Marsa Tunisia Email: info.mena@fes.de Responsible for content: Thomas Claes| Director| Economic Policies for Social Justice https://mena.fes.de/themes/economy-and-social-justice Contact/Orders: info.mena@fes.de Design: Marc Rechdane The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung. Commercial use of media published by the Friedrich-Ebert-Stiftung(FES) is not permitted without the written consent of the FES. Publications by the Friedrich-Ebert-Stiftung may not be used for electioneering purposes. Attribution 4.0 International Licensed as:

This book is being published at a historical moment similar to that in Latin America at the beginning of the foreign debt crisis in the 1980s, when most developing and low-income countries became tied up in foreign loans that initially seemed relatively inexpensive and necessary for development and growth. A decade later, these debts proved to be a heavy burden for Latin American countries, as the disadvantages of working with the IMF and the World Bank became clearer. The global crisis of 2008 then had the effect of pushing the Bretton Woods institutions to conduct important analyses of, and ultimately make revisions to, the Washington Consensus. We find ourselves today at a similar juncture, with foreign debt service burdens increasing in developing and low-income countries. The IMF and the World Bank have adopted a different discourse than that which they espoused in the 1980s and 1990s. This contemporary discourse is grounded in awareness of the ills of inequality of income and of opportunity, and supports imposing international taxes on giant corporations and progressive taxes on the rich. This discourse also ostensibly supports the Sustainable Development Goals and social protection. The book offers three case studies evaluating the main IMF- backed policies under the theme of Social Protection. As far as we could tell, this is the first book to conduct a systematic study of the major disparities between this discourse from the IMF and the IMF’s practice on the ground, and we hope it will be useful to the reader.