Social Policy Responses to Covid-19 in the Global South: Evidence from 36 Countries

The Covid-19 pandemic has prompted manifold social policy responses all around the world. This article presents the findings of a meta-analysis of thirty-six in-depth country reports on early Covid-19 social policy responses in the Global South. The analysis shows that social policy responses during the early phase of the pandemic have been predominantly focused on expanding temporary and targeted benefits. In terms of policy areas, next to labour market and social assistance measures, the focus has also been on unconventional social policy instruments. The social policy responses of developing economies were often rudimentary, focusing on cash transfers and food relief, and heavily relied on external funding. In contrast, many emerging economies introduced a much broader array of social policies and were less reliant on external support.


Introduct ion
Since its outbreak in early 2020, the Covid-19 pandemic and the extensive public health interventions it has made necessary have massively disrupted the lives and livelihoods of billions of people around the globe. Governments have introduced various social policy measures to mitigate these socioeconomic consequences of the pandemic. Indeed, it is in crisis times like these that the need for a generous and universal welfare state becomes most obvious (Farnsworth and Irving, 2011;Starke et al., 2013).
Social policy scholars across the world have responded quickly and begun to analyse governments' and international organizations' social policy responses to the pandemic (e.g. Béland et al., 2021a;Cook and Ulriksen, 2021;Leisering, 2021). Broadly speaking, differences in countries' social policy responses are rooted in the variable impact of the pandemic across countries as well as different political and economic contexts. While the pandemic has indeed been global and has somehow affected every single country in the world, it has not been an entirely 'common shock', let alone a 'natural experiment', for welfare states across the world (see Moreira and Hick, 2021). In addition, countries' social policy responses have been conditioned by differences in domestic factors, such as political regime, government ideology, or fiscal capacity.
One particular focus of many ongoing research projects on Covid-19 social policy responses has been the establishment of quantitatively oriented policy response 'trackers' (e.g. Cheng et al., 2020;Daly et al., 2020). However, there is still a major lack of in-depth case studies of governments' social policy responses, especially for countries of the Global South. To fill this gap, the Collaborative Research Centre 1342 'Global Dynamics of Social Policy', primarily based at the University of Bremen, has initiated a globally comparative research project to analyse worldwide social policy developments in the wake of the Covid-19 pandemic.
This article provides a meta-analysis of the first thirty-six country reports produced by this research project, which have examined early social policy responses to the Covid-19 pandemic in countries across the Global South-from Argentina to Zimbabwe. The article first reviews the emerging literature on Covid-19 social policy responses, and outlines the sample of thirty-six countries examined during the first phase of the project. The article then summarises and discusses the major findings that have emerged from the thirty-six country case studies. It concludes by sketching promising avenues for further research.

The emerging literature on Covid-19 social policy responses
Since the outbreak of the Covid-19 pandemic, a literature has emerged that seeks to document, evaluate and explain global social policy responses to Covid-19. This line of research clearly has a lot to learn from earlier research on the effects of economic crises on welfare state development (Farnsworth and Irving, 2011;Starke et al., 2013). At the same time, it will be an empirical question to what extent social policy responses to Covid-19 actually resemble responses to earlier, other kinds of crises (see Moreira and Hick, 2021).
Social policy scholars across the world have quickly begun to analyse governments and international organisations' social policy responses. Béland and colleagues (2021a) argue that social policy responses to Covid-19 have generally reflected, rather than broken with, national welfare state legacies. Relatedly, they argue that most social policy responses have aimed at stabilising rather than fundamentally changing the existing economic order-representing a kind of 'emergency Keynesianism'. Leisering (2021) argues that governments' social policy responses have been mostly 'stopgap measures' but suggests that international organisations have been more 'future-oriented', already developing 'normative models [ : : : ] designed for the time after the crisis'. Yet, most scholars seem to agree that the Covid-19 pandemic has not (yet) turned out to be a 'critical juncture' for welfare state development (see Gronbach and Seekings, 2021).
Early research on cross-national differences in social policy responses suggests that such differences are rooted in the variable impact of the pandemic across countries as well as different political and economic contexts. While the pandemic has indeed been a global event and has somehow affected every single country in the world, it remains an underappreciated point that not all welfare states have been affected equally (Moreira and Hick, 2021), as often implied by social scientific conceptualisations of the pandemic as a 'natural experiment' or a 'common shock' (see, for instance, Capano et al., 2020;Hick and Murphy, 2021). Taiwan's economy, for instance, remained largely stable throughout the pandemic thanks to early and effective public health containment measures, allowing Taiwan's government the luxury of a more limited social policy response (Soon et al., 2021;see Shi and Soon, 2020). In contrast, countries that were unable to effectively contain the pandemic, such as Brazil or Peru, faced much larger labour market disruptions and welfare state challenges, as did countries with high shares of informal employment (Jawad, 2020;Greer et al., 2021).
Countries' social policy responses are, of course, also shaped by different domestic conditions (see Greer et al., 2020). The emerging literature on Covid-19 social policy responses has already produced some tentative findings in this regard, which so far appear Social Policy Responses to Covid-19 in the Global South to confirm general theories of welfare state development (see Dorlach, 2021). To begin with, high fiscal capacity seems to have been an important precondition of strong social policy responses (Woo, 2020). On average, left-wing governments appear to have supported more generous social policy responses (Blofield and Pribble, 2021;Hornung and Bandelow, 2022), although the cases of Brazil and Mexico show that the effects of partisan ideology are complex and not always straightforward (Lustig and Trasberg, 2021). In contrast, political institutions with more 'veto points' seem to have inhibited quick and effective social policy responses (Rocco et al., 2020;Béland et al., 2021b;Ramia and Perrone, 2021).
The sheer size of this emerging new literature is testament to the commitment of social policy scholars to quickly contribute to a better understanding of the pandemic and its social consequences. Unsurprisingly, however, many of this literature's findings remain highly preliminary. This calls for further comparative analysis, especially of countries beyond the Global North.

Covid-19 soc ia l polic y res ponse s in thirt y-s ix c ountries
To contribute to a better understanding of initial Covid-19 social policy responses across the Global South, we commissioned thirty-six country experts or expert teams to author detailed country reports on social policy developments during the first year of the pandemic. In determining our sample of thirty-six countries, we used a combination of purposive and convenience sampling.
In line with our interest in the Global South, we focused on countries beyond the twenty-one 'core' member states of the Organisation for Economic Cooperation and Development (OECD). 2 Following the inclusion criteria of the Global Welfare State Information System (WeSIS), which is currently being developed by the CRC 1342, we only considered countries with a population of more than 0.5 million.
In arriving at our final sample of thirty-six countries (Table 1), some convenience sampling was unavoidable, as we relied on the short-notice availability of qualified country experts, a challenge that was exacerbated by the pandemic itself. Accordingly, not all experts who were interested in authoring country reports were also able to do so.
Despite these challenges, we managed to keep our sample of cases highly diverse (see Gerring and Cojocaru, 2016), both in terms of world regions and economic development levels, as per the World Bank's 2021 classifications. In terms of world regions, this sample includes ten countries from Latin America and the Caribbean, nine from Europe and Central Asia, seven from East Asia and the Pacific, seven from

Tim Dorlach
Sub-Saharan Africa, two from South Asia, and one from the Middle East and North Africa. In terms of economic development, the sample includes four high-income countries, fifteen upper-middle-income countries, sixteen lower-middle-income countries, and one low-income country. Each country expert or country expert team was asked to write a detailed essay describing the respective country's social policy response and to complete a systematic data appendix. In the data appendix (for an example, see Shi and Soon, 2020: 13-20), country experts were asked to report all significant national-level legislative activity between January and September 2020 in nine social policy areasnamely, 'pensions', 'healthcare', 'long-term care and disability', 'labour market', 'education', 'family benefits', 'housing', 'social assistance', as well as 'other policies aimed at social protection' (e.g. food subsidies or tax cuts). The inclusion of this final category was important because 'social policy by other means' (Seelkopf and Starke, 2019) continues to be widespread across the Global South. Explicitly excluded from this reporting were public health measures introduced to contain the pandemic, given that our interest was in the social policy measures passed to alleviate the negative social consequences of the pandemic.
The significant legislative activity that needed to be reported included both laws passed by parliament and (especially relevant in countries with high levels of executive control) presidential or cabinet decrees with the power of law. While the qualifier 'significant' gave country experts substantial discretion, it was necessary because the sheer volume of legislative changes made complete reporting in some countries impossible.
To ascertain the quality of reporting, all country reports underwent double-blind peerreview by at least one other country expert.

Meta-a nalysis of thirty-six country c ase studies
Based on a meta-analysis of the thirty-six country reports, this section outlines the main characteristics of Covid-19 social policy responses in the Global South. In a nutshell, countries' social policy responses during the early phase of the pandemic have been predominantly expansionary, but with a focus on temporary and targeted benefits. In terms of policy areas, the focus has been on labour market and social assistance measures as well as a series of unconventional social policy instruments. The social policy responses of developing economies were often rudimentary, focusing on cash transfers and food relief, and heavily relied on external funding. In contrast, many emerging economies introduced a much broader array of social policies and were less reliant on external support.

Expansion and maintenance
One of the most consistent findings from the thirty-six country case studies is that social policy change during the first year of the pandemic has been characterised by expansion and maintenance, with only rare instances of retrenchment. The majority of the 350 social reforms reported in the thirty-six country reports were categorised as expansion of coverage and/or generosity of existing benefits or as the introduction of new benefits. 3 For instance, China expanded existing unemployment insurance and social assistance programmes (ten Brink et al., 2021: 5-6), while countries like the Philippines and Zambia introduced entirely new cash transfer programmes (Pruce, 2021: 7;Ramos, 2021: 7-8).

Social Policy Responses to Covid-19 in the Global South
The fact that social policy change in the aftermath of the pandemic has been predominantly expansionary does not imply that welfare states in the Global South all expanded equally. Instead, a variety of different types of social policy responses can be identified. A few countries, including Costa Rica and Uruguay (Rossel and Gutiérrez, 2021;Voorend and Alvarado, 2021), already had well-developed and relatively universal welfare states, but introduced important smaller adjustments, such as Costa Rica's shortterm Protection Plan (Bono Proteger). A few other countries, with less universal welfare states, implemented broad social policy responses. This occurred most notably in Brazil (Massard da Fonseca et al., 2020), where poverty rates actually fell to a historic low during the pandemic, mostly thanks to the country's generous Emergency Allowance (Auxilio Emergencial). Yet other countries, such as Mexico and Uganda, largely maintained their insufficient social protection systems and implemented only 'extremely limited', 'minimal' or 'rudimentary' social policy responses, leaving their populations in dire need (Baldizón, 2021;Velázquez Leyer, 2021;Tallio, 2021). Most countries in the Global South, however, entered the pandemic without universal welfare states and then introduced social policy responses that were significant but not fully sufficient.

Rare retrenchment
Even though social policy change in the first year of the pandemic has been predominantly expansionary, it is interesting to briefly discuss some of the few cases of retrenchment (see Starke, 2021) that were reported in the thirty-six country case studies. One case of retrenchment, although of planned rather than existing benefits, comes from Azerbaijan, which postponed the introduction of a national universal health insurance programme, originally scheduled for 2020, 'due to the Covid-19 pandemic and rising costs' (Guliyev, 2021: 5). It remains to be seen how long this postponement will last. But the episode serves as a warning that the Covid-19 pandemic might slow down rather than speed up the global campaign for Universal Health Coverage (see Cook et al., 2020).
Another case of retrenchment comes from Costa Rica, which overall displayed a robust and generous social policy response. In March 2020, Costa Rica passed a shorttime work law, allowing 'private sector employers to reduce the working hours (and consequently, the salaries) of their employees by up to 50%, if the company's profits were reduced by at least 20%' (Voorend and Alvarado, 2021: 24). While likely inspired by the German Kurzarbeit scheme (see Greer et al., 2021), the Costa Rican programme was not state-subsidised. While the original German scheme involves a combination of retrenchment (of workers' salaries and rights) and expansion (of public social expenditure), the Costa Rican scheme only adopted the retrenchment part of this model.

Temporary and targeted expansion
While the direction of countries' social policy responses was predominantly expansionary, this expansion was generally temporary and targeted in nature, rather than permanent and universal. While it is understandable that social policy responses to an assumedly temporary pandemic should also be temporary, it is often unclear if benefit expansion will even last until the end of the pandemic, which is still not in sight for many countries in the Global South. In Turkey, for instance, a temporary ban on the firing of workers (Öktem, 2021: 5) expired in July 2021, even though the pandemic has entered yet another major

Tim Dorlach
wave. Cash transfer programmes in many countries, including Bolivia, India, and Indonesia, were explicitly designed as one-time transfers (Borges, 2021: 6;Kühner et al., 2021: 7;Sumarto and Ferdiansyah, 2021: 30). In other words, expectations that governments would use the pandemic as an opportunity to build more permanent and universal social safety nets have not yet materialised.
Many of the country reports also emphasised countries' reliance on status-or incomebased targeting of expanded social benefits, thus often limiting access to social benefits. For instance, Russia's social policy response explicitly targeted certain status groups, such as families with children or the (formally) unemployed, while largely excluding informalsector and rural workers (Tarasenko, 2021: 7). Ecuador, on the other hand, used incometargeting to determine eligibility for a new cash transfer programme, the Family Protection Grant, with the explicit purpose of expanding coverage to informal workers. However, the country's existing means-testing mechanism, the Social Registry, excluded around half of all households in the lowest income decile from access to the programme (Palacio, 2021: 8-9). At the same time, many governments realised that traditional targeting instruments required updating during the crisis to make sure that benefits reach their intended recipients. China, for example, largely digitalised the process of means testing required for accessing social assistance benefits (ten Brink et al., 2021: 6).

Predominance of labour market and social assistance policies
A meta-analysis of the 350 social reforms reported in the thirty-six countries reports also reveals an unequal distribution across policy areas, which means that welfare states in the Global South have used some policy instruments more than others in their pandemic responses. Figure 1 illustrates that the majority of social policy responses occurred in the fields of labour market policy and social assistance. While this type of data disregards the 'size' of policy measures, e.g. in terms of their budget, it still suggests that governments Social Policy Responses to Covid-19 in the Global South were mostly concerned with replacing lost incomes of workers in the formal and informal sector. Fewer reforms were introduced in the fields of (in descending order) healthcare, education, pensions, family benefits, housing, and long-term care and disability.
A high number of social reforms fell under the rubric of 'other policies aimed at social protection' (e.g. food subsidies or tax cuts), pointing to the fact that unconventional (from the perspective of Northern welfare state research) welfare policy instruments (see Dorlach, 2019;Seelkopf and Starke, 2019) have played a significant role in the crisis response of Southern welfare states. For example, Kenya's most immediate social policy response was the introduction of a temporary '100% [income] tax relief for individuals earning a gross monthly income of KES 24,000 [approx. USD 200] or less' with the expressed purpose of 'increasing household incomes' during the pandemic (Ouma, 2021: 5). In Zimbabwe, 'other' social policy measures in response to the pandemic aimed at food poverty reduction through 'direct grain provision' to rural households (Chipenda and Tom, 2021: 7). 4

Differences between developing and emerging economies
The meta-analysis also reveals clear differences in countries' social policy responses by their socioeconomic development level. The most significant line of distinction in this regard seems to be between low-and lower-middle-income countries (or 'developing economies') on the one hand and upper-middle-income and recently 'graduated' highincome countries (or 'emerging economies') on the other hand (see Lavers and Hickey, 2016;Dorlach, 2021). The social policy responses of developing economies were often rudimentary, focusing on cash transfers and food relief, while relying heavily on external funding (see Chipenda and Tom, 2021). In Zambia, for instance, the Emergency Cash Transfer programme, which combined cash and food assistance, was supported by multiple external donors, including the World Food Programme and the World Bank (Pruce, 2021: 7). Emerging economies, on the other hand, often introduced a much broader array of social policies in response to the pandemic and were much less reliant on external support. Argentina, for instance, also introduced cash transfer and food support programmes, but complemented these with wage and loan subsidies as well as a series of regulatory welfare policies, e.g. to protect tenants from rent increases or evictions during the pandemic. The total fiscal cost of Argentina's social policy response measures amounted to an estimated 3-4 per cent of GDP by October 2020 (Arza, 2021: 6). This confirms that broad conceptions of low-and middle-income countries (LMICs) or the Global South, while useful for some analytic purposes (e.g. Schmitt, 2020), mask major differences between the social policy dynamics of developing and emerging economies.

Conc lusion and research agenda
This article has presented meta-findings from a collaborative research project on the Covid-19 social policy responses of thirty-six countries in the Global South. In a nutshell, countries' social policy responses during the early phase of the pandemic have been predominantly expansionary, but with a focus on temporary and targeted benefits. In terms of policy areas, the focus has been on labour market and social assistance measures as well as a series of unconventional social policy instruments. The social policy responses of developing economies were often rudimentary, focusing on cash transfers and food Tim Dorlach relief, and heavily relied on external funding. In contrast, many emerging economies introduced a much broader array of social policies and were less reliant on external support.
In this concluding section, I highlight some issues that warrant particular attention from social policy scholars. Of perhaps most immediate interest is the question if some temporary social policy expansions will be made permanent or when they will be phased out. With limited signs of consolidation, it currently seems more likely that many benefits will remain temporary and expire soon. For instance, Brazil's impactful Emergency Allowance programme was initially implemented in April 2020 (Massard da Fonseca et al., 2020: 9-10; see Greer et al., 2021Greer et al., : 1216. The programme was extended several times, although with increasingly lower benefit generosity, and is currently set to expire in July 2021. Similarly, in the realm of labour market policy, Turkey introduced a temporary ban on the firing of workers in April 2020 (Öktem, 2021: 5), but allowed the legislation to expire in July 2021, likely setting up a wave of dismissals and rising unemployment.
Another major concern will be if (and where) social policy expansion might be followed by a period of retrenchment (see Starke, 2021). Social policy expansion during the first year of the pandemic was possible because of deficit spending and external funding. Countries' ability to continue on such a path might well be limited. The Covid-19 pandemic has had an extremely detrimental effect on economic activity and tax revenues, creating real constraints on government budgets (e.g. Sumarto and Ferdiansyah, 2021). Indeed, many countries purposefully lowered tax rates to provide an economic stimulus (e.g. Ouma, 2021). These budget constraints might create strong pressures on social expenditure. The existing literature on the effects of economic crisis on welfare state reforms (e.g. Starke et al., 2013) should provide a useful point of reference for research along these lines.
Finally, it will be of major interest if the pandemic might also trigger more structural welfare state reforms. Existing research suggests that similar disruptive events, such as wars, can have profound consequences for welfare state development (e.g. Obinger et al., 2018). There are some early signs that the Covid-19 pandemic might have similar effects, at least in some countries. The Moroccan government, for instance, announced plans to create a universal social protection system through an expansion of health and pension insurance as well as family and unemployment allowance schemes (Ait Mansour, 2021: 8). Likewise, South Korea has launched a 'Korean New Deal', which includes ambitious goals for the expansion of unemployment and work injury insurance schemes (Shim, 2021: 8). Many other countries, including South Africa, are at least considering the introduction of innovative Universal Basic Income schemes (Noyoo, 2021: 8). The key question in this context is, of course, if governments will honour their structural reform promises.
In other countries, the pandemic might prove to be much more disruptive of existing welfare state institutions. Chile and Peru, two pioneers of pension system privatisation, have allowed insurees to withdraw parts of their pension fund savings in order to provide income support (Kay and Borzutzky, 2022). In Peru, private pension savings worth 3.5 per cent of GDP were withdrawn between April and August 2020 (Olivera, 2021: 10). While this policy did solve short-term liquidity problems of many households, it will also reduce their future pension entitlements, setting up a social security crisis that may eventually trigger pension system re-nationalisation.

Social Policy Responses to Covid-19 in the Global South
These concluding considerations suggest that the effects of the Covid-19 pandemic on welfare states in the Global South are highly varied and still unfolding. While the pandemic might well trigger the creation of more generous and universal welfare states, it might also usher in a period of retrenchment or in episodes of 'creative destruction'. Whatever policy responses continue to emerge, they will be likely to travel between countries through diffusion mechanisms (see Kuhlmann et al., 2020). In the coming years, a major task for welfare state scholars will be to keep track of governments' evolving social policy responses to the pandemic. A special focus in this context will need to be put on countries beyond the small club of core OECD countries, as the pandemic threatens to halt the recent trend of welfare state expansion in the Global South (see Dorlach, 2021).