Policy Pollination as a Causal Mechanism Explaining Social Protection Adoption in Africa

Abstract

This chapter focuses on the instrumental role of international development agencies in the rapid and sustained rise of social protection as a policy agenda throughout Africa in the early twenty-first century. These agencies acted as “policy pollinators,” using various tactics to persuade African governments to implement social protection programmes. Such tactics included generating empirical evidence that cash transfers have beneficial impacts, financing social protection programmes, building state capacity to deliver social protection, and commissioning national social protection policies. Nonetheless, some governments resisted these pressures and inducements, because their political interests and national priorities were not well aligned with the agendas of development agencies. Questions arise about whether the adoption of social protection by African governments should be understood as a nationally owned or “donor-driven” process.

1 Introduction

This chapter is an amended version of a working paper co-published by the SOCIUM Research Centre on Inequality and Social Policy and the Collaborative Research Centre 1342 “Global Dynamics of Social Policy”, University of Bremen, and the UK Institute of Development Studies (Devereux 2020).

Since the late 1990s, a particular form of social protection has become entrenched in the social policy agenda across the world. At the global level, social protection was not mentioned in the Millennium Development Goals in 2000, but it features in three Sustainable Development Goals (United Nations 2015). At the continental level, no African country had a national social protection policy in 2000, but thirty-five of fifty-five had produced one by 2019. Millions of Africans who had no access to social assistance twenty years ago now receive social cash transfers (SCTs) from their governments every month (UNDP 2019).

The literature on social protection has also grown and evolved during this period. Initially (as discussed below), empirical research focused on impact evaluations to establish the effectiveness of social cash transfers and of alternative design choices (e.g. conditional versus unconditional cash transfers). More recently, academic interest has shifted to trying to understand the causal mechanisms driving the widespread adoption and expansion of social protection as a policy sector, especially in Africa. Most relevant for this chapter is an emerging literature that analyses the diffusion of social protection policies, specifically cash transfers (von Gliszczynski and Leisering 2016; Leisering 2019), with a particular focus on sub-Saharan Africa (Adésínà 2011; Hickey et al. 2019). This literature resonates with theoretical and empirical literatures on social welfare in OECD countries (Esping-Andersen 1990; Schmitt and Starke 2012), on welfare provision in the Global South during and since the colonial period (Gough and Wood 2004; Midgley and Piachaud 2011; Schmitt 2015), and on cross-country policy transfer processes (Dolowitz and Marsh 2000; Dobbin et al. 2007; Obinger et al. 2013).

Policy transfer can be understood as a process whereby “knowledge about policies, administrative arrangements and ideas in one political system (past or present) is used in the development of policies’ administrative arrangements, institutions and ideas in another political system” (Dolowitz and Marsh 2000, 5). Causal mechanisms for policy transfer include learning, competition, and emulation—a tendency for countries to mimic policies such as free primary education that have been socially constructed as desirable at the global level.

A fourth causal mechanism in the policy transfer literature is coercion, which involves the use of power (e.g. financial leverage) by transnational actors to induce policy change by national governments, especially those that are “reliant on those entities for trade, foreign direct investment, aid, grants, loans” (Dobbin et al. 2007, 454). This mechanism has been labelled coercive diffusion. “An often-mentioned case of coercive diffusion is financial aid linked to certain domestic reforms defined by donor countries and international institutions such as the International Monetary Fund or the World Bank” (Obinger et al. 2013, 115).

This chapter identifies a fifth causal mechanism, here labelled policy pollination, that carries elements of all four mechanisms in the policy transfer literature but is most closely aligned to coercive diffusion and, as argued below, persuasively explains the spread of social protection in sub-Saharan Africa. In nature, cross-pollination is the process by which pollen is carried between flowers by bees, birds, and other pollinators. For this analogy, flowers are African countries, pollen is social protection policies, and pollinators are agents (staff, consultants) of international development agencies, flying from country to country to propagate social protection. Fertilisation is achieved when countries adopt and finance social protection policies and programmes, as advised by their pollinators.

The next section introduces the notion of policy pollination as a causal mechanism, drawing on Africa’s experiences with structural adjustment in the 1980s and 1990s, Poverty Reduction Strategy Papers in the early 2000s, and social protection more recently. Next, development agencies are identified as the dominant pollinators in the social protection policy transfer process, led by ILO, UNICEF, and the World Bank. Five strategies for policy pollination are then discussed: (1) evidence-building; (2) financing; (3) capacity strengthening; (4) policy formulation; and (5) domestication of international law. There follows a brief discussion of why some African governments have resisted consistent pressure and incentives to adopt the specific variant of social protection advocated by transnational actors.

2 Policy Pollination as a Causal Mechanism

International development agencies¹ have intervened in social policy formulation in Africa since the colonial period and continue to do so today, decades after former colonies achieved independence (Schmitt 2020). This has been achieved primarily through coercive diffusion, leveraging the hard and soft power conferred by development finance—official development assistance (ODA), concessional lending, and humanitarian relief. In the 1980s, for instance, international financial institutions imposed conditions on structural adjustment loans that required African governments to, among other things, remove food price subsidies and impose user fees on poor citizens for health and education services. A striking feature of these Washington consensus (Williamson 2000) policy prescriptions was their uniformity. A simple economic model for market liberalisation and state withdrawal, devised by a powerful group of transnational actors led by the World Bank and International Monetary Fund (IMF), was implemented reluctantly by many African governments, with consequences for poor Africans that were ambivalent at best (Easterly 2006; Moyo 2009).

In the early 2000s a wave of Poverty Reduction Strategy Papers (PRSPs) spread across the Global South, more than half in Africa. In the year 2000 alone, twenty-five African governments prepared a PRSP document (IMF 2016). Again, this policy process was driven by the World Bank and IMF. Preparing a PRSP was rewarded with concessional loans or debt relief (Ejolu 2008). All PRSPs embodied the same pillars, framed within a neoliberal ideology: good governance, pro-poor economic growth, investment in human capital, and in some cases, social safety nets to protect the poorest. The striking similarities in these documents across diverse countries reveal the extent to which they were designed by external agencies acting as “policy pollinators,” rather than emerging organically from domestic deliberative processes about optimal strategies for reducing poverty in each country context.

PRSPs were criticised as “a primary policy device of international development institutions (that) restrict practical and political options, while exacting heavy establishment and compliance costs” (Craig and Porter 2003, 53). Compliance was enforced through coercive application of conditionalities on loans by the World Bank and IMF. However, their roll-out to dozens of low- and middle-income countries was represented by these agencies as participatory and democratic. The IMF described the adoption of PRSPs as a country-driven process, “promoting national ownership of strategies through broad-based participation of civil society” (quoted by Ejolu 2008, 22). Although PRSPs were implemented in more than thirty-five African countries, in recent years they have almost disappeared, confirming how shallowly grounded in domestic policy processes they actually were.

As PRSPs faded into history or were absorbed into national development strategies, another wave of policy documents started spreading across Africa. Since 2010, thirty African countries have published a National Social Protection Policy (NSPP) or National Social Protection Strategy (NSPS). These documents share many common features. Notably, they all include promises that the government will deliver social assistance to specified target groups: children, older persons, persons with disabilities, and “the poor” (UNDP 2019, 56). As discussed later, these policy statements have been heavily influenced—financed, commissioned, even (co-)authored—by transnational policy pollinators: development agencies and agents working for them.

A statement by two United Nations Special Rapporteurs highlights the disproportionate power that international agencies exercise over social policy formulation in poor countries, and the contradictions between the Washington consensus policies that they imposed in the 1980s (e.g. making poor people pay for education and healthcare) and the expansionist social protection policies (notably giving free cash transfers to poor people) that were advocated by the same agencies in the 2000s.

In the past, major international institutions have pushed States to lower government spending and programming in favour of economic development, opening markets and reducing the size of the State. In the last decade however, many international institutions have begun to address the benefits of social protection systems to development and to promote their adoption. (de Schutter and Sepúlveda 2012, 3)

As with PRSPs, the story of social protection’s rapid rise in Africa is interpreted differently by different observers, some seeing it as evidence of progressive thinking by African governments, others as evidence of the power of international development actors to impose their will on governments they advise and support. This can be represented as two stylised narratives.

• 1. The national ownership narrative:Social protection in Africa is a success story for African social policy in the early twenty-first century. Almost two in three African countries now have a national social protection policy or strategy, from a baseline of zero in 2000. Large-scale government-run social protection programmes have been rolled out to millions of people: the Child Support Grant in South Africa (launched in 1998, reaches twelve million children), Productive Safety Net Programme in Ethiopia (launched in 2005, reaches nine million people), Livelihood Empowerment Against Poverty in Ghana (launched in 2008, reaches 150,000 people), among many others (UNDP 2019). Several governments in Africa now have social protection ministries, agencies, and laws.

• 2. The donor-driven narrative:Social protection in Africa is a success story for the international development community in the early twenty-first century. The development industry (bilateral and multilateral donors, United Nations agencies, management consultants, research institutes, NGOs) has “pollinated” social protection throughout Africa. National policies do not necessarily imply national ownership. Government commitment to social protection is variable across the continent, being lower in the poorest aid-dependent countries, where national social protection strategies are conceived and often drafted by international consultants and most social protection programmes have been designed, financed, and evaluated by development actors.

Support for both positions can be found in the social protection literature. In 2010 a book titled Just Give Money to the Poor presented social cash transfers in its subtitle as The Development Revolution from the Global South (Hanlon et al. 2010), implying that the idea is indigenous rather than imported. By contrast, an article published in the Journal of Social Policy in 2016 was subtitled “How international organisations defined the field of social cash transfers in the 2000s” (von Gliszczynski and Leisering 2016). The next section builds the case in support of this second position, showing how external actors pollinated the ideas of social protection and cash transfers throughout much of Africa.

3 International Development Agencies as Policy Pollinators

Social protection as a development discourse emerged out of social safety nets, a limited set of policy instruments that straddles the gap between social assistance and humanitarian relief. The 1990 World Development Report set out a two-pronged strategy for reducing poverty: labour-intensive economic growth and provision of basic social services to the poor. The report also identified the need for safety net programmes as a complementary third element.

Even if this basic two-part strategy is adopted, many of the world’s poor—the sick, the old, those who live in resource-poor regions, and others will continue to experience severe deprivation. Many others will suffer temporary setbacks owing to seasonal variations in income, loss of the family breadwinner, famine, or adverse macroeconomic shocks. A comprehensive approach to poverty reduction, therefore, calls for a program of well-targeted transfers and safety nets as an essential complement to the basic strategy. (World Bank 1990, 3; emphasis added)

This is possibly the first articulation in the development policy discourse of what later became the two building blocks of social protection systems in the Global South: targeted transfers or social assistance to the chronically poor (who “experience severe deprivation”) and categorical vulnerable groups (“the sick, the old”), as well as safety nets or social insurance for those experiencing livelihood shocks (“temporary setbacks”).² In the late 1990s the World Bank developed the safety net component into its Social Risk Management framework (World Bank 2001), and it continues to use the term safety nets even though it is out of favour with other agencies. As recently as 2018 the World Bank published the third edition of “The State of Social Safety Nets” (World Bank 2018).

In 2000, social protection effectively did not yet exist in the development policy discourse. The period since has been framed by two global position statements about development policy: the Millennium Development Goals (MDGs) and the Sustainable Development Goals (SDGs). As noted earlier, the MDGs did not mention social protection at all. Just fifteen years later, social protection consolidated its entrenchment in the international development policy discourse by being named in three of the seventeen SDGs, as an instrument for ending poverty (goal 1), enhancing gender equality and empowerment (goal 5), and reducing inequality (goal 10).³

The year 2012 was another watershed for social protection. Three international development agencies—the ILO, UNICEF, and World Bank—had emerged as the leading actors and policy pollinators in this sector in the Global South. It was increasingly clear that social protection was not a development policy experiment or fad that would disappear after a few years. Coincidentally, these agencies each released definitive statements of their approach to social protection in the same year. These documents give operational and programming content to the aspirations of the international community to pollinate distinct visions of social protection in aid-recipient countries, and the types of interventions each agency chooses to pollinate, reflecting their mandates and ideological positions as much as technical assessments of social needs and context-specific policy options.

In June 2012, the International Labour Conference endorsed a “Recommendation concerning National Floors of Social Protection,” which presented an explicit rights-based approach to social protection, building on the Universal Declaration of Human Rights of 1948. The Recommendation defined social protection floors as “nationally defined sets of basic social security guarantees which secure protection aimed at preventing or alleviating poverty, vulnerability and social exclusion” (ILO 2012). “The guarantees should ensure at a minimum that, over the life cycle, all in need have access to essential health care and to basic income security” (ILO 2012). The four guarantees are:

• 1. Basic income security for children

• 2. Basic income security for persons in active age unable to earn sufficient income

• 3. Basic income security for persons in old age

• 4. Access to a set of goods and services constituting essential health care

In effect, this set of provisions elaborates and gives rights-based effect to the underdeveloped social assistance component of social protection systems, since the Social Security (Minimum Standards) Convention sixty years earlier had focused mainly on social insurance (ILO 1952).

Also released in 2012, UNICEF’s “Social Protection Strategic Framework” (UNICEF 2012) identified three principles for “integrated social protection systems”: progressive realisation of universal coverage, nationally owned systems and national leadership, and inclusive social protection—tackling social exclusion. The specific focus of UNICEF’s engagement with social protection is revealed in the subtitle: “Enhancing equity for children.” While this reflects UNICEF’s mandate as the United Nations agency for children, it also leads UNICEF to favour a certain set of policy instruments and to prioritise some vulnerable or “at risk” groups above others.

UNICEF’s vision of a social protection system has four components (UNICEF 2012):

• 1. Social transfers: Predictable direct transfers to individuals or households, both in-kind and in cash, to protect against shocks and support accumulation of human, financial, and productive assets.

• 2. Ensuring access to services: Interventions that reduce economic and social barriers to basic social services.

• 3. Social support and care services: Human resource-intensive services that identify and reduce vulnerability and exclusion.

• 4. Legislation and policy reform: Changing policies and legislation to remove inequalities in access to services and/or economic opportunities, addressing issues of discrimination and exclusion.

The World Bank’s “Social Protection and Labor Strategy” (World Bank 2012) links social protection directly to labour markets: “social protection and labour systems … help people and families find jobs, improve their productivity, cope with shocks, and invest in the health, education, and well-being of their children.” The subtitle—“Resilience, Equity, and Opportunity”—reveals that the World Bank sees social protection as performing instrumental functions, beyond being a right or entitlement. It should contribute to building resilience and generate opportunities to escape from poverty, such that beneficiaries will become self-reliant and resilient and ultimately no longer need social assistance.

The World Bank strategy has three overarching goals:

• 1. “improve resilience by helping people insure against drops in well-being from different types of shocks”

• 2. “improve equity by reducing poverty and destitution and promoting equality of opportunity”

• 3. “promote opportunity by building human capital, assets, and access to jobs and by freeing families to make productive investments because of their greater sense of security.” (World Bank 2012, xi)

The strategy envisages a linear progression to building social protection and labour systems, starting with small uncoordinated projects towards fully connected, well-functioning, and efficiently run national programmes. “The strategic direction is to help developing countries move from fragmented approaches to more harmonized systems for social protection and labour.” With this framing, the main prongs of the World Bank’s strategy are social assistance, social insurance, and active labour market programmes (Robalino et al. 2012):

• 1. Social assistance: cash transfers, food programs, public works

• 2. Social insurance: unemployment benefits, health insurance, disability pensions, survivors’ pensions, old-age pensions

• 3. Active labour market programmes: intermediation, counselling, job search and matching, vocational skills training, wage subsidies to firms

While social assistance programmes have been effectively pollinated throughout Africa (UNDP 2019), many with World Bank financial and technical support, social insurance has received less attention and active labour market programmes have often failed. A recent review of evaluations concluded that “active labour market policies are much less effective than policymakers typically assume. Many of these evaluations find no significant impacts on either employment or earnings” (McKenzie 2017).

These position statements by three of the most influential international development agencies are important because they define the focus and parameters of each agency’s engagement with, and support for, social protection policies and programmes in countries where they operate—where they exercise substantial influence over policymaking processes, using policy pollination strategies that are discussed next.

4 Policy Pollination Strategies

The international development community has invested heavily in advocating for social protection in Africa since 2000, using various strategies. Five such policy pollination strategies are discussed here: (1) building the evidence base for social protection impacts, (2) financing social protection programmes and systems, (3) strengthening government technical and operational capacity, (4) instigating national social protection policies, and (5) domesticating international law.

4.1 Building the Evidence Base

In the early 2000s the international development community selected social cash transfers (SCTs) as their preferred instrument for social protection in low- and middle-income countries. The UK’s Department for International Development (DFID) set out the theoretical case for cash transfers.

While poverty is multi-dimensional, low and variable income is central to the problem. Modest but regular income from cash transfers helps households to smooth consumption and sustain spending on food, schooling and healthcare … Over time, transfer income can help households to build human capital (and) save up to buy productive assets … Cash transfers can thus both protect living standards (alleviating destitution) and promote wealth creation. (DFID 2011, i)

However, the evidence base at first was thin. There were few unconditional cash transfer programmes in low-income countries, and even fewer that had been rigorously evaluated. Development partners therefore decided, for strategic reasons, to invest in building the evidence base that cash transfers can protect and promote living standards. The implicit theory of change was that policy adoption is evidence driven, so proving that cash transfers can contribute to desired outcomes such as poverty reduction would convince governments to implement and pay for their own national programmes.

From the very first cash transfer pilot projects in Africa, development agencies allocated substantial funds to monitoring and evaluation (M&E) of projects that they instigated and financed. The technical justification for conducting impact evaluations is to confirm that a policy intervention is achieving its objectives and to generate learning for improved design or delivery. Pilot projects also allowed experimentation with different design modalities. In Lesotho’s Cash and Food Transfers Pilot Project, some beneficiaries received food packages, some were given cash, while others received both cash and food. In Kenya’s Hunger Safety Net Programme, three targeting mechanisms determined eligibility in different communities—high household dependency ratios, community-based wealth ranking, and older persons—and their targeting accuracy was compared (Sabates-Wheeler et al. 2015).

Development agencies also commission evaluations of interventions they support for advocacy reasons, aimed at two politically important audiences. First, donor governments need to justify their spending on development programmes to their domestic constituencies—taxpayers in high-income countries. Showing positive impacts in terms of, say, poverty reduction in Africa is strategically necessary to keep the pipeline of development assistance flowing. Cash transfer projects have immediate demonstrable benefits, not least because giving money to poor people makes them less poor, by definition, so it was relatively easy to quantify these outcomes as attributable achievements of development spending.

Second, as noted, development agencies commission impact evaluations of their projects for advocacy purposes in countries where they work, as one policy pollination strategy. Findings of positive outcomes from pilot SCTs were intended to convince African governments to adopt this idea and scale it up. The UK development agency explicitly acknowledged the link between building the evidence base and building political support. “Robust monitoring and evaluation are crucial both for programme performance and political sustainability. The rapid spread of cash transfers in MICs in recent years has been in large part due to just such high-quality analysis and M&E” (DFID 2011, vii). For this reason, even small-scale pilot projects received disproportionately large evaluation budgets.

The first decade of this century saw a rapid accumulation of evidence that social protection in Africa, specifically cash transfers, “works.” But early evaluations were not always methodologically rigorous. A GTZ-funded evaluation of the influential Pilot Social Cash Transfer Scheme in Kalomo District, Zambia, found improvements for beneficiary households in several indicators relative to baseline, including food security and asset ownership (MCDSS 2007). However, this study was criticised for not including a control group, which meant that findings of positive changes in beneficiaries’ wellbeing could not be unambiguously attributed to the SCT. Following these first-generation social protection initiatives, a more sophisticated wave of impact evaluations was commissioned of larger programmes such as the Productive Safety Net Programme (PSNP) in Ethiopia, the Hunger Safety Net Programme (HSNP) in Kenya, and Concern Worldwide’s “Graduation model” projects in Burundi and Rwanda. Multi-year multi-round household surveys were designed (baseline, midline, endline, sometimes also follow-up) with treatment and control groups, following rigorous quasi-experimental randomised control trial (RCT) protocols.

Numerous publications have synthesised the accumulating experience and evidence with various forms of social protection programmes across African countries. Notable contributions include:

• 1. Social Protection in Africa. This book presents fifteen case study programmes from six southern African countries, drawing on research commissioned by the DFID- and AusAID-funded Regional Hunger and Vulnerability Programme. RHVP’s Regional Evidence-Building Agenda compiled evidence of social protection innovations and disseminated lessons across six countries to promote uptake of best practice, applying a simple “policy influencing” model: “Evidence-building + Capacity-building = Positive policy change” (Ellis et al. 2009).

• 2. Cash Transfers Evidence Paper. Written by staff in the Policy Division of the UK Department for International Development. “This paper provides a synthesis of current global evidence on the impact of cash transfers in developing countries, and of what works in different contexts … While the primary purpose of cash transfers is to reduce poverty and vulnerability, the evidence shows that they have proven potential to contribute directly or indirectly to a wider range of development outcomes” (DFID 2011, i).

• 3. The Cash Dividend: The Rise of Cash Transfer Programs in Sub-Saharan Africa. A book written by a World Bank economist and a World Bank consultant. “In 2009, growing interest in the use of CT programs in Sub-Saharan Africa led the World Bank to initiate a comprehensive desk review of the CT programs that had been used recently in the region. This book presents the results of the review” (Garcia and Moore 2012, 2).

• 4. Social Protection for Africa’s Children. An edited book published with financial support from DFID and UNICEF. “This book includes both ‘quantitative’ and ‘qualitative’ studies of social protection in Africa that either target children directly or have significant impacts on children’s well-being” (Handa et al. 2011, 7).

• 5. Cash Transfers: What Does the Evidence Say? A rigorous review commissioned by UK AID from the Overseas Development Institute (ODI). “This review retrieves, assesses and synthesises the evidence on the effects of cash transfers on individuals and households through a rigorous review of the literature of fifteen years, from 2000 to 2015” (Bastagli et al. 2016, 5). Summarising the impacts reported in 165 evaluations of fifty-six cash transfer programmes in low- and middle-income countries, the authors found “strong evidence” that cash transfers are associated with increases in household expenditure (a proxy for reductions in monetary poverty), food expenditure and dietary diversity (indicators of food security), school attendance, use of health services, household savings, and women’s decision-making power.

• 6. From Evidence to Action: The Story of Cash Transfers and Impact Evaluation in Sub-Saharan Africa. An edited book that was a product of the Transfer Project, co-funded by two United Nations agencies—the Food and Agriculture Organisation (FAO) and UNICEF. “Evidence on the effectiveness of unconditional cash transfers provided through government programmes in SSA has not been substantially documented. … One key objective of this book is to provide an overview of this accumulated evidence” (Davis et al. 2016, 1).

• 7. Realizing the Full Potential of Social Safety Nets in Africa. A World Bank report in its Africa Development Forum series. “This report first presents a snapshot of social safety nets in Africa and the mounting evidence for the effectiveness of these programs in promoting the well-being and productive inclusion of the poorest and most vulnerable” (Beegle et al. 2018, 2).

• 8. The State of Social Assistance in Africa. A report produced by UNDP in collaboration with other UN agencies and the African Union. “The motivation behind this report and data platform has been to provide African policymakers, civil servants, researchers, development practitioners and civil society a comprehensive overview of social assistance in Africa across its legal, financing and institutional dimensions” (UNDP 2019, 11).

None of these publications are products of “blue skies” academic research; they were written or commissioned by international development agencies for advocacy purposes, to promote specific policy positions. The Foreword to “From Evidence to Action” states: “These pages also document the ways in which the Transfer Project has influenced the policy debate in each of the eight countries” (Davis et al. 2016, vi). While the primary intended impact of cash transfer projects is to improve the wellbeing of beneficiaries, the primary intended impact of cash transfer evaluations is to persuade governments to implement and finance national cash transfer programmes.

4.2 Financing Social Protection Programmes

Many African governments were reluctant to introduce social protection programmes, despite pressure and financial incentives from international agencies to do so. One major reason is that these governments see social protection as unaffordable, given fiscal constraints in low-income contexts and competing priorities for public spending (Seekings 2017). In particular, cash transfers at scale are regarded as too expensive, especially if they involve regular transfers of meaningful amounts of cash to all people defined as poor in the country every month, not only for a year or two but indefinitely.

One response to the reality of limited budgets and low prioritisation of social protection in much of Africa was for external actors to provide the seed funding, especially for technical inputs such as design and systems-building, but also in many cases for the cash transfers themselves. This was intended to be an interim measure until governments assumed responsibility for running and financing these projects as national programmes. The anticipated shift in financing from external to domestic sources can be illustrated as a “funding seesaw.” It is striking that the countries where social protection is fully funded by external actors are some of the world’s poorest and most aid dependent. Conversely, all the countries where social protection is fully funded from domestic resources are middle or high income (World Bank 2018, 18).⁴ This provides empirical evidence for a familiar dilemma—that countries with the greatest need for social protection have the least resources to deliver it—and justifies external actors stepping in to fill the financing gap. Sub-Saharan Africa has “the smallest proportion of working-age population in the world” (ILO 2017, 130) and high levels of economic informality, therefore a very small tax base. The funding seesaw predicts that as national incomes rise, governments will collect more taxes and allocate more fiscal resources to social protection, which also allows them to take more control over social protection design and delivery. Until then, development agencies will continue to offer full or partial funding of social protection in low-income African countries, and they will continue to exploit this leverage by shaping social protection in these countries in line with their own mandates and objectives.

A second response by external actors to claims that social protection is unaffordable in low-income countries was to develop the “business case” for “investment” in social protection. This required demonstrating that public spending on social protection generates economic returns to individuals, local economies, and the national economy. Pathways from cash transfers to economic growth include human capital formation (this explains why the World Bank promotes conditional cash transfers that require beneficiaries to send their children to school and health clinics) and local income multipliers (because cash transfers stimulate demand for goods and services). FAO simulated the multiplier effects of cash transfer programmes using a methodology called Local Economy-Wide Impact Evaluation (Thome et al. 2016). The Economic Policy Research Institute (EPRI), a think-tank, used a microsimulation model showing how investment in social protection contributes to economic growth, reducing poverty and requiring less social protection in the future (Samson 2005). The Australian and UK aid agencies each commissioned reviews of the effects of social protection on economic growth (Mathers and Slater 2014; Barrientos and Scott 2008; respectively). While these studies might be characterised as contributions to the evidence-building agenda rather than the financing debate, the primary purpose was advocacy—challenging government perceptions that social protection is wasteful or unproductive expenditure that low-income countries cannot afford.

Related to this was work demonstrating that social protection can generate “value for money” for governments as well as agencies, in terms of cost-efficiency and cost-effectiveness against policy objectives such as poverty reduction. DFID commissioned studies of “measuring and maximising value for money” in social transfer programmes and social protection systems (White et al. 2013, 2015). These manuals offer guidance on how to assess whether social protection programmes and systems actually are reducing poverty, and how to achieve this more cost-efficiently and more cost-effectively.

A third response by external actors was to point out that public spending decisions are political choices and that even the poorest countries can afford some spending on social protection. The World Social Protection Report 2017–19 reveals that spending on social protection as a percentage of GDP averages just 4.5 per cent in sub-Saharan Africa, but 18 per cent in Western Europe (ILO 2017). The ILO argues that African countries should allocate more to social protection, because poverty is lowest in countries that spend the most on social protection, and they produced guidelines setting out how governments can create more fiscal space. Options include increasing tax revenues, reallocating public expenditure, expanding contributory social security coverage, lobbying for international aid, reducing illicit financial flows, drawing on foreign exchange reserves, and borrowing or restructuring government debt (Ortiz et al. 2015). ILO’s research draws on empirical data, modelling, and country case studies, but it was undertaken for explicit advocacy purposes, to persuade governments that they can (and should) spend more on social protection than they do already.

Despite these efforts, the question of who finances social protection programmes remains a contested issue. As Seekings (2017, ii) explains, “international organizations have generally failed to convince national policymaking elites to raise and to allocate scarce domestic resources to social protection programmes. The result is an ‘affordability gap’ between what is advocated for African countries and what these countries’ governments are willing to spend.”

4.3 Strengthening Capacities to Deliver Social Protection

International agencies have invested in building African government capacity to design and implement social protection programmes and to strengthen social protection systems. Strategies include embedding expatriate consultants as advisors within government ministries, arranging study tours for politicians and technical staff to observe social protection practices in other countries, and facilitating high-level dialogues with parliamentarians or regional bodies such as the African Union.

The most popular capacity-building mechanism is training workshops, where government officials convene for periods between one day and two weeks, to learn from “experts” about social protection theory and practice. The workshop setting allows agencies to transfer their preferred approaches to social protection to participants, through positions taken by trainers in relation to specific design choices. Agencies have fundamentally different visions of social protection. For instance, while the World Bank favours poverty targeting using proxy means tests, UNICEF prefers categorical targeting of vulnerable groups, while ILO advocates instead for universal rights-based programmes.

Some agencies run their own training workshops. The World Bank has run a course in Washington every year since the early 2000s, now called the “Social Safety Nets and Delivery Core Course,” for policymakers, analysts, and operational staff from international agencies. This course “aims to provide participants with an in-depth understanding of the conceptual and practical issues on safety nets or social assistance as part of broader social protection systems.”⁵ One session in the 2019 course was called “Making the case for social safety nets,” confirming that the pedagogical objective is policy pollination as much as being purely educational. The ILO runs an annual two-week Academy on Social Security at its International Training Centre in Turin. This course offers “a diversified training package on governance and financing, reforms and extension of social protection systems.”⁶

Online training is also available through TRANSFORM, an inter-agency initiative led by ILO, UNICEF, and UNDP, and hosted by the Virtual Campus of socialprotection.org. TRANSFORM—Leadership & Transformation Curriculum on Building and Managing Social Protection Floors in Africa—explicitly promotes the ILO’s preferred approach. “By the end of this course, you should be able to understand why and how a Social Protection Floor is beneficial to your specific country context and how it can assist social and economic development.”⁷

Training courses convey specific information and technical skills, but they also offer an unparalleled opportunity for policy pollination, by giving trainers who are employed or contracted by international agencies a platform and a captive audience of social protection policymakers and stakeholders. Social protection has many definitions and is characterised by numerous design dilemmas—cash or food? conditional or unconditional? targeted or universal?—that are not technical problems but policy choices that reflect competing ideological positions. Consciously or unconsciously, trainers inevitably communicate their personal biases—and those of their agencies—to course participants, and this causally influences social protection policy processes when participants return home to their offices.

4.4 Instigating National Social Protection Policies or Strategies

The first National Social Protection Policy (NSPP) in Africa was promulgated by Mali in 2002, and only four more countries had followed by 2010. Then in 2011, five more countries published their social protection policies, and by 2019 the number had risen to thirty-five, almost two-thirds of the fifty-five countries in Africa (Fig. 7.1).

Fig. 7.1

New national social protection policies or strategies in Africa by year, 2000–2019. (Source: Author’s compilation)

Nonetheless, it does not necessarily follow that this was a nationally owned process of developing and institutionalising a domestic policy agenda in every case. The short period in which so many countries published these documents suggests that a deliberate policy transfer process was underway. As with the wave of Poverty Reduction Strategy Papers (PRSPs) that spread across the Global South in the early 2000s (as discussed earlier), external actors have been heavily involved in guiding or driving the process of drafting social protection policies in Africa since 2002. Most of these policy documents were drafted or commissioned by international agencies and their expatriate consultants, with varying degrees of involvement of national consultants and relevant government officials. Twenty-two of the thirty-five African countries that have an NSPP are low income and twelve are lower-middle income, all with high levels of involvement of international development partners in domestic policy formulation.

From 2010 to 2012 the World Bank led a process of developing a National Social Protection Policy and Strategy for Togo. The Acknowledgements explain that: “This report was prepared by a World Bank team with support from the Government of Togo and (…) various donors, including the World Bank, UNICEF, ILO, and UNDP (…) This final version incorporates the Government’s comments” (World Bank 2012, i).

The covers of some African social protection policy documents display the institutional logos of international agencies that supported the development of the policy, alongside the country’s coat of arms, clearly signifying that these policies were at best co-produced and co-owned. Examples include the Gambia (UNDP and UNICEF) and Liberia (African Development Fund, European Union, UNICEF, World Bank, World Food Programme, and Concern Worldwide). The content of these policies further confirms that they reflect the ideas, ideologies, and priorities of these agencies, rather than those of national governments. At least twenty-five NSPPs are organised around four imported conceptual frameworks: the life-cycle approach, social protection floor, social risk management, and transformative social protection. These frameworks derive, respectively, from UNICEF, ILO, the World Bank, and the Institute of Development Studies (Devereux and Kapingidza 2020).

4.5 Domestication of International Law

Another plausible explanation for the adoption of social protection by so many African governments almost simultaneously is social construction, which asserts that globalisation has been associated with a growing consensus across the world about appropriate societal goals, and the actors and means to achieve them (Dobbin et al. 2007). A case in point is the rise since World War II in the recognition of individual human rights (as a societal goal), the establishment of the United Nations and the International Criminal Court (as actors), and the overthrow of repressive dictatorships and their replacement with democratic institutions like regular elections and a free press (as means).

Seen through a constructivist lens, the rapid diffusion of social protection throughout Africa could be interpreted as reflecting the voluntary incorporation of globally constructed human rights-based agendas into domestic policy processes by African governments. After all, they are active participants in global policy forums (notably the United Nations General Assembly) where international standards are debated, resolutions are adopted, and conventions are ratified. Support for this view might be found by analysing the application of international law in specific country contexts.

In 1948 the Universal Declaration of Human Rights asserted that “everyone, as a member of society, has the right to social security” (United Nations 1948). In 1952 the ILO’s Social Security (Minimum Standards) Convention 102 established the basis of modern social protection systems with nine branches of social security,⁸ including family benefit, healthcare, unemployment, and old age. In 1966 the International Covenant on Economic, Social and Cultural Rights “recognise(d) the right of everyone to social security” (Mpedi and Nyenti 2015).

Other international conventions affirm the right to social protection or social security for specific vulnerable groups, notably the Convention on the Elimination of All Forms of Discrimination Against Women (1979), the Convention on the Rights of the Child (1989), and the Convention on the Rights of Persons with Disabilities (2006). Most African countries have ratified these three conventions. More recently and most pertinently, in 2012 all member states of the International Labour Conference voted to adopt the Social Protection Floors Recommendation (R202), which advocates for a rights-based package that guarantees access to healthcare for all as well as income security for children, people of working age, and older persons (ILO 2012).

The United Nations appoints Special Rapporteurs who function effectively as policy pollinators, by drafting policy statements that clarify the commitments governments have made under international law and travelling to countries to verify that member states are fulfilling their obligations. In 2012 the United Nations Special Rapporteurs on the Right to Food and on Extreme Poverty and Human Rights co-authored a proposal for a “Global Fund for Social Protection” (de Schutter and Sepúlveda 2012). This document clarifies the content of the right to social protection in international law and the obligations of all member states to “respect, protect, and fulfil the right to social protection” (de Schutter and Sepúlveda 2012, 5) as comprehensively and expeditiously as possible, including by passing appropriate domestic laws.

Under the International Covenant on Economic, Social and Cultural Rights (ICESCR), States must devote their maximum available resources to the fulfilment of economic and social rights, including through the establishment of social protection systems. As recognised under the ICESCR, some dimensions of economic and social rights can only be achieved progressively over time. However, this cannot be invoked as a pretext for delaying action. (de Schutter and Sepúlveda 2012, 6)

An implicit theory of change explains how the globally affirmed human right to social protection might “cascade” down to the realisation of this right by individuals living in specific country contexts (Devereux 2017). First, a government representative signs or ratifies relevant instruments in global (e.g. Union Nations) or regional (e.g. African Union) forums. Next, the provisions of these global and regional instruments are codified in national position statements such as the Constitution or National Development Plan. Third, policies, programmes, and projects are designed and implemented to give effect to these provisions. Fourth, legislation is passed that gives these social protection interventions the status of a justiciable right. Finally, local civil society organisations campaign to hold the government accountable to deliver on the right to social protection.

In reality, this process plays out in very few countries. The road from ratification to implementation to enforcement is very long—in some cases, non-existent. Most global Declarations, Conventions, Covenants, and Recommendations are not legally binding. The United Nations Special Rapporteurs have no legal authority to enforce UN resolutions. No government has yet been prosecuted at the International Criminal Court for violating the right to social protection. At the national level, a right to social security is specified in about half of the constitutions in Africa (fourteen out of thirty examined by Fombad 2013). But in most cases, this right is not justiciable—it cannot be enforced. In South Africa, civil society has taken the government to court to uphold or extend the right to social protection. For example, an alliance of local civil society organisations successfully lobbied Parliament and the High Court to extend the age of eligibility for the Child Support Grant from seven to eighteen years, on the basis that eighteen years is the legal definition of a child in South Africa (Proudlock 2009).

But South Africa is an exception to the rule, for reasons related to its unique history (Devereux 2011). In many other African countries, civil society activities are curtailed and strictly regulated, and they have no freedom to campaign for economic, social, and cultural rights. This illustrates a critical point: that national governments are accountable to their domestic constituencies and are responsive to local political imperatives, rather than to declarations signed in global forums.

5 Understanding Resistance: Why Some Governments Say “No”

Political self-interest drives government policy processes, and this is the lens through which all decisions about national social protection policies should ultimately be analysed. Theories of change asserting that policy choices are evidence based, or reflect the domestication of international law, are not necessarily aligned with political realities. Governments need to be convinced that introducing a new policy and committing resources to new or scaled-up programmes will help them to win votes—in other words, policymaking is driven more by “what’s popular” with key political constituencies than by “what works” for the poor.

In this context, social protection is susceptible to both positive and negative politicisation (Devereux and White 2010). In democratic regimes with accountable governments, delivering benefits to poor people makes governments popular and earns them votes—a positive outcome for poor people and their governments. Negative politicisation occurs if governments manipulate targeting and eligibility criteria to ensure that benefits are disbursed not on the basis of need, but to their own supporters as a reward for their loyalty, or to opposition supporters to induce them to switch their vote in future elections. An example of positive politicisation comes from an election campaign in Lesotho, when opposition parties made a manifesto commitment to double the amount paid to pensioners and the ruling party responded by promising to raise the social pension payment, which they did after winning the election (Croome et al. 2007). On the negative politicisation side, donor agencies in Zambia who argued for rolling out the SCT from poorest to least poor districts, based on poverty headcount rates, faced pressure from politicians who wanted the programme to be launched in their districts first, irrespective of poverty rankings (Harland 2014).

African governments are located at the intersection between external pressure from international development agencies to implement specific policies and programmes in certain ways—exerted through hard or soft conditions on development aid—and internal pressure from constituents and local civil society lobbying on behalf of specific groups of citizens. In this context, the imperative driving policy choices are not necessarily how to reduce poverty rapidly, but political survival. One strategy that governments deploy to balance these competing priorities is to use “development partners” to finance programmes favoured by these agencies—such as social cash transfers—and to commit government resources to interventions that are more popular domestically. Malawi is a case in point. For many years international agencies advocated to scale up social cash transfers, building on a positively evaluated UNICEF-funded pilot project in Mchinji District (Miller et al. 2011). But the government preferred to allocate its funds to supporting farmers with subsidised fertiliser and seed, regarding this as a productive investment in a politically influential constituency. Eventually the social cash transfer did scale up to cover all districts, but mainly through funding provided by development agencies. In 2019, the World Bank, European Union, Germany, and Ireland supported the SCT programme in twenty-seven districts, while the government financed the SCT in just one out of twenty-eight districts (Government of Malawi 2019).

The Malawi case illuminates two important asymmetries between African governments and their development partners. One is political: donor agencies believe (and have invested resources to prove) that cash transfer programmes uplift poor people and that governments should therefore allocate their own resources to such spending. But governments do not necessarily regard marginalised groups among the poor as politically important constituencies. Their political interests lie elsewhere—in this case, with supporting farmers who grow the country’s food—a national strategic priority—and vote. The second asymmetry is about the choice of instruments. Despite years of policy advocacy for social cash transfers by the international development community, many governments remain unconvinced. Pejorative attitudes exist towards the poor and towards social cash transfers everywhere, not only in Africa. Free “handouts” from the state are believed to create “dependency syndrome” (Shepherd et al. 2011), laziness, and wasteful spending (e.g. on alcohol) among beneficiaries, who should instead be given support for productive livelihoods (e.g. agricultural inputs).

It follows that the reluctance of some African governments to implement cash transfers reflects careful political calculations, as well as genuine disagreements between national actors and international development agencies, about the optimal allocation of scarce public resources. From the perspective of policy pollination, these cases reveal that the efforts of international agencies to propagate social protection do not always succeed, even if they invest heavily in all the pollination strategies described above.

6 Conclusion

Many governments in sub-Saharan Africa have implemented social cash transfer programmes for their poor and vulnerable citizens during the last twenty years. This process has been actively promoted by international development agencies that have offered substantial financial and technical support, tied to the adoption of this specific form of social protection. Several plausible explanations for this policy diffusion process have been considered in this chapter, which complement rather than contradict each other.

The first explanation is that this is an evidence-driven policy process. Social protection has proven—through rigorous evaluations of cash transfer schemes—to achieve significant positive impacts on poverty and vulnerability. This accumulation of evidence convinced many African governments to implement, scale up, and ultimately pay for their national social protection programmes. A second possibility is that it is a political choice. Politicians adopt policies that are electorally popular and consolidate their power. Given this context, social protection is expanding throughout Africa because delivering cash transfers is amenable to both positive and negative politicisation. A third factor is that social protection is well established in international law, from the Universal Declaration of Human Rights in 1948 to the Social Protection Floors Recommendation in 2012. The adoption by African governments of national social protection policies, programmes, and laws might simply reflect the domestication of their legal commitments under international law.⁹

Four causal mechanisms are commonly discussed in the policy diffusion literature. In three of these—learning, competition, and emulation—policies are voluntarily chosen by adopting governments, while in the fourth—coercion—policies are imposed on reluctant governments by external actors, using their hard or soft power. This chapter has focused on a mechanism closely linked to coercion, which I call policy pollination. The argument is that social protection has been pollinated throughout sub-Saharan Africa, especially in low-income countries, by international development agencies, using their financial leverage and technical expertise to incentivise policy adoption by African governments and to institutionalise social protection in national policy structures.

The extent to which this policy process was coercively imposed on African governments, and the extent to which it was autonomously chosen with support solicited by these governments from their development partners, varies from country to country. Devereux and Kapingidza (2020, 298) propose that a social protection policy process can be characterised as being externally induced or pollinated by international actors, rather than being domestically driven and nationally owned, when the following conditions apply:

• 1. The policy process is led by external actors, through their staff and consultants.

• 2. Policy drafting and consultation processes are missing or are dominated by national and local elites.

• 3. External actors advocate for the same social protection instruments (e.g. cash transfers) across diverse country contexts.

• 4. An externally supported social protection project becomes the flagship national programme.

• 5. Evaluations of social protection programmes are commissioned by external actors and are conducted by international research institutes.

• 6. A high proportion of social protection spending is financed by external actors rather than the national government.

It could be argued that the mechanisms of diffusion of social protection policies and programmes across Africa are of secondary concern, and what matters is the outcome: that millions of poor and vulnerable Africans now have access to social protection, which is their human right. However, the way that social protection is introduced into a national policy discourse matters—which instrument? who benefits? who decides? who pays? In some countries, external actors appear to be more committed to social protection than national actors. The test for sustainability of social protection in such contexts will come when development agencies withdraw their financial and technical assistance, leaving each government to decide whether it is politically beneficial and fiscally cost-effective to continue prioritising this specific set of social policy instruments.