Until relatively recently, politics has not been accorded a significant role in thinking and policy-making around social protection in international development. In particular, welfare economics has tended to focus on how social protection can
correct market failures and play a redistributive role, and overlook important aspects of political economy (Casamatta et al 2000: 342) and domestic politics (Niles 1999: 3). Work on social protection in Africa has followed suit, arguing that the key problem
for social protection here is simply a lack of financial and administrative capacity. Even where institutional issues are taken seriously (e.g. Mathauer 2004), the role of national politics is seen as purely contextual, to be examined “for the sake of
completeness” (ibid: 16), rather than accorded an explanatory role. However, and in general terms, there “…is no economic law that prevents societies from deciding to allocate more resources to old-age security and less to some other expenditure”
(Beattie and McGillivray 1995: 68, cited in Devereux 2001: 22), and it could be argued that the greater the fiscal constraints, the greater the implications of political attitudes concerning who deserves support, and in what form (Graham 2002: 25). For
example, even where adequate financial resources and administrative capacity are in place, as in the case of Botswana, social protection measures failed to reach the poorest groups due to a lack of political commitment that has its origins in
Botswana’s political economy and related political institutional arrangements (Good 1999, de Waal 1997).
However, if it has become increasingly clear that ‘politics matters’ to the conception, implementation, success and sustainability of social protection – and issues of development more broadly (e.g. Houtzager and Moore 2003) – there is less clarity concerning the specific ways in which politics shapes social protection, or what forms of political analysis are required to understand these relationships. This is particularly the case with Africa,2 not least because if its relative paucity of sustained programmes of social protection.3 The most thorough attempt to understand the politics of social protection in developing countries describes such efforts as “initial work in a new area” (Graham 2002: 1), and notes that we know little about what kinds of social assistance systems are feasible or sustainable in contexts of high levels of poverty or inequality (ibid: 27). Most work on social protection in Africa makes little mention of political concerns. Work that does tends to consider only a
limited range of variables (e.g. political discourse only) or make fairly general references to ‘political commitment’ or ‘political support’, without analysing the basis of this, or examining how it emerged and might be sustained. Examining the politics of development is also associated with a broader set of methodological problems concerns of relevance here. For example, “It is not possible to observe the underlying motivations for a public policy” (Niles 1999: 20), leaving researchers to approximate this by measuring the geographic distribution of policies as against poverty levels and areas of political discontent, to determine economic efficiency and political expediency. ‘Political factors’ tend to be highly contextualised within particular polities, and efforts to abstract general patterns from this risk losing the sense of history and context that shapes the politics of the possible.
Given these constraints, any findings and conclusions concerning the politics of social protection in Africa are necessarily contingent on further theoretical elaboration and empirical testing. The following section (Section 2) examines the broad debates
concerning the links between politics and social protection in Africa, in recent historical perspective, and sets out the framework of analysis to be pursued here. Section Three fills out this framework, revealing and to some extent evaluating the
role that different forms of politics may play in shaping social protection. Section Four discusses the types of political impact that social protection programmes are associated with, before Section Five makes the case for thinking about the politics of
social protection in terms of ‘political contracts’. Sections Five and Six briefly map out tentative policy recommendations and future research themes that emerge from this analysis.
Thanks to Armando Barrientos for insightful and instructive comments on an earlier draft, and to Jenny Peterson, a doctoral student at IDPM, whose excellent bibliographic work provided the basis for this paper. This paper also draws on the author’s research into the politics of chronic poverty in Uganda, carried out over 2002-3 through the Chronic Poverty Research Centre.
Most work on the politics of social protection draws on non-African regions, particularly European, post-Communist and Latin American states. Such research offers useful conceptual guidelines and comparative data, but it is often difficult to directly extrapolate from such studies as they often contain implicit judgements that are of less relevance to Africa (e.g. the presumptions on class allegiance that to some extent underpin discussions if the political support required for social protection). Many are based on the assumption that spending on social protection derives entirely from domestic revenues.
For example, the only universal social pensions are found in Botswana, Mauritius and Namibia – South Africa and Senegal have means-tested versions. Over the 1980s, only Gabon and Seychelles introduced social protection systems in any serious way (Gruat 1990). Bendokat Tovo (1999) note that the state in Togo has offered a limited level of social assistance to vulnerable groups such as widows, orphans and the handicapped.