FAC Working Paper 43
by Colin Poulton
Theories of policy neglect of, or discrimination against, agriculture in Africa include urban bias (Lipton 1977; Bates 1981) and the narrow self-interest of autonomous elites (van de Walle 2001). Whilst structural adjustment removed much of the previous tax burden on African agriculture (Anderson and Masters 2009), the sector also saw declining investment from international development partners and through national budgets (Fan et al. 2009). Whilst there has been some recovery in public investment in agriculture over the past decade, signalled by the 2003 Maputo Declaration (Assembly of the African Union 2003), investment in the infrastructural and institutional public goods needed to support smallholder-led agricultural growth remains disappointing. As a result, the contribution of the agricultural sector to growth and poverty reduction objectives in Africa is widely believed to have been below potential.
In theory, democratisation, which has proceeded unevenly across Africa during the past two decades, should encourage pro-poor agricultural policy, as the majority of voters in many countries remain rural and poor. This paper draws on case studies of recent policy change (attempted and actual) in eight African countries, plus an analysis of the political systems in these countries, to explore the evolving role of competitive electoral politics in agricultural policy making. An important observation is that politicians are as likely to rely on ethnic allegiances and forms of social or political control to secure votes as they are to engage in policy competition. Moreover, the political incentives facing senior policy makers in the agricultural and rural development sphere may be inimical to the development of strong institutions to promote smallholder agricultural growth. Instead the paper finds that it is exogenous factors – macroeconomic dependence on agriculture and, most strikingly, sustained threats to regime survival – that create positive incentives for agricultural investment, even where social or political control is relied on to secure votes. The implications for participants in agricultural policy processes are briefly explored.