Why reforms fail: Political-economy forces and agriculture in Africa

Ataman Aksoy, Bernard Hoekman

15 May 2013



There are many hypotheses on why some nations fail and others become successful (see Acemoglu and Robinson 2012). While the debate rages on, an area of agreement is that the strength of institutions and their ability to adjust to shocks is an important factor. Why the two sides of the US-Mexico border in the town of Nogales evolved so differently is in some ways similar to the question why the coffee and tea sectors in Kenya, located in the same geographic areas and subject to similar reforms implemented at the same time, evolved so differently. A recent study, African Economic Reforms: Role of Consensus and Institutions (Aksoy 2012), asks exactly that question. It uses a combination of comparative analysis and case studies to analyse why a similar set of policy reforms affecting nine export crops in five countries in eastern and southern Africa with similar historical backgrounds had very different outcomes.

The study finds that political-economy forces are a major determinant of observed outcomes. While in almost all instances agricultural reform generated an initial positive supply response, the degree to which these were sustained following international price shocks depends on the extent to which existing elites were able to share in the benefits accruing to producers and the overall significance of a crop for the economy. This research helps to inform recent debates on the importance of ethnic diversity and inequality as a determinant of economic development. Thus, Alesina et al (2013) argue that ethnic inequality may be an important feature of African underdevelopment. Conversely, however, in the case of African agriculture Anderson and Bruckner (2011) argue that there is a significant within-country effect of policy distortions on economic growth, suggesting that the strong ethnic divisions that characterise many sub-Saharan African countries cannot be a major driver of performance because ethnic inequality is largely time-invariant. The contributions to Aksoy (2012) provide evidence that in the case of agriculture ethnic diversity and inequality per se is not a key determinant of outcomes; what matters is how the benefits of reforms are distributed.

African agricultural policy reform in the 1990s

As of the 1980s, most African countries were losing market share in their primary export crops, which were highly taxed by the state, either directly or de facto through marketing arrangements. Most of the public agencies tasked to provide non-price support to agriculture through better seeds, new technologies, better transport, and credit ended up taxing agricultural production through interventions in marketing and processing. Agriculture was also taxed indirectly, as the pattern of policy was skewed towards supporting industry and urban elites. Agricultural policy reforms initiated during the 1990s, along with programs to liberalise economic activity more generally, aimed to significantly reduce export restrictions and taxes, privatise paparastatal marketing monopolies, and restructure smallholder support systems. The core objective was to increase the rate of return for producers. The agricultural reforms coincided with political reforms. In Zambia, Kenya, and Tanzania, long time single-party systems were abandoned and multi-party systems introduced. In Uganda and Mozambique, civil war had ended and attempts were made to also move to multi-party systems. While many of the old leadership stayed in power, new actors and groups obtained some voice and influence.

The reforms were significant: they reduced average taxation of industrial crops in sub-Saharan Africa by some 30 percentage points (Anderson and Masters 2009). Studies evaluating the reforms have found that agricultural supply responses were uneven (Akiyama et al. 2001, Kherallah et al. 2002, Baffes and Garner 2003). Many have interpreted the lack of sustained supply response as nonresponse; others argue that a lack of complementary inputs, especially rural infrastructure, is a key factor.

Managing re-distributional tensions and external shocks

Any assessment of agricultural reform must consider subsequent international price developments and the evolution of the real exchange rates that determine producer prices. Higher export prices post-reforms obviously will provide a better environment than one characterised by global price collapses. Following the agricultural reforms in the early 1990s international commodity prices increased, helping to create a general aura of success. But prices collapsed around 2000, which contributed to growth reversals and the negative evaluations of the reforms undertaken in the early 2000s.

A key question then is what explains success and failure of countries to manage adverse price developments in the post-reform period. Aksoy (2012) provides a framework that ties the sustainability (success) of reforms to the degree of consensus on the reform programmes and the ability of the institutional arrangements established after the reforms to accommodate unexpected or negative shocks. This framework builds on Rodrik (1999) which differentiates between short-run growth accelerations caused by elimination of some constraints and sustained growth episodes that require better and more flexible institutional structures. The hypothesis is that social conflict and the inability to respond successfully to shocks is greater: 

  • When there is substantial disagreement on the initial policy changes.
  • The products affected are important in generating rents that are a major source of income for elites.
  • The beneficiaries from reforms are not connected to the ruling groups.

In instances in which serious disagreements arise over reform policies, shocks that jeopardise gains from the reform face the criticism of those groups that objected to the policy changes. This is followed by attempts to return to the pre-reform environment. This controversy creates uncertainties for producers and reduces the probability of a successful adjustment to the original reform programme. In cases in which agreements were reached on the policy changes, adjustment to a negative shock is easier. The rent-seeking fight is not as intense, and producers’ income increases do not seriously threaten the incomes of groups supported by the government. As important, a precondition for successful reforms is that enough political and economic space is available for the parties to make adjustments to unexpected adverse international developments. In addition, cases of good institutions, institutions that can adapt to changing conditions are associated with less conflict and less losses by the elites.


Research suggests that an important determinant of reform success is the ability of societies to address the distributional impacts coming from an increase in producer prices (see the case studies in the Annex). Ethnicity plays a role insofar as this is correlated with the ruling regime, but what appears to matter more for sustained performance is that the reforms do not threaten politically powerful groups with significant income (rent) losses. Greater political/social consensus on the nature of the reforms helps to sustain them and to manage the impacts of post reform exogenous shocks. It also helps to develop more efficient institutional structures that help to mitigate shocks.


Acemoglu, D and AJ Robinson (2012), Why Nations Fail: The Origins of Prosperity, Power and Poverty, New York, Crown Publishers.

Akiyama, T, J Baffes, D Larson, and P Varangis (2001), Commodity Market Reforms: Lessons from Two Decades, Washington, DC, World Bank.

Aksoy, M Ataman (2012), African Agricultural Reforms: The Role of Consensus and Institutions, Washington, DC, World Bank.

Alesina, A, S Michalopoulos and E Papaioannou (2013), “Ethnic Inequality”, VoxEU.org, 4 February.

Anderson, K and M Brückner (2011), “Price distortions slow economic growth in sub-Saharan Africa”, VoxEU.org, 7 October.

Anderson, K and W Masters (2009), Distortions to Agricultural Incentives in Africa, Washington DC, World Bank.

Baffes, J and B Gardner (2003), “Transmission of World Commodity Prices to Domestic Markets under Policy Reforms in Developing Countries”, Journal of Policy Reform 6(3), 159-80.

Kherallah, M, C Delgado, E Gabre-Madhin, N Minot and M Johnson (2002), “Reforming Agricultural Markets in Africa”, Food Policy Statement 38, Washington DC, IFPRI.

Rodrik, Dani (1999), “Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses”, Journal of Economic Growth 4, 385–412.

Annex: Country case studies

The cases analysed in Aksoy (2012) illustrate how political economy forces determine outcomes. One is the differential performance of tea and coffee sectors in Kenya and Tanzania. Both were the main sources of rent allocation and especially in Kenya, the ruling tribe (Kikiyu) was the major producer of coffee. In Kenya, the new president was from another tribe (Kalenjin) and sought to undermine the political base of the past president by eliminating their main source of income. As a result, reform programs failed and coffee output stagnated. In Tanzania, cooperatives extracted rents from farmers through long-term contractual arrangements and acting as a monopoly. There was little systemic support for the sector, and output stagnated. These two coffee cases illustrate the impact of a political rent fight over an important sector for each economy. The reforms could not overcome the conflict among the groups.

In the tea sector, multinational companies (estates) were strong and created an umbrella under which smallholders received some protection. In Kenya, the tea sector, despite being owned historically by the Kikiyu tribe, was spared because smallholder expansion took place in the region dominated by the new president’s tribe. In conjunction with the strength of the multinationals, this supported a positive outcome post-reform, despite attempts to eliminate an important source of income for the Kikiyu tribe. In Tanzania, the tea reforms were the least contentious among all commodities, in part reflecting the strength of the multinationals as well as the lack of involvement of the political elites in this sector. Most tea estates were not nationalised in contrast to other commodity estates. In this sector reforms were implemented efficiently and the sector performed well.

Like coffee, reforms in the cashew sector proved unsuccessful in both Mozambique and Tanzania. These two countries supplied almost 70% of the world’s cashews in the 1970s. In both countries cashew output collapsed in the 1980s and the countries accounted for less than 7% of world exports and production in the early 1990s. Nationalisation of processing, along with ‘villagisation’ (ujamaa) in Tanzania and civil war in Mozambique, contributed to this outcome. Both countries initiated reforms in the early 1990s in highly contentious environments. Cooperatives that extracted the rents in Tanzania were connected to the ruling party and were its extension in rural areas. In Mozambique, private processors were connected to a political leadership that opposed reforms and wanted to maintain a monopsony on exports. After the reforms, significant increases occurred in producer prices and output doubled in both countries. This expansion came to an end when international prices collapsed around 2000. Local and export taxes were increased in Tanzania to extract more of the rents, while export taxes were maintained at a high level in Mozambique. Today both countries still maintain high export taxes, and government affects the marketing through a warehouse marketing system in Tanzania and through the National Cashew Institute in Mozambique. As of end 2008, output in Mozambique was barely at the 2000 level; in Tanzania, it was far below that.

Two contrasting successful reform examples are tobacco in Tanzania and cotton in Zambia. In both cases parastatals were privatised and replaced by international companies. In both cases, the commodity was small in relation to the economy, and there was little conflict about reforms. The parastatals were bankrupt and their limited staff was absorbed by the companies, which supplied inputs and extension services and created a ready market for output under quasi-monopsonistic conditions. Output increased in both cases, and quality improved despite declining producer prices. In both cases, the companies faced crises due to side-selling, mostly caused by low producer prices and exchange rate fluctuations. However, ways were found to overcome the shocks without interference from governments, and output rebounded dramatically.



Topics:  Development

Tags:  growth, Africa

Consultant in Development Economics Department of the World Bank

Professor and Director of Global Economics, Robert Schuman Centre for Advanced Studies, European University Institute; Research Fellow, CEPR