Conclusion

The concluding chapter brings together the implications and insights from the four empirical chapters, reconsidering the theoretical claim in the light of this evidence. It also discusses the generalizability of the findings, both for Africa as a whole and for new democracies throughout the world. Finally, it attempts to situate the argument and findings in the context of an ever-changing Africa. In doing so, it considers the impact of urban–rural linkages, and the implications of rapid demographic changes taking place across the continent. In a continent that has been experiencing historically unprecedented rates of urbanization, how should we expect political incentives, and subsequent policies, to evolve? How might pro-rural development affect the nature of urban–rural relations across Africa? How in turn might this influence the nature of electoral competition? And finally, how much reason do we have to be optimistic about democracy in Africa?

Democracy’s Rural Dividend

This chapter evaluates the second and third hypotheses identified in Chapter 2, that the distribution of benefits resulting from the introduction of democratic electoral competition should be targeted primarily towards rural areas, and that the extent of any such urban–rural differences should be conditional on levels of urbanization. This is done using quantitative cross-national analysis of data from the Demographic and Health Survey Series. These household surveys provide incredibly detailed individual-level data on a range of essential education and health outcomes, across thirty African countries. This enables investigation of the extent to which these outcomes are affected by the introduction of electoral competition, and importantly, how these effects vary across urban and rural areas. Robust estimates demonstrate that democratic elections significantly increase access to primary education and reduce infant mortality rates, but only for children in rural areas. Moreover, these effects are conditional on the level of urbanization. The analysis provides no support for a series of alternative explanations, thereby strengthening confidence in the validity of the theoretical argument.

Introduction

This chapter introduces the book’s central argument, and discusses the contribution that it makes to the literature. This contribution is twofold. First, Rural Democracy acknowledges that in addition to strategies of electoral fraud, clientelism, and ethnic politics, African rulers have also attempted to compete for votes by providing public goods and services to their citizens. A discussion of the range of strategies that governments use to secure rural votes, the distinctions and interactions between them, and their relative significance, serves to frame the book’s argument, and sets scope conditions to bound its generalizability. The book’s second major contribution relates to the impact of democracy on development more broadly. By theorizing the effects of electoral institutions within the specific contexts in which they are embedded, Rural Democracy suggests a way forward for future research on the impact of democracy on development in Africa, and elsewhere also. In discussing these contributions, Chapter 1 clarifies the minimalist understanding of democracy adopted in Rural Democracy. Although the argument in Rural Democracy is that even minimally competitive elections generate incentives for incumbents to consider voters’ preferences when making policy, the extent to which elections create incentives for the costlier strategy of pro-rural development likely diminishes below some level of democratic quality, or fairness. Chapter 1 also introduces and justifies the book’s threefold empirical strategy. This includes a discussion of case selection, and a consideration of how the results might generalize beyond the particular cases that are included in the analyses.

Urban Incumbent Hostility

This chapter is the first of two that provide quantitative cross-national evidence to support the book’s central theoretical claim. Responding directly to the puzzle of urban incumbent hostility discussed in Chapter 2, this chapter evaluates the expectation that the lower proclivity of urbanites to support incumbents and express satisfaction with democracy should be conditional on levels of urbanization. More specifically, analysis of public opinion data from twenty-eight countries across Africa shows that urban hostility towards incumbents and dissatisfaction with democracy are not only robust to a variety of potentially confounding factors but are decreasing with urbanization. The data come from multiple rounds of the Afrobarometer Survey Series, enabling the use of fixed effects models to estimate robust within-country effects by leveraging changes to urbanization over time. These findings are important because, as highlighted in Chapter 2, the expectation that urban incumbent hostility should be conditional on urbanization is a unique observable implication of the book’s argument. Although scholars have put forward a variety of reasons for why urbanites across Africa should be less likely than rural residents to support incumbents, none of these alternative explanations account for the conditional effects demonstrated in this chapter. As such, this evidence supports the theoretical claim that electoral competition leads African governments to pursue pro-rural development policies.

Mechanisms: Historical Evidence from Botswana

The third stage of the empirical strategy considers a potential counter to the central theoretical argument; it is possible that rather than resulting from electoral incentives, any urban–rural differences in development may result from contemporaneous external forces, such as donor conditionality. If foreign donors encouraged both competitive elections and pro-rural policies, then the link between electoral competition and pro-rural development across Africa may be spurious. Chapter 6 addresses this possibility through a largely qualitative analysis of the historical case of Botswana in the period immediately following independence. This case is useful because uniquely in sub-Saharan Africa Botswana has held competitive elections consistently since gaining independence in 1966. Taking advantage of this, Chapter 6 explores the link between electoral competition and pro-rural development during a period when the international context was very different, and when no such external pressures to introduce competitive elections and rural development policies were likely to have been felt. Using archival resources such as original minutes from cabinet meetings and ministerial correspondence, the chapter traces the ruling party’s responses to electoral outcomes in the early post-independence period. Despite its dominant position, the ruling party responded strongly to losses of rural support with a major program of rural development policies. Interestingly, cabinet minutes document an explicit policy of prioritizing rural development projects that were both highly visible and likely to be completed prior to the next election, suggesting a strong role for electoral incentives. Alongside this archival evidence Chapter 6 also analyses the contents of the ruling party’s manifestos during this period to evaluate changes in policy emphasis during this period. Taken together, the evidence from this historical case strongly supports the proposed theoretical mechanism, suggesting that electoral competition does indeed lead to rural development in Africa.

Elections and Rural Development

Chapter 2 presents the theoretical argument underlying the hypothesis that the introduction of competitive elections across Africa results in pro-rural development. This argument is motivated by a puzzle: despite widespread acceptance that party competition in Africa is dominated by issues of ethnicity and clientelism, not by an urban–rural cleavage, urban residents across Africa are significantly less likely to support incumbents than their rural counterparts. How can this stark urban–rural electoral cleavage be accounted for? After reviewing relevant existing literature, Chapter 2 argues that democratic elections make African governments more responsive to rural interests. This follows from acknowledgement of the particular demographic contexts in which electoral institutions are embedded across Africa. Because a majority of Africans live in rural areas, competitive elections create incentives for governments to implement pro-rural policies to win the support of the rural majority, thereby generating dissatisfaction among urban voters. The argument rests on the assumption that voters in Africa condition their votes on the provision of public goods and services, at least in part. This assumption is crucial, because if it does not hold then there would be no reason to think that politicians should implement pro-rural policies in order to win rural votes. This argument implies an important conditional effect, that incentives to implement pro-rural policies should dissipate as the size of the rural majority decreases. Consequently, the extent of urban–rural differences in development outcomes resulting from electoral competition should be conditional on levels of urbanization. Taken together, the theoretical argument implies three key empirical implications: (1) across Africa, urban hostility towards incumbents and dissatisfaction with democracy should be conditional on levels of urbanization; (2) the distribution of benefits resulting from democratic electoral competition should be targeted primarily towards rural areas; (3) the extent of pro-rural targeting of the democratic dividend should be conditional on levels of urbanization.

Microfoundations: Voting for Public Goods in Ghana

Chapter 5 evaluates the core assumption underlying the argument that voters in Africa engage in evaluative voting and condition their support on the provision of public goods and services. This assumption is crucial because without it there would be no reason to think that politicians should implement pro-rural policies in order to win rural votes. Moreover, it is not obvious that the assumption holds, because existing work on the determinants of voter choice in African elections has focused predominantly on issues of clientelism and ethnic voting. In examining whether voters in Africa use elections to hold governments accountable for their performance in office, Chapter 5 narrows the focus to a single country, Ghana. Doing so accounts for important contextual variations in how goods are provided, which influence the extent to which responsibility for this provision can be attributed to political action. The analysis uses data on two types of goods to investigate whether Ghanaian voters are evaluative. First it looks at education, considering a range of inputs at the district level. Usefully, analysis of this data supports the idea that voters are more likely to condition their support on the provision of goods and services which can be attributed to political action than on those which cannot. Acknowledging limitations to inference resulting from aggregation to the district level, the chapter then turn to roads. Analysis of an original panel dataset containing detailed information on road conditions throughout Ghana provides even more robust evidence that electoral support is affected by the provision of public goods.