Democracy and lived poverty in Africa

Abstract

ABSTRACTPast research on the relationship between democracy and poverty in Africa has produced surprisingly mixed findings. We argue that one source of variation is in the measures of democracy and poverty used by prior studies, which capture different concepts and contain different amounts of error. Using measures that map closely onto theory and which are directly comparable across countries, we show that electoral democracy is robustly correlated with small-but-significant reductions in lived poverty over time. We provide additional quantitative and case study evidence that accountability encourages governments to take swift action on poverty. Finally, we show that our results are sensitive to measurement choices, helping to explain null results in prior literature. Altogether our results suggest that empowering the poor in Africa will ultimately lead to meaningful reductions in poverty.KEYWORDS: Africapovertydemocracyaccountabilityanti-poverty programmesAfrobarometer Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 Harding and Stasavage, “What Democracy Does.”2 Carlson, “The Relevance of Relative Distribution.”3 Keefer, “Clientelism, Credibility and the Policy Choices of Young Democracies.”4 Arriola, “Capital and Opposition in Africa.”5 We intentionally omit a large number of studies that use at their dependent variable either public goods provision or objective proxies of poverty such as night lights. Most of these studies come to the conclusion that democracy reduces poverty. We exclude them primarily because our discussion of survey-based measures does not speak to them or their validity.6 Lake and Baum, “The Invisible Hand of Democracy.”7 Zweigel and Navia, “Democracy, Dictatorship and Infant Mortality.”8 Ross, “Is Democracy Good for the Poor?”9 Garcia, “Democracy is Good for the Poor.”10 Rosenberg, “Political Economy of Infant Mortality.”11 Tebaldi and Mohan, “Institiutions and Poverty.”12 Djeneba, “The Quest for Pro-poor and Inclusive Growth.”13 Saha, “Legislative Democracy, Economic Growth and Multidimensional Poverty.”14 Khodaverian, “The African Tragedy.”15 Ramos, Flores, and Ross, “Where has Democracy Helped the Poor?”16 Wullert and Williamson, “Democracy, Hybrid Regimes, and Infant Mortality.”17 Such as Bueno de Mesquita, et al., The Logic of Political Survival or Siegle, Weinstein, and Halperin, “Why Democracies Excel.”18 Dahl, Polyarchy.19 Vaccaro, “Comparing Measures of Democracy”; Casper and Tufis, “Correlation versus Interchangeability.”20 Beegle et al., “Methods of Household Consumption Measurement.”21 Kudamatsu, “Has Democratization Reduced Infant Mortality.”22 For Round 5, which we use as a control for Round 6, we generate this measure ourselves by averaging the five component measures. In Rounds 6–8, this calculation is already done and reported with the rest of the data.23 We lag democracy one year to ensure any changes in democracy occurred before poverty was measured.24 These values are estimated from bivariate regressions at the country level using data averaged over the entire survey period, for a sample of 32 countries.25 We choose a subset simply to make the figure less crowded and easier to read.26 We also include fixed effects at the level of the survey round, to account for any changes in survey protocols as well global shocks that might affect poverty rates for all countries in a given round.27 Country-rounds for which there was not a Afrobarometer survey the round before are dropped.28 Lake and Baum, “The Invisible Hand of Democracy.”29 Matschke, “Are there Eelection Cycles in Wage Agreements?”; Milas, “Employment, Output and Political Business Cycle Effects”; Nordhaus, “The Political Business Cycle.”30 The EFW index is produced by the Fraser Institute, which promotes the conclusions of classical liberal economics. Their scoring reflects the organization’s ideals, with better scores assigned to policies consistent with limited government. However, the values in the index are not themselves ideological and are compiled from various existing and widely-accepted datasets.31 Bratton and van de Walle, Democratic Experiments in Africa, 133.32 Budjan and Fuchs, “Democracy and Aid Donorship.”33 Owen and Mitchell, “Why Do Democracies Attract?”34 Escriba-Folch, Meseguer, and Wright, “Remittances and Democracy.”35 Scott and Steele, “Sponsoring Democracy.”36 Miles, “Deploying Development to Counter Terrorism.”37 Budjan and Fuchs, “Democracy and Aid Donorship.”38 Carling, “The Determinants of Migrant Remittances.”39 The World Bank data on resource flow are likely more comparable than data on poverty because they are being reported by financial institutions or sending countries with high statistical capacity.40 Specially we control for whether subjects believe the survey was being conducted by the government. Subjects may report higher poverty to the government if they are trying to get assistance or avoid taxation (Carlson and Fidalgo, “When A Null Pocketbook Effect is the Artifact”).41 The headcount data are binary and therefore it would be appropriate to use logit for this analysis. We continue to use regression for ease of interpretation and to make the results directly comparable to those in Table 2.42 Dörfell and Freytag, “The Poverty Effects of Democratization”; Gao and Yang “Is Democracy Pro-poor?”43 Brown and Hunter, “Democracy and Social Spending”; Lake and Baum, “The Invisible Hand of Democracy”; Stasavage, “Democracy and Education Spending”.44 The EFW gives higher scores to smaller governments. We invert the measure so that larger governments have higher scores.45 For in-depth discussions of these democratic erosions see Hirshel-Burns, “Benin’s King of Cotton” and Kohnert and Preuss, “Benin's Stealthy Democracide.”46 Samson and Houessou, “Lived Poverty in Benin,” 3.47 The Round 8 surveys were administered in Benin in November-December 2020, which fell during the first year of the Covid 19 pandemic. Thus, it is likely that the economic challenges during the pandemic contributed to some of the reported increase in lived poverty in Benin during this survey period. However, Benin took a much “softer” approach to Covid-19 restrictions than many other countries by never instituting a full-scale country-wide lockdown. Instead, the government implemented “sanitary cordon[s] around the affected areas with free movement of the population within it” only from March 30 to May 11. In addition, while the country closed bars and clubs, supermarkets and, notably, restaurants remained open. (See Osseni, “Benin Responds to Covid-19”). Thus, we would expect the impact on livelihoods (and thus, poverty) to be lower in Benin than it was in other sub-Saharan states that adopted stricter Covid-19 restrictions. Yet, Benin saw the sixth highest increase in lived poverty between rounds 7 (2016/7) and 8 (2020/1) out of the countries included in both rounds, rising .28 points compared to a 32-country average increase of .13 points. (See Mattes and Patel, Lived Poverty Resurgent, 7).48 Presidency of the Republic of Benin, Government Action Programme, 2016–2021 Summary.49 Paul et al., “Budgeting Challenges,” 2.50 Presidency of the Republic of Benin, Government Action Programme, 2016–2021 Summary, 25.51 Noret and Yedji, “Neoliberalism and Apathy in Benin”.52 Ibid.53 Ibid.54 Sackey, “Ghana's LEAP Programme.”55 Handa et al., LEAPImpact Evaluation; Palermo et al, Ghana LEAP 1000 Programme.56 Merttens, Uganda Social Assistance Grants for Empowerment.57 Dietrich et al., “Human Capital Returns to Cash Transfers.”58 Abdulai, Rethinking Eelite Committment to Social Protection, 20.59 Ibid., 30.60 Ibid., 31.61 Mugisha and Kitamirike, “Two Decades of Growth,” 87.62 Vokes and Wilkins, “Party, Patronage or Coercion,” 589.Additional informationNotes on contributorsElizabeth CarlsonElizabeth Carlson is an assistant professor of political science and public policy at NDSU and a faculty scholar at the Sheila and Robert Challey Institute for Global Innovation and Growth. Her research focuses on political behaviour and the political economy of development in Africa. Her work has appeared in World Politics, Comparative Political Studies, Journal of Development Studies, Political Science Research and Methods and elsewhere. She holds a PhD from UCLA.Kristin McKieKristin McKie is an associate professor of government and African studies at St. Lawrence University. Previously, she has been a postdoctoral associate in the Program on Democracy at Yale University and a Fellow at the Kellogg Institute at the University of Notre Dame. Her research interests include the politics of presidential term limits, political institutions, and the development of the rule of law across sub-Saharan Africa. She holds a Ph.D. from Cornell University.