Are lions democrats? The impact of democratization on economic growth in Africa, 1980–2010

Abstract

If we look back at the past two decades, timing seems to point to a close connection between democratic reforms and economic growth in sub-Saharan states. Most countries in the area introduced multiparty politics and made dramatic – if incomplete – democratic progress between 1990 and 1994. Quite strikingly, it is exactly from 1994 to 1995 (and particularly from 2000) that the region began to undergo a period of significant economic progress. Because of the undeniable temporal sequence experienced in the region – that is, first political reforms, then economic growth – some observers pointed to a nexus between democratic progress and economic performance. But is there evidence in support of a causal relationship? As of today, no empirical research has been conducted on the democracy–growth nexus in the early twenty-first century's so-called “emerging Africa”. To fill this gap, we discuss the different arguments claiming an economic advantage of democracies, we present our theoretical framework and carry out an empirical analysis of the growth impact of political regimes in 43 sub-Saharan states for the entire 1980–2010 period. Our findings confirm that African countries, many of which had long suffered the combination of authoritarian rule and predatory practices, derived some economic dividends from democratic progress.

Keywords:

• democratization
• economic growth
• Africa
• African politics
• political regimes

Funding

This work is part of a research project on “The economic, social and political consequences of democratic reforms: a quantitative and qualitative comparative analysis (COD)”, funded by a Starting Grant of the European Research Council [Grant Agreement no. 262873, “Ideas”, 7th Framework Programme of the EU].

Notes on contributors

Giovanni M. Carbone is associate professor of political science at the Department of Social and Political Sciences of the Università degli Studi di Milano. His research focuses on the consequences of democratization and on politics and economic development in sub-Saharan Africa.

Vincenzo Memoli is assistant professor of political science at the Department of Political and Social Sciences at the Università degli Studi di Catania. His research focuses on political behaviour and satisfaction with democracy, and quantitative methodology.

Lia Quartapelle is a former researcher (Africa desk) of the Istituto per gli Studi di Politica Internazionale (ISPI) in Milan. She is currently an elected member of the Italian parliament, with a particular focus on international relations and politics.

Notes

1. Easterly and Levine, “Africa's Growth Tragedy”; Artadi and Sala-i-Martin, “The Economic Tragedy of the XXth Century.”

2. Morten Jerven's recent works sparked a debate about Africa's growth data, as the reliability of GDP estimates is negatively affected by the weak capacity of national statistical offices, political interferences, and donors’ pressures. These are important issues and caveats for the study of the continent's economic performance. Yet their true implications are still unclear. As Fenske points out with direct reference to Jerven's analysis, for example, “correcting classical measurement error will often strengthen a result by removing attenuation bias, rather than weakening it,” and it is quite unlikely that such corrections would overturn major research findings. See Jerven, Poor Numbers; Fenske, “The Causal History of Africa,” 126. See also Young, The African Growth Miracle.

3. McKinsey Global Institute, Lions on the Move; Radelet, Emerging Africa.

4. Ndulu and O'Connell. Policy Plus; Radelet, Emerging Africa.

5. Wantchekon, “Democracy and African Development,” 197.

6. Sirowy and Inkeles, “The Effects of Democracy on Economic Growth and Inequality.”

7. For an overview, see Tavares and Wacziarg, “How Democracy Affects Growth,” 1344ff.

8. Abdiweli and Said, “Political Freedom and the Stability of Economic Policy”; Przeworski et al., Democracy and Development; Rodrik and Wacziarg, “Do Democratic Transitions Produce Bad Economic Outcomes?”; Quinn and Woolley, “Democracy and National Economic Performance.”

9. Tavares and Wacziarg, “How Democracy Affects Growth”; Gasiorowski, “Democracy and Macroeconomic Performance”; Oatley, “Why is Stabilization Sometimes Delayed?”; Serieux, “Democratization and Structural Adjustment in Sub-Saharan Africa”; Przeworski et al., Democracy and Development.

10. Barro, “Democracy and Growth”; Brunetti, “Political Variables”; Plumper and Martin, “Democracy, Government Spending, and Economic Growth.”

11. Sirowy and Inkeles, “The Effects of Democracy on Economic Growth and Inequality,” 150.

12. Pinto and Timmons, “The Political Determinants of Economic Performance”; Tavares and Wacziarg, “How Democracy Affects Growth,” 1344.

13. Plumper and Martin, “Democracy, Government Spending, and Economic Growth.”

14. Kriekhaus, “Democracy and Economic Growth”; Kriekhaus, “The Regime Debate Revisited.”

15. Sirowy and Inkeles, “The Effects of Democracy on Economic Growth and Inequality”; Przeworski et al., Democracy and Development; Doucouliagos and Ulubaşoğlu, “Democracy and Economic Growth.”

16. Nkurunziza, and Bates, Political Institutions and Economic Growth in Africa.

17. van de Walle, “Economic Reform in a Democratizing Africa”; van de Walle, African Economies.

18. Ferree and Singh, “Institutional Duration and Growth in Africa,” 32.

19. Narayan, Narayan, and Russell, “Does Democracy Facilitate Economic Growth.”

20. Tiruneh, “Regime Type and Economic Growth in Africa,” 14.

21. Lewis, “Growth without Prosperity in Africa.”

22. Rodrik and Wacziarg, “Do Democratic Transitions Produce Bad Economic Outcomes?,” 54.

23. Ferree and Singh, “Institutional Duration and Growth in Africa,” 42.

24. Doucouliagos and Ulubaşoğlu, “Democracy and Economic Growth,” 73.

25. Artadi and Sala-i-Martin, “The Economic Tragedy of the XXth Century.”

26. Fosu, “Democracy and Growth in Africa.”

27. Lewis, “Growth without Prosperity in Africa.”

28. Kriekhaus, “Democracy and Economic Growth.”

29. Ibid.

30. Gerring et al., “Democracy and Economic Growth.”

31. Ferree and Singh, “Institutional Duration and Growth in Africa.”

32. Bratton and van de Walle, “Neo-Patrimonial Regimes and Political Transitions in Africa”; Bach, “Patrimonialism and Neopatrimonialism.”

33. Jackson and Rosberg, Personal Rule in Black Africa.

34. Joseph, “Class, State and Prebendal Politics in Nigeria.”

35. Bates, “State Failure.”

36. Bates, When Things Fell Apart, 52.

37. Bach, “Patrimonialism and Neopatrimonialism,” 276.

38. Jackson and Rosberg, Personal Rule in Black Africa, 424.

39. Data in this paragraph are calculated from the authors’ Leadership Change Dataset. While our dataset also includes instances of “interim leaderships” (that is, African leaders in office for less than 12 consecutive months), we do not count them in the data presented in this article.

40. Carbone, “Do New Democracies Deliver Social Welfare?”

41. Stasavage, “Democracy and Education Spending in Africa”; Stasavage, “The Role of Democracy in Uganda's Move to Universal Primary Education”; Kjær and Therkildsen, “Elections and Landmark Policies in Tanzania and Uganda.”

42. Cassani and Carbone, Citizens’ Wellbeing in Africa's Competitive Authoritarian Regimes.

43. A score of 7 or more on the Polity scale is often taken as the threshold necessary to classify countries as democratic. Yet the very authors of Polity IV at times lower this threshold to 6 points. Marshall and Cole, for example, explain that “countries with Polity scores from +6 to +10 are counted as democracies in tracking ‘Global Trends in Governance, 1946-2010’” – see Marshall and Cole, Global Report 2011. We believe that in the African context, which has been historically dominated by nondemocratic rule, a score of 6 signals a remarkable democratic achievement.

44. Data on the level of schooling (a variable much used in the literature) include many missing observations and have therefore been discarded from the analysis.

45. Easterly and Levine, “Africa's Growth Tragedy”; Alesina et al., “Fractionalization.”

46. Ndulu and O'Connell. Policy Plus.

47. Following the economic growth literature on Africa, we experimented with various standard growth regressors, including investments, trade openness, conflicts, landlockedness, or decade. However, these variables present serious problems of multi-collinearity.

48. Steenbergen and Bradford, “Modeling Multilevel Data Structures.”

49.  See Greene, Econometric Analysis.

50. Since empirical evidence supports our main hypotheses when we employ our preferred RE model (see below in the text), we decided to also re-run our analyses using a fixed-effects (FE) regression model as an additional robustness check. While, as we already pointed out, this forced us to drop three theoretically important variables that we know to be invariant with regard to any specific country (namely, the log of GDP at 1979, the degree of ethnic fractionalization, and the British colony variable), we wanted to see whether our findings hold even after controlling for all the unobservable country-specific time-invariant factors that may affect the relationship under examination. The decision to verify our results by re-running the analyses with an FE model is in line with the outcome of the Hausman test, which suggested we employ a FE model. The empirical results of the FE model confirmed the positive impact of the degree of democracy variable – that is, our primary hypothesis – both when we look at GDP growth as well as when we look at GDP growth per capita, whereas we do not find evidence of a link between democratic duration and economic performance.

51. When, following suggestions by anonymous reviewers, we tested our model by including a debt relief variable, as well as by excluding the CFA Franc countries from our sample, we found that the variable central to our analysis – the level of democracy – remained significant (not so the duration of democracy variable). When we ran the analysis replacing “ethnic fractionalization” data with an “ethnic polarization” measure we found that the ethnic polarization variable was never significant in directly explaining Africa's growth.

52. Ndulu and O'Connell, Policy Plus; Radelet, Emerging Africa; Aryeetey et al., The Oxford Companion to the Economics of Africa.