On the Probability of Military Rule, Africa 1970-2007

1. On the Probability of Military Rule, Africa 1970-2007 Raul Caruso, Catholic University of Milan Jacopo Costa, University of Florence Roberto Ricciuti, University of Verona and CESifo

2. Outline • Introduction • Models • Data and methodology • Results • Conclusions

3. Introduction • Since the start of the so-called third wave in 1974, and the acceleration after the fall of the Berlin Wall in 1989, democratization has been impressive. • According to the Economist Intelligence Unit (2011), one-half of the world’s population now lives in a democracy of some sort. More specifically, 12.3% lives in full democracies, 37.2% in flawed democracies, 14.0% in hybrid regimes, and still 36.5% in authoritarian regimes. • Authoritarian regimes, in which the military plays a direct (when a junta rules the country) and indirect role (when the army guarantees the monopoly of violence for a civilian despotic government), are the second largest regime group. A similar ancillary role can be found in hybrid regimes.

4. Freedom House (2009) reports that among the 48 countries of sub-Saharan Africa, 10 (21%) were rated Free in calendar year 2008, while 23 were rated Partly Free (48%) and 15 were rated Not Free (31%). • This compares with the results in 1980, when 4 were rated Free (9%), 15 were ranked Partly Free (32%), and 27 were ranked Not Free (59%), showing some improvements.

5. Theory • Acemoglu, Ticchi, and Vindigni (2010), Besley and Persson (2010): agency problem between the military and the government. • Main variables affecting military rule: income, external threat, ethnic fragmentation.

6. The economy consists of two groups: the rich elite and the citizens, distinguished by their incomes. • Democracy leads to redistributive policies, in particular, to the provision of public goods that are beneficial for the citizens and costly for the rich elite. • Starting from an oligarchy in which they hold power, the elite are unwilling to allow a transition to democracy. The only way they can prevent this is by creating a specialized unit of the state, the military, responsible for using force and repressing demands for democratization. • A powerful military also creates a political moral hazard problem because it can turn against the elite and take direct control of the government.

7. The elite have three potential strategies in oligarchy: no repression, thus allowing a rapid transition to democracy; repression, while also paying soldiers an efficiency wage so as to prevent military takeovers; and repression without significant concessions to soldiers, thus opting for nonprevention or facing the risk of a military takeover.

8. If a democracy inherits a large military from the previous nondemocratic regime, then it will also be confronted with a choice between making concessions to the military and facing a coup threat. • The decisive voter in a democracy wishes to prevent coups, but if soldiers realize that democracy will reform the military, reducing their rents. Since democracy cannot commit to not reforming the military, it can only make current concessions to soldiers and current concessions may not be sufficient to compensate the soldiers for the prospect of a military dictatorship. • Consequently, societies in which nondemocratic regimes in the past have chosen large militaries may have difficulty consolidating democracy and may end up with military dictatorships.

9. In an oligarchy, whether the elite prefer to set up a large military depends on the effectiveness of the military and on the extent of inequality. When the military is not very effective or inequality is limited, the elite prefer to allow a smooth transition to democracy, because such a regime will not be highly redistributive (while repression is likely to fail). • When military repression is likely to be effective and there is sufficient inequality, the elite may prefer to build a large military for repression and deal with the political agency problem by paying the military an efficiency wage.

10. Natural resources increase the political stakes because soldiers will be able to capture the natural resource rents if they take power. • As a result, natural resource abundance makes democracies more likely to fall to military coups.

11. Model and data • We estimated the following panel data random effect probit model: Militaryit= α1 + α2Xit+ α3Zit + α4Wit + α5Pit + α6Sit + α7Cit + εit • We consider 48 African countries over the period 1970-2007.

12. The dependent variable is a dummy equal to one in case the ruler is a Military and zero otherwise (Database of Political Institutions 2010). • The vector Xit includes GDP per capita, derived from Penn World Tables 6.3, and the added value of the agricultural, manufacturing and mining sector as percentage of GDP, using the UNCTAD database. • The vector Zit includes variables concerned with ethnic fragmentation, distinguishing between polarization and fractionalization (Reynal-Querol). • Wit is a vector including variables concerned with the external sector: openness, from the Penn World Tables 6.3) and the intensity of external treat, defined as level of hostilities on a 1-to-5 scale (Militarized Interstate Disputes 3.10).

13. The vector Pit includes the Agricultural Raw Price, taken from Free Market Price Index, and the Crude Oil Price, derived from Free Market Price Index (calculated as the average of Dubai/Brent/Texas equally weighted ($/barrel)) from UNCTAD. • Sit is a vector consisting of urban violence – using the “Urban Social Disturbance in Africa and Asia” data, reporting the unrest of social violence in the more relevant city into every state (PRIO). We have also considered the growth rate of population from Penn World Tables 6.3 and a dummy for if landlocked countries. • Cit is a vector of dummy variables describing the colonial rule of a country. • εit is a random error. • Continuous variables are in logs.

14. Existence of Military Rule, baseline results

15. Existence of Military Rule, baseline results 2

16. Existence of Military Rule, interactions

17. Existence of Military Rule, lagged independent variables

18. Conclusions • Income per capita negatively affects MR, with some nonlinearities. • Manufacturing and primary are positively related with MR. • Polarization and fractionalization have opposite effects (+/-). • Openness negatively affects MR.

19. External threat becomes significantly positive only when lagged. • Crude oil price negatively affect MR, whereas agricultural price are usually non significant.