1. Dictatorship, Patronage and Public Good Provision
Karim Khan, PIDE, Islamabad
and
Anwar Shah, Quaid-i-Azam University, Islamabad

2. Introduction
Dictatorship has been taking dfferent shapes and is experienced by almost all of the existent civilization.
Dominant Coalition in Dictatorship with regulated economic competition and Rents
Persistence despite Undesirability, 68% of the world’s countries with non-democratic regimes during the last half of the 20th century (Deacon, 2009), Three-fourth of the countries in the world have experienced direct military rule since 1945 (Mulligan et al. 2004).
Different strategies of different dictators; however, military, violence and patronage are the most important instruments in sustaing the power.
Due to higher spending on patronage or targeted transfer, lower is spent on the provision of public goods.
Qualitative Evidence—countries with lack of legislature enroll only only 20% of their school age populations as compared to 81% in countries with effective legislatures (Deacon, 2009)--the proportion of children staying in school to the fourth grade fell from 81% to 72% and childhood disease immunization rates fell by more than one-half in Nigeria when the military rule came in 1983 (Deacon and Saha, 2005).
We test the hypothesis that dictators rely more on military for political support instead of relying on the wide cross-section of society--Consequently more spending on military as compared to the spending on the provision of public goods

3. Framework of Analysis
Our measure of patronage is military spending which, if our hypothesis was true, should be higher in dictatorships than in democracies.
Our measure of public good provision is secondary school enrollment rate which, given our hypothesis, should be significantly and negatively affected by the persistence of dictatorships.
In both cases, our major emphasis is on the explanatory power of dictatorship while controlling for a bunch of other possible explanatory variables.
Cross-sectional regression which is based on averaged annual data. Three justifications-imbalanced panel, persistence of instituions, endogeneity issue with some of the control variables.
Two different measures of dictatorships:
polity score which is based on Polity IV database (Marshall and Jaggers, 2000) which rates countries based on the degree of political competition, the openness and competitiveness of executive recruitment, and the extent of legislative and judicial constraints on the chief executive.
The second measure, denoted by dictatorship2, is based on Golder (2005) which measure regime type by a dummy variable where democracy takes a value 0 while dictatorship takes a value of 1.

4. Data and Summary Statistics

5. OLS Results for Secondary School Enrollment Rate

6. 2SLS Regression for SSE

7. OLS Results for Military Spending

8. 2SLS Regression for SSE

9. Conclusion
Motivation from the recent literature on institutions and economic development—absolutist institutions
dictatorship has a significant negative effect on the secondary school enrollment rate and a significant positive effect on the military expenditure as percentage of GDP
we conclude that dictators tend to rely more on military for political support instead of relying on the wide cross-section of the society.
More future research is needed to draw some general propositions for policy recommendations regarding institutional reforms in the third world countries.
We have taken a very narrow approach by indexing the dictators’ behavior towards patronage with military expenditure. Future work may develop an index for patronage that can capture the effects of targeted transfers both to the private interest groups as well as to the state’s privileged groups like military, bureaucracy, and the judiciary.
In addition, more econometric analysis is clearly needed in order to understand the exact channels of causation.