Urska Zagar posted two very interesting empirical articles (here and here) about the connections between corruption, economic freedom and economic welfare. The first two indicators are qualitative while the third one is rather easily measurable. In a sample of 50 advanced, developing and least developed countries, she found a positive and robust correlation between corruption perception and GDP per capita, meaning that higher GDP per capita, on average, reduces the scope of corruption. The relationship between corruption and economic development is in fact even more intriguing than empirical figures suggest.
The impact of corruption on economic growth is an important theoretical and empirical theme in the economic literature. In this theoretical and empirical post, I will briefly review the main literature on corruption and development and discuss the empirical studies.
I would firstly refer to the article by Rodrik, Subramanian and Trebbi (2002) where the authors discuss the primacy of institutions for economic development over the exogenous factors such as geography (link). There has been an extensive amount of literature on the role of geography in economic development. The empirical strategy is quite simple. Usually, the basic equation includes the list of explanatory variables such as distance from equator, the percentage of land used for agricultural production and so forth. For example, the basic equation might take the following form:
log(GDP)=b1+b2(Dist)+b3(Land)+b4(Dummy_SubAfrica)+e
where GDP stands for GDP per capita, Dist for distance from equator, Land for the percentage of fertilized land and Dummy_SubAfrica is a dummy variable, taking the value of 1 if a country is located in the Subsaharan Africa and 0 if otherwise. In addition, e represents stohastic error unexplained by the selected predictors.
However, the basic problem with geographic approach to explaining economic development is that the approach does not, by itself, distinguish between the endogenous features of economic development. In regressing GDP per capita on the distance from equator, the empirical estimates usually result in a modestly negative correlation between the two variables. However, the distance from the equator cannot itself explain the nature of economic development and its significance over time mainly because of its low predictive power in explaining the evolution of institutions and governance.
An interesting approach has been incorporated by Acemoglu, Johnson and Robinson (2001). They incorporate a historical and institutional perspective in the empirical framework of the explanation of economic development. The authors used mortality rates of colonial settlers to explain the institutional quality. They further argue that where settlers encountered few health hazards compared to European settlement, they established solid institutions, strong enforcement of property rights and a robust system of law. In other areas where health hazards frequently occured, colonizers focused on the extraction of natural resources and showed little or no interest in building high-quality institutions. Rodrik, Subramanian and Trebbi (2002) further estimated the impact of geographic variables, institutions and openness on macroeconomic variables.
The main findings in their study were the following. First, each degree distance towards equator, on average, reduces the income per capita by 0.94 percent. The geographical location by the equator is also negatively related to capital per worker (-1.68), human capital per worker (-0.25) and total factor productivity (-0.32). Second, the quality of institutions is a very good measure of the economic welfare. In their panel, the authors found that each additional improvement in the rule of law, on average, leads to 2.22 percent increase in income per capita. The result is statistically significant at 1 percent. The improvement of institutional quality is both positively related and statistically significant when capital per worker, human capital per worker and total factor productivity are regressed on the institutional quality.
Back in 1997, Peterson Institute published an extensive study, searching the causes of corruption (link). The author estimated that the highest cost of corruption in regressing a series of macroeconomic variables on corruption index is lower education expenditure and a decrease in private investment. The estimate of the total macroeconomic cost of corruption is a difficult and daunting empirical task.
The main causes of corruption are mostly endogenous and related to the institutional evolution ranging from legal origins, colonial historical variables to the public sector efficiency variables. In search of the grain of truth, it should be noted that capturing the stylized effect of corruption of economic growth and development requires an interactive empirical approach. In other words, it would be impossible to establish "cause-and-effect" connection between corruption and development indicators without extracting a large amount of historical data and regressing it on the key endogenous variables.
Clearly, the role of geographical characteristics should not be neglected. However, the primacy of institutions in explaining the contemporary patterns of economic development is rather undisputable. Daron Acemoglu recently wrote an in-depth article on the distribution and explanatory factors regarding world poverty (link) where it clearly stressed the lack of institutions of human capital in the evolution of world poverty. When discovering the true causes of corruption, there should nonetheless be a clear and distinct rationale underlined by the evolution of institutional quality over time as the most significant measure in explaining the evolution, causes and effects of corruption on economic development.
No comments:
Post a Comment