Three essays on democratic capital and per capita income.

摘要

The correlation between incomes and democracy is well documented, but the exact nature of the relationship and direction of causation, if any, is still being debated. The empirical conceptualization of democracy to include the accumulation over time of social and political infrastructure upon which economic policy is devised and implemented is a novel and promising approach. We present three essays that extend our understanding of the relationship between democracy, expressed as "democratic capital," and economic growth.;In the first essay, we put together a data set of various functional forms for democratic capital, with annotations, summaries, and the correlation among them, and we explore their properties as measures of democracy. We find that, compared to democratic capital based on a full scale of democracy index, democratic capital built on the binary form produces a better fit as a predictor of both the incidence of democracy and of incomes. This result suggests that the qualitative difference between democracy and autocracy (as opposed to the extent of democracy) has more explanatory power about the relationship between democratic capital and economic growth. The correlation between democratic capital and levels of income is positive and statistically significant but negative and statistically significant in relation to income growth. This result supports the notion that democratization is a feedback mechanism for higher incomes, not the initiator of it.;In the second essay, we drop the assumption of zero initial democratic capital as neither theoretically necessary nor reasonable in most if not all situations. We also vary the depreciation rate used in the accumulation of democratic capital. Using democratic capital based on the binary indicator of democracy, we estimate the relationship with levels of income per capita and examine the fit statics for various combinations. Our results show that the best fit combination of initial democratic capital and depreciation rate is one that assumes that countries enter the data set with a mostly autocratic history. This best fit initial democratic capital is, however, also greater than zero. Furthermore, we find the depreciation rate that produced the best fit estimates to be between three to six percent. Depending on the initial democratic capital--depreciation rate combination, the estimates of the coefficients on democratic capital vary from positive to negative and from no statistical significance to highly statistically significant estimates. Finally, the top ranked (by fit statistic) initial democratic capital--depreciation rate combinations have estimates for democratic capital that are positive and statistically significant in relation to levels of income but negative and statistically significant for income growth.;In the third essay we test the relationship between democratic capital and economic growth before and after countries have taken off into industrial growth. We use the best fit combination of underlying binary indicators of democracy, initial democratic capital (about a quarter of the maximum possible), and a depreciation rate (4%) obtained in the first and second essays as our preferred variable for democratic capital. We find that among countries that have taken off into industrial growth, electoral democracy has a negative relationship with income per capita before the takeoff into industrial growth in contrast to a positive one after. The relationship with annual Gross Domestic Product (GDP) per capita growth rate also changes but the results are not as robust. Our results were sustained for two sets of criteria for takeoff into industrial growth.;Taken together, our investigations provide further evidence for endogenous democratization. That is, as countries experience economic growth, the pressure for political liberalization also grows. That the estimates on democratic capital are positive after takeoff into industrial growth, but negative before, even while democracy has a generally slightly negative effect on economic growth may suggest that political reforms consolidate the economic gains of growth, but do not produce them.