Income inequality and economic progress: an empirical test of the institutionalist approach

American Journal of Economics and Sociology, The,  Jan, 1996  by Kang H. Park

• E-mail
• Print
• Link

Z = c + [d.sub.1]Y + [d.sub.2]I + v [2]

where

Y = log of deaths from political violence, 1968-77

[X.sub.1] = the Gini index, various years

[X.sub.2] = log of population, 1975

[X.sub.3] = log of GNP per capita, 1975

[X.sub.4] = the separatism index, 1975

Z = the average growth rate of GNP per capita, 1968-77

I = the average growth rate of gross domestic capital investment, 1968-77,

and

u and v are error terms.

One of the major difficulties in the empirical analysis of institutional economics is the selection of appropriate indicators representing the variables that the institutionalist describe. Peach (1987) noted that

A reasonable argument can be made that institutionalists must be more actively involved in the process of developing alternative measures of distribution and in applying such measures to the available data. There are few institutionalist studies of income and wealth. Since institutionalism is a "fact-based" theory, more such studies should be encouraged.(9)

In response to Peach, this study investigates empirically the relationship between income distribution and economic progress, but is not intended to develop alternative measures of them. This study is limited to the use of currently available indicators regarding distribution and economic progress.

The most well-known compilation of political and social indicators was done by Taylor and Jodice (1983). They compiled various indicators of social and political instability such as armed attacks, deaths from political violence, riots, and political strikes of 155 countries for the period of 1948-1977. "Deaths from political violence" is chosen as a proxy variable for socio-political instability because this variable represents not only the incidence of political and social unrest but also the intensity and magnitude of such unrest better than other variables. The most recent 10-year time span 1968-1977 is used to observe deaths from political violence for each country. Given the general perception that income inequality is a fundamental rather than a precipitating condition for social and political unrest, the premise of this study is that a long time span should be considered to see the effect of income inequality on socio-political instability. Also in statistical terms, the distribution of the variable, deaths from political violence, moves closer to normal distribution when a 10-year time span is used.(10)

Taylor and Jodice also compiled various measures of income inequality such as the top 10% income share, the top 20% income share, the bottom 40% income share, and the Gini index.(11) The Gini index is used for this study because this index, despite its deficiencies, is still a better measure of inequality among the available indicators.(12) Jain (1975) also compiled income distribution data. Gini coefficients from Jain's data are correlated with Gini coefficients from Taylor and Jodice at .983 for a matching sample of 65 countries.(13) A factor limiting the sample size is the income inequality measure. While data on deaths from political violence are available for 155 countries, the data on income inequality are available for only 65 countries.