CLUSTERING DEMOCRACY AND INEQUALITY

This research paper applies a technique based on data symbolization of regimes to detect clusters in the relationship between Democracy and Income Inequality for a panel of countries during the period 1970- 2012. Our results are the following: two clusters are identified and for both of them the Granger causality is unidirectional, going from democracy to inequality for cluster 1 (consolidated democracies and low levels of economic inequality) and from inequality to democracy for cluster 2 (weak democracies and high levels of economic inequality); the ARDL econometric technique shows a U-inverse relationship between the variables, with threshold values of democracy over which in cluster 1 inequality starts to decline, whereas for cluster 2 levels of inequality above the threshold value cause the democratic levels to fall drastically. The implications of such empirical results are that in the first case an advanced level of democracy can imply a progressive reduction of inequality, but still the free market mechanism has to be implemented with policies on three pillars: fiscal, educational and institutional. In the second case, inequality becomes intolerable and socially disruptive, and allows democracy to deteriorate, with the need of gradual and selective reforms to create a sensitivity to the dynamics of change.