First Name: Tim
Department Dept of Economics
Supervisor(s): Dr. Tony Bryant , Dr. Roselyne Joyeux
Using panel cointegration methods to understand rising top income shares
Rising top income shares in developed countries have recently generated a great deal of political and economic discussion. While many theories have been proposed to explain this increase, there has been little robust econometric testing of those explanations. This paper seeks to understand top income shares using econometric modelling.
The paperís originality lies in the range of explanatory variables offered in the model, as well as the unique approach of modelling the top end of the income distribution using a panel time series dataset (allowing for cointegration econometrics).
Using a variety of data sources this paper will employ panel cointegration methods, including the FMOLS and PDOLS estimators, to understand the long run relationship between the top 1% income share and a number of explanatory variables.
Our analysis finds that economic openness, the size and ideology of Government, the development of financial markets, top marginal tax rates, and the strength of unions are all important determinants of top income shares. By contrast, there is no evidence to support the notion that skill-biased technological progress has led to rising top income shares.
Practical and Social implications
These results offer a comprehensive explanation of why top income shares have been rising in a number of developed economies in the last three decades. The implications on economic and social policy extend towards potential avenues for reform in addressing/counteracting the rise, depending on the degree to which policymakers and voters believe that this rise has adverse effects on the economy and wellbeing.