The Gini and the tonic: Understanding the dynamics of inequality measurement
LSE Sir Arthur Lewis Building (SAL.LG.04) and on Zoom
Refreshments served at 12.15pm before seminar commences at 12:30pm.
Speaker:
Professor Sanghamitra Bandopadhyay, Professor of Development Economics and the Deputy Director of the Centre for Globalisation Research, Queen Mary University of London and Visiting Professor, LSE III
Understanding how to accurately measure the dynamics of inequality is of utmost importance to social scientists. In this paper, for the first time, I identify which inequality measures are best suited to capture the dynamics of inequality, especially for panel regression applications. To undertake this study, I generate a dataset of twelve inequality measure types for up to 100 years across 34 countries using annual data on mortality distributions. Upon modelling inequality as a fractionally integrated process I find that inequality measures that are independent of the mean have more stationary cases, a vital requirement for regressors in panel regression analysis. Mean-dependent measures like the Gini, however, are mostly non-stationary, making them wholly unsuitable for panel regressions. This result is confirmed by estimating the "inequality and growth" relationship, where only mean-independent measures generate a stable relationship. Using a VAR approach, impulse responses show that mean-dependent measures are more likely to carry the effect of a shock, making them less suitable for panel regression analyses. Tests of normality and volatility function as excellent "marker" tests whether a chosen inequality measure is suitable for dynamic contexts. The findings suggest that no inequality measure should be used for dynamic purposes without rigorously testing their suitability.
For online attendance please register here: https://lse.zoom.us/meeting/register/tZ0qfuyrqDwqEtzOlSbpcMC-7Yo5HR3GE--v