Convergence in income inequality in the United States: a nonparametric analysis

Abstract

This note investigates the spatial distribution of income inequality in the US over the period 1969 to 1999. Taking into account the methodological limitations of traditional convergence analysis, a non parametric approach is applied to examine the dynamics of the entire cross-sectional distribution. The study reveals the presence of a process of convergence in income inequality across the US states throughout the sample period, as a result of the evolution experienced by those sates located at both ends of the distribution in 1969. Nevertheless, the estimates performed suggest that this process will not continue indefinitely.

Notes

¹ For an exception, see Benabou (1996) and Ravallion (2003).

² Gaussian kernel functions were used, while the value of the smoothing parameter was selected following Silverman (1986, p. 48).

³ For a formal definition of these instruments, see Durlauf and Quah (1999).

⁴ Gaussian kernel functions were employed in the estimations, while the smoothing parameter was determined according to Silverman (1986, p. 86).

Fig. 2. Stochastic kernel and contour plot of the distribution of income inequality in the US states

⁵ The robustness of these findings was checked using a single 30-year transition. The results, which are available from the authors upon request, were in the two cases very similar.