FLORENT DUBOIS

The Contribution of Residential Segregation to Racial Income Gaps: Evidence from South Africa

Florent Dubois, Christophe Muller

Abstract

Persistent racial income disparities cannot only be explained by differences in socio-economic characteristics. In this paper, we contend that local segregation should be an essential component of the determination of socio-ethnic income gaps using the contemporary White/African gap in South Africa. First, we complete Mincer wage equations with an Isolation index. Second, we decompose the income gap distribution into detailed composition and structure components. Third, we explore the heterogeneity of segregation effects along three theoretical lines: racial preferences, labor market segmentation, and networks effects. Segregation is found to be the main contributor of the structure effect, ahead of education and experience, and to make a sizable contribution to the composition effect. Moreover, segregation is detrimental to incomes at the bottom of the African distribution, notably in association with local informal job-search networks, while it is beneficial at the top of the White distribution. Only minor influences of racial preferences and labor market segmentation are found. Specific subpopulations are identified that suffer and benefit most from segregation, including for the former, little educated workers in agriculture and mining, often female, immersed in their personal networks. Finally, minimum wage policies are found likely to attenuate the segregation’s noxious mechanisms.

Mot(s) clé(s)

Post-Apartheid South Africa, Distribution Decompositions, Income Distribution, Residential Segregation

Local Whittle Analysis of Stationary Unbalanced Fractional Cointegration Systems

Florent Dubois, Elena Ivona Dumitrescu, Gilles de Truchis

Abstract

In this paper we propose a local Whittle estimator of stationary bivariate unbalanced fractional cointegration systems. Unbalanced cointegration refers to the situation where the observables have different integration orders, but their filtered versions have equal integration orders and are cointegrated in the usual sense. Based on the frequency domain representation of the unbalanced version of Phillips’ triangular system, we develop a semiparametric approach to jointly estimate the unbalance parameter, the long run coefficient, and the integration orders of the regressand and cointegrating errors. The paper establishes the consistency and asymptotic normality of this estimator. We find a peculiar rate of convergence for the unbalance estimator (possibly faster than root-n) and a singular joint limiting distribution of the unbalance and long-run coefficients. Its good finite-sample properties are emphasized through Monte Carlo experiments. We illustrate the relevance of the developed estimator for financial data in an empirical application to the information flowing between the crude oil spot and CME-NYMEX markets.

Mot(s) clé(s)

Unbalanced cointegration, Long memory, Stationarity, Local Whittle likelihood