Income distribution in urban China: An overlooked data inconsistency issue

Article history: Received 28 May 2013 Received in revised form 17 February 2014 Accepted 19 February 2014 Available online 26 February 2014 The Urban Household Income and Expenditure Surveys, conducted by the National Bureau of Statistics, are extensively explored in income distribution studies. However, we find that a survey coverage expansion that includes migrant residents in the urban sample may induce serious data inconsistency before and after the year 2002. To further unveil the inconsistency, we construct a random walk hierarchical beta-2 distribution model, estimated by the unscented Kalman filter, to investigate the magnitude of the structural break. Results show that the gaps of Gini coefficients and the shape of the distribution can be bridged by a counterfactual time series that coherently measures the urban China income distribution. © 2014 Elsevier Inc. All rights reserved. JEL classification: D31 C32 C46