Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 208 - Survey Estimation
Type: Contributed
Date/Time: Tuesday, August 4, 2020 : 10:00 AM to 2:00 PM
Sponsor: International Statistical Institute
Abstract #312763
Title: Maximum Constrained Pseudo-Likelihood Estimation of Income Distributions, Combining Sources
Author(s): Victor Bustos*
Companies: INEGI-MEXICO
Keywords: Inequality; Pseudo-likelihood; Comparability; Tax records; SNA results; Survey data
Abstract:

Despite their usefulness, accurate estimation of income distributions from survey data and other sources has proven elusive and comparisons among them unsuitable. Also, statistics derived from them (e.g., Ginis) cannot be compared. Other data sources are available but seldom used. We present an approach for their estimation from such data, allowing for both income under-reporting and underrepresentation of large income households, by using other statistical sources and reconciling all of them under different conditions. Our proposal selects the best distributional model using a Maximum Constrained Pseudo Likelihood (MCPL) criterion. Our procedure is applied to Mexican data from 3 sources: (a) the National Household Income and Expenditure Survey, (b) Mexico’s System of National Accounts, and (c) income tax records, sources that may produce differing results regarding total household current income. Results for all 32 Mexican states and some years between 2010 and 2016 show remarkable stability, allowing us to make comparisons.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2020 program