Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/190175 
Year of Publication: 
2018
Series/Report no.: 
WIDER Working Paper No. 2018/128
Publisher: 
The United Nations University World Institute for Development Economics Research (UNU-WIDER), Helsinki
Abstract: 
This paper develops a new imputation methodology applied to missing incomes values in PNAD. PNAD is the main Brazilian household survey, but it has no imputation. The imputation process starts by fitting regression models applied to different income sources considering the complex sampling design of the survey. Later this procedure is combined with stochastic methods. In 2015, 2.5 per cent of the sample had per capita incomes imputed, resulting in slightly higher levels of inequality. Inequality and poverty changes were not affected by imputation. We took advantage of the methodology proposed to input rents and to preserve pressure points in income distributions associated with Brazilian institutional features such as minimum wages.
Subjects: 
Imputation methods
stochastic imputation
inequality
poverty
missing incomes
JEL: 
C81
D31
D63
I3
Persistent Identifier of the first edition: 
ISBN: 
978-92-9256-570-1
Document Type: 
Working Paper

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.