Abstract:The objective of this Working Paper is to estimate the likelihood of the exit of households from poverty and identify the determinants of this transition, taking into consideration the length of time that households have spent in poverty. Our focus is to analyze whether short-term changes in the labour market affect the probability of exiting or remaining in poverty. We use the only panel data that are available in Brazil for carrying out this kind of analysis: the Monthly Employment Survey (PME), which was conducted from March 2002 to May 2007. However, since this survey follows households for a very short period of time, we had to adopt estimation techniques that control for cases of right- and left-censoring. The most important results in this Working Paper are: 1) the longer the spell of poverty, the lower the probability of exiting it; 2) households that entered into poverty with zero income (namely, their poverty income gap was equal to one) are not those with the lowest probability of exiting this condition; 3) changes in the unemployment rate of household members do not directly affect the duration of the household’s poverty; and 4) the increase of the average wage of informal workers has a significant, positive effect on the probability of the exit of poor households from poverty.

Keywords:Duration of poverty spell; Poverty exit; Labour market; Survival models for left-censored data
Publication Date:
Type/Issue:Working Paper/44

This publication can also be founded in a condensed and/or expanded format: