Syndicate content

How persistent is poverty in the short run?

Joao Pedro Azevedo's picture

Poverty is often measured using repeated cross-sectional surveys that provide a snapshot of the poverty status of a given household at a particular point in time. Such designs call for interviewing different respondents in each round, and because individuals or households are only observed once, we cannot always tell whether their poverty status is enduring or transitory.

This begs the question: how often do households enter and exit poverty in the short-run?

As discussed in our previous post, there can be large fluctuations in the prevalence of poverty by season. Given the aggregate changes we see within the year at the national level, it is very important to understand to which extent such fluctuations are also affecting different individuals and households in different ways within the year. Indeed, research by Martin Biewen at IZA finds that, in mostly developed economies, household transitions into and out of poverty can be substantial.

Panel survey designs that repeatedly visit the same respondents enable such detailed monitoring of trends within the year. In the case of Tajikistan’s Household Budget Survey, the same households are interviewed in each quarter, allowing us to see whether households remain poor over long periods, or whether there is “churning” into and out of poverty from one quarter to the next.



Note: Author’s calculation using the TJK HBS

As the graph above shows, in Tajikistan, about 50% of households are never poor in a given 4-quarter period, and about 10% are consistently poor. The remaining 40% of households are at times poor, while at other times they are above the poverty line for the quarter. These estimates imply that there is room to improve the targeting of poverty-related initiatives by taking these fluctuations into account.

This churning has clear data collection quality implications, and illustrates the additional value of investing in continuous household panel surveys where it is possible. If panel surveys are not feasible for a particular country, analysts and policy makers should at least bear in mind that when you survey people matters, both for data comparability over time, as well as for identifying who is in most need of particular programs and policies.

These trends also underscore the importance of applying suitable methods for identifying social program beneficiaries that are able to distinguish those entrenched in poverty from those who are moving in and out of poverty. One common method is the Proxy Means Test approach, which is designed to identify those who are chronically poor. However, when using this approach, there are potential estimation challenges, especially in contexts where many people move in and out of poverty on a regular basis.

Add new comment