Labor markets and poverty in village economies

Oriana Bandiera, Robin Burgess, Narayan Das, Selim Gulesci, Imran Rasul and Munshi Sulaiman


This paper reports the results of a randomized evaluation of the nationwide roll-out of a program that transfers assets and skills to the poorest women in villages in Bangladesh, studied through a detailed labour survey that tracks over 21,000 households, drawn from the entire wealth distribution in 1,309 rural Bangladeshi villages, four times over a seven year period. The survey gathers detailed data on hours worked, days worked and earnings for each labour activity of each household member.

The authors find that, at baseline, the choice of labour activity for women is limited as they allocate over 80% of hours worked to three activities: maid services, agricultural labour and livestock rearing. These labour activities are strongly correlated with poverty: poor women engage mostly in casual wage labor as maids and agricultural labourers, while wealthier women specialize in livestock rearing. The key question examined in the paper is whether enabling the poorest women to take on the same work activities as the better-off women in their villages can set them on a sustainable path out of poverty. The paper evaluates BRAC’s Targeting the Ultra-Poor (TUP) program that provides a one-off transfer of assets and skills to the poorest women with the aim of instigating occupational change.

The authors find the program transforms the labour activity choices of ultra-poor women. They increase hours devoted to livestock rearing by 361% after four years post-transfer, and reduce hours to agricultural labour and maid services by 17% and 36% respectively. Aggregating across labour activities, the net effect on hours worked and days worked is an increase of 22% and 25%, respectively, suggesting poor women had idle work capacity and that the program enables them to put it to a productive use by taking on livestock rearing activities.

The reallocation of labour supply across work activities by the ultra-poor leads their earnings to increase by 37% after four years relative to controls, and there is a 15% fall in the probability that an ultra-poor household is below the $1.25 extreme poverty line. Per capita consumption expenditure increase by 10% and the value of household durables by 110%, with both effects being larger after four years than after two. In line with this, earnings from livestock rearing are not entirely consumed, but rather also used to save and invest further in productive assets. Four years post-transfer, savings of the ultra-poor have increased ninefold and they are more likely to receive and give loans to other households.

Overall, the results show that one-off asset and skills transfers to the ultra-poor, enable them to overcome barriers to accessing high return labour activities. These reallocations of labor supply across work activities leads to increases in their consumption, and a diversification of their asset base, especially through accessing land, and that this process sets them on a sustained trajectory out of poverty. The paper concludes by comparing results for Bangladesh to those from six pilot studies in Ethiopia, Ghana, Honduras, India, Pakistan and Peru.


Bandiera, O., Burgess, R., Das, N., Gulesci, S., Rasul, I., & Sulaiman, M. (2016). Labor markets and poverty in village economies (Development Discussion Paper No. EOPP 058). London: STICERD.