Research Department Working Paper n°20

Active labour market programmes in Latin America and the Caribbean: Evidence from a meta analysis

This paper presents a systematic collection and assessment of impact evaluations of active labour market programmes in Latin America and the Caribbean

The paper delineates the strategy to compile a novel meta database and provides a narrative review of 51 studies. Based on these studies, the quantitative analysis extracts a sample of 296 impact estimates, and uses meta regression models to analyse systematic patterns in the data. In addition to analysing earnings and employment outcomes as in previous meta analyses, we also code and investigate measures of job quality, such as the effects on hours worked and formality.

We find that ALMPs in LAC are particularly effective in increasing the probability of having a formal job, compared to other outcomes. Our results also show that training programmes are slightly more effective than other types of interventions. Moreover, when looking at the sample of training programmes alone, we observe that formal employment is also the outcome category that is most likely to be impacted positively by these programmes. In terms of targeting, we find that ALMPs in the region work better for women than for men, and for youth compared to prime-age workers.

Finally, medium-run estimates are not more likely to be positive than short-run estimates, while programmes of short duration (4 months or less) are significantly less likely to produce positive effects compared to longer interventions.