In Simple Words
Imagine a street vendor who earns $10 a day selling fruit. When a temporary government construction job opens paying $15 a day, the vendor accepts it but must close the fruit stand. The vendor does not gain $15 in brand-new income; net earnings only increase by $5 because $10 of fruit sales were given up to do the government work.
Other members of households in the Congo trial
- Study comparison households
- Comparison households
- Family member in public works
- Work participation fell by 1.5 percentage points
Eastern Congo, over a year after the program. Other household members’ work fell relative to comparison households; no claim all participants lost income.
See source →Research question
Evaluations of public works programs often assume that participant wage payments represent net economic gains. However, because public works require heavy physical time commitments, participants and family members frequently reduce existing productive labor, offsetting gross program wage transfers.
Participants
Randomized controlled trials involving thousands of program participants in Côte d'Ivoire and the Democratic Republic of the Congo.
Study Design
Individual and intra-household randomized controlled trials measuring daily time allocation, self-employment days, and total monthly household earnings.
Results
In Côte d'Ivoire, net participant earning gains accounted for only 53 percent of total public works wages due to foregone lower-paying work. In the Democratic Republic of the Congo, when one family member participated, other household members worked 1.5 percentage points less, contributing to a 9 percent decline in total monthly household earnings.
Takeaway
Calculate net economic program benefits by subtracting estimated foregone private earnings and household labor reductions from total public wage disbursements.