present is detrimental to the welfare outcomes of the child (McLanahan and Sandefur, 1994; Case
and Paxson, 2001; Ginther and Pollak, 2003). This paper finds substantially different outcomes
for children from rural Burkina Faso. The results are based on an estimation strategy, household
and child fixed effects, that can address the endogeneity of fostering. This paper also describes the
advantages of a research methodology, tracking both households involved in the fostering exchange,
without which I could not examine the impact of fostering on the biological siblings who were left
behind.
This analysis is informative for understanding why families choose to adjust their structure
and reallocate resources between two households (sender and receiver) in such a way as to make
everyone better off in terms of school enrollment. However, while there is strong evidence of a
short-run Pareto improvement in schooling for all children associated with fostering and a long-
run improvement in the welfare outcomes for the foster child, future research needs to examine
additional welfare measures to see if fostering also has a positive impact along other dimensions,
such as health and nutrition.
24