Working Paper: NBER ID: w26957
Authors: Mark R. Rosenzweig; Rafael J. Santos Villagran
Abstract: We exploit variation in the composition of local fish catches around the time of birth using administrative and census data on adult cognitive test scores, schooling attainment, and occupation among coastal populations in Colombia to estimate the causal effects of early-life consumption of methylmercury (MeHg) and DHA, elements contained in fish, on cognitive development. Using an IV strategy based on an equilibrium model of fish supply that exploits time-series variation in oceanic SST anomalies on both coasts of Colombia from 1950 to 2014 as instruments, we find that net of cohort and municipality fixed effects increases in high-(low-)MeHg fish catches around a cohort’s birth negatively (positively) affect the cohort’s verbal and math test scores upon exiting high school and their likelihood of continuing their schooling, while increasing (decreasing) the likelihood the cohort is disproportionally represented in manual-labor occupations.
Keywords: methylmercury; DHA; cognitive development; fish consumption; Colombia
JEL Codes: J13; O15; Q22; Q53
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
SST anomalies (Y10) | fish catches (Q22) |
fish catches (Q22) | cognitive outcomes (D91) |
high MeHg fish catches (Q22) | cognitive outcomes (D91) |
low MeHg fish catches (Q22) | cognitive outcomes (D91) |
high MeHg fish catches (Q22) | schooling continuation (I21) |
low MeHg fish catches (Q22) | schooling continuation (I21) |
high MeHg fish catches (Q22) | manual labor occupation (J46) |
low MeHg fish catches (Q22) | manual labor occupation (J46) |
fish supply composition at later ages (Q22) | adult outcomes (I26) |
total fish catches around birth (Q22) | cognitive skill measures (G53) |
total fish catches around birth (Q22) | height (Y60) |