Effect of Family Size on IQ of Children Instrumental Variable Estimation
| OLS (no controls) | OLS (controls) | First Stage | Second Stage | N | |
|---|---|---|---|---|---|
| Instrument: twin at second birth (Sample: First child in families with two or more births) | 0.717* (0.019) | 173,706 | |||
| Number of children in family | −0.052* (0.005) | 0.006 (0.005) | −0.149* (0.052) | ||
| Instrument: twin at third birth (Sample: first and second children in families with three or more births) | 0.839* (0.018) | 135,970 | |||
| Number of children in family | −0.082* (0.008) | −0.002 (0.007) | −0.170* (0.052) | ||
| Instrument: twin at fourth birth (Sample: first, second, and third children in families with four or more births) | 0.903* (0.036) | 44,695 | |||
| Number of children in family | −0.046* (0.012) | 0.009 (0.010) | −0.115 (0.080) | ||
| Instrument: first two children same sex (Sample: first and second children in families with two or more births) | 0.082* (0.003) | 304,755 | |||
| Number of children in family | −0.045* (0.004) | 0.014* (0.004) | 0.065 (0.074) | ||
| Instrument: first three children same sex (Sample: first, second, and third children in families with three or more births and first two births same-sex) | 0.070* (0.005) | 94,578 | |||
| Number of children in family | −0.077* (0.010) | 0.008 (0.009) | 0.076 (0.171) |
↵* indicates statistical significance at the 5 percent level. Standard errors (in parentheses) allow for correlation of errors within family. N represents number of individuals. The regressions also include indicators for age, test year, mother’s age, mother’s age at first birth, mother’s education, father’s education, father’s IQ, and log birth weight of the children studied.