Table 6.
IV robustness analyses: different specifications of the variable of interest.
| Dependent variable | (1) KS3 |
(2) Systolic blood pressure |
(3) Diastolic blood pressure |
|||
|---|---|---|---|---|---|---|
| Coeff | p-Value | Coeff | p-Value | Coeff | p-Value | |
| Panel A: Pr(Underweight) | ||||||
| Coefficient and p-value | −5.861 | 0.315 | −19.26 | 0.021 | −10.56 | 0.065 |
| 95% confidence interval | [−17.3, 5.6] | [−35.66, −2.85] | [−21.77, 0.64] | |||
| First stage F-statistic | 19.91 | 18.22 | 18.22 | |||
| p-value, Hausman test | 0.205 | 0.012 | 0.060 | |||
| Panel B: Pr(Overweight) | ||||||
| Coefficient and p-value | 3.916 | 0.310 | 13.31 | 0.009 | 7.117 | 0.044 |
| 95% confidence interval | [−3.64, 11.47] | [3.46, 22.49] | [0.21, 14.03] | |||
| First stage F-statistic | 45.49 | 40.24 | 40.24 | |||
| p-value, Hausman test | 0.106 | 0.139 | 0.242 | |||
| Number of observations | 4844 | 4047 | 4047 | |||
Notes: See notes to Table 3. Panel A presents the estimates for the effect of being underweight; Panel B shows the estimates for being overweight.