Table 3

Continuity in Exogenous Characteristics

Dependent Variable
Child Is a Girl
(1)
Height (Centimeters)
(2)
Height-for-Age (Z-Score)
(3)
Adult Is Black
(4)
Panel A: Mothers Sample
1{Ai ≤ 15}−0.005
[0.004]
−0.079
[0.335]
−0.012
[0.052]
0.001***
[0.000]
F-test 1.14  0.055 0.054 9.69
p-value 0.303  0.818 0.820 0.007
Observations91,4802,031 2,027 91,718
Means 0.5160.2−0.5770.997
Panel B: Fathers Sample
1{Ai ≤ 15}0.021
[0.012]
−0.486
[0.676]
−0.082
[0.113]
0.002
[0.001]
F-test 3.04 0.517 0.5191.67
p-value 0.102 0.483 0.4830.215
Observations50,0261,6481,648 50,267
Means 0.5171.31.2760.995
  • Notes: Each cell represents an OLS estimate from regressing a different dependent variable on the discontinuity. The dependent variable is binary in Columns 1 and 4 and continuous in Columns 2 and 3. The height variables are taken from the 2010–2011 Zimbabwe DHS. Cluster-robust standard errors appear below the estimates in brackets. Clustering is at year of birth, that is, age in 1980. All regressions include linear splines in parent age (coefficients not shown). Statistical significance: *p < 0.10, **p < 0.05, ***p < 0.01.