Table 6.
Continuity-based analysis: primary outcomes applying the donut-hole approach
Donut-hole radius | RD estimator | MSE-optimal bandwidth [meters] | Robust inference |
CER-optimal bandwidth [meters] | Nl | Nr | ||
---|---|---|---|---|---|---|---|---|
95% CIrbc | SE | p-value | ||||||
Outcome: average household wealth ($) | ||||||||
10 | − 157,882 | 593.148 | [− 345,971, – 10,895] | 85,480 | 0.037** | 488.513 | 17 | 66 |
50 | − 328,490 | 497.881 | [− 814,696, 171,742] | 251,647 | 0.201 | 732.984 | 7 | 42 |
Outcome: mean BMI (kg/m2) | ||||||||
10 | 34.16 | 288.62 | [− 18.14, 75.76] | 23.95 | 0.229 | 239.14 | 9 | 49 |
50 | − 0.298 | 852.939 | [− 5.14, 6.07] | 2.859 | 0.87 | 707.55 | 27 | 79 |
Discrete analysis using cluster standard errors; bandwidth is the distance (in meters) from yellow/red border to grandparent’s census block; fully adjusted model including age, year, gender, race; all models use a first-degree polynomial; Nl (left) and Nr (right) indicate the effective number of observations within the MSE bandwidth used for estimation; wealth is rounded to the nearest whole dollar
p < 0.10;
p < 0.05;
p < 0.01