Table 5.
Effect of Attending a Catholic School on Log of Students’ Wages per $100 in 1974 (Semiparametric Method with Instrumental Propensity Score), Simulation Study Using Data From the Wisconsin Longitudinal Study, 1957–2011
| Effecta | Methods | Overall | Stratum 1 | Stratum 2 | Stratum 3 | Stratum 4 |
|---|---|---|---|---|---|---|
| Estimate (95% CI) | Estimate (95% CI) | Estimate (95% CI) | Estimate (95% CI) | Estimate (95% CI) | ||
| CACE | OLS | 0.101 (0.052, 0.150) | ||||
| 2SLS | 0.151 (0.034, 0.268) | |||||
| SEM | 0.147 (0.052, 0.257) | 0.172 (−0.115, 0.370) | 0.028 (−0.171, 0.357) | 0.301 (0.038, 0.490) | 0.066 (−0.065, 0.401) | |
| CQCE | SEM | |||||
| 0.10 | SEM | 0.105 (0.010, 0.223) | 0.105 (−0.087, 0.347) | 0 (−0.140, 0.355) | 0.221 (0.018, 0.730) | 0.054 (−0.105, 0.580) |
| 0.25 | SEM | 0.083 (0, 0.167) | 0.140 (−0.074, 0.254) | 0.024 (−0.113, 0.250) | 0.185 (0.008, 0.300) | 0.025 (−0.051, 0.262) |
| 0.50 | SEM | 0.102 (0, 0.134) | 0.098 (−0.095, 0.222) | 0.033 (−0.154, 0.208) | 0.170 (0.012, 0.263) | 0.007 (−0.030, 0.223) |
| 0.75 | SEM | 0.105 (0, 0.201) | 0.130 (−0.095, 0.288) | 0.049 (−0.174, 0.272) | 0.241 (0.029, 0.336) | 0.040 (−0.042, 0.288) |
| 0.90 | SEM | 0.197 (0, 0.310) | 0.288 (−0.192, 0.464) | 0.035 (−0.293, 0.460) | 0.405 (0.041, 0.547) | 0.087 (−0.049, 0.511) |
Abbreviations: 2SLS, 2-stage least squares; CACE, complier average causal effect; CI, confidence interval; CQCE, complier quantile causal effect; OLS, ordinary least squares; SEM, semiparametric method.
a Comparison on the treatment distributional effect between compliers under treatment and control: P = 0.5456 (stratum 1); P = 1.0000 (stratum 2); P = 0.0007 (stratum 3); P = 0.4647 (stratum 4).