Heterogeneous effects by birth parity: Impact of the PAJE reform on fertility
| . | Announcement Effect . | Implementation Effect . | ||||||
|---|---|---|---|---|---|---|---|---|
| . | First-born . | Parity 2+ . | First-born . | Parity 2+ . | First-born . | Parity 2+ . | First-born . | Parity 2+ . |
| . | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . | (7) . | (8) . |
| Half Benefits | –.005 | .001 | .000 | .006 | ||||
| (.003) | (.018) | (.000) | (.009) | |||||
| Zero Benefits | –.002 | –.018* | .000 | –.007† | ||||
| (.002) | (.008) | (.000) | (.004) | |||||
| Number of Observations | 1,890 | 1,890 | 2,786 | 2,786 | 1,890 | 1,890 | 2,786 | 2,786 |
| R2 | .011 | .020 | .008 | .019 | .000 | .019 | .000 | .015 |
| Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| . | Announcement Effect . | Implementation Effect . | ||||||
|---|---|---|---|---|---|---|---|---|
| . | First-born . | Parity 2+ . | First-born . | Parity 2+ . | First-born . | Parity 2+ . | First-born . | Parity 2+ . |
| . | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . | (7) . | (8) . |
| Half Benefits | –.005 | .001 | .000 | .006 | ||||
| (.003) | (.018) | (.000) | (.009) | |||||
| Zero Benefits | –.002 | –.018* | .000 | –.007† | ||||
| (.002) | (.008) | (.000) | (.004) | |||||
| Number of Observations | 1,890 | 1,890 | 2,786 | 2,786 | 1,890 | 1,890 | 2,786 | 2,786 |
| R2 | .011 | .020 | .008 | .019 | .000 | .019 | .000 | .015 |
| Year Fixed Effects | Yes | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
Notes: Robust standard errors are shown in parentheses. Each cell presents a regression discontinuity (RD) estimator using household-level data and a first-order polynomial control function. The dependent variables are dummy variable indicators for the probability of birth at the household level, for births that occurred nine months after the law was first discussed in the National Assembly in March 2013. “Half benefits” is a dummy variable indicator equal to 1 for households that receive half the amount of PAJE benefits and equal to 0 for households that receive the full PAJE benefits. “Zero benefits” is a dummy variable indicator equal to 1 for households that receive zero PAJE benefits and equal to 0 for households that are eligible for full or half the PAJE benefits. Regressions also include a linear control function, the interaction term between the control function and the RD estimator, and a year fixed effect.
†p < .10; *p < .05