| Male |  | Female |
---|
Age |
Ï„
|
z-stat | Â |
Ï„
|
z-stat |
---|
6 | 0.268*** | 4.581 | Â | 0.172*** | 2.794 |
7 | 0.095* | 1.867 |  | −0.002 | −0.349 |
8 | 0.135*** | 4.098 | Â | 0.029 | 0.998 |
9 | 0.031 | 0.69 | Â | 0.042* | 1.893 |
10 | 0.008 | 0.389 |  | −0.011 | −0.803 |
11 | −0.005 | −0.419 |  | −0.010 | −0.903 |
12 | 0.002 | 0.082 | Â | 0.001 | 0.046 |
13 | −0.004 | −0.395 |  | 0.024** | 1.974 |
14 | 0.014 | 1.264 | Â | 0.023** | 2.256 |
15 | −0.014 | −1.192 |  | 0.001 | 0.015 |
16 | −0.013 | −0.558 |  | 0.015 | 1.565 |
- This table shows the difference between children born on January 1 and on December 31. The dependent variable is equal to one if the student is in the wrong grade for his age and zero otherwise. The first columns show the difference for males, while the last columns show the difference for females. Data come from PNAD for cohorts born between July 1985 and June 1991, living in the Northeast region. The sample is weighted to resemble the parental education of the college candidates. Functions are estimated using a triangular kernel with the bandwidth selection procedure proposed by Calonico et al. (2014). Ï„ is the robust RD estimate
- ***, **, * represent statistical significance at the 1, 5, and 10 % levels, respectively