Appendix Table 6.
Predicting Empirical Kink Points
Dependent Variable: Empirical Kink Point | ||||||
---|---|---|---|---|---|---|
White Non- Hispanic |
White Non- Hispanic |
Black Non- Hispanic |
Black Non- Hispanic |
Hispanic | Hispanic | |
Unweighted | Weighted | Unweighted | Weighted | Unweighted | Weighted | |
I | II | III | IV | V | VI | |
Reference Group Median |
0.8211** (0.1482) |
0.8196** (0.1905) |
0.6371** (0.1938) |
0.5676 (0.3820) |
0.6930* (0.2573) |
0.6579 (0.4770) |
Year*Ed Group Dummies | yes | yes | yes | yes | yes | yes |
CMSA Dummies | yes | yes | yes | yes | yes | yes |
| ||||||
Number of Observations | 1171 | 1171 | 474 | 474 | 255 | 255 |
R-Squared | 0.540 | 0.671 | 0.573 | 0.595 | 0.564 | 0.604 |
Notes: Standard errors in parentheses are clustered on CMSA.
indicate staistical significance at the 10, 5, and 1 percent level respectively. The unit of observation is the metropolitan area-race/ethnicity-education-year cell. Weighted regressions are weighted by estimated reference group population.