Parameter Estimates from Earnings Regressions, Allowing the Effect of Average School SAT to Vary by Parental Education
| Variable | Parameter Estimates | |||
|---|---|---|---|---|
| 1976 Cohort | 1989 Cohort | |||
| Basic | Self-Revelation | Basic | Self-Revelation | |
| School SAT score/100 | 0.126 (0.035) {0.058} | 0.041 (0.036) {0.055} | 0.206 (0.073) {0.110} | 0.117 (0.075) {0.112} |
| Average years of parental education | 0.066 (0.027) {0.041} | 0.063 (0.027) {0.038} | 0.128 (0.053) {0.097} | 0.106 (0.053) {0.097} |
| Years of parental education * school SAT score/100 | −0.004 (0.002) {0.004} | −0.004 (0.002) {0.003} | −0.009 (0.004) {0.008} | −0.008 (0.004) {0.008} |
| Effect of a 200-point increase in school SAT score if | ||||
| Average years of parental education = 12 (equivalent to high school graduate) | 0.148 | −0.025 | 0.193 | 0.052 |
| Average years of parental education = 16 (approximately equivalent to college graduate) | 0.113 | −0.060 | 0.120 | −0.009 |
| Average years of parental education = 19 (approximately equivalent to graduate degree) | 0.087 | −0.087 | 0.065 | −0.055 |
| Sample size (unweighted) | 12,053 | 12,053 | 6,466 | 6,466 |
Source: C&B data and Detailed Earnings Records from the Social Security Administration.
Notes: Parameter estimates drawn from a weighted least squares regression. In addition to the variables listed in the first column of the table, both the basic and self-revelation models control for race, gender, student’s SAT score, a dummy indicating missing student SAT score, student’s high school GPA, a dummy indicating if high school GPA was missing, and a dummy for college athlete; the self-revelation model also controls for the average SAT score of the schools to which the student applied and for the number of applications the student submitted. Two sets of standard errors are reported, one in parentheses and one in brackets. Standard errors in brackets are robust to correlated errors among students who attended the same institution. The dependent variable for the 1989 cohort is the log 2007 earnings and for the 1976 cohort is the log (median of 1993 annual earnings through log 1997 annual earnings). Individuals are excluded if their annual earnings (for the 1989 cohort) or median annual earnings (for the 1976 cohort) were less than $ 13,822 in 2007 dollars. The parental education measure is the average of the mother’s and father’s education; if data were missing for one parent, the average was set equal to the years of education of the parent with data. Individuals with no parental education data for either parent are excluded. The average (standard deviation) of years of parental education was 15.23 (2.70) for the 1976 cohort and 16.18 (3.24) for the 1989 cohort.