African American and Hispanic wage income compared with wage income of similarly educated white workers, in the private sector versus the state and local public sectors
|Race/education level||Private sector||State and local public sectors|
|African Americans with less than high school||-12.6%||+8.3%**|
|African Americans with a high school diploma||-13.1%***||-5.7%***|
|African Americans with some college||-11.0%*||-5.2%|
|African Americans with a bachelor’s degree||-13.3%||+0.4%**|
|African Americans with an advanced degree||-19.5%***||+1.7%**|
|Hispanics with less than high school||-7.4%**||+6.7%*|
|Hispanics with a high school diploma||-10.2%***||-4.0%|
|Hispanics with some college||-9.3%||-6.2%|
|Hispanics with a bachelor’s degree||-19.3%***||-0.6%|
|Hispanics with an advanced degree||-19.6%***||-2.0%|
Notes: Controls for all models include education, experience, gender, race, marital status, organizational size, metropolitan status, citizenship, Census region, full-time status, and total work hours. Full regression results are included in the appendix tables. See the “Table and figure notes” section of the briefing paper for more detail.
*Probability estimate 0 is >.1. (Interactive models show significance of the interaction term.)
**Probability estimate 0 is >.05.
***Probability estimate 0 is >.01.
Values describe percentage difference in annual income from wages compared with white workers in the same sector, using the equation dlog(y)/dx =100% x (e≈í‚â§-1). In these interactive models, the described effects reflect the sum of the coefficient on each race indicator variable and the respective race-education interaction variables. The probability estimates denote the statistical significance of the interaction terms. Thus for models with statistically insignificant interaction terms, such as the "African Americans with less than high school" interaction in the private sector, this can be interpreted to mean that either the white-African American wage gap at this education level is not statistically different from the wage gap at the "African Americans with a high school diploma" education level, or that the wage penalty of not having a high school degree is not different for African Americans and whites.
Source: Authors’ analysis of Current Population Survey Annual Social and Economic Supplement microdata, pooled years 2006 and 2007
This chart appears in: