The Bureau of Labor Statistics reported another month of solid job growth in September, bringing this year’s average to over 225,000 jobs per month—the highest average monthly job growth since 1999. As the jobs recovery consistently grinds on, more attention is shifting toward the absence of wage growth in the recovery. Wage stagnation is part of a longer-term trend that has been well-documented in EPI’s research. Last month’s Census Bureau report on income, poverty and health insurance coverage in 2013 provided more evidence of weak wage growth following the recession. In my previous analysis of data from the Census report, I also identified how uneven that growth has been for different groups of women. Median real (ie, inflation-adjusted) annual earnings for African American women working full-time full-year in 2013 were 3.3 percent below the 2009 level, compared to 0.2 percent and 0.5 percent lower for white and Hispanic women, respectively.

Given the magnitude of that disparity, I sought to confirm it using hourly wage data for full-time full-year workers from the CPS ORG files, the data source for EPI’s signature research on wage trends. Based on that analysis, I identified similar racial disparities in women’s hourly wage growth.

Overall median real hourly wages for black women working full-time full-year declined 5.1 percent between 2009 and 2013, compared to a decline of 0.2 percent and 2.4 percent for white and Latina women, respectively. Measuring from the last business cycle peak before the Great Recession—from 2007 to 2013—real wages for black women grew a meager 0.8 percent compared to 3.0 percent for white women and 1.8 percent for Latinas.

Since the focus is on full-time full-year workers, the difference in wage growth is not explained by variation in the share of part-time versus full-time workers. Education also seems to have little to do with the observed differences by race and ethnicity. Educational attainment has increased for all groups of women since 2007 and women at every level of education saw real wages decline between 2009 and 2013. However, wages fell the most for women with some college education (a category that includes those who have attended college but not completed a 4-year college degree and those with a 2-year degree). In 2013, 36.3 percent of black women, 30.4 percent of white women and 29.6 percent of Latina women who worked full-time full-year fell into this some college category. The fact that African American women in this category experienced a larger decline in wages than whites or Latinas offers the first clue about what happened to median wages for black women. The second clue comes from a comparison of the occupational distribution of these women in 2007 and 2013.

In making this comparison, I first ranked the ten major occupation categories used in the CPS from highest to lowest median wage based on the BLS Occupational Employment Survey. Next, I identified occupations with the largest shares of women’s employment. At the top of the wage scale were management and professional occupations, in the middle were office and administrative occupations and at the bottom were service and sales occupations.

In 2007, one-third of women with some college were employed in office and administrative occupations. Between 2007 and 2013, the percentages of all women employed in these mid-level occupations fell and the share employed in lower paying service and sales jobs increased, but that is where the similarities end. For whites and Latinas, this shift from mid-wage to low-wage jobs was buffered by the fact that the shares in higher paying management and professional occupations also increased. For black women, there was no such buffer. The share employed in management and professional occupations declined along with the share employed in office and administrative occupations.

The net result for black women was clear occupational downgrading, or relatively large movements out of high- and mid-wage occupations into lower-wage jobs. Though women at other levels of education experienced some occupational shifting as well, changes across occupations were smaller and less concentrated toward lower-paying jobs than they were for black women with some college education. This pattern of occupational downgrading among traditionally middle-wage earning black women is consistent both with racial disparities in median wage growth overall and by level of education.