Racial Disparity in Earnings

Today, even as I am writing this, people are gathered on the Mall in Washington, DC, to mark the fiftieth anniversary of the march on Washington at which Dr. Martin Luther King, Jr., gave his most famous speech. It’s useful to remember that the actual name of that event was the March on Washington for Jobs and Freedom. I thought this would be a good occasion to take a look at the employment situation of African American workers.

Figures on unemployment are relatively easy to obtain. For example, you can quickly find that 13.8 percent of the African American population in the labor force is unemployed, compared to 7.2 percent in the White population. (These are actually 2012 figures.)

What I thought would be more interesting would be to look at the racial mix of various occupations and see the impact on earnings. Recently, I have been doing a lot of analysis using correlations, so I decided to see how racial presence in occupations is correlated with median earnings in those occupations. Understand that correlation is not the same as causation, but it indicates that two things are happening together, for whatever reason. In this case, I was trying to determine whether concentration of any race tends to happen together with the level of median earnings.

What I found was not surprising but also not pleasant to contemplate. The correlation between percentage of African American workers in occupations and median income in those same occupations was –0.37. If you’re not familiar with correlation, let me explain what this negative correlation means: To some extent (specifically, 37 on a scale of 0 to 100), the greater the percentage of African Americans working in an occupation, the lower the median income of that occupation is likely to be. For Hispanic/Latino workers, the negative correlation is actually even greater: –0.48.

For Whites and Asians, however, the correlations are positive: 0.24 for Asians and 0.43 for Whites.

Note that this simple analysis masks a lot of information. It does not tell us what positions the workers of each races are holding within the occupations. It also does not tell us the full-time or part-time status of the workers. (Actually, a slightly higher percentage of White workers are part-timers.) It does not include earnings of self-employed workers. It does not account for loss of earnings among those who are counted in an occupation but who currently are unemployed. For actual earnings comparisons, a better indicator might be that full-time African American male workers are currently earning 75.3 percent of the earnings of White male workers. For women, the figure is 85.0 percent. These actual earnings comparisons are consistent with what I found about tendencies in occupations.

These statistics are one more indication that we do not yet live in a postracial society. Understand that the solution to this situation is not simply a matter of achieving colorblindness in hiring, although that certainly would help and is something we have not yet achieved. In an experiment described in a paper (PDF) called “Are Emily and Greg More Employable than Lakisha and Jamal?,” resumes were randomly assigned African American– or White-sounding names and sent to employers in the Chicago and Boston areas. The resumes with White-sounding names resulted in 50 percent more callbacks for interviews.

Even if hiring were not biased, the career prospects of African Americans are damaged by a justice system that stops, arrests, and imprisons African Americans at a much higher rate than Whites, even for offenses that are known to be committed at equal rates. Imprisoning people not only puts a stain on the convict’s record that reduces employability but, especially for young people, breaks up families and thus damages the prospects of the next generation.

I am not blind to the advances in racial justice that have been made over the past 50 years. But our nation has a lot further to go to realize Dr. King’s dream, and doing so will take positive action, not passive waiting around for change to occur.