African-American students need to complete two more levels of education to have the same probability of getting a job as their white peers, a new study by Young Invincibles finds.The following graph is also included in the write up:
The researchers looked at data mainly from the Bureau of Labor Statistics and the U.S. Census, isolating the effects of race and education on unemployment. They found that an African-American male with an associates degree has around the same chance of getting a job as a white male with just a high school diploma. “At every level of education, race impacts a person’s chance of getting a job,” Tom Allison, a research manager and one of the study’s authors, told ThinkProgress.
From this data we see that whites* who've graduated from high school but not attended any college courses do about as well as blacks who have an associate's degree do, and whites who have some college but no degree do about as well as graduate-level blacks do.
Hold these two observations in your head as we take a look at some relevant GSS data.
The following graph shows average IQ, converted from Wordsum scores^, by educational attainment and by race among whites and blacks. For contemporary relevance all data is from 2000 onward and respondents born outside of the US are excluded:
The study mentioned in the article notes that white high school graduates do are as successful finding employment as blacks with associate's degrees are.
Because the GSS measures by years of education, distinguishing between "some college" and "associate's degree" is dicey.
If we assume blacks with associate's degrees fall midway between blacks with some college but no degree and blacks with bachelor's degrees, however, we get a black average that is almost identical to the average among whites with a high school diploma but no college experience.
On the first observation, then, we see that employers tend to treat white high school graduates as about as valuable as blacks with associate's degrees because the two groups tend to be roughly equal in ability.
And we see the same with regards to the second observation--the average IQ among whites with some college is virtually identical to the average IQ of graduate-level blacks.
What happens in the real world then is just what we'd expect to find if employers are rationally hiring based on ability rather than on educational credentials alone.
This aforementioned study spun this--and these studies and subsequent media reports on them always do--as an affront to social justice etc etc, but it's merely the consequence of businesses making prudent, reasonable broad-based assumptions about potential employees based on the actionable information they have available to them.
The GSS allows us to evaluate this using more than just Wordsum scores. The following tables are from selected results from the social survey's science module. Again all responses are from 2000 onward and limited to those born in the US.
Results from the question on electrons being smaller than atoms:
From the question on human evolution:
From the question on the scientific validity of astrology:
From the question dealing with the efficacy of using antibiotics to treat viruses:
The patterns we see with IQ as measured by Wordsum scores generally hold here as well--in fact, the modestly educated whites from our two previous observations actually best more highly educated blacks on these science questions rather than just matching them as they do in IQ.
It's true that most employers will prefer a white applicant over a black applicant if both have similar levels of educational attainment.
It is not true that this is an illustration of irrational discrimination on the part of those employers, however. To the contrary, affirmative action practices--both implicit and explicit--are so deeply embedded in the fabric of American life that this outcome is, given current societal circumstances, inevitable.
See treatments of related ideas from The Alternative Hypothesis here and here.
GSS variables used: YEAR(2000-2014), RACECEN1(1,2), BORN(1), WORDSUM, EDUC(0-11,12,13-15,16-17,18-20), ELECTRON, EVOLVED, ASTROSCI, VIRUSES
* All data is for non-Hispanic whites and non-Hispanic blacks.
^ Assuming a national mean IQ of 98 and a standard deviation of 15.