Home
> Uncategorized > Race and the race to the top
Race and the race to the top
January 22, 2016
Bloomberg has a pretty amazing article today with two fantastic graphs. Here’s the article, but the graphs pretty much say it all.
Categories: Uncategorized
mathbabe 
Ouch!
LikeLiked by 1 person
At first it looks like a stark message of “no chance for people of color, no matter what.” But it’s too crude a measure and raises many questions. It can’t be just the white support system (connections), since Asians show the benefits of education. It can’t be just tribalism (“whites only need apply”) for the same reason. If there are cultural differences between ethnicities — Asian, Latino, black — and distinct mainstream cultural perceptions of those ethnic cultures, then there might be a better explanation for the disparities beyond just skin color. They should have distinguished American blacks and Caribbean and East African blacks to test a culturally-based hypothesis.
LikeLike
Sorry – West African.
LikeLike
Fascinating… I’m disappointed but not surprised by the reported differences in chances between whites, blacks, and hispanics. Also disappointed that the article didn’t seem to bother at all to attempt to interpret the huge difference in the width of the confidence interval between the “Asian” category and the other three. Could this possibly point to the fact that these categories are grossly oversimplified, arguably themselves artifacts of the racist society that such studies seek to expose?
LikeLike
Not all degrees are created equal as this analysis tacitly assumes. It’s quite possible that if you restricted these graphs to a particular university or type of university (say, flagship state schools) the race effects would be far less dramatic,if they exist at all. A Simpson’s Paradox of sorts could easily explain these graphs because racial composition varies substantially between universities.
The real story here is that minorities are disproportionately enticed into poorer-quality, often for-profit educational institutions by those promoting the “all degrees are created equal” canard. Victims leave with debt loads that virtually ensure they never reach millionaire status. A college degree may be a road to riches but where you get it determines whether you’re on an interstate highway or a dirt path.
LikeLike
Are the confidence intervals different because of sampling?
LikeLike
My attention was drawn to the fact an Associate’s Degree is a negative predictor for Asians, compared to stopping with a High School diploma.
LikeLike
No control for test scores or major = garbage study.
I am surprised you’re so credulous here.
LikeLike
Disagree. It’s just different information than if they had controlled for those. I’m surprised you’re so quick to judge!
LikeLike
The huge intervals for the Asian category combined with the break in monotonicity of the Associate’s Degree makes me a bit skeptical on the research quality of the paper, at least as far as it applies to Asians. Also, not all degrees are created equal, so these numbers are probably capturing more than just the race effect.
LikeLike
Rob, *all* those groups show the benefits of education. In each group, e.g., a bachelor’s degree means something like 2-4x the chance of being a millionaire compared with just finishing high school, and a master’s means something like 2x on top of that. The *absolute* increases are smaller for the Hispanic and black populations just because all their numbers are 3-4x lower than for the white population.
Colby, my guess is that the large variation in “Asian” results is either just smaller numbers. The intervals seem like they might be roughly proportional to 1/sqrt(population). Nathan, I wonder whether your second sentence is the answer to your first: it doesn’t seem impossible that “Asians” includes, let’s say, Group A who for some cultural reason seldom take associates’ degrees and who for some entirely different reason are more likely to end up millionaires, and Group B who happen to go the other way on both those things. (E.g., it seems like there might be all kinds of differences in culture and outcomes between people of Chinese and Indian descent.)
LikeLike
The government is trying to increase the middle-class by subsidizing things that middle-class people have: If middle-class people go to college and own homes, then surely if more people go to college and own homes, we will have more middle-class people.
The problem is that home ownership and college educations aren’t causes of middle-class status, they are merely markers of class status; and subsidizing the markers does not produce the traits, if anything, it undermines them as recent history shows.
LikeLike
There’s also no reason to assume that the financial benefit of higher education is due to any skill demonstrated other than maybe discipline and endurance. If everyone has a Master’s, then the system will continue to create higher credentials to differentiate. Cathy, this gets at my earlier tweet to you about quasi-experiments. Even if you can statistically tease out confounding variables, how does data like this show that education has financial value beyond the credential?
LikeLike
Nathan — Associate degrees are frequently, maybe usually, a mark of initial failure, not academic progress. It’s a failure to get into a senior college, since senior college is open to HS grads. All the other degrees are a progress: can’t get a bachelor’s without a high school degree, can’t get a masters without a bachelor’s and even a Ph.D. gets an MPhil on the way automatically. So the depressed value for associate’s degrees among Asians says to me that Asians typically fail less educationally, and those that do are selected for higher economic failure as well. That need not be true of other ethnicities including whites, especially poor whites who often progress through the smaller increments.
G — To me the graph is not intended to show the benefit of education but to show a pure measure of economic attainment. Give different races the same economic tool (education), the results remain racially discriminatory.
Greg mentions that the tool given is not the same — not all degrees are equal — but the Asian category belies that criticism of the graph. The graph still shows that education as a broad tool does not break the racial bias in our society. The cause may be that blacks and Latinos are given less valuable degrees, but isn’t that part of what we mean by racial discrimination? Iow, I don’t see Greg as criticising the graph’s import for American racist discrimination so much as describing an element of American racist implementation.
LikeLike
rob – Your hypothesis would actually imply a rather nefarious detriment of being Asian which isn’t addressed – that failures are more punishing than for other races. Who knows if it is true (although the cynical side of me is inclined to believe), but the point being that studies like these only create more questions than answers, and those answers may not be the message it’s trying to convey at face value.
LikeLike
While I agree there are lingering questions, I think these graphs convey a hell of a lot.
LikeLike
Nathan — I’m not sure what you mean, but it’s not punishing, it’s a selection effect. So imagine someone from a privileged background who gets straight F’s in high school despite trying. That’s a student headed for an associates at best and a lower chance of economic success. Now imagine an underprivileged student who does reasonably well in a crappy public school. That student is motivated but hindered by a poor educational background and so is also headed for an associate degree, but is more likely to succeed with it.
So if you assume that one demographic inclines towards the top of its successes, the failures among them will appear as anomalous. Similarly, one would expect that white privileged students graduating from public higher ed would show less of an economic disparity with underprivileged graduates from public higher ed compared with overall privileged vs underprivileged graduates since the privileged public grad is a selection of the bottom of the white demographic, while the non white public grads are representing their demographic more widely and evenly, given that the privates remain racially discriminatory.
So again the graph says to me that Asians are educationally focused, and their educational failures are anomalous.
LikeLike