Thursday, January 5, 2017

Critical Race Theorists at WaPo desperately doing math

From The Asian American ‘advantage’ that is actually an illusion by Jeff Guo.

If you dig beneath the "America is racist" presumption of the article and the dyed-in-the-wool social justice warrior mindset, there are actually a couple of interesting issues.

Guo starts with the SJW schtick.
For decades, the data on median household incomes have shown the same, persistent racial disparities: Asians beating out whites at the top, while Hispanics and blacks hover near the bottom.

Click to enlarge

Asian Americans seem to offer proof that minorities can prosper — and even leapfrog whites — if they work hard and jump through the right hoops. For that reason, Asian Americans have often been invoked as a way to excuse the income gaps between whites and blacks or whites and Hispanics.
Except for the last sentence, all pretty true. The first sentence of the last paragraph is indeed a common argument "Asian Americans seem to offer proof that minorities can prosper — and even leapfrog whites — if they work hard and jump through the right hoops." However, Guo's ideological colors are showing with the second sentence, "Asian Americans have often been invoked as a way to excuse the income gaps between whites and blacks or whites and Hispanics." It is possible that some people have made that argument but I doubt it. I see it always cast as education plus hard work equals success. It is not about "excusing" anything, it is about observing and identifying the causal mechanisms of success.

Guo later comes back to the SJW claim: "The fact that Asian Americans outearn white Americans on average has often been used to deny claims that whites enjoy special advantages in America." This is critical race theory in the raw.

So what is there to support Guo's claim that the Asian American ‘advantage’ is actually an illusion?

This is one of the interesting bits. Guo is pointing out that when comparing groups, you have to compare apples-to-apples. I agree wholeheartedly. His argument is:
But why do typical Asian American households outearn typical white households? Like many statistics showing an Asian American advantage, this fact proves illusory upon closer examination. A common explanation is that Asian Americans are better educated. While that’s true, there’s another factor that can completely account for the income gap between Asians and whites.

It has to do with where people reside.

Prices and rents vary wildly in different parts of the country. The cost of living near Jonesboro, Ark., for instance, is about 18 percent below the national average, while the cost of living near San Francisco is about 21 percent above the national average.

White and African Americans are more likely to live in cheaper locales, while Asian and Hispanic Americans are more likely to live in pricier ones. The contrast between whites and Asians is particularly stark. Nearly 1 in 5 white Americans reside in rural counties, where a dollar goes a lot further. But 97 percent of Asian Americans live in or near a major city, where the cost of living is higher.
Excellent points. Guo then spends several paragraphs detailing the adjustments that need to be made. What are the results, having taken into account where people live?

Click to enlarge

Asian Americans are down by 9%, everyone else is roughly the same +/- 3%.

Guo, at the beginning, seemed to be preparing us for the evidence that once you make adjustments, the new numbers would refute "Asian Americans seem to offer proof that minorities can prosper — and even leapfrog whites — if they work hard and jump through the right hoops." But after the adjustments, Asian Americans still make more than white Americans and both groups are still dramatically ahead of African Americans and Hispanics.

Its as if Guo thought his approach would refute the claim he disliked, worked hard on the numbers, found that they didn't refute the claim, but wrote the article as if they did anyway.

While I intensely dislike the reformed Marxist identity politics, I would argue that Guo is headed in the right direction but didn't go far enough. Critical race theorists actually never want to go in this direction of comparing like-to-like because it so often undermines their argument.

Yes, where people live is one factor undergirding income differences. But so is group age, education attainment, majors achieved (STEM are going to earn more than poets), familial structure, immigration recency, work status (full time, part time, etc.), work duration (how many years continuously employed in the US), etc.

Just comparing race groups based on geographic location is about as equally unrevealing as race on its own, or gender, or education attainment, etc. If you want to compare like-to-like, you have to hold everything equal, i.e. you have to match all the other factors as well. Most of the differences in income are almost certainly going to be explained by differences in those other factors. For example, in terms of education attainment, 36% of non-hispanic whites have college degrees. Asian Americans? 54%. African Americans are 32% and Hispanics are 23%. That will drive an immense amount of difference between the ethnic groups, not even taking into account the differences in types of degrees (STEM and non-STEM). Age is another variable that has to be controlled for. Whites are older (42 years), Hispanics are on average, 27 years old. Asian Americans are 34 years old and African-Americans are 31 years old. Very roughly, the older you are, the more you earn. You have to control for age.

Guo wants to argue that there is white privilege and that Asian Americans are not quite as successful as the unadjusted numbers would show. But if you are obsessed with racial privileges, in order to understand what is really going on, you have to compare like-to-like. Given that there are more than a dozen material determinant variables, getting to real apples-to-apples comparisons would involve a lot of work. Guo hasn't done that. He hasn't even refuted the original argument he wanted to overturn.

In the meantime, those who observe and who measure are probably still on pretty firm ground in making the claim that hard continuous work, cultural values, and quality of education explain virtually all differences between individuals (the balance made up by factors like geographic location, years in the US, age, etc.) and that differences between groups are driven primarily by differences in those factors rather than discrimination or privilege.

The Triple Package by Amy Chua and Jeb Rubenfeld covers a lot of this ground.

No comments:

Post a Comment