Inequality, racism, and framing.
by William Saletan
May 4, 2009
People vary in their abilities based in part on genetic differences. Suppose these differences at the individual level sometimes add up to differences in average ability between people of one race and people of another. Should we say so?
Here are three perspectives on the question. On Wednesday, the New York Times ran the following story:'No Child' Law Is Not Closing a Racial Gap ...
On Thursday, I raised a question about the Times story:Why categorize and measure students by race? Aren't there better ways to organize the data? … [Parts of the test report] organize the data by factors that can help us target and adjust educational policy: kids with low scores, kids in public school, kids in high school, kids whose parents didn't graduate. … But race? Does that category really help? And what message does it send to kids when headlines assert a persistent "racial gap"?
On Friday, Steve Sailer, the founder of the Human Biodiversity Institute, responded to my question. He argued that I was wrong to propose to "stop counting" scores by race:
The reason people all over the world and of all different ideologies can't help but be interested in race is [that] a racial group is, fundamentally, an extended family. So, race is about who your relatives are, which is an inherently interesting topic.
Saletan has been arguing that we should just group people by looking at one gene at a time. (Of course, on average, individual gene differences will tend to follow racial lines.) But, more fundamentally, what he doesn't get is that racial groups have an existence independent of genetics. They are fundamentally genealogical entities—who begat whom. Unsurprisingly, when you stop and think about it, the genes tag along with the begats.
Sailer, like the Times, is embracing racial averaging of test scores. But unlike the Times, he's doing so in the belief that differences in the resulting averages are in large part genetic. He's arguing not just that some people do better than others based on inherited ability (the genetic question) and that this ability is more prevalent among people of one race than among people of another (the distribution question), but that this is how the data should be aggregated, averaged, and compared (the framing question).
To be precise, I am arguing that this is how the data is aggregated, averaged, and compared ... by law. The No Child Left Behind legislation godfathered by Ted Kennedy and George W. Bush is explicitly concerned with narrowing racial achievement gaps.
More generally, that mainstay of the civil rights industry, the concept of "disparate impact" -- as exemplified by the EEOC's four-fifths rule, which, in the Supreme Court's Ricci case was cited by the city of New Haven to justify throwing out a firefighter promotional test that no blacks passed -- requires the government to maintain vast statical offices for sorting employees by race. Similarly, the Community Reinvestment Act requires millions of mortgages to be sorted by race in the government's giant Home Mortgage Disclosure Act database in order to lean on the mortgage industry to lend more money to minorities. (How's that working out for us lately?)
Should the government count by race? In 2002, I endorsed and voted for Ward Connerly's California initiative that the state government should stop counting by race. I reasoned by analogy to religion. In the 1950s, the Census Bureau proposed adding a religion question to the Census, but Jewish groups protested, so the Census doesn't count people by religion. And that makes it very hard to file a disparate impact lawsuit over purported religious discrimination based on statistical differences. There simply aren't any government statistics on religion today, so religious discrimination cases require direct evidence of discrimination, so there are fewer lawsuits over religious discrimination than over racial discrimination, and so employers seldom impose religious quotas on themselves.
But, Connerly's initiative to eliminate data collection by race went down to defeat badly, and I haven't expressed much of an opinion on the subject of whether or not the government should collect data by race since. But if the government's going to collect collossal amounts of data by race and impose legal differences by race, then I think it's my duty as a citizen to look at the government's numbers and see what they say.
It's important to separate these three questions. We know that genes influence many abilities. We also know that some of these genes vary considerably in prevalence between ethnic groups. One example is the RR variant of ACTN3, a gene that affects fast generation of muscular force and correlates with excellence at speed and power sports. The opposite variant of the gene is called XX. Tests indicate that the ratio of people with RR to people with XX is 1 to 1 among Asians, 2 to 1 among European whites, and more than 4 to 1 among African-Americans.
We shouldn't overstate the case. Genes don't determine everything, and most genes don't vary significantly between populations. But research is constantly finding new gene-trait correlations and group differences. If your faith in equality depends on an ethnically or racially even distribution of all ability-influencing genes, you're in trouble.
That's why the framing question matters. People of your race may be on average faster, smarter, or more volatile than people of my race. But the opposite pattern may turn up if you and I are classified in some other way. My dad was black, my mom was white, I was born in Hawaii, I was raised in a broken home, I grew up in Indonesia, I went to private school, I played basketball, I used drugs, my grades were unspectacular, and I went to Harvard Law. Guess my IQ.
Rather than focus on an exotic such as the President, who wrote a 460-page book (helpfully subtitled A Story of Race and Inheritance so that you don't miss the point) justifying to himself that he was "black enough" to be a leader of blacks, I think it's more helpful to state what I've often pointed out: "Somewhere around eleven million Hispanics and seven million African Americans have higher IQs than the average white American."
I put a lot of effort a decade ago into trying to come up with broad evidence for Saletan's argument that the government's system of asking people to check off little race and ethnicity boxes is too error-prone and illogical to work, but I eventually had to admit to myself that, on the whole, it was good enough for government work. Sure, there are more than a few exotics like Tiger Woods (who came up with 1 word to describe himself: "Caublinasian") and Barack Obama (who came up with 150,000 words in Dreams from My Father to rationalize his claim to being "black enough"), but most of the time, the government's system kind of sort of works.
The distribution question doesn't settle the framing question, because race is just one way in which ability can be unevenly distributed. To answer the framing question in the affirmative, you have to show something more. You have to show that classifying and comparing by race, rather than using some other classification system or judging each person as an individual, does more good than harm.
It's Ward Connerly's view that the government classifying people by race does more harm than good. Judging from the Obama Administration's amicus curiae brief in the Ricci case, it's definitely not Barack Obama's view. Perhaps Mr. Saletan should take up his argument with the President of the United States rather than with me.
Sailer's argument is that racial classification is natural—that we "can't help but be interested in race" because we tend to define others as in or out of our extended family. I think he's right about that. We're prone to tribalism. But that's not a reason to encourage racial classification. It's a reason to beware it.
In other words, Steve Sailer will more or less win on the scientific grounds any debate over race he choose to engage in seriously, so it's best not to debate the topic at all.
Fine. But can we first get rid of all the government's laws, institutions, and regulations that not only count by race but then discriminate by race, such as the EEOC, the four-fifths rule, the CRA, and so forth?
In other words, Sailer has all the government data on his side, but that just makes it worse!
Consider Sailer's views on immigration. A few months ago, he wrote:
Typically, the two most important factors influencing the long-term success of an organization are the quantity and quality of people involved. … This is particularly true for a country. Yet there has been barely any discussion in the U.S. prestige press on the implications of the demographic change imposed by immigration. … Is adding 100 million Latinos to the U.S. population a good idea? …
And there has been little change in the racial disparities in crime rates. Racial and ethnic differences of all kinds have been strikingly stable since the 1970s. In particular, the word that best sums up Latino America is inertia. Things just sort of keep on keeping on in the general direction that they were already moving. What we do know is that all of these troubles are exacerbated by the mass immigration of people with low human capital.
This is what can happen when you constantly look for racial angles in data on crime, IQ, and other measures of the "quality of people." You start aiming policies at ethnic groups. But I don't think this kind of racism is a product of uneven distribution. It's a product of bad framing.
By this point poor old kicked-around Saletan has finally collapsed into just plain pointing and sputtering about how I, and anybody else who notices the massive demographic changes brought about by our Establishment's immigration policy, is some kind of evil racist.
(I will admit that it's also possible that Saletan has come around to agreeing 100% with me and he's just picking a fight with me to give my sensible views more publicity.)
Read the rest of Saletan's article here.