Update, May 10, 2013: On Friday afternoon, after this story had already been published, Jason Richwine resigned from the Heritage Foundation.
David Weigel is a reporter for the Washington Post.
Four years ago, long before he’d join the Heritage Foundation, before Marco Rubio was even in the Senate, Jason Richwine armed a time bomb. A three-member panel at Harvard’s John F. Kennedy School of Government accepted Richwine’s thesis, titled “IQ and Immigration Policy.” In it, Richwine provided statistical evidence that Hispanic immigrants, even after several generations, had lower IQs than non-Hispanic whites. Immigration reformers were fools if they didn’t grapple with that.
"Visceral opposition to IQ selection can sometimes generate sensationalistic claims—for example, that this is an attempt to revive social Darwinism, eugenics, racism, etc,” wrote Richwine. “Nothing of that sort is true. … an IQ selection system could utilize individual intelligence test scores without any resort to generalizations.”
This week, Heritage released a damning estimate of the immigration bill, co-authored by Richwine. The new study was all about cost, totally eliding the IQ issues that Richwine had mastered, but it didn’t matter after Washington Post reporter Dylan Matthews found the dissertation. Heritage hurried to denounce it—“its findings in no way reflect the positions of The Heritage Foundation”—and Richwine has ducked any more questions from the press.
His friends and advisers saw this coming. Immigration reform’s political enemies know—and can’t stand—that racial theorists are cheering them on from the cheap seats. They know that the left wants to exploit that—why else do so many cameras sprout up whenever Minutemen appear on the border, or when Pat Buchanan comes out of post-post-post retirement to write another book about the “death of the West?”
Academics aren’t so concerned with the politics. But they know all too well the risks that come with research connecting IQ and race. At the start of his dissertation, Richwine thanked his three advisers—George Borjas, Christopher Jenks, and Richard Zeckhauser—for being so helpful and so bold. Borjas “helped me navigate the minefield of early graduate school,” he wrote. “Richard Zeckhauser, never someone to shy away from controversial ideas, immediately embraced my work.”
Yet they don’t embrace everything Richwine’s done since. “Jason’s empirical work was careful,” Zeckhauser told me over email. “Moreover, my view is that none of his advisors would have accepted his thesis had he thought that his empirical work was tilted or in error. However, Richwine was too eager to extrapolate his empirical results to inferences for policy.”
Borjas’ own work on immigration and inequality has led to a few two-minutes-hate moments in the press. He wasn’t entirely convinced by Richwine, either.
“I have never worked on anything even remotely related to IQ, so don't really know what to think about the relation between IQ, immigration, etc,” Borjas told me in an email. “In fact, as I know I told Jason early on since I've long believed this, I don't find the IQ academic work all that interesting. Economic outcomes and IQ are only weakly related, and IQ only measures one kind of ability. I've been lucky to have met many high-IQ people in academia who are total losers, and many smart, but not super-smart people, who are incredibly successful because of persistence, motivation, etc. So I just think that, on the whole, the focus on IQ is a bit misguided.”
But Richwine had been fascinated by it, and for a very long time, in an environment that never discouraged it. Anyone who works in Washington and wants to explore the dark arts of race and IQ research is in the right place. The city’s a bit like a college campus, where investigating “taboo” topics is rewarded, especially on the right. A liberal squeals “racism,” and they hear the political correctness cops (most often, the Southern Poverty Law Center) reporting a thinkcrime.
I saw this for the first time in 2006. During the backlash to the McCain–Graham immigration bill, the young paleo-conservatives Marcus Epstein and Richard Spencer founded the Robert Taft Club, a debating society that would welcome taboo ideas and speakers. They did so, and the SPLC then branded them “hateful”—it was the way of things. But I’d sometimes attend those events, as a reporter. In 2006, they invited American Renaissance founder Jared Taylor to a debate on race and conservatism. Years later, a few reporters condemned James O’Keefe because he’d been in the room. They botched the story, accusing O’Keefe of planning the event, when he’d merely shown up. The lesson everyone took away from this? Well, of course the left would blow up anything you said about race into a controversy. That was no reason to stop doing it.
Richwine either relished in the controversy or didn’t care. In 2008, while at the American Enterprise Institute, he joined a panel discussing a new book from Mark Krikorian, executive director of the Center for Immigration Studies. “Decades of psychometric testing,” said Richwine, “has indicated that at least in America you have Jews with the highest average IQ, usually followed by East Asians, and then you have non-Jewish whites, Hispanics, and then blacks. These are real differences. They’re not going to go away tomorrow.”
Even in that room, conservatives tried to distance themselves from Richwine’s remark. “It's looking at America in 1965 and assuming that's what we always were,” said Krikorian. Even Richwine added some caveats. “I point out that Ayaan Hirsi Ali was given an IQ test in the Netherlands and did very poorly,” he said. “[It’s] hard to imagine someone brighter.”
But Richwine was winning fans on the nativist right. Marcus Epstein was in the audience, asking a question, then writing the event up favorably at the anti-immigration site VDare.com. Over the years, VDare’s Steve Sailer would point to Richwine’s work and charts to reveal cold truths about racial IQ differentials. In March 2009, he shared Richwine’s calculations “from the 2003 New Immigrant Survey of the backward digit span subtest from the Wechsler IQ test.” Immigrants from Mexico had IQs, on average, 18 points lower than those of white Americans.
Throughout 2009, Sailer pointed readers to Richwine’s research on IQ and crime rates. He doesn’t recall whether he talked to Richwine personally, but Richwine never really put up a wall between his work and the endorsements of nativists. While at AEI he got to know Richard Spencer, the other Taft Club founder, and another thinker who laughed at the constant denunciations of “hate watchers.” In 2010, as first noticed by Yahoo News’ Chris Moody, Richwine wrote a couple of pieces for Spencer’s new white nationalist magazine Alternative Right. His debut story demolished a piece by the pro-immigrant legalization conservative Ron Unz. “His reason is superficially plausible—the sole offense of some Hispanics in federal custody may be an illegal border crossing,” wrote Richwine. Alas, as numbers from the Bureau of Justice Statistics would prove, “Unz is wrong when he says that Hispanics are no more criminal than whites. Hispanics are, in fact, substantially more likely than whites to commit serious crimes, and U.S.-born Hispanics in particular are about two and a half times more likely.”
At his day jobs, on the mainstream right, Richwine wasn’t usually this blunt. He might cite the Bell Curve in an article for AEI’s magazine, but there was no thinkcrime there—he’d thanked Charles Murray in his dissertation. When he joined Heritage, Richwine wrote only rarely about immigration, applying his statistical acumen more often to public pension crises and student loans. “His mistake is that he wrote about a taboo subject,” Charles Murray told the New Republic yesterday. “And to write about IQ and race or ethnicity is to take a very good chance of destroying your career. And I really hope that doesn’t happen.”
It’s happening right now. According to Politico, Heritage is on the hunt for a PR guru who might spin away the Richwine story. On the paleo right, that’s being interpreted as a nolo contendere admission of thinkcrime. VDare’s writers have defended Richwine as a statistician whose work people prefer to hyperventilate about than to debunk, because they can’t debunk it. “The forces of orthodoxy have identified a heretic,” wrote John Derbyshire, who was laid off from National Review last year after writing that he’d educated his children about racial differences. “They’re marching on his hut with pitchforks and flaming brands. The cry echoes around the Internet: ‘Burn the witch!’ ”
Anyone could have predicted it. Richwine didn’t mind taking on taboos or talking to taboo people. That’s how immigration reform foes talk amongst themselves. That’s not how they’re going to stop the bill.
“In my estimation, our School gives too much emphasis on moving from findings to policy implications in scholarly work,” said Harvard’s Richard Zeckhauser. “In many cases, merely presenting the facts would be a preferable way to go. That makes it much harder for one’s opponents to dismiss what you say, or to accuse you of manipulating facts to reach policy conclusions. Moreover, I believe that policy conclusions usually rest on one’s underlying values. If one complements one’s empirical assessments with values issues, those assessments get questioned, particularly if one addresses a controversial realm of policy, as Richwine surely did in his dissertation. In many contexts, one’s work will have a long run greater influence on policy if the facts are left to speak for themselves.”
On Tuesday, ThinkProgress ran a story by Zack Beauchamp on Dr. Jason Richwine’s graduate dissertation on Hispanic IQ and immigration titled “The Inside Story Of The Harvard Dissertation That Became Too Racist For Heritage.” Thursday night, Dr. Richwine reached out to provide his side of the story. What follows is Richwine’s letter and Beauchamp’s response.
Jason Richwine writes:
This may disappoint some people, but there is no fascinating inside story of how I was awarded a PhD. The simple, boring explanation is that my dissertation is a solid piece of research. The “errors and omissions” that Zack Beauchamp claims to have uncovered exist only in a caricature of my dissertation. He knocks down a lot of straw men, but he doesn’t land any blows on my actual work.
Two factual corrections: First, my wife is not an immigrant. Second, I took the normal five years to complete my degree, not four, so readers can forget all that innuendo about sacrificing quality and depth for the sake of rushing.
Now for the substantive critiques. The extent to which self-identified Hispanics share a common genetic heritage is not important to my argument. As I explain on pages 76 and 77, the average IQ difference between Hispanics and non-Hispanic whites should be of concern because it is persistent over generations. Whether that persistence is due to genetics, environment, culture, or some other factor does not change the fact that the difference exists. It would be necessary to explore the biological basis for Hispanic identity only if my argument depended on a genetic transmission of IQ differences. It doesn’t.
I understand that Professor von Vacano has written extensively on the topic of Hispanic identity. And I also understand that scholars have a tendency to think their own specialty is hugely relevant to what everyone else studies. But, in reality, a long discussion of Professor von Vacano’s research interest would add little value to my dissertation.
I’m a bit bewildered by the rest of the critiques because they aren’t really critiques at all. The environment’s role in shaping IQ, the limits of IQ as a predictor of individual success, and the importance of non-cognitive abilities are all mentioned in my dissertation, sometimes in considerable detail. It’s difficult not to conclude that Beauchamp has intentionally ignored or downplayed my coverage of these issues in order to falsely portray my work as “sloppy.”
Take, for example, my conclusion regarding attempts to raise IQ. Beauchamp eventually acknowledges that I’m correct — that is, it is very difficult and perhaps impossible to permanently and substantially raise IQ through intervention programs. However, in what is supposed to be a devastating rebuttal of me, Beauchamp says these programs may still provide non-cognitive benefits. Strange — that sounds a lot like me! Page 70, footnote 20 of my dissertation:
This is not to say that Head Start or any other intervention inherently lacks value. Some programs may help children make non-cognitive gains in educational achievement and reduce their chances of committing crimes. These programs should be evaluated, using proper cost-benefit analysis, with all their strengths in mind, even if IQ enhancement is not one of them.
Or how about page 84:
When comparing individuals, the effect of IQ differences is often small. A large number of personality attributes, many of which are unrelated to IQ, affect a person’s ability to succeed in life. For that reason, an individual’s IQ score is merely a probability of future success, not a prediction from a crystal ball. For example, a person’s IQ affects his likelihood of completing college, but some college graduates are not very smart. Betting that an individual person with an IQ of 100 will complete more years of schooling than a person with an IQ of 95 is a risky gamble. The less intelligent person may be a very hard worker, while the smarter person could be lazy and unmotivated.
Does this look like the writing of, in Beauchamp’s words, an “IQ fundamentalist” who thinks IQ is “an almost-perfect guide to someone’s prospects for success in life”?
IQ is not the only important human trait — not by a long shot. Nevertheless, it remains an important predictor, on average, of many socioeconomic outcomes we care about. There can be no denying this. I continue:
However, if presented with two groups of 100 random Americans, one group with average IQ 95, the other group at 100, it is a virtual certainty that the smarter group will have higher educational attainment. In this way, IQ scores can be thought of as individual probabilities that aggregate into certainties in large groups.
That’s the crux of the issue.
The general claim that I ignored contrary evidence simply can’t be supported by a fair reading of the text. For example, much is made of my prominent citation of Richard Herrnstein and Charles Murray’s The Bell Curve. But I also discussed two major critiques of that book. On pages 82 and 83 of the dissertation, I even draw this conclusion: “It appears that Herrnstein and Murray’s critics have succeeded in establishing a larger role for the environment, without proving a lesser role for [IQ].” Is that something that a blind follower of Charles Murray would write?
Beauchamp seems to have decided a priori that my dissertation is one-sided, then viewed the entire work through that mental filter. He says I was “forced to concede” that environmental deprivation can adversely affect IQ. I did include environmental influences in my long discussion of what factors impact IQ differences, as any careful scholar would. Why Beauchamp characterizes this as a forced concession is not clear.
Regarding the quality of the datasets, that’s discussed in depth in chapter 2. The samples vary in size, but they all yield results pointing in the same direction. Furthermore, Beauchamp seems to think I haven’t noticed the critiques of Lynn and Vanhanen’s national IQ data. See pages 27 and 28 for a full discussion, in which I cite eight different academic references on that topic.
I could go on, but I’m already getting repetitive. Beauchamp ignores what’s actually in my dissertation so that he can say it’s full of omissions.
Substantive issues aside, another disappointing element in the article is the treatment of the quote from Christopher Jencks, who was my third committee member. The article uses the quote to imply that I ignored important parts of Jencks’ critique of my dissertation.
That never happened. In reality, my interaction with Professor Jencks was as normal as the rest of the process I followed in producing my dissertation. Like my other advisors, he gave me extensive written comments and suggestions. I revised the dissertation accordingly. I then sent Professor Jencks a 33-page document that detailed exactly how I revised the text in response to every single concern that he had expressed. In no case did I ignore a comment or fail to make revisions as I thought appropriate.
In response, Professor Jencks wrote to me in an email: “I think you have done a thorough and conscientious job of dealing with my comments, criticisms, and suggestions, and I am happy to approve it as it stands.” This didn’t mean he agreed with everything. He went on to say that he continued to be concerned with my use of ethnic categories like “Asian” and “Hispanic,” which he believes are inappropriately broad when talking about culture, and he felt that I left too little room for the differential effects of IQ on culture within ethnic groups. “That said,” he concluded, “I’m not asking for more revisions, just making suggestions for you to think about in the future.”
May I suggest that this is a completely normal situation in PhD programs? It would be a rare committee indeed if every member agreed with every data interpretation and policy judgment in the dissertation that they approved. My interactions with all my dissertation advisors, including Professor Jencks as the third reader, followed normal protocol from beginning to end.
Here is the truth about my dissertation: It’s a careful empirical analysis, firmly grounded in the mainstream of psychological science, vetted by a team of respected scholars, well researched, fully sourced, and a valuable contribution to policy discussions. I know, I know — what a boring reason to be awarded a PhD!
My thanks to Dr. Richwine for the factual clarifications. If only his treatment of my article, and his own dissertation, had been so forthright.
On the issue of his incomplete definition of the term “Hispanic,” Richwine suggests the only thing that matters is that the persistence of low Hispanic IQ on tests over generations. As it happens, I addressed this potential rebuttal at length in my original piece. The reason the definition matters, even if some pattern can be shown inside a group, is that it’s impossible to identify what that pattern means about the group and whether that pattern will continue unless you know what makes that group unique. As I put it:
Why do definitions matter if Richwine succeeds in showing a deep, persistent difference between so-called “Hispanics” and “whites?” Aside from the fact that it makes it impossible to figure out the scope of the dissertation (are Mexicans of largely European descent likely to have low IQs? What about African-descendent Brazilians?)…Without a proper definition of what he means when he says Hispanic, we have no way of evaluating the role that immigrants’ “Hispanicness” — whether that means shared genes, culture, or national background — plays in determining their IQ. Put differently, in order to know whether and how being Hispanic matters for IQ, we need to know what it means to be Hispanic. That, in turn, makes it impossible to evaluate how meaningful Richwine’s conclusions about the persistence of the IQ gap are or how they apply to any particular group of immigrants.
The purportedly exculpatory email from Professor Jencks he provides makes this point for me. In Richwine’s own summary, Jencks “continued to be concerned with my use of ethnic categories like ‘Asian’ and ‘Hispanic,’ which he believes are inappropriately broad when talking about culture.” This inappropriate broadness is precisely the point — they are so broad, I argue, as to make generalization about them meaningless without ample defense of why such a generalization is appropriate in this case. Richwine provides none, choosing to ignore the overwhelming literature on the social construction of race.
Similarly, Richwine misses my point on early childhood interventions and non-cognitive skills. The argument does not depend on whether Richwine mentions these factors occasionally in his dissertation — as Richwine points out above, I address his arguments on interventions specifically. Rather, my point was that he ignores the way in which such factors fatally frustrate his attempt to make broad predictions about immigrants based on their IQ. As I put it, “there’s simply no reason to think IQ matters enough to provide the juice for sweeping theories about the life prospects of entire groups of immigrants.” The proof of IQ fundamentalism is in the pudding.
For instance, on the issue of early childhood interventions, he does not attempt to explain whether or not the non-IQ related gains they produce (gains he consigns to a footnote) might be able to make up for any of the costs he associates with low-IQ immigration. For instance, on page 93, he argues that “Hispanics become less willing to play by the rules of the middle class when their low average IQ prevents them from joining it,” thus explaining why Hispanic immigration will produce more “underclass” behavior like dropping out of school and criminality. However, early childhood interventions can improve educational attainment and reduce criminality down the line — as he notes in his own footnote! Richwine pays lip service to factors other than IQ scores being important, yet edits them out of his substantive analysis.
This pattern repeats itself on the broader issue of non-cognitive traits. Richwine argues that (p. 100) “IQ has been linked to possessing middle class values, having a future time orientation, and cooperating in competitive games” in order to make his argument that Hispanic immigration will further lower social capital and “trust” inside the United States. These qualities bear intimate resemblance to non-cognitive traits like Conscientiousness or Agreeableness that either aren’t all that closely linked to IQ or, on some accounts, actually explain certain levels of performance on IQ tests. Yet Richwine never attempts to explore the connection between social trust and non-cognitive traits, or even establish that Hispanics lack the relevant non-cognitive qualities.
Essentially, Richwine suggests the supposedly lower Hispanic IQ will predict bad behavior without bothering to establish whether the immigrant populations might have or be able (with education) to get to higher levels of other traits that would counterbalance any IQ deficit. That sounds pretty “one-sided” to me.
I could go point-by-point on the other, lesser charges — for instance, his discussion of the flaws in the Lynn and Vanhanen data is hardly “full,” and he doesn’t consider criticisms of The Bell Curve in each case where it might be warranted. But, in Richwine’s words, “I’m already getting repetitive.”