Sometimes a narrative is just too good to give up, even when the facts don’t support it. This seems to be the reason why some supporters of racial preferences in college admissions keep citing bad research in legal briefs before the Supreme Court.  

In the two cases challenging the race-based admissions practices of Harvard College and the University of North Carolina, more than a dozen briefs cite “The Shape of the River,” a 1998 book by William Bowen and Derek Bok, to support such race-based admissions.

That research, however, largely has been discredited. The authors of the briefs citing it either haven’t done their research or think they can pull a fast one on the Supreme Court as the justices consider the two cases, Students for Fair Admissions v. President and Fellows of Harvard College and Students for Fair Admissions v. University of North Carolina.

The Shape of the River” hit bookshelves in 1998 to great fanfare in liberal media outlets because it purported to prove that racial preferences in higher education help black students to make more money after graduation. Indeed, it claimed that racial preferences were responsible for the growth of the black middle class, and that without those policies, blacks would suffer.

The book also purported to disprove the “mismatch effect”—the documented phenomenon that lowering admissions standards for racial minorities actually reduces the number of minorities entering academia and high-paying professional jobs.

But the book did no such thing.

The authors, Bowen and Bok, studied the graduation rates and post-graduation careers of students at 28 colleges, concluding that racial preferences were a great benefit to black students. In fairness to them, their study did prove one thing: Racial preferences favored black applicants at the expense of white and Asian applicants. That was a big deal because up to that point, many liberals denied this result. 

But from there on out, the book’s major conclusions weren’t reliable. Its core claim—that the mismatch effect didn’t exist—was based on several serious mistakes.  

First, the authors failed to separate black students admitted to elite institutions because of academic merit from those who were admitted under racial preferences. The authors could have disaggregated that data, but they didn’t. As a result, the study didn’t include data about the specific group at issue—an error that made the study, in the words of economist Thomas Sowell, “the statistical equivalent of ‘Hamlet’ without the prince of Denmark.”

Second, the authors looked only at students’ SAT scores and considered no other academic credentials. Is a student at Penn State with a score of 1200 as academically advanced as a student at Princeton with a 1200? Probably not.

Students are more than a single test score—Bowen and Bok argued as much before they wrote “The Shape of the River”—so we learn little, if anything, by assuming that two students with the same test score are the same.

Third, Bowen and Bok didn’t even compare students with the same SAT scores; they compared scores in broad bands. As professor Gail Heriot, a leading expert on racial preferences explains, comparing bands “is what statisticians do when they set out to muddy the waters.”

This error meant that Bowen and Bok didn’t actually test the mismatch hypothesis that they claimed to disprove. The hypothesis, to quote Sowell again, is that “the larger the differential in academic qualifications between black and white students at a given institution, the larger the racial differential in failure to graduate tends to be.”

To test this hypothesis, Bowen and Bok would have needed to look at the data from individual institutions. Instead, they looked at aggregations of data from institutions with different students and different SAT levels.

Incidentally, another study, “America in Black and White,” actually did look at data from individual institutions, and it confirmed the mismatch effect.

It gets worse for Bowen and Bok, however, because their own data demonstrated the mismatch effect despite their efforts to hide it.

Another study, “Reflections on ‘The Shape of the River,’” looked at Bowen and Bok’s report and found some shocking hidden conclusions. For example, Bowen and Bok reported that 8 of 10 black students in the study collected a diploma—a number well above the national average.

However, flipped on its head, this same statistic tells a different story. Whereas 2 of 10, or 20%, of black students failed to graduate, only 6% of white students failed to graduate. In other words, Bowen and Bok’s own report showed that even at elite institutions, the dropout rate for black students is 3.3 times that of white students.

One wonders what other revelations might be hidden in Bowen and Boks’ raw data. Researchers are unlikely ever to know because, in bold defiance of academic transparency, the authors refuse to make their raw data publicly available.

These are only a few of the errors in “The Shape of the River.” So many studies have done so much damage to it that serious researchers of affirmative action won’t rely on its claims about the mismatch effect.

And yet, the book remains a favorite citation for ideologues and activists. In the current cases at the Supreme Court, more than a dozen amicus briefs cite “The Shape of the River” in defense of racial preferences.

This looks like a commitment to narrative over facts.

The great irony of this wrongheaded commitment is that it actually harms the black students it claims to want to help. If not for racial preferences, we’d have more black doctors, engineers, and professors than we have today, as Heriot explains here.

It’s no good that this commitment to a false narrative is winning out over reality, at least for some people. But for others with more open minds, it provides an important warning: Beware of experts peddling statistics that confirm your beliefs; doubt them always, and double-check their work.

Have an opinion about this article? To sound off, please email and we’ll consider publishing your edited remarks in our regular “We Hear You” feature. Remember to include the url or headline of the article plus your name and town and/or state.