Harvard's Going Test-Optional. There's a Better Way.

SAT test preparation books sit on a shelf at a Barnes and Noble store June 27, 2002, in New York City. / Getty Images
December 22, 2021

The big test is going bust. Last week, the Washington Post reported that the proverbial "school in Boston" will extend its test-optional admissions policy, which it adopted in 2020, for another four years. Harvard is not the first elite university to stop requiring applicants to submit SAT or ACT scores—the University of Chicago made the change in 2018 and the University of California will no longer consider standardized tests at all. But Harvard's decision shows testing critics have seized the pinnacle of the ivory tower and licenses other institutions to follow its example.

The critics argue standardized tests are biased against poor and minority students and unreliable predictors of success in college. But the truth is, debates about equity are mostly about public relations. As I argued recently in The Week, the main reason Harvard is dropping test requirements is that it's good for Harvard.

There's a better way, though. A lottery of qualified applicants would recognize achievement without fetishizing small and easily massaged differences. By randomizing an element of the process, it would also reduce the anxiety of applicants who needn't worry about pleasing anonymous gatekeepers. Above all, a lottery would transform college admissions from a rigged competition to a game of chance that resists corruption.

To hear Harvard tell it, you'd think the decision to go test-optional was altruistic. The official announcement attributed the policy to the COVID pandemic, which has interfered with test administration as well as interrupting instruction for many students. But it's hard to believe political and legal considerations weren't big factors, too. Like most American universities, Harvard has enthusiastically adopted the rhetoric of "diversity, equity, and inclusion" and devoted significant administrative resources to those principles. It's also currently defending itself from a lawsuit alleging discrimination against Asian applicants, which could reach the Supreme Court.

Neither of those explanations is sufficient. Although Harvard has wide discretion over admissions, it does want to keep average test scores high in order to maintain a top ranking with U.S. News and World Report and other influential surveys. That makes it risky to accept too many applicants with low scores.

Test-optional policies solve that problem because applicants who don't submit scores don't contribute to the average. Under less pressure to sustain the ranking, admissions officers can more easily reward applicants with other qualities that they consider desirable. They can also entice students with little chance of admission to apply, further pumping up the application pool and reducing the admission rate.

Public debate focuses on the implications of these policies for "underrepresented" minorities—that is, blacks and Hispanics. Even when scores were required, a 2019 study by academic economists (including an expert witness for plaintiffs in the federal case against Harvard) found these applicants enjoyed significant advantages over Asian (and to a lesser extent, white) applicants.

Yet the perennial debate about affirmative action distracts from the larger distortion. Whatever their race, legacies, athletes, and other specially recruited applicants enjoyed much higher admit rates than their test scores would predict. To put it simply, getting into Harvard is not that difficult for sports heroes or the children of alumni and donors with otherwise marginal credentials. For all the talk about equity,  more than a third of Harvard's class of 2022 is composed of legacies alone.

Test-optional admissions compounds these applicants' existing privileges. Enabled by experience, connections, or money to accumulate good grades, bespoke internships, and polished essays, they'll find no difficulty generating strong applications without a meritocratic paper trail.

Defenders of test-optional policies argue admissions officers can easily distinguish between the simulacra of achievement and the real thing. Even if that's true, it ignores the role of institutional interest. Universities want to enroll students with family connections, athletic ability, and donor potential because that's how they raise money. Contrary to the common assumption that selective universities are looking for geniuses, academic achievement is just one criterion among many.

But it's easier to complain about existing admissions practices than to propose an alternative. After all, Harvard and its most exclusive peers offer only about 2,000 undergraduate places each year. With around 25 applicants for every seat, they need some way to distinguish among them.

That's where a lottery comes in. Proposals have circulated in the academic press since the late '80s and attracted renewed attention as acceptance rates have dropped into the low single digits. The basic idea is simple. All qualified applicants would have their names placed into a hat, from which the appropriate number of admits would be randomly selected.

Many differences among proposals revolve around so-called threshold criteria. Minimal versions could be limited to a cutoff score on the SAT or other standardized test. More elaborate strategies might allow applicants to choose from a list of qualifications, perhaps including GPA, or non-academic activities. Since selective universities receive far more qualified applications than they can accommodate, though, the final decision would always be determined by chance.

Shifting the decision from admissions officers to a roll of the dice would have finally principal advantages. The first is that it would seriously reduce the burdens involved in applying. Competitive admissions currently involves a formidable amount of paperwork. Online applications have somewhat reduced the burden, but a lottery based on one or a few criteria would be far easier. Imagine if all you had to do to apply to college was submit a test score, perhaps in combination with uploading a transcript.

Ease of application doesn’t mean much in itself, though. A greater benefit would be relieving some of the scrutiny applicants feel. Despite ritual protestations that colleges don't assess individual worth, the present system makes applicants feel as if their whole being is subject to judgment. Having organized their lives around pleasing hypothetical admissions officers, they can be devastated to learn they haven’t made the cut. The unfavorable outcome of a lottery is still disappointing, but doesn't feel like a personal judgment.

A greater benefit attends those who do get in, though. Elite admission practices encourage the fantasy of moral superiority. Rather than recognizing the good luck that almost certainly provided background conditions for their academic success, those admitted to Harvard or its peers can be forgiven for thinking they've proved they deserve the heightened economic opportunities and social status they'll almost certainly enjoy. If anything, the parents are worse—as vendors of college-branded sweatshirts and bumperstickers well know.

Finally, a lottery would restrict institutions' ability to "craft the class" in self-serving ways. Despite their tributes to the moral and educational benefits of ethnic and cultural diversity, the enormous advantages enjoyed by children of donors and alumni show what elite universities think is important.

A lottery is no panacea, of course. One of the most powerful objections is that it wouldn't do much to address the correlation between academic success and socioeconomic status. Even if applicants only had to meet a minimal cutoff of qualification, richer applicants would be more likely to meet any plausible threshold. By replicating the distribution of luck in life, a lottery could be even more random than we'd think.

The unfairness of life isn't an objection to the lottery in particular, though. As the journalist Freddie deBoer has argued, educational outcomes are stratified because American society is stratified—by income, education, and intellectual ability. The demand that college admissions somehow unravel those conditions by refusing to consider relatively commensurable, stable measures of ability like standardized tests is, at best, magical thinking.

Contrary to standard criticisms, moreover, there's considerable evidence that standardized tests do predict applicants' undergraduate performance. Last year, a faculty taskforce at the University of California concluded that the tests were more reliable than high school GPA and other measures. That's partly because variation in grading standards makes it hard to compare students with similar transcripts. Widespread grade inflation makes the problem worse. Although scholarly debate continues, it's hard to avoid the conclusion that the reason elite universities want to reduce the influence of test scores is so they can recruit less qualified students who nevertheless bring money, prestige, or political credibility along with them.

Of course, those same considerations make it unlikely that Harvard and its counterparts would voluntarily adopt an admissions lottery. Discretion, secrecy, and nebulous moralism aren't only in their material interest—they're an important part of their mystique. Like the Wizard of Oz, they'll have to be forced to remove the curtain. Maybe Senator Tom Cotton and other politicians who've threatened to tax big endowments can help.

Samuel Goldman is an associate professor of political science at George Washington University.