Asymmetric Information and Test Optional College Admissions Policies

| Chrissy Ramakrishnan | December 18th, 2018 |

Issues surrounding the use of standardized test scores for college admission have become particularly salient over the past decade. Many students view them as unfairly narrow assessments of their potential, and scores have been shown to be correlated with a student’s socioeconomic background. Because of the controversies surrounding standardized testing, test optional admission policies have been growing in numbers; test optional schools do not require students to submit a standardized test score with their application. These policies could attract academically strong candidates who may struggle with the strict expectations of standardized tests and generally aim to help students from disadvantaged backgrounds.

Recently, the highly selective University of Chicago announced that they were dropping their ACT/SAT score requirement, becoming the first top-10 research university to do so. The policy, packaged as the UChicago Empower Initiative, aims to “enhance the accessibility of its undergraduate College for first-generation and low-income students”. Clearly, a narrative has emerged that test optional admissions policies can give rise to the sort of socioeconomic diversity that most universities seek in the 21st century.

Research on whether or not test optional policies increase institutional diversity reveals mixed results, often because researchers employ very different methods of evaluation. Simply looking at descriptive statistics of universities before and after they adopt a test optional policy does not address their true impact, as the policies usually come coupled with larger initiatives to serve low income and minority students. However, the few studies that have attempted to isolate the effects of test optional policies on institutional diversity have found that they have modest, if not minimal, influence. For example, one study from the American Educational Research Association (AERA) used quasi-experimental techniques to assess the relationship between test-optional policies and growth in the proportion of low income and minority students at policy adopting universities. They found that, on average, the policies on their own do not enhance the diversity of test optional universities. They affirm that “test-optional policies overall have not been the catalysts of diversity that many have claimed them to be.

Viewing these results through the lens of an economic issue about asymmetric information offers an explanation for why these results may not be surprising. Asymmetric information refers to situations where one party in an exchange, the “buyer”, does not have as much information as the “seller”. This can result in form of market failure, known as adverse selection. Adverse selection is the tendency for the unknown information to make the good or service undesirable to the uninformed party. Two ways to combat adverse selection include screening, which involves the individual with less information attempting to learn as much as possible about the other, and signaling where the informed individual attempts to set themselves apart via translating information deemed important to the buyer.

Colleges are interested in students that will succeed and graduate, and therefore it is in their best interest to admit academically successful students. For them, “undesirable” would be admitting students who are not prepared to succeed at their college. College admissions departments are entrenched in a market for information, applicants signal to the college that they are worthy of admission, and colleges set criteria for screening them.

However, is eliminating the amount of information available to admissions offices, namely removing a test score, creating undesirable outcomes? That answer depends on whether or not there will be additional pressure on the existing information. Therefore, we must pay special attention to the shifts in demand for other signals at test optional colleges. In the absence of test scores, what sorts of substitutes exist to determine the quality of an applicant?

Belasco, Rosinger and Hearn, the authors of the AERA paper, note that the signals that admissions officers must rely on in the absence of a test score, are subject to the same sorts of issues that are alleged to weaken the value of standardized test scores. In fact, test-optional admission policies may perpetuate stratification within the postsecondary sector, in particular, by assigning greater importance to credentials that are more accessible to advantaged populations. Without access to standardized test data for every applicant, test-optional colleges may rely more heavily on school-specific measures, such as the strength of curriculum or extracurricular involvement, to draw comparisons between prospective students. However, the availability of advanced (AP, IB, and honors) courses is unequally distributed across socioeconomic groups. Data from the Office of Civil Rights, a subagency of the US Department of Education, confirms that minority students have less access to rigorous college preparatory classes.

Test optional admission policies may result in perpetuating the same drivers of inequality as those requiring standardized test scores do, namely they favor those wealthy and socially knowledgeable enough to create the right set of signals that tell the college they will be successful. Less advantaged students may have the same potential but lack the access to these sorts of signals.

These conclusions parallel findings from another study that found following the introduction of a ban on credit checks for job applicants, the likelihood of unemployment in African Americans actually rose. The authors explicitly conclude “it appears that the prohibition of credit screening and the increased emphasis on other signals may actually, relatively, hurt minority applicants”.

Additionally, if we were truly to believe that students who do not submit test scores were evaluated the same as their submitting counterparts, why would there be such a buzz among applicants about whether it is advantageous to them or not to submit their test score? We would think that test optional policies would lead students to forgo taking expensive and stressful tests. Furthermore, not submitting a test score may signal to an admissions officer that the applicant has a lower score. This means that the other signals that the admissions officers are relying on to determine application quality might need to be even stronger for that of a non-score submitting applicant if they are to be competitive.

In our eagerness to eschew standardized tests from admissions criteria, we must not forget to truly consider what the effects of these policies will be. Is the alternate always the better option? This is not meant to make us retreat from test optional policies, which can do a great deal of good in terms of communicating that a college is mindful of the challenges that applicants face. However, this is a way of framing their potentially unfavorable effects in a way that allows us to understand and address them. This serves to caution us and warn that a test optional admissions policy or even abolishing the use of standardized testing in admissions may not necessarily result in a better process.

The potential pitfalls of test optional policies are not to be taken lightly. Data on test optional colleges has shown that underrepresented minority, first generation, and female students are less likely to submit scores. Therefore, if the type of signal substitution described above is at play it could be disproportionately affecting underrepresented minority students in the absence of other policies.

This mainly theoretical musing on the application of an economic concept to college admissions just may have very real implications. Unless test optional policies are coupled with targeted initiatives in other areas of the application process such as holistic admissions and need-based financial aid programs, they may do little to increase institutional diversity, especially at selective institutions. Other studies have shown that colleges do see increases in applications from underrepresented minority and low-income students after adopting test optional policies, but they also found that the policies place a greater financial-aid demand on colleges. If additional financial aid is not offered to support students, test optional policies can do little on their own. We are quick to praise the University of Chicago for their apparent commitment to diversity, but is this immediately warranted without knowing how their admissions process is changing as a result of the lost signal? Colleges must think carefully about what their process will look like with a test optional policy, which may become increasingly relevant in the near future as selective colleges start to consider changing their admissions practices.

Ultimately, this question boils down to a discussion of what we think the informative value of a standardized test score is relative to other pieces of an admissions profile. More information is not always better if the processes that produced that information are ineffective and the new information offers little independent value. This points to the idea that discussions around test optional policies are truly a debate about the signaling quality of standardized tests. Test scores are meant to inform admissions departments of a student’s level of college readiness, and perhaps we are not so sure of their capacity to do so anymore.

Chrissy Ramakrishnan is a junior majoring in Economics and Sociology from Franklin, Wisconsin. Chrissy is interested in higher education policy and enjoys studying issues related to education from multiple perspectives.