To Dismantle Structural Racism in College Recruiting, Policymakers and Advocates Must Play Offense, Not Defense
Following the Supreme Court decision to ban race-conscious admissions, which tools can be used to uphold our commitment to racial justice in college access? Many higher education leaders have highlighted the importance of identifying and recruiting prospective students earlier on, since the Court’s decision applies to admissions decisions.
More broadly, the movement for racial equality in college access is at a crossroads. The conventional progressive approach is to identify alternative practices, which can collectively serve as a substitute for affirmative action. Some have suggested that colleges target zip codes with large shares of Black, Latino, and Native American students in their recruiting efforts.
We recommend a more radical, critical approach, which views affirmative action as a racially progressive practice sitting atop a structurally racist system of college access. Rather than adding bits of progressivism to this system, we should dismantle the racist structures that determine whether and where students enroll in college.
This approach is motivated by scholarship from the sociology of race, which suggests that when overt prejudice declines, systematic racial inequality is maintained by structural racism. Structural racism is a form of systematic racial bias whereby processes viewed as neutral or common sense systematically advantage dominant groups and disadvantage marginalized groups. In his seminal statement on structural racism, American sociologist Eduardo Bonilla-Silva states that “the only way to ‘cure’ society of racism is by eliminating its systemic roots.” As we describe below, the process of recruiting students to college is built upon a foundation of racial segregation and, therefore, amplifies the effects of historic and contemporary racial segregation.
We have been studying recruiting for several years, focusing on the central rather than peripheral recruiting efforts of colleges. Following the Supreme Court ban on race-conscious admissions, many colleges have indicated they will expand outreach to underrepresented students. When colleges refer to ‘outreach’ programs for marginalized students, these efforts often happen in isolation — and often in separate offices — from central recruiting efforts spearheaded by the office of admissions or the office of enrollment management. We situate recruiting within the process of enrollment management, and summarize results from our research, which highlights subtle, everyday structures that disadvantage students of color. We hope that this information encourages college leaders and policymakers to interrogate racialized structures within their central recruiting pipeline rather than adding progressive practices to the periphery.
Figure 1 shows the “enrollment funnel,” a conceptual model used by the enrollment management industry to describe stages in the process of recruiting prospective students. The top of the funnel is the pool of all “prospects,” who are the prospective students that colleges would like to convert into enrolled students. “Leads” are prospects whose contact information has been obtained by a college. “Inquiries” are prospects who have reached out to a college, either on their own (e.g., sent ACT scores) or in response to a solicitation (e.g., a targeted email). The practical purpose of the enrollment funnel is to inform interventions that increase the probability of “conversion” from one stage to another. For example, colleges use financial aid packages to convert admits to enrolled students. Whereas affirmative action focuses on which applicants are accepted, the enrollment funnel shows that debates about racial equity in college access should consider all stages that contribute to college enrollment.
Figure 1: The Enrollment Funnel (Source)
Our research has examined two recruiting interventions at the top of the enrollment funnel that impact the college access and success of millions of students each year: purchasing “student lists,” which contain the contact information of desirable prospects (i.e., this is how colleges obtain leads); and off-campus recruiting visits, in which admissions representatives travel to local high schools, attend college fairs, and host hotel receptions in an effort to generate student interest. Visits influence where students apply and enroll. Colleges also value visits as a means of maintaining positive relationships with guidance counselors at feeder schools.
In one study, we analyzed the characteristics of public high schools that receive recruiting visits by public research universities. The recruiters from most universities in our sample made more out-of-state recruiting visits than in-state visits. These out-of-state visits systematically targeted affluent, predominantly White public high schools.
In a spatial analysis of recruiting visits, we found that high schools in White communities were more likely to be visited by college recruiters, receive multiple visits from the same university, and be visited by more than one university than high schools in communities of color with comparable income and education levels. Findings also reveal that universities engage in “recruitment redlining” — the circuitous avoidance of predominantly Black and Latino communities along recruiting visit paths. For example, Figure 2 shows public high schools in the Los Angeles metropolitan area that were visited (filled circle markers) and not visited (hollow circle markers) by recruiters from the University of Colorado at Boulder. Color keys indicate population race/ethnicity at the census tract level. The visit path in the Los Angeles area follows predominantly White coastline communities that are historically rooted in Jim Crow segregation and skirts Black and Latino communities in the city center that are stereotyped by perceived problems of poverty and crime. These visit patterns are based on the historical educational disenfranchisement of communities of color and amplify the effects of historical disenfranchisement.
The recruiting visits of selective private colleges are more problematic than those of public research universities. In a 2018 New York Times op-ed, we noted that many selective private colleges target schools in high-income neighborhoods. For example, public high schools visited by Connecticut College were in zip codes where the average household income was about $122,000. By contrast, public high schools that did not receive a visit were in zip codes where the average household income was about $65,000.
Figure 2: Recruitment Redlining of Black and Latinx Communities in the Los Angeles MSA (Source)
Another analysis shows that selective private college recruiters visit a dramatically disproportionate number of private high schools, given the number of students these schools enroll. Public high schools vastly outnumber private schools, and only about 9% of K-12 students in the U.S. attend a private school. For example, Figure 3 shows that Tulane University made 963 visits to high schools in 2017 and 48% of these visits were to private schools. However, if Tulane recruiters visited public and private high schools at the same rate, then only 18% of visits would have been to private schools.*
On average, private high schools enroll a much higher proportion of White students than public high schools. And most selective private universities enroll a large number of White students who attended a private school. This is no coincidence, but rather the intended result of recruiting efforts that target affluent, predominantly White private schools.
Figure 3: Actual versus Proportional Number of Private School Visits for Selective Private Universities (Source)
We also studied the student list business. Colleges identify and recruit prospective students by purchasing “student lists” from the College Board, ACT, and other vendors. These lists contain the contact information of prospective students who satisfy the “search filter” criteria (e.g., test scores, GPA, zip code) specified by the college making the list purchase. Student lists are a key input for undergraduate recruiting campaigns because purchased “names” make up the pool of prospects who receive subsequent recruiting interventions (e.g., mail, email, social media) designed to push them through the inquiry, application, and enrollment stages of the enrollment funnel.
Our recent study analyzing student lists purchased by public universities finds that Black, Latino, and Native American students are more likely to be systematically excluded from list purchases. These exclusionary patterns are driven by two factors. The first source of exclusion is racial inequality in the underlying database of prospective students. For example, the College Board’s database largely consists of students who took one of their tests. However, rates of test-taking differ by race.
The second source of exclusion is the “search filters” that enable universities to control which prospect profiles they purchase. Although colleges choose filters based on their preferences, which may include equity-driven goals, these choices are structured by which behaviors the filter allows and which behaviors the filter encourages. Colleges motivated by equity-driven goals may directly filter by race and ethnicity. However, other search filters are highly correlated with race, and colleges may be unaware that their purchases systematically exclude students from communities of color. For example, a university may filter prospects based on AP exam score, but who attends schools with robust AP curricula? A university may purchase prospects in high-income zip codes, but residential redlining has often excluded people of color from living in high-income zip codes.
We highlight search filters that are based on geodemography as particularly problematic. Geodemography is a branch of market research that classifies consumers based on information about where they live. For example, College Board’s Segment filter integrates demographic, academic, and historical college-going data to categorize every U.S. Census tract into one of 33 “neighborhood clusters” and every high school into one of 29 “high school clusters.” For example, a Census tract in Bethesda, MD, may be in the same neighborhood cluster as a Census tract in Evanston, IL. Universities may utilize Segment to select all prospects from neighborhood cluster 33 and/or high school cluster 22. Because geodemographic filters are based on the characteristics of schools and neighborhoods and because schools and neighborhoods are racially segregated, these filters may perpetuate racial inequalities in college access.
For example, Figure 4 taken from our study shows the racial (and income) composition of student lists from eight orders by a public research university that utilized Segment filters in combination with academic filters (see Figure 4 notes for filter details). Figure 4 compares racial characteristics of prospects whose profiles were purchased via Segment to the characteristics of all public high school students in four metropolitan areas (right column). In New York, for example, White and Asian students comprised 58% and 27% of the prospects whose profiles were purchased via Segment, respectively, but make up 30% and 9% of students in public high schools in the metropolitan area. By contrast, Black and Hispanic students comprised just 1% and 8% of prospects, respectively, despite making up 26% and 34% of students in public high schools. Figure 4 also shows that the prospects whose profiles were purchased via Segment were also from households that earned much more than the average median income in New York (left column). Similar racial and income patterns are evident across the other three metropolitan areas.
Figure 4: Segment Filter Prospects by Metropolitan Area (Source)
The findings from this research have implications for college leaders and policymakers. College leaders who are serious about achieving racial equity in college access should begin by interrogating their own recruiting practices. For example, how do the high schools and community colleges they visit differ racially, socioeconomically, and geographically from those they do not visit? How do the prospect profiles they purchase differ racially, socioeconomically, and geographically from those they do not purchase? However, rooting-out structural racism from the core enrollment funnel requires greater commitment than adding outreach practices to the periphery. Public universities with stable state support are prepared to make these changes. Unfortunately, the business models of selective private colleges are inexorably tied to serving affluent, White constituencies, and we anticipate that these institutions will make marginal changes.
For the policymakers and civil rights advocates who have spent decades defending affirmative action, now is the time to play offense. Backdoor substitutes for affirmative action — such as buying names of prospective students who live in predominantly non-White zip codes — are likely to face conservative legal challenges in the future. And targeted recruiting in schools and communities of color seems like a tacit agreement to ignore the widespread and longstanding recruiting practices that target affluent, predominantly White schools and communities and benefit White students. The more practical and effective approach is to target structural racism in the broader system of college access, which is the root of the problem.
Over the past 30 years, the enrollment management industry has grown dramatically in scale, sophistication, and importance to college access. While researching student list purchases by public universities, we learned that many universities outsource student list purchases to enrollment management consultants, such as EAB, Ruffalo Noel Levitz, and others. Ruffalo Noel Levitz, for example, offers a product that calculates the number of names to buy in each zip code, based on predictive modeling of past cohorts. Unfortunately, predicting future behavior based on the past amplifies historic inequities and reinforces existing structural barriers to equality of opportunity. Thus, such recruiting practices and products are structural barriers to equality of opportunity.
The broader enrollment management industry — including consultancies, vendors, and the products they make — should, therefore, be subject to greater legal and regulatory scrutiny from federal regulatory agencies — including the Federal Trade Commission and the Consumer Finance Protection Bureau — that are tasked with enforcing laws that protect consumers from unequal and unfair treatment. We recommend that higher education legal experts transition from defense to offense by raising legal challenges to products and practices in the broader enrollment management industry.
*Visiting public and private high schools at the “same rate” means that the total number of visits to high schools would remain the same but each public high school and each private high school would have an equal probability of receiving a visit. When calculating the hypothetical, proportional number of visits to public versus private schools for each university, we only consider high schools in states that received at least one visit to a public or private high school from that university. Our rationale for this decision is that is unhelpful to include schools in states that university ignored entirely. See our study for more details on methodology.
Karina Salazar is an assistant professor in the Center for the Study of Higher Education in the College of Education at The University of Arizona.
Ozan Jaquette is an associate professor of higher education in the University of California Los Angeles Graduate School of Education & Information Studies.