The Iron Triangle of College Admissions: Institutional Goals to Admit the Perfect First-Year Class may Create Racial Inequities to College Access

It is back-to-school season and college campuses across the country are buzzing with excitement to welcome students! But the average person is unaware of the decisions and tools used to determine who gets an acceptance letter and who does not. We created a short video to demystify this process, known as enrollment management, and the use of predictive analytics to create a first-year class.

The two-minute video shows how the automation of enrollment management, with the use of algorithms, locks in existing racial inequities to access higher education. Algorithmic racial biases are innately embedded in data which disadvantages Black and brown students in their pursuit of higher education.

Students of color are falling through the cracks in not being recruited or offered acceptance based on decisions from algorithms that do not consider their humanity. Rather, colleges use data points that are highly correlated to race and socioeconomic factors to make these decisions, which disproportionately disadvantage Black and brown students.

If colleges do not act swiftly to mitigate racial biases in their algorithms, students of color will continue to be denied access to four-year institutions and racial educational attainment gaps will continue to widen. This time of racial reckoning in the United States means it is also the time to remove these barriers so all students have equal access to higher education.

This blog post highlights some of the contributing factors to the racial inequities to college access within enrollment management, while providing recommendations for colleges and universities.

The Iron Triangle

At the core of the problem is the fact that many colleges and universities are immersed in an arms race for the most desired students, while juggling institutional priorities. Institutions have to enroll a class that achieves a balance of diverse and highly academically qualified students, while meeting their bottom line.

This juggling of institutional priorities of diversity, academic quality, and revenue puts colleges in a challenging position we refer to as the iron triangle. The theory behind the iron triangle is that the shape of a triangle is fixed. All three points cannot share the top position

The conflict enters when colleges attempt to achieve diversity, academic quality, and revenue at similar rates of priority. A pursuit of one of the goals can disrupt the other two, inhibiting advancement of the others. Either there will be a peak at the top with two level points at the bottom (as displayed in the figure, where the typical priority for institutions is revenue), or there will be a peak at the bottom, with two level points at the top. Most institutions prioritize revenue to keep the university running and often negate diversity.

While tuition keeps the lights on and academic quality gets high U.S. News & World Report college rankings, which generates more applicants from higher-income families. Failing to prioritize racial diversity puts higher education back in its archaic role of sorting and sifting, which perpetuates racial and socioeconomic inequities. It may be doubted that any one institution can realize all three of these goals at once, but all institutions must try.

Institutional Recommendations

Institutions can make transformative changes with three critical actions to mitigate the racial equity gaps to college access within enrollment management:

• Think critically about data.

The use of data points highly correlated with race such as zip codes, SAT/ACT scores, family income, private vs. public high school, and more, has racial equity implications. If institutions do not use a critical lens throughout the process, these predictive models will continue to deliver an output that does not actually reflect a student’s academic potential but rather reiterates systemic inequities that are rooted in the biases of such variables.

• Use multiple models.

Identify models which are less averse to admitting Black and brown students. For example, in addition to traditional methods like OLS regression, institutions should explore more sophisticated statistical strategies like neural networks that give the feature to rank students. Comparing multiple statistical strategies lets enrollment managers make their decisions based on models that promote a racially diverse class.

• Bring multiple departments together.

Build on the strengths and various perspectives of experts across campus. Admission officers, financial aid officers, academic support staff for racially diverse students, institutional researchers, and others all working together can help shift institutional culture toward enrollment goals that prioritize racial/ethnic diversity.

These three actions can help four-year colleges and universities begin to address the challenges of the iron triangle to admit a more racially diverse class.

Enjoy what you read? Subscribe to our newsletter to receive updates on what’s new in Education Policy!