Social scientists tend to aggregate Hispanic data rather than disaggregate by ethnic subgroups, usually for convenience. Running the kinds of studies, often surveys and experiments, that advance social scientific theory and understanding require large sample sizes for groups of people based on demographic characteristics (e.g., gender, ethno-racial identity), and aggregating Hispanic participants into a single group is assumed to facilitate inference at the cost of granularity. (Rodríguez-Muñiz, 2021; Mora, 2014) However, this practice can have real world consequences by limiting collective understanding of important subgroup variation. Research examining this topic in college admissions is an important social process that will likely become more opaque due to heightened scrutiny from the Trump administration and a conservative Supreme Court. Failing to consider subgroup variation within the Hispanic pan-ethnic umbrella could lead to faulty assumptions about who is enrolling in college, an important step toward redressing longstanding social inequalities and improving socioeconomic mobility.
“Hispanic, Other” and College Admissions
A 2024 study described variation in the ways that students identified as "Hispanic, Other" on college applications (Alvero et al., 2024a). When students selected "Hispanic, Other", they were asked to write in how they identified as Hispanic to capture backgrounds and heritages distinct from Mexican, Puerto Rican, and Cuban (the three categories explicitly named in many versions of the Hispanic question). The dataset included data from every single self-identifying Hispanic applicant to a large university system, and 12% of these students selected "Hispanic, Other".
After hand-coding each applicant´s response, the paper identifies an unusual trend: the most common background was directly related to Europe. For example, the most common "Hispanic, Other" response was "Spain", but applicants also wrote in "Portugal", "France", "Italy"; genealogical claims like "Born in Florida to Spaniards” and “38% Iberian Peninsula"; and claims through historical colonial relations, such as "Macanese" (formerly under Portuguese rule) and "Spanish Formosa" (the Spanish Empire ruled the northern tip of Taiwan, then called Formosa, for approximately 16 years in the mid-1600s). Students who included some European identity like this also reported significantly higher average incomes than all other Hispanic applicants. Hispanic identity is complex and rarely straightforward, and it is true that European heritage is integral to Hispanic identity, but these trends belie social and demographic realities in college admissions.
For example, most of the applicants are from California, and, from a purely numerical perspective, it would be reasonable to expect that people referring to Central American identity would comprise the largest group in the "Hispanic, Other" data (they were the second largest). Central Americans, like Mexican heritage people in the western U.S., have faced well documented obstacles and barriers to social equality related to racism, colorism, classism, anti-indigeneity, linguistic prejudice, and a number of other social maladies. The same cannot be said for people with European heritage, especially not at the same scale. But aggregating all groups in the same broad ethno-racial category, leads to a false equivalence in which these differences are flattened rather than heightened.
Most practically, these students also contribute toward enrollment counts, which eventually determine Hispanic Serving Institution (HSI) designations (at least 25% Hispanic undergraduate enrollment; see Garcia, 2019). While the future of the HSI program is up in the air, the program operates by allocating grant money to colleges and universities that enroll a high number of Hispanic students. The basic idea is that the money would go toward developing educational programming and opportunities to serve Hispanic students. But if schools are enrolling high numbers of students who claim Hispanic identity through seemingly dubious connections like "38% Iberian Peninsula", then how schools might serve their Hispanic students becomes less clear. The paper found similar trends, as the European- identifying applicants applied to the most selective campuses at higher rates than other Hispanic students. Disaggregated enrollment data from the university system also showed that these same students were disproportionately enrolled at the same highly selective campuses. As schools, including very selective ones as in this case, approach the threshold to become HSIs, it becomes much less clear whether schools recognize these trends, how they will address them with HSI funding support, and, maybe idealistically, how European-identifying Hispanic students see themselves within the broader Hispanic community on campus.
Technological and Political Implications
New technology used in admissions, both by evaluators as well as applicants, could hinder nuance even more. Another paper found that admissions essays created by generative AI (e.g., AI technology that generates content and is used in online interfaces like ChatGPT) are written most like applicants from more socially privileged backgrounds, such as those with college educated parents and people who come from neighborhoods with high socioeconomic status (Alvero et al., 2024b); essentially, they sound more like the European identifying applicants. The college admissions essay is a highly personal genre that allows admissions offices to develop a clearer sense of applicants. As students use AI more in their writing, it has the potential to elevate some voices while suppressing others. Admissions offices are also using AI at different stages of the evaluation process (though not to make formal, outright decisions), and at each stage, there is a chance these tools will homogenize Hispanic applicants in ways that do not account for important subgroup variation (Alvero et al., 2025).
Much of what is known about Hispanic students and the unique challenges they face to succeed in college is based on aggregated data. But as these and other studies show, analyzing subgroups can yield important insights into assumptions about Hispanic students, and how the pieces of the pan-ethnic puzzle fit together. Federal attacks on higher education could make this type of research more difficult, but academics could help illuminate these discrepancies and show how they hurt or, hopefully, benefit Hispanic students who have faced the most discrimination.
Citations
Alvero, A. J., Giebel, S., & Pearman, F. A. (2024a). Income and campus application disparities among European and non-European heritage Hispanic undergraduate applicants. PNAS nexus, 3(9), pgae337.
Alvero, A. J., Lee, J., Regla-Vargas, A., Kizilcec, R. F., Joachims, T., & Antonio, A. L. (2024b). Large language models, social demography, and hegemony: comparing authorship in human and synthetic text. Journal of Big Data, 11(1), 138.
Alvero, A. J., Sedlacek, Q., León, M., & Peña, C. (2025). Digital accents, homogeneity-by-design, and the evolving social science of written language. Annual Review of Applied Linguistics, 1-19.
Garcia, G. A. (2019). Becoming Hispanic-serving institutions: Opportunities for colleges and universities. Johns Hopkins University Press.
Mora, G. C. (2014). Cross-field effects and ethnic classification: The institutionalization of Hispanic panethnicity, 1965 to 1990. American Sociological Review, 79(2), 183-210.
Rodríguez-Muñiz, M. (2021). Figures of the Future: Latino Civil Rights and the Politics of Demographic Change. Princeton University Press.
About the author
AJ Alvero is an Assistant Research Professor at the Cornell University Center for Data Science for Enterprise and Society with departmental affiliations in Information Science, Computer Science, and Sociology.
