Abstract
Fraternity and sorority participation at large US universities correlates negatively with the size of the applicant pool. In a cross-section of 24 universities, the Pearson correlation between applicants and Greek participation is r = -0.64 (t = -3.9, p ≈ 0.0008). The relationship is concentrated among public institutions (r = -0.85, p ≈ 0.0003); among private institutions it is weaker and not statistically distinguishable from zero (r = -0.52, p ≈ 0.10). Against a fixed panel of 95 schools, applications grew 3.27× between 2001 and 2022 while admit rates fell from 58.4% to 45.1%. Greek participation snapshots show persistent heterogeneity rather than convergence: application scale concentrates, campus social structure does not.
1 The cross-section
The pooled correlation of r = -0.64 is statistically solid: a two-tailed t-test on the correlation gives t = -3.9 with 22 degrees of freedom (p ≈ 0.0008). But Figure 1's coloring carries the substantive point — the slope is a public-university phenomenon. Private universities occupy both extremes (Washington and Lee University at 72% and several schools at essentially zero) with little relationship to applicant volume, and their subsample correlation cannot rule out zero. “Bigger applicant pool, less Greek life” is, above all, a statement about public flagships.
2 The two-decade backdrop
The panel is held fixed across years to avoid trend breaks created by schools entering the data late. Within it, growth concentrated at the top: the largest applicant magnets pulled away from the rest of the top 200, and admit rates tightened from 58.4% to 45.1%. Greek participation is overlaid as snapshot evidence (it is not a federal annual series): the latest cut shows top-50 schools averaging 24.9% participation against 29.0% for ranks 51–200.
3 Heterogeneity across the ranking
The rank strip separates two claims that are often mixed together: application scale is concentrating (true in this panel) and campus social organization is converging (not true). Within every rank band, Greek participation spans nearly its full range — evidence that campus culture is set locally, not by admissions market position.
4 Geography
Top states by current metric
- CA 27
- NY 25
- NC 12
- TX 11
- FL 9
- MA 8
- IL 8
- OH 8
5 Notes on inference
The cross-sectional correlation is descriptive, not causal; public flagships differ from private campuses in housing markets, demographics, and social calendars in ways this design cannot separate. The private-school subsample (n = 11) is small: its non-significant correlation means the data cannot rule out zero, not that an effect is proven absent. All statistics are computed at build time from the published datasets. Caveat on the scatter source: Values are institution-level snapshots and can vary by year, campus and reporting definitions.
Data availability
Datasets: data/us-college-greek-life.json and data/us-top200-longitudinal.json in the story repository. A Jupyter
notebook reproducing the correlation, significance tests, and subgroup analysis can be opened
directly in Google Colab via the “Notebook” link above.
- Urban Institute Education Data API endpoint catalog
https://educationdata.urban.org/api/v1/api-endpoints/?mode=R - IPEDS directory endpoint
https://educationdata.urban.org/api/v1/college-university/ipeds/directory/2022/ - IPEDS admissions-enrollment endpoint
https://educationdata.urban.org/api/v1/college-university/ipeds/admissions-enrollment/2022/ - College Transitions Greek Life (live)
https://www.collegetransitions.com/dataverse/greek-life - Internet Archive CDX API
https://web.archive.org/cdx/search/cdx
Scatter data: Common Data Set 2023-2024 (C1 applicants, F1 fraternity/sorority rates) and institutional reports · Panel updated: 2026-03-03