If you underwrite multifamily in a small market, the question that matters most is who, exactly, your renter is. Demographic averages — median income, age, household size — are useful at the metro level and almost useless below it. A 25,000-person CBSA with one large public university is not really a 25,000-person market. It's a 25,000-person market plus a 22,000-student demand block whose lease decisions are governed by a single institution's enrollment trajectory.
The College Town Index is built to surface that distinction. The headline finding: in 74 U.S. micropolitan CBSAs, a single institution accounts for more than 80% of total higher-ed enrollment in the metro. In another 41 metropolitan CBSAs, one institution accounts for more than 60%. Those 115 markets are concentrated demand stories — and they're not analyzed that way.
What single-institution dominance actually means
Take Pullman, Washington. Total CBSA population is roughly 50,000. Washington State University reports about 30,000 students. WSU is, in our data, 100% of the CBSA's higher-ed enrollment — there is no second institution above the IPEDS reporting threshold. Whatever happens to WSU's enrollment, retention, and out-of-state pricing flows directly through to demand for every rental unit within commuting distance.
The single-institution college town isn't a tertiary market with some students in it. It's a single-tenant industrial property dressed up as a metro.
Other examples in the same category:
- Oxford, MS — University of Mississippi, ~100% of CBSA enrollment
- Starkville, MS — Mississippi State, ~100%
- Boone, NC — Appalachian State, ~100%
- Stillwater, OK — Oklahoma State, 97%
- Athens, OH — Ohio University, 93%
- State College, PA — Penn State, 96%
Each of these has been written up as a "growth market" or "decline market" at various points. Few have been underwritten as what they functionally are: concentrated bets on one institution's competitive position.
The decline pattern that doesn't show up in metro data
Pullman, Washington is the cleanest example of why CBSA-level enrollment matters at the large-metro tier. Washington State's home metro shed enrollment at a -4.06% 5-year CAGR (Fall 2019 → Fall 2024) — the steepest decline among CBSAs with more than 25,000 students. An investor looking at "Pullman CBSA population" or "metro household income" sees nothing unusual; the numbers move slowly. The actual demand-side change — a meaningful contraction in the renter pool that drives 70%+ of off-campus leasing — is invisible without enrollment data.
State College, PA is the same story with an asterisk. Penn State's reported headcount in the State College CBSA fell from ~92K in Fall 2019 to ~65K in Fall 2024 in raw IPEDS numbers — about −29% over five years. Some of that is genuinely demographic (Pennsylvania's high-school cohort has shrunk). Some is methodological — the 2020 EFFY/EF reporting change reclassified online students. The point isn't the exact number. The point is that the multifamily underwriting case for State College in 2024 is materially different from 2019, and the only way to see it is to look at the institution.
Other declines worth pricing in (5-year CAGR, CBSAs with at least 25K students):
- Albany-Schenectady-Troy, NY: −3.48%
- Spokane-Spokane Valley, WA: −3.39%
- Portland-Vancouver-Hillsboro, OR-WA: −2.97%
- Akron, OH: −2.13%
- Athens, OH (Ohio University): −2.08%
The growth pattern that also doesn't show up in metro data
The fastest-growing CBSAs by 5-year enrollment CAGR (Fall 2019 → Fall 2024, markets with at least 15K students):
- Lafayette-W. Lafayette, IN (Purdue): +17.97% — a roughly 2.3x expansion of the dominant demand-side block in five years
- Salt Lake City, UT: +6.10%
- Knoxville, TN (U. of Tennessee): +5.59%
- Baton Rouge, LA (LSU): +4.81%
- Lexington-Fayette, KY (U. of Kentucky): +4.39%
- Madison, WI: +2.83%
Lafayette is the standout. It's the largest 5-year enrollment expansion at any major metro in the dataset, driven almost entirely by Purdue's deliberate scale-up. If you want to know why West Lafayette rents and absorption are doing what they're doing, no aggregate metro report is going to tell you. Enrollment data does.
National context: total U.S. higher-ed enrollment hit 20.36M in Fall 2024 — within 1.7% of the 2015 peak (20.71M) and up sharply from the 2021 trough (18.66M). The recovery is real, but it's concentrated in a handful of metros. Most CBSAs are flat or down.
The online-program distortion you have to remove
Any analysis of college-town enrollment that uses raw IPEDS totals will, without correction, give you misleading answers. Concord, NH is the cleanest example: with Southern New Hampshire University's 180K+ online students included, the Concord CBSA appears to have more students than residents — a 127% enrollment-to-population ratio. Strip out SNHU's distance-only students and Concord's actual physical college-town concentration is 5.15%. The same pattern applies to the host metros of Western Governors University, Grand Canyon, University of Phoenix, Liberty, and a handful of others.
v2 of the College Town Index ships two parallel views — one with all reporting institutions, and a "physical" view that strips out the 19 largest online-heavy institutions (SNHU, WGU, Grand Canyon, U. of Phoenix, Liberty, UMGC, Walden, Capella, Purdue Global, Penn State World Campus, and others). The physical view is what you should use to underwrite local rental demand. The all-institutions view is useful only for tracking enrollment-business trends.
The real college towns, ranked by physical enrollment-to-population ratio (where students are a meaningful share of the resident base):
- Vermillion, SD (U. of South Dakota) — 70%
- Rexburg, ID (BYU–Idaho) — 68%
- Vincennes, IN; Pullman, WA — ~55%
- Brenham, TX; Lafayette-W. Lafayette, IN — 47–50%
- Athens, OH; Boone, NC; Corvallis, OR — 39–45%
- Oxford, MS; Starkville, MS; Ithaca, NY — 36–38%
How investors should use this
Three concrete applications:
- Pre-deal screening. Before underwriting a small-market multifamily acquisition, check the dominant institution. If one school is more than 60% of CBSA enrollment, the deal is fundamentally a bet on that institution. Pull its 5-year enrollment trend from the workbook before you build the rent-growth case.
- Portfolio concentration. If your portfolio has multiple assets in single-institution CBSAs, you have correlated demand-side exposure that doesn't show up in standard geographic diversification metrics.
- Disposition timing. Enrollment is a leading indicator for off-campus rental demand. A 2-year decline at the dominant institution is a meaningful early signal — usually 18-30 months ahead of the rental market response.
How the data was built
The College Town Index v2.0 is built from three public sources:
- NCES IPEDS — Institutional Characteristics (HD), Fall Enrollment (EF), and 12-month Enrollment (EFFY) tables, 2015 through Fall 2024 (Provisional). Pulled directly from the NCES IPEDS data center.
- HUD USPS ZIP→CBSA crosswalk — Q4 2025 release. Each institution's reported ZIP is mapped to its dominant CBSA by total population ratio.
- U.S. Census Bureau — CBSA delineation files (March 2023) and ACS 5-year population estimates (2018, 2019, 2021, 2023 vintages).
The full methodology, including how the EFFY schema change in 2020 was handled and how the 19 online-heavy institutions are flagged, is in the README tab of the workbook. Every input file is linked in the sources tab. The whole thing is reproducible from public APIs.
The next release will layer ACS 1-year median rent data and HUD vacancy estimates onto the same CBSA framework, so the demand and supply pictures sit in one workbook. If there's a public dataset you'd find useful joined to this — RealPage student-housing inventory, NCREIF returns by metro, anything else — drop a note: hello@tjyresearch.com.