The Enrollment Cliff by State: Where the Freshman Supply Is Already Falling

The enrollment cliff gets talked about like a national event: one year, the incoming class shrinks everywhere at once. The data does not look like that. Pull the state of residence for every first-time freshman in IPEDS and the country splits in two. Between fall 2016 and fall 2024, Texas added 34,426 first-time freshmen per year, California added 30,084, and Florida added 14,277. Over the same eight years, New York lost 15,716, Pennsylvania lost 13,106, and Michigan lost 6,426. Same country, same window, opposite directions.

This analysis uses the IPEDS Fall Enrollment residence component (the EF C file), which records every first-time, degree/certificate-seeking undergraduate at every Title IV institution by their state of residence. Read as a state series, it measures each state's freshman supply: the pipeline of new students a state produces, wherever they end up enrolling. Fall 2024 is the newest complete collection, so 2016 to 2024 is the comparison window. Every number below came from a live query against that file; the SQL and returned rows are logged in the repository at docs/data-evidence/enrollment-cliff-by-state-where-freshman-supply-is-already-falling/queries.md.

Nationally, the cliff has not shown up yet. First-time enrollment fell from 3.24 million in 2012 to a COVID trough of 2.75 million in 2020, then climbed back to 3.11 million by fall 2024, within five percent of the all-time high. If you only looked at the national total, you would conclude the sector is fine.

It is not fine, because the total is an average of states moving in opposite directions. WICHE's Knocking at the College Door projections have shown for years that the number of high school graduates peaks nationally in the mid-2020s and declines roughly 13 percent into the 2040s, with the steepest projected losses in the West and the Northeast while Texas and much of the South keep growing. The IPEDS series shows what that divergence looks like in actual enrollment, before the projected peak even arrives: the states on the wrong side of the projection are already underwater.

The losses concentrate in the Northeast and the Midwest, which is where WICHE's projections have been pointing since the last edition. New York's decline alone is bigger than the entire annual freshman class of 40 states. Add Pennsylvania and Ohio, and three states account for 34,000 missing first-time freshmen per year, more than the entire enrollment of most state flagship systems.

One measurement caveat worth naming: first-time freshmen by residence is a proxy for the high school pipeline, not a pure count of it. The measure includes older first-time students and excludes recent graduates who never enroll. It moves with the pipeline, which is what matters for planning, but it is not the same number WICHE projects.

Alaska lost a quarter of its freshman supply in eight years, a decline with no parallel in any other state. The rest of the steepest-drop list is the map you would draw by hand if you started from demographics: the northern New England and Appalachian pipelines, plus the older industrial states. West Virginia and Mississippi have been losing 18-year-olds for years, and the enrollment data now carries the same shape.

Growth concentrates in the South and the Mountain West, with Nevada, Utah, and Idaho (up 27.4 percent, the fastest in the country) at the front. The California row deserves its own sentence: the state's freshman supply was still growing through fall 2024, even though WICHE projects California to face one of the steepest post-peak declines of any state. The West Coast's hardest years are ahead of it, not behind it. That is the difference between reading the rear-view mirror and reading the projection, and it is why state-level trajectory matters more than the national total.

A shrinking home state does not have to sink an institution, because institutions can recruit elsewhere. What determines the damage is how much of the first-time class comes from inside the state. The same EF C file supports that calculation directly: for the institutions located in a state, divide the first-time freshmen who are residents of that state by all their first-time freshmen. We ran it for fall 2024.

The two tables are two different business models. Texas and California institutions run on their own pipelines, at 92.5 and 90.3 percent in-state. The District of Columbia, New Hampshire, and Vermont run on everyone else's: 10.8, 20.8, and 29.1 percent in-state. High reliance is not a virtue or a fault by itself. It is a sensitivity setting. It tells you whose demographic curve your institution is actually riding.

This table is the answer to the question in the title. These are the states where the freshman pipeline is already shrinking and where institutions mostly draw from inside the state anyway. New York is the clearest case: the biggest absolute loss in the country, with institutions that still source three quarters of their first-time class in-state.

Note what is missing from the list: California, Texas, and Florida. Their institutions are the most in-state-reliant in the country, but their pipelines are still growing, so the sensitivity has not cost them anything yet. That is exactly the arrangement WICHE's projections threaten. If California's graduate count follows the projected post-peak decline, the 90.3 percent reliance goes from a strength to the state's biggest planning problem.

The low-reliance states look safe, and in one sense they are: New Hampshire's institutions could lose every New Hampshire high school graduate and still fill most of a class. But insulation is not the same as safety. An institution drawing 79 percent of its freshmen from out of state is exposed to the weighted average of everywhere it recruits. When Pennsylvania loses 13,000 freshmen a year and New York loses 15,700, the New England privates do not read about it in the news; they see it in the yield.

The honest framing is that every institution is exposed to some state's curve. The question is which ones, how much, and whether anyone has done the multiplication. Most enrollment conversations stop at "we recruit nationally," which is a description, not a number.

Two facts sit next to each other uncomfortably. First, national first-time enrollment in fall 2024 was within five percent of its all-time high, because growth in Texas, California, and Florida has been offsetting the Northeast and Midwest the whole time. Second, WICHE's projections have the national high school graduate count peaking in the mid-2020s and falling about 13 percent into the 2040s, with the steepest declines in the West and the Northeast.

Put those together and the risk is not a single cliff year. It is the moment when the states doing the offsetting start shrinking too. The 2025 and 2026 collections, when they publish, are the first ones that can show whether California's freshman supply has turned, and those two files deserve more attention than any national total printed next to them.

Every number in this post came from three queries against the IPEDS Fall Enrollment residence file: the national series by year, the 2016-versus-2024 comparison by state of residence, and the in-state reliance join with the institution directory. No file downloads, no Excel, no reconciliation of survey components. This is the standard workflow we describe in how to ask an IPEDS question: the question in plain English, the query against the source, and the answer with the table, year, and calculation attached.

If you want the same cut for your own sourcing footprint, the states you actually draw from, weighted by your own enrollment, that is a demo away.