The Decade-High That Proves Nothing

Your persistence rate is a lagging indicator.

Persistence just hit a decade high. That is not the good news your dashboard thinks it is.

Executive summary

National persistence reached 77.1 percent for the fall 2024 cohort, the highest in ten years, with Black, Hispanic, Native American, and part-time starters all posting decade highs.

The number rose in the same month Fitch warned that private college margins are deteriorating and Johns Hopkins cut 110 jobs. Retention up, institutions strained. That divergence is the signal, not the headline.

Persistence measures a student behavior, re-enrollment one year later. It does not measure whether the institution is viable or whether the program the student returned to earns its place.

Aggregate persistence is an average. It hides the program-level spread where every consequential decision actually lives.

The risk this summer is not the data. It is the inference. Leaders will read a decade high as proof the current model works, and relax the pressure on portfolio decisions that the number was never built to inform.

The framing

Here is what most leaders will do with this report. They will treat it as vindication.

The narrative for three years has been crisis. Enrollment cliff, retention emergency, a sector in structural decline. Then the National Student Clearinghouse Research Center publishes a decade high, and the temptation is immediate. The interventions are working. The advising redesign paid off. The model holds.

That reading is comfortable. It is also wrong.

Persistence is a lagging indicator. It tells you a student came back. It tells you nothing about why, nothing about what they came back to, and nothing about whether the institution can afford to keep the doors open around them. A rate can climb for reasons that have nothing to do with anything you did, and a climbing rate can sit directly on top of a balance sheet that is failing.

The institutions that closed this year had persistence rates too. Some of them were rising.

So the question the report should provoke is not "are we retaining students." You are, slightly better than last year, along with everyone else. The question is sharper, and the answer is not in this dataset. If persistence is at a ten-year high and institutions are still cutting staff and merging programs, then persistence was never the metric that decided survival. What is it actually a measure of, and what decision does it license?