Skip to content
Chimera readability score 0.5092 out of 100, reading level.

Last year, I argued that the decline in college enrollment throughout the 2010s was a correction, not a crisis. This is because almost all of the decline occurred at low-quality colleges—those with poor graduation rates and loan repayment outcomes—which students would probably be better off avoiding. Between 2010 and 2023, the number of students seeking degrees at the worst tranche of colleges dropped by nearly half, while student numbers at the best schools actually increased.
The National Center for Education Statistics recently published enrollment figures by institution for fall 2024, so I updated my analysis with the new data, hoping to see a continuation of past trends. Enrollment at the lowest-quality schools remains well below its apogee in fall 2010. However, there are troubling signs of a comeback.
Enrollment at the worst fifth of institutions reached a nadir in fall 2022. Since then, student numbers at the lowest-quality schools have risen by over 120,000. The University of Phoenix, a for-profit megaschool in the bottom quintile of student outcomes, added nearly 19,000 degree-seeking undergraduates between 2022 and 2024. Enrollment grew at high-quality institutions as well, reflecting a broader rebound in college enrollment, but the rise at lower-quality institutions remains concerning.
To classify institutions by quality, I rely on student outcomes data from the early 2010s, when my period of analysis begins. It is possible some schools classified as low-quality have improved their outcomes and no longer deserve to be in the bottom quintile. To examine this possibility, I re-ran the analysis classifying schools according to their outcomes as reported in the most recent edition of the College Scorecard, and evaluated only the last two years of enrollment changes.
By this updated measure, the recent rise in college enrollment looks even more skewed towards lower-quality institutions. The worst quintile of colleges increased total enrollment by six percent between fall 2022 and fall 2024, as did the second-worst quintile. Enrollment at top-quality schools rose at around half that rate.
This is also concerning because bottom-quintile colleges remain a poor deal for students. According to the most recent outcomes data, students attending schools in the bottom quintile complete their programs at a rate of 26 percent and repay their student loans at a rate of 31 percent. Former students of bottom-quintile schools earn around $32,400 six years after enrollment—comparable to the earnings of workers with only a high school diploma.
Consumer choice is a powerful tool: students were fleeing low-quality colleges and reshaping the higher education system long before regulators caught on. But the rebound in enrollment at low-quality institutions is a sign that policymakers still have a role to play in holding federally-funded colleges accountable for poor outcomes. Otherwise, student numbers at bad schools may continue to climb—meaning a whole new generation could have to deal with the fallout of broken promises.

Facts Only

College enrollment declined throughout the 2010s, primarily at low-quality institutions with poor graduation and loan repayment rates.
Between 2010 and 2023, enrollment at the worst-performing schools dropped by nearly half.
Enrollment at top-quality schools increased during the same period.
Fall 2024 data shows enrollment at the lowest-quality schools has rebounded, rising by over 120,000 students since fall 2022.
The University of Phoenix, a for-profit school in the bottom quintile, added nearly 19,000 degree-seeking undergraduates between 2022 and 2024.
Enrollment also grew at high-quality institutions, but at a slower rate than at lower-quality schools.
Bottom-quintile colleges have a 26% program completion rate and a 31% student loan repayment rate.
Former students of bottom-quintile schools earn a median of $32,400 six years after enrollment.
The analysis initially used student outcomes data from the early 2010s but was updated with the most recent College Scorecard data.
The updated analysis shows the worst and second-worst quintiles of colleges increased enrollment by six percent between 2022 and 2024.
Top-quality schools saw enrollment growth at roughly half the rate of lower-quality institutions.

Executive Summary

College enrollment trends have shifted significantly over the past decade, with a notable decline in attendance at low-quality institutions—those with poor graduation rates, low loan repayment rates, and weak earnings outcomes for graduates. Between 2010 and 2023, enrollment at the worst-performing schools dropped by nearly half, while top-tier institutions saw increased student numbers. However, recent data from fall 2024 shows a concerning rebound in enrollment at low-quality colleges, with over 120,000 additional students attending the bottom quintile of schools since 2022. The University of Phoenix, a for-profit institution in this category, added nearly 19,000 undergraduates in just two years. While enrollment also grew at high-quality schools, the rate of increase was lower. Updated analysis using recent College Scorecard data confirms that the rebound is disproportionately concentrated in lower-quality institutions, where students face completion rates of just 26% and median earnings of $32,400 six years post-enrollment—comparable to high school graduates. This trend raises questions about the effectiveness of market-driven accountability in higher education and suggests a potential need for stronger regulatory oversight to prevent students from investing in programs with poor outcomes.

Full Take

The strongest version of this narrative highlights a market-driven correction in higher education, where students initially avoided low-quality institutions due to poor outcomes, only for enrollment to rebound in recent years. This suggests that while consumer choice can drive accountability, it may not be sufficient without regulatory oversight. The analysis deserves credit for using updated data to test its initial assumptions and for clearly defining the metrics used to classify institutional quality.
However, the framing risks implying that all low-quality institutions are uniformly predatory, which may overlook nuanced cases where schools serve non-traditional students or face systemic barriers to improvement. The emphasis on earnings as a primary metric of value also assumes a narrowly economic view of education, potentially undervaluing other benefits like social mobility or personal growth.
Root cause: The narrative reflects a tension between market-based accountability and the need for regulatory guardrails in higher education. It assumes that students act as rational consumers, yet the rebound in enrollment at low-quality schools suggests that information asymmetry, financial constraints, or aggressive recruitment tactics may distort decision-making.
Implications: If this trend continues, students—particularly those from marginalized backgrounds—may disproportionately bear the costs of poor outcomes, reinforcing cycles of debt and limited opportunity. The second-order consequence could be a further erosion of public trust in higher education as a whole, even as top institutions thrive.
Bridge questions: What structural factors might explain why students are returning to low-quality institutions despite poor outcomes? Could alternative metrics (e.g., social mobility, non-monetary benefits) provide a more complete picture of institutional value? How might policymakers balance accountability with access for non-traditional students?
Counterstrike scan: A coordinated influence campaign might exploit this narrative to push for deregulation (framing oversight as paternalistic) or, conversely, to justify aggressive crackdowns on for-profit schools. The actual content does not match this pattern, as it presents a measured argument for accountability rather than ideological extremism.
Patterns detected: none

Sentinel — Human

Confidence

The article shows strong signs of human authorship, including personal voice, erratic sentence structure, and detailed methodological transparency.

Signals Detected
low severity: Sentence length variance is erratic, with a mix of short and long sentences, inconsistent with AI-generated uniformity.
low severity: Text exhibits passionate emphasis on specific points (e.g., 'troubling signs of a comeback'), which is atypical of AI-generated content.
low severity: No evidence of template-driven argumentation or verbatim repetition of talking points across sources.
low severity: Claims are supported by specific data sources (e.g., National Center for Education Statistics, College Scorecard) with clear methodology.
Human Indicators
Idiosyncratic phrasing (e.g., 'apogee in fall 2010', 'broken promises')
Personal voice and analytical perspective (e.g., 'I argued', 'I rely on')
Nuanced discussion of potential limitations (e.g., reclassification of schools by updated metrics)
Are Low — Arc Codex