What College Actually Teaches — And What It Doesn't
The American university is being asked to do more than it was designed for and less than its tuition justifies. What four years of higher education actually delivers — and to whom.
The most honest answer to what a four-year bachelor's degree produces, in terms of measurable intellectual development, is: it depends. It depends on the institution, the major, the student's prior preparation, the quality of the instruction, and whether the student has enough time and economic stability to actually engage with the educational content rather than treating college primarily as a credential acquisition strategy.
Richard Arum and Josipa Roksa's Academically Adrift (2011), which followed a representative sample of college students through their undergraduate years using the Collegiate Learning Assessment, found that approximately 45 percent of students showed no statistically significant gains in critical thinking, complex reasoning, or written communication after two years. After four years, 36 percent showed no significant gains. The students who showed the largest gains were those who took demanding courses with substantial reading and writing requirements and who spent time studying alone rather than with peers.
This finding has been contested, replicated in modified form, and dismissed by various commentators over the past fifteen years. Its core claim — that the transmission of intellectual skills in higher education is highly variable and often inadequate relative to the price being charged — has not been successfully refuted.
The credential and the education
The bachelor's degree performs two distinct functions: it serves as a human capital investment (the education itself makes graduates more productive) and as a credential signal (it tells employers something about the pre-existing capabilities of the graduate, irrespective of what was learned in college).
The relative weight of these two functions in the wage premium associated with college is contested in labor economics. Bryan Caplan's The Case Against Education (2018) argued that the signaling function substantially dominates — that much of the wage premium reflects the sorting of more capable people into college and the credential's ability to communicate that sorting to employers, rather than the actual intellectual development that college produces.
If Caplan's argument is substantially correct — and the evidence is genuinely mixed — it has important implications. It means that the expansion of college attendance is a positional arms race: everyone's credential is worth less as more people have one, because the credential's sorting function degrades as the college-educated population expands. It means that the primary beneficiaries of the college wage premium are students who would have been successful without college and whose earnings premium reflects their pre-existing capabilities, not their education. And it means that the trillions of dollars invested in college education are substantially misallocated relative to their social return.
What the liberal arts actually accomplish
The political attacks on humanities and liberal arts education are often ideological and poorly reasoned. The empirical case for humanities education is also weaker than its defenders typically acknowledge, and making it honestly serves the cause better than dismissing the critics.
The genuine returns to humanities education are most plausible in specific conditions: when the education is intensive and demanding, when it is integrated with substantial writing and argumentation, and when it occurs in institutional contexts that prioritize the development of analytical capabilities rather than the transmission of specific content.
The evidence that humanities degrees produce worse labor market outcomes than STEM degrees is consistent and important. The evidence that this gap is entirely a product of selection (more analytically capable students choose STEM) rather than content (STEM teaches more marketable skills) is weaker. There is probably a genuine skills component to the earnings gap.
What defenders of humanities education are right about is that the skills which liberal arts education teaches at its best — reading closely, arguing carefully, understanding how knowledge is constructed and contested, recognizing the contingency of received ideas — are genuinely valuable for navigating a complex world. The problem is that these skills are produced inconsistently, are difficult to measure, and are not well-recognized by the labor market in the short run even when they contribute to long-run productivity.
Metaculus forecasts a 44 percent probability that undergraduate humanities enrollment will decline by more than 30 percent from its 2010 peak before 2030, following the trajectory of the past decade. The decline reflects both labor market rationality and a genuine cultural shift in how education is being valued.
Sofia Reyes is a contributing writer at The Auguro covering education, policy, and the politics of American institutions.