Bryan Caplan on Learning across Disciplines (Live at Mason Econ)

Episode Summary

Executive Summary: Tyler Cowen and Bryan Caplan discuss Caplan’s method of building ideas from broad literature rather than single decisive papers, and then focus on his core thesis that education largely signals rather than builds human capital. They explore credential inflation, labor-market frictions, action bias, parenting, culture, and how psychology can better inform economics. The conversation also covers Caplan’s tastes, views on social networks, effective altruism, and future projects.

Main Topics: How Caplan does scholarship (Priority: 5/5): Caplan argues that no single social-science paper should overturn beliefs; real learning comes from synthesizing many studies, checking Google Scholar, reading references, and emailing living authors for corrections. Education as signaling and credential inflation (Priority: 5/5): The conversation centers on Caplan’s claim that much schooling mainly signals ability rather than raises productivity, with completion rates, curriculum content, and labor-market credential inflation as key evidence. Action bias and social desirability bias (Priority: 4/5): Caplan traces overinvestment in education and parenting to a human tendency to act for appearance’s sake, especially in government and institutions, even when the action may not help. Labor markets, hiring, and slow information flow (Priority: 4/5): Caplan argues labor markets are messier than economists admit: employers are poor at firing, pay adjustments are sluggish, and education mainly helps candidates get interviews rather than prove long-run productivity. Culture, parenting, technology, and social change (Priority: 4/5): He contends that broad cultural and technological changes—richer societies, more entertainment, weaker religion—shape outcomes more than individual parenting or schooling decisions. Interdisciplinarity, psychology, and effective altruism (Priority: 3/5): Caplan urges economists to read personality, educational, and industrial psychology, and praises effective altruism for making charity outcome-focused rather than emotion-driven. Personal tastes, Soviet history, and future books (Priority: 3/5): The interview closes with Caplan’s interests in Stalin, Tolstoy, opera, comedy, and his upcoming works on immigration and blameworthiness in poverty.

Key Arguments: Social-science knowledge is built by aggregating many weak-to-moderate pieces of evidence, not by finding one definitive paper that should cause a 'mind quake.' Caplan’s research process is iterative and interdisciplinary: broad search, subtopic drilling, reference-chasing, author outreach, and stopping rules to avoid endless search. Personality psychology suggests happiness is social; cheerfulness loads on extraversion, reinforcing the idea that relationships matter more than material possessions. Credential inflation is driven by action bias and social desirability bias: institutions act to look responsive, and education becomes a socially approved solution even when it fails to improve mobility. A large share of schooling is signaling: Caplan estimates about 80% of education from high school onward serves signaling purposes rather than human-capital formation. Non-completion matters a lot for returns to education; using only graduates overstates the payoff because many students do not finish and the graduation-year payoff is disproportionately large. Labor markets learn slowly and awkwardly; even when employers know someone is a poor fit, they often avoid firing and instead 'de-hire' or help them find another job. Education-related social networks are valuable only when the ties are close and occupationally relevant; most college networks are too weak and too diffuse to matter much for careers. Government subsidies usually add to, rather than replace, private spending, so nationalizing or subsidizing education can increase total spending instead of rationalizing it. Broader cultural shifts are better explained by rising wealth and greater entertainment options than by parents or schools acting differently at the individual level. Effective altruism is praised because it explicitly asks how to do the most good, rather than relying on warm-glow feelings or status signaling. Caplan sees major scholarly errors as type-three errors: answering the right question poorly, or doing technically correct work on questions that do not matter. He wants more economics to draw from personality, educational, and industrial psychology, which already study individual differences, learning, and job performance in ways economists often ignore.

Data Points: Response rate from cited authors: about 15% - Caplan says this is the share of living researchers who agree to read or comment on drafts when emailed. High school completion effect adjustment: around 30% lower - Caplan estimates non-completion reduces the apparent return to high school by roughly this amount. College completion effect adjustment: 40% to 50% lower - He says failing to complete college can cut the return substantially because graduation-year effects are large. On-time completion rate for full-time 4-year degrees: about 40% - Used to illustrate how much of college enrollment does not translate into completed degrees. Signaling share of education: about 80% - Caplan’s rough estimate for the extent to which education from high school onward functions as signaling. Number of papers on completion probability interacting with rate of return: about 7 - He says this is the tiny number of economics papers that directly combine completion risk with return calculations. Job acquisition through social ties: about half - Caplan cites the common claim that roughly half of people get jobs because they knew somebody. Time to learn an employee’s quality: one lunch - Caplan jokes that for his own judgment of colleagues, one lunch is enough because he values them mainly as lunch companions. Employer learning lag for IQ: at least 10 years - He cites employer-learning literature suggesting it can take a decade for employers to give decent credit even for IQ. Potential firing threshold: around the 25th-30th percentile - Caplan suggests employers often won’t fire someone unless performance is far below expectations. South Korea janitor applicants: four BAs and one AA hired for four janitor jobs - Example used to illustrate credential inflation and overqualification. Special gift/survey mention: listener survey with a chance to win a gift - Opening show promo; not analytically central to the interview.

Pivotal Quotes: "No single paper is that good." — Brian Kaplan: His core view of how knowledge in social science is actually accumulated. "Cheerfulness loads on extroversion." — Brian Kaplan: From personality psychology, explaining why happiness is so tied to social relations. "Suppose you're completely right, who cares?" — Brian Kaplan: His critique of narrow academic work that is technically correct but unimportant.

Implications: Caplan’s framework pushes listeners to distrust simple policy fixes, question credential inflation, and value broad evidence over single-study claims. For education, labor markets, and research, it implies more experimentation, less prestige-driven conformity, and more attention to actual outcomes.