Noah Smith on Whether Economics is a Science

Episode Summary

Executive Summary: Russ Roberts and Noah Smith debate ideology, pragmatism, and the limits of economics as a science. They agree incentives matter, but differ on how much reliable causal knowledge economics can produce, especially in macroeconomics. Both see strong reasons for humility: many policy claims rest on weak identification, yet some natural experiments and cumulative evidence do shift views.

Main Topics: Ideology, priors, and pragmatism in economics (Priority: 5/5): They discuss how economists form beliefs, whether ideology should be explicit, and Smith’s self-description as a pragmatist/Humean rather than a consistent ideologue. Big government, living standards, and caution about radical reform (Priority: 4/5): Smith argues rich countries with large governments still perform well, so reform should be incremental; Roberts worries about long-run drift, weak opportunities for the less skilled, and policy mistakes like minimum wage laws. What counts as science in economics (Priority: 5/5): The conversation centers on whether economics can generate reliable causal knowledge. Smith defends natural experiments and the credibility revolution, while Roberts stresses fragility, non-replicability, and limited practical certainty. Macro versus micro evidence (Priority: 5/5): Smith argues macroeconomics has fewer clean natural experiments and is therefore more dependent on priors; Roberts pushes back with historical examples like WWII demobilization and QE, questioning whether macro evidence can really settle debates. Minimum wage evidence and identification (Priority: 4/5): They use minimum wage research to illustrate the strengths and weaknesses of empirical economics. Smith says typical minimum-wage hikes usually show little immediate employment effect; Roberts stresses that literature disputes remain and identification is never perfect. When studies change minds (Priority: 4/5): The hosts debate whether any study can decisively shift opinion. They cite examples of economists changing views, including lottery-based school studies, Posner on Keynesian stimulus, Williamson on QE, and Lucas revising his macro beliefs. Humility and uncertainty in policy debates (Priority: 3/5): The episode closes with agreement that economists should acknowledge uncertainty privately and publicly, even when their incentives push them toward overconfidence and ideological certainty.

Key Arguments: Ideology is widespread in economics and especially in media commentary because strongly motivated people are more likely to seek attention and be selected for it. Roberts’ methodological bias is that incentives and emergent order are powerful, but they must be used cautiously because bad incentives can produce harmful outcomes. Smith’s position is pragmatic: policy should be judged case by case, with attention to what has worked in comparable countries rather than grand theory alone. Large, prosperous countries with substantial governments suggest that big government is not obviously incompatible with high living standards, so reform should be incremental. Roberts argues that even if today’s outcomes are acceptable, policy can still be dangerous if it slowly accumulates distortions over time or blocks opportunity for lower-skill workers. Smith supports wage subsidies/EITC-type policies over the minimum wage as a better short-run way to help low earners. Natural experiments improve economics relative to simple correlations because they exploit quasi-random variation, but they are still weaker than controlled lab experiments and must be evaluated for exogeneity. Macroeconomics is especially hard to identify causally because shocks are economy-wide and linkages make clean comparisons rare; thus priors matter more. Roberts contends that major macro episodes, like the end of WWII or QE, are interpreted in radically different ways, showing that evidence often fails to resolve debates. Smith agrees that some studies are more convincing than others and cites lottery-based education studies as examples where evidence can genuinely alter beliefs. Both speakers acknowledge that economists sometimes change their minds in response to evidence, but only gradually and rarely in dramatic, universal ways. Smith emphasizes that science itself is not as decisive as laypeople imagine; even physics can contain deep disagreements about fundamental laws.

Data Points: Podcast date: December 1, 2015 - Episode introduction Roberts age: 61 - Roberts mentions his age while discussing how his views changed over time Government share of GDP: 40%+ - Roberts argues the U.S. can have strong outcomes even with government at roughly this share of GDP Historical timeframe for Roberts’ training: Late 1970s - He contrasts his econometrics training with newer methods QE balance sheet expansion: $800 billion to $4 trillion - Roberts cites the Federal Reserve’s balance sheet growth after the Great Recession Years of Great Recession-era policy: 2009 - Roberts references the $800 billion stimulus package in the aftermath of the Great Recession World War II end: 1945 - Used as a natural experiment for the effect of demobilization and reduced military spending Psychology replication concern: 60% - Roberts references Brian Nosek’s discussion that about 60% of psychology experiments do not hold up Public school/lottery study: Chicago education lotteries - Smith cites lottery-based evidence on school choice and outcomes

Pivotal Quotes: "I think of myself as a pragmatist." — Noah Smith: Smith defines his stance on ideology and policymaking as case-by-case rather than doctrinaire "Economics isn't much of a science in the way that people in everyday English think about it." — Russ Roberts: Roberts frames his critique of economics’ ability to produce reliable policy guidance "if deep down, when pressed, we'll admit that we really are less certain than we think, that's a good thing." — Noah Smith: Closing reflection on humility, evidence, and the limits of certainty

Implications: Listeners are left with a case for intellectual humility: evidence can move economists, but rarely decisively. Policy debates should favor incrementalism, careful identification, and skepticism toward grand claims, especially in macroeconomics.