Written with AI: Statistician Andrew Gelman has a long-standing reputation for using his blog, Statistical Modeling, Causal Inference, and Social Science, to dismantle “junk science” and “p-hacking” wherever he finds it. While he often focuses on psychology, he has directed his most pointed criticisms toward economists who treat noisy data as definitive proof for social engineering.
His critiques generally fall into three categories:
The “Nudgelords” and Behavioral Economists
Gelman has been a relentless critic of the “Nudgelord” faction—prominent scholars who believe they can solve social problems through small, behavioral interventions.
Brian Wansink: Perhaps his most famous target. Gelman spent years detailing the hundreds of errors and “p-hacking” in Wansink’s research on eating behavior (such as the “bottomless soup bowl” study). He famously mocked Wansink’s “deep dive” approach to data, where Wansink admitted to “tweeking” datasets until they showed statistically significant results.
Richard Thaler and Cass Sunstein: While he respects their intellectual contributions, he frequently skewers the “Nudgelord” attitude that assumes experts are “sophisticates” while the public is “naïve.” He argues that their policy recommendations often rest on fragile, non-replicable psychological experiments.
The Prestige Cartels and “Cargo Cult” Science
Gelman often targets high-status economists who use complex econometric models to mask simple, noisy observational data.
James Heckman: Gelman has criticized the Nobel laureate for what he calls “Type M” (magnitude) errors—claiming massive, life-changing effects from small sample sizes. He specifically targeted Heckman’s work on early childhood intervention, arguing the effects were likely “over-the-top” estimates derived from noisy data.
Steven Levitt: He has occasionally poked holes in the “Freakonomics” style of research, particularly when it relies on “cute” identification strategies (like the “himmicanes” vs. “hurricanes” naming study) that fail to hold up under rigorous statistical scrutiny.
Thomas Piketty: Gelman has pointed out the “garden of forking paths” in wealth inequality research. He warns that when researchers have a strong moral goal, they often (unconsciously) make data-processing choices that ensure the “correct” political conclusion.
The Methodology Gatekeepers
Gelman also targets the institutional structures that protect these alliances.
The American Economic Review (AER): He has criticized the flagship journal for its “gatekeeping” and for publishing studies with obviously flawed statistics simply because the authors are high-prestige figures.
The “Discovery” Hawkers: He frequently skewers “TED Talk-ing hawkers” who present preliminary, shaky findings as mature, settled science to influence government policy.
Gelman’s “iron law” of criticism is that he rarely frames these as moral failures; instead, he calls them “intellectual errors” born from a “statistical fantasy world.” He argues that these scholars have fooled themselves into believing in the “law of small numbers”—the idea that a small sample can reliably represent a massive population.
Aside from Andrew Gelman, a loose alliance of “data vigilantes,” forensic economists, and open-science advocates has emerged to challenge the prestige cartels. These critics often operate outside the traditional peer-review system, using blogs and social media to bypass the gatekeepers of the “big five” economics journals.
The Forensic Methodologists
This group treats economic papers like crime scenes. They do not care about the “moral narrative” of the research; they care if the numbers in the tables are mathematically possible.
Uri Simonsohn, Joe Simmons, and Leif Nelson: Known for their blog Data Colada, this trio invented “p-curve” analysis to detect p-hacking. They famously exposed fraudulent data in behavioral economics papers, including the work of Dan Ariely. They represent the “Behavioral Realism” that Berkeley claims to value, but they apply it to the scientists themselves.
Nick Brown and James Heathers: While they often work in psychology, their “GRIM” (Granularity-Related Inconsistency of Means) test is a terror to economists. They use simple arithmetic to show that reported means in many papers are mathematically impossible given the sample sizes.
The Replication Specialists
These critics focus on the “iron law” of the lab: if it is true, it should happen twice.
The Replication Network (TRN): This is a dedicated hub for economists who attempt to recreate the “landmark” studies of the field. They frequently find that when they use the original author’s data but change a single, arbitrary assumption, the “significant” results vanish.
Brian Nosek and the Center for Open Science: While broader than economics, Nosek’s work on the Reproducibility Project forced the field to reckon with the fact that its most “exciting” findings—often the ones used to brief the IMF or the California legislature—frequently fail to replicate.
The Public Intellectual Skeptics
These figures use their own high status to point out when the “emperors” of the field have no clothes.
Nassim Nicholas Taleb: He is perhaps the most aggressive critic of “junk economics.” He skewers the use of Gaussian (bell curve) models in macroeconomics and finance, arguing that they ignore “fat tails” and “Black Swan” events. He treats the entire field of econometrics as “Cargo Cult Science” that uses complex math to hide a lack of skin in the game.
Stephen Turner: As you know, Turner critiques the “tacit knowledge” and expertise claims of the social sciences. He points out that when economists claim “neutral expertise,” they are often just laundering political preferences through technical jargon.
Noah Smith: Though he is a former academic and a mainstream commentator, Smith often calls out “Macro BS” and the tendency of economists to cling to models that have been “empirically thrashed” by real-world events.
The “Inside Baseball” Insurgents
There are also economists within the system who have made a career out of calling out their peers.
Angus Deaton: The Nobel laureate has become increasingly vocal about the “dark side” of randomized controlled trials (RCTs). He argues that the “gold standard” of evidence often produces “useless” results because it ignores the social and political context—the very “alliances” Pinsof describes.
Edward Leamer: He wrote the classic paper Let’s Take the Con out of Econometrics. He argues that most empirical work is “whimsical” because researchers keep trying different models until they find the one that gives them the result they wanted to see.
These critics form a counter-alliance. They use “open science” as a purification ritual of their own, claiming that by being transparent, they are more “honest” than the prestige cartels at places like Berkeley or Harvard.
This list confirms the “iron law” of academic status: critique from outsiders like Gelman rarely results in a loss of position. Instead, the “prestige cartel” typically responds by absorbing the critique (the “pivot” to transparency or open science) or by ignoring it until the critic lacks the social capital to continue the fight.
The “iron law” of academic status functions as a self-correcting immune system for elite institutions. When an outsider like Andrew Gelman identifies a catastrophic flaw in the work of a Nobel laureate, the institution does not collapse. Instead, it engages in what David Pinsof would call strategic coordination to preserve its monopoly on prestige.
This process moves through three distinct phases:
Phase 1: The Ritual of Indifference
The first move is always silence. In a prestige cartel, attention is the most valuable currency. By ignoring a critic, the cartel denies them the social capital required to enter the “alliance.” If a Berkeley chair ignores a blog post, they are signaling that the post is “not even wrong”—it is socially nonexistent. This is why you rarely see direct rebuttals; a rebuttal is a transfer of status.
Phase 2: The Transparency Pivot
If the criticism becomes too loud to ignore—as seen during the replication crisis—the cartel performs a purification ritual. They do not admit the original work was “BS.” Instead, they become the champions of the “new standards” designed to fix the problem.
The Move: They launch “Open Science” initiatives or “Transparency” centers.
The Result: They now control the definition of “rigor.” By being the ones to “solve” the replication crisis, they ensure that they remain the gatekeepers. They effectively say, “Yes, science had a problem, and we are the only ones qualified to fix it.”
Phase 3: Absorption and Defanging
The cartel eventually absorbs the critic’s vocabulary but strips it of its sting.
Gelman’s “Type M” errors become a standard “limitation” section in a new Berkeley paper.
“P-hacking” becomes a technical hurdle that their new, more complex (and harder to critique) models have supposedly overcome.
This absorption serves a dual purpose: it makes the critic look like they are “harping on the past,” and it updates the “patchwork narrative” to look more scientifically robust without changing the underlying power structure.
Why Status is Immutable
The reason figures like James Heckman or the Labor Bloc at Berkeley do not lose their positions is that their value to the alliance is not their “truth” but their utility. They provide the “intellectual ammunition” for state governments and global NGOs. As long as Sacramento needs a justification for the minimum wage, and the IMF needs a narrative for inequality, the Berkeley alliance remains indispensable. The “iron law” ensures that as long as the patronage flows, the prestige remains intact.
