Where Truth Lies

Episode Summary

Executive Summary: The episode examines misinformation as something subtler than outright lies: selective interpretation, confirmation bias, narrative fallacies, and data mining can make smart people believe distorted claims. Through cases on CEO pay, diversity investing, breastfeeding, and the 10,000-hours rule, Alex Edmans argues that curiosity, skepticism, and "consider the opposite" can help us evaluate evidence more accurately.

Main Topics: Misinformation beyond obvious lies (Priority: 5/5): The show opens by arguing that misinformation often isn’t blatant fraud but subtle distortion shaped by psychology, identity, and selective evidence use. Corporate scandals and selective interpretation (Priority: 5/5): Edmans recounts the BHS and Sports Direct controversies and how a parliamentary committee embraced evidence that fit its pre-existing belief that CEOs were overpaid and harmful. Data mining and confirmation bias in research (Priority: 5/5): The conversation shows how researchers and advocates can search many variables until one supports a desired conclusion, as in the gender diversity investment example. Narrative fallacy and emotionally appealing stories (Priority: 4/5): The Belle Gibson case illustrates how people accept inspiring but false stories because they fit cherished beliefs about healing, grit, and natural living. Misreading correlation as causation (Priority: 5/5): Breastfeeding is used to show how strong correlations can be misleading when confounders like home environment, maternal education, and smoking are not controlled for. The myth and limits of the 10,000-hours rule (Priority: 4/5): Edmans revisits the popular performance myth, explaining that the original evidence was narrower and more qualified than the public version. Practical defenses against misinformation (Priority: 5/5): The episode closes with tools like curiosity, asking whether a conclusion is desired or extreme, and deliberately considering the opposite explanation.

Key Arguments: Misinformation is usually not just deliberate lying; it often emerges from people sincerely seeing what they want to see. Confirmation bias makes us scrutinize claims we dislike while accepting claims that align with our values. Researchers and institutions can unintentionally mislead by relying on outdated studies, selective citations, or post hoc interpretations. Data mining occurs when one starts with a desired conclusion and searches many measures until one appears to support it. The Belle Gibson story succeeded because audiences wanted an inspiring narrative about self-cure and rejection of conventional medicine. Breastfeeding studies often show correlations, but those links weaken or disappear when controlling for confounding factors. The popular 10,000-hours message was overstated; the original study focused on violinists, perceived success, and did not prove simple universal causation. Curiosity is a better antidote than intelligence alone because it encourages balanced evaluation across political and personal beliefs. The most effective defenses against misinformation involve slowing down, checking assumptions, and considering alternative explanations.

Data Points: AP tweet market impact: Dow plunged more than 100 points - After a hacked AP Twitter account falsely reported explosions at the White House and Obama being injured. BHS pension crisis: Thousands of former employees - BHS bankruptcy left workers without their full pensions. Superyacht value: $150 million - News reports of Philip Green awaiting delivery of the yacht Lionheart intensified public outrage during the BHS scandal. CEO-worker pay study versions: 2010 and 2013 - Edmans found the committee had cited the 2010 paper, while the updated 2013 version had the opposite result after peer review. CEO-worker pay recommendation year: 2016 inquiry; later became law - The parliamentary report relied on the earlier result and recommended pay-gap disclosure. Employee satisfaction study span: 28-year period - Edmans referenced research linking employee satisfaction and shareholder return over decades. Investment diversity study measures: 24 different measures - He tested multiple indicators of female friendliness against long-term financial returns. Women-friendly measures with positive result: 2 of 24 - Only two diversity measures showed positive relationships, and one was insignificant. Breastfeeding recommendation: Exclusively breastfeed for at least the first 6 months - This was the guidance given by the WHO and the National Childbirth Trust. 10,000-hours rule: 10,000 hours - The popular practice thesis discussed in relation to Malcolm Gladwell and Anders Ericsson.

Pivotal Quotes: "The greater the gap between CEO and worker pay, the better the company performance." — Alex Edmans: Describing the result he found in the 2013 version of the paper, which contradicted the version cited to Parliament. "The narrative fallacy is to see some events and believe that there is a link between the two." — Alex Edmans: Explaining why the Belle Gibson story felt compelling despite being false. "It means that we are able to show discernment. If there's a study that we don't like, we can come up with a whole host of reasons for why it's unreliable." — Alex Edmans: On why people accept favorable evidence uncritically while rejecting unfavorable evidence.

Implications: Listeners should treat appealing claims as hypotheses, not facts. For journalists, policymakers, and consumers alike, the lesson is to check versions, control for confounders, and use curiosity to resist stories that merely feel true.