Scott Alexander explores the potential link between falling testosterone levels and declining crime rates, finding the hypothesis intriguing but ultimately insufficient to explain observed crime trends. Longer summary

Scott Alexander explores the hypothesis that falling testosterone levels might be related to the secular decline in crime. He examines studies showing a significant drop in testosterone levels in American men from 1987 to 2004, and research linking higher testosterone to violent crime. However, he identifies several problems with this hypothesis, including timing discrepancies between testosterone decline and crime reduction, contradictory evidence, and the fact that crime is dropping in women at the same rate as in men. While intrigued by the potential connection, Scott concludes that the hypothesis doesn't fully explain the observed crime trends. He suggests further research, including studies comparing testosterone levels in violent criminals to the general population and proposing an examination of digit ratios in historical skeletal remains to track testosterone levels over time. Shorter summary

Scott Alexander discusses the lack of scientific research on nerds, exploring potential explanations for their distinct traits, including the possible role of testosterone. Longer summary

Scott Alexander explores the lack of scientific study on nerds, despite their distinct characteristics. He lists nine common traits associated with nerds and questions why these seemingly unrelated traits often appear together. The post considers various explanations, including social factors and biological determinism, with a focus on the potential role of testosterone. Scott examines research linking low testosterone to several nerdy traits in men, but also acknowledges conflicting evidence. He concludes by noting the complexity of sex hormones and their effects, mentioning observations about transgender individuals, and questioning whether female nerds form a distinct cluster like male nerds do. Shorter summary