Record Low Crime Rates Are Real, Not Just Reporting Bias Or Improved Medical Care

Establishes that US crime (murder possibly at a 250-year low; violent and property crime near 50-year lows) is genuinely low, then defends that against two deflationary objections. Reporting bias is ruled out by the NCVS victimization survey (which bypasses police and matches their trend), by car theft (always reported for insurance) and murder (always reported) declining at similar rates, and by reporting rates having risen over time. The medical-care artifact (better trauma care converting murders into assaults) is challenged by Eckberg (2014, NCVS shows stable lethality) and by evidence that gun injuries have grown more severe (more wounds, more critical regions, 'increased magazine size'), roughly offsetting medical gains — so the suspicious cancellation favors taking the data at face value. Surveys candidate explanations (lead, mass incarceration, abortion, cell-phones-replacing-street-dealers, cameras/DNA, psychiatry, the BLM backlash-to-the-backlash) under a 'we're a safetyist culture' frame, and teases why intuitions clash with the stats.

Quality 73 · Strong. Strong band, upper. Rigorous, well-sourced myth-busting that settles a genuinely contested empirical question and carefully dismantles the standard objections. Upper-Strong: high-value data analysis rather than a novel framework.

Claude’s paradigm shift 35 · Slight. Slight. Empirical synthesis of crime statistics and the lethality literature; no new idea.

Real-world impact 3 · Moderate. Rigorous, well-sourced myth-busting that settles a genuinely contested empirical question — US crime really is at multi-decade lows — by dismantling the reporting-bias and improved-medical-care objections (NCVS, car theft, Eckberg). Real topical relevance to a material public question, but high-value data analysis informing understanding — modest RWI.