In the original post, Scott said that although murder was down in El Salvador, the statistics didn't really show a decline in theft.

This claim is about what Scott's earlier post on El Salvador's crime statistics said. It is judged by that post's text. True: the original post says that murder fell but the statistics did not clearly show a fall in theft (or property crime). Mostly true: the post makes that point but words it a bit differently or hedges it, for example by saying the theft data were unclear or unavailable. Mixed: the post mentions theft only vaguely, or partly suggests theft fell. Mostly false: the post mostly implied crime in general, theft included, had fallen, with at most a passing caveat. False: the post said the statistics showed theft falling, or never mentioned theft.

Set before the research began.

1. E1 Prison And Crime: Much More Than You Wanted To Know Astral Codex Ten (Scott Alexander) · 27 November 2024 Own record

   “I can’t find good statistics on other crimes like theft, but the crappy statistics I find say it hasn’t budged (1, 2).”

2. E2 Prison And Crime: Much More Than You Wanted To Know Astral Codex Ten (Scott Alexander) · 27 November 2024 Own record

   “Either my statistics are bad, or the gangs that the government cracked down on weren’t in the theft business.”

3. E3 Prison And Crime: Much More Than You Wanted To Know Astral Codex Ten (Scott Alexander) · 27 November 2024 Own record

   “As incarceration rate quadrupled, homicide went down by a factor of twenty:”

4. E4 Crime Impacts of El Salvador's Crackdown Policy (Working Paper No. 20) School of Government and Public Transformation, Tecnológico de Monterrey (Santos, Silverio-Murillo, Balmori-de-la-Miyar, Rodríguez) · 16 June 2026

   “We detect no measurable effects on assault, larceny, or motor vehicle theft.”

The claim is about what the original post said, so that post's text is the record being judged. In its El Salvador section, the post says homicide fell by a factor of twenty as incarceration quadrupled (E3). It also says it could not find good statistics on theft, but the poor statistics it did find showed theft "hasn't budged" (E1), and it offers explanations for why theft might not have fallen (E2). That matches the claim that murder was down but the statistics didn't really show a fall in theft.