Crime Stats Are A Lie

Get my Math Box: https://www.melscience.com/kevin
Use code KEVIN for 50% off your first box

We’ve been collecting and analyzing social statistics for about 200 years now. In some ways, we’ve made a lot of progress -- but in others, we’re just as clueless now as when André-Michel Guerry first set forth his laws of tracking crime. Mixing math and morality introduces so many possible variables that our interpretations of the data can range from insightful to harmful. Comparisons of violent crime between Scotland, India, and Estonia show just how difficult it is to make sense of social statistics, especially if we try to judge them in relative terms.

A handful of additional problems plague our analysis: data and its inferences that are good today might not hold up in the future, and a study of Boston-area youth shows how surprising long-term outcomes can be. Perverse financial incentives, underreporting, and catching minor criminals instead of the major drivers of crime all make our understanding of antisocial behavior that much more tenuous. Throw in the media’s use of our social statistics and it feels like the more we know about crime, the less we understand.

*** ADDITIONAL READING ***

Analysis of A-M Guerry’s “Essay on the Moral Statistics of France,” by Michael Friendly: https://www.datavis.ca/gallery/guerry/

English Translation of Guerry’s Essay, by Whitt & Reinking: https://www.datavis.ca/gallery/guerry/Guerry1833-Whitt-translation-OCR.pdf

“The Mismeasure of Crime,” by Mosher, Miethe, and Phillips (2011): https://sk.sagepub.com/books/the-mismeasure-of-crime-2e

“The Cambridge-Somerville Study: A Pioneering Longitudinal Experimental Study of Delinquency Prevention,” McCord, J. (1992). https://psycnet.apa.org/record/1992-98237-009

“Crime Waves and Moral Panics,” by Morgan Godvin: https://daily.jstor.org/crime-waves-moral-panics/

*** LINKS ***

Vsauce2:
TikTok: https://www.tiktok.com/@vsaucetwo
Twitter: https://twitter.com/Vsauc