Thread by @cmkoeb, I got a burr under my science saddle and fact checked whether [...]

I got a burr under my science saddle and fact checked whether the “more police less violent crime” studies actually say what Yglesias said they say and WHEW BOY he really misrepresented those studies

First one he cites, he admits the violent crime that goes down is car theft NOT homicide or any other bodily harm

Next up U Penn police.... let’s go to the study. Authors say no effect on homicide, if anything an insignificant but NEGATIVE correlation between inc policing & homicide

Here’s the money figure. The sig results are ROBBERY, which sucks but it’s not assault, which is a much less significant correlation. Only one regression produces sig correlation w/sexual assault

But this is how Yglesias writes his summary? Uh buddy, it doesn’t match the data in the paper you linked to

Ok study 3 he cites, Mello. Uh oh. Robbery, larceny & auto theft were sig but only suggestive evidence for homicide according to the ABSTRACT

Ok study 3 he cites, Mello. Uh oh. Robbery, larceny & auto theft were sig but only suggestive evidence for homicide according to the ABSTRACT

Here’s how Matt represents the study. “Drop in violent crime” doesn’t bother mentioning that we’re talking about car theft not murder

Now we’re to this one & homicide in NYC is already super low so not surprisingly they lump all violent crime into one category & cannot look at homicide separately. In fact the word homicide appears once in the whole paper. This is not a paper about homicide rates

Final paper that he links to and the SSN basic facts he links to doesn’t even have the word “gun” in it. Honestly he showed me he’s a sloppy citer, summarizer & scholar so I’m not surprised this one doesn’t say what he said it said either.

Now their is some good scholarship on homicide rates but it’s often correlated with GUN ACCESS for example for every 1% inc in gun ownership in a state, there’s a correlated 1% higher femicide rate https://www.mic.com/articles/133656/women-are-more-likely-to-be-killed-in-states-with-higher-rates-of-gun-ownership

Women Are More Likely to Be Killed in States With Higher Rates of Gun Ownership

Women are more likely to be killed by someone they know in states with high gun ownership, a new study says. The paper, written by researchers at the Boston University School of Public Health, is the...

https://www.mic.com/articles/133656/women-are-more-likely-to-be-killed-in-states-with-higher-rates-of-gun-ownership

Now for everyone’s favorite, 90s Murder spikes are correlated with cheap handguns! https://www.thetrace.org/2018/08/guns-supply-shock-crack-epidemic-murder-rates/

Guns, Not the Crack Epidemic, Drove Murder Rates in the ’90s, According to These Economists

In recent years, a spike in murders also followed a boom in handgun production.

https://www.thetrace.org/2018/08/guns-supply-shock-crack-epidemic-murder-rates/

So if you want to drop the homicide rate, it really stands to argue you’d want to tax guns & ammunition, maybe put a high level of burden on handgun ownership & make sure that the profession w/DV rates 2-4x the general pop is disarmed. http://kutv.com/news/local/40-of-police-officer-families-experience-domestic-violence-study-says

Do 40% of police officer families experience domestic violence? It's complicated

A study claims that 40% of police officer families experience domestic violence, but where is that number coming from? A meme circulating on Facebook is making the claim based on that study, but is...

http://kutv.com/news/local/40-of-police-officer-families-experience-domestic-violence-study-says

So in summary Matt Yglesias is a very sloppy citer who doesn’t appear to understand the papers he’s citing because NONE of them have anything to do with the murder rate. NONE NOTHING NADA.

Special shout out to the best reporter on gun violence @AlexYablon who wrote up the 90s murder spike study for the Trace. Your reporting is missed!

Latest Threads Unrolled: