Data Talks

It's #NBERday !

First up:

Q: Did COVID lockdowns increase domestic violence?

A: A lot of people say yes. But maybe.... no? #NBERdayhttps://t.co/wSyc25fsQ0

— Lyman Stone 石來民 (@lymanstoneky) November 9, 2020

Click for the thread. 

This reminds me of so many media reports within the first few weeks of Trump's 2016 election, claiming that hate crimes were up dramatically.  All of it was premature nonsense.  Hate crimes are a vanishingly  small percentage of crimes, committed in every possible configuration (i.e. not all hate crimes are committed by a single group), and the volume of hate crimes drifts up and down by season and media focus.  

Most critically - journalists reporting on increased hate crime were at best projecting a fantasy of theirs.  Hate crimes are spottily reported locally and the national hate crimes data is annual.  There was no way for journalists to see any trend in hate crimes before and after the election.  It was pure speculation combined with confirmation bias on their part.  

Covid and Domestic Violence?  We still don't understand much about Covid, much less lock-downs or other intrusive policies.  A lot of research to be done but there is little which can be claimed with much degree of confidence.