[that image]

Political scientist Alan Abramowitz writes:

In case you missed it, Paul Krugman just cited the rather thoroughly discredited claim that shark attacks have influenced election results.

Some background on that debunked study of shark attacks and elections is here with followup here.

There’s also a larger point that these sort of arguments (voters being swayed by shark attacks or college football games or hormones or whatever) have an anti-democratic tinge, in that the implication is that voters are capricious so don’t trust democracy. Of course, if it were true that voters were swayed by shark attacks etc., then we should be told. But given that there’s no good evidence for such claims, I think it can be damaging to repeat them. There are various schools of social science that really want to find such large effects from tiny inputs, and they are able to use statistical methods to apparently find these things, but that’s a problem with how statistics is used, and it’s an example of the replication crisis in science.

Abramowitz also points out that Krugman’s main argument, that McConnell’s goal is to sabotage Biden’s agenda, can be made without relying on this discredited claim.