Very detailed and helpful paper. Upshot:
- found a small (non-significant) *decrease* in full vax rates in Ohio post lottery
- found a small (non-significant) increase over all lottery states
Agree: "unlikely there are hugely positive or hugely negative effects"
- found a small (non-significant) *decrease* in full vax rates in Ohio post lottery
- found a small (non-significant) increase over all lottery states
Agree: "unlikely there are hugely positive or hugely negative effects"
https://twitter.com/DavidnLang/status/1412161173215084544
As a non-expert, I like the preregistration & use of synthetic control. Some remaining questions
- is theirs the right counterfactual? (see thread)
- is the proper outcome to measure full vax, as they did, or 1 dose (I'd actually think 1 dose is the more lottery-relevant outcome)
- is theirs the right counterfactual? (see thread)
- is the proper outcome to measure full vax, as they did, or 1 dose (I'd actually think 1 dose is the more lottery-relevant outcome)
Disappointingly, the quote tweets of the paper mostly seem to read "lol, vaccine lotteries didn't work"
People should read the whole thread & embrace the acknowledgement of substantial uncertainty that the authors recognize! (I know this runs contrary to twitter norms.)
People should read the whole thread & embrace the acknowledgement of substantial uncertainty that the authors recognize! (I know this runs contrary to twitter norms.)
My guess is we'll soon see many papers assessing vaccine lotteries, using different counterfactuals (which is good!).
Some will find + effects, some fail to reject a null hypothesis, some may even find - effects.
Similar to challenge of assessing if min wage affects employment
Some will find + effects, some fail to reject a null hypothesis, some may even find - effects.
Similar to challenge of assessing if min wage affects employment
Also, as with the min. wage debate, the way people--especially lay popularizers--interpret evidence will probably end up being driven heavily by their prior hunches, and/or their normative beliefs about vaccine lotteries' acceptability or lack thereof
Something else that would be really interesting: comparing the vaccine lotteries that at first glance appear to work well vs. not work:
https://twitter.com/DavidnLang/status/1412161187689615372?s=20Is the difference between states the design of the lottery? Demographics of the state population? Some other factor?
• • •
Missing some Tweet in this thread? You can try to
force a refresh
