At the moment I’m going to share a simulation I’ve been engaged on as an instance one thing I wrote about yesterday. The brief model of what I stated was nearly like a collection of easy deductions, one after one other, which began as soon as I noticed what inhabitants weighting does and doesn’t do. At the moment I’m going to assessment that perception that I had written about, however I’m additionally going to introduce code in order that you can also mess around with this outcome and see for your self that relying on the choice mechanism, parallel traits could not maintain for each the weighted and unweighted case, and as a consequence it might not be potential to establish each the weighted and unweighted ATT goal parameters utilizing the identical diff-in-diff design. It’s not in all probability correct to name it an impossibility outcome, as a result of it’s certainly potential, however the situations do place restrictions on the kind of choice into therapy.
However first, let’s seek the advice of the magic 8-ball to see if I might be paywalling this put up. As you recall, I randomize the paywall as I battle as a rule with paywalling extra typically — evidenced by round 6 months now of posting all my content material with out any paywall. So we go to my python script, we flip a coin thrice, and if it’s heads 2+ occasions, I paywall, and if it isn’t, I don’t. And at the moment it was, like yesterday, heads two out of three which implies paywall!
Thanks all for supporting the substack. It’s reader supported, I’m making an attempt to get again to extra common posts about econometrics, and might be randomly paywalling once more as earlier than. I admire all of your help (each subscribers and never), and hope that this put up is insightful and attention-grabbing. Although Alfred Adler would say that your emotions and ideas are your life process, and never mine, so possibly I simply provide this up as a present and you are able to do with it what you need.
