17.8 Joint tests

  • event study combines event-study effect & model for counterfactual
  • significance tests test both!
  • partial answer: placebo tests
    • test something where we know there’s no effect
    • if we find something, the counterfactual model is wrong