Causal Inference -- 17/23 -- Difference-in-Differences Basics I

Поділитися
Вставка
  • Опубліковано 10 гру 2024

КОМЕНТАРІ • 7

  • @c.comploj3775
    @c.comploj3775 Рік тому +3

    This playlist is a gold mine for PhD students. Thanks again, Ben.

    • @ben_elsner
      @ben_elsner  Рік тому

      Happy to hear you found it useful. Spread the word!

  • @annawilson3824
    @annawilson3824 4 місяці тому

    21:12

  • @jayagupta9239
    @jayagupta9239 2 роки тому

    @ben elsner the coefficient of interaction term in regression equation is equal to ATT given that our parallel trend assumption is satisfied ?

    • @ben_elsner
      @ben_elsner  2 роки тому

      Correct

    • @jayagupta9239
      @jayagupta9239 2 роки тому

      @@ben_elsner thank you soo much . Can you also explain about the treatment intensity ( the fraction of teens as discussed in anguist book ). You don't need to make a video on that just explain in comments

    • @ben_elsner
      @ben_elsner  2 роки тому

      @@jayagupta9239 Not sure what you mean by treatment intensity. In most DiD applications the treatment is binary, such as a policy either switching on or off. There is newer research on DiD with a continuous treatment, but that's a lot harder. For a reference, see here: arxiv.org/abs/2107.02637#:~:text=This%20paper%20analyzes%20difference%2Din,to%20the%20binary%20treatment%20setup.