a crossword for economists

12345678
Across
  1. 3. [BLANK] trends
  2. 4. unobserved variable causing both treatment and outcome
  3. 6. the variable whose causal effect you want to identify
  4. 7. ? A trio tint, panel dropout
  5. 8. [BLANK] bias: non-random sorting into treatment
Down
  1. 1. fake treatment test
  2. 2. the F-stat on your excluded instrument tests this
  3. 4. ? An anagram of mic loper - follows the instrument.
  4. 5. ideal design