Reading questions for empirical papers

[based on John Asker’s old IO syllabus]

What is the research question?
- Why is it worth asking?
- How does the research question relate to existing theoretical and empirical literature?
What’s the ideal data set?
- Given the question, what would the ideal observables and variation be?
What data do they actually have?
- Where does the data come from?
- What is the unit of observation?
- How is the data set in this paper different from that ideal data set?
What is the empirical strategy for answering the research question?
- If the paper is reduced form, can you map the equation to a model?
- If the paper is structural, what is the assumed data generating process?
  - Are the modeling decisions sensible?
  - Do they allow for interesting results with respect to the research question?
- What are the sources of exogenous variation?
  - Where is identification coming from?
What econometric techniques are being used in this paper?
- What is the central estimating equation (or equations)?
- How does it map to the theoretical framework?
- What is in the unobservable component?
  - What do the agents know that the econometrician doesn’t?
What are the main results of the paper?
- Do they move your prior? Are they plausible?
What do we learn from this paper?
- Do we care about this setting/ estimate per se, or are the results interesting because they relate to a broader phenomenon?
  - If the latter, do you believe they are externally valid?
- What are the economic implications of the results?
- Are there policy implications?
What questions does this paper leave unanswered?
- How might you answer them?