Reading assignments:
Please study the article before coming to class.
Answer the following questions:
a. Explain how the paper uses cross-sectional data to construct a district panel in order to estimate the effect of rainfall on human capital outcomes. Think of another type of causal effect you could analyze with a similar data structure.
b. Starting from a simple panel equation Y_ijt=b R_ijt +e_ijt, how do the authors address the problem of measurement errors on rainfall?
c. How do they address the problem of sample selection (in that only the healthier kids are observed as adult in the survey)?
d. Use results in table 4 to explain how they establish the mechanisms through which rainfall at birth has long-term welfare effects. Look at order of magnitude in the results.
Answer the following questions:
a. Derive proposition 1 in a standard model of behavior under risk, where second period output is function of input and (variable) rain, f(x, R). Then derive input choice when R is still variable but known. Use these to get the intuition of the paper on the response of investment and profit to forecasting.
b. Make sure that you can fully derive the comparative static results of one of proposition (2), (4), (6), or (7).
c. Establish the exact correspondence between the propositions and the empirical analyses of Tables 3 and 4. Are there discrepancies between the theoretical variables and those reported in the Tables (added variables, missing variables, variables that are substituted)
d. Discuss the results of Table 6. State the results derived from comparing columns 4 and 5.
October 20: Banerjee, Abhijit, and Esther Duflo. Do Firms Want to Borrow More? Testing Credit Constraints Using a Directed Lending Program. Review of Economic Studies, (2014) 81 (2): 572-607
Answer the following questions:
a. Define credit rationed and credit constrained, and explain the predictions of the demand model using Figures 1 and 2
b. What does Table 5 show in terms of the effective expansion/contraction of credit with the reforms?
c. Use Table 7 and the demand model predictions to support the argument that firms are credit constrained
d. What do the OLS and IV estimates in Table 9 columns 5-6 measure? Use the theory of the supply behavior to explain why the IV estimate is higher that the OLS estimate.