Homework 3.1

Author
Affiliation

Alex Cardazzi

Old Dominion University

All materials can be found at alexcardazzi.github.io.

Completion Requirements: Complete the following questions in RStudio via the homework template. When you are ready, submit your rendered html to Canvas.

Grading Criteria: Full credit will be given to correct, well formatted, and detailed answers. Partial credit will be given if I can follow your work and/or see your thought process via code, comments, and text. Point totals are listed next to each question.

This homework explores the question, “do police reduce crime?” by replicating Di Tella & Schargrodsky (2004). This study addresses this question by exploiting a terrorist attack on a Jewish institution in Argentina, after which all similar institutions received additional police for security. Please read through the paper before completing this Homework.1

Question 1

Summarize the main research question in the paper, the application analyzed, and why DiD is appropriate for this research question. (1 Point)

Question 2

  1. Read in ditella_2004.csv and name it argentina. (0 Points)

  2. Tabulate the month_raw variable. Notice the “extra” observations, and use code to eliminate those rows. Re-tabulate the month_raw variable to confirm. (1 Point)

  3. Identify (in words) the variables indicating “treatment” and “post.” If one (or both) do (does) not exist, construct them. (0 Points)

  1. Tabulate the number of observations in each of the following groups: treatment group pre-treatment, treatment group post-treatment, control group pre-treatment, control group post-treatment. (1 Point)

  2. For each of the group-periods above, find the average number car_thefts. (1 Point)

  3. Produce a visual of the average number of car thefts by treatment group over time. Be sure to include points, lines, a vertical line representing the timing of the terrorist attack, a legend, and axis labels. (1 Point)

  4. Notice that there are 3 neighborhoods contained within the data. Which variables vary by neighborhood? Produce a summary statistics table to address this, and provide a discussion. (1 Point)

Question 3

  1. Discuss the general assumptions underlying the causal inference method used in the empirical analysis and how they apply in this research setting. (1 Point)
  1. Discuss threats to a causal interpretation of the results. How do the authors address these assumptions in the paper (if at all)? (1 Point)

Question 4

  1. Construct an indicator variable that measures whether a block is within two blocks of a Jewish institution (call is jewish_inst_within_2_blocks). (0 Points)

  2. Replicate columns (A), (B), and (C) of Table 3. (2 Points)

  3. How does the interpretation change as one moves from columns (A), (B), and (C)? (1 Point)

  1. Estimate and plot an event study version of column (A). Add a horizontal line equal to the treatment effect from (A). What does this figure tell you? (2 Points)
  1. Re-estimate column (C), except with a LPM. Interpret the coefficients quantitatively and the overall results qualitatively. (2 Points)

Question 5

  1. Test whether the treatment effect varies by neighborhood. (1 Point)
  1. Estimate a triple difference specification of column (A) with bank, public building, and gas station indicators like in Table 7. Interpret the coefficients. (2 Points)

Question 6

  1. Do you believe the results? What would you have done differently or have liked to see tested? (2 Points)

Footnotes

  1. The PDF will be made available on Canvas.↩︎