Replication: Does the Election of a Female Leader Clear the Way for More Women in Politics?

What you need:

• Download the paper “ElectionFemaleLeader.pdf”
• Download the Readme which includes information on variable names

Get Started (18 points)

Read the Introduction of the paper.

1. (2 points) What is the research question? ANSWER

2. (2 points) Why is this question important, according to the authors? ANSWER

3. (2 points) Where does the data come from? ANSWER

4. (2 points) What is their empirical strategy to answer the research question? ANSWER

5. (10 points) Read the rest of the paper and summarize the main findings here.

ANSWER

Data and Summary Analysis (17 points)

1. (2 points) Import the dataset main_dataset.Rdata.

#ANSWER

2. (5 points) Replicate the left hand part of Table 1 of the paper (all candidates). It does not need to look exactly the same, but it should be easy to read and convey the same information. I recommend subsetting the data (there is many ways to do this, but I use the dplr package) so that you keep only the variables listed in the table. Then using stargazer to create summary statistics. It is okay if you include more statistics than is in the original table. Do not worry about changing the variable names.

#ANSWER

3. (5 points) Replicate the right-hand part of Table 1 (female candidates only). It does not need to look exactly the same, but it should be easy to read and convey the same information. I recommend filtering the data so that only female candidates remain and then subsetting the data so that you keep only the variables listed in the table. Then using stargazer to create summary statistics. It is okay if you include more statistics than is in the original table. Do not worry about changing the variable names.

#ANSWER

4. (5 points) Summarize the tables in a narrative. This should be original work and not quoted from the author.

ANSWER

Empirical Results (20 points)

1. (5 points) In this part, you are going to create a scatterplot that looks similar to Figure 2. I recommend using rddtools, in particular the functions rdd_data() and plot(). This will not look identical to Figure 2, but will look similar.

#ANSWER

2. (5 points) Discuss what is shown in the plot from Question 1. How is it different than the one in the paper?

ANSWER

3. (5 points) In this section you will estimate the following RDD Model:

\[ gewinn\text{_}norm = \beta_0 + \beta_1 margin\text{_}1 + \beta_2 Female\text{_}Win + \varepsilon\]

where

\[ Female\text{_}Win=\begin{cases} 1 & \text{if } margin\text{_}1>0 \\ 0 & \text{if } margin\text{_}1<0 \end{cases} \]

I recommend that you use the rddtools package, in particular the rdd_reg_lm() function. Display your results using print() or stargazer(). Your output will not match any of the columns in Table 2 precisely, but will be close. Include as an option in the function rdd_reg_lm(...,bw=10.1).

#ANSWER

4. (5 points) Interpret the coefficient of interest (\(D\)) from the above regression within the context of this paper and Regression Discontinuity. Be precise in your language.

Assumptions (5 points)

Do you think that the continuity assumption is satisfied? (Recall, this means that in absence of the “treatment”, there is no reason why the \(Y\) variable would jump at the cutoff point.) Explain your reasoning.