Data Analysis Assignment 3

Data Assignment 3 is due October 20.

Overview

For this assignment, we will be using Rounds 5, 6, 7, and 8 of the Afrobarometer Surveys. The Afrobarometer contains information on political attitudes, values, and beliefs from representative population samples in 34 African countries. We will combine the Afrobarometer data with information about female heads-of-state in Africa to estimate the impact of female leaders on gender attitudes and other outcomes in a two-way fixed effects framework.

Getting Your Data

You can download the Afrobarometer here. There is a single merged file for each round (containing data from all the countries included in that round). You need to download the data for Rounds 5, 6, 7, and 8. The data is in SPSS format, but you can convert it to Stata format using Stata’s import spss command. Converting the data from SPSS to .dta format is time-consuming, so I recommend only doing this once. It makes sense to save a separate Stata data set for each round because the same variable names are sometimes used for different variables in different rounds.

Data on female heads of state in Africa is available here. The data is in Excel format.

Cleaning Your Data

Create a cleaning do file that is separate from the do file that converts the raw SPSS files into .dta format. Record each step involved in cleaning your data in the do file(s), so that you can replicate the entire process easily.

Keep only the variables you need from each round of the Afrobarometer data. The following variables are common to all the Afrobarometer data sets:

• COUNTRY is a numeric country code
• DATEINTR is the date of the Afrobarometer survey interview
• URBRUR is an indicator for urban vs. rural locations
• Q1 is the respondent’s age
• Q101 is the respondent’s gender

Labels sometimes change across survey rounds, so I recommend using Stata’s decode command to convert labeled numeric variables into string variables containing the text of the labels. You can then generate the numeric variables needed for analysis after combinging all of the data sets.

In addition to the independent variables listed above, you also need to keep data on key outcomes of interest. We will be looking at two measures of political attitudes. The first measures support for female leaders. In Rounds 5, 6, and 7, respondents were asked whether they agreed more with the statement

Men make better political leaders than women, and should be elected rather than women.

or the statement

Women should have the same chance of being elected to political office as men.

Responses to this question appear as variable Q22 in Round 5, as Q18 in Round 6, and as Q16 in Round 7. You will use this variable to estimate the impact of female heads of state on support for female political leaders.

The surveys also contain information about trust in the head of state (Q59A in Round 5, Q52A in Round 6, Q43A in Round 7, and Q41A in Round 8). From these, you can construct a measure of trust in the head of state (you can decide how to code it).

In addition to your analysis of political attitudes (i.e. support for female political leaders and trust in the head of state), you need to choose one additional set of outcomes to analyze. Some suggestions include:

• The presence of security forces as measured by the variables EA_SEC_A (police), EA_SEC_B (army), and EA_SEC_C (roadblocks). These variables can be analyzed individually or converted into an index.
• The provision of health and education facilities, as measured by the variables EA_FAC_B (schools) and EA_FAC_D (health clinics). These variables can be used to test the hypothesis (discussed here) that female politicians prioritize the provision of social services.
• Beliefs about the economic situation in the country and the prospects for the future (use lookfor economic to see a list of these variables).
• Other Afrobarometer outcomes of your choice.

To build your combined data set, keep only the variables you need from each round and then use Stata’s append command to put the round-level datasets together. Generate new, clean variables for use in your empirical analysis - setting any variables to missing as needed (for example, if someone answers “Don’t Know” or refuses to answer a question).

Female Heads of State

The variable DATEINTR is in Stata’s date format, which can be a challenge to work with (though there is plenty of online documentation). To create numeric variables for the month and the year in which a survey interview took place, use the code:

gen year = year(dofc(DATE))
gen month = month(dofc(DATE))

To generate a dummy variable for interviews that occured when a female head of state was in power, you might want to merge in the data from the spreadsheet on female leaders. Or, you might just want to code this variable by hand, since the number of female leaders is small. To do this, you will need to use Stata’s date format (again). Here is how I would code up a dummy variable for observations in Malawi during the period when Joyce Banda was the president of Malawi:

gen joycebanda = 0
label var joycebanda "Joyce Banda was president"
replace joycebanda = 1 if (country=="Malawi" & dofc(DATE)>td(7apr2012) & dofc(DATE)<td(31may2014))

You can extend this code to instead indicate whether any female leader was in power.

Analysis

Your analysis should include the following:

1. A summary statistics table
2. A regression table reporting the results of two-way fixed effects estimation documenting the impact of female heads of state on political attitudes (as described above)
3. A regression table reporting the results of two-way fixed effects estimation using any other set of outcome variables form the Afrobarometer
4. An event study figure estimating the impact of female heads of state by year and testing the common trends assumption (see discussion of event study figures and sample code hear)

Your summary statistics table should include data on all the variables used in your analysis. It should be nicely formatted (that goes without saying at this point).

Your regression tables should report estimates of the causal impact of having a female head of state on your outcome variables of interest (political attitudes in Table 2 and outcomes of your choice in Table 3). In addition to the independent variable of interest, you should include country and year fixed effects and cluster your standard errors at the country level. You should also control for respondent age, gender, and (optionally) educational attainment.

Presentation

You should present your results in a 3 to five page note (i.e. a mini-paper), as you have for your other assignments.

Your note should begin with an Introduction. Motivate your analysis in reference to the research question(s). Do not begin with a narrative opening sentence like “For this assignment, I decided to look at…” In your first sentence, describe your big question and explain why your analysis is of interest to an economist studying gender. Then use the remainder of your introductory paragraph to explain what you will do, your lever, and why you think your results are important.

A Data section should come next. Describe the data sets used in your analysis, and present your summary statistics. How do country-years with a female head-of-state differ from the rest of the sample?

The rest of your note should describe your Analysis. Present your results. Devote at least a paragraph to each table of results. Explain the regression specifications that you use. Discuss your coefficient estimates and their level of statistical significance, and summarize the main conclusions from each table.

Finally, you should offer a Conclusion paragraph at the end of the document. Make sure to cite any refenences in a Bibliography section.

You can upload your completed data assignment and replication files here.