Replication Page for "Public Opinion and Senate Confirmation of Supreme Court Nominees", with Jeffrey Lax and Justin Phillips, 2010, Journal of Politics, 72(3):767-84
With the data and code described below, researchers can replicate the results from the paper. Note that all the files referenced below, including csv versions of all the datasets, can be found in this zip file.
- Data
- There are a number of datasets we used in the paper. A codebook describing the variables in each is available here.
- A list of the polls we used to compile each nominee "megapoll" is available here.
- Public opinion on individual nominees.
- As described in the paper, for each nominee we compiled a "megapoll":
- Census files for postratification.
- To create the estimates of state-level opinion, we use according to the population frequencies derived from the “5-Percent Public Use Microdata Sample” in the Census. We use the microdata samples from the 1980, 1990, and 200 census, depending on the nominee.
- Estimates of state-level opinion for nominees.
- Here are the state-level estimates for each nominee, as given in Table 1 of the paper
- Here's a "long" version of the data, which also includes the support of all respondents and the percent of respondents saying do not confirm, in addition to support among those with an opinion (which we used in the paper). This is the version of the data we used in our R script (linked to below).
- Here's a "short" version of the same data:
- Separate files of estimates for each nominee can be found in the zip file containing all files from this page.
- State-level data
- Here are the state-level data we used throughout both sets of analyses:
- Roll call data
- Finally, we merged our opinion estimates with roll call data for senators' votes on each nominee. The data for all nominees except Alito and Sotomayor were provided by Lee Epstein, Rene Lindstadt, Jeffrey Segal, and Chad Westerland, whom we thank. We collected data for the votes on Alito and Sotomayor.
- Scripts
- All the analyses were performed using R. There are separate scripts for a) creating the state-level estimates and b) merging the estimates with the roll call data and performing the statistical analyses used in the paper and c) creating each graph. Note you will need to install the "arm" library to run each script.
- Merged Data
- As noted above, we merged the datasets together in R. Here is a version of the merged dataset, which allows for replication our of roll call analyses in other statistical programs: