United States Chamber of Commerce votes

American Federation of Labor-Congress of Industrial Organizations votes

Data used in ``A Panel Probit Analysis of Campaign Contributions and Roll Call Votes'':

The data to replicate the analysis is separated by vote type and congressional session. Each line in the file contains the data for a member of the House for all votes in a given session. The records contain the following variables (in order):

• vote1 vote2 ... voteT (=1 if member voted in favor of position; =0 otherwise; T represents the total number of votes in a session)
  
  
• labor1 labor2 ... laborT (ln of labor PAC contributions during month of vote)
  
  
• corp1 corp2 ... corpT (ln of corporate PAC contributions during month of vote)
  
  
• unemp1 unemp2 ... unempT (unemployment rate during month of vote)
  
  
• psh1 psh2 ... pshT (measure of presidential support in district during month of the vote; not used in analysis, nor would I recommend using it!)
  
  
• party (party affiliation of member)
  
  
• committee (=1 if member is on committee with relevant jurisdiction)
  
  
• ICPSR code for member
  
  
• Congressional district identifier
  
  
• ok1 ok2 ... okT (dummy variables indicating whether the observation is valid in that particular month or not)

**In looking over these data files many years after conducting the analysis (and after losing the original GAUSS code in a hard disk crash), it looks like all of the votes for a given Congress (i.e., both the 1st and 2nd sessions) are included in each of these data sets. Thus, we would need to subset the relevant columns from each data file to replicate the by-session analysis.

102nd Congress:
AFL-CIO votes: 1st session and 2nd session .

USCC votes: 1st session and 2nd session .

103rd Congress:
AFL-CIO votes: 1st session and 2nd session .

USCC votes: 1st session and 2nd session .

104th Congress:
AFL-CIO votes: 1st session and 2nd session .

USCC votes: 1st session and 2nd session .

Several programs are necessary to replicate the analysis, and each will need some tweaking to get them to work correctly depending on which part of the analysis is being replicated. Please contact me via email to obtain the programs to replicate specific parts of the analysis.