

PanelRegressionModels.do estimates the panel regression models reported in the paper. It also calculates the various
substantive effect calculations of the panel models and creates the figures reported in the paper.

IndividualVoteModels.do calculates the marginal effects for each vote and then graphs these quantities.
All files are written to the folder IndVotes that a user must create.


Vote level variables:
rc_vote highskillpcnt lnCapitaltemp2DigEstab33ES lnEmployManuf    bankPACpcntPAC_3_ corpPACpcntPAC_3_ labPACpcntPAC_3_ PrezVotePercRepub  PrezSuppSamePty_new    MktValAgProdII Wdummy MWdummy  southdummy unemployedpcnt avg_UnemployChange2yr lnmdincm   pcntForBorn pcntBlack cathad_adhperc jewad_adhperc mainad_adhperc evanad_adhperc FECproblems votetype 

upper level variables:
FECCodeNumeric VoteID session state_distL





note repeated observations within session due to multiple votes!

to check outliers on district level characteristics run:
duplicates drop FECCodeNumeric session, force

district level:
duplicates drop state_distL session, force


Write do file that calculates diferences session by session for the variables and the geographic or legislator unit, and then look at outliers.




VoteID gives the congressional session*10000 + Voteworld ID number. E.g., 1040343 is vote 343 from the 104th Congress.
Vote vargins (rc_vote is the vote variable) can be checked with VoteWorld.



