These files include data and estimation code for:

Boehmke, Frederick J. N.d. “The Influence of Unobserved Factors on Position 
Timing and Content in the NAFTA vote.” Political Analysis (forthcoming).

Compiled on April 10, 2006 by Frederick J. Boehmke (frederick-boehmke@uiowa.edu).


INCLUDED:

  1 Stata data set:

   timing.do 			The Box-Steffensmeier, Arnold and Zorn (1997) 
				data set from the ICPSR's publication-related 
				archive (#1126). Downloaded January 27, 2004.

  3 Stata do files:

   boehmke2006pa-mkdata.do 	Creates the data set (boehmke2006pa-mkdata.dta) 
				used for estimation.   
   boehmke2006pa-analyze.do 	Runs the models in the paper.
   boehmke2006pa-graph.do 	Generates the graph used in the paper.

  1 README file:

   boehmke2006pa-README.txt	Describes the included data and files.


NOTES:

1. The included data set, timing.dta, is an unaltered download of the 
   Box-Steffensmeier, Arnold and Zorn (1997) data set from the ICPSR's 
   publication-related archive. The study number is 1126 and the URL is 
   http://webapp.icpsr.umich.edu/cocoon/ICPSR-STUDY/01126.xml. The included
   copy was downloaded on January 27, 2004 and has not been modified. Users
   interested in updates to the original data set, should there be any, 
   should check the ICPSR web site.

2. Variable manipulations are performed in boehmke2006pa-mkdata.do, which creates
   the data set boehmke2006pa.dta used for the analysis. 

3. The models are estimated by boehmke2006pa-analyze.do, which includes the SUDCD
   estimators for Weibull and log-normal likelihood functions. 

4. The graph (Figure 1 in the paper) is generated by boehmke2006pa-graph.do. 

5. The programs are written for Stata version 7 or above, though they may run
   in previous versions. Final models were run by the author in Stata 8.


VARIABLES:

These are the variables in the final Stata data set. See the paper for additional 
information on variable sources and construction. 

Here are the variables used in the analysis:

	vote numdiff commdiff contdiff mexbordr pscenter hhcenter partyid
  	timing corptpct labtpct mexbordr dleader rleader ncomact ideol pscenter 
	inter1 hhcenter inter2 rtcensor position

Here are descriptions for all the variables. The first set is from the original
data set; the second set is the variables created for my analysis.

district	District number
name		MC's last name only
state		Two-Letter Postal Code
partyid		1=Dem., 0=Rep.
corptpct	Corp. PAC $ / all $
labtpct		Labor PAC $ / all $
mexbordr	1=on Mexican border, 0=else
timing		Time To Position, in Days
dleader		1 if Dem. Leadership, 0 else
rleader		1 if Rep. leadership, 0 else
ncomact		1 if on com. w/NAFTA action
vote		1=vote for NAFTA, 0=vote agains
pscenter	Centered union membership
inter1		PSCENTER * IDEOL
ideol		Ideology (COC-based dummy)
hhcenter	Centered household income
inter2		HHCENTER * IDEOL
perotsq		PECENTER squared
pecenter	Centered Perot vote
mexstate	1 if state has Mex. border

numfor		Members of Congress declaring support for NAFTA by this day
commfor		Number of NAFTA Committee members declaring support for NAFTA by this day
commopp		Number of NAFTA Committee members declaring opposition to NAFTA by this day
numtot		Members of Congress declaring a position for NAFTA by this day
numopp		Members of Congress declaring opposition to NAFTA by this day
numdiff		Net favorable positions declared by this day
commdiff	Net favorable NAFTA committee member positions declared by this day         
rtcensr		Position declared on day of NAFTA vote
position	Position declared before day of NAFTA vote
contdiff	Net Labor PAC contributions