Ji Li and Lung-fei Lee, "Binary Choice under Social Interactions: an 
Empirical Study with and without Subjective Data of Expectation", 
Journal of Applied Econometrics, forthcoming.

README file for Section 7

RE_test.txt: 1022 observations, 24 variables;


Variables in RE_test.txt:

MRID: Identification Number of main respondent;

CaseID: 5 digit identification number of discussant;

Actual_Choice: 1 if voted/intended to vote for Dole; -1 if voted/intended to
  vote for Clinton

Sub_Expectation: 1 if expected by main respondent to vote for Dole; -1 if
  expected by main respondent to vote for Clinton

Married: 1 if currently married; 0 otherwise

Women: 1 if woman; 0 otherwise

Christian: 1 if Christian (Protestant or Catholic); 0 otherwise

Go_to_service: 1 if go to religious service once a week or almost so; 0
  otherwise

Employed: 1 if currently employed; 0 otherwise

LT_HS: 1 if highest level of education is less than or equal to 11th grade;
  0 otherwise

HS: 1 if highest level of education is high school; 0 otherwise

Some_college: 1 if highest level of education is some college; 0 otherwise

College: 1 if highest level of education is bachelor's degree; 0 otherwise

Age: Actual age

Squared_Age: Squared age divided by 100

Black: 1 if African American; 0 otherwise

Hispanic: 1 if of Hispanic origin; 0 otherwise

HH_Income: 1 if annual household income is <$15000; 2 if $15000-24999; 3 if
  $25000-34999; 4 if $35000-49999/$50000-74999; 5 if >=$75000

Father_R: 1 if father considered himself a Republican when respondent
  growing up; 0 otherwise

Father_D: 1 if father considered himself a Democrat when respondent growing
  up; 0 otherwise
  
Mother_R: 1 if mother considered herself a Republican when respondent
  growing up; 0 otherwise

Mother_D: 1 if mother considered herself a Democrat when respondent growing
  up; 0 otherwise

Better: 1 if own household is financially better off compared to last year;
  0 otherwise
  
Worse: 1 if own household is financially worse off compared to last year; 0
  otherwise
