Sukjin Han and Sungwon Lee, "Estimation in a Generalization of
Bivariate Probit Models with Dummy Endogenous Regressors", Journal of
Applied Econometrics, Vol. 34, No. 6, 2019, pp. 994-1015.

This readme file describes files used for the simulations and the
empirical application. All data files and associated Matlab files for
the empirical application are zipped in hl-application.zip. All code
files for the simulations are zipped in hl-simulations.zip.

****************************************************************
*******Files in folder "Simulation\Main Simulation Codes"*******
****************************************************************

Main_Correct_Specification.m
- Estimates both semiparametric and parametric models with true
  marginal distributions being standard normal

Main_Missspecification_Mixture.m
- Estimates both semiparametric and parametric models with true
  marginal distributions being a mixture of normal

Main_Misspecification_T3.m 
- Estimates both semiparametric and parametric models with true
  marginal distributions being t(3) *Standard normal distribution
  function for function G in semiparametric models 

Main_Misspecification_T3_GT.m
- Estimates both semiparametric and parametric models with true
  marginal distributions being t(3) *The distribution function of t(3)
  for function G in semiparametric models


*******************************************************
*******Files in folder "Simulation\Bootstrap CI"*******
*******************************************************

MainCode_Bootstrap_CIs.m
- Estimates both bootstrap CIs using a normal approximation and
  bootstrap percentile CIs in a semiparametric model

MainCode_Coverage_Calculation.m
- Computes the coverage probabilities. *This file should be run after
  obtaining CIs from "MainCode_Bootstrap_Cis.m"


*******************************************************************************
****Files for the Empirical Application (in folder "Empirical Application")****
*******************************************************************************

MainCode_Semi_Empirical.m
- Estimates a semiparametric model and provides bootstrap standard errors

MainCode_Par_Empirical.m
- Estimates a parametric model and provides bootstrap standard errors

011719_Dataset.xlsx / 011719_Dataset.csv
- Contains the data; Continuous variables are standardized


******************************************************************************
****Details on the Dataset file (011719_Dataset.xlsx / 011719_Dataset.csv)****
******************************************************************************

column 1: family size 
column 2: =1 if living in MSA
column 3: age
column 4: =1 if male 
column 5: years of education

column 6: hourly wage (not used)
column 7: industry group (not used)

column 8: number of employees (instrument 1)
column 9: =1 if a firm has more than one location (instrument 2)

column 10: =1 if having a private insurance (treatment variable; D)
column 11: Number of visits (not used)

column 12: =1 if vistied a doctor in January (outcome variable; Y)

column 13: =1 if living in Northeast
column 14: =1 if living in Midwest
column 15: =1 if living in South
column 16: =1 if living in West

column 17: =1 if white 
column 18: =1 if black 
column 19: =1 if Minority
column 20: =1 if Asian

column 21: =1 if ever married 

column 22: =1 if physical health is below good
column 23: =1 if mental health is below good

column 24: monthly income

column 25: standardized family size 
column 26: standardized age 
column 27: standardized income 

column 28: standardized number of employees (instrument 1)

column 29: standardized education level (standardized years of education)
column 30: standardized hourly wage (not used)

column 31: standardized sick day 1 (sick32)
column 32: standardized sick day 2 (sick34)

column 33: =1 if physical health is excellent
column 34: =1 if physical health is very good
column 35: =1 if physical health is good
column 36: =1 if physical health is fair
column 37: =1 if physical health is poor

column 38: =1 if mental health is excellent
column 39: =1 if mental health is very good
column 40: =1 if mental health is good
column 41: =1 if mental health is fair
column 42: =1 if mental health is poor