************************************************************************************************** *** Additional materials for the paper *** “Quantitative Analysis of Health Insurance Reform: Separating Regulation from Redistribution” by Svetlana Pashchenko and Ponpoje Porapakkarm September, 2012 ************************************************************************************************** The folder contains three groups of files: the Fortran code, input files needed to run the code (files with extension .in) and output files produced by the code (files with extension .out). The Fortran code consists of the following files: - main1.f90 – the main program - mod_calibration1.f90 – sets the parameters and initial guesses of the model. This file also contains indexes that determine whether the model is a baseline or an experiment, and whatcomponents of the reform are included in the experiment. - mod_computation1.f90 – defines key variables, sets the grid points, transition matrixes and computes the distribution over some state variables. - mod_global_var1.f90 – defines some variables that are used in many subroutines and functions. - Solve_NR1.f90, value_funNR1.f90 – solve for the value function of people of the working age. - Solve_Ret1.f90, value_funR1.f90 – solve for the value function of retirees. - initvalue_fun1.f90 – solves for the value function for the last period of life. - mod_service1.f90 – contains some technical subroutines and functions used for interpolation and optimization. The files should be compiled in the following order: mod_types.f90 mod_service1.f90 mod_calibration1.f90 mod_computation1.f90 mod_interface1.f90 mod_global_var1.f90 main1.f90 To run a baseline, the following indexes in the mod_calibration1.f90 file should be set to 1: Baseline = 1 nCRpool = 1 indPen = 1 indSub = 1 indMCD = 1 The code will produce the files baseline.in, GammaE.in, GammaJR.in, GammaR.in, VEbaseline.in, VJRbaseline.in, and VRbaseline. The first file contains calibrated parameters of the baseline model and the last six files – distribution of individuals over the state-space and their baseline value functions. These files are needed to run an experimental case. To run an experiment, an index baseline should be set to 2. The following indexes should be also changed to consider different versions of the reform: - nCRpool – indicates whether there is community rating in the individual market. If it is equal to 0 the market is unregulated. If it is equal to wnages (which is equal to 40), there is age-adjusted community rating. - indPen – indicates whether there is a penalty for uninsured. 1 means no penalty, 2 means there is a penalty. - indSub – indicates whether there is a subsidy. 1 means no subsidy, 3 means there is a subsidy as in the reform. - indMCD – indicates whether there is a Medicaid expansion. 1 means no expansion, 2 means expansion. To compute steady-states the parameter TRperiod (the number of transition periods)in mod_calibration1.f90 should be set to 0. To run the transition it should be set to 85. The key output files are the following: -calculation.out1 contains information about employment, insurance statistics and aggregate variables. -welfareSS.out contains welfare statistics for the case of the steady-state. -welfareTR.out. reports the consumption equivalent variation when the transition is taken into account. -Emplyrate.out and InsStat.out report more detailed insurance and employment statistics for the steady-states. -TR_employment.out and TR_insurance.out report detailed insurance and employment statistics during the transition period. -Transa0.out, TranscE.out, TransCRp.out, TranspESI.out, and Transrr.out report how the tax function parameter a0, the wage deduction cE, community rated premiums in the individual market, the ESI premium and the interest rate adjust during the transition period. We can provide additional clarification about this code and about the replication of the results if something is not clear. Please contact us at sap9v@virginia.edu or ponpojep@umac.mo.