There are two folders: numerical and empirical For running the programs in the numerical folder: ------------------------------------------------- Software: Matlab2020b Operating system used: MacOS or Windows Expected computation time: it takes about 3 hours to run the calibration search program (calibration_fimsch.m), and less than 0.5 hour to run each of the search programs in the "Policy" section. It takes less than 5 minutes to run the rest of the programs. To run these programs, one needs to complete the programs under "Calibration (Section 4)" first before running the rest of the programs. Information on the files in the "numerical" folder is provided below. Calibration (Section 4): "calibration_fimsch.m": this file searches value of parameters in calibration. "objfun.m": the objective function for the search file above "calibration.m": solves the numerical model with parameters solved from the search file “fun_hivpr.m”: function to solve hiv prevalence rates Counterfactual (Section 5): CF_fert.m: eliminates premarital fertility motives. fun_hivpr_CFfert.m: function to compute his prevalence rates in CF_fert.m CF_fert_contra.m: eliminates both fertility motives and contraception costs (all 3 psi's=0) CF_contra1.m: reduce disutility of contraception by half for all sex CF_contra2.m: reduce disutility of contraception by half for premarital sex only CF_contra3.m: reduce disutility of contraception to 0 for all sex CF_mar1.m: increase lambdaM_Bar by 30% CF_mar2.m: double lambdaM_Bar in BM CF_mar3.m: add a positive utility value to marriage (30% of pop average (V0-Vd) for all groups) CF_mar4.m: add a positive utility value to marriage (100% of pop average (V0-Vd) for all groups) CF_mar5.m: add a positive utility value to marriage (30% of (V0-Vd) for each group) CF_mar6.m: add a positive utility value to marriage (100% of (V0-Vd) for each group) Policy (Section 6): TR_BM.m: computes total revenues from benchmark model PL_inc_sch.m: search the subsidy rate for income subsidy policy (subsidise all women) PL_inc_obj.m: objective function of the search program above PL_inc.m: solve the model with income subsidy policy (subsidise all women) using the subsidy rate searched from the program above PL_incL_sch.m: search the subsidy rate for income subsidy policy (subsidise less-educated women) PL_incL_obj.m: objective function of the search program above PL_incL.m: solve the model with income subsidy policy (subsidise less-educated women) using the subsidy rate searched from the program above PL_incpm_sch.m: search subsidy rate for income subsidy policy (subsidise unmarried women) PL_incpm_obj.m: objective function of the above PL_incpm.m: solve the model with income subsidy policy subsidise unmarried women) using the subsidy rate searched from the program above PL_edu_sch.m: search subsidy rate for education subsidy policy (allow all women to continue education and subsidise all women) PL_edu_obj.m: objective function of the above PL_edu.m: calibrate the model with education subsidy policy (allow all women to continue education and subsidise all women) PL_ART_sch.m: search subsidy rate for ART subsidy policy PL_ART_obj.m: objective function of the above PL_ART.m: solve the model with ART subsidy policy PL_ART2.m: solve the same model with ART subsidy policy but assume no behaviour change from BM Extensions (Section 7): calibration_ex08.m: calibrate 2008 model, assuming part of women have no HIV knowledge (lambdaI=lamdbaI_bar) calibration_ex98.m: calibrate 1998 model, assuming more women have no HIV knowledge (lambdaI=lamdbaI_bar), and psi_u increases (more for less-educated) fun_hivpr_ex.m: function to solve hiv prevalence rates in the extended model extension.m: draw figures for the Extension section. HIV prevalence_data.mat: HIV prevalence for adults, women, young women in 1990-2013 from UNIAIDS data. extension_data.mat: stores data used in the extension section. For running the programs in the empirical folder: ------------------------------------------------- Software: STATA14 Operating system used: MacOS or Windows Expected computation time: the do files can be run within 5 minutes each. Below is the information for files in this "empirical" folder (the do files are independent and can be run in any order). Kenya_2008.do: this file computes/constructs key statistics and moments from Kenya DHS (2008) for calibration and extraction of other useful information. Kenya_allyears.do: this file computes/constructs key statistics and moments from Kenya DHS (1989, 1993, 1998, 2003, 2008) that provides a cross-year comparison and targeted moments for Section 7. Kenya_reg.do: this file constructs a sample for regression analysis using Kenya DHS (1989-2008) and runs regression in Section 2 and Appendix. KEwomen_compiled.dta: this is a data file that combines all years (1989-2008) data compiled from Kenya DHS using the codes provided. KEwomen_compiled2.dta: this is a data file that combines all years (1989-2008) data compiled from Kenya DHS using the codes provided, including women's birth order data. KEwomen"year"_compiled.dta: this is a data file that includes data compiled from Kenya DHS of a particular year using the codes provided. KEwomen"year"_compiled2.dta: this is a data file that includes data compiled from Kenya DHS of a particular year using the codes provided, including women's birth order data. cross country_allyear.dta: this is the sample used for cross-country regressions. cc_reg_allyear.do: this file runs cross-country regressions.