This file explains how to use the code for Taxes and Female Labor Supply by Remzi Kaygusuz.

Below you will also find an explanation about the data files and stata do file.

****************************************************************************************************

There are 5 matlab files: mainR.m, married1.m, married2.m, parameters.m,
and txf.m .

mainR.m is the main body of the code. When you type mainR.m in the command line
the code asks you to input data years for wages, distribution, and taxes. For instance,
if you type 1980 for a wage year, the code uses the wages from 1980 US data. Similarly,
if you type 1990 for tax year the code uses 1990 tax rates.

The outputs are the values of the variables in the Tables 12, 13, or 17 in the paper.
In order to reproduce Table 10, you need to tpye LFP80 and then distmat in the command
line after the code completes the computations. The benchmak labor force partitipation rates
are reproduced only when the wage, distribution, and tax years are entered as 1980.


Parameters of the model are initailized in parameters.m. This file is called by mainR.m .
parameters.m calls the following data files. These are compiled from either census or cps data.
The Stata dofile used to compile this data is included in the package. 


The following are data files for female labor force participation in 1970,
1980, 1990, and 2000, respectively.

LFP70.TXT, LFP80.TXT, LFP90.TXT, LFP00.TXT.

The following are data files for wages in 1970,
1980, 1990, and 2000, respectively.

wagedata1970.TXT, wagedata1980.TXT, wagedata1990.TXT, wagedata2000.TXT.

The following are data files for distributions in 1970,
1980, 1990, and 2000, respectively.

marriagedata1970.TXT, marriagedata1980.TXT, marriagedata1990.TXT, marriagedata2000.TXT.


married1.m is negative of utility function for two earner households (ignores the utility cost), and
married2.m is negative of utility function for single earner households.

********************************************************************************************************


dofile.do compiles the data used in the paper. I used Stata 10.1 to run this code.
You have to change the working directory of Stata to the directory where you save the data files.
4 census data files (for 1970, 1980, 1990, 2000) and 4 cps data files are used (for 1970, 1980, 1990, 2000).
These files are from http://ipums.org/  . Please refer to this web site for the data. If you set the directory
accurately and run the do file, the code is going to deliver all the statistics used in the paper. 
For variable definitions see ipums web page. Below you will find the description of the variables created by 
the author. There is a also a tax data file (compiled from IRS publications). See the paper for the references.

edlvl= education variable as described in the paper.
nage= age variable as described in the paper.
part= indicates labor force participation.
yhrswrk=annual hours worked
hwage= hourly wages
wwage= weekly wages
agi90=adjusted gross income in 1990 (similarly for other years)
avgtax90= average tax payment in 1990 (similarly for other years)
tax90=total taxes paid in 1990 (similarly for other years)
taxret90=number of taxable returns in 1990 (similarly for other years)
returns90=number of all tax returns in 1990 (similarly for other years)