help nopomatch
-------------------------------------------------------------------------------

Title
    nopomatch -- Decomposition of Nopo (2008)

Syntax
        nopomatch matchvar [matchvar(s)] [if] , outcome(varname) by(varname)
                 [options]

Parameters
      matchvar(s):                   list of variables used to match
                                      observations of the groups defined by
                                      by().
      outcome:                       outcome for which the gap will be
                                      computed. (e.g.: wage)
      by:                            dummy (0/1) that discriminates the
                                      sample to perform the computation of
                                      the gap. (e.g.: gender).


    options                         Description
    -------------------------------------------------------------------------
    Main
      sd                             computes standard deviations.
      replace                        deletes the _match and _supp variables
                                      previously created (if any).
      reportby(varname)              split the results for different groups
                                      of observations.
      fact(factor-variable)          specifies the use of frequency weights.
      filename(file-name)            saves output in STATA file.
    -------------------------------------------------------------------------


Description
    nopomatch -- module to decompose the gap in an outcome between two groups
    using matching, following the methodology developed by Nopo (2008).
    nopomatch performs the matching, computes the gap decomposition and
    provides the standard deviation of the unexplained component of the gap.

Options
        +------+
    ----+ Main +-------------------------------------------------------------
    sd computes standard deviations of the unexplained components of the gap.
        Note that the usage of this option substancially increases
        significantly the computation time.

    replace deletes the _match and _supp variables previously created (if
        any).

    reportby(varname) repeats the command for each group of observations of
        each category of varname.

    fact(varname) specifies the use of a weighting factor variable (frequency
        weight) in the data.

    filename(string) creates an output STATA file (string.dta) with the
        results of the decomposition.

Output

A typical output of the nopomatch module looks like this:

****************************************************
*****  Gap in matchvar(s) decomposition 
****************************************************

      D = #     is the total gap, the difference between averages in the
                  outcome-variable of the comparing groups. It is expressed
                  as a percentage of the average of outcome-variable for the
                  group in which by-variable=0
      D0= #     is the unexplained part of the gap.
      DM= #     is the part of the gap that can be explained by the
                  differences in characteristics between two groups of the
                  individuals with by-variable=1, that are in and out of the
                  common support of the variables defined by
                  match-variable(s)
      DF= #     is the part of the gap that can be explained by the
                  differences in characteristics between two groups of the
                  individuals with by-variable=0, that are in and out of the
                  common support of the variables defined by
                  match-variable(s)
      DX= #     is the part of the gap that can be explained by differences
                  in the distribution of characteristics of the two groups
                  discriminated by by-variable in the common support of the
                  variables defined by match-variable(s)
****************************************************
      percM= #  is the percentage of the individuals with by-variable=1 that
                  are in the common support of the variables defined by
                  match-variable(s)
      percF= #  is the percentage of the individuals with by-variable=0 that
                  are in the common support of the variables defined by
                  match-variable(s)
****************************************************


Note that D = DO + DM + DF + DX. For technical details see Nopo (2008). 

Examples

    . nopomatch age education, outcome(wage) by(male)

    . nopomatch age education, outcome(wage) by(male) fact(freqweight) sd
        replace filename(example1)

    . nopomatch age education occupation sector, outcome(wage) by(male)
        fact(freqweight) reportby(education)

    . nopomatch age education occupation sector, outcome(wage) by(male) sd
        replace fact(freqweight) reportby(firmsize) filename(example2)


References

Nopo, H. 2008. Matching as a Tool to Decompose Wage Gaps Review of Economics an
> d Statistics, 90(2): 290-299.

Acknowledgements

This code was built from previous versions developed by Sebastian Calónico and 
> Luana Marques.

Disclaimer

This program is provided without warranty of any kind. The authors are not resp
> onsible for any cost derived by the usage of this program.

Authors

Atal, Juan Pablo; Hoyos, Alejandro and Hugo Nopo. If you observe any problems m
> ailto:hugon@iadb.org