Counterfactual decomposition in Stata

[Maud Wester]

I am trying to analyse the change in wage differentials between men and women in the Netherlands, using OSA 1990 and 2010.
Based on the model proposed by Chernozhukov, V., I. Fernández-Val, and B. Melly (2013), I want to do a counterfactual analysis based on a distribution regression.
I converted the variables gender (female=1), marital status(married or living together =1), country of origin (dutch = 1), education (high education=1) and age (older than 38=1) into dummy variables.
However, when executing the counterfactual command in Stata 17 as follows:
counterfactual log_income marital origin education old group(gen) method(logit)
I get the following error:
factor-variable and time-series operators not allowed (101)
Can somebody help me solving this problem or proposing a different approach?

Thanks in advance!

[Øyvind Snilsberg]

use a comma before the options,

counterfactual log_income marital origin education old, group(gen) method(logit)

[William Lisowski]

The counterfactual command is apparently a community contributed Stata command, rather than part of base Stata.

The paper you cite is behind a paywall. For those lacking access, the authors also gave a presentation to the 2016 Swiss Stata Users Group Meeting in Bern

https://www.stata.com/meeting/switze...tzerland16.pdf

which explains their methodology and demonstrates the use of the cdeco and counterfactual commands, but unfortunately does not give a source from which the commands can be downloaded.

Further searching of the web finds a Google site for the author Blaise Melly on which at

https://sites.google.com/site/blaise...-distributions

the commands are described and installation instructions are given.

If the command you ran was

counterfactual log_income marital origin education old group(gen) method(logit)

then you neglected the comma that must separate the variable list from the options. From the output of help counterfactual we see

Syntax

       counterfactual depvar indepvars [if] [in] [weight] [, options]

which tells us that if any options are to be provided, the must be preceded by a comma. When I ran one of the author's examples given in the help output and omitted the comma, I got the same unhelpful error message you did.

[Maud Wester]

Thank you both so much!

[Kadir Ruslan]

I am doing the decomposition of wages difference using cdeco. How do I get the results of decomposition for all independent variables not only the total difference, explained and explained? I want also to identify the contributing factor along the distribution.