cdeco command in Stata 16

[Olena Ch]

Hello Everyone.

I am trying to use cdeco command to decompose the gender difference in wages. The command was written by Chernozhukov, Fernández-Val and Melly (2013, “Inference on counterfactual distributions”). The command works well in case of a continuous dependent variable. I am trying to decompose a binary dependent variable, which is "1", in case wage is below the threshold and "0" otherwise. The method used is "probit". The error message I receive is the following:

est_probit(): 3200 conformability error
<istmt>: - function returned error

Do you know how can I solve this issue? Any feedback is appreciated.

[Sven-Kristjan Bormann]

You cannot use the cdeco command for a binary dependent variable because a binary variable does not have quantiles by definition. The option "method" is used to indicate which method should be used to estimate the conditional distribution for the counterfactual distribution. You can choose the method "probit" to estimate the quantiles of the conditional distribution but the dependent variable still needs to be continuous.

You could look at the normal Oaxaca-Blinder decomposition or a Fairlie decomposition instead. See the codes below install these commands instead.

ssc describe oaxaca
ssc install oaxaca

ssc describe fairlie
ssc install fairlie

[Olena Ch]

Thank you very much for your clarification. You are right about decompositional issues with the binary dependent variable.

I have another issue with cdeco command. I am trying to decompose the earnings variable. The decomposition works well without industry dummies. Once I am adding them (15 dummies for each industry), the following message shows up:

est_probit(): 3200 conformability error
<istmt>: - function returned error

The number of observations for each covariate used in decomposition is the same. Also, when I am trying to add just one industry dummy, the decomposition works as well. Do you have any idea about this problem?

[Sven-Kristjan Bormann]

If I remember correctly, then I did not specify dummy variables separately but only as one variable. This would be the industry variable in your case.
I might have encountered the same problems as you but I don't want to re-run the old
For the decomposition results it should not matter if you have dummies for each category separately or only one variable containing these categories.
Have tried to use a method like method(qr) or method(loc)?

[Farogat WIUT]

Hi all,

I want to use counterfactual decomposition method with a binary dependent variable. I used cdeco, but stata gives me an error message. What do you recommend me to do?

[Clyde Schechter]

#5 is a duplicate post, asked and answered at https://www.statalist.org/forums/for...position-issue.

[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.