Counterfactual decomposition issue

[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?

cdeco owns_dwelling male age_group high_ed married divorced separated widowed log_monspending hhsize capital, by(country) method(probit)

command cdeco is unrecognized
r(199);


Thanks,
Farogat.

[Clyde Schechter]

I don't know what the counterfactual decomposition method is, and I have never heard of cdeco, but this problem has a generic solution. Whenever Stata tells you that a command is unrecognized, unless you have mistyped its name, it means that it is not an official Stata command, nor is it among the user-written commands you have installed on your computer.

The response to that is to locate the user-written command you are looking for. This you do by running:

findit cdeco

When you do that, a Viewer screen will pop up showing you a list of possibilities. (In this case the list consists of only one possibility.) Pick the one that looks like what you are looking for and click on the link for it (shown in blue). That link will take you to a page describing the program in more detail. If that's the one you want, click on the "click here to install" link, it will install itself, and then you should be good to go.

[Jared Greathouse]

Farogat WIUT You're looking for

net install counterfactual, from("https://sites.google.com/site/mellyblaise/") replace

Clyde Schechter Apparently, the command "cdeco uses the counterfactual distribution to decompose the differences between two observable quantile functions. This is similar in principle to the famous Oaxaca/Blinder decomposition of the quantile differences".

Really wish I was smart enough to know why we'd need a quantile function in this scenario, or what one even is and how it works, but it appears to be some cool-ish treatment effects estimator.