avoiding duplicates from propensity score matching

Question

I am using matchit for my firm level analysis.

psm_out <- matchit(treat~firm_variables, data=firm_data, 
method="nearest", exact=c("year","industry"), ratio=1)

The generated result gave me a lot of duplicates. For example, treatment sample is firm A's year 2016. And within the matched control group, I have firm A's year 2010. I want a matched control group without treatment group's other years. How can I get a desired matching result?

Answer 1

You can solve this quite easily by removing from your control group firms that appear as treated for one year at least. Once you have corrected this, you can run the command you posted.