Hi everyone,
Hope all is well.
I am hoping someone might be able to give some guidance on the following:
I have repeated cross-sections for firms from 2015-2022. I have a treatment (experiencing a shortage), and I want to see how this affects things like fuel intensity and efficiency.
I have tried CEM in a very general sense; however, I don't believe I have taken care of the time aspect, so to ensure I'm only matching like with like, and end up with a matched sample with firms that are the same across years.
E.g. I may have a firm visible in 2016,2017,2019,2020, where they reported shortages in just 2017 and 2020. (i.e shortage ==1).
Is it a case of (1) identifying the firms in the year in which there is a shortage and keep if the shortage ==1 so there is one row per treated firm; (2) save these firms in a temp (3) and merge back to my full dataset where I can create a new variable if _merge ==3, it gets a value of 1, otherwsie, it is 0. Then from there carry out the matching ..?
Any help/advice would be greatly appreciated.
Thank you!
Gráinne
Hope all is well.
I am hoping someone might be able to give some guidance on the following:
I have repeated cross-sections for firms from 2015-2022. I have a treatment (experiencing a shortage), and I want to see how this affects things like fuel intensity and efficiency.
I have tried CEM in a very general sense; however, I don't believe I have taken care of the time aspect, so to ensure I'm only matching like with like, and end up with a matched sample with firms that are the same across years.
E.g. I may have a firm visible in 2016,2017,2019,2020, where they reported shortages in just 2017 and 2020. (i.e shortage ==1).
Is it a case of (1) identifying the firms in the year in which there is a shortage and keep if the shortage ==1 so there is one row per treated firm; (2) save these firms in a temp (3) and merge back to my full dataset where I can create a new variable if _merge ==3, it gets a value of 1, otherwsie, it is 0. Then from there carry out the matching ..?
Any help/advice would be greatly appreciated.
Thank you!
Gráinne
