I would greatly appreciate your advice regarding using psmatch2 for matching treatment from two different years. Using a Nearest Neighbour PSM approach without replacement, I would like to match treatment observations across the two time periods (2010 and 2014) and create a "pseudo panel" of treatment. I pooled cross-sectional data for 2010 & 2014 and keep only treatment ==1.
My question is whether I should include time variation here or not. I used the following code but I am not sure if it is correct or not
Or I should use the following code:
Thanks in advance for your advice
My question is whether I should include time variation here or not. I used the following code but I am not sure if it is correct or not
Code:
psmatch2 year14 $xlist, outcome(CHE10) common logit ate noreplacement
Code:
psmatch2 treatment year14 $xlist, outcome(CHE10) common logit ate noreplacement
