I am using a fixed effects model for a monthly panel data set, which is of the form:
Code:
xtreg ln_hrs did time treated i.month married age, fe
I am doing a difference-in-difference analysis where the policy starts in 201610 and time = month > 201609

Now, I am being advised to include a lagged DV. Specifically, the change in work hours after time should include the effects of working hours in the preceding month time-1 (201609), because I expect that individuals change in work hours are influenced by the previous month's level.

My question is how to go about implementing this lagged DV? I want to use the difference between my outcome variable ln_hrs at period time (201610) and time−1 (201609) as my DV for period time

Moreover, I'd like to see the distribution of these individual's work hours between the these 2 periods in a kdensity or histogram.