How are DiD regression models interpretted?

[Abdul Malik Awan]

My apologies for the beginner question. I'm new to STATA so still learning.

My dataset looks at city level time-series on employment rate from '69 to '09. I am trying to see the effect of a minimum wage introduction in selected cities in a country. I am only using data from '74 and '75 in my evaluation.

Here is the code I tried:

---

clear all
set more off
capture log close
use "C:\StateCrime.dta"
tab year, gen(y)
gen minwage = 0
replace minwage = 1 if cityID == 11 | cityID == 23 | cityID == 45 | cityID == 57 | cityID == 60 | cityID == 61
reg empr minwage if year==1975
gen treat = cityID == 11 | cityID == 23 | cityID == 45 | cityID == 57 | cityID == 60 | cityID == 61
gen after = year >= 1975
table treat after if year == 1974 | year == 1975, c(mean empr)
gen did = treat * after
reg empr did treat after if year == 1974 | year == 1975

---

I want to interpret the coefficients of my model. I have attached a picture of the output.

Would I be right in saying:

Intercept is there is 85.8% is avg employment rate in affected cities in '74.
DiD estimate of -0.2% so min wage reduced employment rate by this much.
Treatment estimate of -1.2% so min wage reduced employment rate by this much.
After estimate of +0.6% so min wage increased employment rate by this much.

I can claim that the DiD result is not statistically significant at any level. However the treatment and after results are statistically significant at a 5% level. I arrived at this conclusion because if the t-value produced is less than 0.05 then I can reject the H0 stating there is no significant effect of introducing a minimum wage policy on employment rates, and consequently I accept the H1.

[George Ford]

There is no effect on employment by did variable. Treat merely measures the means difference between the treated and untreated prior to the treatment. after measures the change in the dv post treatment (for all firms, on average).

[Abdul Malik Awan]

There is no effect on employment by did variable. Treat merely measures the means difference between the treated and untreated prior to the treatment. after measures the change in the dv post treatment (for all firms, on average).

Thank you George for the clarification. You're the man. Have a blessed day!