Hi all,

I try to establish a standard DiD model with fixed effect to investigate the impacts of a minimum quality standard impact on the change of coverage rates of 2 different types' vaccines. One type is high coverage rates vaccine and another one is low coverage rates. The specification is as follows:
Array

The outcome variable outcome (Yit) is the coverage of the immunisation. "post" identifies the period before and after the effective year of the policy, that is year 2005 (i.e., post = 1 means after Year 2005, post = 0 means before Year 2005). The key thing is in my DiD, both groups (2 type's vaccine) are affected by the policy impact. Is that possible that I refine the definition of treatment and control group, in other words, In my case, I defined that in the country with the policy impact, both before and after the change, the treatment group is “High” type quality vaccine, and control group is the “Low” type quality vaccine? and it is shown as a "high" variable in the model ( it should be a "treat" variable in normal DD analysis), and if high = 1 --> high type, high = 0 --> low type vaccine.

After the new definition of two groups, then I can investigate how the policy gives two groups "additional" change based on the original coverage level. *Parallel Assumption has been met. So after running the regression, as per the result of DiD:

Table 3: Difference-In-Difference Estimation Results
Outcome var. Coverage (%) SE t P>t
Before
Low Quality 83.423
High Quality 93.846
Diff (H-L) 10.423 0.757 13.99 0.000***
After
Low Quality 91.815
High Quality 97.422
Diff (H-L) 5.607 0.406 13.40 0.000***
DiD -4.816

1. Both types' coverages increase after the policy.
2. Low type increases more on coverage than the high type
3. The DiD estimator (-4.816) gives me intuition that compared to the pre-intervention period, the gap between high and low type vaccine are contracted after the policy.

In conclusion, my concern is just whether this could be a DiD after I redefining the definitions of the T & C groups?

Thank you so much!