Hello all,
I am a new member and I appreciate all your help and comments.
I am interested in evaluating the effect of temporary tax cut policies on firm performance. Specifically, I have annual data at the firm-level in 2000-2017. The first intervention was a CIT cut in 2008 and 2009, some firms (TreatCIT) only need to pay 70% CIT in 2008 and 2009. The second one was the VAT cut in 2009, the tax rate reduces 50% in some goods and services. All the firms were affected by at least 1 intervention. Particularly, in 2009, the total number of firms was 3675, 1970 affected by CIT cut only, 175 affected by VAT cut only, and 1530 affected by both kinds of the tax cut (no firm is unaffected, so there is no normal control group).
I want to see the effect of each kind of tax cut, and the total effect of both. The specification is as follow:
yit = firmi + yeart + α*TreatCIT*posCIT + β*TreatVAT*posVAT + γTreatCIT*TreatVAT*posVAT + Xitδ + u
In which: TreatCIT is a dummy for firms received CIT cut
TreatVAT is a dummy for firms received VAT cut
posCIT is a year dummy for years from 2008-2017
posVAT is a year dummy for years from 2009-2017
My questions are:
1. Can I obtain the effect of each kind of tax cut and the total effect of both with the above regression?
2. If not, can I run 2 separate generalized DID regressions to evaluate each policy impact? How to obtain the total effect?
Thank you in advance.
Viet
p/s: I attached a graph for illustration.
Related Posts with Difference in differences with multiple interventions and multiple treatment groups
how could the difference of number of a variable in a panel data be shown in statahello, i want to see how many people' living status(dummy variable) have been changed using a 2-peri…
create the day before a dateHello! I'm still working with dates and now have a new problem. If I have a date 22dec2014, I want …
Turning variable labels into variable names with a suffixHi I have data where my years are recorded in letters from H to CC (representing 1948 to 2021). The …
Creating event time variableHi, I have a time series of stock returns with multiple event date indicators, such as the following…
Adding multiple labels to regsaveI am saving coefficients using regsave (from ssc) and I am able to use the 'addlabel' option to add …
Subscribe to:
Post Comments (Atom)
0 Response to Difference in differences with multiple interventions and multiple treatment groups
Post a Comment