Dear researchers,
I have unbalanced panel data for a set of firms for the year extends from 2000-2019. In 2010 a new standard has been issued, and accordingly, firms started to adopt the standard, but the adoption process is not simultaneous. I mean group of firms adopted in 2010, other groups adopted in 2011, other groups adopted in 2012, and so on. The main variables of interest are growth and leverage. Thus, my main aims are as below:
  1. To examine the relationship between growth and leverage before the adoption of the standards.
  2. To examine the relationship between growth and leverage after the adoption of the standards.
  3. To examine the impact of the standard on the relationship between growth and leverage.
Kindly note that the adoption of the standards measured as a dummy variable (Event) equal 1 in the years of adoption and zero, otherwise.
At first, and to examine the first and second aim mentioned above, I have used the following codes:
Code:
ge id= _n
encode Companyname, gen(COMPANY)
xtset COMPANY Year, yearly
First model:
Code:
Xtreg Growth Leverage Age Size i.Year if Event==0, fe cluster (COMPANY)
Second model:
Code:
Xtreg Growth Leverage Age Size i.Year if Event==1, fe cluster (COMPANY)
Third model DID:
Code:
xtreg Growth i.Event##(c.Leverage c.Age c.Size ) i.Year, fe cluster ( COMPANY)
I know that the coefficients before the interaction show me the relationship with the dependent variable before the adoption of the standards. While the coefficients of the interaction between Event and each independent variable show me the impact of Event on the relationship between the dependent and independent variables. So, I have two questions:
  1. If the DID model will show me the relationship between dependent and independent variables before the adoption of the standards (i.e. before the interaction), then do you recommend using the first model.
  2. If the DID will show me the relationship before the adoption (i.e. before the interaction), and the impact of the Event on the relationship, then how to know the relationship between the dependent and independent variable after the adoption. I know that the summation of the coefficient before the adoption and the coefficient after the interaction will show me the relationship after the interaction, but I need to see the Sig-value.
  3. Can I re-group the model by industry and year as a robustness check, if yes, what is the code?
Many thanks in advance.