Different Variables but same Coefficients and Standard Errors

[Ferhat Demirci]

Hello, community;
I am working on panel data to explain the effects of political risk on corporate investment. I use some independent variables at the firm level, such as RAO, leverage and firm size, etc. Also, I have added some macroeconomic variables to the country level. I have five models. The main model explains overall political risk, and the others explain sub-components. After running the first model, I changed the overall political risk variable with sub-component variables and took interesting results. Coefficients and standard error values stayed the same. I think it happened because of the time dummies. But I have to show time effects.

[Carlo Lazzaro]

Ferhat:
this omission usually happens because the predictor is perfectly collinear with the fixed effect.

[Ferhat Demirci]

Mr. Lazzaro,
Thanks for your interest. Do you mean omitted time variables? If yes, I did not ask that.
I replaced the political risk variable with conflict, but the results (r squared, coefficients, standard errors) are always the same as you see above.

[Carlo Lazzaro]

Ferhat:
it s difficult to reply positively without taking a look at an example/excerpt of your data, that you can share with intererested listers via CODE delimiters.

[Ferhat Demirci]

Ferhat:
it s difficult to reply positively without taking a look at an example/excerpt of your data, that you can share with intererested listers via CODE delimiters.

Hello, again Mr. Lazzaro,
I am posting a sample of my dataset and do file that includes my commands. I hope useful for you. Thanks a lot for your interest.
Please notice that GDP Growth, PoliticalRiskRating, and Conflict don't differ for cross-sections. These are country-level variables and the same for all firms.

Sample Dataset Sample.xlsx
Do file(commands) Commands.do

[Carlo Lazzaro]

Ferhat:
1) your .do file seems ok;
2) probably nobody on this forum would ever open spreadsheets coming from unknown sources due to the risk of downloading active contents. Please share an example/excerpt of your data via CODE delimiters. Thanks.