Hello,

I am working on my MSc thesis in studying the impact of lobbying on company financial performance. I have access to an unbalanced panel dataset with 300 companies and 8 years of data, with information including lobbying expenses, company, and sector controlling variables.

I am interested in testing whether the impact of lobbying expenses (independent var.) on profitability measures (dependent var.) is greater in the short or long term. To do so, I am wondering whether it is appropriate to:

A-) make use of the panel data and lag the independent variable (for the 8 years) using Fixed Effects, First Differences and Random Effects methodologies, OR;

B-) do I have to restrict myself to using pooled OLS and/or OLS via cross-sectional analysis.

I know that I have several tests available that support the decision of which ones to pick between FE, FD, RE, and pooled OLS, might question is more on whether it makes sense to lag the independent variable while using these methods.

Also, one of my concerns is how to interpret the results when dealing with approach (A) - since I am including several years for the dependent variables, how do I interpret the lagged variable coefficients? Is it as straightforward as cross-sectional OLS?

Thank you