Hello,

I am working with panel data for US-listed manufacturing firms for the period 2000-2019. I have used this data to conduct a firm-level fixed-effects regression to study the impact of Geographic Diversification (GSD) on Firm Performance (ROA). The result of this firm fixed effects regression is an inverted U relationship between GSD and ROA. The maximum point of this inverted U corresponds to GSD of 0.6. (This may be considered to be the optimum level of Geographic Diversification ie the level at which firm performance is maximized.)

While this is fine, it may be argued that the "optimum level" of Geographic Diversification could be different for different firms. It could certainly be different across industries.

To answer the question of whether this optimum Geographic Diversification differs across industries, I wanted to include the 2-digit industry as one of the terms. However, given the firm fixed effects model, it gets dropped off.

Would you have any other recommendations for me? Is there any way I can integrate "industry" into the firm-level fixed effects model. (Essentially I am trying to get 2 levels of fixed effects .... firm-level and industry-level.) Thank you.