Suppose you have panel data on 1500 firms between 2001 and 2020. That is, you observe data on each firm, annually, for 20 years. You want to test the prediction that larger firms are more productive, and you decide to regress firm productivity on number of employees (size) and control for firm fixed effects