You want to gage the relationship between education and income and see how government spending impacts education. You have data for 100 US cities between 2000-2020, and run the regression:

10_it = ₁ + _t + ₁ money_education_it+ ₂population_it + _it

money_education is how much city i in year t spends on education

populaiton is the population in city i and year t

income10 is the average income in that same city i 10 years later (aka in year t+10)

a_i and _t are the city and year fixed effects a_i equals one for city i and is zero otherwise for all i, _t is equal to one in year t and zero for all other years for t= 2001, ..., 2020. 2000 is not included.

You restrict the regression by dropping all area fixed effects and instead a single constant is added. This is to gage the relevance of the area fixed effects. Let's say the f-statistic = 129.67.

What are the degrees of freedom?

What F Table critical value should we use?

What possible endogeneities are do we still have left even if we run this model that has time and city fixed effects?