Dear all,

I am getting the error "too many base levels specified" when running the following lines:

Code:
forvalues x = 0/1 {
        eststo d_laborrights`x': reg working    active_50 future_50 age age2 urban i.intyear i.region if d_laborrights==`x'
                 test active_50-future_50=0        
                 estadd scalar difference = _b[active_50]-_b[future_50], replace
                 estadd scalar F_t = r(F), replace
                 estadd scalar p_f = r(p), replace
    }
suest d_laborrights0 d_laborrights1, cl(cluster)
My unit of analysis are individuals. working, active_50, future_50 are dummies.

d_laborrights is a dummy indicating two subgroups of countries that have lower labor rights (d_laborrights=0) and stronger labor rights (d_laborrights=1).
Through my internet search, I already learned that the issue may be that there is no value in i.region that is represented in both subgroups (i.region indicates subnational regions).

I am mainly interested in the active_50-future_50 test result. Is there a way to compare the test results between the two country subgroups?

With suest, I would have liked to run the following test:
Code:
test [d_laborrights0_mean]active_50 - [d_laborrights0_mean]future_50 = [d_laborrights1_mean]active_50 - [d_laborrights1_mean]future_50
Thank you very much!