Dear All,

I hope I can get a suggestion from someone in the field of spatial econometrics, although the question is not strictly related to Stata usage. I estimate the following model:
Array

The estimation reports that zeta is positive, while rho is negative, this indicating, if I am correct, that there exists a negative spatial dependence in the error term. I read different papers documenting that such a situation, although little explored, is not unusual. Nonetheless, I still have some doubt about the interpretation of such a result. Could you please offer me some insight about how to interpret those results?

Thanks in advance for your kind help.

Dario