Using estimated coefficient on GWR regression as dependent variables.

[Salahudin Al Ayyubi]

I conduct research on the factors causing spillover in local government spending over 5 years across 514 municipalities as well as the impact of the spillover on regional productivity convergence. My plan is to manually create a spatial variable for government spending and utilize Geographically Weighted Regression (GWR) to calculate spatial coefficients for each region. The coefficients obtained from the GWR regression will then serve as proxies for spillover variable. Is that possible? or does anyone know a more appropriate method for measuring the magnitude of spillovers in the context of public policy?