Abstract (Translated):
This paper developed and applied the Stochastic Frontier Geographically Weighted (SFGW) to access the tax efficiency of the Brazilian states. The estimated coefficients obtained by this method were close to those from the traditional Stochastic Frontier estimation allowing, however, the examination their geographical variations. It has been found that southern states have, as well as Rio de Janeiro e São Paulo, a greater capacity of transforming economic product into tax collection. These states also show a greater marginal productivity in the tertiary sector, while the northern states have a relatively greater productivity in the secondary sector. The tributary productivity of the primary sector grows, in a distinct pattern, from east to west. The comparison between sectors, the tertiary sector showed the greater tributary productivity. The total potential increase in the state tax collection in year 2006 was estimated in R$ 39,6 billons, and the states that showed greater room for improvement were, respectively, Rio de Janeiro (R$ 9,3 billons), São Paulo (R$ 8,4 billons) e Santa Catarina (R$ 3,8 billons).