초록 close

Estimating only decision making units’ efficiency can’t give important information to decision makers or policy makers; and the methods that estimate efficiency at the first stage and analyze deterministic factors of efficiency at the second stage by using OLS or Tobit have several innate problems. Therefore, this paper tries to recover these innate problems by using kernel regression at the second stage.Efficiency estimate results of each provinceㆍcity of China by the first stage analysis show that while Shandong, Henan, and Guangdong get the highest efficiency score, one, Shanghai, Beijing, and Tianjin get very low efficiency scores even though these cities have high population density. Efficiency’s five environmental deterministic factors are considered at the second stage; the results show that the relationship between efficiency and tax revenue per capita of each provinceㆍcity are reverse; the relationship between efficiency and extra tax revenues are positive; efficiency and expenditure divided by total revenues are reverse; efficiency and disposable income are irrelevant; and the relationship between efficiency and the number of college or upper level- educated people are positive.