0023-6152

AA00697606

22077

Unrefereed

0023-6152

AA00697606

Using an input–oriented DEA model with the assumption of VRS, this paper develops a framework on agricultural production efficiency including 2 outputs and 6 inputs. The data source is a survey to 99 household farms of Hebei province, China, conducted by the authors in 2010. According to the efficiency scores of DEA, the sampled farms are divided into three types, and the status of returns to scale for each farm is examined as well. Slack analysis of the outputs shows that comparing with net profit, ratio of net profit can be increased with a larger margin. Meanwhile, the radial and slack adjustments indicate the inefficient and redundant amounts of the inputs, respectively. In the second stage, effects of a variety of social and natural determinants are assessed, with the adoption of an Ordinal Logistic Regression model. Based on the empirical findings, policy recommendations are put forward, concerning the circulation of basic agricultural factors of land, labor, etc.; strengthening the construction of public agricultural infrastructures, particularly the irrigation facilities; deepening the institutional reforms to extend quality seeds and advanced agricultural technologies; and the closer integration of the public services and credit markets supporting the agriculture.