Test of Convergence in Agricultural Factor Productivity: A Semiparametric Approach

We tested for club convergence in U.S. agricultural total factory productivity using a sigma convergence test. We used the same club of states as used by McCunn and Huffman as well as different states within 10 clubs identified by the cluster analysis. Results showed convergence was evident only in a few club groups. Clusters group identified using a statistical methodidentified only converging clubs. Variables affecting total factor productivity among states were identified using parametric, semiparametric and nonparametric methods. Semiparametric and nonparametric methods gave a better fit than a parametric method as indicated by the specification test. Our results indicated that health care expenditure, public research and extension investment, and private expenditure are important variables impacting total factorproductivity differences across states.