Evaluating the Contribution of Infrastructure to U.S. Agri−Food Sector Output

The effect of infrastructure investment or capital on private sector output and productivity has been widely discussed over the past two decades. However, only limited studies have focused on the contribution of infrastructure to the output of U.S. agricultural and food sector. Considering the importance of agriculture in U.S. economy and its strong dependence on infrastructure, two empirical analyses were made in this thesis to evaluate the output impact associated with infrastructure in agricultural and food sector in the United States. The first study examines the spillover effect of two major transportation modes, roads and rails, on agricultural output across the United States. Applying a spatial econometric model that considers spatial dependence in both output and explanatory variables to agricultural output in 44 states from 1981 to 2004, the analysis suggests that spending on road system has significant and positive direct effect on agricultural output. However, the spillover effect of road infrastructure on agricultural output in the neighboring state varies along with the order of neighboring states considered in the model. No significant direct and spillover effect of rail infrastructure is observed. The second study estimates the dynamic effects of public infrastructure on the output of 34 U.S. food manufacturing industries during the period of 1958−2005. The contribution of public infrastructures to food industry output growth is assessed using heterogeneous dynamic methods. Results suggest a positive and significant contribution of public infrastructures to the output growth in food manufacturing industry. Also, a long-run relationship is confirmed between food manufacturing output, input use and public infrastructure. Moreover, the output effect of public infrastructure is homogeneous among the food manufacturing industries in the long run but varies in the short run. The findings of both studies in this thesis have several policy implications.