Investment and expectations in the U.S. farm sector

Abstract

The objective of this study was to analyze the interrelationship between expectations patterns and investment behavior of producers in the U.S. farm sector. A major problem confronting U.S. farm sector is overinvestment and the asset fixity of farm physical assets. An understanding of farmer's investment behavior for capital goods is needed by both farm financial intermediaries to evaluate potential loan demand and by manufactures of farm capital inputs to formulate production and marketing strategy. A review of previous studies reveals one or more of the following major shortcomings: (1) no explicit treatment of expectations on the empirical application of investment theory, (2) a lack of studies addressing the effects of farm program policy on the investment in the U.S. farm sector, (3) no justification of the choice of expectations in the agricultural economic studies. The neoclassical investment model is chosen over other approaches in this study because of direct linkages it enables us to establish between fanners' investment decision and government policies, including monetary policy, fiscal policy, and farm program policy. To incorporate farm policy programs on aggregate investment equations, the expected implicit revenue in crop production was developed. To measure the expected implicit revenue, expected price and expected yield for each crop was estimated under alternative expectations specifications. The five major crops considered in this study are wheat, com, sorghum, cotton, and soybeans. Eight expectations specifications for price and yield of each crop are as follows: (1) the naive expectations, (2) the extrapolative expectations, (3) the adaptive expectations, (4) the ARIMA expectations, (5) VAR expectations, (6) the augmented adaptive expectations, (7) the error correction expectations, and (8) the expectations derived from a general equilibrium model emphasizing agriculture. Net investment equations for equipment and structures as a function of expected implicit revenue and expected implicit rental price of capital were estimated. Once the net investment equations were estimated, the performance of the alternative expectations hypothesized in farm sector investment equations for equipment and structures was validated using multi-dimensional criteria. Finally, the estimated equations for the alternative expectations patterns were endogenized into the AG-GEM model to examine the extent to which each hypothesis altered the impact of two topics of current interest: the impact of no chemical use and the impact of high deficits.