UW-Madison UW-Madison

Staff Paper No. 413 - Abstract

Recovering a Soil Quality Measure from Crop Trials Data: A Dynamic Econometric Method

Kwansoo Kim
Bradford L. Barham [barham@mailplus.wisc.edu]
Ian Coxhead [coxhead@wisc.edu]

Staff Paper No. 413, August 1997, 26p.


The definition and measurement of soil quality is a challenge faced by all researchers concerned with the environmental consequences of agricultural land use. Conceptual difficulties, lack of suitable data and the need for a methodology capable of capturing long-term changes are all serious constraints. In this paper we develop a new approach to measuring soil quality from longitudinal data on yields and certain farm management practices. These data are widely available from crop trials experiments. The methodological core is a dynamic model that, in its most general form, is capable of linking the evolution of soil quality, as the key state variable, to farm management practices and important observed outcomes such as yields. Implementation makes use of a non-linear time-series estimation procedure that exploits both the recursive properties of the dynamic model and the availability of longitudinal crop trials data. Estimation results shed light on the hypothesized evolution of soil quality in the short and long run under different crop rotations and fertilizer applications, and reveal some interesting dynamic relationships. As such, our measure has potential applications in the design of public policy and in the valuation of agricultural land.
Last updated on Wed, Jun 21, 2006 2:37pm