| The purpose of this analysis is to examine the technical efficiency of Illinois grain farms and the factors which influence its measurement. The effects of applying a parametric versus a nonparametric estimation technique, modeling the frontier as deterministic or stochastic and using a single aggregate-output versus a multiple output measure are examined using a homogeneous sample of Illinois grain farms over three time periods. In addition, the effects of government farm programs and farm size are assessed.;Five different frontier methods are used. The models are the nonparametric single-output, nonparametric multiple output, deterministic half-normal, stochastic half-normal and the 95th regression quantile.;The nonparametric models result in higher measures of efficiency than the parametric models and their efficiency distributions are different. Furthermore, efficiency ratios are not highly correlated between the two types of procedures. Hence, farms which are classified in one analyses as highly efficient may not be when another procedure is used.;Differences also exist between the parametric stochastic and deterministic models. The distribution of the efficiency estimates of the stochastic half-normal model is very narrow. The deterministic and quantile regression models result in more disperse and plausible efficiency distributions.;The effect of using a single- versus a multiple-output frontier has a small influence on measured efficiency. Efficiency ratios and their distributions are very similar. It appears that studies which examine Illinois grain farm efficiency using aggregate output data are not likely to be biased due to output aggregation.;The results also indicate that increases in the level of the set-aside requirement of USDA corn programs did not affect to any degree the efficiency of the farms. This result is fairly consistent across size class and frontier estimation procedure.;Small farms are generally found to be significantly less efficient than farms of medium or large size. Medium-sized farms are in general not found to be significantly less efficient than large farms. This result indicates that the existing trend in Illinois agriculture towards larger farms may not necessarily be accompanied by increases in efficiency. |
