composition. First, the use of COLS approach,
which categorizes all deviations from the frontier as
inefficiency, may be too sensitive to outliers. Even
after eliminating several observations that seemed
dramatically different from the sample and using
various temporal aggregates, φe resarch found rela-
tively large levels of inefficiency. Second, the speci-
fication of the RHF appears to be imposing rather
high levels of decreasing returns to scale for the
larger farms. This occurs because of the relative
constancy of the factor shares and because the scale
measure varies inversely with output. For the larger
firms, higher levels of scale inefficiency tend to
offset increases in pure technical efficiency. For the
six-year average data, total farm efficiency initially
rises but does not increase significantly for farms
larger than 400 acres. Also, for several individual
years, total efficiency declines for large size opera-
tions.
Clearly, additional research is needed to identify
under what circumstances particular methods
should be employed to measure farm efficiency.
Perhaps more accurate measurements of the level of
inefficiency should involve the use of stochastic
frontier procedures that permit deviations from the
frontier to be due to random events as well as to
technical inefficiency. Also, more care needs to be
taken in the applications of specific functional
forms. The use of the ray-homothetic function in the
literature has not been based on statistical criteria.
Instead, it has been used because it permits the
optimal size of farm to vary with factor intensity, a
unique characteristic of the function. For those tech-
nologies and samples where the factor intensities do
not vary appreciably across firms, perhaps more
emphasis needs to be placed on statistically deter-
mining the “best” functional form prior to generat-
ing measures of efficiency. This is especially
significant in an environment where returns to scale
are hypothesized to be important determinants of
efficiency and the distribution of farms. Even when
total efficiency is accurately assessed, errors in the
measurement of the decomposition can lead to inap-
propriate recommendations, strategies, and policies
to ameliorate its presence. Finally, direct compari-
sons with efficiency measures from procedures that
incorporate multiple output technologies may pro-
vide additional insight into the assessment of firm
behavior in the agricultural sector.
REFERENCES
Aly, Hassan Y., K. Belbase, R. Grabowski and S. Kraft. “The Technical Efficiency of Illinois Grain Farms:
An Application of a Ray-HomotheticProduction Function.” So. J. Agr. Econ., 19(1987):69-78.
Bagi, F.S. and C.J. Huang. “Estimating Production Technical Efficiency for Individual Farms in Tennessee.”
Can. J. Agr. Econ., 31(1983):249-256.
Battese, G.E. and T. J. Coelli. “Prediction of Firm-Level Technical Efficiencies with a Generalized Frontier
Production Function and Panel Data.” J. Econometrics., 38(1988):387-399.
Bravo-Ureta, B.E. and L. Rieger. “Alternative Production Frontier Methodologies and Dairy Farm Effi-
ciency.” J. Agr. Econ., 41(1990):215-226.
Bymes, P., R. Fare, S. Grosskopf and S. Kraft. “Technical Efficiency and Size: The Case of Illinois Grain
Farms.” Euro. R Agr. Econ., 14(1987):367-381.
El-Osta, H.S., R.A. Pelly and G.W. Whittaker. “The Technical Efficiency of Com Farms in the Com Belt:
A Methodological Note.” Paper Presented at the 1990 Joint AAEA∕WAEA∕CAEFMS∕AERE Annual
Meeting, Vancouver, B.C.
Elyasiani, E. and S. Mehdian. “Efficiency in the Commercial Banking Industry, A Production Frontier
Approach.” AppliedEcon., 22(1990):539-551.
Fare, R., S. Grosskopf and H. Lee. “A Nonparametric Approach to Expenditure Constrained Profit Maximi-
zation.” Am. J. Agr. Econ., 72(1990):574-581.
Fare, R., L. Jansson and C.A.K. Lovell. “Modelling Scale Economies with Ray-Homothetic Production
Functions.” Rev. Econ. Stat., 67(1985):624-629.
Fare, R. and B.J. Yoon. “Returns to Scale in U.S. Surface Mining of Coal.” Resources and Energy,
7(1985):341-352.
Grabowski, R. “Curve Fitting Versus Hypothesis Testing: A Comment.” Applied Econ., 22(1990):427-429.
Grabowski, R. and K. Belbase. “An Analysis of Optimal Scale and Factor Intensity in Nepalese Agriculture:
An Application of a Ray-Homothetic Production Function.” Applied Econ., 18(1986):1051-1063.
120