measures of performance.
The deviation of a producer from his or her potential level of output, Farrell’s definition
of inefficiency, provides one such measure. Many empirical applications of this notion of
inefficiency exist in agricultural economics, including Bravo-Ureta and Rieger in dairy, and
Akridge and Hertel in agricultural supply cooperatives, among others.1 These applications
involve the estimation or construction of a stochastic production frontier where the random error
about the maximum level of output for a given bundle of inputs includes deviation resulting both
from truly random factors and a measure of idiosyncratic inefficiency. However, this approach
presumes that the producer’s objective is to achieve a maximum level of output using this bundle
of inputs. However, when producers use inputs to both increase output and to lower the variance
of output (Just and Pope), then it is more plausible to define their objective in terms of both the
mean and the variance of output. In this context, efficiency is achieving the optimal tradeoff
between risk and return. If the presence of insurance causes a producer to deviate from this
optimal tradeoff, then this provides more conclusive evidence of moral hazard.
The existence of moral hazard and, to a lesser extent adverse selection, in specialty crop
insurance mean that insurance markets either do not exist, or are extremely thin (Lee, Harwood,
and Somwaru). Many of the major fruit and vegetable crops were only brought into the FCIC
fold by the 1980 Federal Crop Insurance Act, and still suffer from poor participation rates.2
Because of the absence of insurance markets, price and quantity data that are usually used to
1 Much of this literature prior to 1992 is reviewed by Battese.
2 This list includes almonds, cranberries, grapes, onions, peppers, popcorn, and walnuts, or “...any agricultural
commodity grown in the United States (Gardner and Kramer). Insurance for potatoes, tomatoes, peaches and citrus was
available prior to the Act.