Empirical Calibration of a Least-Cost Conservation Reserve Program

programs are $24.098, $38.489, and $38.547 million. Gains from policy reform are small,
averaging about one penny per acre enrolled. Efforts undertaken to eliminate information
asymmetry could be much more valuable, achieving at most a 37.5 percent reduction in total
program costs relative to the second best.

V. Conclusion

Several recent articles have shown the usefulness of stochastic frontier techniques
for analysis of regulation under asymmetric information. Here, I extend earlier results in
two directions. First, I develop a GMM-based methodology for estimating a stochastic cost
frontier for a profit-maximizing producer. This approach differs from earlier GMM and ML
frontier techniques in that it accommodates multiple equations (in this case a cost equation,
expenditure share equations, and the ratio of revenue to cost) and is robust to arbitrary
cross-equation correlation, heteroskedasticity, and geographic clustering. Further, it is easy
to estimate as it does not require non-linear optimization.

Second, I extend the empirical contract theory literature by using stochastic frontier
analysis to estimate the technology of a heterogeneous sample of producers in an unreg-
ulated sector. Although the econometrician cannot directly observe producer type, this
approach permits estimation of the probability distribution of types in the population. This
information provides the necessary material for empirically specifying an optimal contract
mechanism in the Baron and Myerson framework.

Application of this methodology to a simulated model of a voluntary environmental
program for land retirement in the agricultural sector yields some interesting results. The
simulation permits comparison of an optimal second best mechanism to the optimal full
information mechanism and a sub-optimal Pigouvian subsidy mechanism. In reality, actual

18

<<   16   17   18   19   20   21   22   >>

More intriguing information