"Estimating the Uncertainty of Policy Effects in Nonlinear Models," Econometrica, 1980.Paper: pdf file
A method is described in this paper for estimating, by means of stochastic simulation, the asymptotic variances of multipliers for nonlinear models. It is used to estimate the uncertainty of the results of eight policy experiments for a particular model.
This paper presents a method for estimating the uncertainty of estimated policy effects in econometric models. I have used this method to estimate the uncertainty of policy effects in the US model. One application is in 1981#2. The method is also discussed in Chapter 9, Sections 9.3 and 9.4 in 1984#2.
An important change in the method is discussed in Chapter 10, Section 10.2, in 1994#2. The original method required that a stochastic simulation be done within a stochastic simulation, and in 1984 S.G. Hall pointed out in private correspondence that this is not required. The method that is discussed in Section 10.2 in 1994#2 incorporates Hall's suggestion, and so this section is a replacement for the earlier discussions. The newer method is used in Chapter 11, Section 11.3, in 1994#2.
The method (i.e., the newer method) is programmed into the Fair-Parke program, where the command is STOMULT.