all commodities and productive factors in the model. Endogenous prices adjust until factor and
commodity market equilibrium conditions are satisfied. Compared with fixed-price models, CGE
methodology is more consistent with modern economic theory, allows greater flexibility in the
specification of economic behavioral relationships, and generally produces more moderate
estimates of economic impact.
Implementation of a regional CGE model does not require the a priori designation of an
economic base and basic activity is not necessarily limited to a few traditional manufacturing
sectors. In a CGE model, economic change is governed by “supply-side” (e.g. available quantities
of productive goods and services) and trade-related constraints, rather than by backward linkages
transmitted via changes in final demand. Also, since the opportunity to substitute among resource
inputs is subject to diminishing returns in a CGE model, measures such as employment multipliers
are variable depending on the resource input level.
Figure 1 traces the basic steps of a regional CGE modeling effort. First, data are
collected, organized, and reconciled into a benchmark equilibrium data set (the social accounting
matrix).4 Next, behavioral and accounting relationships are specified, and the model parameters
are calibrated given the benchmark data. Finally, the policy change scenarios are introduced, and
counterfactual equilibria representing the situation under the new policy regimes is calculated.
Impacts are estimated by comparing the counterfactual equilibria against the benchmark scenario.
4The data set for the Northeast Oregon CGE was assembled mainly from IMPLAN-
generated regional product accounts (Alward, 1999), state of Oregon tax and expenditure data,
and REIS county income and employment estimates (U.S. Dept. of Commerce, 1999). Timber
harvest estimates (Oregon Department of Forestry) and livestock grazing statistics (Bedell; Bedell
and Stringham, 1994; Hewlett, Cross, and Hart, 1987) were also used to estimate benchmark
resources flow levels.