5.1 Alternative Approaches for Economic Impact Analysis

   5.1.1    Behavioral vs. Nonbehavioral Models

The analyst can choose among two fundamentally different approaches in conducting an EIA:

• nonbehavioral: engineering and financial analysis of compliance costs, and
• behavioral: analysis of economic responses to compliance requirements, usually in a market setting

In both cases, "engineering" estimates of the costs of compliance for actual or model plants in the regulated industry are the driving factor. With the nonbehavioral approach, the impacts of the regulation are simply assumed to fall on the entities owning the facilities faced with the compliance responsibilities. Analysis takes the form of gauging the severity of impacts, typically using accounting measures of profit and loss. Alternatively, the behavioral approach explicitly recognizes, for example, that owners of the affected facilities are economic agents that can, and presumably will, make adjustments such as changing production rates or altering input mixes that will generally affect the market environment in which they operate. One likely market consequence is a change, typically a rise, in price that will then induce behavioral responses by consumers and producers, which affect resource allocation.

Recognizing the conceptual advantages and the underlying ability to provide the information requirements of an EIA and EA, ISEG has traditionally employed the behavioral economic model approach when feasible under data and other resource constraints. Given this history, the discussion now moves to issues of market model scope and length of run.

   5.1.2    Market Model Scope: Partial vs. General Equilibrium

A strictly partial equilibrium model tracks the effect of a regulatory action in one market; all other possible market interactions are ignored. A strictly general equilibrium model tracks the effects of a regulation in all sectors of the economy; no intersectoral linkages are ignored. In between these two extremes are models that capture a finite set of pre-defined important market linkages, ignoring effects in all other markets. These models are sometimes referred to as "general" equilibrium models (e.g., Just, Hueth and Schmitz, 1982), but they will be referred to here as multimarket partial equilibrium models.

Because the scope of many OAQPS regulatory actions is sector-specific and typically not large enough to substantially affect other sectors of the economy, partial equilibrium models are typically employed in estimating the economic impacts. These partial equilibrium models may be strictly partial (one product market) or may be extended to multiple markets when more than one market is directly affected by a regulation, and/or other related product markets are potentially indirectly affected by the regulation. For example, the EIA conducted for the Pulp and Paper Cluster rule modeled the interactions between the directly affected markets for pulp inputs and the indirectly affected final paper and paperboard products. Although these models may be extended to include multiple markets, they generally do not account for interactions with the entire U.S. economy, as does the general equilibrium approach.

   5.1.3    Length-of-Run Considerations

In developing the partial equilibrium model, the analyst must determine the alternatives available to economic agents in response to regulation and the context within which these choices are made. This is largely dependent on the time horizon for which the analysis is performed. Three benchmark time horizons are considered here: the very short run, the long run, and the intermediate run.

   5.1.3.1    Very Short Run

In the very short run, all factors of production are assumed to be fixed, leaving the directly affected entity with no means to respond. Within a short time horizon, regulated producers are constrained in their ability to adjust inputs or outputs due to contractual, institutional, or other factors. In essence, this is equivalent to the nonbehavioral model described earlier. Therefore, the impacts of the regulation fall entirely on the regulated entity. Producers incur the entire regulatory burden as a one-to-one reduction in their profit. This is referred to in the nomenclature of EIAs as the "full-cost absorption" scenario.

Unfortunately, there is no hard and fast rule for determining what length of time constitutes the very short run. Nonetheless, under most conditions, economic entities have some flexibility to adjust factors of production. And, as indicated above, ignoring those adjustments tends to overstate producers' losses.

   5.1.3.2    Long Run

In the fullness of time, all factors of production are variable and producers can be expected to adjust production plans in response to cost changes imposed by a regulation. Figure 5-1 illustrates a typical, if somewhat simplified, long-run industry supply function. The function is horizontal, indicating that the marginal and average costs of production are constant with respect to output. This horizontal slope reflects the fact that, under long-run constant returns to scale, technology and input prices ultimately determine the market price, not the level of output in the market. Market demand is represented by the standard downward-sloping curve. The market is assumed here to be perfectly competitive; equilibrium is determined by the intersection of the supply and demand curves. The implications of imperfect competition on market equilibrium are discussed below. In this case, the upward parallel shift in the market supply curve represents the regulation's effect on production costs. The shift causes the market price to increase by the full amount of the per-unit control cost (i.e., from P0 to P1). With the quantity demanded sensitive to price, the increase in market price leads to a reduction in output in the new with-regulation equilibrium (i.e., Q0 to Q1). As a result, consumers incur

Taken together, impacts modeled under the long-run/full-cost pass-through scenario reveal an important point. Namely, under fairly general economic conditions, a regulation's impact on producers is transitory. Ultimately, the costs are passed on to consumers in the form of higher prices. However, this should not be used as a justification to completely dismiss producer impacts in an EIA. For one, the long run may cover the time taken to retire all of today's capital vintage--perhaps decades. Therefore, transitory impacts could be fairly protracted. Given the call to discount the costs of a policy (see Section 8), transitory impacts could dominate long-run impacts in terms of present value. Moreover, the statutes and EOs referenced in Section 2 implicitly call for an assessment of impacts on current producers and workers; thus, a purely long-run approach is moot for addressing transitory, but important, concerns such as facility closures, capital displacement, and worker dislocation. Given the previously referenced limits of a very short run perspective, the analyst should ideally consider some intermediate case between the very short and long run to gauge economic impacts.

   5.1.3.3    "Intermediate" Run

The "intermediate" run can best be defined by what it is not. It is not the very short run and it is not the long run. In the intermediate-run, some factors are fixed; some are variable. The existence of fixed production factors generally leads to diminishing returns to those fixed factors. This typically manifests itself in the form of a marginal cost (supply) function that rises with the output rate, as shown in Figure 5-2.

Again, the regulation causes an upward shift in the supply function. The lack of resource mobility may cause producers to suffer profit (producer surplus) losses in the face of regulation; however, producers are able to pass through the associated costs to consumers to the extent the market will allow. As shown, in this case, the market-clearing process

   5.1.4    Static vs. Dynamic Models

At its core, the partial equilibrium model simulates a single period market outcome without and with regulation. Each solution of the model is, therefore, static in nature. The model can, in principle, be extended to incorporate temporal dynamics such as exit and entry over time and adoption of new technology. Then, time paths of market equilibria without and with the regulation can be simulated to track regulatory effects over time. This approach has some obvious attractions but is typically difficult to implement. Dynamic models of firm entry, exit, and technological change are difficult to estimate and generally quite sensitive to model specification and parameter values. Therefore, the common practice is to stimulate regulatory effects in a static setting, using a representative year as the basis for estimating annual regulatory impacts. More information on selecting the analytical baseline and time period of analysis is presented in Section 8.

   5.1.5    Standard Approach for ISEG Industry-Level Analyses

The remainder of this section focuses on EIAs conducted for regulations of existing sources in a single regulated industry. This single-industry focus characterizes rulemakings such as the imposition of a MACT standard for a NESHAP. The standard approach for addressing these problems is the development and application of partial equilibrium/ intermediate-run/static (PEIS) models to assess impacts in the potentially affected market(s). The role of new sources and the impacts of new source performance standards, while not part of this document's focus, can easily be accommodated within the methods described here. However, that is not the case for industry-specific regulations that have costs large enough to substantially affect other sectors or regulations, such as NAAQS standards, that cut extensively across industries. Those extensive regulations need more serious consideration of general equilibrium and macroeconomic models to appropriately gauge economic impacts. Those modeling topics are beyond the scope of the current document but may be included in future revised editions. Thus, the section continues by describing the conceptual underpinnings of the PEIS modeling approach.