21.6 Forecasting Economic Impacts for Purposes of the Assessment
315
being modeled, that is, it loses generality; (2) considerable cost and effort is involved in determining
the appropriate constraints; (3) it is difficult to determine appropriate objective function(s) since they
apply to a region rather than an individual decisionmaking unit (i.e., a firm); and (4) if the best technological mix is a major study objective, much effort, time, and cost are required to construct
appropriate cost functions for each alternative technology. For an application of this type of model,
see Chappelle et al. (1982).
An approach suggested by Chappelle (1993, p.
16) may be useful in quantifying nonpriced goods
and services in a comprehensive model:
[E]xcept by measuring transactions required to experience nonpriced goods and services, those [nonpriced]
products would be ignored in the analysis. Therefore,
many amenity and other nonpriced products will continue to be ignored if accounting systems are not expanded to include them as entities. However, these products can be included in the analysis by entering
appropriate constraints into a demand-driven mathematical programming model based on input-output accounts.
The model could be made dynamic by embedding the
mathematical programming model within a simulation
modeling system that could continually update exogenous
variables along their forecasted trends to form a simoptimization model. This type of model could generate scenarios exhibiting consequences of policy changes in the
face of changing conditions.
21.6 Forecasting Economic
Impacts for Purposes of
the Assessment
The following steps must be considered when forecasting comprehensive economic impacts in assessments:
1. Decide whether changes need to be made in the
input-output model to reflect expected changes
over the forecast period (e.g., in technology, in
the household consumption function). If changes
are needed, strategies will have to be devised
for this.
2. Develop forecasts of exogenous variables (e.g.,
final demands of all goods and services) for each
forecast year. These forecasts should be based
on authoritative reports if at all possible.
3. Calculate forecasts of endogenous variables by
running the model. Using an input-output
model, the forecasted endogenous variables
consist of production levels for each sector of
the economy.
4. If at all possible, provide some information on
the likely precision of forecasts. This will be difficult because of the lack of statistical information on many inputs. From the standpoint of
evaluating alternative scenarios, the client
should be able to surmise whether there are any
operationally significant differences between the
scenarios. A major conclusion of the assessment
may be that data are so poor (in quantity or quality) that significant differences between alternative scenarios cannot be adequately appraised.
After a decision has been made regarding
sources of economic impact information, it can be
applied to the assessment. This generally involves
use of input-output multipliers. The objective in
this phase is to quantify all economic effects-direct (which are data to the input-output analysis),
indirect, and induced. In this way, economic impacts likely to occur with changes to the regional
environment can be quantified, which will then permit planners to weigh these impacts against other
types of impacts (e.g., ecological, social, and political). Most impact assessments involve making
forecasts-both of current conditions and likely future conditions if current direction continues.
Forecasting is carried out in three stages: (1) using multipliers to forecast total impacts from direct
impacts, (2) forecasting future production schedules, given expected final demands for all goods
and services, and (3) forecasting future distributions of transactions, given forecasted production
levels.
21.6.1 Use of Multipliers in
the Assessment
Applications of the various multipliers, whether
derived in the assessment or simply taken from
some other study, are fairly straightforward. Since
all multipliers are in the form of total effect divided by the direct effect, for forecasting purposes
all we need is the direct effect, which we must remember is input data for the input-output model
(or implied by the transactions data). Of course,
to use multipliers or any other formulation, it is
necessary that the assumptions associated with input-output modeling can safely be incurred. If the
assumptions can be made, direct effects of
changes in the region can be used with the multipliers to derive estimates of total effects. For example, if it is envisioned that an additional 145
jobs in the lumber industry will be created because
of the increase of regional timber resources, jus-
315
being modeled, that is, it loses generality; (2) considerable cost and effort is involved in determining
the appropriate constraints; (3) it is difficult to determine appropriate objective function(s) since they
apply to a region rather than an individual decisionmaking unit (i.e., a firm); and (4) if the best technological mix is a major study objective, much effort, time, and cost are required to construct
appropriate cost functions for each alternative technology. For an application of this type of model,
see Chappelle et al. (1982).
An approach suggested by Chappelle (1993, p.
16) may be useful in quantifying nonpriced goods
and services in a comprehensive model:
[E]xcept by measuring transactions required to experience nonpriced goods and services, those [nonpriced]
products would be ignored in the analysis. Therefore,
many amenity and other nonpriced products will continue to be ignored if accounting systems are not expanded to include them as entities. However, these products can be included in the analysis by entering
appropriate constraints into a demand-driven mathematical programming model based on input-output accounts.
The model could be made dynamic by embedding the
mathematical programming model within a simulation
modeling system that could continually update exogenous
variables along their forecasted trends to form a simoptimization model. This type of model could generate scenarios exhibiting consequences of policy changes in the
face of changing conditions.
21.6 Forecasting Economic
Impacts for Purposes of
the Assessment
The following steps must be considered when forecasting comprehensive economic impacts in assessments:
1. Decide whether changes need to be made in the
input-output model to reflect expected changes
over the forecast period (e.g., in technology, in
the household consumption function). If changes
are needed, strategies will have to be devised
for this.
2. Develop forecasts of exogenous variables (e.g.,
final demands of all goods and services) for each
forecast year. These forecasts should be based
on authoritative reports if at all possible.
3. Calculate forecasts of endogenous variables by
running the model. Using an input-output
model, the forecasted endogenous variables
consist of production levels for each sector of
the economy.
4. If at all possible, provide some information on
the likely precision of forecasts. This will be difficult because of the lack of statistical information on many inputs. From the standpoint of
evaluating alternative scenarios, the client
should be able to surmise whether there are any
operationally significant differences between the
scenarios. A major conclusion of the assessment
may be that data are so poor (in quantity or quality) that significant differences between alternative scenarios cannot be adequately appraised.
After a decision has been made regarding
sources of economic impact information, it can be
applied to the assessment. This generally involves
use of input-output multipliers. The objective in
this phase is to quantify all economic effects-direct (which are data to the input-output analysis),
indirect, and induced. In this way, economic impacts likely to occur with changes to the regional
environment can be quantified, which will then permit planners to weigh these impacts against other
types of impacts (e.g., ecological, social, and political). Most impact assessments involve making
forecasts-both of current conditions and likely future conditions if current direction continues.
Forecasting is carried out in three stages: (1) using multipliers to forecast total impacts from direct
impacts, (2) forecasting future production schedules, given expected final demands for all goods
and services, and (3) forecasting future distributions of transactions, given forecasted production
levels.
21.6.1 Use of Multipliers in
the Assessment
Applications of the various multipliers, whether
derived in the assessment or simply taken from
some other study, are fairly straightforward. Since
all multipliers are in the form of total effect divided by the direct effect, for forecasting purposes
all we need is the direct effect, which we must remember is input data for the input-output model
(or implied by the transactions data). Of course,
to use multipliers or any other formulation, it is
necessary that the assumptions associated with input-output modeling can safely be incurred. If the
assumptions can be made, direct effects of
changes in the region can be used with the multipliers to derive estimates of total effects. For example, if it is envisioned that an additional 145
jobs in the lumber industry will be created because
of the increase of regional timber resources, jus-