252
strained when applied to environmental analysis,
particularly those of "fixed production and emission coefficients, lack of substitutability, fixed
supply channels, fixed tastes and mechanistic behavior. . . . The linearity assumption, which is a
strong one regarding economic commodities and
processes, is considerably more severe for the ecologic commodities. Interactions between different
pollutants and between pollutants and the environment at different concentrations and at different
levels of economic activity are typically nonlinear."
They further point out that these problems can be
overcome, but only with side calculations and modifications of coefficients used in the 1-0 model. A
rather notable hurdle in applying such a model is
to analytically describe ambient environmental
quality. This requires the use of diffusion processes
and effectively modifies the scheme of ecological
regions, generally requiring that they be very small.
There are a number of approaches to including
environmental sectors with the economic sectors in
a single model, including the (1) Cumberland
model, (2) Isard-Daly models, (3) Leontief model,
and (4) Victor (1972) model. Space does not permit describing these models here, but the reader
may consult Richardson (1972) or Miller and Blair
(1985) for details.
Development of data sets to input these models
as part of an ecological assessment is almost totally
out of the question for both time and funding considerations. To my knowledge, not one of these approaches has been attempted by assessment teams,
and for good reason. Implementation of any of
these models would constitute a large-scale research project in itself. However, these models are
important in that they provide clarification as to
how the joint ecological-economic system operates. The stumbling block in all cases is the lack of
an integrated set of ecological accounts showing
flows into and out of the environment. When an
analyst wishes to include ecological sectors in a
model, it is necessary to distill a mass of case study
results into a coherent, compatible data set. For a
very interesting application of this approach, see
Isard et al. (1972).
17.8 Integration of Economic
Impact Estimates with Other
Social Impact Estimates and
Ecological Impact Estimates
There are some variables that link economic impacts with social impacts, for example, employment and, by implication, population. Population
Economic Linkages to Natural Resources
(including immigration) and per capita consumption forecasts are needed to calculate future demands for all goods and services. These exogenous
variables are used in 1-0 analysis to forecast future output of all goods and services, both final and
intermediate sales. Also, a number of social indicators are quantified in economic accounts in monetary terms, including unemployment compensation, retirement benefits, medical payments, and
welfare payments. Social impacts of changes in
these economic indicators should be explored in the
social impact section of the assessment. Impacts in
monetary terms have to be translated into values
that measure human suffering, well-being, and
promise. Also, a cumulation of values that provides
a measure of community well-being is necessary in
the social impact analysis.
Linkages between economic and ecological impacts are less clearly defined. The difficulty, as
noted previously, is that physical units of resources,
or even resource constraints, do not appear anywhere in the typical 1-0 analysis. Rather, resources
have to be considered in side calculations. For example, if we calculate output forecasts of resources,
using sectors in an 1-0 analysis, it is a rather simple process to check if these production levels are
feasible, given expected regional resource stocks.
However, often it is very difficult to determine the
amounts of each resource truly available at prices
implicit in the 1-0 model. There are ways to modify price assumptions, but these are difficult, as well
as somewhat arbitrary (see Miller and Blair, 1985,
pp. 351-357). Also, outflows of pollutants implied
by estimated production levels forecasted by the
1-0 model may be checked against available environmental assimilative capacity data, although such
data are generally both limited and fragmentary.
Feedback of environmental quantity and quality
likely to constrain future economic production is
particularly difficult to measure. Ideally, resource
stocks at every year in the forecast period should
be measured in terms of ecological impacts of the
previous year's production, but this degree of finetuning is not feasible given available data resources. About the best that can be done now is to
temper future production schedules with expected
changes in natural resources.
It is important that economic impact analysis be
viewed from the standpoint of political economics
and social analysis, rather than from a narrow technical viewpoint. The point of the analysis is to measure the extent to which natural resources influence
human welfare, in both the short and long term.
Measurement scales noted above (i.e., output, income, and employment) have broad social implications. More output, income, and employment are
strained when applied to environmental analysis,
particularly those of "fixed production and emission coefficients, lack of substitutability, fixed
supply channels, fixed tastes and mechanistic behavior. . . . The linearity assumption, which is a
strong one regarding economic commodities and
processes, is considerably more severe for the ecologic commodities. Interactions between different
pollutants and between pollutants and the environment at different concentrations and at different
levels of economic activity are typically nonlinear."
They further point out that these problems can be
overcome, but only with side calculations and modifications of coefficients used in the 1-0 model. A
rather notable hurdle in applying such a model is
to analytically describe ambient environmental
quality. This requires the use of diffusion processes
and effectively modifies the scheme of ecological
regions, generally requiring that they be very small.
There are a number of approaches to including
environmental sectors with the economic sectors in
a single model, including the (1) Cumberland
model, (2) Isard-Daly models, (3) Leontief model,
and (4) Victor (1972) model. Space does not permit describing these models here, but the reader
may consult Richardson (1972) or Miller and Blair
(1985) for details.
Development of data sets to input these models
as part of an ecological assessment is almost totally
out of the question for both time and funding considerations. To my knowledge, not one of these approaches has been attempted by assessment teams,
and for good reason. Implementation of any of
these models would constitute a large-scale research project in itself. However, these models are
important in that they provide clarification as to
how the joint ecological-economic system operates. The stumbling block in all cases is the lack of
an integrated set of ecological accounts showing
flows into and out of the environment. When an
analyst wishes to include ecological sectors in a
model, it is necessary to distill a mass of case study
results into a coherent, compatible data set. For a
very interesting application of this approach, see
Isard et al. (1972).
17.8 Integration of Economic
Impact Estimates with Other
Social Impact Estimates and
Ecological Impact Estimates
There are some variables that link economic impacts with social impacts, for example, employment and, by implication, population. Population
Economic Linkages to Natural Resources
(including immigration) and per capita consumption forecasts are needed to calculate future demands for all goods and services. These exogenous
variables are used in 1-0 analysis to forecast future output of all goods and services, both final and
intermediate sales. Also, a number of social indicators are quantified in economic accounts in monetary terms, including unemployment compensation, retirement benefits, medical payments, and
welfare payments. Social impacts of changes in
these economic indicators should be explored in the
social impact section of the assessment. Impacts in
monetary terms have to be translated into values
that measure human suffering, well-being, and
promise. Also, a cumulation of values that provides
a measure of community well-being is necessary in
the social impact analysis.
Linkages between economic and ecological impacts are less clearly defined. The difficulty, as
noted previously, is that physical units of resources,
or even resource constraints, do not appear anywhere in the typical 1-0 analysis. Rather, resources
have to be considered in side calculations. For example, if we calculate output forecasts of resources,
using sectors in an 1-0 analysis, it is a rather simple process to check if these production levels are
feasible, given expected regional resource stocks.
However, often it is very difficult to determine the
amounts of each resource truly available at prices
implicit in the 1-0 model. There are ways to modify price assumptions, but these are difficult, as well
as somewhat arbitrary (see Miller and Blair, 1985,
pp. 351-357). Also, outflows of pollutants implied
by estimated production levels forecasted by the
1-0 model may be checked against available environmental assimilative capacity data, although such
data are generally both limited and fragmentary.
Feedback of environmental quantity and quality
likely to constrain future economic production is
particularly difficult to measure. Ideally, resource
stocks at every year in the forecast period should
be measured in terms of ecological impacts of the
previous year's production, but this degree of finetuning is not feasible given available data resources. About the best that can be done now is to
temper future production schedules with expected
changes in natural resources.
It is important that economic impact analysis be
viewed from the standpoint of political economics
and social analysis, rather than from a narrow technical viewpoint. The point of the analysis is to measure the extent to which natural resources influence
human welfare, in both the short and long term.
Measurement scales noted above (i.e., output, income, and employment) have broad social implications. More output, income, and employment are