Input-Output Analysis

Jon D. Erickson and Melinda Kane

Introduction
Ecological economics is built … Structural foundations …
Technologies as recipes …
Important consideration of substitutability of inputs …

These structural considerations of ecological economics made the methodological

approaches of input-output analysis a good fit … In fact, in a seminal article entitled "On
Economics as a Life Science," Herman Daly sketched a view … drawing on the empirical
analyses of Leontief, Stone, and others characterized by … Scholars who today identify
with ecological economics, including Hannon (), Herendeen (), Duchin (), Victor ()
worked in the 1970s and 80s to … This work was some of the earliest to lay out
empirical foundations to the central premise of ecological economics of an economy
embedded in ecosystems, fundamentally dependent on environmental source and sink.
Contemporary research presented and published in ecological economics on X, X, and X
continue in this tradition.
This chapter builds the methodological foundation of this structural approach to
economic analysis. The basic input-output model is presented, followed by a number of
extensions relevant to ecological economics. These include the development of social
accounts, hybrid environmental input-output tables, total flows analysis, and the supply-
side model. The chapter concludes with a short assessment of the family of structural
economic models as a renewed platform for a methodological approach to ecological
economics.

Standard Input-Output Model

Nobel laureate Wassily Leontief (1936) is credited with having developed input-
output analysis in the 1930s, a time when economists were researching the root causes of
the Great Depression and concerned with supply bottlenecks as the nation prepared to
enter World War II. However, the examination of interdependence between economic
sectors is much older than Leontief’s work, dating back to Quesnay’s Tableau
Economique in 1758, while the notion of production coefficients, a familiar feature of the
input-output system, dates back to Léon Walras’ work in 1874. Today virtually all
developed countries maintain I-O accounts to complement their national income
accounts, while the United Nations has supported a standardized system of economic
accounts in order to facilitate the use of input-output analysis as a planning tool for less
developed countries.
The most basic input-output framework is the static model, which shows the
flows between industries during a one-year time period. The core of the model is a set of
linear relationships describing each industry both as a producer and consumer of goods
and services. Industries purchase inputs from other industries (inter-industry
transactions), and also from other factors of production including payments to labor,
government taxes, interest payments on capital, profits, and adjustments to inventories.
The I-O system also shows final demand for an industry’s output, generally consisting of
household consumption, government demand, private investment and foreign demand
(the components of gross domestic product).
The heart of input-output modeling is the transactions table showing the purchase
and sale of intermediate inputs between industries. Each cell of the matrix represents
both a sale (reading across rows) and a purchase (reading down columns). Figure 1
illustrates the basic form of the static IO model. Total economic output equals total
input, both the sum of inter-industry transactions (Z) and final demand (Y), as well as the
sum of Z and value added (W).

Final Total
Intermediate + Demand = Output
Production (Z) (Y) (X)

+
Value Added (W)

=
Total Input (X)

Figure 1. Basic Input-Output System

Much of the literature on input-output analysis is described through linear algebra,

so a brief mathematical description of the core model is helpful in preparation of further
exploration. With n industries and aggregated final demand and value-added, Figure 1
can be organized in a set of linear equations in matrix form. Assuming an index i for
selling sector (row), and index j for purchasing sector (column) each inter-industry
transaction is denoted as zij. Total output of any particular industry sector is then defined
as the sum of all sales to both intermediate (industry) and final consumers:
X i  z i1  z i 2    z in  Yi (1)

or, in more general terms,

n
X i   z ij  Yi (2)
j 1

In order to analyze the impacts of changes in a particular sector on the larger

economy, information on the ratio of various inputs to the level of outputs is needed.
These ratios, denoted aij, are the technical coefficients of the model and are calculated as:
z ij
aij  (3)
X j

For example, if in producing $1000 of output for the livestock sector, ranchers spent $300
on cattle feed, the technical coefficient would be $300/$1000 or 0.3. The coefficient is
interpreted as the amount of input i required for the output of $1 of sector j. These
technical coefficients are assumed fixed in nature (i.e. constant returns to scale) so that if
livestock output were tripled, purchases of feed by ranchers would also triple. Also, it is
assumed that inputs are used in fixed proportions (i.e. strictly complimentary), such that
if more feed was purchased but additional labor was not available, the amount of
livestock output could not be increased. Only when all inputs are increased in constant
proportions can total output increase.
Critics of the input-output approach most often cite the limitations of this fixed
proportion assumption. However, it is the assumption of complimentarity of inputs
(particularly over short-term planning horizons) that allows for very detailed description
of economic production and empirical study of technological and resource constraints.
Compared to the popular Cobb-Douglas production function, typically with only a single
labor and single capital input, an IO approach is able show detailed flows within a large
production system.
If the assumption of fixed technical coefficients is accepted, equation 1 can be
rewritten substituting for each zij:

X i  ai1 X 1  ai 2 X 2    ain X n  Yi (4)

A system of n such equations, one for each sector, then fully describes the inter-industry
transactions within the economy and shows the technical coefficients that dictate how
much of each sector’s output sector j will buy to produce each dollar’s worth of its own
output.
By further manipulating each equation, grouping the Xi’s and leaving final
demand on the right-hand side, the following form is reached:

1  a11 X 1  a12 X 2    a1i X i    a1n X n  Y1

 a 21 X 1  1  a 22 X 2    a 2i X i    a 2 n X n  Y2

(5)
 ai1 X 1  ai 2 X 2    1  aii X i    ain X n  Yi

 a n1 X 1  a n 2 X 2    a ni X i    1  a nn X n  Yn

The entire system can be rewritten in matrix form as

I  AX Y (6)

where

 a11 a12  a1i  a1n   X1  Y1 

a a 22  a 2i  a2n   X  Y 
A   21 X   2 Y   2
        
     
a n1 a n 2  a ni  a nn   Xn Yn 
, ,
and I is the n x n identity matrix. The input-output system of equations is most often
represented by further manipulating the matrix notation, pre-multiplying each side of the
equation by (I-A)-1 to show the entire output of the economy as a function of final
demands, resulting in:

X  I  A Y
1
(7)

where (I-A)-1 is known as the Leontief inverse.

This inverse captures both the direct and indirect requirements of production. For
example, in producing one dollar’s worth of grain for final consumption, the agricultural
sector requires inputs from both itself (perhaps in the form of seed and other inputs) and
from other sectors (perhaps a tractor or fencing). These are the direct requirements. But
in producing those direct requirements, each sector uses inputs from other sectors; the
tractor producer needs steel and tools and fuel, while the seed company needs sorting
equipment and labor among other things. These are the indirect requirements for
producing grain. In the standard model, the inverse captures both direct and indirect
requirements in calculating the total amount of production necessary to fulfill any extra
final demand.
National databases are constructed with industry survey data that characterize
these relationships. In the U.S., the Bureau of Economic Analysis (BEA) publishes
national benchmark tables every five years using the North American Industry
Classification System (NAICS). Private companies such as IMPLAN
(www.implan.com) and REMI (www.remi.com) assemble data into software packages
used for national, state, and regional analysis (down to the U.S. zip code level).
Disaggregated models typically include over 500 sectors, the majority of which are part
of manufacturing, but data can be built in standard one and two-digit aggregations, or
custom aggregation to suit the questions at hand. Since each industry may produce more
than one commodity, data is published and incorporated into models as both make and
use tables.

Extensions of Input-Output Analysis

Descriptions of the standard Leontief model and it's many extensions are covered
in texts such as Miller and Blair (1985), Miller et al. (1989), and Rose and Miernyk
(1989). Extensions have dealt with the both the limits of the strict technological
assumptions and applications to broader research questions, including the foundations of
dynamic economic approaches such as computable general equilibrium (CGE) models
(Rose 1995). A few extensions with significance to central research questions in
ecological economics are summarized here.

Social accounting matrix

Traditional input-output analysis focused on the structure of production, the
matrix in the upper left corner of Figure 1, but did not provide much detail for final
demand, described only in terms of the four major components of household
consumption, investment, government consumption, and trade. While the IO table can
describe production in terms hundreds of industries, in the standard model households are
represented by a single sector in final demand, even though they account for most of final
demand and are the major driving force in the economy. This restricted treatment of
households limits the ability of the IO model to address such issues as income
distribution or more specific impact analysis of changing patterns of household spending.
The need for a more detailed treatment of households in order to address major
economic policy questions led researchers, beginning with the work of Richard Stone in
the 1960s, to expand the IO system to a fuller social accounting matrix, or SAM. Figure
2 provides a simplified version of a SAM. As consumers, households are disaggregated
from one column into many according to criteria relevant to the policy question at hand,
such as income level or education status. As suppliers of labor and capital, the value-
added block is also disaggregated, often by different wage groups or occupations. The
connections between the source of income (household supply of labor and capital) and
consumption (household purchases in final demand) is represented by a new matrix of the
flows between final demand and value-added.

Final Demand

Intermediate + Low Medium High Other = Total

Production Income Income Income (Gov.’t, Output
HH HH HH Exports)

+
Low Wage Earners
Value Medium Wage Earners Flows between Final Demand
Added High Wage Earners and Valued Added
Other (Capital, Imports)
=
Total
Input

Figure 2. Social accounting matrix

In the typical SAM model, when industries purchase new labor, labor then spends
new income in the economy, contributing to new rounds of inputs and outputs. By
including households in the endogeneous part of the model, an "induced" component to
economic multipliers is added to the direct and indirect impacts of new demand. For a
full overview, history, and application of SAM models to planning, see the World Bank
publication by Pyatt and Round (1985).

Regional models
Early input-output studies reflected concern with the national economy, however
the framework has been extended to apply to sub-national regions and to deal with the
inherent openness and trade between regions. In absence of regional tables, much of the
research has involved various approaches to regionalize national accounts. Miller and
Blair (1985) summarize non-survey techniques that have estimated regional input
coefficients through the adjustment of national technical coefficients using published
information on regional employment, income, or output by industry. For example,
regional supply percentages can be calculated as a ratio of the locally produced available
amount of an input to the total amount of the input available in the region, to transform
the national technical coefficients matrix, A, to a matrix of regional input coefficients.
Techniques have also developed to estimate location quotients, supply-demand pooling
methods, and regional purchase coefficients.
The importance of interregional feedbacks for open regional models has also led
to development of interregional and multiregional models. To overcome severe data
constraints in these models, various empirical research has incorporated non-survey and
partial survey (or hybrid) methods to account for changes in spatial coverage (e.g.
regionalizing national tables) or temporal situations (e.g. accounting for changing
technologies or the introduction of completely new products). A number of commercial
software packages use various incarnations of regionalization techniques, described in
part by Brucker et al. (1990).
Isard (1951) and Leontief (1951) are credited with the seminal work in
development of regional models. Jensen (1990) offers an excellent discussion of more
recent progress in applying an input-output approach to regional economies.

Environmental input-output analysis

The policy questions of the 1960s and 1970s presented a growing need for
analysis that linked economic and environmental effects, and the basic IO/SAM
framework was expanded to incorporate environmental and natural resource accounts
(Victor 1972). This additional set of information includes resource use, such as land of
different kinds or location, and various kinds of environmental pollution associated with
production by each industry and final use designed to fit with the IO/SAM framework
(United Nations 1993; Lange 1998, 1999)
The first environmental models were framed as interregional models based on the
notion that the two “regions” were the human economy and the natural environment.
Isard (1967) and Daly (1968) both developed fully integrated models that encompassed
inter-sectoral flows within and between the economy and the regional environment.
However, data limitations have led to more modest efforts at showing the links between
the economy and certain environmental media by adding vectors to account for the by-
products of pollution and additional input requirements from natural resources. These
vectors are incorporated into the transactions matrix and have their multipliers in the
corresponding inverse (i.e. pollution or natural resource multipliers) to show the total
pollution output or resource requirements associated with a given level of final demand.
As one example, Kane (2003) reviewed research on water quantity and quality
illustrative of these integrated approaches, for example through the development of water
use coefficients to calculate direct, indirect, and induced demands on the resource; total
outputs of waste products using pollution coefficients; and interregional trade in water.
These integrated approaches have also expanded beyond analysis of dollar flows alone.
For example, IO approaches have also been developed with physical measures of flows,
particularly in analysis of economic dependence on energy inputs. This approach was
pioneered by Herendeen (1974) and Hannon and Blazeck (1984) in the Energy Research
group at the University of Illinois. Also of interest is the application of decomposition
analysis using a BTU-based IO table (Gowdy and Miller 1987; Ang 1995; Casler and
Blair 1997; Rose 1999).
Supply-side model
In the traditional Leontief model, demand is viewed as the limiting factor and the
source of any impacts on the economy. However, a mirror image of this demand-driven
process must also hold on the supply-side. Often called a Ghoshian model after Ghosh
(1958), the source of economic change can also derive from changes in value-added
inputs (W). Here W becomes the limiting factor and the source of economic change, and
Y is now the sink. Technical coefficients are again fixed, but are derived by dividing
output into rows of the transaction matrix, rather than down columns as in the Leontief
system. In contrast to the fixed input requirements in the demand-driven Leontief
system, the supply-driven Ghoshian system assumes fixed output coefficients.
While the Leontief model is most useful when describing a short-run economic
system with idle resources and thus very elastic factor-supply curves, the Ghoshian
system is plausible under conditions of resource scarcity and very inelastic factor-supply
curves, as in the case of supply-side shocks. For instance, Giarratani (1976) estimated a
Ghoshian input-output model to investigate supply linkages associated with U.S. energy
production and the impact of oil supply allocation schemes on the U.S. economy. While
energy shocks are often estimated in dollars, the model can be easily transformed into
energy units such as quads or BTUs by incorporating sector and fuel specific energy
intensities.

Future Directions
Heyday of IO research related to energy and environmental issues was in the
1970s and 80s … much of this work in the United States was defunded into the 1980s
and 1990s, however the United Nations made significant progress on creation of
integrated accounts … Today IO is more on incorporation of environmental data into the
structure of IO. For example, Carnegie Mellon … While the application of IO was no
longer in vogue in recent years, the work of ecological economics do provide some
exceptions, for example … European examples from ISEE conference …
One extension of the IO model that has received far less attention connects to the
roots of the original call from Daly (1968) in his seminal article "On Economics as a Life
Science" …

The application of the IO model with physical units highlights distinct parallels
between structural analysis of ecological and economic systems, an early theme within
IO research dating to Isard (1967) and Daly (1968), and today an important direction for
ecological economics. For example, the assumption of no net storage in ecological
systems parallels the assumption of no net investment or no net change in inventories in
economic systems. Also, energy transfers in ecosystems are modeled in much the same
way as inter-industry transfers in economies. However, the straightforward application
of the economic principles of IO to ecological questions has also been faced with several
problems, since ecology has no analogous concept to final demand. As Szyrmer (1984)
notes:

… in ecology the 'useful' intrasystem flows are regarded as much more important
than the 'useless' respiration outputs, while in economic systems the main
 emphasis is placed upon the final demands rather than upon the internal
transactions, the latter being valuable only insofar as they provide directly or
indirectly the goods and services for final consumption (p. 40).

Thus, an ecologist may place greater importance on the total flows (direct and indirect)
from one system element to another (e.g. between sun and zebra in the African savannah
ecosystems) or between one system element and the system as a whole (e.g. the impact of
lion extinction on the African savannah ecosystem).
With these corollaries in mind, Szyrmer developed total flow analysis to quantify
the relationship between any two system elements, or between any one element and the
system as a whole (Szyrmer 1984, 1992; Szyrmer and Olanowicz 1987). A central
question from a total flow perspective is: what happens to element i activity or to total
system activity if element j is decoupled from the system? In an economic context, what
happens to other sectors in a regional economy and to that regional economy as a whole
if one of its sectors begins importing all of its input requirements? Such a “hypothetical
extraction” technique can be used to determine the dependence of a particular sector or
the region as a whole on a given sector. This technique is equivalent to zeroing all the
entries in a given sector’s column, re-computing the Leontief inverse and total output,
and subtracting the new total output figures from the previous case.
In terms of impact analysis, total flow can be viewed as an upper bound for the
perturbation that a system element may exert upon itself, upon other elements, or upon
the system as a whole. When considered from the perspective of the entire system, this
impact can be expressed as a scalar measure of sectoral significance. In other words, if all
sector inputs from a particular sector are externalized, estimates can be made for the
decline of total system activity. Beyond the traditional Leontief output multiplier, total
flow provides many additional ways to rank sectors, depending on one’s perspective of
what constitutes a “key” sector characteristic. Sectors with high transit flows may be
“key” in that they function to maintain the connections between other sectors, while those
with large total flows might be considered “key” by virtue of the sheer magnitude of
flows or the share of total throughput that are embodied in it as a sector. Analysis of total
output dependence is important in economic systems, especially in highly specialized
production structures, for example when commodity production is completely dependent
on processing sectors (e.g. crude oil producers and refineries).

Concluding Remarks
 If the methodological literature reviewed here seems a bit dated, it's only the
frontiers of IO has been less about methods and more about data in recent years.
Also, the world of CGE came into favor by many economists, leaving the
structural approaches behind.
 Much momentum in the 1970s and 1980s to integrated approaches was lost with
political changes, particularly in the U.S. Development of energy tables dropped.
 However, international work continued to develop. Today set of ENRA at
international levels, use of IO in life-cycle approaches (for example, Carnegie
Mellon), plus recent European initiatives. There's also the incorporation of spatial
data, for example, with the application of land-based IO models (Hubacek).
  Expansion to new questions, with better integration of economic and ecological
relationships

Key Further Reading

Daly, H. (1968). “On Economics as a Life Science.” The Journal of Political Economy
76(3): 392-406.
Lange, G. (1999). "How to make Progress toward Integrating Biophysical and Economic
Assessments." Ecological Economics 29(1):29-32.
Miller, R. E. and P. D. Blair (1985). Input-Output Analysis: Foundations and Extensions.
Englewood Cliffs, New Jersey: Prentice-Hall, Inc.
Miller, R. E., Polenske, K. R. and A.Z. Rose (1989). Frontiers of Input-Output Analysis.
New York: Oxford University Press.
Rose, A. Z. and W. Miernyk (1989). “Input-Output Analysis: The First Fifty Years.”
Economic Systems Research 1(2): 229-271.

Other Literature Cited

Alward, G. S. and C. J. Palmer (1981). IMPLAN: An Input-Output Analysis System for
Forest Service Planning. Fort Collins, Colorado: U.S. Forest Service.
Ang, B.W. (1995). “Multilevel Decomposition of Industrial Energy Consumption.”
Energy Economics 17, 39-51.
Batey, P. W. J. and A. Z. Rose (1990). “Extended Input-Output Models: Progress and
Potential.” International Regional Science Review 13(1 & 2): 27-49.
Brucker, S. M., S. E. Hastings, et al. (1990). “The Variation of Estimated Impacts from
Five Regional Input-Output Models.” International Regional Science Review
13(1&2): 119-139.
Casler, S. and P. Blair (1997). “Economic Structure, Fuel Combustion, Pollution
Emissions,” Ecological Economics, 22, 19-27.
Carter, H. and D. Ireri (1972). “Linkage of California-Arizona Input-Output Models to
Analyze Water Transfer Patterns” in A. Carter and A. Brody (eds.) Applications
of Input-Output Analysis. Amsterdam: North-Holland.
Ghosh, A. (1958). "Input-Output Approach to an Allocative System." Econometrica
25(97): 58-64.
Giarratani, F. (1976). "Application of an Interindustry Supply Model to Energy
Issues." Environment and Planning A 8: 447-454.
Gowdy, J. and J. Miller (1987). “Technological and Demand Change in Energy Use:
An Input-Output Analysis.” Environment and Planning A, 19, 1387-1398.
Hamilton, J. R., Robison, M.H. Whittlesey and J. Ellis (1994). “Interregional Spillovers
in Regional Impact Assessment: New Mexico, Texas, and the Supreme Court.”
Growth and Change 25: 75-89.
Hannon, B. and T. Blazeck (1984). “The Marginal Cost of Energy Goods and
Services.” Energy Systems and Policy 8, 85-112.
Herendeen, R. (1974). “Affluence and Energy Demand.” Mechanical Engineering
96, 1-16.
Hewings, G. J. D. and R. C. Jensen (1986). “Regional, Interregional and Multiregional
Input-Output Analysis” in P. Nijkamp (ed.) Handbook of Regional and Urban
Economics, Vol. I - Regional Economics. Amsterdam: North-Holland.
Isard, W. (1951). “Interregional and Regional Input-Output Analysis: A Model of a
Space Economy.” Review of Economics and Statistics 33(4): 318-28.
Isard, W. (1967). “Linkages of socio-economic and ecological systems.” Papers and
Proceedings of the Regional Science Association 21: 79-99.
Isard, W. (1968). “Some Notes on the Linkage of the Ecologic and Economic Systems.”
Papers of the Regional Science Association 22: 85-96.
Isard, W. and R. E. Kuenne (1953). “The Impact of Steel upon the Greater New York-
Philadelphia Industrial Region.” Review of Economics and Statistics 35(4): 289-
301.
Jensen, R. C. (1990). “Construction and Use of Regional Input-Output Models:
Progress and Prospects.” International Regional Science Review 13(1&2): 9-
25.
Johnson, M. H. and J. T. Bennett (1981). “Regional Environmental and Economic Impact
Evaluation.” Regional Science and Urban Economics 11(1981): 215-230.
Kane, M. (2003). Shared Futures of City and Country: a Total Flows Analysis of
Urban-Rural Ecological Economic Dependence in New York City’s Catskill
Delaware Watershed. Ph.D. Dissertation, Department of Economics,
Rensselaer Polytechnic Institute, Troy, NY.
Lange, G. (1998). "From Data to Analysis: the Example of Natural Resource
Accounts Linked with Input-Output Information." Economic Systems
Research 10 (2):113-134.
Leontief, W. (1936). “Quantitative input-output relations in the economic system of
the United States.” Review of Economics and Statistics 18: 105-125.
Leontief, W. (1951). The Structure of the American Economy, 1919-1939. New York:
Oxford University Press.
Lofting, E. and P. McGauhey (1963). Economic Evaluation of Water, Part III: An
Interindustry Analysis of the California Water Economy. Water Resources
Center Contribution No. 67. Berkeley, CA: University of California.
Miernyk, W. (1967). An Interindustry Forecasting Model with Water Quantity and Water
Quality Constraints. Proceedings of the Fourth Symposium on Water Resources
Research, Water Resources Center, Ohio State University.
Moore, F. T. and J. W. Peterson (1955). “Regional analysis: an interindustry model of
Utah.” Review of Economics and Statistics 37: 363-83.
Polenske, K. R. (1980). The U.S. Multiregional Input-Output Accounts and Model.
Lexington, Massachusetts: Lexington Books.
Pyatt, G and J. Round (1985). Social Accounting Matrices: A Basis for Planning.
Washington, DC: The World Bank.
Roberts, D. (2000). “The Spatial Diffusion of Secondary Impacts: Rural-Urban Spillovers
in Grampian, Scotland.” Land Economics 76(3): 395-412.
Rose, A. (1995). Input-output economics and computable general equilibrium
models. Structural change and economic dynamics, 6(3), 295-304.
Rose, A. (1999). “Input-Output Structural Decomposition Analysis of Energy and
the Environment.” In: J.C.J.M. van den Bergh (ed.), Handbook of
Environmental and Resource Economics, Edward Elgar, Cheltenham,
pp.1164-1179.
Schaffer, W. A. and K. Chu (1969). “Non-survey techniques for constructing regional
interindustry models.” Papers and Proceedings of the Regional Science
Association 23: 83-101.
Shao, G. and R. E. Miller (1990). “Demand-side and Supply-side Commodity-Industry
Multiregional Input-Output Models and Spatial Linkages in the U.S. Regional
Economy.” Economic Systems Research 2(4): 385-405.
Stevens, B. H., Treyz, G. I., Ehrlich, D. J. and J.R. Bower (1983). “A New Technique for
the Construction of Non-Survey Regional Input-Output Models.” International
Regional Science Review 8(3): 271-86.
Stone, R. (1970). "Demographic Input-Output: An Extension of Social Accounting." In:
A.P. Carter and A. Brody (eds.). Contributions to Input-Output Analysis, Volume
1. North-Holland Publishing, Amsterdam.
Szyrmer, J.M. (1984). Total Flow in Input-Output Models. Dissertation submitted to
the Department of Regional Science, University of Pennsylvania.
Szyrmer, J. M. (1992). “Input-Output Coefficients and Multipliers from a Total
Flow Perspective.” Environment and Planning A 24: 921-37.
Szyrmer, J. M. and R. E. Ulanowicz (1987). “Total Flow in Ecosystems.” Ecological
Modeling 35(1-2): 123-136.
United Nations (1993). The System of Integrated Environmental and Economic
Accounts. UN, New York.
Victor, P. (1972). Pollution: Economy and the Environment. London: Allen &
Unwin.
Waters, E. C., Holland, D. W. and B.A. Weber (1994). “Interregional Effects of Reduced
Timber Harvests: The Impact of the Northern Spotted Owl Listing in Rural and
Urban Oregon.” Journal of Agricultural and Resource Economics 19(1): 141-160.
West, G. R. (1990). “Regional Trade Estimation: A Hybrid Approach.” International
Regional Science Review 13(1&2): 103-118.