308
4. In demand-driven models, final demands for
each sector are exogenous variables and intermediate transactions are endogenous variables.
In supply-driven models, payments for inputs
for each sector are exogenous variables and intermediate transactions are endogenous variables.
5. Quantities of inputs purchased by each sector
are a function only ofthat sector's level of output. No substitution of factors (e.g., capital for
labor) is permitted (i.e., the relative proportion
of factors when forecasting with the model remains identical to that mix represented in the input-outputs accounts).
6. "Supply and demand in each market are
equated, not through changes in price and resulting movements along supply and demand
curves, but through a horizontal shift in the demand function of each industry resulting from
changes in production levels in other sectors ....
The assumption of maximizing behavior, which
is central to partial equilibrium analysis [the primary tool of analytical economics], plays no explicit role in the Leontief system [i.e., the 1-0
system of equations] .... [Rather] it is assumed
that producers have little or no choice as to factor proportions in the short run and react to demand changes by changing output rather than
price ... (Chenery and Clark, 1959, p. 4).
However, there is an implicit objective function built into the structure of an input-output
model. Basically there are two options: (1) a
supply-driven model in which the implicit objective function is to find the production schedule for all sectors that will exactly exhaust the
exogenous supplies of resources (as defined in
the payments sectors of the transactions table),
or (2) a demand-driven model in which the implicit objective function is to find the production schedule for all sectors that will exactly
meet the exogenous schedule of final demands.
Most input-output models, even those that focus on natural resources, are of the demand-driven type. There are a number of reasons for this,
primarily because such seems more logical
given the economic system and the fact that inputs become more specialized as they proceed
through the production system (e.g., a tree has
more alternative uses than a 1- by 4-inch piece
of lumber). Also, for the most part, resources
are actually valued by their expected contributions to the value of final products. For a comparison of supply- and demand-driven models,
see Hoover and Giarratani (1984, p. 330) and
Schallau and Maki (1983).
Methods of Economic Impact Analysis
7. Any resource constraints (e.g., natural resource
stocks, industrial capacity, labor supply) must
be handled by side calculations that are exogenous to the input-output model. Expanding the
input-output model to a linear programming
model permits building in such constraints, but
only at great costs in terms of model development and generalization of results (see Section
21.5).
8. "The total effect of carrying on several types of
production is the sum of the several effects. This
is known as the additivity assumption, which
rules out external economies and diseconomies"
(Chenery and Clark, 1959, p. 33).
21.2.1 Brief Description of an
Input-Output Model*
Input-output accounts form the basis for the transactions table, which indicates the dollar value of
each sector's transactions with all others in the regional economy. Input-output accounting results in
a transactions table summarizing all economic linkages in the regional economy for a specific time
(usually a year). A simple hypothetical transactions
table is shown in Table 21.1.
This table shows all economic linkages in the region priced at the producer level. It is easily read
and is a description of the regional economic structure at a given time. A row is read as describing
how the sales of a sector are distributed. For example, the agriculture sector has a gross output of
$100 million, which is distributed as follows: it
sells $25 million to itself (i.e., transactions between
firms in the sector itself), $30 million to industry,
$10 million to trade (within region) and services,
$5 million directly to households, and $30 million
to exports (outside region) and other final demand.
In a realistic model, a vastly larger number of sectors would be required (a bare minimum would be
about 40 for a developed industrial economy, and
most examples include hundreds of sectors, some
as many as 500, with elaboration particularly of industrial sectors). Also, capital accounts would be
included to trace changes in inventory, depletion,
and depreciation.
The columns describe how costs are distributed.
For example, in our simple table, industry buys $30
million of inputs from agriculture, $130 million
from itself, $90 million from trade and services,
$115 million from households (i.e., primarily labor), and $115 million from imports (from outside
*Quoted sections from Chappelle et al., 1986.
4. In demand-driven models, final demands for
each sector are exogenous variables and intermediate transactions are endogenous variables.
In supply-driven models, payments for inputs
for each sector are exogenous variables and intermediate transactions are endogenous variables.
5. Quantities of inputs purchased by each sector
are a function only ofthat sector's level of output. No substitution of factors (e.g., capital for
labor) is permitted (i.e., the relative proportion
of factors when forecasting with the model remains identical to that mix represented in the input-outputs accounts).
6. "Supply and demand in each market are
equated, not through changes in price and resulting movements along supply and demand
curves, but through a horizontal shift in the demand function of each industry resulting from
changes in production levels in other sectors ....
The assumption of maximizing behavior, which
is central to partial equilibrium analysis [the primary tool of analytical economics], plays no explicit role in the Leontief system [i.e., the 1-0
system of equations] .... [Rather] it is assumed
that producers have little or no choice as to factor proportions in the short run and react to demand changes by changing output rather than
price ... (Chenery and Clark, 1959, p. 4).
However, there is an implicit objective function built into the structure of an input-output
model. Basically there are two options: (1) a
supply-driven model in which the implicit objective function is to find the production schedule for all sectors that will exactly exhaust the
exogenous supplies of resources (as defined in
the payments sectors of the transactions table),
or (2) a demand-driven model in which the implicit objective function is to find the production schedule for all sectors that will exactly
meet the exogenous schedule of final demands.
Most input-output models, even those that focus on natural resources, are of the demand-driven type. There are a number of reasons for this,
primarily because such seems more logical
given the economic system and the fact that inputs become more specialized as they proceed
through the production system (e.g., a tree has
more alternative uses than a 1- by 4-inch piece
of lumber). Also, for the most part, resources
are actually valued by their expected contributions to the value of final products. For a comparison of supply- and demand-driven models,
see Hoover and Giarratani (1984, p. 330) and
Schallau and Maki (1983).
Methods of Economic Impact Analysis
7. Any resource constraints (e.g., natural resource
stocks, industrial capacity, labor supply) must
be handled by side calculations that are exogenous to the input-output model. Expanding the
input-output model to a linear programming
model permits building in such constraints, but
only at great costs in terms of model development and generalization of results (see Section
21.5).
8. "The total effect of carrying on several types of
production is the sum of the several effects. This
is known as the additivity assumption, which
rules out external economies and diseconomies"
(Chenery and Clark, 1959, p. 33).
21.2.1 Brief Description of an
Input-Output Model*
Input-output accounts form the basis for the transactions table, which indicates the dollar value of
each sector's transactions with all others in the regional economy. Input-output accounting results in
a transactions table summarizing all economic linkages in the regional economy for a specific time
(usually a year). A simple hypothetical transactions
table is shown in Table 21.1.
This table shows all economic linkages in the region priced at the producer level. It is easily read
and is a description of the regional economic structure at a given time. A row is read as describing
how the sales of a sector are distributed. For example, the agriculture sector has a gross output of
$100 million, which is distributed as follows: it
sells $25 million to itself (i.e., transactions between
firms in the sector itself), $30 million to industry,
$10 million to trade (within region) and services,
$5 million directly to households, and $30 million
to exports (outside region) and other final demand.
In a realistic model, a vastly larger number of sectors would be required (a bare minimum would be
about 40 for a developed industrial economy, and
most examples include hundreds of sectors, some
as many as 500, with elaboration particularly of industrial sectors). Also, capital accounts would be
included to trace changes in inventory, depletion,
and depreciation.
The columns describe how costs are distributed.
For example, in our simple table, industry buys $30
million of inputs from agriculture, $130 million
from itself, $90 million from trade and services,
$115 million from households (i.e., primarily labor), and $115 million from imports (from outside
*Quoted sections from Chappelle et al., 1986.