Geoscience Reference
In-Depth Information
et al. (
2009
). Once initial trade flow matrices are estimated for sectors based on a
reconciliation of CFS and IMPLAN data, these can then be iteratively refined via a
Fratar model. However, the conventional Fratar model (Sheffi
1985
) cannot esti-
mate the diagonal (intrastate flow) values, so NIEMO incorporates a DFM (doubly-
constrained Fratar Model) to supplement the off-diagonal flow estimates from the
standard Fratar model, providing consistent estimates for both on- and off-diagonal
values.
Usually, input-output models measure the direct (final demand), indirect (inter-
mediate input flows) and induced (secondary consumption associated with direct
and indirect employment) effects of changes in economic activity. In applying
NIEMO we chose, in most cases, to measure the direct and indirect impacts only,
for two reasons. First, it is a convention in MRIO to ignore induced effects (Miller
and Blair
1985
), presumably because induced consumption is less likely to cross
interregional, e.g. interstate boundaries. Second, although there are local induced
impacts associated with local indirect effects, the local and the imported indirect
impacts are typically allocated in a MRIO model via some assumption that falls
short of the accurate allocations of the “ideal” interregional input-output model
(Isard
1951
). For example, imported inputs of an origin sector are allocated to a
particular destination sector in the same proportion as local inputs.
In pursuing our research goals, the choice of approaches involved difficult trade-
offs. The use of linear economic models is justified by various factors, including the
richness of the detailed results made possible at relatively low cost; NIEMO, for
example, has approximately six million multipliers. The principal insight that
drives our research is that, with some effort, it is possible to integrate data from
IMPLAN's state-level input-output (I-O) models with the CFS data and other
sources for all individual States, making it possible to build an operational multi-
regional input-output model.
The drive behind the development of NIEMO was two-fold: to assess the
interstate impacts of events analyzed with our regional model [SCPM (the Southern
California Planning Model)]; and to allow us to extend the range of problems that
may be studied at the interstate level.
4.2
Building NIEMO
The U.S. Commodity Transportation Survey (CTS) on inter-regional trade flows
have been available since 1977 but reporting was discontinued for some years. For
the years since 1993, this data deficit can be met to some extent with the recent
(CFS) data from the Bureau of Transportation Statistics (BTS), but these data are
incomplete with respect to interstate flows. A key step was to create conversion
tables to reconcile the CFS and IMPLAN sectors.
The primary requirements for building an interstate model for the U.S. of the
Chenery-Moses type are two sets of data, regional coefficients tables and trade
coefficients tables (Miller and Blair
1985
). Models of this type can be used to
estimate inter-state industrial effects as well as inter-industry impacts on each state,
