Geoscience Reference
In-Depth Information
Table 8.5
Washington unemployment rate and net migration
1968
1969
1970
1971
1972
1973
1974
Unemployment rate (%)
4.8
5.3
9.2
10.1
9.5
7.9
7.2
Net migration (thous.)
66.0
43.0
44.4
0.4
23.1
13.4
52.7
Source
: Washington projection and simulation model data base
at least based on the economic history of Washington. While the state has grown
two-fifths faster than the United States over the past 40 years, there is no evidence
that it has lost competitive advantage because of higher production costs due to the
inelasticity of the supply of labor or capital.
4
Indeed, there is substantial evidence in the state's boom and bust history that
labor supply is highly responsive to labor demand. When the economy expands, the
improved conditions, as signaled by new jobs, lower unemployment (at least
temporarily), and higher wages, trigger a migration of people into the region.
On the other hand, when the economy performs poorly, people move out.
Between 1969 and 1971, for example, Washington lost 82,100 jobs during the
Boeing Bust, boosting the unemployment rate to 10.1 %, almost twice the national
rate (Table
8.5
). In response to the downturn, 23,100 people on net left the state in
1972. The bust was followed by a 10-year expansion that created 662,200 new jobs
and attracted 534,200 migrants on net into the state.
Figure
8.1
illustrates not only the mobility of labor but also the distinct time lag
between employment change and net migration. The peaks and troughs of employ-
ment change consistently lead the peaks and troughs of net migration by 1 or
4
At the risk of oversimplificaton, economic modeling of regions in the United States is divided
into two camps: demand-oriented models based on the assumption that labor and capital are
mobile; and supply-side models that assume the supply of labor and capital is inelastic. What
regional economists presume about labor and capital mobility has a significant bearing on the
formulation of their models and the results of their analyses. Early versions of the Massachusetts
Economic Policy Analysis Model (Treyz et al.
1977
and Treyz and Duguay
1983
), described as an
endogenous wage determination model, yielded long-run employment multipliers close to one
principally because of the inelasticity of labor supply. According to the model, if an industry
expanded, it would generate jobs in other sectors through the traditional multiplier mechanism.
However, because the labor force—net migration in particular—did not fully respond to the
increased demand for labor, the new jobs permanently lowered the unemployment rate, drove
up wages, increased business costs, and reduced the demand for the state's exports. This in turn led
to a loss of jobs across the economy that offset most of the initial indirect employment gains. In a
simulation with the Massachusetts model, the total impact of 10,000 new jobs in the electrical
machinery industry amounted to only 13,786 jobs (including the initial 10,000 workers) 10 years
later. The implied long-run employment multiplier was 1.38. Also contributing to the small size of
the multiplier was a seemingly high estimate of the elasticity of demand for exports with respect to
production costs. Termed the “location response,” it was estimated to be
4.28. It would seem that
a model with such low employment multipliers because of the negative feedback effects would
have a difficult time explaining the Massachusetts Miracle in the late 1980s, which created
500,000 wage and salary jobs in a 6-year period, one sixth of the state's total payroll employment
today. It certainly would not have predicted the economic boom in Washington between 1983 and
1990, which added 600,000 jobs or nearly one-fifth of today's total employment.
