Economic Restructuring, Moehr
26
migration flows, probably indicates two very different socio-economic systems. Second,
counties are problematic units of geography because they are both too big and too small.
I have tried to develop this paper with the understanding that local places made a
difference during economic restructuring – places had different resources to draw upon,
different chances for successful restructuring, and different levels of ‘attractiveness’ for
people and capital. However, places do not necessarily equate with county boundaries.
Chicago may be thought of as a single place, but it covers numerous counties across three
states. On the other end of the continuum, San Bernardino County, CA includes some
exurban areas of metropolitan Los Angeles and a good portion of the Mojave Desert.
Using a better geographical unit – one that more accurately approximates the socio-
economic system of real places – should help to reduce the disturbance terms in the
model. Third, it is impossible to tell how much the model is affected by the data
suppression required for CBP confidentiality standards. More importantly, this effect is
disproportionately large in counties with fewer reporting businesses. Therefore, less
populated counties are contributing less information to the restructuring factors when in
fact they may have experienced more restructuring relative to counties with larger
populations.
Even with these measurement limitations in mind, though, I think the findings of
this paper are important. Structural modeling has usually not taken into account the
spatial distribution of the variables under consideration, and this paper offers a useful
example of the advantages found in explicitly considering the geography of the socio-
economic system. Substantively, I find that the new economy is not beneficial for all
people and places. In fact, gains in the financial sector were not significantly related to
Convention