Geography Reference
In-Depth Information
of production that are dif erent from the regional average. In general, exiting plants tend
to be less ei cient than average, while entrants tend to have techniques that are more
widely dispersed.
Separating the inl uence of these processes on aggregate measures of technology can
make use of a variety of decompositions. The most general of these, avoiding covariance
terms, is outlined here. Thus, the change in the labor coei cient in region
r
, between time
t
and time
t
+ 1 is:
(
l
p
2
l
r
)
D
s
r
l
r
t
11
2
l
r
t
5 D
l
r
5
s
p
D
l
r
p
1
p
[
I
p
[
I
p
(
l
r
t
11
p
2
l
r
)
2
s
r
p
(
l
r
p
2
l
r
)
1
s
r
t
11
p
(2.1)
p
[
N
p
[
X
In equation (2.1),
s
represents market share and is dei ned at the plant level as
s
r
p
5
x
r
p
/S
p
x
r
p
, and
x
p
represents the output of plant
p
. In turn, D represents the change in
the value of the following variable between years
t
and
t
+ 1 and the
2
notation represents
the average value of the variable over the years
t
and
t
+ 1. The summation subscript
I
denotes the set of incumbent plants, the summation subscript
N
represents entrants,
and the summation subscript
X
represents exiting plants. In turn, the terms on the right
hand side of equation (2.1) capture the following inl uences on aggregate technology:
technological change in incumbent plants; changes in the market share of incumbent
plants; plant entry and plant exit. Changes in the aggregate capital coei cient can be
decomposed in exactly the same way.
The theoretical frameworks developed by Fisher and Price help us understand the
broad movement of the economy from a population perspective (Mayr, 1984). Building
on these foundations, more detailed arguments outline how processes of competition
and cooperation give rise to the particular coni gurations of technologies and institutions
that we observe. In terms of technology, considerable attention has been given to the way
in which heterogeneity shapes both the direction and pace of change. From the work of
Habakkuk (1962), David (1975), Metcalfe and Gibbons (1986) and Webber et al. (1992),
we know that the form or shape of technological variety inl uences the direction of imita-
tion within an economy, and how innovation is guided by relative prices. Those prices
are part of the selection environment within which plants compete, they are generated by
the choices of individual plants, and through market processes of supply and demand, as
well as through political contest, particularly in the case of wages.
The extent of variety within a selection environment is also thought to control the
pace of aggregate change, after Fisher (1930). Much less clear, however, are the precise
linkages between aggregate change, the strength of selection and the generation and
destruction of heterogeneity. In particular, how does the variance in plant characteristics
inl uence the pressure of selection; to what extent can inei ciency (perhaps thought of as
a measure of variety) be subsidized as a hedge against the lock-in of characteristics that
may prove unproi table in the long run; how does lock-in occur and how is it overturned?
Here there is great need for careful historical accounts of industrial and regional devel-
opment, of the generation of new products and processes of production, of competition
between i rms and technologies and resulting aggregate dynamics. Ideally, that history
would trace the strategies and fortunes of individual business units, linking economic
data for establishments and industries in particular places with i rm ethnographies and
