|
|
|
SHIFT-SHARE
TECHNIQUE
Recall that the
purpose of the Constant Share and Shift-Share Techniques
is to project future employment in a given area. To do
this the Constant
Share Technique assumes
that each industry will continue to maintain a
"constant share" of the reference region's
employment into the future. If an industry has 5% of the
employment share now, it will have 5% it is assumed that
this industry will have 5% in the future as well.
Clearly, this is a very debatable assumption and one that
is not likely to hold for the vast majority of industries
in a local economy.
In response to this
significant shortcoming in the Constant Share approach,
economists have developed the Shift-Share
Technique. An offshoot of
the Constant Share Approach, the Shift-Share Technique
recognizes that some local industries are likely to be
growing at a faster rate when compared to the reference
region and other industries will be growing more slowly.
To quote from Klosterman:
"The
shift-share projection technique modifies the
constant-share projection formula by adding a shift
term to
account for differences between local and reference
region growth rates that cause an industry's
employment to shift into or out of a region."
(pps. 176, 178)
Like the Economic
Base Analysis Techniques
that expanded upon and improved upon previous methods,
the Shift-Share approach represents a useful refinement
of the Con-Share Technique. Rarely does the local economy
remain static when compared to a reference economy. The
local economy may specialize in certain industries and
attract employment in that industry, thereby increasing
its employment in that sector. For example, the Seattle
Metropolitan area has specialized (and to some degree
continues to do so) as a center for Transportation
Manufacturing. At the other end of the spectrum, an area
may lose employment in certain industries to regions that
have specialized in those products and services.
Seattle's waterfront has lost some employment to Tacoma
and other West Coast cities due to a variety of factors
(among them better transportation linkages and newer
facilities).
The Constant Share
method cannot and does not take these changes (or shifts)
into account in any way. The Shift-Share Technique
attempts to account for these "shifts" in
employment by recognizing certain trends and accounting
for them in the calculation of employment projections.
The Shift-Share
approach is based upon the modification of the
Constant-Share method. As with the Con-Share method, a
local economy is compared to a reference region economy.
However, a shift term is calculated to account for
changes in the local economy.
The basic formula for
the Shift-Share approach:
Projected
Employment
Industry I=
|
(
|
1+
|
Reference
Region Projected Growth Rate, Industry I
|
+
|
Projected
Local Employment Shift Term
|
)
|
X
|
Industry I Local Employment
|
Again,
note how the Shift-Share approach assumes that projected
local industry employment is directly related to the
Reference Region employment growth rate, but that the
calculation includes a "shift term". The only
modification to the Constant Share formula is the
addition of the Shift Term to our calculation. The
calculation of the shift term will be addressed shortly,
but for now let's go through a simple example and compare
the Shift Share technique to the Constant Share approach.
Using the above
formula and the following information what would be the
projected employment for Industry I?
Ref.
Region Growth Rate
|
0.3055
|
Shift Term
|
0.0741
|
Industry I Local
Employment
|
2,500
|
The
calculation would be as follows:
Shift Share Projected
Employment= (1+ 0.3055 +0.0741) X 2,500 = 3,449
In our local economy,
having calculated a Reference Region growth rate of
0.3055 (30.55%), an additional industry shift term of
0.0741 (7.41%), and a local economy employment on 2,500,
Industry I would be projected to have 3,449 jobs in a the
future year.
Using the same
information with the Constant Share Formula the projected
employment would be estimated to be:
Con Share Projected
Employment= (1+ 0.3055) X 2,500 = 3,264
Interpreting Our
Results
In our example we can see that the use of the Shift Term
in the Shift Share method means a difference in our
projections of almost 200 jobs. (3,449 - 3,264 = 185)
This difference results directly from our Shift term. A
positive shift term, in this case over 7%, indicates that
the growth rate in Industry I is greater in the local
economy than in the reference economy. Again, to
reiterate, the shift term attempts to account for local
conditions.
For example, airplane
manufacturing is growing faster in King County than in
the State as a whole. Therefore the local shift term must
be positive. Further, the larger the industry-specific
shift term the greater the divergence from the overall
regional growth rate. A shift term of 0.0891 will have a
greater effect than a shift term of 0.0221.
This Shift Term is
calculated by comparing the growth rates in the local and
reference region. As Klosterman notes, it is important
that these calculated growth rates be for the same period
of time. If they are for differing numbers of years a
correction will need to be employed in our formulas (see
Klosterman pps. 178-184)
The Shift Term
formula is very simply:
Shift Term
=
|
Local Growth Rate
for Industry I for Period X
|
-
|
Regional Growth Rate
for Industry I for Period X
|
And, from the Constant Share
page, a growth rate is calculated using the formula :
Growth
Rate Formula=
|
(
|
Projected
Employment Value
|
-
|
Previous
Employment Total
|
)
|
/
|
Previous Employment Total
|
Using
the following information, then, calculate the Shift Term
for our Industry I:
Year
|
Local
Employment Industry I
|
Reference
Region Employment
Industry I
|
1985
|
1,826
|
72,984
|
1995
|
2,500
|
89,267
|
The Growth
Rates for our two regions, from 1985 to 1995 would be:
Local GR= (2,500 -
1,826) / 1,826 = 0.3691
Regional GR= (89,267
- 72,984) / 72,984 = 0.2231
And the Industry I
Shift Term would be:
Shift Term = 0.3691 -
0.2231 = 0.1460
Interpreting Our
Results
In interpreting these results we would determine that the
local economy is growing faster in Industry I than the
regional economy, because the Local growth rate is
greater than the Regional growth rate. Therefore, there
is some evidence that the local economy might be
specializing in this Industry and we therefore might want
to account for that in our calculations. We would then
use this Shift Term in our Shift Share projection formula
to account for this information.
UNDER CONSTRUCTION
The
assumptions of the Shift Share technique
The
calculation of industry specific Reference Region
Growth Rates, Local Growth Rates
The
calculation of industry specific Shift terms
The
calculation of projected local industry
employment using the Shift Share technique
Interpretation
of the Growth Rates and Shift Terms
The
fundamental difference between the Con Share and
Shift Share techniques and the rationale behind
this difference
As should be
clear by this time in the course, planners often
use simple, but powerful analysis techniques to
estimate current conditions and to make
projections about the future. The simplicity and
elegance of these techniques also lead to
modifications to these techniques. However it is
important to note that these modifications are
often themselves quite simple and quite elegant.
The Shift Share projection technique is an
example of this. The simple Constant Share
approach has been modified to account for local
conditions by adding one factor to the Con Share
formula, the Shift Term. Again, simple
modifications to standard practices should be the
first means of adding complexity to an analytical
tool.
Another key
lesson of the Shift Share approach is that
generic formulas are powerful, but limited tools.
Modifications to standard analytic methods,
especially those that take into account local
conditions, are often the first means of applying
these generic techniques to a local situation.
"The
computational simplicity of these projection
models (Con Share and Shift Share), relative to
more sophisticated techniques, provides a
pragmatic justification for using them to prepare
quick and reasonably accurate short-run
projections. They are also appropriate for
preparing long-term 'baseline' forecasts
examining the implications of continuing past
employment trends into the future." (Klosterman,
p. 186)
What would a
negative shift term indicate? Give an example of
an industry in King County that might show a
negative shift term when compared to the State
economy. Why did you choose this industry?
Why is it
necessary to have a comparable time period when
calculating the Local and Regional growth rates?
What modification does Klosterman recommend when
the two time periods are not the same?
What dangers
do you see in using the Shift-Share technique? In
what conditions might the Shift-Share technique
be a poor choice?
Why have the
Constant Share and Shift Share techniques come
under attack? What shortcomings have researchers
identified as regards these models? (pps.
184-186)
Emmerson, Richard R.,
R. Ramanathan, and Wolfgang Ramm. 1975. "On the
Analysis of Regional Growth Patterns." Journal
of Regional Science 15: 17-28.
Greenberg, Michael R.
1972. "A Test of Alternate Models for Projecting
County Industrial Production at 2,3,4 Digit SIC
Levels." Regional and Urban Economics 1:
397-418.
Klosterman, Richard
E. 1990. Community and Analysis Planning Techniques.
Rowmand and Littlefield Publishers, Inc. Savage,
Maryland. See Chapter 12.
Klosterman, Richard
E., Richard K. Brail, and Earl G. Bossard. 1993. Spreadsheet
Models for Urban and Regional Analysis.
|