How to Analyze the Regional Economy with Occupation Data
Jun Koo1
Cleveland State University
Prepared for Economic Development Quarterly
Abstract
Occupation is an important aspect of the regional economy since regional
competitiveness has become increasingly dependent upon local knowledge
bases and worker quality. It has been forgotten, however, in most regional
economic analysis. This study proposes a new approach that utilizes occupation data available in the public domain in analyzing the regional
economy. The study defines occupation clusters and presents three occupation analysis approaches (i.e., overview analysis, occupation cluster
analysis, and occupation-based industry targeting analysis). The Cleveland metropolitan area is used for the illustration of analysis techniques.
Acknowledgments: The author would like to thank Ziona Austrian, James Robey, and Mark Rosentraub for
their comments and advice. Funding for this study was provided by the Greater Cleveland Partnership and
the Cleveland State University Presidential Initiative.
1
Jun Koo is an assistant professor in the Maxine Goodman Levin College of Urban Affairs at Cleveland
State University. His research focuses on technology-based economic development, industry clusters, and
workforce development; 2121 Euclid Avenue, Urban Building, Cleveland State University, Cleveland, OH
44115; Tel 216.687.5597, Fax 216.687.9277; E-Mail [email protected]
1
1.
Introduction
For most traditional economic development scholars and professionals, regional
economic analysis means examining industrial strengths and weaknesses and developing
strategies to replace declining industries and build regional competitiveness. In other
words, industry has served as a dominant unit of analysis for almost all applied regional
economic analyses.
This trend is changing now that researchers have begun paying attention to occupations as well as industries. Recently, regional competitiveness has become increasingly
dependent upon local knowledge bases and worker quality. Thus, examining regional
economies from a different angle (i.e., occupations) can provide important insights for regional development. For instance, although the automobile and chemical industries produce completely different products, software engineers in the two sectors often perform
similar tasks and are relatively interchangeable. Because workers who perform similar
tasks can easily move between industries with minimal retraining, strategies focusing
solely on industries are likely to overlook occupation-based opportunities across industries.
For this reason, policymakers need to pay as much attention to the functions that local
workers perform as to the output that they produce.
This paper proposes a new approach to analyze the regional economy with new occupation data that are available in the public domain. Economic characteristics and prospects of a region can be examined through the lenses of types of jobs that local workers
perform. This paper proceeds as follows. Section 2 sets the stage by briefly reviewing
theoretical discussions on occupations as a framework for regional analysis. Section 3 describes the knowledge-based benchmark occupation cluster approach for regional analysis
and presents a set of occupation clusters. Section 4 presents a technique for an occupationbased overview regional analysis. Section 5 and 6 illustrate more specific occupationbased analysis techniques (i.e., occupation cluster and occupation-based industry targeting
analyses) that can be applied to states or metropolitan areas. Occupation employment data
for the Cleveland metropolitan area are used for the illustration of analysis techniques.
Lastly, section 7 discusses implications for research and policy.
2
2.
Industry, Occupation, and Regional Economy
Industry has been a dominant focus in most regional economic analyses over the
last several decades because of Isard’s (1960) early work on industrial analysis and the
popularity of the economic base theory. Many regional analysis techniques, such as location quotients, shift-shares, and input-output analysis, were developed to examine local
industrial structure, industry cycles, industrial linkages, etc. Many state and local governments established industry task forces and launched strategic plans to develop specific target industries that could improve the business environment, create new jobs, and eventually boost the overall economic performance of a region. The development of the steel industry in Chicago and the polymer industry in Akron provide examples of this kind of development (Markusen, 2004). Despite industry’s dominant position in regional analysis,
however, history has proven that industry-based economic development strategies, such as
industry targeting, can lead either to success or disaster. In other words, the effectiveness
of industry targeting as a policy tool for regional development is questionable.
The limitations of industry as a primary unit of regional analysis were well recognized by Thompson and Thompson (1985, 1987, 1993). They argued that industry-based
analysis is a good start but can overlook another important aspect of the regional economy,
namely, occupations. To remedy this problem, they introduced the “occupationalfunctional approach,” which defines five distinct roles of a regional economy (i.e., entrepreneurship, central administration, research and development, precision operation, and
routine operations) and introduces occupations into the analysis. If the occupation-based
functional approach is used in conjunction with industry-based regional analysis, two regions with similar industry mixes can show very different characteristics and prospects depending on the functions performed in each industry. A recent study showed that some
industries do have different geographic patterns of innovation and production activities
(Koo, 2004). Rubber, plastic, and ceramic related industries, for instance, have high concentration of innovation activities in the Northeastern states whereas their production activities are heavily concentrated in the Carolinas. Such contrast in geographic distributions
of innovation and production activities implies that rubber, plastic, and ceramic related in-
3
dustries in the two regions might have very different occupational structures, which cannot
be captured by industry-based analysis only.
A regional economy indeed has two dimensions (i.e., industry and occupation). In
fact, the failure of many previous industry-targeting strategies can be attributed to the lack
of understanding of the two-dimensional regional economy. Policy approaches based on
one-dimensional thinking (i.e., industry-based strategies) for the two-dimensional regional
economy are likely to fail. As Thompson and Thompson (1993) argued, one needs two
“cross-hairs” to hit a target in the two-dimensional economic space.
The occupational aspect of the regional economy has become an increasingly important issue because today’s development process is heavily dependent upon human capital (Mather, 1999; Reich, 1991). Many previous studies showed that human capital is a
crucial factor for knowledge production (Anselin, Varga, & Acs, 1997; Griliches, 1979;
Jaffe, 1989). In particular, since knowledge is tacit and its movement depends on knowledge workers, human capital can serve as an intermediate agent in the knowledge spillover
process. Lucas (1988) showed that the accumulation of human capital can generate positive externalities since new skills acquired by each worker can be shared or can spill over
to others in the same location, eventually making the entire labor pool more productive.
Black and Henderson (1999) also argued that the accumulation of human capital promotes
endogenous economic growth. In addition, old style industry-targeting strategies accompanied by huge benefit packages for firms (e.g., tax incentives) have now proven to be futile (Greenstone & Moreti, 2003; McGuire, 203) because firms tend to focus more on the
quality of the local labor force in their location decisions (Calzonetti & Walker, 1991).
Therefore, the economic success of a region in the new economy hinges on whether its
economy has the right mix of workers to produce and disseminate new knowledge. Given
that the availability of skilled workers is one of the most important determinants of firm
location decisions, an in-depth analysis of regional labor endowments based on the occupational structure is an important element for any regional economic analysis.2
2
I do, however, acknowledge that the occupation is not a perfect measure of the workforce. The degree to
which occupations represent the quality of workers depends on knowledge and skill requirements of each
occupation. Occupations that require specific levels of education, such as doctors and engineers, have closer
fit with workers who fill the job than occupations with more general and lower levels of knowledge and skill
4
Although the new approach that Thompson and Thompson introduced several decades ago seems fundamentally valid and the emphasis on occupations is more relevant than
ever in today’s knowledge economy, only a handful of studies have used both industry and
occupation systematically for regional analysis. McKee and Froeschle (1985) and Theodore and Carlson (1998) used occupation data to identify local job opportunities. Ranney
and Bencatur (1992) proposed an approach that utilizes occupational skills and training
resources for developing economic development strategies. More recently, Feser (2003)
and Markusen (2004) demonstrated the potential for an occupation-based analysis as an
anchor for community and economic development strategies. Although some work has
been done, there is a relative dearth of research on occupations mainly because reliable regional-level occupation data were not available until recently.
3.
Knowledge-Based Benchmark Occupation Clusters
This study relies on two relatively new occupation databases. The first primary da-
tabase is the Occupational Employment Survey (OES) from the Bureau of Labor Statistics
(BLS). The OES database is published annually and has employment and wage information for almost 700 occupation categories at different geographical levels (e.g., metropolitan area, state, and nation).3 Although the database provides rich information on occupational mixes at different geographic scales, its size and complexity make it difficult to utilize for regional analysis. Besides, workers often move from one occupation to another
with little effort needed for retraining. In other words, many occupations share core
knowledge and skills and are thereby interchangeable. This study defines benchmark occupation clusters according to their knowledge requirements to foster more meaningful
analysis at a manageable level. Subsequent regional analysis is conducted based on
knowledge-based benchmark occupation clusters derived in this section.
requirements, such as clerical workers and machine operators. Since the latter occupations do not necessarily require specific qualifications, the education, knowledge, and skill levels of workers in the latter occupations are likely to be more diverse than those of workers in the former occupations.
3
The OES data are available from as early as 1996. However, the BLS changed its occupation classification
system in 1999. Therefore, the OES data before and after 1999 are not comparable.
5
Second, to create knowledge-based occupation clusters, the Occupational Information Network (ONET), is introduced. ONET (ver. 5.1 published in 2003) is a comprehensive database of worker attributes and characteristics developed based on nationwide surveys. It describes over 900 occupations in terms of 33 knowledge variables.4 Occupation
categories in the OES and ONET databases are roughly comparable. When ONET has
more detailed occupation categories, however, they are aggregated so that all ONET occupations match the OES occupations. A total of 48 OES occupations do not have comparable ONET occupations. Those unmatched OES occupations are dropped at the statistical
clustering stage. They are, however, added back later to potentially related occupation
clusters based on judgment. ONET occupation categories are then grouped into 20 occupation clusters based on their knowledge requirements.
The procedure for deriving benchmark occupation clusters proceeds as follows.5
The final ONET database prepared for the statistical clustering step has 33 knowledge
variables for 661 occupations. Common data reduction techniques are then applied to
knowledge variables and occupation categories in the ONET database. First, a principal
component factor analysis is conducted to reduce the number of knowledge variables and
thereby obtain more interpretable occupation cluster definitions. Derived principal components of knowledge variables are rotated using a varimax solution for better interpretation of the results. Knowledge factors with loadings of at least 0.5 are used for the interpretation of each factor. A total of 13 knowledge factors are extracted, depending upon
eigenvalues and interpretability.
I then conduct a statistical cluster analysis to group occupations based on 13 derived knowledge factors. Ward’s (1963) agglomerative hierarchical cluster algorithm is
applied to 661 occupations with 13 knowledge factors.6 This step yields a set of knowl4
See Feser (2003) for more detailed discussion of the ONET database.
This procedure relies heavily on Feser (2003). Feser’s analysis is based on the old BLS occupation classification system. However, the old classification system has been updated, and most occupation-related data
published by federal and state agencies since 1998 are based on the new standard occupation classification
system. The procedure implemented in this paper is also based on the new classification system.
6
A cluster analysis is conducted based on 13 knowledge factors instead of 33 knowledge variables because a
large number of dimensions for clustering (e.g., 33 knowledge variables) can dilute the unique characteristics
of occupations groups. In fact, the comparison of results based on 13 knowledge factors and 33 knowledge
variables shows that the former produce more intuitive and interpretable results.
5
6
edge-based benchmark occupation clusters that draw on the same set of knowledge requirements. The most difficult task in a statistical cluster analysis is determining how
many clusters need to be extracted. A large number of clusters are more representative but
may lack simplicity. On the other hand, if the number of derived clusters is too small,
comprehensiveness may be sacrificed. One of the most common criteria is an R-square
that represents the proportion of variance accounted for by clusters. The examination of Rsquares at each level of cluster hierarchy reveals that the statistical clustering procedure
may stop at around 17-21 clusters. After careful review of the five sets of results, a total of
20 occupation clusters are retained based on their interpretability. Table 1 shows the final
set of benchmark occupation clusters, their mean knowledge intensity, and the number of
U.S. total employment.7 Following Feser and Koo (2001), the knowledge intensity of occupation clusters is derived as follows:
Si =
∑ j ( K ij ) 2
ni
(1)
where Si is the mean knowledge intensity for occupation cluster i, Kij is the knowledge requirement j for occupation cluster i, and ni is the number of occupations in occupation
cluster i.
[Table 1 Here]
In terms of size, clerical workers and semi-skilled laborer and service workers
make up the two largest occupation clusters, representing 22 and 17 percent of the total
U.S. employment, respectively. The most knowledge-intensive occupation clusters are
social scientists (118.6); engineers, technicians, and architects (99.2); doctors, biomedical
scientists, and technicians (78.2); computer scientists and related specialists (68.0); and
healthcare specialists (64.9). An effective regional analysis could be conducted by examining the unique knowledge characteristics of occupational clusters and regional labor
force endowments.
7
A detailed list of occupations and knowledge variables in each benchmark cluster is available upon request.
7
4.
Overview Analysis
A regional economy can be defined by what it does (i.e., functions) as well as what
it makes (i.e., products). For instance, regions that produce computer chips and drugs can
be different in many ways from those that produce tobacco products and furniture. Similarly, regions that perform extensive research and development and therefore have a large
number of scientists and engineers are likely to be unique in many different aspects compared to heavy manufacturing regions where skilled production workers account for a significant share of the local labor force. Therefore, as the industry composition of a region
can provide a snapshot of the regional economy, local occupation distribution patterns can
also offer important insights about the region for economic development professionals. In
particular, since occupations differ in terms of how much knowledge and skills are required, the examination of occupation distributions can reveal how knowledge-oriented a
region is. This section presents two occupation-based analysis approaches that can provide
economic development professionals with an overall picture of the regional economy.
Occupation Trends
Changes in regional occupation distributions over time can depict the past, present,
and future of the regional economy. As the global and national economic structure
changes from labor-intensive to knowledge-intensive, types of functions that local workers
perform are also expected to evolve accordingly. Therefore, the underlying structural
changes of the regional economy are well reflected in occupation trends.
Since the OES data before and after 1999 are not comparable, this study uses the
1999 and 2001 data for a trend analysis. Although the timeframe is relatively short for a
reliable trend analysis, it is still meaningful to demonstrate its values. As the OES data
accumulate over time, the occupation trend analysis will become more relevant and valuable. A major issue that economic development professionals and academics may face
when using the OES data is how to handle suppressed cells. BLS does not report occupation estimates when estimated values do not satisfy the agency’s confidentiality or reliabil-
8
ity criteria. When occupation estimates are not reported at the state or metropolitan levels,
I estimate suppressed values based on national occupation shares.8
[Table 2 Here]
Table 2 presents the occupation trends of 20 knowledge-based industry clusters in
the nation, Ohio, and Cleveland from 1999 to 2001. The examination of occupation trends
at different geographical levels shows national as well as regional economic forces that
shape the characteristics of the regional economy. In the Cleveland metropolitan area,
only seven out of 20 occupation clusters experienced some growth during this period. The
most important drivers of growth are mostly service-related occupations such as financial
and legal personnel (2.7%), healthcare specialists (2.2%), and sales, marketing, and advertisement personnel (3.8%). These are relatively well-paid occupations that may shape the
future of Cleveland. Local financial and legal personnel and sales, marketing, and advertisement clusters outpaced the national trends in terms of their growth rates. The healthcare specialists cluster, however, lagged behind the nation with national employment
growing twice as fast as local employment.
On the other hand, there have been significant employment declines in a wide range
of occupation groups. In particular, manufacturing workers took the hardest hit (e.g.,
skilled laborer and machine operators). Note that Cleveland’s doctors, biomedical scientists, and technicians cluster declined by 2.4 percent during this period whereas the national cluster grew by over seven percent. In combination with the lackluster growth of the
healthcare specialists cluster, this implies some serious challenges to growing healthcarerelated industries in the Cleveland area. Occupation trends in Table 2 suggest that the
growth in healthcare-related industries in Cleveland is occurring mainly in low- to midlevel occupations. Job growth in more knowledge-intensive occupations (e.g., doctors and
8
The OES data have 22 occupation groups that cover over 700 detailed occupation categories. Occupation
group estimates include all suppressed individual occupations. Therefore, the distribution of occupation
group employment according to national shares can yield reasonable estimates of suppressed values. The
share is calculated as Oij*(Okn/Oin), where Oij is the number of employment of occupation group i in region j,
Okn is the national employment of occupation k, and Oin is the national employment of occupation group i.
To evaluate the validity of this technique, we calculate estimates for nonsuppressed cells and correlate them
with real values. The correlation coefficient between hypothetical and real values is 0.97.
9
biomedical scientists), which is potentially related to the development of the biotech industry in the future, is not observed.
The region suffered a significant loss not only in the doctors, biomedical scientists,
and technicians cluster but also in other knowledge-intensive and high-paying occupation
clusters such as computer scientists and related specialists and engineers, technicians, and
architects. Cleveland lost over 20 percent of its workforce between 1999 and 2001 in both
clusters, whereas the nation on average saw growth of 8.6 percent and 1.6 percent, respectively. Given the importance of these occupations for the development of knowledgebased industries in the new economy, Cleveland’s economic future may be stifled.
Metropolitan Knowledge Index
The knowledge intensity of occupations is meaningful information about the overall quality of the regional labor force and the readiness of the regional economy for the 21st
century. Cities and states with a high concentration of relatively knowledge-intensive occupation groups are in a better position to identify and nurture more knowledge-intensive
future industries. In other words, the quality of the regional labor force, when measured by
occupation mixes and their knowledge intensities, can indicate the region’s potential in the
new economy. Based on benchmark occupation clusters, this paper proposes a knowledge
index for metropolitan areas as follows:
Mk =
∑ i S i Eik
nk
(2)
where Mk is the metropolitan knowledge index of city k, Si is the mean knowledge intensity
for occupation cluster i, Eik is the employment share of occupation cluster i in city k, and nk
is the number of occupation clusters present in city k.
[Table 3 Here]
Table 3 lists the top 20 metropolitan areas in the country in terms of labor force
quality in 2001. Cities with higher knowledge intensity scores produce jobs that demand
relatively more knowledge and skills. As expected, so-called high-tech regions, such as
Boston, San Jose, Washington, Raleigh-Durham, appear at the top of the list. Relatively
small cities that are not often considered technologically advanced, such as Lowell (MANH), Stamford (CT), Tallahassee (FL), and Gainesville (FL), are also at the top 20 list.
10
The Cleveland-Lorain-Elyria metropolitan area has the highest knowledge intensity score
in Ohio, but it ranks only 43rd out of 337 metropolitan areas in the U.S. This implies that
Ohio’s overall labor quality is relatively low.
To examine the quality of Cleveland’s regional labor force in more detail, the study
compares Cleveland with two of the most knowledge-intensive cities (Boston and RaleighDurham) and two peer cities (Columbus and Cincinnati) in Table 4. As expected, Boston
and Raleigh-Durham have strong concentrations of knowledge-intensive occupations. For
instance, the location quotients of doctors, biomedical scientists, and technicians in the two
cities are 1.32 and 1.94, respectively. Those of computer scientists and related specialists
are 2.01 and 2.16, and those of engineers, technicians, and architects are 1.37 and 1.15, respectively.
[Table 4 Here]
In contrast, all three Ohio cities have quite generic occupation mixes overall, similar to those of the nation (i.e., their location quotients are close to 1 for most occupation
clusters). Cleveland has a slight edge over the other two peer cities in financial and legal
personnel (LQ=1.11), skilled laborer and machine operators (LQ=1.42), healthcare specialists (LQ=1.08), law enforcement and safety workers (LQ=1.10), and engineers, technicians
and architects (LQ=1.02). Surprisingly, despite having a strong healthcare industry, Cleveland does not have a particularly concentrated cluster of doctors, biomedical scientists, and
technicians when compared to the nation (LQ=0.89). Given that it is the most important
occupation group for developing the biotech industry on which Cleveland has set its sights,
a lower than expected concentration of doctors, biomedical scientists, and technicians cluster can limit the region’s future development strategy of developing the biotech industry.
5.
Occupation Cluster Analysis
The value of overview analysis can be greatly enhanced when it is accompanied by
more detailed analysis of occupation clusters. Since every region differs in some respects,
an overview analysis can help economic development professionals identify more relevant
occupation clusters to nurture for the region. A close look at detailed occupations in important regional occupation clusters can provide further information about the region’s
11
economic adaptability and prospects. In particular, a detailed analysis of regional occupation clusters can reveal occupational gaps and opportunities that a region faces. Drawing
upon the 2001 OES data, this section presents an illustrative analysis of three knowledgeintensive occupation clusters in the Cleveland area: doctors, biomedical scientists, and
technicians; computer scientists and related specialists; and engineers, technicians, and architects.
Doctors, Biomedical Scientists, and Technicians Cluster
The doctors, biomedical scientists, and technicians cluster consists of 20 occupations and is one of the most knowledge-intensive occupation groups. Some of the occupations require more extensive knowledge than others. The doctors, biomedical scientists,
and technicians cluster relies on four distinct but related knowledge fields: biology, chemistry, medicine and dentistry, and mathematics. These knowledge variables define the
characteristics of this cluster. Occupations included in this cluster share these knowledge
bases although their requirement levels may vary by occupation.
[Table 5 Here]
Table 5 illustrates the concentration levels of the 20 occupations in Cleveland.
Physicians in many different specialties are overrepresented in Cleveland; the location
quotient of general practitioners is as high as 1.88. Given the strength of Cleveland’s
healthcare industry, this is not surprising. However, the cluster has a lower than expected
concentration level because of the dearth of related professions other than doctors in the
region. For instance, professionals who are more likely to be involved in biomedical research activities, such as medical scientists (LQ=0.15), material scientists (LQ=0.83), microbiologists (LQ=0.30), biological science postsecondary teachers (LQ=0.51), health specialist postsecondary teachers (LQ, 0.42), medical and clinical technologists (LQ=0.86),
and biological technicians (LQ=0.40), are seriously underrepresented in the Cleveland
area. In other words, the presence of a strong healthcare industry accounts for a high concentration of physicians, but the region lacks other research-oriented biomedical professionals. Such an unbalanced distribution of occupations explains the lower than expected
level of overall concentration of the doctors, biomedical scientists, and technicians cluster
and may hamper the region’s strategy to develop the biotech industry.
12
Computer Scientists and Related Specialists Cluster
The computer scientists and related specialists cluster includes 11 occupations and
is defined by two very distinct knowledge fields: computers and electronics and mathematics. When compared to other occupation groups, the knowledge requirements for this cluster are more narrowly defined and have significant depth. Occupations that require indepth specialized knowledge bases often pose a significant challenge for economic development professionals because unique knowledge requirements make it difficult for existing
workers to transition to other occupations through retraining programs. Therefore, developing a sizable local pool of workers qualified for such occupations can be costly and
time-consuming.
[Table 6 Here]
Table 6 shows that Cleveland significantly lacks professionals in the computer scientists and related specialists cluster. Nine out of eleven occupations are underrepresented
(i.e., the location quotients are less than 1). Database administrators (LQ=1.23) is the only
occupation with a significant level of concentration in Cleveland. However, the knowledge intensity of database administrators is among the lowest in the cluster. Concentration
levels of other more knowledge-intensive occupations, such as computer hardware engineers (LQ=0.20), computer and information scientists (LQ=0.32), computer software engineers, systems software (LQ=0.35), and computer software engineers, applications
(LQ=0.53) are far below the national averages. The extent of the underrepresentation of
this occupation cluster in the Cleveland area is such that any future attempt to develop strategic industries that demand high-quality computer specialists is likely to be seriously undermined.
Engineers, Technicians, and Architects Cluster
The engineers, technicians, and architects cluster is the second most knowledgeintensive group and consists of 26 occupations. Compared to the previous two occupation
groups, this cluster requires relatively broader knowledge bases. Four knowledge fields,
engineering and technology, design, mathematics, and physics, define the cluster. Such
broad knowledge requirements can pose a challenge for regions because training workers
for these occupations can be costly as well as time-consuming.
13
[Table 7 Here]
Table 7 presents the concentration levels of the 26 occupations in Cleveland. The most
noticeable characteristic of this cluster is its concentration patterns. Cleveland has a strong
presence of relatively more knowledge-intensive engineering occupations. For instance,
industrial engineers (LQ=1.54), chemical engineers (LQ=1.38), material engineers
(LQ=1.36), mechanical engineers (LQ=1.33), engineering managers (LQ=1.27), and electrical engineers (LQ=1.05) are all more concentrated than the national averages. This is
probably due to the historically strong presence of the manufacturing sector. The table,
however, also shows that mid-level knowledge occupations (i.e., technicians and drafters)
are underrepresented. This is an interesting occupation distribution pattern that undermines the popular belief that the region lacks workers in more knowledge-intensive occupations. In addition, from an economic development policy perspective, such a pattern
poses a relatively less serious challenge because training technicians and drafters, if necessary, costs less than training engineers.
6.
Occupation-Based Industry Targeting Analysis
Traditional industry-targeting strategies usually focus on existing employment con-
centrations in a region or on the latest trendy industries without any serious consideration
of the region’s capacity to attract, nurture, and develop certain industries. When trying to
develop knowledge-intensive high-tech industries, however, state and local governments
need to pay more attention to their regional endowments, i.e. their capabilities to build
strength locally in such industries. The most important regional asset when developing
specialization in a certain sector is labor force. Whether a region has the right mix of
workers is a critical question in many firms’ location decisions. That is, the success of industry targeting relies heavily on the match between the characteristics of local workforce
(i.e., supply) and the labor needs of target industries (i.e., demand). Therefore, by studying
occupation and industry information together, economic development professionals can
make better-informed policy decisions. This section presents a two-step approach for occupation-based industry targeting and illustrates its application to the Cleveland metropolitan area.
14
Matching Industry Labor Needs and Regional Occupation Mix
To examine how well Cleveland’s labor force fits different industries, the region’s
occupation mix should be compared with industry labor needs. If Cleveland has an occupation mix that closely matches the labor needs of a certain industry, the city may indeed
be well prepared to attract, nurture, and develop that industry. To implement this strategy,
this study uses the location quotients (LQ) of the 20 occupation clusters in Cleveland and
knowledge-weighted industry labor needs. Knowledge-weighted industry labor needs are
obtained as follows:
Lij = K j S ij
(3)
where Lij is the knowledge-weighted labor need for occupation cluster j of industry i, Kj is
the knowledge intensity of occupation cluster j, and Sij is the employment share of occupation cluster j.
I use the Spearman correlation as a goodness-of-fit measure. This measure correlates the relative importance of occupation clusters for a certain industry measured by
knowledge-weighted industry labor need with the relative strengths of local occupation
clusters measured by location quotients. High correlation coefficients therefore imply that
a regional occupation mix (i.e., supply) is close to industry labor needs (i.e., demand). In
other words, the measure can provide valuable information as to whether the region has the
right mix of workers to develop a certain type of industry. Tables 8 and 9 list the top 10
best-fit manufacturing and service industries for Cleveland’s labor endowments. In manufacturing, the occupation mix of Cleveland bears a strong resemblance to the labor requirements for traditionally strong industries in Cleveland, such as rubber and miscellaneous plastic products (SIC 30), fabricated metal products (SIC 34), petroleum refining (SIC
29), and primary metal (SIC 33). In services, finance and insurance service related sectors
(SIC 63, 64, 61, 62, 67, 60) top the list.
[Table 8 Here]
[Table 9 Here]
A comparison of manufacturing and service sectors in Tables 8 and 9 reveals an
interesting pattern. Most best-fit manufacturing sectors are already saturated in the region,
15
whereas best-fit service sectors seem to be emerging now. None of the best-fit service sectors are as strongly concentrated as many of the best-fit manufacturing sectors. This implies that having the right mix of labor force to develop such manufacturing industries
might reflect current industrial concentrations in Cleveland rather than indicate future development potential. That is, this approach can be useful in identifying potentially emerging industries given local labor force distributions. However, one needs to exercise some
caution since the results can be overshadowed when there is already strong presence of certain industry groups.
Identifying Key Occupations
Once target industries are identified, the next important question for economic development professionals is how to attract, nurture, and develop them. Since workforce is
the most critical element of building knowledge-intensive industries in particular
(Calzonetti & Walker, 1991), identifying key occupations for target industries is an important first-step. Key occupations are defined as those that are critical to the development
and expansion of an industry because, first, they require a high level of knowledge and
skills to perform essential functions (these occupations are therefore not easily replaceable), and second, the industry demands a significant number of workers in those occupations. Therefore, knowledge intensity and the size of occupations determine how critical
they are to the development and expansion of a certain industry. The knowledge-weighted
industry labor need index presented in equation 3 can effectively implement this idea. The
index represents the relative importance of occupation clusters. Therefore, higher values
mean that they are more important to the development and expansion of a certain target
industry.
Tables 10 and 11 present key occupation clusters for the top 10 best-fit manufacturing and service industries and their location quotients in Cleveland. All top 10 manufacturing industries share almost identical key occupations. Skilled laborers and machine operators make up one of the most important occupation groups in the top 10 manufacturing
industry targets. Supervisors and management personnel and engineers, technicians, and
architects are also critical to those industries. All three occupation clusters are relatively
well-represented in the Cleveland area. In particular, the skilled laborer and machine op-
16
erators cluster is highly concentrated in the region (LQ=1.42). Given the strong presence
of the manufacturing sector in Cleveland, this finding is hardly surprising.
[Table 10 Here]
[Table 11 Here]
On the other hand, two relatively more knowledge-intensive industries, chemical
and allied products (SIC 28) and electronic and other electrical equipment and components
(SIC 36), demand significant numbers of doctors, medical scientists, and technicians and
computer scientists and related specialists. Both occupation clusters are, however, underrepresented in Cleveland. In particular, an in-depth analysis of detailed occupations in the
doctors, biomedical scientists, and technicians cluster in the previous section revealed that
although there is a relatively strong presence of medical practitioners in the region, research-related professionals such as microbiologists and medical scientists are seriously
underrepresented. The analysis also showed that virtually all occupations in the computer
scientists and related specialists cluster are underrepresented in Cleveland. Therefore,
unless significant efforts are made to develop a critical mass of those key occupation
groups, the development of the chemical and allied products and electronic and other electrical equipment and components industries will likely to face serious limitations because
of the lack of local labor force in relatively more knowledge-intensive key occupation
groups.
In the service sector, supervisors and management personnel, clerical workers, and
sales, marketing, and advertisement personnel clusters are in high demand among a wide
range of industries. The financial and legal personnel cluster is particularly important in
emerging finance and insurance service related sectors (SIC 63, 61, 62, 67, 60), and doctors, biomedical scientists, and technicians and healthcare specialists clusters are critical
for the health service industry (SIC 80). Cleveland has about average levels of concentration in all those occupation groups except for the financial and legal personnel group
(LQ=1.11). The emerging status of many finance and insurance service related industries
and the relatively strong presence of financial and legal personnel in the Cleveland area
suggest that there will be more room for future growth in those industries. Therefore, the
region needs to pay constant attention to educating and retraining workers in those key oc-
17
cupations. On the other hand, Cleveland might face a challenge in developing health service-related industries because the city lacks higher-end occupations in the doctors, biomedical scientists, and technicians cluster.
7.
Conclusion
Occupation is an important aspect of the regional economy; it has been forgotten,
however, in most studies of regional economies. This paper presents various approaches
that economic development professionals can apply to analyze the regional economy with
occupation data. Using the Cleveland metropolitan area as an example, the paper provides
three illustrative occupation analysis approaches (i.e., overview analysis, occupation cluster analysis, and occupation-based industry targeting analysis) and discusses how they can
inform strategic development planning practices. Each set of analysis can offer economic
development professionals unique occupational perspectives of the regional economy. As
illustrated, in particular, the occupation analysis in conjunction with industry information
can be a powerful new tool for the regional economic analysis.
Although this paper limits its contents to the three proposed occupation analysis
approaches in one metropolitan area, potential permutations of occupation-based regional
analysis can be virtually limitless given the depth of occupation-related data available in
the public domain. For instance, some industries share similar key occupations as presented in Table 10 and 11; therefore, a complete set of labor-based industry clusters can be
derived.9 This can be both a theoretically and practically valuable undertaking since the
agglomeration economies theory predicts that firms can benefit from locating near other
firms that rely on workers with similar characteristics. A deep pool of skilled workers allows firms to adjust their employment levels easily according to the business cycle.
In addition, the combination of a key occupation analysis and the knowledge requirement data can provide useful information to design regional workforce training programs. The examination of key occupations for a certain industry reveals where the labor
gap lies (i.e., types of regionally underrepresented occupations and essential knowledge
requirements for those occupations). Therefore, state and local governments can design
9
See Feser and Koo (2001) for more detailed discussion of labor-based industry clusters.
18
training programs accordingly to develop underrepresented key occupations that are essential to attract and nurture certain target industries.
Since occupations can be grouped according to the types and depths of knowledge
requirements, an examination of knowledge requirements for individual occupations can
also provide critical information for identifying career ladders. Workforce training programs designed along well-defined career paths can help workers better prepared as they
move up career ladders.
19
Reference
Anselin, L., Varga, A., & Acs, Z. (1997). Local geographic spillovers between university
research and high technology innovations. Journal of Urban Economics, 42, 422488.
Black, D., & Henderson, V. (1999). A theory of urban growth. Journal of Political Economy, 107, 252-284.
Calzonetti, F. J., & Walker, R. T. (1991). Factors affecting industrial location decisions: A
survey approach. In H. W. Herzog & A. M. Schlottmann (Eds.), Inustry Location
and Public Policy (pp. 221-240). Knoxville: University of Tennessee Press.
Feser, E. J. (2003). What regions do rather than make: A proposed set of knowledge-based
occupation clusters. Urban Studies, 40(10), 1937-1958.
Feser, E. J., & Koo, J. (2001). Labor-based industry clusters.Unpublished manuscript,
Dept. of City and Regional Planning, Univ. of North Carolina at Chapel Hill.
Greenstone, M., & Moreti, E. (2003). Bidding for industrial plants: Does winning a 'million dollar plant' increase welfare? NBER Working Paper, 9844.
Griliches, Z. (1979). Issues in assessing the contribution of research and development to
productivity slowdown. Bell Journal of Economics, 10, 92-116.
Isard, W. (1960). Methods of Regional Analysis. Cambridge: MIT Press.
Jaffe, A. (1989). Real effects of academic research. American Economic Review, 79, 957970.
Koo, J. (2004). In search for knowledge-based industry clusters. Paper presented at the Association for Public Policy Analysis and Management 26th Annual Research Conference, Atlanta, GA.
Lucas, R. E. (1988). On the mechanics of economic development. Journal of Monetary
Economics, 22, 3-42.
Markusen, A. (2004). Targeting occupations in regional and community economic development. Journal of the American Planning Association, 70, 253-268.
Mather, V. (1999). Human capital-based strategy for regional economic development.
Economic Development Quarterly, 13, 203-216.
McGuire, T. J. (203). Do taxes matter? Yes, no, maybe so. State Tax Notes, June, 885-890.
McKee, W. L., & Froeschle, R. C. (1985). Where the Jobs Are. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.
Ranney, D. C., & Betancur, J. J. (1992). Labor-force-based development: A communityoriented approach to targeting job training and industrial development. Economic
Development Quarterly, 6, 286-296.
Reich, R. (1991). The Work of Nations. New York: Knopf.
Theodore, N., & Carlson, V. L. (1998). Targeting job opportunities: Developing measures
of local employment. Economic Development quarterly, 12, 137-149.
Thompson, P. R., & Thompson, W. R. (1985). From industries to occupations: Rethinking
local economic development. Economic Development Commentary, 9, 12-18.
20
Thompson, P. R., & Thompson, W. R. (1993). Cross-hairs targeting for industries and occupations. In D. L. Barkley (Ed.), Economic adaptation: alternatives for nonmetropolitan areas (pp. 265-286): Westview Press.
Thompson, W. R., & Thompson, P. R. (1987). National industries and local occupational
strengths: the cross-hairs of targeting. Urban Studies, 24, 547-560.
Ward, J. H. J. (1963). Hierarchical grouping to optimize an objective function. Journal of
American Statistical Association, 58, 236-244.
21
[Table 1] Benchmark Occupation Clusters and Knowledge Intensity
Occupation Cluster
Knowledge Intensity US Emp (2001)
Financial and legal personnel
44.7
3,486,110
Social scientists
118.6
289,520
Artists and performers
28.2
997,610
Doctors, biomedical scientists, and technicians
78.2
1,535,000
Transportation and mining workers
22.8
5,081,120
Computer scientists and related specialists
68.0
2,647,320
Supervisors and management personnel
59.5
10,437,320
Specialized mechanics, repairs, and technicians
24.8
5,511,610
Semi skilled laborer and service workers
11.6
20,512,370
Clerical workers
29.2
26,476,580
Skilled laborer and machine operators
18.6
7,904,440
Healthcare specialists
64.9
9,167,680
Construction workers
15.0
3,799,720
Special educators and teachers
58.1
6,540,090
Sales, marketing, and advertisement personnel
54.3
6,152,340
Food preparation workers
14.6
3,162,430
Law enforcement and safety workers
37.6
2,962,330
Engineers, technicians, and architects
99.2
2,589,010
Farming and agricultural workers
46.9
443,480
Earth scientists
63.3
188,970
Source: OES, ONET, and Author’s Calculation
22
[Table 2] Occupation Cluster Trends 1999 – 2001
Cleveland US Emp
Emp (2001) Chg 99-01
Financial and legal personnel
33,677
2.2%
Social scientists
2,349
6.1%
Artists and performers
6,496
4.5%
Doctors, biomedical scientists, and technicians
13,512
7.4%
Transportation and mining workers
39,515
0.7%
Computer scientists and related specialists
16,970
8.6%
Supervisors and management personnel
85,933
-1.3%
Specialized mechanics, repairs, and technicians
45,061
3.7%
Semi skilled laborer and service workers
171,103
0.5%
Clerical workers
227,978
3.3%
Skilled laborer and machine operators
99,293
-9.2%
Healthcare specialists
85,994
5.1%
Construction workers
25,526
5.1%
Special educators and teachers
51,070
5.2%
Sales, marketing, and advertisement personnel
56,462
1.0%
Food preparation workers
24,751
-3.2%
Law enforcement and safety workers
28,378
0.3%
Engineers, technicians, and architects
23,360
1.6%
Farming and agricultural workers
1,052
0.2%
Earth scientists
1,567
9.7%
Source: OES and Author’s Calculation
Occupation Cluster
US Wage State Emp State Wage Cleveland Emp Cleveland
(2001) Chg 99-01 (2001)
Chg 99-01 Wage (2001)
54,570
-1.4%
51,002
2.7%
54,619
55,430
18.5%
50,205
21.6%
55,764
38,380
-6.1%
34,469
-9.5%
34,233
74,670
1.3%
74,491
-2.4%
69,826
39,281
-6.1%
36,696
-14.0%
37,436
63,237
-5.8%
58,624
-22.4%
59,945
52,510
-7.2%
48,952
-9.3%
51,905
33,707
-2.9%
32,849
-0.2%
35,045
24,997
-1.1%
24,405
-0.2%
24,183
27,745
1.7%
26,087
-2.4%
27,164
31,108
-9.8%
31,483
-17.4%
31,948
38,816
2.2%
38,222
2.2%
38,878
31,623
-6.6%
32,742
-19.1%
36,232
39,650
6.6%
38,170
8.1%
41,483
50,115
7.5%
47,676
3.8%
49,866
25,297
-2.4%
24,814
11.8%
21,068
39,104
-3.0%
38,235
6.7%
41,007
60,443
-8.7%
56,148
-25.4%
55,272
32,744
12.7%
36,255
-18.5%
31,248
53,509
-17.7%
46,466
-6.8%
41,143
23
[Table 3] Metropolitan Knowledge Index
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
43
Metropolitan Area
Boston, MA-NH PMSA
San Jose, CA PMSA
Washington, DC-MD-VA-WV PMSA
Lowell, MA-NH PMSA
Raleigh-Durham-Chapel Hill, NC MSA
Boulder-Longmont, CO PMSA
Seattle-Bellevue-Everett, WA PMSA
Austin-San Marcos, TX MSA
Stamford-Norwalk, CT PMSA
Philadelphia, PA-NJ PMSA
Hartford, CT MSA
Tallahassee, FL MSA
Kansas City, MO-KS MSA
Houston, TX PMSA
Gainesville, FL MSA
Baltimore, MD PMSA
Worcester, MA-CT PMSA
Albany-Schenectady-Troy, NY MSA
Rochester, NY MSA
Nassau-Suffolk, NY PMSA
Cleveland-Lorain-Elyria, OH PMSA
Source: OES, ONET, and Author’s Calculation
Total Emp (2001) Knowledge Index
1,971,030
37.50
950,600
37.10
2,673,890
36.93
124,200
35.70
661,370
35.68
180,980
35.66
1,331,820
35.64
658,400
35.23
207,750
35.04
2,342,100
35.02
612,480
34.91
153,910
34.87
939,300
34.85
2,062,570
34.84
118,790
34.82
1,208,970
34.70
230,250
34.69
444,770
34.66
527,420
34.65
1,184,580
34.40
1,108,600
33.37
24
[Table 4] Occupation Cluster Mix in Selected Metropolitan Areas*
Occupation Cluster
Boston Raleigh-Durham Cleveland Columbus Cincinnati
Financial and legal personnel
1.57
0.96
1.11
1.04
0.90
Social scientists
1.73
2.39
0.76
0.79
0.83
Artists and performers
1.35
0.86
0.72
0.82
0.79
Doctors, biomedical scientists, and technicians 1.32
1.94
0.89
0.85
0.79
Transportation and mining workers
0.60
0.65
0.80
0.94
0.99
Computer scientists and related specialists
2.01
2.16
0.74
1.31
1.01
Supervisors and management personnel
1.12
1.05
0.94
1.00
0.99
Specialized mechanics, repairs, and technicians 0.80
0.85
0.87
0.84
0.96
Semi skilled laborer and service workers
0.94
0.83
0.95
1.07
1.02
Clerical workers
0.99
0.97
0.98
1.03
0.97
Skilled laborer and machine operators
0.69
0.67
1.42
0.83
1.07
Healthcare specialists
1.12
0.97
1.08
0.97
1.06
Construction workers
0.67
0.89
0.76
0.77
0.78
Special educators and teachers
0.98
0.94
0.88
0.75
0.79
Sales, marketing, and advertisement personnel
1.15
1.07
1.06
0.99
1.07
Food preparation workers
0.84
0.89
0.84
0.98
0.94
Law enforcement and safety workers
0.99
0.86
1.10
0.93
0.91
Engineers, technicians, and architects
1.37
1.15
1.02
0.91
0.94
Farming and agricultural workers
0.36
0.56
0.20
0.33
0.32
Earth scientists
0.61
1.84
0.66
0.61
0.47
Source: OES and Author’s Calculation
* Values are LQs of occupation clusters.
25
[Table 5] Occupations in the Doctors, Biomedical Scientists, and Technicians Cluster
Occupation
Family and General Practitioners
Pediatricians, General
Medical and Clinical Laboratory Technicians
Podiatrists
Surgeons
Internists, General
Chemists
Obstetricians and Gynecologists
Physician Assistants
Pharmacists
Veterinarians
Chemical Technicians
Medical and Clinical Laboratory Technologists
Materials Scientists
Environmental Scientists and Specialists, Including Health
Biological Science Teachers, Postsecondary
Health Specialties Teachers, Postsecondary
Biological Technicians
Microbiologists
Medical Scientists, Except Epidemiologists
Source: OES, ONET, and Author’s Calculation
LQ
1.88
1.82
1.27
1.21
1.19
1.15
1.10
1.07
1.03
1.00
0.98
0.94
0.86
0.83
0.54
0.51
0.42
0.40
0.30
0.15
Knowledge Intensity
119.24
119.24
50.64
55.22
84.82
119.24
66.50
119.24
57.92
86.68
69.68
43.54
62.43
67.68
47.29
140.41
117.42
15.01
47.02
82.09
26
[Table 6] Occupations in the Computer Scientists and Related Specialists Cluster
Occupation
Database Administrators
Computer Science Teachers, Postsecondary
Network and Computer Systems Administrators
Computer Systems Analysts
Computer Support Specialists
Computer Programmers
Network Systems and Data Communications Analysts
Computer Software Engineers, Applications
Computer Software Engineers, Systems Software
Computer and Information Scientists, Research
Computer Hardware Engineers
Source: OES, ONET, and Author’s Calculation
LQ
1.23
1.01
0.92
0.86
0.83
0.81
0.62
0.53
0.35
0.32
0.20
Knowledge Intensity
47.75
93.93
39.12
68.10
35.89
67.25
40.53
95.69
95.69
93.93
95.69
27
[Table 7] Occupations in the Engineers, Technicians, and Architects Cluster
Occupation
Mechanical Drafters
Industrial Engineers
Chemical Engineers
Materials Engineers
Mechanical Engineers
Engineering Managers
Industrial Engineering Technicians
Electrical Engineers
Architects, Except Landscape and Naval
Landscape Architects
Electricians
Physicists
Mechanical Engineering Technicians
Civil Engineers
Chemistry Teachers, Postsecondary
Civil Engineering Technicians
Architectural and Civil Drafters
Electrical and Electronics Drafters
Urban and Regional Planners
Health and Safety Engineers, Except Mining Safety Engineers and Inspectors
Aerospace Engineers
Statisticians
Physics Teachers, Postsecondary
Electronics Engineers, Except Computer
Environmental Scientists and Specialists, Including Health
Environmental Engineering Technicians
Source: OES, ONET, and Author’s Calculation
LQ Knowledge Intensity
2.32
64.55
1.54
104.52
1.38
151.67
1.36
69.47
1.33
102.86
1.27
134.69
1.11
79.75
1.05
153.90
1.03
146.82
1.03
93.61
1.03
74.69
0.96
112.33
0.95
65.22
0.82
138.63
0.76
87.43
0.75
58.86
0.73
61.81
0.68
64.27
0.67
110.59
0.67
52.82
0.65
157.81
0.61
48.77
0.59
91.87
0.56
121.83
0.55
47.29
0.51
58.86
28
[Table 8] Top 10 Best-Fit Manufacturing Industries
SIC Industry
30 Rubber and Miscellaneous Plastics Products
Fabricated Metal Products, Except Machinery and Transportation
34
Equipment
26 Paper and Allied Products
29 Petroleum Refining and Related Industries
Measuring, Analyzing, and Controlling Instruments; Photographic,
38
Medical and Optical Goods; Watches and Clocks
33 Primary Metal Industries
39 Miscellaneous Manufacturing Industries
28 Chemicals and Allied Products
Electronic and Other Electrical Equipment and Components, Except
36
Computer Equipment
37 Transportation Equipment
Source: ES202 and Author’s Calculation
LQ
1.40
Correlation
0.62
2.41
0.60
0.98
1.23
0.58
0.58
1.03
0.58
2.77
0.97
1.83
0.56
0.55
0.54
0.87
0.54
1.11
0.54
29
[Table 9] Top 10 Best-Fit Service Industries
SIC Industry
63 Insurance Carriers
64 Insurance Agents, Brokers, and Service
61 Non-Depository Credit Institutions
65 Real Estate
73 Business Services
62 Security and Commodity Brokers, Dealers, Exchanges, and Services
67 Holding and Other Investment Offices
76 Miscellaneous Repair Services
80 Health Services
60 Depository Institutions
Source: ES202 and Author’s Calculation
LQ
1.11
0.72
1.04
0.91
0.79
0.62
1.00
0.88
1.15
1.18
Correlation
0.62
0.60
0.55
0.54
0.54
0.53
0.46
0.46
0.45
0.44
30
[Table 10] Key Occupations for Top 10 Best-Fit Manufacturing Industries*
Occupation Cluster
SIC 30 SIC 34 SIC 26 SIC 29 SIC 38 SIC 33 SIC 39 SIC 28 SIC 36 SIC 37
Financial and legal personnel
58.6 75.1 60.3 188.2 169.0 59.5 86.3 108.2
Social scientists
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Artists and performers
2.3
2.3
1.7
0.0
12.7
0.8
29.0
1.7
Doctors, biomedical scientists, and technicians
27.4
6.3
26.6 199.4 74.3 38.3
7.0 726.5
Transportation and mining workers
63.8 60.4 117.0 155.7 9.8 107.2 29.4 59.1
Computer scientists and related specialists
34.7 38.8 42.8 185.0 361.1 44.2 70.0 115.6
Supervisors and management personnel
577.2 584.3 533.7 701.5 524.8 577.2 589.6 675.9
Specialized mechanics, repairs, and technicians 116.1 96.2 173.8 187.0 85.1 202.4 54.1 185.3
Semi skilled laborer and service workers
122.0 66.2 105.3 40.4 57.2 89.4 184.1 65.0
Clerical workers
204.4 246.7 209.1 237.1 308.4 183.7 372.3 278.6
Skilled laborer and machine operators
947.5 952.9 863.4 509.5 616.2 843.0 642.3 606.7
Healthcare specialists
6.5
4.5
9.1
28.6 24.7
5.8
3.9
27.3
Construction workers
3.6
44.6
1.4
12.5
4.4
10.1 29.1
1.8
Special educators and teachers
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.2
Sales, marketing, and advertisement personnel 120.0 161.8 169.4 172.1 271.5 94.5 315.5 213.9
Food preparation workers
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.3
Law enforcement and safety workers
4.5
7.9
7.9
35.0 12.0 13.2
6.4
33.1
Engineers, technicians, and architects
273.8 386.9 266.8 530.7 1237.0 429.5 226.2 450.4
Farming and agricultural workers
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.8
Earth scientists
0.0
0.6
0.6
13.3
0.0
0.0
0.0
0.6
Source: OES, ONET, and Author’s Calculation
* Values represent the relative importance of occupation clusters; they do not have any absolute meaning.
114.4
0.0
8.2
9.4
21.9
352.2
489.1
99.9
45.0
239.4
826.6
13.6
3.9
0.0
202.0
0.0
6.0
928.5
0.5
0.0
106.4
0.0
16.6
7.8
42.2
132.6
430.8
188.0
49.0
173.2
660.7
18.8
30.3
0.0
96.7
0.0
21.8
958.3
0.0
0.0
Cleveland
LQ
1.11
0.76
0.72
0.89
0.8
0.74
0.94
0.87
0.95
0.98
1.42
1.08
0.76
0.88
1.06
0.84
1.1
1.02
0.2
0.66
31
[Table 11] Key Occupations for Top 10 Best-Fit Service Industries*
Occupation Cluster
Cleveland
LQ
670.1
1.11
0.0
0.76
0.6
0.72
0.0
0.89
0.5
0.8
229.2
0.74
702.1
0.94
4.5
0.87
102.8
0.95
1481.3 0.98
0.0
1.42
25.3
1.08
0.0
0.76
0.0
0.88
214.5
1.06
0.0
0.84
20.7
1.1
4.0
1.02
0.0
0.2
0.0
0.66
SIC 63 SIC 64 SIC 61 SIC 65 SIC 73 SIC 62 SIC 67 SIC 76 SIC 80 SIC 60
Financial and legal personnel
502.0 282.5 1101.0 195.8 66.2 826.5 716.5 39.3
Social scientists
0.0
0.0
0.0
0.0
7.1
0.0
8.3
0.0
Artists and performers
2.3
0.6
1.1
1.4
29.9
8.7
14.9
2.5
Doctors, biomedical scientists, and technicians
10.9
0.8
0.0
0.0
4.7
0.0
13.3
0.0
Transportation and mining workers
0.7
0.2
2.1
9.6
42.0
1.6
13.0 47.0
Computer scientists and related specialists
456.3 148.2 221.0 23.8 768.4 363.8 381.5 21.1
Supervisors and management personnel
578.9 556.9 635.5 1024.0 392.7 555.7 985.3 613.4
Specialized mechanics, repairs, and technicians
6.2
3.0
4.0 323.6 43.2
2.7
44.1 874.2
Semi skilled laborer and service workers
127.8 177.2 149.1 211.8 265.8 65.7 81.4 79.9
Clerical workers
1124.2 1113.4 989.3 726.8 543.4 790.7 802.7 461.1
Skilled laborer and machine operators
2.4
0.6
0.7
4.1
96.3
0.9
3.9 202.0
Healthcare specialists
142.1 48.0 38.3 49.3 175.2 22.7 67.5
1.3
Construction workers
0.3
0.2
0.2
27.2 18.5
0.0
0.9
55.1
Special educators and teachers
1.2
0.0
0.0
14.5
7.0
0.6
19.2
0.0
Sales, marketing, and advertisement personnel 525.1 1274.4 415.4 546.3 375.2 1441.1 444.7 172.1
Food preparation workers
0.0
0.0
0.0
10.1
4.7
0.0
1.6
0.3
Law enforcement and safety workers
19.2 10.5 12.8 89.5 264.3 23.7 16.9
8.3
Engineers, technicians, and architects
24.8 12.9
4.0
31.7 89.3
6.0
69.4 123.0
Farming and agricultural workers
0.0
0.0
0.0
2.3
4.7
0.5
10.3
0.0
Earth scientists
0.6
0.0
0.0
1.9
3.2
0.0
1.9
0.0
Source: OES, ONET, and Author’s Calculation
* Values represent the relative importance of occupation clusters; they do not have any absolute meaning.
27.3
0.0
0.0
513.0
4.6
36.0
217.2
17.4
81.0
551.9
8.7
3338.5
0.8
60.4
11.9
27.9
18.4
6.0
11.7
0.0