A Look at Connecticut's Industry Clusters
By Daniel W. Kennedy, Ph.D., Senior Economist, Patrick McPherron, Ph.D., Economist, and Nicholas A. Jolly, Economist, DOL

The state of Connecticut has been active in the identification and support of the industry clusters located within its borders. In 1998, a task force of business leaders was established that identified six industry clusters in the State.¹ Following research conducted by this task force, legislation was implemented in 1998 that led to Connecticut's Industry Cluster Initiative under the direction of the Department of Economic and Community Development (DECD). There are currently nine industry clusters identified in Connecticut and supported through seed money provided by various sources. The nine industry clusters are:

• Aerospace
• Agriculture
• Bioscience
• Insurance and Financial Services
• Maritime
• Metal Manufacturing
• Plastics
• Software and Information Technology
• Tourism

Dr. Michael Porter of the Harvard Business School defines industry clusters as "geographic concentrations of interconnected companies, specialized suppliers, service providers, and associated institutions in a particular field that are present in a nation or region." Porter asserts that a sophisticated local demand can encourage competitive industries to maximize the benefits of geographic aggregation, providing the economies of scale necessary for the cluster's net exports to create wealth in the region. Essentially, he is applying fundamentals of the academic model, which combines intense competition to publish in the best journals with strong informational spillovers, to the business world. The prototypical examples may be Wall Street or Silicon Valley.

A key component of Connecticut's Strategic Two-Year Workforce Investment Plan is economic development promoted through the State's Industry Cluster Initiative. The industry cluster groups provide valuable information to State and local workforce development providers regarding the training and hiring needs of businesses in the cluster industries. To effectively plan to address these needs, the regional workforce investment boards need to have information about the cluster industries in their regions. To identify various aspects of regional employment in the industry clusters, the industries that make up each cluster had to be identified. Analysts and economists within the Department of Labor's Office of Research reviewed the criteria for industry clusters and the approaches taken in other states, and settled upon a set of industries (based on the North American Industry Classification System) for each Connecticut cluster. While the component industries identified for each cluster may be modified, this initial work has made it possible to provide needed information about the industry clusters for workforce development planning.

Out of this effort, a report was developed (Connecticut's Industry Clusters, http://www.ctdol.state.ct.us/lmi/occ_papers.htm) that provides perspectives of employment in each of the State's industry clusters, including overall employment, number of businesses, and average wages, the concentration of employment in the State relative to the nation (location quotients), and the top occupations in terms of level and growth in demand for workers. Following are some of the findings from that report.

Employment

Employment in the nine clusters represented approximately twenty-seven percent of total State employment in 2003. There was no one individual cluster that comprised more than ten percent of total State employment. Table 1 shows the percentages listed in descending order. As can be seen, the insurance and financial services cluster comprised 8.0% of total State employment in 2003, making it the largest of all nine clusters.

Employment Concentration in CT

The location quotient (LQ) of employment identifies the relative concentration of employment in an area compared to a larger area. In this case, the industry cluster employment in Connecticut was compared to U.S. employment. When the LQ is greater than one, the cluster is typically thought to have a possible local specialization and competitive advantage. These clusters generally provide products and services to consumers outside of their region and, in turn, return revenues (wealth) to the region. Table 2 gives the LQ for the nine clusters for 2003 (the latest year for which US annual average data was available when the calculations were made). As can be seen, five industry clusters had a higher concentration of employment within Connecticut than in the nation as a whole, identifying these as export-producing industries. Aerospace had the highest employment concentration in the State, with an LQ of 5.20.

Businesses, Jobs and Wages

Table 3 presents 2004 data for each of the industry clusters. Things to consider when viewing this data are that these reflect all businesses, large and small, and all workers, from executive management to maintenance and clerical support. With respect to the percent change figures for 2000 to 2004, readers should be aware that the State's economy was at its peak in 2000 and entered a downturn in 2001 from which it has not yet fully recovered the jobs lost. These will affect the interpretation of the data. For example, the insurance and financial services cluster shows, by far, the largest number of businesses, and the number has grown between 2000 and 2004. One reason is that this cluster includes many independent agents. Another example is software and information technology, which is the only cluster to show a decline in average wages. The dot.com bust and end of the Y2K frenzy of activity in the late 1990's are the prime reasons for the decline in wages, as well as in the number of businesses and jobs in the sector. Clearly, more analysis needs to be done to better understand the trends of the industry clusters in these areas.

Occupations

The primary occupations used in each cluster were also identified. This can be viewed in several ways: those occupations providing the most current jobs, those for which the most new jobs are expected to be created, those for which jobs are not plentiful but are expected to grow the fastest, and those in which the most openings are anticipated. Each perspective is provided in the report.

Table 4 ranks the top five occupations within the clusters in terms of numbers of jobs in the cluster in 2002. It also shows each cluster's projected annual need (most openings) for workers in the occupations through 2012. These openings would include jobs in which departing workers will need to be replaced as well as openings arising from the creation of new positions. Also shown, where available, is the average annual wage paid to all Connecticut workers in each occupation.

Clusters are not Frozen in Time

Having described and detailed these elements of Connecticut's nine industry clusters, it is important to address that clusters are not frozen in time. That is, a cluster is a motion picture, not a snapshot. Once formed, the set of inter-firm networks changes through time. Clustering is the result of interactions between or among a number of firms, usually within geographic proximity, and in the same, or related industries. Once this dynamic reaches 'critical mass' and is sustained, it passes through a series of stages. Thus, factors that once gave rise to the birth and rapid development of a cluster may no longer exist. Further, two major threats may endanger the continued viability of a cluster: (i) Internal Threats arising from structural rigidities, and (ii) External Threats that can arise from cyclical disturbances, fundamental technological changes, and competition from other clusters.² In addition, two important longer-wave phenomena can heavily impact a cluster's birth, life span, and death: the Product Cycle and the Technology Cycle, including Technological Discontinuities. A case in point is Boston's Route 128 in the late 1980's.³ Rapid increases in the computing power of PCs in the last half of the eighties suddenly rendered the minicomputer obsolete. Overnight, the thriving minicomputer cluster along Route 128, which included such successful firms as DEC and Data General, came to an abrupt end in a fit of 'creative destruction.'

On the other hand, clusters can reinvent themselves, and come back from the brink. Further, declining employment may not necessarily indicate the maturity or impending decay and death of a cluster. It may reflect a period of rapid technological change, where member firms are substituting capital for labor and raising productivity, while maintaining, or even increasing, their output and sales. The point to keep in mind is that, just because a cluster exists does not automatically imply that it is in its embryonic, or rapid-growth stage. Thus, for any selected time-period, a given industry cluster can be in any stage of its life cycle, depending on the factors discussed above.

Summary

Workforce development is increasingly critical to economic development aimed at making industry clusters more competitive. Industry cluster analysis is a demand-driven labor force initiative that assists the workforce investment system in identifying the training and recruitment assistance needed to supply businesses with skilled workers, as well as the training and support needed to increase worker opportunities for sustained, rewarding employment. Additionally, identifying clusters helps industries to identify and foster strategic relationships with other industries to reap the benefits of informational spillovers. Finally, the analysis of industry clusters allows for an additional informative perspective of workforce demand and supply in the marketplace.

¹ Industry Clusters. "Industry Cluster Initiative." Department of Economic and Community Development. August 2, 2004.

² Karlson, Charlie, Börje Johnsson, and Roger R. Stough, "Industrial Clusters and Inter-Firm Networks: An Introduction," INDUSTRIAL CLUSTERS AND INTER-FIRM NETWORKS, Charlie Karlson, Börje Johnsson, and Roger R. Stough Editors, Edward Alger Publishers: Northampton, MA (2005).

³ Buenda, Fernando, "Toward a System Dynamics-Based Theory of Industry Clusters," INDUSTRIAL CLUSTERS AND INTER-FIRM NETWORKS (2005).