APPENDIX 2

A Comparison of Micro and Meso-level Approaches

It is possible to illustrate contrasting methodological approaches for identifying clusters by examining the structure of one drawn from the same general region: 13 counties in North Carolina and Virginia that represent part of the Appalachian Regional Commission region (ARC study) versus Western Economic Development Partnership Region (NCACTS study), which consists of 22 North Carolina counties. Apparel and textiles dominate the economies of the common counties in these two regions (44 percent of NC counties in ARC study area also part of NCACTS study). Each approach examined the same general cluster: a knitting mills cluster in the ARC sub-region and a knitted goods cluster in WEDP region.

Liston et al. (1997) produced an in-depth case study for ARC that focused on four main industry sectors (number of firms): knit fabric (8), knit under and outerwear (30), hosiery (88), and miscellaneous knit (6). Suppliers of various key inputs and machinery from related sectors also received attention, all of which together are said to employ about 20,000 workers (percent of regional economy unstated). The four main knitwear sectors listed above were of principal interest when the team selected firms for in-depth interviews. In addition, the ARC knitting mills case study documented the role of various service industries, distributors, industry membership organizations, and public services and infrastructure. The many and various connections, linkages and cooperative agreements among all the contributing cluster elements were detected from face-to-face interviews and reviews of existing documents, as summarized in Exhibit A2.1. The ARC study is a micro-level study while the NCACTS study is meso-level.

We can now compare the meso-level regional trade cluster approach to see how it differs from a micro-level analysis in results and interpretation. First, there are some obvious similarities simply because firms in some of the same cluster sectors are studied: knit fabric (14), knit outer and underwear (3), and hosiery (28) firms are the equivalent sectors. Second, the NCACTS approach includes a more extensive and detailed group of additional sectors that trade with the main knitted goods sectors and with each other: apparel made from purchased knit fabric (74), yarn mills (19), narrow fabric mills (5), pleating and stitching (4), thread mills (4), and a scattering of other sectors that routinely trade with the WEDP knitted goods cluster (Click here for a map of location of WEDP knitting goods establishments.)

Exhibit A.2.2 reports which specific segments of the full knitted goods cluster are concentrated in this region, and which are far less fully represented, relative to the national trade cluster. Since the sectors that trade with each other within the cluster are known, regional development policy might be focused on the most beneficial cluster segments, particularly those that deserve special attention to retain production in the face of overall cluster declines. Recruitment policies can also be adjusted to focus on missing or under-represented cluster segments provided those segments complement or take advantage of existing comparative advantages of the region. Regional development officials can augment their knowledge of regional comparative advantages by considering additional relevant facts, particularly the relative strength of trade connections among all the sectors of this regional trade cluster (as indicated by strength of industry "loadings" similar to those reported above Exhibit 3.7).