The purpose of the index -- and of this issue of IndustryWeek -- is to identify world-class manufacturing regions; not necessarily the fastest growing, most diversified, or the lowest cost regions, but those that are best at manufacturing. Our efforts differ from those that search out the "best" place to locate a branch plant in the United States (too often those are exercises to find the lowest operating costs for a production plant, ignoring the complexity and diversity of manufacturing in a global economy). IW's effort at identifying world-class manufacturing regions is dedicated to capturing the full scope of modern manufacturing practices. There are five disciplines to manufacturing, or parts of the production process, that influence corporate productivity:
- Managing the product.
- Researching, designing, and deploying the product into the marketplace.
- Producing the product.
- Selling and marketing the product.
- Shipping and distributing the product.
Real productivity takes place through the interaction between demand for the company's product, the company's strategy for fulfilling that demand, execution of the strategy, and where the company locates each of the five functions. Manufacturing is a diversified activity, and the type of community that excels at the creation of one type of product may not be the same as one that is world-class for other types of products, or may not be as competitive for all five parts of the production process. Places that are desirable for large foundries, stamping plants, and petrochemical refineries are not the same as those that are good for software or pharmaceutical production. Places that are good for management and sales and marketing may not be the most efficient at logistics or production. Due to this complexity, our index is based on market outcomes so that our prejudice over what is a world-class environment will not distort what the economy is telling us. We let economic outcomes identify world-class manufacturing communities.
The index results are calculated for Metropolitan Statistical Areas (MSAs). These are regional labor market areas centered on one or more central cities. In the United States an MSA is defined as the county that contains the central city with a minimum population of 50,000 plus adjacent counties, as long as at least 15% of those who are employed and live in the adjacent county commute into the central county for work. At times the commuting picture becomes more complicated, where some suburban counties send their workers into two or more counties, and each county contains a central city. Then the terminology changes. In these cases each central city forms the core of a Primary Metropolitan Statistical Area (PMSA) and, together with abutting PMSAs, they form a Consolidated Metropolitan Statistical Area (CMSA). We treat each PMSA on the same footing as an MSA and ignore the CMSA in our calculations. We also calculated the index for a larger unit of geography that is based on MSAs and PMSAs and includes the nonmetropolitan rural counties that have economic linkages to the MSA. These are Consolidated Economic Areas (CEAs) and they are defined by the U.S. Dept. of Commerce's Bureau of Economic Analysis. The reason CEAs are larger than MSAs is that the rural counties that are not part of an MSA, but are part of a CEA, do not have large flows of commuters entering the MSA. These rural counties do have trading and other economic relationships with the MSAs. We have two reasons for examining CEAs. First, an increasing share of investment in standardized manufacturing production is taking place in these counties -- taking advantage of a rural workforce while retaining the advantages of proximity to the economic assets of the adjoining metropolitan area (such as airports and logistic infrastructure). The second is that the economic performance of the rural United States is split into three groups. Those that abut metropolitan areas tend to be economically healthy and growing. Those that are either tourist or retiree destinations, or have strong demand for their natural resources, appear to be performing well. Those that do not share the attributes of the first two groups are losing population and experiencing economic decline. Our interest in the performance of these nonmetropolitan counties as sites of world-class manufacturing led us to calculate the index for just these counties. Therefore, we prepared three databases as part of the
World-Class Communities research and produced three indices: The main index is based on Metropolitan Statistical Areas (MSAs) and Primary Metropolitan Statistical Areas (PMSAs); high performance nonmetropolitan counties were the object of the second database; and the third database is for Consolidated Economic Areas (CEAs).
The World-Class Formula
The World-Class Communities index is composed of five variables, each directly connected to manufacturing productivity. The contribution of each manufacturing worker to manufacturing's contribution to Gross Metropolitan Product (GMP) is our approximation of productivity in 1996. The second variable identifies if a community is truly a specialized manufacturing community, and this is the percent of employment that is in manufacturing industries in 1996. The third variable is the three-year average growth rate in manufacturing employment from 1993 to 1996 -- a result of the decision making of manufacturing firms that shows the areas where industry is placing its bets. The fourth variable is a measure of the importance of the region's manufacturing output to the national economy. We include the metropolitan area's contribution, in percentage terms, to Gross Domestic Product from manufacturing in 1996 because it emphasizes the value of the product rather than employment. The last variable is the three-year dollar change in the metropolitan area's contribution to Gross Domestic Product from manufacturing from 1993 to 1996. We included this variable because a world-class manufacturing community is one that is growing, not declining, and growth in the value of goods produced is key to economic development. Each variable in the formula carries equal weight.
Calculating The Variables
Statistics that directly measure manufacturing's contribution to GMP do not exist, so we estimated them for all of the 314 metropolitan areas. (We repeated these steps separately for the CEAs and the rural areas that are associated with CEAs, but are not a part of MSAs.) We first added together total income at the county level for all counties in an MSA to yield total metropolitan income; this was repeated for manufacturing income. We then obtained estimates of Gross State Product (GSP) and manufacturing's contribution to GSP from the Bureau of Economic Analysis of the U.S. Dept. of Commerce. To estimate GMP we multiplied GSP by the metropolitan area's share of total state income, and to estimate manufacturing's contribution to GMP we multiplied manufacturing's share of GSP by the MSA's share of income generated from manufacturing in the state. In those cases where the MSA crosses state lines, as happens in Kansas City, Cincinnati, Philadelphia, and other places, we calculated GMP and GMP from manufacturing for the part of the metropolitan area that is in each state and then added them together. In the last step in calculating our approximation of manufacturing productivity, GMP from manufacturing for each manufacturing employee, we divided GMP from manufacturing by manufacturing employment in the MSA. The county-level data on total income and income from manufacturing, total employment and manufacturing employment, and population were downloaded from the Regional Economic Information Service's Web site that is maintained by the University of Virginia. Gross State Product also was obtained from this source. The second component of the index is the three-year average rate of growth. Manufacturing-employment growth rates are an important indicator of which MSAs manufacturers think are the most productive. Economists examine "marginal" decision making as a way of determining the direction markets are taking. In the case of our index, manufacturing-employment growth is the best indicator we have of the places where growth in manufacturing competitiveness is taking place. The third variable that entered the index is manufacturing's share of total employment in the MSA. This is a straight-forward measure of the degree to which the MSA is specialized in manufacturing and is a measure of how important manufacturing is to the region's economic base. This variable tends to favor small to midsized metropolitan areas and recognizes those places that specialize in manufacturing production. The fourth variable entered into the index is a direct measure of the absolute importance of the regional economy to the manufacturing base of the U.S. economy. This variable recognizes the facts that the command and control functions of manufacturing and the centers of many interfirm and intrafirm economic relationships are concentrated in larger metropolitan regions. This variable recognizes the nonproduction aspects of American manufacturing. The fifth variable examines the three-year growth absolute dollar change in manufacturing GMP from 1993 to 1996.
The five components of the index are measured on different scales. The first component is dollars per worker; the second the percentage change over a three-year period; the third is the percent of the workforce; the fourth is the metropolitan area's percent contribution to U.S. GDP from manufacturing; and the fifth is the dollar change over a three-year period. In order to combine the five elements, each was converted into a standardized statistical distribution called a "z-score." Z-scores are standardized units of measure indicating the distance each metropolitan area is from the mean, or average, for the variable in question for all metropolitan areas. A metropolitan area's z-score on a particular variable provides its position in the distribution of that variable. Once variables that are measured in different units are converted into z-scores, they can be added together because they now share a common unit of measure -- their position along the statistical distribution. In the case of our competitiveness index, each z-score is given equal weight.
Hill is senior research scholar, The Urban Center, Levin College of Urban Affairs, Cleveland State University, Cleveland, Ohio. Brennan is research associate, The Urban Center, Levin College of Urban Affairs, Cleveland State University.