The Consequences of Metropolitan Manufacturing Decline: Testing Conventional Wisdom

Alec Friedhoff, Howard Wial, and
Howard Wial Former Brookings Expert
Harold Wolman
Harold Wolman Director, George Washington Institute of Public Policy, George Washington University

December 16, 2010

Between 1980 and 2009 the United States lost 7.1 million manufacturing jobs, about 38 percent of its manufacturing base. It lost most (two-thirds) of these jobs between 1980 and 2005, prior to the Great Recession. More than 61 percent of these lost jobs were in 114 indus­trial metropolitan areas—metropolitan areas that strongly specialized in manufacturing in 1980. This report examines the ways in which the industrial composition of those areas changed dur­ing the 1980–2005 period and the consequences that those changes have had for wage and employ­ment levels in those areas.

Although scholars, policymakers, and journalists have extensively analyzed and debated the causes of deindustrialization and its consequences for displaced workers, they have paid much less attention to the kinds of jobs that have replaced lost manufacturing jobs and to the consequences that industrial shifts have had for metropolitan areas. As a result, assertions about these phenomena have abounded. Some of these have been supported with evidence while others have been backed up only with theory or have simply been asserted without support.

An analysis of employment and wage data for 114 metropolitan areas that specialized in manu­facturing in 1980 and lost manufacturing jobs from 1980 to 2005 finds that:

Two-thirds (76) of the 114 metropolitan areas, mostly in the Midwest, performed worse than the nation as a whole in both job growth and average wage growth from 1980–2005. Only three (Charlotte, Manchester, and Portland, ME) performed better than the national aver­age on both. In general, metropolitan areas in the Northeast had slow job growth but relatively rapid wage growth, those in the South had rapid job growth but slow wage growth, and those in the Midwest had slow growth of both jobs and wages.

The metropolitan areas that lost the fewest manufacturing jobs gained the most non-manufacturing and advanced service jobs. Regardless of whether there is a cause-and-effect relationship between manufacturing and non-manufacturing jobs, the two are complementary rather than competitive.

Between 1980 and 2005, the 114 metropolitan areas typically had faster growth in transportation and warehousing and slower growth in advanced services, tourism, and government than the nation as a whole. The shift of employment toward transportation and warehousing accords with popular perceptions of employment change in these metropolitan areas, although it is uncertain whether that shift contributed to job or wage growth.

Of the 114 metropolitan areas, the typical region’s 1980–2005 job growth rate was 12.8 percentage points lower than it would have been if all its industries had grown at their respective national rates. The sluggish job growth in many of the 114 metropolitan areas, including those that specialized in autos and auto parts, was due more to slow job growth within the areas’ existing industries than to their specialization in industries that grew slowly throughout the nation.

Of the 114 metropolitan areas, the typical region’s 2005 average wage was 6.1 percent lower than it would have been if its industry composition had not changed since 1980. In the typical metropolitan area, employment in low-wage industries grew by 42.5 percent from 1980–2005, while employment in high-wage industries increased by only 11.7 percent. Because of changes in the industrial composition of employment, including the loss of manufacturing jobs, wages in most of the 114 metropolitan areas were lower than they would otherwise have been, but these wage-lowering industry shifts were similar to those that occurred nationwide.

The inflation-adjusted average wage grew by 16.9 percent between 1980 and 2005 in the 38 metropolitan areas that were most industrially diverse in 1980 but by only 9.5 per­cent in the 38 that were least industrially diverse. However, there was no difference in job growth between the most and least industrially diverse metropolitan areas.

The findings hold implications for economic development policy in deindustrialized metropolitan areas. By allowing us to examine and analyze common conceptions of American manufacturing, the findings reveal a solid baseline for economic development policymakers.