by Alexander Klein

A re-examination of trends in the Manufacturing Belt across the 19th and 20th centuries

Our research on economic geography and agglomeration economies in the United States focused on the decades at the turn of the 20th century. This period was known for large spatial inequalities in the location of manufacturing, when more than 80% was in the so-called Manufacturing Belt - a region including New England, Middle Atlantic, and the Midwest. Nick and I tried to understand what drove the industrial location into the Manufacturing Belt, the economic forces that made it a successful industrial region from the late nineteenth century until the 1960s. Our approach was empirical and for that purpose, we built large data sets by digitising the US Census of Manufactures from 1860 until 2007.

The first paper in this research agenda was published in the Journal of Economic Geography in 2012 and titled ‘Making Sense of the Manufacturing Belt: Determinants of U.S. Industrial Location 1880-1920’. It explained the economic forces behind the existence of the Manufacturing Belt in its formative years at the turn of the 20th century. Specifically, we found that market potential was central to its existence and that its impact came through scale economies and industrial linkages. Natural advantage played a role, but only in the late 19th century and its importance faded away after that. This paper challenged the prevailing opinion that the manufacturing location was driven solely by natural advantage. We resurfaced the older explanation going back to Harris, the father of the market potential concept, that transportation costs and market potential mattered and they were almost as crucial as endowments of natural resources. Our empirical examination confirmed the educated opinion of scholars working in the 1950s and 1960s as well as contemporaries that the Manufacturing Belt was indeed the product of economic forces going beyond natural resources.

This paper was a springboard to our further examination of US manufacturing at the turn of the 20thcentury. In particular, we wanted to know how the location of industries in the cities influenced their labour productivity. For that, we turned to the literature on agglomeration economies which inspired us to look at the role of industrial specialisation and diversification respectively. As before, we used the Census of Manufactures as the primary data source and found that initially, greater industrial specialisation was associated with faster productivity growth and that industrial diversification mattered only for very large cities. Quantitatively, the industrial specialisation in 1890 raised the level of labour productivity by about 4% by 1920. This paper was published in the Economic History Review in 2020 with the title ‘Agglomeration Externalities and Productivity Growth: US Cities 1880-1930’.

After these two papers, we decided to look at the Manufacturing Belt and cities during the entire 20th century. We discovered that before plunging into a century-long investigation, the basic long-run trends in spatial concentration of US manufacturing needed re-examining. New advances in geographical indices had occurred since the conventional long-run geographical indices were produced, and we applied them to the geographical concentration of manufacturing sector between 1880 and 2007. The results showed that whilst post World War II patterns were robust to new geographical indices, the interwar period was different - we did not find an increase in spatial concentration. On the contrary, the geographical concentration was declining. Overall, the long-run geographical concentration of manufacturing sector did not exhibit an inverted U-shape curve, as was thought previously, but a continuous decline. This paper was published in the European Review of Economic History in 2021 with the title ‘Spatial Concentration of Manufacturing Industries in the United States: Reexamination of Long-Run Trends.

This research agenda has been interrupted by Nick’s untimely death. We had been working on a research monograph to explain the life-cycle of the Manufacturing Belt, and its ultimate demise in the 1970s and 1980s the ‘Rust Belt’. The monograph will carry his ideas, and the insights he brought into US economic geography.

About the author

Alexander Klein is Professor of Economic History at the University of Sussex.