Ohio History Journal

JON GLASGOW

The Westward Expansion

of the Manufacturing

Belt: The Ohio Machine Tool

Industry in the Late

Nineteenth Century

For more than sixty years the Manufacturing Belt has been recog-

nized as an important feature of the human geography of the United

States. Early studies, mostly by geographers,¹ delimited the current

areal extent of the region-roughly a quadrilateral with corners at St.

Louis, Minneapolis, Portland (Maine), and Richmond. Later, attention

shifted from description to explanation as geographers were joined by

scholars from other disciplines who were also interested in finding

answers to questions about the locational aspects of the Manufacturing

Belt.² Why did it originate in New England rather than elsewhere?

Jon Glasgow is Associate Professor of Geography at The College at New Paltz, State

University of New York.

1. Sten DeGeer, "The American Manufacturing Belt," Georgrafiska Annaler, 9

(1927), 233-359; Richard Hartshore, "A New Map of the Manufacturing Belt of North

America," Economic Geography, 12 (1936), 45-53; Clarence Jones, "Areal Distribution

of Manufacturing in the United States, Economic Geography, 14 (1938), 217-22; Alfred

Wright, "Manufacturing Districts of the United States," Economic Geography, 14

(1938), 195-200.

2. Robert Aduddell and Louis Cain, "Location and Collusion in the Meat Packing

Industry," in Business Enterprise and Economic Change, ed. by Louis Cain and Paul

Uselding (Kent, Ohio, 1973), 85-117; Fred Bateman and Thomas Weiss, "Comparative

Regional Development in Antebellum Manufacturing," The Journal of Economic

History, 35 (1975), 182-208; John Borchert, "America's Changing Metropolitan Re-

gions," Annals of the Association of American Geographers, 62 (1972), 352-73; Andrew

Burghardt, "A Hypothesis About Gateway Cities," Annals of the Association of

American Geographers, 61 (1971), 269-85; Edward Duggan, "Machines, Markets, and

Labor: The Carriage and Wagon Industry in Late-Nineteenth Century Cincinnati,"

Business History Review, 51 (1977), 308-25; Irwin Feller, "The Diffusion and Location

of Technological Change in the American Cotton-Textile Industry, 1890-1970," Tech-

nology and Culture, 15 (1974), 569-93; Jean Gottman, Megalopolis: The Urbanized

Northeastern Seaboard of the United States (Cambridge, Mass., 1961); David Meyer,

20                                                        OHIO HISTORY

Why it expanded westward as far as St. Louis and Minneapolis but no

further? Why did it become internally differentiated with particular

cities specializing in the manufacture of some products rather than

other ones? With the decline of the dominance of the Manufacturing

Belt, recent works have offered prescriptions for retaining existing

industries and for attracting new ones, or policies to ameliorate the ill

effects of industrial decline.³ However, the task of explanation is not

complete; for some important events involved in the extension of the

Manufacturing Belt across the Appalachians into the Midwest remain

to be explained. One such event is the emergence, between 1880 and

1900, of a trans-Appalachian machine tool industry centered in Ohio

(Table 1). This coincided with the first appearance of firms that

specialized in the production of machine tools and with advances in

machine tool technology that were prerequisities for the later develop-

ment of assembly line production of automobiles.⁴ Without the pres-

ence of nearby and innovative machine tool producers, it would have

been difficult for turn-of-the-century automobile firms of the Midwest

to challenge those of New England for supremacy in the national

market, and the subsequent geographical evolution of the Manufactur-

ing Belt might have been very different.

The premise of this paper is that the development of machine tool

production in Ohio was a manifestation of the Industrial Frontier

Hypothesis. In a previous study of the time and place of origin of three

manufacturing industries,⁵ it was shown that in each case: (1) crucial

financial backing had been provided by local entrepreneurs who

"Emergence of the American Manufacturing Belt: An Interpretation," Journal of

Historical Geography, 9 (1983), 145-74; Allan Pred, The Spatial Dynamics of U.S.

Urban Industrial Growth, 1800-1914; Interpretive and Theoretical Essays (Cambridge,

Mass., 1966): Mary Beth Pudup, "From Farm to Factory: Structuring and Location of

the U.S. Farm Machinery Industry," Ecomonic Geography, 63 (1987), 203-22; David

Ward, Cities and Immigrants: A Geography of Change in Nineteeth Century America

(New York, 1971).

3. Barry Bluestone and Bennett Harrison, The Deindustrialization of America (New

York, 1982); Ann Markusen, Regions: The Economics and Polities of Territory (Totowa,

NJ, 1987): Richard Peet, "Relations of Production and the Relocation of United States

Manufacturing Industry Since 1960," Economic Geography, 59 (1983), 11243; Neil

Smith and Ward Dennis, "The Restructuring of Geographical Scale: Coalescence and

Fragmentation of the Northern Core Region," Economic Geography, 63 (1987), 160-82.

4. Victor S. Clark, History of Manufactures in the United States. Vol. IlI,

1893-1928 (New York, 1929), 153; Frederick Geier, The Coming of the Machine Tool

Age: The Tool Builders of Cincinnati (New York, 1949), 11; Joseph Roe, English and

American Tool Builders (New Haven, 1916), 261; Nathan Rosenberg, "Technological

Change in the Machine Tool Industry, 1840-1910," The Journal of Economic History, 23

(1963), 414-43.

5. Jon Glasgow, "Innovation on the Frontier of the American Manufacturing Belt,"

Pennsylvania History, 52 (1985), 1-21.

22                                                  OHIO HISTORY

geographical sequence and is, therefore, consistent with the idea of a

westward moving Industrial Frontier within which conditions were

especially favorable for industrial innovation. Two other pertinent

concepts are David Meyer's idea of "Replicated Systems"⁶ to under-

stand the development of trans-Appalachian manufacturing in general,

and George Wing's "Steamboat Theory"⁷ to explain the development

of machine tool production in Cincinnati in particular. Wing maintains

that skilled labor was critical in explaining the growth of the machine

tool industry, and that Cincinnati had a surplus of skilled machinists in

the 1890s because the recent decline of the local steamboat manufac-

turing industry meant that "... individuals who had received their

training in the old industry had to seek other avenues for their

talents."⁸ This interpretation is plausible for Cincinnati, but it does not

explain why the other major trans-Appalachian centers of steamboat

production-Pittsburgh, Louisville, and St. Louis-did not develop

into rival centers of machine tool production.

According to Meyer, the major industrial cities of the Midwest

originated as producers of manufactured products for relatively self-

sufficient market regions which he called "replicated systems."⁹ All

such regions had similar demands for manufactured products; there-

fore, each nascent regional center had about the same mix of manu-

facturing industries, including producers of machine tools. Differences

in industrial structure among these cities began to develop in the 1850s

and 1860s as individual firms in particular cities took advantage of

declining freight rates and increasing scale economies in production to

capture multi-regional and even national markets.

Each industrial system [i.e. market region] was a potential location for a new

industry. Its location depended on ties to existing multiregional/national

industries that conveyed advantages to some industrial systems and on a

random location of inventors and innovators ... .10

The Wing and Meyer interpretations have interesting parallels with

and differences from the Industrial Frontier Hypothesis. As in the

Industrial Frontier Hypothesis, Wing proposes that the decline of an

established industry, by releasing resources that are not readily avail-

6. Meyer, "American Manufacturing Belt."

7. George Wing, "The History of the Cincinnati Machine-Tool Industry," (unpub-

lished D.B.A. dissertation, Indiana University, 1964).

8. Ibid., 44, 59-60. See also Clark, Manufactures in the United States, Vol. II, 360;

and Roe, Tool Builders, 266-67.

9. Meyer, "American Manufacturing Belt."

10. Ibid.. 161.

Westward Expansion of the Manufacturing Belt                        23

able elsewhere, can provide an advantage for attracting new industries

to a region. A significant difference is that Wing specified skilled labor

released from the demise of an obsolete manufacturing industry rather

than capital and entrepreneurship driven out of primary industries by

the depletion of natural resources. Although Meyer's interpretation

and that of the Industrial Frontier both recognize the importance of the

location of invention and innovation in explaining the location of

manufacturing industry, Meyer describes innovators and inventors as

having been randomly distributed among places. Contrarily, the idea of

an Industrial Frontier connotes that innovators and inventors, along

with skilled laborers and entrepreneurs, migrated to the Industrial

Frontier in sufficient numbers to create a region that was especially

conducive to the creation of new manufacturing industries.

Combining the ideas of Meyer and Wing with that of the Industrial

Frontier provides the following plausible, but as yet unsubstantiated,

explanation for the emergence of the Ohio machine tool industry. The

first machine tools produced in the United States were made as a

sideline by Philadelphia and New England firms that needed machine

tools themselves to manufacture products such a textile machinery,

firearms, and locomotives. Some of the more innovative of these firms

capitalized on the expanding market for their sideline by shifting to

machine tools as their major product." As demand for manufactured

products increased in the trans-Appalachian west, Meyer's "replicated

systems" emerged including firms that would evolve into machine tool

producers by the same process. Then, just as critical advances were

being made in machine tool technology, conditions conducive to

industrial innovation spread westward with the Industrial Frontier into

Ohio, accounting for its rise to first place in machine tool production.

Following Wing's idea, Cincinnati became the leading city in machine

tool production, surpassing rival Ohio industrial center Cleveland,

because of the special advantage of skilled labor released by the

cessation of steamboat production in Cincinnati.

11. Thomas Cochran, Frontiers of Change: Early Industrialization in America (New

York, 1981), 61; George Gibb, The Sacco-Lowell Shops: Textile Machinery Building in

New England 1813-1944 (New York, 1969), 179; Guy Hubbard, "The Machine Tool

Industry," in The Development of American Industries: Their Economic Significance ed.

by John Glover and William Bouck (New York, 1933), 510-19; Roe, English and

American Tool Builders, 114-20; Nathan Rosenberg, Technology and American Eco-

nomic Growth (New York, 1972), 99; Merritt Smith, "John Hall, Simeon North, and the

Milling Machine: The Nature of Innovation among Antebellum Arms Makers," Tech-

nology and Culture, 14 (1973), 573-91; W. Paul Strassman, Risk and Technological

Innovation: American Manufacturing Methods during the Nineteenth Century (Ithaca,

N.Y., 1959), 117-30; Robert Woodbury, Studies in the History of Machine Tools

(Cambridge, Mass., 1972).

24                                                     OHIO HISTORY

Perhaps the most conjectural part of the above explanation is the

assertion that, as early as 1880, conditions in Ohio were unusually

favorable for the establishment of firms specializing in the production

of machine tools for national markets. This study provides evidence

that such favorable conditions did exist, as indicated by the concen-

tration in Ohio of what are hereafter referred to as Precursor Indus-

tries-industries presumed to have been strongly linked to machine

tool production and which provided the preconditions for the develop-

ment of a machine tool industry with a national market.

Eighteen-eighty has been selected as the appropriate year for this

study, for the following reasons. Consistent with the idea of an

Industrial Frontier, 1880 has been identified as the year in which

Skilled mechanics in the workshops of New England and Philadelphia ...

began to cross the Alleghenies in increasing numbers. Settling in ... Cincin-

nati, Cleveland, and Hamilton [a satellite of Cincinnati] they prospered so well

that Cincinnati . . . ultimately ousted Philadelphia to become the machine-tool

making capital of America.¹²

Furthermore, 1880 has been described as ".. a significant bench-

mark ... in terms of technical innovation leading to mass production

...;"¹³ the year in which " ... a group of firms whose principle product

was machine tools ..." first appeared;¹⁴ and as the initial year in a

three-decade era "... characterized by an immense increase in the

,,15

development of machine tools for highly specialized purposes ...."15

From the 332 industries for which data are published in the 1880

Census of Manufacturing,¹⁶ twenty-two were identified as being pre-

cursors of the machine tool industry. Nine of these were described in

Fitch's seminal 1880 census monograph¹⁷ as being either industries that

required automatic or precision metal working, or used interchange-

able parts: agricultural implements, ammunition, clocks, cutlery and

edge tools, firearms, hardware, railroad and street cars, sewing ma-

chines, and watches. Seven industries were added because of their

apparent similarity to, or relationship with, the first nine: watch cases,

12. Lionell Rolt, A Short History of Machine Tools (Cambridge, Mass., 1965), 176.

13. John James, "Structural Change in American Manufacturing, 1850-1890," The

Journal of Economic History, 43 (1983), 440.

14. Ross Robertson, Changing Production of Metalworking Machinery: 1860-1920,

Studies in Income and Wealth, Vol. 30 (New York, 1966), 484.

15. Rosenberg, "Technological Change," 433.

16. Department of the Interior (Census Office), Report on the Manufactures of the

United States at the Tenth Census: (June 1, 1880): General Statistics (Washington D.C.,

1883), "Table II: The United States by Specified Industries," 19-24.

17. Charles Fitch, "Report on the Manufactures of Interchangeable Mechanism," in

Department of the Interior, Manufactures of the United States, 1880.

26                                                        OHIO HISTORY

1880; and the foundry supply industry for its presumed connection to

the foundry and machine shop industry.

Data pertaining to nineteen Precursor Industries¹⁹ were compiled for

Cincinnati, Cleveland and ten other cities²⁰ selected to create three

groups of four cities each (Table 2). The Northeast group consists of

the four cities that ranked second through fifth, after Cincinnati, in

machine tool production in 1900.21 Cincinnati and the three other major

trans-Appalachian centers of steamboat production constitute the

group of River Cities.²² The Lake Cities include Cleveland and the

three other cities on the Great Lakes that had, in 1880, numbers of

manufacturing employees most similar to the number employed in

Cleveland.23

It is reasonable to assume that the most favorable conditions for the

development of machine tool production were in those cities with labor

forces possessing the greatest variety of skills and expertise in Precur-

sor Industries as measured, initially, by the total number of Precursor

Industries found in a city (Table 2, column 1). Note that by this

measure Cincinnati and Cleveland each rank first in their respective

groups and that their numbers are within the range of the cities of the

Northeast group. Since a Precursor Industry in a city may have

consisted of only one small and inefficient firm with a market limited to

that city, the numbers listed in column 1 of Table 2 may be misleading.

Therefore, an attempt was made to identify cities in which Precursor

Industries served larger than local markets and cities in which Precur-

sor Industries were unusually productive. In a city in which a Precur-

sor Industry's percentage of the local manufacturing labor force

exceeded that industry's percentage of the national labor force (i.e.,

cities in which the location quotient²⁴ for that industry exceeded 1.0),

19. Three of the twenty-two Precursor Industries (ammunition, clock cases, and

watches) were not found in any of the cities examined in this study.

20. Although the more familiar city names are used throughout this study, the data are

for the counties in which the cities are located. County data were preferred to city data

(both were available in the 1880 Census of Manufacturing) so that factories located just

beyond the corporate limits of cities would be included in the study.

21. Edward Sanborn, "Metal-Working Machinery," in U.S. Census of Manufactur-

ing 1900, Vol. X., Part IV, Special Reports on Selected Industries (Washington, D.C.,

1902), 385.

22. Louis Hunter, Steamboats on the Western Rivers: An Economic and Technolog-

ical History (New York, 1969) 105-07.

23. Francis Walker, "Remarks on the Statistics of Manufactures," in Department of

the Interior, Manufactures of the United States, 1880: General Statistics, xxiv-xxv. The

appropriate data are Cleveland 21,724; Milwaukee 20,886; Buffalo 18,021; and Detroit

16,110.

24. The location quotient is a commonly used measure of the extent to which

phenomena are concentrated in particular places. See, for example, George Schnell and

28                                            OHIO HISTORY

numbers in the third column of Table 2 do not follow the pattern that

appears in the first two columns. However, it is worth noting that Cin-

cinnati ranks first in this column, and a close second in the other two.

In general, the data arrayed in Table 2 show that Cincinnati and

Cleveland had numbers of Precursor Industries comparable to the

major machine tool centers of the Northeast, and distinctly higher

numbers than the other cities in the "River" and "Lake" Groups. The

validity of this generalization becomes more apparent by examining

Table 3 in which Cincinnati and Cleveland have been merged with the

Northeast to form a group of six Machine Tool Cities, leaving the

remaining three River Cities and three Lake Cities to be combined into

a group of six Non-machine Tool Cities. Note that this regrouping of

cities precisely discriminates between cities with higher and lower

numbers of Total Precursor Industries (Table 3, column 1), and misses

by only one city (Buffalo) in discriminating between cities with higher

and lower numbers of Extensive Market Precursor Industries (Table 3,

column 2). Except for Cincinnati, however, the Machine Tool Cities do

not appear to be significantly different from the Non-machine Tool

Cities in numbers of More Productive Precursor Industries (Table 3,

column 3). This somewhat unexpected outcome with respect to the

More Productive Precursor Industries led to the following refinement

in classifying Precursor Industries (Table 4).

Table 4 shows the Precursor Industries arranged in descending order

by the number of sample cities in which each occurred in 1880. Since

they were found with equal frequency in Machine Tool and Non-

machine Tool Cities, the first eight industries are called General Pre-

cursor Industries. The other eleven industries, each found in less than

half of the sampled cities, are located almost exclusively in Machine

Tool cities. Evidently, these eleven Specialized Precursor Industries

had demanding labor, technology or market requirements that were

found in only a limited number of manufacturing centers in 1880.

Machine Tool and Non-machine Tool cities are nearly identical with

respect to numbers of General Precursor Industries (Table 5). In the

aggregate, each group has exactly the same total number (41), very

nearly the same number with Extensive Markets (18 and 21), and very

nearly the same number of More Productive ones (24 and 21).

Likewise, examining the numbers of General Precursor Industries

listed for individual cities in Table 5 reveals only slight differences

between the two groups of cities. The total numbers range from 5 to 8

for Machine Tool cities and from 4 to 8 for Non-machine Tool cities;

the number with extensive markets range from 2 to 4 and from I to 6

respectively; and the number of More Productive, General Precursor

Industries ranges from 1 to 7 for Machine Tool Cities and from 0 to 5

Westward Expansion of the Manufacturing Belt               33

ized Precursor Industries, none of which occurred in enough cities to

permit valid rank-order correlation of within-industry relationships, the

sixth test used data for eleven individual firms distributed among six

separate Specialized Precursor Industries in seven different cities-the

six Machine Tool cities plus Milwaukee. These are the eleven cases in

which a specialized Precursor Industry was represented in a city by a

single establishment.²⁶ The seventh test includes data from all twelve

of the sample cities for all manufacturing industry in the aggregate, not

limited to Precursor Industries. As expected, all coefficients between

value added per employee and capital intensity were positive, although

only two of these were significant at the 0.05 level. Surprisingly, five of

the seven correlations between productivity and size of establishment

were negative, with three of the negative correlations significant at the

0.05 level. None of the correlations between capital intensity and

establishment size were statistically significant, but five of the seven

coefficients had the expected positive sign. Of course, the positive

correlations between productivity and capital intensity, using these

cross sectional data, cannot be interpreted as verifying the idea that

increases in capital intensity over time caused increases in productiv-

ity; however, they do suggest that it would be worthwhile to acquire the

necessary longitudinal data. Likewise, negative correlations between

size of establishment and productivity need not be interpreted as

evidence that as establishments increased their labor productivity they

reduced the size of their labor forces. Perhaps 1880 was a year in which

the most productive establishments in these industries had the fewest

employees because they were the newest establishments, with their

higher productivity based upon a greater willingness of such newer

establishments to accept the risks involved in being innovative.

It is obvious that more needs to be done to provide conclusive

answers to the questions raised. Nevertheless, it is hoped that some

progress has been made in understanding the process by which the

Manufacturing Belt was extended across the Appalachians into the

Midwest. Specifically, a set of industries has been identified as being

ones that provided preconditions for the development of a national

market machine tool industry in a city. Furthermore, consistent with

the idea of a westward moving frontier of optimum conditions for

industrial innovation, evidence has been provided that, in 1880,

Precursor Industries were not randomly or uniformly distributed

among trans-Appalachian cities; rather they were concentrated in

26. The disclosure rule forbidding the publication of data for individual firms was not

in effect in the 1880 Census of Manufacturing.

34                                            OHIO HISTORY

Cincinnati and Cleveland-historically and geographically between the

innovation of mass produced steel, by the Bessemer process in

Pittsburgh in the 1870s, and the innovation of mass produced automo-

biles on assembly lines in Detroit in the 1900 to 1910 decade.