Urbanization in the Early American South: A Theoretical Framework (Nov. 1989)

 

 

Introduction

Since the mid‑1970’s, as part of the fallout from the “new urban history,” southern urban history has received a tremendous amount of attention. Unlike most southern historians, these new southern urban historians have not simply focused on how the South was unique, but more importantly on how southern urbanization fit into the national pattern of urbanization. Social science theory has played a major role in developing organizational frameworks to objectively study southern urbanization. However, to date, none of these models has been rigorously tested in a Southern context. The purpose of this paper is to develop a theoretical framework and a set of empirically testable hypotheses of urbanization in the early American South (1607‑1860), derived from available theories and case studies.

Definitional problems common to urban studies everywhere hamper any theoretical framework of early Southern urbanization. What is a city? What is urbanization? To urban historians who define “urban” in terms of late 19th century northern cities fueled by industrialization and immigration, the whole question of southern urbanization is moot. However, such a narrow view of “urban” distorts the understanding of urbanization as a process because it eliminates the critical importance of pre‑industrial urbanization.

The dichotomy between colonial and antebellum studies creates another barrier to the proper study of early southern urbanization. This dichotomy is very puzzling because colonial and antebellum urban historians often seem to be asking the same questions and, indeed, coming up with the same answers. But somehow antebellum historians do not consider the colonial South distinctly “southern” and colonial historians, who have made the most sophisticated use of social science theory, are unwilling to invade the sacred territory of the antebellum South.

The dichotomy between northern and southern urban studies presents an almost equally artificial barrier to the study of early southern urbanization. Most of the comparative studies between North and South have revolved around studies of urbanism, how the quality of urban life was similar or different between North and South. Although southern urban historians might give a footnote or two to northern urbanization research, no southern urban historian has to date used any of the sophisticated theory and methodological tools developed by northern urban historians.

One of the most healthy developments in the “new urban history” has been the increasing use of the regional approach to urban studies and the attention to the effects of the hinterland and city‑systems on urbanization. However, how to define “region” remains problematic. To simply assume the South was one region is simply too facile and contrary to much evidence. A better approach might be to analyze the South, not as a separate economic region, but as part of a greater national region. Scarcely interdependent southern cities shared a parallel “colonial” status in the national city‑system. Assumed similarities in southern urbanization over time and space and assumed differences between northern and southern urbanization qualify the South as a separate unit of analysis for the purposes of this study, but eventually these assumptions of similarities and differences will have to be tested within a national (or perhaps even international) framework.

In this paper, I will try to break down the artificial dichotomies limiting the study of early southern urbanization through the even greater use of social science theory and the development of more sophisticated methodological tools for testing that theory. However, I barely tap the potential for comparative analysis, sticking fairly closely to an American context. Anthropologists like Carol A. Smith, using the same theories presented in this paper, have done studies of modern third world urbanization which show tremendous comparative potential for studies of pre‑industrial American urbanization. This paper will hopefully help push urban historians, both North and South, even further towards the development of an interdisciplinary, comparative approach to understanding the historical process we call urbanization.

Definitions

What is a city? What is urbanization? Although there is no general consensus among historians about qualitative and quantitative measures of urbanization, following the example of Philip Hauser (9) and John Sharpless (5), I will use Hope Tisdale Eldridge’s simple definition of urbanization:

Urbanization is a process of population concentration. It proceeds in two ways: the multiplication of points of concentration and the increase in size of individual concentrations…cities may be defined as points of concentration. (Eldridge 338)

As Eldridge notes, the definition “comprehends the totality of the process both in time and space” (338). Such a simple, “neutral” definition, Leo Schnore and Eric Lampard agree, “‘is not only close to usage but also leaves open the question of the socioeconomic causes and consequences of urbanization'” (Schnore 25). There need be no “lower limits to the size and density which qualify a concentration as a city. There is no clear‑cut level of concentration at which a city suddenly springs into being” (Eldridge 338). The differentiation between “urban” and “rural” inolves an arbitrary cut point (Hauser 9). “Urbanization is a process of becoming” (Eldridge 339).

If, as many urban historians have noted, natural increase played a minor role in early American urban population growth (due to small urban populations, low fertility, high mortality, unbalanced sex ratios, etc. [Pred,1980,30]), then the analysis of urbanization boils down to the study of population redistribution through migration (Morrill 4‑5). Nevertheless, most studies of urbanization ignore migration and simply assume a potential urban population source, a regular migration flow that the city can tap into. Migration theory, which has not progressed much beyond Ravenstein although Everett S. Lee has taken Ravenstein about as far as such theory can go (Lee 47‑57), has not helped the study of urbanization very much. Undoubtedly a greater understanding of the complex relationship between migration and urbanization will require more work along the lines of Morrill’s Monte Carlo modeling.

Urban population growth must be examined in terms of the urban capture of the general migration flow which includes rural‑rural, rural‑urban, urban‑urban, immigrant‑rural, and immigrant‑urban migration. This urban capture is a function of relative social and economic opportunities between rural and urban areas which change over time. In early America, rural‑rural migration by far dominated the migration flow but the urban component became increasingly more important with time. But why population concentration? And why at any particular point, old or new? Most urban historians associate the ability to hold and capture migration with general economic potential. As Pred so aptly states, “the degree of success that cities of [the antebellum] era had in expanding their populations presumably depended mostly on the extent to which the ever changing supply of employment opportunities in their local economies permitted them both to attract migrants and to hold in‑migrants and potential out‑migrants away from other competing cities [or competing towns and rural areas]” (1980,31).

The greatest single contribution to antebellum urbanization came ultimately from immigration, although rural‑urban migration and urban‑urban migration can not be ignored. Why did most of the immigrants settle in cities, particularly since a great number of them were probably rural people in Europe? Even if “push” factors played the dominant role in forcing Europeans to migrate to the United States, where those migrants would initially settle would depend on many factors: initial advantage of major ports of entry, kith and kin who had previously migrated, urban boosters and land speculators, and a wide variety of other sources of information. But eventually the population would distribute itself based on economic opportunites in various regions, urban and rural. Although the great port cities had an initial advantage, their ability to capture a share of migrants would eventually ride on their ability to provide employment opportunities for those migrants.

In terms of the definition of urbanization employed here, most available theory focuses simply on the interrelationship between economic opportunity and point concentration. The main questions asked are why some points of concentration flourish and others do not and what effect that concentration has on the surrounding hinterland and other potential points of concentration. The first part of the paper will examine four major theories used to explain point concentration and its effects: central place theory, Vance’s mercantile model of settlement, staples theory, and von Thunen’s theory.

Central Place Theory

Central place theory provides the most universal theoretical framework for understanding point concentration and the functional relationships between points of concentration. Classical central place theory developed out of the work of German geographer Walter Christaller in the 1930s. To Christaller, place meant the market place:

Market places are sites with social, economic, cultural and other referents where there are a number of buyers and sellers, and where price offered and paid by each is affected by the decisions of the others. (Berry 1).

Assuming a continuous distribution of population over an isotropic plain, competing market places, and rational consumers who will visit the closest market place to save on transportation costs, Christaller projected a hierarchical pattern of central market places. Equidistant between any three central market places would be another market place of the next lower order. Each market place would have a corresponding hexagonal market area; the higher order market place would have a market area three times greater than the next lower order. Businesses and services would locate in market places providing the “minimum‑sized market area necessary to make each of them just profitable” (Berry 63). The highest‑order market places would provide all businesses and services; each lower‑order market place would have fewer and fewer businesses and services. Market places achieve centrality in relation to their market area and thus become central market places, or more simply, central places.

Although highly abstract, Christaller’s theory has been successfully applied across many nations and cultures. In the best overview of modern central‑place theory Geography of Markey Centers and Retail Distributuion, Brian J. L. Berry shows that “the general features of this system repeat themselves in each country, regardless of divergent historical traditions” (89). For Berry, the geography of retail and service business is the “vital, equilibrating interface” between the geography of production and the geography of consumption, and is inseparable from urban and transportation geography. “Intercity exchanges are the connecting strands of a complex economy, and the cities are the points about which such an economy is organized. It is in the cities that the geographies of production and consumption interlock” (Berry 2).

“Cities and towns may arise as specialized producers themselves, but many are supported exclusively by their role as market centers. As such, they are neither more nor less than a cluster of retail and service establishments located in a place that provides a convenient point of focus for consumers who visit to purchase the goods and services they need” (Berry 3). “At any point in time, the geographic distribution of retail and service business in central places approximates an equilibrium adjustment to the geographic distribution of consumers” (Berry 4). In an historical examination of rural America, Berry recognizes that central places were often initially not associated with retailing, but rather with “grist mills, post offices, county seats, and railroad stations.” However, competition between these incipient market centers, under conditions of differential “accessibility to the consuming population,” eventually resulted in a central‑place system (Berry 5).

Although strict adherence to Christaller’s theory would require that central places of equal rank would have the same population, modern central place theorists have often adopted G.K. Zipf’s “rank‑size rule” to describe the great stability in urban population distribution over time (Berry 76). The rule basically states that the population of any regional city is equal to the population of the largest regional city divided by the regional rank of the city in terms of population. Although theoretically weak, “much analysis in the quarter‑century since Zipf did his work leads to the conclusion that the regularity will result wherever the rate of relative population growth of centers at any level of the hierarchy is, on the average, a constant fraction of the rate of relative population growth of the entire hierarchy of centers” (Berry 76). Less developed nation are most likely to disobey the rank‑size rule.

Berry tests his central‑place theory by establishing the existence of a systematic hierarchical relationships. Along with tests of the rank‑size rule, Berry also analyzes the relationship between total market area, total population served, and functional level, “as measured by the number of central functions (i.e., the number of different kinds of retail and service business) provided” (27). The functional hierarchy for the United States, from lowest order to highest order are: hamlet, village, town, small city (county seat), regional city, regional metropolis, and national metropolis. The retail functions of this hierarhy (from village through regional metropolis) are equivalent to: local convenience center, full convenience center, shopping goods center, specialty goods center and secondary wholesale center, and primary wholesale center. Berry notes that the population sizes of each order will vary over different parts of the country, and presumably over time (1967,21).

Mercantile Model of Settlement

Although more universally applied and accepted than any other theory of urbanization, central place theory has been criticized by many geographers, economists, and historians. James E. Vance, Jr. extends the most articulate critique in his excellent monograph, The Merchant’s World: The Geography of Wholesaling. Rather than a deductive theoretical approach like central place theory, Vance proposes an inductive theoretical approach based on the historical record. To Vance, central place theory is a mechanical, deterministic, special case model “transformed by disciples into a general one” (5), predicated upon origins in feudal, self‑sufficient independent agricultural villages (145).

Many American and Canadian geographers have shown that central place theory does not explain frontier urbanization. Focusing on external rather than internal forces, Vance proposes an exogenic “mercantile model of settlement” to explain the development of “New Lands.” (This model is very similar to the “dendritic marketing system” analyzed by anthropologists working with third world development [Smith 34‑36].) “[M]ost interior cities in the United States and Canada have not grown out of the soil but, quite the contrary, were sited in terms of the trading function whose long external links were clear and full before the first lot was platted” (Vance 153). General merchants (unrecognized in central place theory), not requiring as large a threshold as a specialty merchant, flourished on the frontier as the demand for imported goods and the willingness to pay high prices led to the quick establishment of wholesale trade (Vance 81).

Whereas Berry concentrates on retail and service centers, Vance believes the key to urbanization lies in wholesale centers. The “pull of scarcity” or the “push of abundance” or both led self‑sufficient communites into “an external, open‑system for trade,” relying on information exchange and agents of trade (Vance 5). Wholesaling does not follow the endogenic limits of retailing because face‑to‑face contacts are no longer necessary. “The wholesale trade of any city (entrepot) will extend outward as far as the limits of commercial intelligence available to that city’s merchants” (156).

Vance points out the strong “initial advantage” that the first wholesalers have. “By institutionalizing the line of trade, custom replaces chance, and it is posible to maintain ties over a considerable distance without the risk of competitive intrusion” unlike retailing (Vance 21). “[W]holesaling works not on hierarchical lines but on historical ones. Once the ties between the wholesaler and the retailer have been created, they become institutionalized‑‑not on the basis of reducing the intervening distance to the minimum, which is the strong factor in retail trade, but rather on the basis of experience” (Vance 78). Vance also notes that such established wholesalers had great flexibility in adjusting prices to undercut any rising competition (83). In America, this initial wholesaling advantage helped create a “two level city structure, comprised of basic towns and large merchant cities,” with no “middle class of towns” as central‑place theory would predict (Vance 83). “Only with the elaboration of commercial agriculture, in the nineteenth century, and the creation of a manufacturing belt was there the growth of any clear settlement hierarchy” (Vance 155).

The earliest Atlantic coastal cities “were traders’ towns built at what might be called the unraveling point of trade, that place which represented the last point common to a sufficient number of supply lines to the frontier to assure a reasonable institutional income to the trader” (Vance 81). Vance’s “unraveling points” are very similar to Burghardt’s “gateway cities.” “The central place, by definition, is located towards the center of its tributary area, whereas the gateway city is located eccentrically toward one end” (Burghardt 269‑270). “Gateway cities often develop in the contact zones between areas of differing intensities or types of production, along or near economic shear lines..In contrast, the central place, at least in its idealized format, lies within a relatively homogeneous productive region… Since entrance into an extended hinterland is of the essence of a gateway, such a city will tend to be located on a site of considerable transportational significance, i.e. either at a bulk‑breaking point or at a node of transport lines” (Burghardt 270).

Although Vance heavily criticizes central place theory, he believes that the best overall explanation of North American settlement will be found in a joint central place‑mercantile model of settlement, with primacy given to the mercantile model (164).

[I]nitial settlement took place in terms of a mercantile model…As the scale of the trading system enlarged, the mercantile towns grew, which in turn increased the demand for hinterland provision for those towns and for the collection of staples in greater quantity…Only at this stage in the externally‑based system did the central‑place model begin to chracterize settlement. That characterization was limited (1) to the later additions to the settlement pattern, and (2) to those areas chracterized by areally‑based staple production…Subsequently there was parallel growth of settlement in accordance with both models. (Vance 165)

Staples Theory

What Vance calls the mercantile model of settlement can be subsumed in what most geographers, economists, and historians call “staples theory.” Vance himself acknowledges that staples collection usually precedes wholesale distribution and recognizes the important work of Harold Innis, the father of modern staples theory. Staples theorists analyze how economic, environmental, and technological factors, through their deterministic effect on staple production, processing, and marketing, inevitably impact on urbanization. In practice, Vance’s theory and staples theory represent two sides of the same coin: the former concentrating on distribution and the latter on collection.

Robert E. Baldwin’s article “Patterns of Development in Newly Settled Regions” published in 1956 provides an excellent introduction to staples theory. Although Baldwin was primarily writing about modern third world development, many historians have used Baldwin’s framework to great avail in the study of early American development. Baldwin creates a hypothetical comparative model between two regions identical in every respect except “natural resource conditions” (165). The climate and soil of one region is conducive to a labor intensive “plantation crop” and the other region conducive to a capital intensive “non‑plantation crop.” The differences in “production functions” cause the two regions to diverge quite quickly.

In the “plantation crop” region, increasing returns to scale make large plantations the optimum and create a demand for “the labor of very low income groups” capable of performing the simple tasks required” (Baldwin 166‑167). In the “non‑plantation crop” region, the “family‑size farm gives an efficient scale of production” which calls for wealthier migrants who can afford the initial costs of transportation and production (Baldwin 167‑168). In the plantation district, “a reasonably stable hierarchy of export producers emerges within the economy. At one end stand the plantations employing large quantities of low wage labor. The other end of the scale is composed of many small, family cultivators who operate under a tenure system or perhaps own their land. The income level of these farmers is not much higher than the plantation wage, and the possibilities for expansion by these producers is not favorable” (Baldwin 171).

In plantation economies, domestic industries do not arise to capture the import and domestic markets due to the lack of a trained labor force, the flow of capital to foreign investors, “high costs of internal transportation”, and “the lack of other marketing facilities” (Baldwin 172‑173). Also, planters form poor entrepeneurs for leadership in manufacturing. “This [planter] group, because of the unique non‑pecuniary advantage of the plantation life, tends to develop a social antipathy towards occupations in the manufacturing field. And, the low income group possesses neither sufficient training nor the social and economic opportunities necessary to provide more than the occasionally successful entrepeneur. All of these factors tend to restrain the economy from breaking out of its predominantly export‑oriented nature” (Baldwin 174). In the non‑plantation region, “the more equitable distribution of income, which arises as the economy develops its export production, is more favorable for the induced development of domestic industry” (Baldwin 176).

Douglass C. North, the other major figure in modern staples theory, also examines forward and backward linkages generated by the production of various staples. A staple export requiring “substantial investment in transport, warehousing, port facilities and other types of social overhead investment” can have a barrel rolling effect, inducing other exports and the growth of domestic industries, promoting urbanization and increased specialization which will hasten these processes even further (North,1959,946‑947). In general, such linkages do not exist in plantation economies “which will merely develop a few urban areas devoted to the export of the staple commodity and the distribution of imports” (North,1959,946).

Von Thunen’s Theory

Any study of urbanization must examine the relationship between the city and its hinterland. As the city grows and develops, so does the city’s hinterland, and vice versa. The hinterland is the market area of the central market place, the source of final demand, and the growth of the hinterland instigates urban growth through increased demand for urban goods. But the hinterland is also producer and what the hinterland produces affects and is affected by urban growth.

In 1826, Johann Heinrich von Thunen published a theory of the spatial organization of agriculture around an urban market which continues to demand attention. This theory has much in common with central place theory, in Thunen’s assumptions of:

(1) a uniform physical environment; (2) a completely commercial economy in which the farmer was both desirous and capable of maximizing his profit; (3) only one means of land transportation and that with costs directly proportional to distance; and (4) an area in which both market and hinterland were solely dependent on each other for their existence (the ‘isolated state’ as Thunen expressed it). (Gregor 57)

Thunen’s theory predicts that the presence of an urban market will result in concentric zones or rings of declining agricultural intensity as one moves outward from the city, “the outer boundary of each zone being determined by the distance at which farmers practicing a particular farming system could no longer make a profit because of increased transportation costs and thus would have to switch to a more extensive system” (Gregor 57‑58).

Although Thunen never assumed that his model was anything more than an “ideal type” which would allow him to isolate the effects of non‑transportation variables, nevertheless, his theory sparked a world‑wide search for Thunen rings that continues to this day. Although confirmed by most research, “several forms of real‑life modifications of Thunen’s zones have been advanced by researchers: centrifugal and centripetal movements; coalescence; distortion and fragmentation; and most recently, reversal of intensity gradients” (Gregor 62). In particular, technological improvements in transportation have seriously reduced “the economic advantage of proximity to markets relative to other locational forces‑‑principally climate, soil, and terrain” (Gregor 64).

Because of the basic assumption of a “closed system,” Thunen’s theory suffers from many of the same problems as central place theory. However, J. Richard Peet shows that Thunen’s theory applies equally well to an “open system.” “The idea of competition between crops for the occupancy of land forms the basis of the von Thunen theory of crop location” (Peet 284). Thus all crops, whether grown for the local market or for export, must compete for proximity to the market. “[T]o gain a position near the market a crop must give a high rent, and, despite exceptions, this is usually derived from large revenues per acre which are the result of considerable capital and labor inputs per acre” (Peet 284).

Thunen’s theory, concentrating on hinterland response to urbanization rather than on urbanization itself, addresses a different set of questions than either central place theory or the mercantile model of settlement and is compatible with either theory. Whether an urban area is a wholesaling or a retail center is not critical information for determining the impact of urbanization on hinterland agricultural production.

Most of the evidence of Thunen rings in the Unites States occurs in 19th century cities. James Lemon reports that “a search for von Thunen rings [in 18th century Pennsylvania] fails to disclose obvious regional patterns” (185). Lemon believes that early Pennsylvania violated too many of Thunen assumptions to qualify: the land was not homogeneous in quality; wagons competed with river boats as means of transportation; there were two major market centers, Philadelphia and Baltimore, and several smaller ones; exports were a significant factor so the market was not closed; and “most farmers were not optimizers.” “[M]ost farmers continued to produce a wide range of crops and livestock, much for domestic consumption, and emphasized wheat, ‘the standing staple of the province,’ for export. These were key factors obscuring potential patterns of specialization. Had farmers not been so dependent on wheat for their incomes, von Thunen rings of specialization might have appeared more distinctly than they did” (Lemon 185‑186). Besides distance from market, religious affiliation and quality of the land affected crop choice. Nevertheless, Lemon believes that “specialized types of agriculture developed, if only faintly, in certain areas” (193).

Diane Lindstrom, who studied early 19th century Philadelphia, believes that Lemon missed the Thunen rings because the rings were too small. She identifies, for Philadelphia in 1840, radiating outward from the urban core, zones of market gardens, dairy products, and livestock fattening; forestry products; grain without fallow; grain with fallow; butter, cheese, and distilling; and stock raising and industrial materials (111‑112).

Michael P. Conzen finds strong support for Thunen’s theory in his mid‑nineteenth century study of Blooming Grove. In general, land prices and per‑acre farm values increased with proximity to the urban core (1971,74). “That agricultural land on the urban fringe began with wheat, but changed to market gardening and later to incipient dairying, seems fitting evidence that, at the very local scale considered here, von Thunen’s model continues to have relevance for historical investigation” (1971,96).

 

Because of the omnipresence of cotton and tobacco, most of the research on urbanization in the early American South ignores central place theory in favor of a standard staples explanation. In order to understand how we might test for the effects of central places in the South and to identify problematic areas of analysis, we should first examine the more sophisticated attempts by northern historians and historical geographers to find a balance between central place theory and staples theory. Historical geographers like Edward Muller, Allan Pred, and Michael Conzen, concentrating on northern antebellum cities, have helped to fill out Vance’s mercantile model. These geographers have shown the importance of transportation, commercial and information flows, city‑systems, and development of the hinterland in city growth. However, each inevitably faces one of the most nagging problems in the historical application of central place theory: What is “centrality” and how do we measure it? The difficulty of defining centrality has led many geographers to talk instead about “nodality,” but this term can be equally as problematic.

Edward Muller develops a regional urbanization model that he applies to a study of selective urban growth in the Middle Ohio Valley for the years 1800‑1860. “A network of approximately 148 urban centers with populations of more than 100 gradually emerged to service this frontier economy” (Muller,1976,183). Using a regression analysis of the absolute increase in towns’ populations as a function of initial population, he then analyzes why some towns grew more and some less than the general trend. For the larger towns, Muller finds that “the nodality of their locations on the developing transportation network with respect to the pace of rural settlement and expansion of agriculture in the local hinterlands” best explains the “differential growth performances” (184).

Manufacturing activities played only a minor role in the selective growth of towns before 1830, but by 1850 Muller notes that “manufacturing had become an important component of the assortment of functions in the larger towns, and urban population was positively related to the total value of manufacturing production” (188). For the 1850’s, “the distribution of nodality among the large towns was clearly related to the location of new nonprocessing manufactures at the time when these activities were becoming the foundation of continued urban growth” (192). But Muller does not try to establish whether manufacturing led to population growth or vice versa, or whether nodality explained both manufaturing and population growth. Muller can not seem to decide whether nodality is a dependent variable or an independent variable. He equates nodality with population growth rates, the faster growing towns always have the greatest nodality. At first, transportation connections seemed to explain this nodality but by 1850 the burden has fallen on the location of nonprocessing manufacturing.

Clearly, through the concept of “nodality,” Muller is trying to capture the essence of the “centrality” in central place theory, a knowledge of which is essential to any test of locational effects on urbanization. Making reference to H. S. Perloff, et al., Muller states that:

A node is a centering point of component parts. Thus, as the state of being a node, nodality combines both the accessibility and connectivity of a location within a circulation network, which is defined by not only the formal parts of the network but also the characteristics of circulation within it and between other networks. Since these characteristics of movement are partially dependent upon the activities for which movement occurs, nodality at any time is also defined by the circulation requirements of an area’s important activities. (Muller,1977,23n)

As a measurable variable, this concept of nodality is quite imprecise. Perloff et al. are actually much more precise in their definition, noting three components of nodality: first the position with respect to “the over‑all configuration of population nodes or to the national market”; “second, distance from a major population node or metropolitan center”; “and, third, distance from a number of minor nodes or the extent of urbanization within a given broad area” (Perloff 83‑84). “Through the use of ‘nodal indexes’ representing the three aspects of nodality, the role of space in the development of the economy can be analyzed in terms of various measures of activity within the various economic sectors, such as agriculture, mining, manufacturing, etc.” (Perloff 84). However, Perloff et al. note that these distances are more a matter of “relative transportation costs” than miles (85). Although such a definition of nodality is still not entirely unambiguous (e.g., how do we define “major” population node), still Perloff et al.’s concept of nodality is empirically far superior to Muller’s definition.

Robert Higgs in a study of urbanization in four Midwestern states‑‑Missouri, Iowa, Kansas, and Nebraska between 1870‑1900 develops a simple central‑place model. Urban population growth rates are equated with changes in hinterland population density, hinterland per capita farm income, and nodality. Higgs points out the inherent problems in defining the hinterland, which he call the “trade area.” He settles on defining the trade area for a city smaller than 10,000 as the population is the county in which the city is located; for a city greater than, 10,000, he adds all contiguous counties; then he proceeds to make exceptions for the very largest cities. Higgs begs off other possible definitions of hinterland due to the constraining county level data, but reports that “experiments with the data seem to support” that modifications of the definition have little effect (1969,372).

Higgs also contributes to the definition of nodality. “The index of nodality is defined as the number of railroad lines and navigable rivers emanating from a city’s location…For example, a city situated on the Mississippi River and without rail conections would have an index of 2; should it be reached by a railroad from the East its index would increase to 3; and if the railroad should pass through the town and on to the West the index would become 4” (372).

Using multiple regression analysis, Higgs concludes that “the main source of city growth was, as central place theory would have it, the growth of population in the surrounding countryside” (1969,369). Higgs proxies real per capita farm income using censal value of farm products (deflated) but the variable proves statistically insignificant, possibly due to the failure of census data to measure income directly or perhaps due to instability in farm prices (1969,373‑374). He confirms the work of Williamson and Swanson which finds that industrialization played little role in city growth but also finds that “the acquisition of railroad lines‑‑that is, increased nodality‑‑made only a small contribution to city growth” (1969,369). Although Higgs uses his model to support central‑place theory, he inevitably supports staples and mercantile theory. “Settlement in the western prairie region was carried out mostly by people seeking to establish farms for the production of cash crops to export outside the region…the extent and timing of settlement in a given area were quite independent of whether a city in that vicinity were growing fast or slowly” (375). But inevitably the rising farm population stimulated the growth of cities through demand for the city’s goods.

Bonnie Barton argues that the weaker concept of nodality has obscured Christaller’s original concept of centrality. “[C]entrality is defined as the surplus of importance of a place, or the ability of a place to provide goods and services in excess of the needs of its own residents. ‘Neither area nor population very precisely expresses the meaning of the importance of the town.’ This surplus of importance of a place thus emerges in terms of its command over its hinterland” (Barton 34).

“The importance of a town in colonial New England was intimately linked to the town’s involvement with entrepeneurial activity” (40‑42). Barton identifies three levels of centrality in colonial New England towns: agricultural towns with limited mercantile involvement, like Andover; towns “providing services specifically oriented to consumption by the mercantile community” (e.g., inns and taverns), like Dedham; “towns in which the entrepeneurial sector is concentrated [providing] an excess of goods and services created exclusively by these agents,” like Boston and Newport (43‑44). “The relationship between the three levels of centrality are fueled by the most important entrepeneurial towns. Ties between the largest towns and other places are direct. The least important agricultural villages, for example, do not interact with the largest places through the intermediary towns but rather supply their occasional surpluses directly to the mercantile sector. Economic interactions between towns of lower levels of centrality appear to be far less important than their direct access to the most influential places” (44). Unlike Muller, Barton shows that centrality does not correlate very well with population growth (40‑44). Likewise, Ernst and Merrens point out that for the colonial South the two criteria of urban importance‑‑size and function‑‑were not synonomous because “the usual functions of exchange, colletion, storage, and distribution of commodities and manufactures were commonly performed in the smallest urban centers” (555).

Barton’s study of colonial New England urbanization supports Vance’s mercantile model. Indeed she claims that “Vance’s wholesaler…is simply a special case of the entrepeneur” (44). But by placing so much power in the hands of the entrepeneur, Barton bypasses the question of why entrepeneurs succeeded in some places and failed in others and ignores the importance of transportation, communication, nodality, and distance. Also, the hierarchy of functions by which she defines centrality appears a little arbitrary. Why is a tavern necessarily a higher function than a saw mill? Barton provides little evidence for her conclusions about the economic interaction between towns beyond simply noting what functions existed in various towns.

The most promising attempt to relate centrality to nodality in a test of central place theory is Michael P. Conzen’s study of midwestern urbanization from 1850 to 1910. Conzen makes quite creative use of empirically determined measures of interurban interaction, metropolitan accessibility, and hinterland development. As a proxy for interurban interaction, for example, Conzen uses data on stage and train frequencies because “passenger traffic reflects a broader phenomenon of settlement integration than wholesale or retail trade would produce on their own” (1975,380). Conzen also develops factors of metropolitan accessibility (air‑line distance, time‑distance, frequency of mail service between county seats and central city, and sector regionalization of the urban region centered on the metropolis) and local centrality (number of transport routes converging on individual counties from outside and the number and size of towns in a county) which seem strikingly similar to Perloff et al.’s factors of nodality. As measures of hinterland development, Conzen examines “twenty characteristics of population, economic production, trade and service organization” drawn from federal census and state business directories to “represent a sample of primary, secondary and tertiary economic activity as mirrored in structural characteristics of areas. Factor analysis ‘reduced’ these characteristics to a few basic dimensions, whose areal variation summarized the stucture of these three main categories of economic activity” (Conzen,1975,375).

Although offering a very promising approach, Conzen’s confusing analysis and sweeping conclusions are problematic. Most of the accessibility and economic development conclusions are based simply on the case of Milwaukee, since correlations for Indianapolis, Chicago, and Des Moines either show little significance or contradict his conclusions. The conclusions on transportation development seem to contradict the conclusions on hinterland development as conflicting “trends” and “counter‑trends” fight for supremacy. For example, “the external relations of cities in the early Midwest as reflected in some important aspects of transport organisation appear to support Vance’s assertion that settlement maturation involved the later development of a central place system within a broader long‑range mercantile one, and that this process was powered from the upper levels of the regional hierarchy down rather than from the lower ranks upwards” (Conzen,1975,382). Yet, at the same time, Conzen finds strong evidence of early central place development as “smaller centres…’carve out’ their own hinterland” (1975,381). His multiple regression analysis of the effect of accessibility on hinterland agricultural and population development (the “Metropolitan factor”) shows that “a clear shift occurred between 1850 and 1890 towards the increased importance of distance measures related to the central metropolis. This suggests an initial period when hinterlands were essentially unorganzied by their nascent metropolitan centres, despite the importance of the latter as headlinks with external markets” (1975,378‑379). Overall, although Conzen’s approach offers the best hope at present for testing central place theory, his analysis merely proves the difficulty in testing historical central place development and does not constitute a proper statistical test of either staples or central place theory. Perhaps the problem reflects either a poor model or poor data. Or perhaps conflicting trends and counter‑trends truly explain the interdependence of city and hinterland and the need for grander theory.

One of the greatest problems facing historial geographers is the lack of a consistent methodology. As shown with Muller, Higgs, Barton, and Conzen, every geographer seems to reinvent the wheel. Nobody applies the same definition of nodality or centrality. Conzen does not even cite Higgs’s study of the late nineteenth century Midwest in his study of the same region and time period. As often as Conzen’s article is cited, no one has bothered to apply his methodology to any other region or time period. Higgs’s methodology provides the distinct advantage of simplicity but is too simple. Conzen’s methodology provides the required complexity but is too complex. I believe that a middle ground between Higg’s and Conzen’s approaches might be a reasonable alternative. But, regardless of the “best” approach, clearly only by continuing to build on existing approaches will we make progress in sorting out the many enigmas of urbanization that these many approaches have already brought to light.

Alternative Economic and Demographic Growth Models

Economic historians drawing upon theories of urban economics have developed a whole alternative group of models for measuring urban economic and demographic growth in nineteenth century American cities. A general assumption among many urban historians is that the growth or decline of urban areas reflects changing economic opportunities. Rural people move to the city because of relatively increasing economic opportunity in the city. Real growth only comes with real opportunities. Although people might immigrate to the city for various socio‑cultural reasons, these were secondary compared to economic reasons.

These economic models are much less concerned about questions of spatial distribution, centrality, and nodality and rarely draw on geographical theory. (Robert Higgs’s work mentioned earlier, although employing the concept of nodality, really fits much better into this group of economic and demographic models.) However, by both focusing on central market places and market relations, economists and geographers have much in common as both assume rational buyers and sellers and accept neo‑classical economic theory.

Williamson and Swanson find that economic growth in northeastern cities between 1820 and 1870 was unrelated to size of cities and only related to the industrial composition of the labor force for the period 1840‑1850. The economic growth rate was negatively related to age of the city and positively related to hinterland population growth. “[T]he growth of the local rural population normally exerts a positive influence on the relative expansion of the city, but…this ‘hinterland’ effect diminishes over time” (1966,56). Williamson and Swanson believe this shift occurs as falling transportation costs allow a single city or group of cities to dominate the provision of services to rural communities to the detriment of small cities (1966,56‑57), which supports the mercantile model.

Williamson and Swanson find that “the older cities should have had a much reduced likelihood of obtaining the growth which younger cities enjoyed, since their age implies a set of industrial characteristics and social overhead facilities which are removed from recent technological change…The greater the historical discontinuity in structural and technological change in a specific decade, the more likely that newly‑formed cities will have an advantage over the old and that older cities will suffer from an equal disadvantage regardless of age” (1966,57).

Robert Higgs questions Williamson and Swanson’s negative findings of urban scale economies and their measure of economic growth. Although he agrees with their “hinterland hypothesis,” he also criticizes their “extremely crude” measure of hinterland, their use of “the rate of growth of the rural population of the state in which a city is located as a proxy for the growth of its hinterland demand” (1970‑71,209). However, as shown earlier, Higgs proposes only a slightly more sophisticated method of measuring hinterland.

Instead of measures of economic growth, John Sharpless focuses on the much more easily measured demographic growth of American and British cities from 1820 to 1861. He organizes his study around “three interrelated theoretical dimensions‑‑spatial location, aggregate city size and economic structure” (3). Overall, the rate of city growth, as reflected by demographic concentration, “declines systematically with increase in size up to a minimum growth point, then begins to rise again across the very largest cities” (Sharpless 177). The “minimum growth point” for the United States lay between 28,000 and 41,600 population for the period 1830‑1860. This finding supports a limited “initial advantage.”

Similar to Williamson and Swanson, Sharpless finds that “the relationship between labor force mix and urban demographic growth was very weak in all decades under analysis” (220). Overall, the effect of manufacturing activities had an uncertain effect on city growth. However, in a study of Boston and Liverpool, he finds that “there was some indication that changes in the migratory contribution to city growth were related to changes in the tertiary [commercial activity and services] sector of the city’s economy, while changes in the secondary [manufacturing and construction] sector conditioned the nature of the indigenous contribution to total demographic growth,” (255) thus stressing the need to understand the two different components of population growth.

The economic models of Higgs, Williamson and Swanson, and Sharpless agree on some key points. The greatest single factor impacting nineteenth century American urbanization was hinterland population density, although Williamson and Swanson note a declining impact with time. Industrialization had little to do with either the economic or demographic growth during this time period. They disagree on the importance of initial advantage and economies of scale with Higgs taking a pro position, Williamson and Swanson con, and Sharpless an intermediate position recognizing that the very largest cities above a certain minimum population did have an initial advantage. Other factors such as disposable income and nodality were not found to be significant.

Urbanization in the Early American South: Historical Applications of Theory

Goldfield’s concept of “urbanization without cities,” although not very precise, probably reflects an accurate picture of early southern urbanization. “An urban place inhabited by fewer than four thousand persons (by 1860) was more characteristic of the antebellum South than of any other region.” This begs the question “Why?” Goldfield clearly aligns himself with a staples interpretation. The lack of cities was “consistent with the relatively few economic functions such cities performed in support of a staple crop economy” (Goldfield,1981,1015‑1016).

Staples theory clearly dominates the theoretical interpretation of urbanization in the early American South. According to this theory, the South, with a general dearth of industrialization and urbanization, was a classic example of the plantation economy where outside of a “few large cities where the marketing facilities for the export commodity are located…few other trading cities spring up” (Baldwin 173). Vance notes that such an economy “could not support a very elaborate trading structure.” In the South this resulted in “the spreading of a rather even net of lowest order trading towns across the agricultural parts of the South. Above this basic net there was only a small number of wholesaling entrepots at points where goods could be brought in from outside manufacturing regions and effectively distributed throughout the agricultural area” (Vance 100). Because of the initial advantage factor, the existing wholesaling entrepots will “reflect the pioneer pattern of the true entrepot location. Thus, the port alignment, the river alignment, and the regional boundary site will tend to be perpetuated even after the forces that made the site distinctive may have ceased to operate” (Vance 100). “Port gateways,” such as Charleston, dominated the southern hinterland (Burghardt 269). Pred finds that “comparatively few hinterland places of a pronounced urban character were economically interdependent with Charleston, simply because of the low level of urbanization associated with South Carolina’s economy, which was dominated by enclavelike plantations whose owners generally purchased their goods and services directly from the seaboard rather than nearby” (Pred,1981,116).

In applying staples theory to the South, economic historians have noted other specific factors that hindered urbanization. Morton Rothstein claims that social and cultural contexts were also important. “[T]he South, like many other plantation economies, developed a rigid social climate and value system which discouraged many kinds of social investment as well as small‑scale entrepeneurial activity. This is a major point in explaining the relative absence of towns and cities in the cotton region” (93).

Rothstein shows that forward linkages were negligible in the cotton South. Cotton processing did not go beyond simple ginning and pressing into bales on the plantation. “Furthermore, plantation owners frequently provided these services for their smaller neighbors, and made direct shipments to the merchant at the seaport. Little of the processing, storing, and shipping took place off the farm or plantation, and consequently there was much less need for secondary mercantile activity” (Rothstein 94). Warehouses, elevators, and other storage facilities were unknown in the cotton trade; no commodity exchange existed in the South until after War; and cotton factors showed little “differentiation of functions” (Rothstein 95‑98).

Perhaps because of the strength of the staples argument, nobody has yet successfully applied central place theory or even von Thunen’s theory to the early American South. Barton notes that Christaller’s theory and the concept of centrality do not apply to “rural landscapes with small villages…of a similar typical size” which characterize the South (34). McCusker and Menard complain that “central‑place theory, with its emphasis on retailing, its assumption of a largely closed system, and its inability to account for the spatial impact of foreign trade and the demands of particular exports, seems inappropriate to early British America” (251).

But surely central place theory as applied to the nineteenth century North is much more sophisticated than these complaints would lead us to believe. Although central place urbanization theory has not been rigorously tested in the context of the early American South, neither has staples urbanization theory. Historians have used staples theory more as a descriptive tool than an analytical tool. The work of the best northern urban historians indicates that staples and central place theory overlap in studies of actual cities. As Williamson and Swanson, Higgs, and Sharpless all find that hinterland population density played the most significant role in determining northern and midwestern urban growth, surely this factor had some impact on urbanization in the early American South. Rothstein notes that the low population density in the antebellum South did deter commercial and industrial growth (94).

Staples theory offers only a general economic context in which many cultural, economic, and demographic factors, only indirectly linked to the primary staple, impact urbanization. Even the most avowed staples practitioners greatly modify received theory to explain actual historical cases. For analyzing the growth of the earliest American cities, Jacob Price points out the importance of an indigenous merchant class and the location of the “entrepeneurial headquarters,” but shows that these characteristics were most affected by the legal and political norms of the metropolis‑‑the Navigation Acts, the complications of the re‑export trade to the Continent, and long‑term credit availability‑‑that all gave London tobacco merchants an initial advantage which was hard for indigenous merchants to overcome (163‑173). Similar conditions existed in the rice economy of South Carolina (161‑163).

Carville Earle argues that the town as an indispensable English institution superceded its role in staples trade. “Historians of the Carolinas, the Jerseys, Maryland, Pennsylvania, and Virginia tell us that these colonists invariably established towns first and then got to the frustrating business of finding a staple crop” (Earle,1977,35). Political‑economic and demographic variables, particularly timing, quantity, and quality (free, indentured, or slave; single or family; wealthy or poor) of immigration dictated the success or failure of the first cities (Earle,1977,34‑50).

If we are ever to isolate the effect of export staples, we will have to do so in a comparative framework. If we compare northern and southern urbanization, then we will need to analyze significant northern variables in a southern context. To what degree did southern urbanization differ based simply on a lower population density and reduced immigration? Did lower densities lead to differences in degree or differences in kind? Lower population density undoubtedly has something to do with staples production, but density is also a function of social attitudes towards mobility, land holding patterns, natural increase and immigration, and a host of other factors only in part related to staples production.

A major reason to use central place theory is to test the hypothesis, implicit in all historical examinations of southern urbanization, that the early American South was a colonial outlier of a greater urban system, whether English or northern. “The interaction between colonialism and urbanization in the South is evident from the inability of geographers successfully to apply central place theory to the Southern urban system, which evinces a colonial rather than a modern, industrial pattern” (1981,1032n). Regardless of whether one accepts or rejects staples theory, central place theory has a major role in the study of southern urbanization to describe if, when, and how the “city‑system” of the early American South began to diverge from the general American experience.

The South as Colonial Outlier

Historians have long curiously observed that the South in the antebellum era (or indeed the postbellum era) never seemed to grow beyond the colonial status established in the earliest years of the first British empire. A dependence on a “one‑staple” economy made the South a colony first of Great Britain and then of the North and made southern cities more “a colonial outlier of the north‑eastern [or British] regional city‑system than a participant in a southern regional city‑system” (Pred,1981,116). Goldfield, one of the leading proponents of a colonial interpretation of Southern urbanization, states that “Southern cities served as collection points for and funnels to Northern centers and as distribution points for the return flow. This system limited capital accumulation in the region, which reduced the opportunities for the region to develop beyond its colonial economy. And everything limited urbanization” (1980,1029). Implicit in this analysis, although Goldfield would disagree, is the assumption that the South was somehow backwards. But how different was the South? Were not all of the original thirteen colonies in a colonial status to the British crown? Was not the Midwest equally a colony of New York City? In antebellum America, “regardless of its regional location, each major center was involved in one or more important sets of interdependencies in which New York City was a very prominent element. This was in no small part attributable to the credit‑extension practices of members of New York’s wholesaling‑trading complex and the related hegemony of the city with respect to the nation’s importing activity and foreign trade in general” (Pred,1980,121).

The mercantile model of settlement posits that in the earliest years of development, the entire American hinterland was subservient to a few major entrepots. Only with additional economic and demographic growth and development did a true central place system arise. The growth of a central place system mirrors the degree to which regions escape colonial dependency. Central place theorists note that the rank‑size rule generally applies only to more developed nations and would posit that as long as the South lacked a true city‑system and intra‑regional ties to other southern cities, as long as a few major southern port cities acting only as middlemen for London and New York continued to dominate the southern urban landscape, the South could never escape its colonial status.

Was the South then doomed to be a less developed “banana republic” unless a true city‑system developed? Or was a true city‑system concomitant with economic and demographic development? Would southern cities remain colonial outliers as long as the South remained a one‑staple economy, as staples theory suggests? Was southern urbanization dependent on hinterland development as much as northern urbanization? Diane Lindstrom’s study of Economic Development in the Philadelphia Region, 1810‑1850 provides an excellent base from which to understand how southern development differed from northern development. In her “Eastern demand model,” Lindstrom shows how Philadelphia’s initial industrialization developed through the intraregional “extensive reallocation of productive resources” (Preface vii). The development of canals and subsequent falling transportation rates in the early antebellum era allowed the development of mining, specialization in agricultural production, and the abandonment of home manufactures which led to increased demand for urban manufactures. The hinterland was depopulated as changing economic conditions forced farmers to either specialize, move to the city, or move out West. After 1840, interurban trade within the Northeast accelerated this modernization trend in the Philadelphia region (and throughout the Northeast) as Philadelphia was tranformed into a modern industrial city.

Could such a situation as Lindstrom describes have occurred in the South? An examination of some of Lindstrom points from a staples perspective indicates that such changes could not have occurred in the South.

(1) Would the development of canals or railroads have helped promote urbanization in the South?

Even more than the canals, one of the main causes of Philadelphia’s transformation was the ever‑increasing cheap supplies of wheat coming from the West which forced the Philadelphia region’s wheat farmers to become coal and iron miners, truck and dairy farmers, or else abandon their farms. (Further developments along these lines will be examined in the section on von Thunen’s theory in the early American South.) Although agricultural reformers promoted greater specialization and more intensive techniques to improve worn out soils in the older parts of the South, in general competition from the West did not destroy the profitbaility of the older staples and did not lead to an economic crisis. “The South was characterized by its concentrated production for the market of a single export staple with a comparative advantage so great that even in periods of low cotton prices, resources could not receive an equal return from alternative types of economic activity” (North,1959,948).

As long as cotton was profitable, the South could not diversify to other agricultural production; few could foresee any other future than cotton. In such a situation changes in transportation such as canal building would have tended to “reinforce dependence” upon staple crop production and to “inhibit diversifed economic activity” (North,1959,947). Thus the building of a regional railroad system in the South would have done little to diversify the southern economy.

(2) What opportunities existed for specialization and increased demand for local manufactured goods?

According to Pred, Charleston’s failure to diversify was due to inability of Charleston’s merchants to tap into the direct foreign import trade, shipping, or local manufacturing (109). Pred provides a long litany of reasons for this failure provided by “nineteenth‑century observers and twentieth‑century scholars” which include: lack of capital and credit, weather, volatility of cotton trade, conservatism, aristocratic notions, low level of urbanization in the hinterland, preference for “Yankee” goods, lack of skilled immigrants, lack of local water power, and “internal [intraregional] transportation difficulties” (1981,110). However, Pred seems to place most blame on a failure of leadership. Charleston’s “business community, by refusing to place greater emphasis on manufacturing or to diversify its trading practices, failed to provide the leadership necessary to make the South Carolina major center part of a highly integrated regional economy” (1980,113‑114).

But an analysis of the structure of demand in the South shows that such industrial enterprise may not have been the wisest course for an entrepeneur. The South had a large market for simple commodities that could be produced on self‑sufficient plantations and farms and a small market for high quality, imported commodities. “What was missing from the southern market was the demand for manufactures for which the technology permitted production on a modest local scale, in small shops scattered across the countryside” (Parker,1970,117), a “middle‑income market” (Parker,1970,121). “The lack of demand meant a lack of supply,” even had capital, labor, and skills been available (Parker,1970,117).

Moreover, unlike Lindstrom’s Pennsylvanians who bought Philadelphia’s goods, southerners did not prefer southern goods and thus increased demand for manufactured goods would not have promoted southern industrialization. According to Goldfield, “a textile mill owner in Augusta, Georgia, complained that southern consumers invariably preferred northern goods over home products. South Carolina industrialist William Gregg noted ironically that his textile products were ‘more popular in New York adn Philadelphia than at home.’ In frustration and for profit, some southern industrialists pirated northern brand names and marketed their goods as such” (Goldfield,1977,245‑246). So the antebellum southern planter and farmer were quite satisfied to export profitable staples, import “Yankee” goods, and maintain self‑sufficiency in foodstuffs.

Pred focuses on information and commercial flows which inevitably promote innovation (1980,142‑165). But surely the major port cities of the South were directly linked to northern ideas and innovations through trade with northern ports. Pred’s content analysis of antebellum New Orleans newspapers, while supporting his point of a lack of a southern regional city‑system, also shows how intricately the South was linked with northern interests (159). The southern demand for “Yankee” goods indicates a familiarity with northern tastes. Yankee entrepeneurs actually lived in many southern cities. And hinterland cities were in direct contact with the major southern seaports. So why did not northern attitudes towards urbanization and economic development filter down? Amos shows that the presence of entrepenurial talent had little effect on diversification in Mobile. According to staples theorist Jacob Price, “the precise locus of entrepeneurial activity was not primarily a cultural phenomenon (i.e., reflecting a different business ethos in different places), but represented to a considerable degree a rational adaptation to the availability of capital and other resources and the marketing problems of specific commodities” (Price 173).

(3) Did the existing presence of urban areas based on the wheat trade allow Philadelphia to urbanize more effectively than the cotton South?

Any study of industrial urbanization must recognize the historical roots of pre‑industrial urbanization. Earle, Rothstein, and others have shown that the wheat trade promoted much greater urbanization than the typical southern staples. Wheat was much bulkier than other southern staples and demanded greater transportation networks, flour mills, hostelries, etc., which promoted urban growth. There are several anomalous southern examples of urbanization based on the wheat trade‑‑Baltimore, Norfolk,and the many small cities in Virginia’s Great Valley. Thus Philadelphia and the surrounding region would have been much more highly urbanized than most of the South simply based on the wheat trade. With such an urban framework, further urbanization and industrialization would have been muh easier than if Philadelphia had started with a village‑strewn rural landscape. Note how much the Chesapeake, which had developed in the late colonial and antebellum era based on the wheat trade, differed from the rest of South.

(4) How did the presence of slavery inhibit southern urbanization?

Urban historians have destroyed the myth that the slave and slavery were incompatible with urbanization. Most economic historians note simply that slavery made possible the plantation economy which prevented diversification and thus urbanization. For example, when cotton prices were low, some planters would find work for their slaves in the city; but when times were better, the planters depopulated the urban slave population in order to get as many hands in the fields as possible.

Another key historical theme is that the competition from slave labor inhibited free immigration to southern cities. Although in tracing the lack of immigration, it would seem difficult to distinguish between initial advantage of northern port cities, European preference for a more temperate climate, racial prejudice, and non‑competitive southern wages, undoubtedly the presence of slavery was significant. “[E]ven the most successful Southern manufacturers had to recruit labor vigorously in the North, and found the presence of slaves in the labor force made such hiring especally difficult” (Wright 123). Incentives to foster immigration might have been introduced, but “because of slavery, the South had no unified private or social interest in encouraging immigration” (Wright 125). Most Southern planters were “no keener on free labor immigration than on allowing new slave imports” (Wright 125) because new labor would drive down the value of the old labor in which planters had heavily invested. In contrast, land speculation in the North fostered immigration. “All over the North there were people who stood to gain from immigration, from the small‑scale operators and town builders to big landowners and railroads. The effects of these incentives are ultimately related to the rise of cities and towns in agricultural areas” in the North (Wright 124).

(5) How much would improved intraregional and interurban trade have promoted southern industrialization and urbanization?

As Pred might note, if only Charleston could have shipped its rice to Savannah, they could have avoided the middleman and grown rich! Pred shows that the South lacked a regional city‑system as none of the four largest southern cities had substantial trade with any of the others, but it is hard to imagine how trade between New Orleans and Charleston would have changed southern urbanization. What defines the South as a “region?” Does a southern regional city‑system provide economies of scale and opportunites for specialization? If each part of the South is self‑sufficient in foodstuffs, what kind of opportunity for trade is there in foodstuffs? Will one city’s trade suffer at the hands of the other? The key seems to be lack of diversification of production and manufacturing, not lack of a regional city‑system. Economies of scale suffered due to the initial advantage of Northern manufactures, opportunities for specialization were few since the South was already a specialized one‑staple economy, and abandonment of self‑sufficiency was unthinkable.

As mentioned, all this argument implies that the South was somehow backwards. Historians analyze why the South did not urbanize, not in an objective manner, but because they feel the South should somehow have urbanized. Why should the South have urbanized? As part of an independent confederation of states, there were obvious disadvantages in a lack of diversified manufacturing and commercial activities. But as part of greater national city‑system there was no need. In an analysis of the largest American cities between 1820‑1860, John Sharpless identified “the movement of the American distribution toward what is conventionally defined as a ‘well‑ordered’ rank‑size hierarchy.” This occurred “despite strong evidence of continuing regionalization of the urbanization process,” as the South became relatively less urbanized and the North and West more urbanized (159). The United States was clearly moving towards an integrated national city‑system of which the South was an integral part. Only by focusing on the South as a separate region does the lack of urbanization stand out as an anomaly. The South had a great world advantage in the production of certain staples. The South had slaves which made those staples profitable. The South needed neither cities nor manufacturing. Efforts at such were fruitless unless the South was to stand alone. A regional southern economy made no economic sense. The South was only a “region” in the sense that the separate sub‑regions of the South shared a parallel “colonial” status in the national system. The similarities in southern urbanization over time and space and differences between northern and southern urbanization make the South a proper subject of analysis but not as a “region” in the sense implied by studies of regional city‑systems.

Analysis of southern economic development employing any of the theories discussed in this paper would find, albeit for different reasons, that southern urbanization lagged behind northern urbanization. But was the South backwards? Only from a comparative perspective based on modern capitalist standards. Inevitably, though, pejorative terms like “backwards” can only hinder our understanding of the complexities of early American society and how people made sense of and responded to the world around them.

Works Cited