The Market Network in the American West

In my last post, I discussed how John Mohr used network analysis to reveal the cultural logic of welfare provision in New York City in the late nineteenth and early twentieth centuries. This time around, I want to talk about how we can use these same tools to help us think about markets and the political economy of market building—a central topic in my work on the Populist movement and the origins of electoral populism in the American West. As the notion of market building suggests, the key idea going forward is that markets are made, not given. In the case of the United States, the construction of a national market was virtually synonymous with the expansion of the railroad. As scholars such as Richard Bensel and Heather Cox Richardson brilliantly describe, this process was carried out over the course of the late nineteenth century as part of a broader Republican-led development program, which offered public land to both settlers and railroad corporations alike. This led to a dramatic increase in the scale at which agricultural commodities such as wheat—the classic frontier crop—were produced and distributed. The effects of this process were reflected in the relationships around which the market came to be organized.

To get a better feel for how things changed in the lead up to the Populist movement, it is worth thinking about markets from a historical, if highly stylized, perspective. To paraphrase Richard Swedberg, when we talk about markets in history, we talk about concrete places in which individual people gathered to exchange goods with one another. Against this backdrop, the emergence of national markets can be understood as the product of legal and technological innovations that allowed trade to be sustained over long distances. From this viewpoint, it is easy to see where the railroad came in. But the railroad did more than simply stretch the geography of the marketplace. This was especially true in the United States, where private companies were more or less left alone to dictate the trajectory of infrastructural development. In the absence of serious regulation, railroad companies were free to construct a system suited to their needs—needs that were very different from those of the communities and agricultural producers that they served. To put it another way, the railroad network in the United States was, for all intents and purposes, designed to maximize corporate profit. The result was a system that privileged high volume and long hauls, contributing to the formation of a steep urban hierarchy, with vast quantities of grain, and other such products, flowing from the rural hinterland into distant terminal markets.

Railroad companies played an active role in determining where lines would go and, just as importantly, where trains would stop. In places such as southern Dakota Territory, railroads were at the forefront of the settlement process. One of the primary ways that railroad corporations encouraged settlement was to create townsites. In addition to attracting settlers, townsite development in Dakota Territory meant linking agrarian producers to would-be grain buyers. By the time the territory was settled, the amount of grain being produced was much greater than the amount that could actually be consumed locally, leaving the excess to be sent to urban centers such as Minneapolis. This created opportunities for various market intermediaries, whose primary job was to buy, sell, and ship grain.  Historically, grain was shipped in sacks. This practice was progressively abandoned following the advent of the grain elevator, which allowed large quantities of grain to be loaded directly into railcars. The problem for farmers was that these elevators were often owned and operated by large grain merchants who colluded with railroads to influence the terms of trade and limit the degree of access to the market. This worked to the disadvantage of agrarian producers, who had littler choice but to accept whatever prices the local elevator operator was willing to offer. When economic fortunes began to decline in the late nineteenth century, farmers throughout the agrarian periphery turned to organizations such as the Farmers’ Alliance and the People’s Party in effort to encourage regulation, if not the outright nationalization of transportation infrastructure.

In much the same way that Mohr thought about the cultural logic of welfare provision in terms of the bipartite relationship between types of people and types of relief, we can think about the structure of the South Dakota grain market in terms of the tripartite relationship between rail lines, towns, and grain elevator owners. More importantly, we can actually recreate this network using information available in historical documents! This is precisely what I did in my book, which examines, among other things, the connection between market building and the rate of Populist organizing in the East River region of South Dakota in 1890. Toward this end, I combined information on the connection between rail lines and towns from the Rand McNally Business Atlas and Shipping Guide with information on the connection between towns and elevator owners from the annual report of the South Dakota Board of Railroad Commissioners. The resulting network can be seen in the figure below.

I use solid lines to demarcate sets of nodes that belong to the same rail system. In some cases, the system consists of a single, such as the Chicago, Milwaukee, and St. Paul (CMSP). In other cases, however, the system consists of multiple lines bound together through interlocking directorates (a topic that I will likely return to in a future post), as was the case with the Chicago and Northwestern and the Chicago, St. Paul, Minneapolis, and Omaha (CSPMO). The first thing that stands out about this figure is the sheer size of the CMPS and CNW + CSPMO clusters. This is a byproduct of the amount of rail line that each company controlled, keeping in mind that these data refers solely to lines being run in South Dakota. The story of market building in South Dakota was, in many respects, a story about the rivalry between the Milwaukee and the North Western. The other thing that stands out about this figure is the extent to which towns (in blue) and owners (in green) tended to be tied to a single line (in pink). Looking at the graph, we see that there were only a handful of owners that partnered with more than one line. Most towns were connected to a single line, though there were a number of burgeoning towns and cities that benefit from multiple connections.

There is little ambiguity about the fact that railroads were the organizing force behind the western grain market. In fact, their influence is so strong that the network above continues to hang together even if we drop all the rail lines from the data. What makes this even more remarkable is the fact that nearly three-quarters of the elevator owners in the data were tied to a single locale, which means that they did nothing to hold the network together. The reason why the network maintains its structure in the absence of rail lines is because of the large companies that operated elevators throughout region. The relationship between towns and owners was characterized by a fundamental duality, in the sense that position of any given town can be understood in terms of the owners to which it was tied, while the position of any given owner can be understood in terms of the town to which it was tied. In an effort to highlight these connections, I use a two-mode blockmodel to represent the market network. The advantage of this approach is that it retains all of the original data. Rather than converting data on the relationship between towns and owners to a one-mode similarity matrix the way that Mohr did, I treat the data as a bipartite graph, which allows me to calculate similarities for both towns and owners at the same time.

The key to making this work is selecting a similarity measure that ensures that towns and elevators have a similarity score of zero, thus maintaining the distinction between the two modes. With this in mind, I use Jaccard similarity, which measures the number of common ties between any given pair of entities as a share of the total number of unique ties between them. I then use a community detection algorithm to identify to groups of structurally equivalent nodes. Once nodes are assigned to groups, we can construct a simplified representation of the underlying network based on the density of ties between groups, as shown below. If the density of ties between a pair of blocks is greater than the density of ties in the graph as a whole, those blocks are treated as tied. For the sake of visualization, I have omitted minor blocks containing five nodes or less. The graph looks more or less like what would expect, in the sense that ties between towns and owners are concentrated among nodes that are tied to the same line, even though the lines themselves are no longer part of the data.

At the same time, the structure of the market network in South Dakota was not wholly reducible to a question of which line a town or owner happened to partner with, as evidenced by the appearance of three distinct blocks within the Milwaukee system. The blocks in question could be differentiated from one another in terms of geography, suggesting that the elevator owners partnering with the Milwaukee in, say, southern South Dakota, were not the same as the elevator owners partnering with the Milwaukee in the central or northern parts of the state. In contrast, the North Western system was, in fact, a coherent relational system bound together by the region’s largest elevator companies. If we view the market network from the perspective of towns, the question is whether a town’s position in this network was connected to the rate of Populist organization in that area. This is going to have wait for another time. For now, I want to close by noting that the Populist program was imbued with a relational understanding of economic hardship and the origins of structural inequality. In many respects, the push for regulation in the late nineteenth century can be understood as an effort to build a freer market by making transportation infrastructure into a neutral conduit for trade, privileging public interest over private profit.



  • Adam Slez

    I am an Assistant Professor in the Department of Sociology at the University of Virginia. My research interests lie at the intersection of historical and political sociology, spanning a range of topics including the relationship between politics and markets, state-building and the formation of the political field, and the spatial organization of political cleavage structures. I explore these themes at length in The Making of the Populist Movement, which examines the origins of electoral Populism in the American West during the late nineteenth century. My previous work on state and party formation in the United States Constitutional Convention of 1787 has appeared in the American Sociological Review.

5 thoughts on “The Market Network in the American West

  1. Is anyone doing something similar for the Canadian side of the border? Or, are the two dominated by the US lines?

    Do you know if there is work on changing networks before and after treaties change/adjust the US/Canada border?

    1. Great questions. There may be work being done on these fronts, but I’m not aware of it.

      I would imagine that the analysis I did could be replicated. For example, I just came across this today: This is virtually identical to the railroad maps that I used. I was able to get information on the location of grain elevators from licensing data compiled as part of a public report issued by the South Dakota Board of Commissioners. It seems possible that Canadian regulatory agencies might have done something similar, though I don’t know the case well enough to say. It has been awhile since I have read Lipset’s Agrarian Socialism!

      I’m also not sure about who is working on the treaty/border question, though I would love to see something on the Red River fur trade, which may be in line with what you are asking about.

      1. Working on a book on US-Canada relations as seen through US-Canada/Britain treaties. I’m interested in the many ways they affect economy and security. Got to thinking networks might an interesting additional way to see effects. On fur trade. Not sure about Red River, but Lawrence Hatter’s recent book, Citizens of Convenience about the early border with what is now Canada is pretty darn interesting and might also have relevant citations. Very much worth a read.

Leave a Reply