Mapping spatial diversity in the United States of America
We turn our attention to mapping spatial distributions for large diverse social systems. We focus on the continental United States (excluding Alaska and Hawaii) as an example having 3111 units (counties) that differ greatly and are located in a vast ecologically diverse geographical space. The social diversity across these units can be viewed as being driven by the presence of different peoples, living in them over long periods of time and carving out places to live in different ecosystems. These areas also have quite different cultures, life styles and preferences contained within them. Attempting to understand the spatial distribution of the social, economic, political, and physical features of such a large system is a daunting task.
One way of doing this is to put a primary emphasis on contiguous areas and study these directly linked areas. Examples of this approach include 1) the work of the US Census Bureau with its broad regions, 2) a provocative book, The Nine Nations of America, by Garreau (1981), and 3) American Nations: A History of the Eleven Rival Regional Cultures of North America, an in-depth historical examination of the formation of the United States and its change over time by Woodard (2011). A second approach to understanding the spatial distribution of America's diversity is to focus on carefully selected indicators, summarize them, and describe areal units in terms of these summaries. Chinni and Gimpel (2010) did this in Our Patchwork Nation: The Surprising Truth about the 'Real' America. Together, these three books form the departure points for our depiction of US spatial diversity.
Geography plays two very different methodological roles in these two broad approaches to studying spatial distributions. Garreau (1981) and Woodard (2011) as well as the Census Bureau adopt an approach where delineating broad contiguous regions takes center stage as an organizing principle. In contrast, Chinni and Gimpel (2010) focus on the attributes of places while ignoring geography completely until they plot the summarized statistical data
Figure 9.1 The standard census map of regions.
Source: U.S. Government.
in geographic space. The former approach appears to emphasize spatial contiguity too much while the latter emphasizes it too little. We steer a middle course between these approaches. Our goal is clustering spatial units in terms of their attributes while, at the same time, taking into account the contiguity (adjacency) of the units. The methodology used builds on the constrained clustering approach developed by Ferligoj and Batagelj (1982, 1983).
Mapping nations as spatial units of the United States
Woodard (2011) makes a distinction between nations and states. A state is a sovereign political entity and, as such, it is eligible for membership in the United Nations (UN). Clearly, the United States of America is a state in this sense. In contrast, 'a nation is a group of people who share - or think they share - a common culture, ethnic origin, language, historical experience, artifacts and symbols,' Woodard (2011): 3. This conception allows him to claim the existence of eleven nations in North America. Earlier, Garreau (1981) suggested that there are nine such nations. Both claim 'The United States of America' is composed of nations and fragments of larger nations. Both authors assert also that our understanding of the USA is too limited when we use only the Census Bureau's classification of units shown in Figure 9.1.
Figure 9.2 The Garreau's 'nine nations' of North America.
Garreau (1981) threw down a gauntlet when he wrote 'forget about the borders dividing the United States, Canada, and Mexico, those pale barriers so thoroughly porous to money, immigrants and ideas. Forget the bilge you were taught in sixth-grade geography about East and West, North and South, faint echoes of glorious pasts that never really existed save in sanitized textbooks,' (1981: 1). From his perspective, the map shown in Figure 9.1 is of little value for describing the spatial diversity of the United States. In its place, he defined nine distinct nations covering Canada, the contiguous 48 states of the USA, Northern Mexico, and some Caribbean islands.
These nine nations are: New England; The Foundry; Dixie; The Islands; MexAmerica; Ecotopia; The Empty Quarter; The Breadbasket, and Quebec.4 Drawing upon partial histories, local interviews, regional folk wisdom, and a wide variety of documents, he lays out the map shown in Figure 9.2.
Woodard (2011) presents an alternative set of eleven nations, for which we have drawn the schematic map shown in Figure 9.3. While he locates these nations in the same overall geographical space as Garreau, he also delves into the long histories of how these nations were formed and, equally important, the identities of their founders. The eleven nations he presents are: Yankeedom; New Netherland; The Midlands;5 Tidewater; Greater Appalachia; The Deep South; New France (most of which is in Canada); El Norte; The Left Coast; The Far West, and The First Nation (also in Canada).
There are considerable overlaps between the two sets of nations. El Norte is essentially the same as MexAmerica. The Left Coast is the same as Ecotopia. Woodard's New Netherlands is very close to New York City (with a few more counties included), one of Garreau's aberrations. In Canada, New France includes Garreau's Quebec. However, Woodard separates
Figure 9.3 The Woodard's 'eleven nations' of North America.
the Cajun parishes of Louisiana to put them into (a separated part of) New France. This was reasonable given the historical migration from the Canadian part of New France to the Cajun area in southern Louisiana. Also, Louisiana's legal code is based primarily on the French legal code rather than on English common law. Confining attention to the USA, The Far West is approximately The Empty Quarter.
There are clear differences in their two regional narratives. Woodard separates Greater Appalachia from Garreau's Dixie as a separate nation. He then expanded this nation to include areas in a southwesterly direction. Tidewater has also been removed from Dixie as a separate nation. Yankeedom includes the US part of Garreau's New England but is expanded to include large areas of Garreau's The Foundry. Neither The Foundry nor The Breadbasket is preserved within Woodard's scheme and Garreau does not include The First Nation.
Woodard's account is attentive to the founders of nations and is more overtly political in the sense of looking at conflicts between pairs of nations and the alliances they formed over time. He sees Yankeedom and The Deep South as locked into a permanent struggle for the control of the US federal government. Our partial summary (in Chapter 6) of the political factions at the Philadelphia meeting where the US Constitution was written is fully consistent with this argument. Greater Appalachia and Tidewater were uneasy allies of the Deep South while The Left Coast and New Netherland sided with Yankeedom in this struggle, one that has lasted for centuries. There are more recent sharp differences. The Left Coast is seen as the political womb of the environmental movement while The Deep South is an integral part of the core of resistance to environmental regulation along with Texas. These conflicts point to deep divides across geographic space regarding culture, values, and attitudes towards government in these nations of the USA.
While depicting broad regions sharing some important characteristics makes considerable sense, the nations in both Woodard's and Garreau's accounts do not map neatly in the divisions used by the Census Bureau. At face value, both the Nine Nation and Eleven Nation partitions seem more persuasive than the somewhat arbitrary and simplified administrative definition of regions used by the Census. Garreau included maps of each of his nine nations containing boundaries, some of which were within the US states rather than simply between them. This allowed us to construct the map shown in Figure 9.2, which represents those nations (or parts of nations) wholly or partially contained in the Continental USA.
Woodard (2011) provides a county level map which we used to construct another regional map. This is shown in Figure 9.3. Both Figures 9.2 and 9.3 are drawn in a format we use hereafter. Note also that there is again a distinction within Florida with a southern tip. However, this seems more like a blank within Woodard's narrative than part of a 'nation'.
We will attempt to 'reconstruct' these maps by clustering quantitative descriptions of areal units while, as noted above, constraining the clustering of these units by their contiguity. We note that large cities, and the counties containing them, are often very different from the adjacent counties surrounding them.
-  This eliminated all guesswork about county membership in nations, in contrast to Garreau's maps where some boundaries between nations were located within states.