Musings on Maps

Any maps create quite powerful tools to knit regional divides into coherent forms.  The order of maps serve both to fabricate harmony of a whole from its parts and for showing its divides, in ways data visualizations make increasingly clear.  If the origins of “chorographies” were illustrations of communities, and aestheticizations of their harmony, we’ve become increasingly interested in and attuned to visualize divides, as much as visualize unity, to explain the symptoms of divisions where they might be most evident, and use the format of the map to demonstrate compelling evidence of the fractures that divide, and explore how ideology ever came to “cluster so predictably,” in with geography,” as Steven Pinker nicely put it, and if different conceptions of human nature exist in different parts of the country. We’ve all been all too familiar with the recent remapping of these divisions, increasingly evident  in our political discourse, and affirmed in different ways by posing a set of choropleth maps as points of departure in order to chart, gloss and explain the extent of our divides by the range of data available on local differences across our fragile union.

As we approach Christmas and the season of gift-buying upon us, it makes sense, as Philip Bump, with some help from folks at Esri and Stamen, suggests, to consider the stores where we realize shopping habits and alternatives as criteria that increasingly divide us–and consider the consequences such divisions hold.  Indeed, the divided commercial topography in many regions of America reveal–even more than real estate–a deeply drawn set of cultural divides rarely considered in attempts to understand our fractured political landscape:  so dense are the clustering of large stores in certain areas that they obscure the legibility of place names.  In ways that reflect something like a Google Maps query “shopping at private business near [ZIP code]”, the distribution of data suggests how the landscape of shopping shifts in regions of the United States that provide a deeply qualitative record of our society.  But rather than generate a list of sites, or map all stores, selective parsing of large scale vs. independent shops provides a neat filter to measure the commercial contours of the landscape in ways that won’t so much correlate as reflect a composite image of the lower forty-eight. What’s the new political topography of the United States?  Well, perhaps it’s less red states vs. blue states, here shown in two views of the divisions between Democrats (blue) and Republicans (red)–by data here parsed at county level and by intensity of affinities–map a deep set of social divides that have strikingly played out in weirdly clear geographical blocks in presidential elections.

As Steven Pinker put it, “Regardless of who wins the presidential election, we already know now how most of the electoral map will be colored, which will be close to the way it has been colored for decades.”  Why, indeed, are some red states so red, and other states so blue?  Is this a sign of a divide between rural and urban inhabitants and political preferences–or is it deeper, and, if so, how did we become so divided in our imaginary of what the nation should be?  Are these maps creating differences, or do they reveal some of the deeply-set social divisions codified by Mason and Dixon as a latitudinal divide, but whose division is now reflected ban urban/rural divide? We all know that the last few elections have been stunningly fragmenting–

In a coy reference to collective mapping, Aaron Black of the Washington Post used Gallup’s annual rankings of which states are the most conservative and which are the most liberal to create a heat map of populations across the nation, employing the popular cartographical device so often adapted in infographics to chart virtual a spectrum of political opinion, which, while less intended to emphasize the fixity of divides within current public opinion, showed that outside of Alabama and Wyoming, the number of self-declared “conservatives” in the country does not follow such a close divide:  yet it is hard to find a less changing term than “conservative” and “liberal,” and these don’t map onto the breaks in voting that have recurred across four presidential elections.

Washington Post

But the pressing question of where these divides actually lie have led folks to return to maps with considerable zeal–both as powerful heuristic vehicles of statistical measurement and projection, and to find possible answers for these real divides.  The prevalence of infographics and talk radio have conspired to make this a real issue of political debate for the The American Conservative, where Patrick J. Buchanan placed the question of secession on the front burner this year, corralling red state secessionists and blue-state secessionists into the same category as folks who acknowledge we either face increasing ungovernability due to geographic expanse or having little in common:  the issue of what we have in common, not much of a question for the very notion of a nation or state was in the past understood to balance multiple interests, may be less the question than the “spirit of intransigence” that we face in most political debates.   Does this have to do with the vociferous defense of issues that won’t be debated, or the fact that we have been all too often poising debates in quasi-imperial essentializations of what is un-American?  Or is it due to “lifestyle”?

Can we remap these divisions along lines of population density, as Robert Vanderbai of  Princeton University has argued, whose compellingly recast the electoral map to reveal how two alternate styles of sociability came to inhabit one land.  By throwing the stark red/blue political topography into 3D relief, Vanderbai invites us to look at that divide, not fractured by different data, but mapped onto density so that we might better understand our electoral maps:

The divisions in the electorate have also, of course, encouraged the weirdly popular but loosely credibly parsing of local attitudes across the nation.

Despite the hugely informative value of this density distribution, the question of how the distribution of the landscape we have created might be as well illustrated by the commercial topography we have created, and the space that they create.   For the interesting map that charts a distribution between the majority dominance of large box-stores or chain-stores over smaller privately-owned stores suggests a unique landscape that we have become all too used to inhabit, and a measure of density all its own–but not necessarily in tandem with Vanderbai’s glorious and striking distribution of a two-color bar-chart.  Although Vanderbai’s chart makes the compelling case that our distribution of electoral votes no longer in fact reflects the population distribution in the country–and indeed the possibility of the winner of a popular vote losing the number of votes in the electoral college–the map is less than informative of the landscape we have created in the United States–a landscape that is still being repeatedly surveyed.  To be sure this reflects or is a reflection of density, but also offers an alternate index of social topography.

Let’s peak at the regions that Bump has created, in far more fuzzy fashion, by agglomerating the number of stores across its 40,000 zip codes, and taking a measure of those regions where large retailers show a clear dominance to the extent that they have pushed out all but one privately owned retailer, noted in reds of varied intensity, versus those blue dots where privately owned retailers constitute the majority, and the corridors that this creates, despite its lack of clear contours.  While the ZIP-code breakdown is reflective of the new concentration of wealth on the eastern corridor, the relative decline in incomes provides a clearer index of where such box stores (predominantly discount retailers, who appeal to the pocketbooks of their customers) will tend to concentrate.

The contours of these divides provide a basis for re-examining the arguments implicit in most recent choropleths, and thinking about the nation we are poised to become.  Although not exactly mirroring the levels of declining income, one can detect in the areas of the spread of the big red superstores a reflection of the yellow states with median incomes below $22,126 a year, showing a selective density of big bog scores pushing in toward Denver, and congregating around Michigan, the Great Lakes, Iowa and near Arkansas.

Starting small, for example, take the data that Natural Earth and OSM mappers help reveal about the new topography of shopping in the Midwest, that both reflects density but even more the sorts of incomes that can support private businesses:  as one moves outside of urban areas, the proportion of zip codes composed almost or entirely of exclusively large retail stores rises in clusters, often clustering (predictably) about pockets in precisely those cities also able to support small businesses:

The blurring of the map, in a far fuzzier version at larger scale, suggests a veritable field of red that is almost aligned in belts, and creates a strong contrast to many parts of the country lying further west . . . where they seem to choke off privately owned stores.

Returning to view the southern reaches of the country a greater scale, the discreet stores in much of South appear as oases of small businesses in relatively open land–presumably where small businesses don’t crest above a level of 50%– in Nashville, Atlanta, and Memphis:

Looking at the odd distribution of the Eastern seaboard, a region where large stores appear to dominate the landscape of all but the coastal corridor, and gain a distinct density as one moves inland and approaches a different economy.

A more fine-grained view of the same region–focussing on New York to Baltimore–reveals a more complicated topography, but echoes the general division between small pockets of private stores and box stores that choke out the competition, being the chain stores that they are, which, with relatively cheaper real estate at malls that can be bought up and leased out by developers, creates a new division between our populations that may be deeper than–if it is created by–a gap in income or disposable wealth.  Interestingly, an abundance of privately held, smaller stores is often concentrated around smaller towns, as well as taking on density at large cities.

What sort of social topography does this create?

The distribution of stores has not been indexed for growth rates, but might be measured against recent census findings about declining incomes.  The depressing image of the appropriately colored “brown” marking declining median household incomes in a period of just five years nicely captures the tone of depression of much of the country, and the difficulty of political classes to capture much backing outside the venting of folks whose incomes have declined.  Despite the aversion of a serious depression after the banking crisis, we’ve been hardly hit.

But the choking off of private businesses might be a sad consequence of these declining incomes, and the road trip searching for big bargains at Dollar Discounts that might be taking up the weekend breaks of much of the country.  While demanding further investigation, the number of chain stores that have emerged across in the country surely reflect the declining incomes in the nation, as much as changing costs of real estate.  And these m

The proliferating choropleths that have emerged to explain our social divisions, something of a tour of data visualizations that have duplicated like rabbits on the web will reveal, have in fact tried to uncover divisions in moods as dividing the country–although this map, with the authority of a DSM, seems to assign fixed moods to the boundaries between states that we have come to internalized, so nurtured are we in reading data visualizations as records of objective truth.

On one extreme, the heuristic value of those boundaries between states are fetishized in a spectacular fashion in America’s Mood Map, which uses statistics originally published in the Journal of Personality and Social Psychology to detect underlying differences of constitution–the personal, rather than the political–that creates an odd visualization whose  very terms (“relaxed and creative”; “friendly and creative”; “uninhibited”) make it sound like we’d all really get along quite well, aside from difficulties of understanding tone, and downplays any political division by recasting the map in terms of something like affect.  By ostensibly mapping divides of personality types and tabling the political, this map even invites you to find the state that best matches your mood, if one is to believe self-reported data.  (The division was parsed, with a straight face, as reflecting weather–hypnotized by the ubiquity of the weather-map and seeking to invest something like objectivity in the odd color-scheme below as something that revealed actual differences to invest the odd graphic with similar credibility–for more on the notion of mapping how “different regions have different personalities” in a data visualization, look here.)

 

Regional essentialization has proceeded, in an attempt to map the country as made of different nations.  In the drive to map national divides, Frank Jacobs parsed nine nations of North America on largely economic criteria, rather than personality, distinguishing regions of similar economic production, and casting each as a separate GDP, based on concepts of the uniqueness of our nation from WaPo journalist Joel Garreau, and again granting them stark objectivity via a choroplethic map that similarly served to instantiate diverse regions of the continent:

 

Is this a proposal for introducing new tariff lines?  If so, how much such a distinction makes sense–what about that “Empty Quarter”?  what on earth do they do, and how can they exist?–can be tabled.  And the future of the “foundry,” if it rests on metallurgy, is certainly on the rocks.  And what does Quebec produce, aside from fleurs-de-lis? Colin Woodard sought to describe national divides into some eleven groups, arguing that we’ve “we’ve never been a nation-state in European sense,” for all who imagined the possibility of American political unity, and wrote a slim book that recast the Progressive notions of a ‘melting pot’ by explaining that each region reflected its own distinct stock, and illuminate differences to which our democracy, supposedly, had unfortunately rendered us blind–but whose divisions Woodard appears to suggest might explain all American history, including the Civil War, up to the spread of the Tea Party.  Woodard’s pedantic mapping of distinct regional variations combines racial essentialism and a “revenge of geography” to point up our innate fragmentation:

The distinctions were starkly mapped in alternative iterations of the “American Nations” were presented as a key to better understand ourselves, as if to hold a mirror to the land:

 

 

Sean Wilkinson Design Woodard’s work extrapolated from the historian David Hackett Fischer’s argument about the differences in the settlement of different regions–particularly the northeast and south–by different Anglo cultures, and the effects of Anglo farmers who settled the northeast from the Scots-Irish herders who settled the Deep South, and from the work of many anthropologists and the psychologist Richard Nisbett on the culture of honor in the South.  But despite some  neatness in registering cultural differences in the map, it doesn’t take huge skill to reveal the culture of separateness in the Deep South, and a floating barrier of “Appalachia”–the ‘light South’–around it;  yet problems arise when one moves from the map to individual experiences. Despite the evocative names with which other regions are endowed, from Tidewater to the New Netherlands, their relative stability or homogeneity as entities comes to break down, as does the notion that their boundaries are so firm.  The odd coherence that he supposes in such regions as “New France”–would Québécois see anything similarly to folks in Louisiana or even New Orleans, let alone accept them as affines?–and other regions–how isolated and cut apart is this “Left Coast” today?  What could ever be mapped as weirdly as Yankeedom, and is it still alive?–seem animated by an ethnic essentialism belied by the shifting map of ethnic divides and affiliations in the United States.

And it is not clear that these areas will retain meaning with and over time–although the deep danger of which the map is something of a symptom is a pronounced lack of national belonging across the land!  Woodard’s sense that American history “all goes back to who settled those regions and where they came from” flattens history, and would not be readily recognized by any local inhabitants. No doubt these can be traced to the deep and unavoidable social divides and fissures that exist in our country, and to the different parts that different regions have played in the culture wars.  Perhaps it might make sense to adopt an approaching demographic forecasting, less entrenched in the categories used today, and the apparent generation of megaregions.

 

 

 

There are, of course, true socioeconomic divides that fragment the country in all too real ways.  The frighteningly influential role of real estate in shaping the country is sadly evident when we map residential space by income, and illuminate truly early modern concentrations of wealth.  Using the American Community Survey to map a choropleth by variables of income, real estate, and educational levels attained, we can perform something like an archeology of the national map, digging deeper to examine the real divides among its inhabitants.  Dividing regions of the country by median level of income, clear pockets of navy blue surprise the eye and remind us about the unavoidability of underlying social divides–that, I’ve elsewhere argued, mirror local reception of the Affordable Care Act.

 

 

Or, again with an eye to the opposition to the ACA, long after the Great Society, we can divide our choros by clearly drawn concentrations of poverty as much as wealth, using a distribution of  local populations living below the poverty line in each of the nation’s counties that reveals a similar concentration of poverty in the deep south and southwest:

 

 

Not to mention the map–which somewhat scarily illustrates parallels–of the percentages of adults over 25 years of age who did not complete High School, in a sort of national map of educational attainment as they are reported in US Census figures:

 

 

These stunning choropleths, courtesy Calvin Metcalf, Kyle Box and Laura Evans, reveal a Southwest and deep South we knew, but weren’t ready to acknowledge.  They tell an interesting story as to why Virginia and Florida belong less to the Old South–or to what it has become.   To look more closely at a comparison of other local trends in the Community Survey, look here. Yet we might do well to look elsewhere to map national divides, and new databases to grasp our emergent geographic divisions and continental divides.  While real estate markets have much to do with the divides, we seem to have individuated the intersection between density and real estate across state-lines, by taking the presence or absence of small businesses in our communities as a sort of tipping point.  For the benchmark of zips with one or more privately owned businesses reveal a major fracturing of the local landscapes in the country, oddly resonant with our divided electorate.  The huge regional discrepancies of coasts and interior needs to be read with density, in ways that few openly consider, as a means to map the habitation of the landscape, shifting from questions of bodies, ethnic composition, or other criteria, toward the nature of interaction between self and place that shops afford, and even the sorts of relations to objects that are synthesized in the contractual relations of the marketplace, not to mention the stimulation and relative criteria of judgment that a diverse marketplace affords customers. The below map of the nation is limited to those spots with no or one small businesses within the ZIP in question (the red dots, which often cluster and overlap, and are more intensely red to reflect the intensity of that concentration) and zip codes which are distinguished by a majority of small businesses (blue dots).  While this omits large regions of the country–where there are an intermediate percentages of small businesses, and they don’t register–those are of course the very regions where large stores are equivalent or exceed small businesses, making them likely sites of the inevitable large box stores.  And the selective optic of the map is precisely its virtue as a portrait of contemporary agora, or our evolving landscape of exchange.

 

 

In ways of considerable sociological significance, the map reflects the landscape we’ve made for our country, or we make for our selves, by mapping some of the sites where we define ourselves to what is left of the public sphere.  The map may indeed suggest some interesting sociological detail about where the country is heading, with interesting correlations to level of educational completion below which it lies.  There is something that goes beyond or beside a decay of discrimination, but leads to a shift in one’s relation to the market and how one is a participant or spectator in a collective activity, in the geography of box stores that can be mapped across America.   The dominance in a mental universe of Home Depot, Walgreens, Wal Mart, Dollar General, Target, Lucky, Thrifty, Family Dollar, Dollar Tree and even Sports Authority exposes a topography of disorientation and dislocation, rooted in the offshore production of goods and plastics, undermining face-to-face encounters in 24/7 Emporia that offer the best deals may well conceal endemic conditions of placelessness and timelessness:  in a country where 80% of the stores nearest to anyone are large retail chains, one has to interrogate the ramifications of what sense of social space we retain:   as we become consumers in another Holiday Season, bracing ourselves for sales, what sort of social change lies in no longer having a corner store? Where’s the analysis of this shift of commercial topography when you need it?

In ways rich with sociological significance, the map tells much about how store-goers orient themselves to the publish sphere, direct consciousness to the relation of goods’ producers, and regard the social contract or social nexus as a given or negotiated part of the status quo.  For as much as it lies in our pocket-books, perhaps the deepest divide among us of all lies in the landscapes where we spend it.  Perhaps the deepest divides of all rest on not so much in levels of income or consumption across the lower 48, so much as the shops available for perusing or spending time–and the attachments to place, to space, to objects, and indeed behavior pursuant from them, which inevitably cascade from how much time inhabitants devote to patrolling the aisles of brightly lit box stores of Sam Walton–rather than looking in at local businesses to greet their owners and examine their wares.  Based on a search for the places where “small business really thrives” in the lower forty-eight among the nation’s 40,000 zip codes, data from OpenStreetMap has allowed a map of the sorts of stores in each–noting in red those zip codes hosting none or one small businesses, whereas those where small businesses constituted an outright majority enjoy varied shades of baby blue.  The resulting small-business texturing of the country provides a valuable index for looking at–or reading, if not actively interpreting–the nature of its communities, and an alternate American Community Survey of its own.

While this doesn’t exactly correspond to a political divide, it may tell us more about what sort of a nation we’ve become–and what sort of commercial contacts that we value, or personal contacts we are able to cultivate to the owners of shops, places more likely to apt local products.   The division of red vs. blue is less reflective of a political divide, but seems to offer itself as an alternative metric to those above introduced to measure social divides, and belongs in the company of other social indicators taken as prognostics of political divides in the electorate, if not of an alternate community survey.

 

 

In ways that do not map directly onto the distribution of real estate–shown in the first map from data of the American Community Survey above–as one might expect, the topography of shopping that it indicates suggests deep social divides to which Nate Silver might pay some attention:  while the coasts are brightly blue, in California we can detect the conservative central valley and interior; the absence of blue in much of the central states is shocking, and preponderance of red throughout the south–save coastal Florida–overwhelms, although the Northwest seems reliably blue.  In Minnesota and Wisconsin, the red blotches shade to pink and dissipate; Texas seems divided with deep red concentrations.  Indeed, the problematic position of Pennsylvania in the national picture is nicely illustrated by its deep red isolation, whereas privately run business dominates the Northeast Atlantic seaboard, particularly north of New York, as well as in much of New England:

 

 

The irony is hopefully apparent, given that the defenders of Republicans are presented as the champions of free enterprise.  Maybe they are:  they are not, it seems, defensive of independent shops or businesses. The sea of red in the middle of the country, but seems a growing miasma that doesn’t necessarily respect political affiliations, but also bodes pretty badly for the survival of a culture of localism, at least in terms of the future (and profitability) of independent business:

 

 

The map, in the image it offers of the Midwest, reveals a striking clumping of regions without independent businesses–clumps not necessarily paralleling cities–the like of which is probably a major shift in our society the likes of which the world may have never seen:

 

 

Those places with greater income may simply be ready to foot the extra bill for local boutiques, some will say, or forget the offers of Amazon’s free shipping:  but we see a deeper divide, I’d argue, shaped by the market as much as responding to lines of wealth, and conditioned by all those aisles of shopping centers illuminated by starkly lit flourescent lights, lured by the slogan “Save Money.  Live better,” and names like Dollar Tree and Family Dollar that remind shoppers to keep the bottom line first and foremost in their shopping experience, rather than cultivating stores that might provide more jobs to their neighborhoods.   (Even if their parking lots may use LED lighting, most stores have a sort of blinding or blanching flickering flourescent light so recognizable that it seems potentially addictive as a unique neurological experience.) We can see a new mapping of the landscape around cities, which, if we take only the midwest, emerge as something like islands with private stores, surrounded by areas of box stores that cluster around the nearby highways and urban belts, to take Chicago and St. Louis; smaller towns are dominated by zip codes dedicated to large chains:

 

And in the deep interior, many zips feature the near-absence of anything but larger super-chains:

And we can see the fracturing of a political landscape in divided states–such as Colorado–where diverse concepts of the geography of sales, as it were, present different models of a lived landscape, almost cheek by jowl, alternating between spaces of chain stores and a telling density around Boulder, Denver, and Colorado Springs–with a few exceptions–of small businesses.  In some ways, this reflects the deep debates in the state on such issues as gun-control, birth-control, ecology and a more nuanced or complex relation to federalism.

 

 

Where we shop is increasingly what we are not only because of our attachment to objects or commodities, but how we are oriented to the world through the market–and as we map our relation to markets in different ways,  the ongoing onslaught of this attack of the big box stores has deep repercussions for our society’s unity.  Go forth and map as you do your holiday shopping . . . To be honest, some of the maps derived from this data don’t make immediate sense to me, or at least jibe with my own experience, such as the large number of small businesses found in Las Vegas, although I suppose this reflects the definition of a “small business.”

 

But the image of the Bay Area is somewhat recognizable, and recognizably blue.

 

Yet a broader landscape and context emerges as one expands to adjacent ZIP codes in order to include bordering strip malls, including those inevitable and all too visible nearby strings of outlets, and sections of highway that are dotted exclusively with box stores.