Tag Archives: Francis Amasa Walker

Infographics in America

If the nineteenth century America has been often described as an era of geographic integration, perhaps no one more than the ambitious statistician Francis Amasa Walker created a new way of seeing the nation that foregrounded both the local differences that continued to divide the nation, but staked out the challenges for integration that the country faced in geographical terms.  His extremely influential 1874 Statistical Atlas, based on the unprecedented 1870 US Census that he directly supervised and made the case before Congress about the undeniable need for funding, whose maps created an image of the challenges of national unity that remained in the republic in the wake of the Civil War in which he had fought:  the Statistical Atlas extends the enterprise of the expanded Census, validating how statistics present a synoptic picture of the political economy–illustrating relations of the local to the polity across continuous United States as if processing part of a mental effort of consolidation.

Even before the unprecedentedly bitter electoral divide of 1876, Walker advanced maps as providing a new way to embody the polity through the visual records derived from statistical aggregates.  Although Walker’s subsequent tabulation of the data on immigrants in the late nineteenth century led him to fear their arrival as threatening the nation’s productivity, based on his perception of the depth of racial differences to the national polity, and encouraged others to do so, he advanced the embodiment of statistics in geographic maps, in ways no doubt influenced by his close collaboration with his father, the economist Amasa Walker.  In ways that prepared a basis for his use of maps to express the contested electoral results of 1876, Walker treated maps as coherent statements about the nation’s divides otherwise not able to be articulated, as a basis to start debate about the well-being of its political economy.  (His maps were so convincing, indeed, in framing a question of geographical organization, that they may have encouraged a narrative of geographical integration only recently challenged by investigations into the geographies continued to be produced on the local or regional level.)

Walker's Image of the Nation's Population

Mapping Population Density in the US, 1870

Susan Schulten has recently advanced the quite compelling argument that Walker’s innovations constituted gave the “invention of the infographic” distinctly American roots.  Her argument spoke appositely to the almost obsessive return in recent elections to infographics that suggested the likely tendencies of voters, and indeed often reframed a narrative of division into “red” and “blue” states–and even designated some “purple”–in ways that revealed an undeniable undercurrent that verged on obsession of questions of national unity and division that were of the very sort that Walker had similarly sought to address when he undertook the reformation of the decennial Census in 1869 at an amazingly young age of twenty-nine–no doubt with insight as to the ability to advance and illustrate the distinct distribution of space that the nation occupied.

Even as a staggering proliferation of maps of electoral zones flooded airwaves, newswires and web during the 2012 election, Schulten traced the invention of the American infographic to the innovative visualizations of data and government statics by an enterprising statistical mapper who after working to organize the 1870 Census, not only drew up a comprehensive reform of the census but treated its findings to create a “statistical survey” that came to  embody the nation’s political and economic unity.  While earlier Censuses were strikingly unscientific, Walker advanced issues of political economy in maps as an extension of his expansion of the decennial census, organizing the tabulation of population, agriculture, mortality and manufacturing data on 39 million Americans, and placing prime importance on geographically orienting statistics as tools to better visualize the nature of social and economic divisions after the Civil War in which he had fought and been grievously injured.

Walker’s maps framed the issue of integration in legible fashion–and produced them to allow the fate of the nation’s unity and division to be processed for a wider audience than would have otherwise confronted them–they did so since they readily processed statistics that went far beyond physical phenomena to chart the racial composite by which the national economy could be understood, moving beyond existing models of its physical geology–which he also included in the Statistical Atlas

walker-map-geologyPrinceton University (from Statistical Atlas, 1874)

–to attempts to embody the composition of its human inhabitants whose aggregates were earlier not clearly understood, in what was indeed the first public census to count African Americans who were former slaves as part of the nation’s fabric.

Colored in Aggregate

Colored Populations in the United States (1874) (Courtesy of Princeton University Library)

Walker’s advocacy of such choreographic statistical maps as snapshots of the political economy led to an invitation to join the editorial board of the New York Times–which he declined, probably since he continued publishing in competitors from Scribners’ Monthly to Harpers’ New Monthly, emerging as a public intellectual of originally progressive bent.  For Walker had convinced the US Congress to adopt a variety of projects that used recent lithographic techniques and statistical correlations to use the results of the 1870 US Census map an coherently compelling image of the nation’s situation for public debate.  If all maps reflect both the character and competency of their makers, Walker’s maps reflect the excitement and tenacity of mining data from the Ninth US Census of 1870 that he had compiled with congressional authority, compiling, correlating, and refining the image of the distribution of wealth, illness, and health to a degree that had not ever been earlier achieved.

He engaged in mapmaking in print as a form of public discourse that elevated the statistical map as a tool for envisioning the nation as an aggregate.  Walker’s early involvement with late nineteenth-century newspapers like the Springfield Republican Newspaper as well, from the late 1860s, at the Atlantic Monthly, in fact no doubt encouraged his trust in the power of such organs of public debate–and the power of printed supplements based on the US Census, several of which he published for Harpers’ New Monthly, as well as in Scribner’s Monthly, The Century and North American Quarterly, in ways that no doubt led to his conviction in the infographic as a way to shape public debate on political economy, population density, home servitude, and the working classes.  Walker’s position as Chief of the Bureau of Statistics and Taxes may have helped him use his position as Superintendent of the fifth US Census at just twenty-nine to present the project of the first Statistical Atlas of the country, a project which he expanded in the 1880 Census, whose unprecedented twenty-two volumes collected an even greater range of information than ever previously collated and greatly refined the unscientific nature of previous decennial censuses.

Francis Amasa Walker saw the U.S. Census not only as a way to view populations of states, but to expand the vision of and the likeness of the nation by the more arduous measure of density per square mile, and to then use that image to chart the distribution of the national population as a demographic tool.  He worked with the census to conceive of the map as a measure for mapping complementary sets of data, by mapping relations between density and select quantifiable variables, mapping population density against wealth distribution, literacy, childbirth rates and disease–but not voting preference–in maps that created a legible record of the country, whose public good he convinced the US Congress to fund Atlases in 1872 and 1873.  By transferring statistical observations into a detailed picture of the nation, he began from a base layer of the contours by which its population was distributed, without focussing on jurisdictional bounds, in ways that effectively augmented the independent authority of maps as media of sociological investigation and public communication to an extent that had never before been the case, but established a central place for detailed choropleths in the American Grain.

Walker's Image of the Nation's Population

The maps are stunning choropleths that exploit lithographic techniques to picture the national population in great detail.  Walker started from the initial map of density, using the census of 1870, of which he was superintendent at the age of 29, and whose possibility for converting into cartographical form he seems to have readily perceived.  The plans led to the publication in 1874 of the Census’ data, in an initial atlas that totaled just fifty-six pages, each map of which has a degree of detail never seen as a visual embodiment of the nation, but emphasized its distribution of population and industry to an extent never realized–and, at least among its readership, posited questions of national coherence that concretize concerns for the country’s political economy:

Mapping Density in US

Density of Population in 1870–including African-Americans as well as Whites

For the first time, they also offer a mode to integrate African-American former slaves within a national economy, and posit a detailed, comprehensive and analytic image of population density across the north and south that suggest the value not only of statistics but of imaging the nation and its divisions in a decennial census.  The embodiment of issues formed as Assistant Secretary of the Treasury placed issues of political economy before the eyes of their readers in especially effective ways.
The fine grain of variations provided a new way to look at the nation, as the compelling lithographic choropleths of Alexander Bache, by using line-engraving to chart the population rather than topography as their concern, or coastlines or hydrography, than the shifts that the 1870 census revealed.  If riverine paths were noted, the South looked distinct when one saw the vast “gaps” or absences of population in much of the rural areas of Kentucky, Tennessee, and Georgia, and indeed the concentration of the population most entirely along rivers where agricultural trade grew.
Population Distribution chloropleth
The image of population density dramatically grew in Ohio, Indiana, and Illinois, in ways that charted the new image of the nation, and must have posed questions of how  population density mapped onto the distribution of wealth, or, an extension of it, childbirth rates, in order to refine and better understand the picture of national population it put forth.
Virginia to Georgia Chloropleth

Things got more complicated and more exciting when Walker mapped the distribution of wealth among these densities of population, in ways that helped reveal the uneven distribution of incomes across the post-war nation, and revealed a conspicuous ongoing divide between northern and southern states that related to industry but also to the extreme impoverishment of much of the population in the southern states that would continue to be a contradiction and difficult conundrum in the nation:

Mapping the Distribution of Wealth

The sort of graduated shaded in maps that Walker created provided something of a counter-map to the symbolic uses of mapping as a way to envision national unity, and pinpointed questions of the intractable nature of differentials of wealth.  If pockets of wealth extended deep West into Iowa and Kansas, and lit up Illinois, Ohio, and Pennsylvania in deep auburn hues, the Mason Dixon line was a divide in how wealth was distributed whose light ochre and pockets of white were rarely interrupted by redoubts of wealth, almost entirely along the coast or select segments of the Mississippi River.  His maps adopted the recent precision of mapping population density against its distribution that had been pioneered by Alexander Bache’s compelling visualization of the records of the 1860 Census with the German emmigre engraver Edwin Hergesheimer

The detailed cross-tabulation of population and wealth seems a bit of an odd ancestor of the modern infographic, although their kinship is clearly recognized.  For the image of Walker demanded the sort of detailed attention as a picture of the nation to which the compacting of information or metadata in the range of infographics generated by GIS programs rarely provide.  If Walker anticipated GIS in a fashion, the rapidity of generating the infographic that synthesized metadata with amazing facility and rapidity rarely demands the sort of attention that Walker’s images command.  In part, the lithographic medium habituated viewers to the parsing of refined distinctions in the economic landscape that variations of shading revealed:  the starkest of our blue/red maps with some pink or light blue question marks are removed from the fine-toothed sort of distinctions drawn in the census, or from a similar statistical subtlety.

But the proliferation of the infographic so readily produced and tabulated provides less of a reference tool than an attempt to hold onto the permanence of a snapshot provides within the rapidly shifting and changing landscape that often seems adrift in the electoral sea.  The questions that Walker asked–distribution of wealth, literacy, disease rates–are rarely raised in the infographics we see most frequently, even excepting Fox News’ proclivity for the infographics that perpetuate stark divisions.  The ‘mediazation’ of the modern infographic as a labor saving device not for observers, but whose construction demanded a more limited investment of detailed attention, has created a new assembly-line production of images of limited refinement, whose authority rests on their mapping substrates, rather than on the measurements they mediate and encode.

The limited subtlety of the infographic suggests not only their ideological points, but a shift in what might be called, with the late Michael Baxandall, a “period eye” to express the tastes of interpreting images from media to the viewing map–the habitual practices by which we look to maps in an age of the rapid-fire production of new infographics each day for broad consumption and, also, ratings appeal–or how infographics snappily  process an argument in a bottom-line nature, rather than approach the social topography whose complexities we’d maybe rather not want to detect or explore at close range:


Most items that bear the name of infographics might not be data visualizations, or even use the map as a tool to construct the meaning of its contents, since the map is so often and so readily abstracted from the territory. Maybe a map might even suggest what we might want to keep distance from, rather than to consider up-close:

Top Party Schools

These pilfered objects are a bit extreme, but their divorce from any sense of geographic meaning is somehow telling.  Sure, these are something like visual jokes, but they make the point I want to make about the liabilities and deceptiveness that the using maps to organize contemporary infographics reveal that adopt maps to bolster their suasive abilities as much as frame a problem.  The mock-maps–the second based on an episode of Ira Glass’s “This American Life” by E. J. Fox, from a series he titles “This American Infographic”– illustrate the problem of the need that cartographical infographics fill, of both adopting the authority of the map (without containing much geographic information) and of using it to display the ready access to metadata that most images of GPS presume–or, in the most banal but most common case, the weather maps on the Weather Channel–and that most folks expect.  Data-mining becomes replaced by graphic design tools imported from the world of advertising, and the maps are a blunt instrument to make blunt arguments, or present an image of the status quo:  the big parties happen at big state universities.

Maps are especially powerful tools to process information for viewers.   Some less ridiculous examples of infographics reveal some surprisingly similar attempts at using mapping forms or mapping syntax to preserve an illusion of omniscience and often to illuminate or make a comment about national unity.  But they also often use maps in ways that, unlike the maps  derived from government censuses that Walker examined with considerable care to demographic variables, conceals an absence of  analytically meaningful argument.  They treat the map as a form of metadata that reduces analytic specificity–from a map of the “battleground states” that effectively uses a format of mapping in order to suggest either the limited support for the Democratic party in the nation (by using a shade of yellow closer to red than blue) or indicate the deeply flawed nature of the democratic process to their viewers.

Recall the sort of maps that were all the rage when Schulten described Walker’s innovative practice, if you can bear it.  Rather than resting on numeracy or the tabulation of relative measures of difference by a statistical model, the map foregrounds the indeterminate nature of places where polling was within a margin of error, rooted less in mathematical literacy but in pollsters’ relative ability of prediction.  The mediazation of the map is removed from a mathematics of mapping or an expectation of refinement, for it is almost more of a symbolization of a politics of stasis or an electoral divide:


The greater refinement of other maps that shaded “tending towards” in lighter colors foregrounded the unpredictable nature of voters’ preferences, more than the composition of the electorate, as they seem to table the question of national coherence or cohesiveness as a whole other issue.


In each of these admittedly varied cases, mapping indicates an aura of accuracy or invests a sense of stability in the face of indeterminate data.  The map is a totem by virtue of its processing of information from varied diverse sources, but the map blunts potentially far more precise tools that seem to divide the polity and focusses on electoral results.   The questions of numeracy that divide the nation are less based in a tabulation of data or statistical familiarity; the block-hues of states mute meaning analysis.

Blocks of red in the map seem possible of being emptied of geographic meaning.  One famous FOX infographic purported to identify a strategy for Republican victory, but undermined the very legitimacy of this potential scenario for the attentive few by mislabeling the states it purported to count with accuracy–as well as deceptively reinterpreting polling trends:

%22Western Path%22

The absence of an expectation of reading measurements beyond numerical addition are evident in a map more reminiscent of the refined criteria of a jigsaw puzzle than in puzzling questions of national unity or ideological difference:

electoral colors map

In part, ideological differences are just not that pronounced, and the maps are oriented to processing polling numbers that were changing like a stock-market ticker-tape, rather than providing a firm basis for a national portrait.  But the father of the infographic would most certainly not be pleased.  The adoption and diffusion of mapping forms in infographics provide metadata constructions perhaps most significant for how they quash related questions or discussions, by ordering a massive amount of data whose impression of preponderance is more likely to take away one’s breath than pose a question, and is almost always likely to conceal an argument.

To some extent, of course, the new elevation of infographics is the creation of the new media economy.  There’s an odd dynamic of devaluing of the analytic power of the map at the same time as elevating its explanatory power.  The map, in an age of reduced news content, seems to substitute for the strength of an analytic news story, as a GPS program produces a snappy infographic that seems both content-heavy and a pleasing amuse-bouche.  The need to process different news sources or on-the-ground informants might be both excused and avoided, where we can come up with a symbolic rendering of what happened, even if we don’t need to look so closely at what its causes or its actual ramifications were.  The absence of analytics in the infographic–which presents, as with the weather, the state of things as they are as an actuality that does not need further analysis or attention to local variation–is perhaps its most pernicious feature as a medium.  They stand at a remove from the maps that the great nineteenth century statistical geographer Francis A. Walker so valuably labored to design.

We develop infographics such as the following by crunching some obtainable numbers–in an image that unsurprisingly perhaps uses the residue of a weather map as its base–to tell a story that collapses multiple different narratives into a single set of information that the viewer can quickly process.  The iconic map that was diffused after the last election was less about fault lines or divisions in the nation, than a cartogram of the new image of alliances in the nation, where the entire midwest stood as a block of blue with the Western states:


That is not to say that infographics can tell a subtler story of similarly chorographic proportions, to describe the image of unemployment in 2003, for example in the country:

choropleth of unemployment in US 2003

But at what cost?  Choropleths such as the below seem to remove individual experience from their comprehensive picture–and provide a “big picture” that is actually difficult if  to comprehend for all the metadata they synthesize.  And they present an intractable image of a social divide whose dark bands of dark blue reveal a density of those out of work that only grew by 2008 in the very same areas:

choropleth of unemployment in US 2008

There was a similar crunching of numbers at a remove from individual experience of tragedy in a map of electric outages suffered as a result of Hurricane Sandy, providing a purview of outages from Augusta to Raleigh.  The map is powerful and striking, but also elided the stories of its destructions or narrative of its meteorology with an easy infographic of a sort of least common denominator to everyone can easily relate of the lights going out.

infographic on power outages from Sandy

These maps erase their inhabitants. So what, then, we might ask, is the territory?

The ghost of Walker, and continued prestige of his aesthetics, have led Nathan Yau of Flowing Data to provide a comparable set of visualizations that embody our national territory, based on the ongoing statistical surveys of the American Community Survey of 2010, to “revive” the project of a Statistical Survey in the footsteps of his august predecessor, noting with some evident pain the absence of any plans by the US government’s Census Bureau to produce one after 2000, perhaps due to the high costs of the Census itself–and the recent Republican-led effort to even claim that the decennial Survey is an unconstitutional invasion of privacy–even though it provides the best basis for the apportionment of government funds, and one of the clearest demographic portraits of the country–that tarred the survey in no uncertain terms as “intrud[ing] on people’s lives, just like the Environmental Protection Agency or the bank regulators,” according to Daniel Webster, an inspiringly named first-term Republican congressman from Florida, who questioned the random nature of the survey as illustrating its “unscientific” value, despite its assessment of over three million American households in considerable demographic detail about their occupations, housing, literacy, languages spoken at home and at work, and levels of education, as well as their approximate computer use.

Perhaps with some premonition of the dangers of resting our democracy on the thin infographics consumed by watchers of television news, the self-published Survey Yau published online imitated the august elegance and clarity of Walker’s maps to point up the absence of needed visualization of the data that the Census compiles.  The images–able to be bought individually as posters–suggest the deep presence of Walker’s idiom of visualization within our current media circus, when the proliferation of news maps from various outlets and Graphics Depts. seem dislodged from the interest of the public good.  Yau’s project was may seem to have obviated need for an impartial assessment of all the data that the Census compiled.  Indeed, while the Times, which once offered Walker a seat on its editorial board, created a brilliantly colored set of interactive visualizations based on data from 2005-9, Yau offered nostalgic images that embodied the images of the nation that the government has puzzlingly withheld, with climatological and agricultural data that provide a similarly detailed atlas of the coterminous United States which, in an age overflowing with data visualizations, remind us of the need to preserve a picture of the nation to ensure we keep the public informed.



The detail of the maps, set in a sepia background with shadings that somewhat approximate the exacting palette of Hergesheimer’s half-tones, provide a set of gradations of population as revealed in the data of the 2013 ACS,



Or, to depart from the demographic and verge into the statistics of the environment, the distribution of levels of rainfall,




or landcover:


The data waits to be visualized, and the simple monocular visualizations capture its complexion with quite understated elegance.  While less rooted in concerns of political economy, the visualization hearkens back to a unified choreography that we often seem to lack–even if it only pictures tornadoes–in ways that go beyond the existential qualities of weather maps as records of the present day.



To be sure, the political economy of the nation has become so fragmented that it is hard to visualize by such clearcut lines or shadings, even though the Times’ visualization of the American Community Survey similarly uses the map as a surface on which to throw the composition of the nation–and the nation’s cities–into relief, here drawing on the Survey to present the complexity of the population of an actually quite segregated New York:

ACS 2005?

from New York Times–“Mapping America: Every City, Every Block

Should such mages serve to whet your appetite for better visualizations that embody an accurate image of the nation, folks based in the Bay Area might want to check out how the students at UC Berkeley’s I-School stack up against Fox News, Huffington Post, the New York Times, and even Francis A. Walker, by looking at the results that are now showcased at http://www.ischool.berkeley.edu/newsandevents/events/finalprojects2013.  (The website promises to present recent graphic charts of Changing BehaviorsEnhancing Information Systems, and Information Organization and Tools that have been refined over the last three years.)

Leave a comment

Filed under data visualizations, GIS, infographics, political economy, statistics

The Density and Habitability of Urban Space

How to map the compelling question of urban livability?

The journalist, blogger and economist Paul Krugman sought to qualify exactly which pressures drive the costs of urban housing recently in “Conscience of  a Liberal.”   Krugman qualified how population density effects urban housing prices, by noting the importance of mapping density in new ways.  Scarcity of space or land constraints might seem to shape prices from a point of view of supply and demand, but Krugman argued that we can attend to a new urban geography by looking at aggregate rather than standard population density alone:  in response to Noah Smith’s argument population density was less relevant than proximity to economic centers, he argued that the rise in the prices of extra-urban areas across America, and “hiving off” of exurban areas in such cities as Phoenix or Houston, where the relative scarcity of land does not determine prices and population density in cities’ centers can change independently from the prices on its peripheries.  Following  Richard Florida, the mapping of “population-weighted density” of census tracts that can capture the way that Americans tend to be moving toward less densely weighted areas of habitation, even if the population, nation-wide, is rising.  (While population rose 10% in the US 2000-2010, it also spread out across urban areas, which became less densely weighted over time.)

This new picture of the United States that has cultivated a “lifestyle” of a less dense mode of living seems to be based on the increased valorization of “inhabitability” as a metric for home ownership in ways that are for more widespread than the movement to the suburbs in the 1970s:  we are all moving toward less weighted centers of population.  The hiving off of new suburban regions on the map that Krugman describes have not only pumped-up real estate is a reaction, in part, of voting with one’s feet against the over-density of urban space.  To be sure, although New York City property values are high, if we all lived in at the amazing density of New York City, the entire 7 billion inhabitants of the globe could fit into a geographic space similar to Colorado–a state which currently has a low density in much of its expanse, but whose pockets of population density are interestingly clustered close to adjoining regions of medium density.


One thinks of the much-circulated infographic, repopulating the world within the United States at the density of global metropoles, that suggests the exceeding density that urban cities have come to acquire, and the continued density that a city such as New York or Paris holds, in comparison to San Fransisco or Houston:


The variation of cities suggests some reasons for, say, the greater “livability” of certain places, as well as the distinct nature of the urban experience in a broad selection of cities across the world.

So what does this “hiving off”–distinct from suburbanization, as it suggests an exurban experience even outside of cities with relatively low population density, entail?  The 1970s phenomenon of widespread suburbanization, allowed by the creation of major highways and arterials that fed by commutes to urban centers like New York or Los Angeles, which served as an endangering withdrawal from many public schools in cities and urban centers, reacted to the intense pace of urban growth and in-migration.  Rather than based on suburban homes, the hiving off of communities echoed a similar demand for “habitability,” but was now about creating new spaces where exurban communities could be created, and how we could use the concept of aggregate density to measure them.

Krugman offered a more specific updating of James Kunstler’s compelling critique of urban sprawl.  when it was launched in the 1993, Kunstler decried the unwarranted expansion of suburban housing and malls in a manifesto that responded to America of the late 1980s, The Geography of Nowhere, as “the greatest misallocation of resources the world has ever known”–what he called “dead spaces” that he saw as produced by the new attitudes to space that the automobile allowed, stripping vitality from urban centers, and creating the rise of quasi-spaces of strip malls as suburban heterotopia, displaced amusement parks of purchasing, lined with unproductive spaces of parking lots.  This partly architectural perversion of lived space, for Kunstler, led to an evacuation of vitality.  Krugman focusses rather on the place that “habitability” has gained among the economic forces that shape the prices of urban homes:  his argument begs to be mapped, or graphically modeled as well as understood by statistical indices.

Economic thought might be playing catch up with cartographical modeling here:  the critical mapping of density, after all, began right after the 1870 census, when Francis Amasa Walker, superintendent of the census, offered a new model for mapping the population density of the nation not according to county, or electoral districts, but according to the actual contours by which the citizenry inhabited space:  dividing each square mile into its relative density of inhabitants, Walker distinguished those zones with less than 5 inhabitants/sq. mile from those with 5-15, 15-40, 40-75, 75-125, and over 125 inhabitants/sq. mile in 1872, and from that year undertook a number of projects with congressional funding that mapped that distribution against wealth  per capita, to create the first map to record the distribution of wealth in the nation’s geographic space, revealing the dynamics of income distribution along spatial axes–in what Susan Schulten aptly described as the nation’s first example of an infographic.  (Never one to not consider the uses of mapping populations, he created similar maps for the US Congress that mapped population density against disease, illiteracy, public debt, ethnic groups, and the birthrate.)

Desnity of Population [of US]

Such a clear figuration of density does not correspond to actual urban areas, given the former density of rural populations–or have fine enough grain to distinguish property value in them.  But it roughly reflects property values in the nation.  But its generalization of density across each county provides a useful metric to understand variations in the generation of income, tax revenues, and birth rates.  While a model of density-independent population growth may be applicable to animals–plants, insects, mammals, and other seasonally reproducing species, population growth is rarely independent from density in human populations:  and with the increasing segregation of age cohorts in contemporary America–and the graying of some cities, and lower age-level of others, the metrics of growth are especially not uniform.

A more contemporary, but less statistically exacting, version of the infographic–stated in the bluntness of th modern infographic, might provide the following portrait of the relative density of individual states in the union:

sq kilometer

At a finer grain, we find a density map as the following, which provides far more refined detail:

specialreports_2edb.population density usa

And we can zoom in for density of regions, in ways that provide a clearer picture of the Southwestern states in terms of how they broke down by counties, over time, around metropoles:


To be fair, such a general level of density is not that relevant to the points that Krugman addresses in this specific issue, where he advocates the need for moving away from such generalizations of state-wide population densities, which are less meaningful for microeconomic questions, to examine the organization of the city as a unit in ways that will allow us to distinguish the changing character and economics of urban areas.  Krugman considers how changes in urban geography across the US reflect a new paradigm or model of population density, he asks us to consider new tools to map where ‘average’ Americans live by the “weighted” density of population distributions similar to the sort of maps Walker first innovated, and focus attention on the relative density of different zones not of the nation but individual urban areas.

He argued that overall ‘density’ means less today than ‘aggregate density‘ or the relative density of neighborhoods or sectors of new urban growth, and used this as a dynamic structure to  map population density according to weighted density to better  understand how the “average” American lives.   He does so to offer a critique or assessment of how his fellow-economist Bill McBride suggested attending to population density as they were correlated with housing costs his competing finance blog, Calculated Risk.  Does overall ‘density’ mean less as a measure of housing cost, Krugman wonders, than ‘aggregate density’ (or relative density) within selective regions of cities?

Like Paul Krugman, I would question the meaning of density as an index of value, but I would suggest we need to develop tools to understand the dynamics of density that involve a more fuzzy or less quantifiable variable of habitability, which I would argue if less easily quantifiable becomes more evident in maps.   And in the remainder of this post, I’d like to show how maps offer a measure of the fuzzy value of ‘habitability’ as well as its inverse, a perceptions of ‘uninhabitability,’ might be loosely understood as a way to map the increase or the radical diminishing of property values rather than density alone–or understandings of how density drives demand.

We might  use these maps to include other variables that shape density, and use maps  to do so.  The question of how we can try to map habitability, both in terms of access to open spaces, access to markets, and attractive environments, might create an even better index of value.  In San Francisco, for example, where property rates have risen dramatically in the past several years, and whose neighborhoods are confined to a small area of land:  Oakland is itself much less dense, if also confined by similar topographic limitations, notably the mountains:  but four of the ten most expensive housing markets in the country, including Santa Clara, San Mateo, San Francisco, and Marin, are in the Bay area, according to the recent March 2013 Report of the National Low Income Housing Coalition–and in the state that requires the greatest hourly wage to rent a two-bedroom apartment.

out of reach hourly wage per two bedroom

To be sure, the variations of real estate and population in a state like California are closely tied, given the huge variations of density across the state.


Yet the profound variations in the Bay Area, to take one example that is clotted with a red of high density, indicates profoundly striking differences in desirability:


And such dramatically differing variations across the region seem to be distorted by the demand for living near a not-densely populated area such as Silicon Valley, where real estate prices are suggested in this Kwelia map of 2012 which illustrates the relative geographic distribution according to median incomes–and suggests the very real distortions in real estate that can occur independently of actual or weighted density.

screen shot 2014-01-17 at 9-2.14.13 am.pngKwelia

Whereas the debate between the economists turned around the measure of density, it’s unclear why in an age of complex data visualization Krugman omits variables beyond density or to map determinants of urban spaces.  As an economist is wont, Krugman tends to confine himself to contrasting the relative value of alternative models of housing markets in urban space.  If we accept the model of scarcity as the motor of rapidly escalating real estate prices, it’s hard to explain the rise of attraction to less densely populated urban areas in urban areas, however, or the re-use of less inhabited areas as new areas of density.  If we might assume rising population means rising housing costs, as Smith has since noted in an update to his post, transit policy and presumably other public policy issues related to urban infrastructure can influence or shape housing costs–issues that introduce habitability as much as density as an influence on housing costs.

If we weigh models of density and populated-weighted density as models of urban environments, population-weighted density might help explain the social geography of urban spaces at a time when average American increasingly live in somewhat less dense neighborhood:  during the first decade of the century, population has broadly “spread out within metropolitan areas” and brought a “hiving off” from cities and led many cities to abandon the model of a vital urban core is often associated with  “quality of life.”   But the generic term does not reflect how urban space becomes mapped as habitable to different populations and demographics:  for just as cities are mapped in new ways by public projects of urban infrastructure, people map areas of settlement around the contours of urban space by preferring zones,  in turn attracting further interest in them.  An interesting liberal strain of response to Krugman’s post bemoans the end of Jane Jacob’s model of the city as constituted by local communities and the positive sense she gave density as causes of the community and vibrancy of individual neighborhoods, which she valued as the basis for urban life.  (Many also critiqued his argument as not acknowledging the economic models of different cities, from how the absence of economic vitality thins cities like Detroit to how hi-tech creates a new geography of work.)  But in mapping population-weighted density only in numeric terms, Krugman may ignore the appeal that the neighborhood–as much as geographic location–has on real estate, and the increasing flexibility with which a neighborhood–even of different but overlapping socioeconomic backgrounds among their residents–is redefined in particularly flexible ways.

Less attention is paid in both models  to the constraints by which people map their own living spaces, even when constrained by a city’s complex and often unfair geography.  Parts of urban environments are recast as more habitable based on different criteria, often in response to the made or ‘created’ geography of the city–or fail to be recast in habitable terms.  The growth of desirable land in Los Angeles responds to the constraints with which freeways have sculpted urban space, and indeed in ways bound by the 405 in a sort of deterministic way, and the creation of the former Cypress Freeway shaped–and in cases destroyed communities in Oakland, California.  Vice versa, patterns of urban settlement and housing prices helped the investment in an urban infrastructure as transportation in Boston or New York, because they led populations to map their lives in urban space.

Does density alone shape market prices?  In a sense, yes:  but density is a reflection of deeper indices that one might better term ‘habitability’ given their manifold causes and considerable variations.  Changes in workplace practices and locations encourage a shifting relation to the urban space, as more people work on-line, of course, and live and work in isolated archipelagos around cities:  the vitality of neighborhoods that Jacobs celebrated has become far less important as a unit of economic meaning with increased geographic mobility and more dispersive work environments, including telecommuting. As a result, the framing of a debate between population-weighted density or the availability of “space” as determining land value omits the notions of habitability shape urban space: rising land prices in LA, for example, reflect the shifting attitudes to space created by a system of freeways that shaped land values–rather than scarcity–and defined a map of urban settlement outside corridors of transport that, more than mountains, shape Los Angeles neighborhoods.

This reflects the desire for a way to maps how areas become understood as habitable:  and in Oakland, where property values have risen not only in response to the rising rents in SF, but as the city has been defining new notions of habitability by restaurants, urban gardens, and public spaces like farmers’ markets.  By trying to identify a model for urban settlement, geography of population density becomes a stick of measurement rather than the remaking of urban spaces:  multiple complex constraints on the habitability of urban environments limit the development of urban land values.  In  the recent recouping of housing prices in the Bay Area, for example, areas of recent price increases were mapped based on data from Zillow.com–and metadata from Google Maps–to find a considerable recuperation of costs in San Francisco and Fremont alike, despite their quite different density, and a mild recuperation in many of the low-density areas of Berkeley and Oakland, as is rendered evident in the regions of light rose coloration below.


Bay Area Property Map

Conversely, if much of Richmond, Brentwood, Antioch, Union City and Hayward remain selling at a far lower price than in the prior boom, many Oakland properties sell at a brisk price, and hi-tech areas like Mountain View–and some areas near Walnut Creek–are positively booming.  To some extent this has to do with access to economic centers:  the less populated but also less desired and more removed Daly City is nowhere near recovering its boom prices, while Mountain View has surpassed boom prices.  But the widespread recovery speaks to issues of habitability, as much as density.

As a counterfactual, let’s look at some actual urban environments where the question of habitability responded to a striking lack of investment in creating a habitable urban space with adequate infrastructure or access to green space–one crucial measure of habitability.  Consider, for example, the considerable distances at which less valuable property lie from green spaces or urban parks in Los Angeles, creating an endemic that led the “LA Streets” Blog Santa to bring the city more green spaces last Christmas.  (African American neighborhoods enjoy .8 acres per thousand people, in contrast to the recommended 10 acres per thousand inhabitants, or the median 6.8 in high density American cities.)  Pink regions here map a distance from park exceeding the LA average, most of which are concentrated away from the shoreline or surrounding national forest:

Access to Parks in Los Angeles


The value of land responds to how we map urban spaces as habitable often shapes property values, as much as the questions of the availability of space. Something like the new Stamen map that focusses on the readability of mapped space by watercolor tiles, pioneered by Zach Watson, provides a new resource to look at urban space to analyze questions of habitability without being distracted by data overload, in what was intended as a visually simpler and pleasurable alternative to the tedium of Google Maps, provides a manner to differentiate neighborhoods in San Francisco–albeit a region whose high real estate prices don’t need much explaining:


SF Greenspace Stamen


Krugman provoked an interesting comment that advances in GIS could provide new tools to include aggregate densities in urban spaces.  Indeed, working with the above Stamen map, or with a time-lapse model of aggregation might provide tools to show the shifts in how space became ‘habitable’ or worthy of inhabitation.  These tools offer the ability to map a new geography of work:  one could model of the shifts in economic redistribution of wealth in urban space, or relative distance of residents to workspace, that might map the relation of neighborhoods to space beyond their relative income or real estate costs.

Flexible GIS models might also help reveal the different maps people create within a city as an environment in powerful ways.  In juxtaposing the alternatives between two models of urban life–the metric of population growth, which  creates competition for limited resources of land in a somewhat Malthusian fashion, or population density, or urban settlement informed by how agglomerations understood in terms of population-weighted density shape demand in extra-urban spaces or urban peripheries for land that changes land prices–the mutable map of the urban fabric is either dramatically oversimplified or ignored.   When land is constrained, the first train of thought goes, rising populations increase the price of land faster than inflation; the questions of how people map land as habitable–and desirable–more informs the second proposition, pointing to the rise of “edge-cities” on extra-urban space that become desired centers of population.

That choice between population and weighted density seems a bit artificial,  because both are abstracted from how people make concepts of neighborhoods out of open space–by mapping it as their own, or by not being able to do so.  Today, New York has of course become the model of a high “population-weighted density” city, based on the average land density of its inhabitants, or the relative density of its populated areas.   But whereas in the 1970s suburbia gained value in the 1970s around greater New York, leading them to abandon a city that became defined as less “habitable,” and to define preserves away from commercial activity, we now value habitability as a relation to a neighborhood, or something like one.

The remapping of the urban core as economically vital paralleled a shift in the understanding of its “inhabitability”– mapping not only its density, but what makes it conducive to habitation or less appealing as a living space.  Could we map the shifts in aggregate population density to investigate patterns of land value, to examine the rise of prices in New York and in regions of uneven density of population spread out within large cities?  One point here would be to chart the spaces made and created in individual cities over time that informed property prices by creating new models of the habitability of the same spaces and of urban life.   To do so, I’ll use maps to illuminate shifting attitudes to space.

Before considering urban geographies, it’s interesting to map a distribution of the predominance of bars to grocery stores in a cute map that I can’t resist–with surprising concentration, the number of bars exceeded grocery stores in urban areas clustered in the Midwest and parts of New England, and areas of the plain states and spots of the Southwest.  It would be interesting to pursue this map into urban areas–does the prevalence of bars over grocery stores suggest a greater role in serving neighborhoods, or a preference for the social function that bars serve in specific urban or rural regions?

Grocery Stores v. Bars

The map may tell us about a culture of sociability–in the colder midwestern north, and New England, bars are important sites of aggregation.  The map speaks about how folks create habitability in urban space in response to different constraints.  We can use this model, more importantly, and this will be the end of the post, to expand the criteria by which we look at urban spaces beyond density that can help raise new questions about the relationships between property value and urban milieu.

Others might reveal the increased livability of different spaces within the city, as this map of the green spaces in New York City, perhaps the city that possesses the densest urban core, in ways that map urban resettlement outside the region of its greatest densities by the foundation of community gardens and urban agriculture, discussed in an earlier post, which notes those places that accept volunteers by blue dots:


Community Gardens NYC region


The map reveals not only an expanding urban green-space, but the re-envisioning of densely occupied areas as of more livable urban spaces, in ways that mapping population density or aggregate density alone don’t suggest.

Still other maps of urban environments reveal the constraints on the very maps we make of urban space that are so central to housing prices–a process perhaps missing from Krugman’s discussion of two attitudes to the analysis of land-value.    Maps are a central tools of real estate to attract attention to  property and shape urban space into neighborhoods, and fit notions of habitability into existing spaces.  Many structures define these spaces of habitability, not necessarily understood in terms of weighted density alone.  Los Angeles is not, as Krugman’s piece suggests,  “hemmed in” by mountains that constrain city’s sprawl; its low density seems to run out to the Valleys, as freeways define the urban space and create  anew map of property values and more habitable and valuable land.  It almost seems land properties rise in their remove from the freeway, where the relation of land to freeway is more easily mapped, much as much of the value of land in Santa Monica seems to recede in value from the Pacific. Freeways or overpasses divide urban space and help structure maps of habitability; freeways of easy commute may, conversely, increase the value of hived off aggregations of “edge” cities, by improving connectedness to other environments.

We can map moreover habitability in different ways.  Cities like Oakland CA offer a sort of counter-example of a space where habitability set off a rapid rise of real estate values and the price of land:  there is no clear confine to urban space, but the spread of restaurants, green spaces, farmers’ markets, urban farms, and open public spaces fashioned a new notion of the inhabitability of distinct areas of its urban space in successive districts, earlier shunned, even when considerable crime continues to plague nearby parts of the city and the city hears the screaming sirens of police racing to the victim or scene of a homicide.  (If some would say real estate prices derive from the flight from the property values of San Francisco, that only seems to be the icing on the cake.)

The value of land is not a simple calculus of limited supply, but of the mapping of urban space–first taken up by real estate agents who sculpted Oakland’s neighborhoods, but informed first by a shifting sense of the habitability of the land, whether in the middle of an old port-city or the more trendy area of the Temescal.  Demographics prefer different models of habitability; the age of the population of Oakland seems exceptionally low, for example, as they might first define the city as more habitable.  The shift of populations toward less densely settled urban areas reflects this shift in how we have come to map habitability, one might say, even if aggregates of population density reflect land value; without numerical data to show for it, I  offer that the value of land reflects the relative scarcity of habitable areas in urban space, but also shifts as notions of habitability emerge.  In New York, the premium on habitability drives the growth of new neighborhoods, from the allure of Soho or the East Village to Brooklyn Heights as more habitable than more crowded urban environments.  Imaginative ways to map public spaces create habitability in urban regions; the lack of parks in Los Angeles fuels a somewhat undifferentiated urban expansion and settlement.

But there are huge obstacles to mapping the habitability of space in the very regions and cities that habitability might be encouraged, as we find habitability to become, in much of California, the preserve of the wealthier few, and urban spaces constrained by barriers to their physical improvement.  Take a simple map of the distribution of parks in the state, and reveals radical differences between the parks/inhabitants: the polemic map ties park-poor areas to income, contrasting darker greens to show greater availability of acreage/thousand inhabitants in the San Francisco than Los Angeles areas:




Income variations allow the possibility of a flexible and inventive relation to maps of habitable space–as the clumping together in aggregates or scarcity of desired urban lands reveals.   As we become more movable in our work, and develop a more flexible notion to workspace, this shifts the relations of local routines of personal and work space, and there is considerable choice about the spaces where we live:  workspace is less dense in different cities, to be sure, as silicon valley or other IT belts like the Research Triangle around Durham NC, but the premium on the flexibility of defining one’s own model of habitability seems to determine the new urban geography Krugman describes.  Further anecdotal evidence, beyond the United States but in North America, might point to the shifting geography in urban Ottawa, where the residential Glebe was distinguished by rising residential real estate.   Within the shifting infrastructural map of the city, one maps one’s own personal space.  Indeed, the bizarre settling of San Francisco’s sky-rocketing real estate prices is partly driven by the arrival of highly paid executives working in Silicon Valley, shuttled to their work in select buses whose routes allow them to select an alternative to public transit, as discussed in an earlier post.  The paths of flights into San Diego, to use another example, skirt more select residential areas like Mission Bay, create a zone marred by the rumble and whine of overhead airplanes.  And the organization of urban space no longer reflects how we map and understand urban geographies as habitable–or come to dismiss them as simply uninhabitable and worth less.

Only by balancing what maps tell us about how an urban region gains value for its habitability can we effectively broaden the discussion and use maps to measure inhabitability against degrees of uninhabitability and will we be able to extend the critical evaluation of value beyond questions of supply and demand.

Leave a comment

Filed under human geography, infographic, mapping density, Mapping Park Poor Neighborhoods, Oakland CA, relative density, urban geography, urban growth, Urban Livability