Geoffrey Skelley compares the percentage of the Democratic primary vote won by Hillary Clinton in 2008 with the percentage she won in 2016: among other things, she was up sharply in the Deep South and down sharply in the industrial Midwest and Appalachia. “While the universe of voters participating in 2008 and then 2016 changed considerably thanks to mobility, interest, and mortality, our map suggests that many ’08 Clinton voters became ’16 Sanders voters, and many ’08 Obama voters became ’16 Clinton voters.” [Daily Kos]
Land does not vote and we can’t judge gerrymanders simply based on geometry. Districts aren’t just abstract shapes on a map, but collections of actual people and voters. Ultimately, the outcomes produced by a particular map matter far more than a map’s appearance. Comparing the actual congressional districts to plausible alternatives in Maryland and other states demonstrates both how gerrymandering is more complex than merely grotesque shapes, and that Maryland is far from the worst partisan gerrymander nationwide.
Last month a Bloomberg story looked at the racial implications of Amazon’s same-day delivery service, which, the story demonstrated in a series of maps, tended to exclude predominantly black ZIP codes.
The exclusions were basically driven by the data: where their customers were, driving distance to the nearest fulfillment centre, that sort of thing. But the issue, it seems to me, is that the demographics behind the data are not racially neutral (something that Troy Lambert’s analysis for GIS Lounge, for example, fails to address): Amazon basically failed to ask its data the next question. Be very careful of why your data is the way it is. In the event, Amazon has since announced that excluded neighbourhoods and boroughs in Boston, New York and Chicago will get same-day service.
(Full disclosure: The Map Room is an Amazon associate.)
Of all the statistics related to gun violence in the U.S. that could be mapped, this has got to be one of the most bizarre: from the Washington Post’s Wonkblog, the number of times a toddler has shot someone since 2015, by state. [Landon Schnabel]
Neil Freeman of Fake Is the New Real takes this whole “reorganize the United States into states with equal population” thing just one step too far:
the United States divided into fifty concentric states with equal population pic.twitter.com/cLQ1wCdfbR
— Neil Freeman (@fitnr) April 12, 2016
My eyes are bleeding again. [Kottke]
In a piece for the New York Times, Parag Khanna—author of the forthcoming book Connectography: Mapping the Future of Global Civilization—argues that super-regions and urban clusters, rather than the 50 states, should be the focus of future planning.
First, there are now seven distinct super-regions, defined by common economics and demographics, like the Pacific Coast and the Great Lakes. Within these, in addition to America’s main metro hubs, we find new urban archipelagos, including the Arizona Sun Corridor, from Phoenix to Tucson; the Front Range, from Salt Lake City to Denver to Albuquerque; the Cascadia belt, from Vancouver to Seattle; and the Piedmont Atlantic cluster, from Atlanta to Charlotte, N.C.
Federal policy should refocus on helping these nascent archipelagos prosper, and helping others emerge, in places like Minneapolis and Memphis, collectively forming a lattice of productive metro-regions efficiently connected through better highways, railways and fiber-optic cables: a United City-States of America.
Note that this isn’t quite the same as, say, reimagining the U.S. as fifty equal states or Pearcy’s famous 38-state thought experiment: this is an argument against using state boundaries for planning purposes. (The EU has similar regions for similar purposes, I believe.) Makes for a very interesting map, though. [Tim Wallace]
In 1916 frost maps that show the average dates of the last spring and first fall killing frost, Slate’s Rebecca Onion sees the history of climate change, given the growth in the length of the growing season since then. (Trying to find a modern-day example for comparison; frost maps don’t appear to be updated as rigorously as, say, hardiness zone maps.) [Slate Vault]
For the first time, USGS forecast maps that measure the potential damage from earthquakes in the coming year now include human-induced earthquakes, such as those caused by hydraulic fracking. (Oklahoma looms large for that very reason.) Maps for the western U.S., where a different methodology is used, presume that all earthquakes are natural in that region. [Max Galka]
Vox’s lead exposure risk map takes a nationwide look at a crisis some might have thought was limited to Flint, Michigan. “The areas where kids are at highest risk of lead exposure—an estimate calculated using government data about the surroundings—are scattered all across the country.” Lead exposure data is hard to come by, so exposure risk is calculated based on Washington State’s methodology, which uses age of housing and poverty as risk factors. [Mapbox]
CityLab’s Laura Bliss has a second post on women and cartography, this time focusing on the work of 19th-century women cartographers, geographers and educators in the United States. The Library of Congress’s map blog, Worlds Revealed, focuses on the work (and maps) of one of those women, Emma Hart Willard.
On Cartastrophe, Daniel Huffman points out the problems in Decision Desk HQ’s interactive cartograms for the U.S. presidential primaries, which maintain a state’s shape but resize the counties as the results come in. “Unfortunately, in so doing, they shuffle the counties around any old which way. The Lower Peninsula of Michigan has 68 counties in reality, the Upper Peninsula has 15. But Decision Desk HQ has shoved most of the counties into the Upper Peninsula, which now has 58, vs. 25 that remain in the Lower Peninsula,” Huffman writes. “This means that we can’t really see spatial patterns, which is sort of the point of having a map.” [via]
There was a time, not too long ago, when our Super Tuesday map would have been impossible to put together and display. Even earlier in the digital era, a complete vote-totals map wouldn’t have been available until every ballot was counted at the end of the night. (Not to mention that in the print-only era, no map would be available until two days after the vote, and then often only in black and white.)
The New York Times graphics department invariably does first-rate work, and their interactive maps of the U.S. presidential primary and caucus results are no exception. You can zoom in, you can get results by county or congressional district (depending on the state), you can choose to view margin of victory (see screencaps below) or each candidate’s vote share.
The Democratic candidates as of March 9:
And the Republican candidates:
Today’s xkcd is a map of the United Sta—wait … what has Randall done? My eyes … they’re bleeding.
FiveThirtyEight maps the Facebook likes of the U.S. presidential candidates: “If Facebook likes were votes, Bernie Sanders would be on pace to beat Hillary Clinton nationwide by a nearly 3-to-1 margin and Donald Trump to garner more support than Ted Cruz and Marco Rubio combined. Anything seems possible this year, but, still, be careful how you interpret these numbers: Facebook likes are not votes.” They ain’t kidding—Ben Carson?! [via]