There’s more than one way to depict data on a map. At the last Esri user conference, Sarah Bell, Kenneth Field and John Nelson demonstrated different ways to map the results of the last U.K. parliamentary election, and how they changed from the previous election. The video of their presentation is attendee-only, but Ken and John have posted about how they each went about their tasks: here’s Ken’s post and here’s John’s; plus, as is his wont, John has posted a video.

The Map Men look at gerrymandering on U.S. electoral district maps. A reasonably comprehensive primer on the subject even if comes from a couple of Brits baffled by the subject. And they finish with a surprisingly sharp point: gerrymanderers wouldn’t know how to draw maps like these if voting intentions weren’t predictable.

Yesterday, U.S. Vice President and Democratic presidential nominee Kamala Harris announced Minnesota Governor Tim Walz as her running mate. Among other things, Walz is a former social studies teacher and early adopter of GIS as a teaching tool, and has nerded out on geography and GIS throughout his political career, both in the U.S. House of Representatives and as governor. Walz even spoke at the 2024 Esri User Conference in San Diego last month (as someone married to a high school teacher, I can say this: he totally talks like a high school teacher). See also this summary of the talk, and Walz’s map nerdery in general, in the Minnesota Reformer.

There has been an outbreak of anti-trans legislation at the state level in the United States, and Erin Reed has spent the last three years tracking it. Her anti-trans legislative risk map measures the extent to which trans people are endangered by such legislation, whether it’s already on the books or could be the offing before the next election. The map reveals, no surprise, a polarized America: one where some states are racing to put anti-trans laws on the books while others enact protections and set themselves up as safe harbours.

Previously: Mapping Safe Washrooms.

Julien Gaffuri’s map of the second-round results of the French presidential election is, as you can see, extraordinarily busy—and, by the way, extremely processor-intensive: it will slow down your machine—because it’s at the commune level and each circle is scaled to population. (News flash: Paris has lots of people in it.) And those circles are striped circles: the proportion of the votes is indicated by the area taken up by a given colour. The map of the first round results shows more stripes (because more candidates) but is by department, so it’s a little easier both to read and to see how the striped circle format works. It’s an interesting alternative to a choropleth map, and a bit less ambiguous.

Le Parisien’s maps aren’t interactive, nor are they particularly large, but they illustrate other aspects of the results, like the abstentions, voter turnout and differences vs. the 2017 vote. The Guardian’s maps are low on detail but provide similar information. Libération’s map, on the other hand, is a cluttered mess, showing each commune as a proportionally sized dot. [Maps Mania]

Previously: 2022 French Presidential Election (First Round).

Previously: Mapping the 2017 French Presidential Election (First Round).

Previously: A Redistricting Roundup; The Washington Post Examines Proposed Congressional District Maps.

Gerrymandering in Texas

The New York Times and Texas Monthly look at the bizarre shapes in the new congressional electoral map of Texas, which gains two new representatives. Texas Monthly’s Dan Solomon: “Across the state, there will be one more majority-Anglo district than under the prior map, and one fewer majority-Hispanic one. The two new seats Texas was awarded for its booming population will be placed in Austin and Houston—and even though non-Anglo newcomers made up 95 percent of the state’s population growth the last decade, both districts will be Anglo-majority.” Kenneth Field has some thoughts. [Maps Mania]

Making Redistricting More Fair

A Surge of Citizen Activism Amps Up the Fight Against Gerrymandering (Bloomberg): “From North Carolina to Michigan to California, voting rights groups, good government advocates, data crunchers and concerned voices are finding new ways into the fight for fair representation, via informational meetings, mapping contests, testimony workshops and new technologies.”

Can Math Make Redistricting More Fair? (CU Boulder Today): “Clelland doesn’t advocate for any political party or for any particular redistricting proposal. Instead, she and her colleagues use mathematical models to build a series of redistricting statistics. These numbers give redistricting officials a baseline that they can compare their own maps to, potentially identifying cases of gerrymandering before they’re inked into law.”

Previously: The Washington Post Examines Proposed Congressional District Maps.

Redistricting—and gerrymandering—is one of the blacker cartographic arts. With the release of data from the 2020 U.S. Census, and the changes in state congressional delegations—some states gain a seat or two, some states lose a seat, others are unchanged—new congressional maps are being drawn up for the 2022 elections. The Washington Post takes a look at proposed congressional district maps in Colorado, Indiana and Oregon, and what their impact may be.

The New York Times maps partisan sorting in America—the tendency for voters to self-select into areas where people think and vote the same way they do—down to the neighbourhood level.

The maps above—and throughout this article—show their estimates of partisanship down to the individual voter, colored by the researchers’ best guess based on public data like demographic information, voter registration and whether voters participated in party primaries.

We can’t know how any individual actually voted. But these maps show how Democrats and Republicans can live in very different places, even within the same city, in ways that go beyond the urban-suburban-rural patterns visible in aggregated election results.

It goes beyond racial, urban vs. suburban vs. rural and house vs. apartment splits, to the point where researchers are wondering whether Americans are “paying attention to the politics of their neighbors” when they decide where to live. This has implications not only in terms of electoral targeting (e.g. gerrymandering, voter suppression), but in terms of basic social cohesion.

The maps are based on research by Jacob R. Brown and Ryan D. Enos published earlier this month in Nature Human Behavior.

Previously: Red and Blue vs. Gray and Green.

Kenneth Field has released a dasymetric dot density map of the 2020 U.S. presidential election results. One dot equals one vote. “Data at a county level has been reapportioned to urban areas. Dots are positioned randomly.” It’s in the same vein as his 2016 map, which went all kinds of viral when he posted it in early 2018. A high-resolution downloadable poster is here; an interactive version is here.

Previously: Kenneth Field’s Dot Density Election Map; Kenneth Field’s Dot Density Election Map Redux.

xkcd did another map thing, so I have to post about it; it’s a rule. This time Randall revisits the design of the map he did for the 2016 U.S. presidential election, in which one figure represents 250,000 votes for each candidate. In a Twitter thread, he explains the rationale for the map:

It tries to address something that I find frustrating about election maps: Very few of them do a good job of showing where voters are. […] There are more Trump voters in California than Texas, more Biden voters in Texas than New York, more Trump voters in New York than Ohio, more Biden voters in Ohio than Massachusetts, more Trump voters in Massachusetts than Mississippi, and more Biden voters in Mississippi than Vermont.

Previously: xkcd’s 2016 Election Map.

Previously: Mapbox Elections.

Kenneth Field’s virtual talk on the cartography of elections, given to University of South Carolina students on 25 September 2020, is now available on YouTube. “It explores the way in which map types and their design mediate the message, and using examples from elections shows different versions of the truth.” Also includes material from Ken’s forthcoming book.

The New York Times uses the colours in aerial images as a proxy for political leanings: rather than red-and-blue electoral maps, the political landscape, Tim Wallace and Krishna Karra argue, is more green and gray.

The pattern we observe here is consistent with the urban-rural divide we’re accustomed to seeing on traditional maps of election results. What spans the divide—the suburbs represented by transition colors—can be crucial to winning elections. […] At each extreme of the political spectrum, the most Democratic areas tend to be heavily developed, while the most Republican areas are a more varied mix: not only suburbs, but farms and forests, as well as lands dominated by rock, sand or clay.

This is a generalization, to be sure, but so are most political maps, and the notion that urban areas tend to vote Democratic while rural areas tend to vote Republican isn’t what I’d call a revelation. Still.

Previously: An Online Atlas of Canadian Election Results; A Cartogram of Canada’s Election Results; More Canadian Election Maps; Mapping the Canadian Election Results: Technical Details.

This week Mapbox launched a tool for the upcoming 2020 U.S. elections: Mapbox Elections, “a resource to help individuals, journalists, and organizations cover the elections, analyze the results, and build modern maps to display it all.” Their first product is a dataset including U.S. presidential election results from 2004 to 2016.

Researchers have released an interactive map showing Canadians’ opinions about climate change—whether it’s happening, and what we should do about it—and, more significantly, the regional variations in that opinion, down to the riding level. Not surprisingly, the oil- and coal-producing regions are much more likely to be climate skeptics.

The map is based on surveys of more than 9,000 Canadians taken between 2011 and 2018, which raised my eyebrows a bit: public opinion can change a lot over seven or eight years, after all. But the researchers did so to get a more accurate sense of regional opinion: opinion polls are usually based on a small national sample; regional breakdowns of that sample have large margins of error, and getting accurate regional samples would be a lot more expensive. More at Global News.

Meanwhile, Mark Gargul writes in response to Ken’s critique of his cartogram of the election results. Mark describes himself as an amateur and readily admits that other cartograms are “clearly more aesthetically pleasing. On the other hand, I was going for something different with my cartogram—specifically, to try to preserve riding-adjacency as much as possible.”

@kennethfield
… On the other hand, what I was going for was preserving, to the extent possible, riding adjacency. If Markham-Stouffville shared a border with Markham-Unionville on a real map, I wanted that border on the cartogram. Hence, the ugliness.

— Mark Gargul (@GargulMark) October 26, 2019

The other thing Mark was going for in his cartogram was to indicate the urban-rural split: metropolitan areas are given a black border: it’s easy to see which ridings are in Montreal or Toronto; seats that are partially urban and partially rural straddle those borders.

I'll summarize for you guys what I let Ken know in more detail: I wasn't going for pretty, but I was going for illustrating the rural-urban split, which doesn't come across well in the other cartograms or maps I have seen

— Mark Gargul (@GargulMark) October 26, 2019

So it’s doing several things at once that may not be immediately apparent.

Previously: A Cartogram of Canada’s Election Results; More Canadian Election Maps.

Update on mapping #CanadaElection2019 by me & @williamscraigm. Data errors fixed. Value-by-Alpha Unique Values, Prop Symbol (both modified by pop density), Dot density (winners), Dot density (all parties >1%). 1 dot=100 votes. Screen grabs. #NotMercator pic.twitter.com/MJbIRRGqsw

— Kenneth Field (@kennethfield) October 23, 2019

I hit “Publish” too soon last night. Kenneth Field and Craig Williams put together a series of maps showing the Canadian election results in a number of different ways: we have a value-by-alpha map, a proportional symbol map, and two kinds of dot density maps: one showing the winners, one showing all votes per constituency. (One dot equals 100 votes; the dots are spread evenly across constituencies, even when people aren’t. You can’t have everything.) And it’s on the Lambert, not the Mercator.

Speaking of the Mercator. Maps Mania’s roundup of Canadian election results maps notes that the Canadian media’s interactive maps (e.g. CBC, Global, Globe and Mail) invariably resorted to Web Mercator, largely because of the mapping platform used. (In-house infographics team? Don’t be ridiculous.) Web Mercator is singularly bad for Canadian election maps, because Nunavut: it’s the largest electoral district by area (1.9 million km²) and the smallest by population (31,906). It’s enough of a distortion on the Lambert: Mercator makes it worse.

As for cartograms, Ken hated the one I posted last night; Keir points to Luke Andrews’s Electoral Cartogram of Canada, which is a bit nicer, and uses only one hexagon per riding instead of seven. Keir also points to this animation that shifts between a geographical map and a cartogram. It’s hard to recognize Canada in cartograms, because it’s difficult for us to grasp just how many people live in southern Ontario.

Previously: A Cartogram of Canada’s Election Results.

Dissatisfied with county-level results maps for the 2019 European parliament elections in Romania, Raluca Nicola built an interactive map that displays the results at the commune level. She explains how she built it here.

This cartogram shows the seat-by-seat results of the federal election held last Monday in Canada. It was uploaded to Wikipedia by user Mark Gargul to illustrate the 2019 Canadian federal election article, and it’s a welcome departure from the usual election results maps in this country.

(An example of the usual results map is Elections Canada’s official map of the unofficial results, in PDF format.)

Canadian election results maps generally use geographic maps, usually the Lambert conformal conic projection that most maps of Canada use (though sometimes it’s the Mercator!) rather than cartograms. Which means that Canadian maps suffer from the same “empty land doesn’t vote” problem that U.S. maps have, though it’s mitigated by the fact that vast rural and northern seats are often won by different parties: you don’t have the same sea of one colour that you get in the States.

That said, Canada is overwhelmingly urban, and so are its electoral districts. Most election results maps resort to using multiple inset maps to show the urban results. (Elections Canada’s map has 29 of them.) Gargul’s cartogram sidesteps both problems neatly; on the other hand, it’s next to impossible to find your own damn constituency (it’s hidden in the mouseover text). If the disadvantage of empty-land election results maps is that the colours aren’t representative, their advantage is that you can tell what regions voted for whom, at least if you know your geography.

So Trump is a serial map-abuser. These three examples clearly show how he uses the map for dominance and to assert his apparent power and possession. This is Trump’s America. He’s simply the latest in a very long line of leaders, politicians, dictators and many others to use maps to try and illustrate a version of the truth that has been cartographically mediated to suit a partisan purpose. Like I said, it’s not wrong to use maps to tell a certain story (apart from when the facts are clearly manipulated which is stretching truth to the realms of plain lies) but it is a case of “reader, beware.”

Previously: ‘A Defilement of a Sacred Trust’; How to Lie with Maps, Third Edition.

In a series of maps, the New York Times explores where the donors to the various Democratic candidates for U.S. president live. The maps are based on data to June 30, and include donations of $200 or more. Bernie Sanders has by far the most donors so far, and they’re distributed broadly, so the second map on the page excludes Sanders donors to tease out where other candidates’ donors are concentrated regionally.