January 31, 2012

Mapping Cyberscapes of the 2012 Republican Presidential Primary

They've given us gems like "I like being able to fire people", suggesting that we replace professional janitors with dozens of children from working-class homes in order to cut costs. And a bunch of other crazy inventive stuff. In an indirect way, they've also given us new vocabulary words, parodies and re-interpretations. They've also provided the raw material necessary for a range of user-generated, web 2.0, prosumptive behavior. So even though you may be a little bit frightened, you should also thank them -- albeit not necessarily with your vote.

But with the Republican presidential primaries already well underway (and today being the Florida primary), we thought it a good time to dig a bit deeper than the superficial soundbites coming from the candidates. So in this post we're understanding the geography of these candidates via pythagoric numerology and haruspicy. Ha! Just kidding, we will be looking at the distribution of geotagged online content like always. After all that's the whole point of the blog and something we've done previously for European political leaders, as well as the 2010 election in the UK and the 2008 US Presidential election. Just sometimes we dream about a change....

So what are the geographies of the 2012 GOP primary like? Is it possible for these cyberscapes to help us predict election outcomes? Are they total hogwash? Just pretty colors?
Mapping references to each of the original eight GOP contenders, one sees that the two current front runners for the nomination, Mitt Romney and Newt Gingrich, actually have very few references relative to the other candidates. It seems evident then that these cyberscapes show a lesser degree of differentiation between candidates than is evident in the primary results thusfar.

One of the clearest patterns is the plethora of references to Rick Perry across his home state and the very few outside of Texas. Given Texas' minimal influence on the nomination process, it's likely that he now wishes he was from Iowa. There is a similar, albeit much smaller, pocket of references to Jon Huntsman in his home of Salt Lake City, while Minnesota has a cluster of references to US Representative Michele Bachmann.

But most evident is the vast swath of territory, with no real conformity to political borders, that is dominated by references to Ron Paul. Libertarians are everywhere! It's nice that the federally funded national highway system is there to help speed their movement!
Even removing the candidates that have now dropped out of the race, Ron Paul's dominance in the cyberscape of the Republican primary field is evident. While Paul's prevalence in geocoded references isn't reflected in the polling numbers in the real election, it isn't entirely surprising. As we saw over two years ago with Barack Obama's disproportionate prominence in Google Maps content, Ron Paul's prominence online is certainly a reflection of his campaign's use of the internet as a primary organizing tool.

But since the electoral system in the US is really so dependent upon what happens at the state-level we thought it worthwhile to stray from our usual method of measuring cyberscapes on a more flexible, point-by-point basis and instead aggregate references on a state-by-state basis [1].
It doesn't seem, however, to make much of a difference in the relative visibility of the different candidates. Ron Paul's seeming dominance over the virtual landscape remains a fact, while Mitt Romney wins only his home state of Massachusetts as well as Utah and Alabama and Gingrich winning just his home state of Georgia. Santorum's "win" in Oregon is primarily due to his "Google problem", with numerous references to the alternative meaning in Eugene [2].

So, if one were to use this map as a prediction of victories in GOP primaries, Ron Paul would easily be the next Republican presidential candidate. Indeed, according to references alone, Ron Paul would have won each of the three primaries that have already taken place (of which he actually won zero).

This aptly highlights the difference between online activism and offline activism. Not that we really needed a reminder after all the protest events of last year. Moreover it will be some time (thank goodness!) before Google Maps can be used to predict presidential elections. Although we're sure that someone is developing an app as we speak.

But one thing is clear, based on these maps we feel that there despite his love of conspiracy theories about the New World Order, Ron Paul might actually be the one controlling the internet.
[1] States shaded grey are representative of no clear "winner" in the number of geocoded references to the candidates. Either there were no references to the candidates' names or at least two candidates were tied for the greatest number of references -- essentially the same reasoning as the many points with no dots on the other maps above.
[2] Which brings up the role of Google and code in how places are represented online.

January 25, 2012

New Article on Wikileaks Published in Antipode

For those interested in Wikileaks, a new publication (co-authored with Sue Roberts and Anna Secor) based on the Wikileaks mapping we did in December 2010 is now available. Drop me a line if you don't have library access.

Critical Infrastructure: Mapping the Leaky Plumbing of US Hegemony
by Sue Roberts, Anna Secor, Matthew Zook
Antipode Volume 44, Issue 1, pages 5–9, January 2012

Geopolitical mappings of the world can say as much about the vulnerabilities of hegemony as about aspirations to power. Mappings of US geostrategic interests are no exception. Recent national security priorities, the details of which were revealed in leaked diplomatic cables, include the identification of sites around the world deemed critical to the US (US Department of State 2009). From beaches where trans-oceanic cables emerge, to factories making vaccines, to maritime routes and ports, sites of particular vulnerability are assembled. The cartographic effect of this assemblage is a partial and highly distributed mapping of the fragile material underpinnings of US power.

January 18, 2012

Protest the Stop Online Piracy Act (SOPA) and the Protect IP Act (PIPA)

Today across the Internet, web sites and services are protesting the Stop Online Piracy Act (SOPA) and the Protect IP Act (PIPA) currently in play in the U.S. congress.

Floating Sheep is joining this protest because the bills (as currently written) give extraordinary and wide sweeping power to major media companies as would be bad for the exchange of ideas and content that is the Internet.

Wikipedia has a nice overview on the issues surrounding these bills and currently has a nice form that provides contact information for U.S. representatives and senators.

January 11, 2012

Announcing "Iron Sheep": Map and Hack Day, February 26th @ AAG

Announcing "Iron Sheep"

Pack up your laptop, grab your data and head to New York for the first annual "lightning mapping of user generated information" event. Dubbed Iron Sheep (at least until someone objects) the event seeks to mimic the format of the “Iron Chef” television series. This workshop challenges participants (grouped into teams with members from diverse backgrounds and skill sets) to produce meaningful analysis and fun, evocative mash-ups from the same sets of user-generated, geo-coded data within a four hour time frame. The goal is to provide a semi-structured environment where participants can socialize and work in a fun yet socially meaningful project. Participants will be drawn from academic, industry and artistic communities from around the world.

Teams will be assigned a targeted question (e.g., What is the most “out of shape” location in the U.S.? or How can we visualize the Occupy Wall Street movement? or Where is the most likely place for the zombie apocalypse to start? or Where is the origin and destination of Super PAC money?) and use crowdsourced data to create a new geo-visualization. Teams can also choose their own questions. The exact questions and datasets depends on the participants who join (see below).

When: Sunday February 26th 2012, 5 pm to 9 pm; Happy Hour(s) to follow
(During the Association of American Geographers Meeting in NYC)

A panel discussion of the event will take place at the AAG conference at 8 am on Tuesday (Feb 28th) in Conference Room E, Lower Level, Sheraton Hotel

Where: New York City, exact location to be announced (depending upon number of attendees)

Confirmed Attendees
* Will attend the panel on Tuesday but not the "lighting mapping" event on Sunday

How to attend
This event is open (register here) although the price of entry is that each participants contributes a dataset that is not commonly available (ideally crowdsourced) or one that has been enhanced in a meaningful and useful way. Data that is relatively comprehensive for the U.S. or world would work best, although something focused specifically on New York City would be appropriate as well. The goal is to have enough commonality between datasets that it is possible to do mashups. There will also likely be a "secret sauce" dataset (again borrowing from the Iron Chef idea) that all mappers would be challenged to include.

The organizers will likely be contributing the following data sets:
  • The retail cost of marijuana at the city, county and state level (based on PriceofWeed.com reports);
  • Busted meth labs;
  • Some slice of geo-coded tweets;
Serious or offbeat datasets are both extremely welcome as we are hoping for the serendipity that comes from the new mashing of data – such as the Beer Belly of America. For example, we'd love to end up with a combination of data sets that includes things like:
  • Super-PAC contributions;
  • Flickr photos tagged with the word "cat";
  • Precinct level voting records;
  • VW bug ownerships;
  • Sales of Twinkies and/or iPhones;
  • Foursquare check ins at bars versus grocery stores;
  • Spending on political ads.
  • Grindr.com and/or BarebackRT.com check-ins;
  • OK Cupid dating information

January 05, 2012

Good vs. Evil: One Final Holiday Map

With the various rounds of travel and vacation at end of December break we ran out of time to post the entries for the second round of Satan/Santa maps. This is an issue that we now want to make right and since we received only one addition entry it makes choosing a winner very simple! But this does NOT mean that the map or the technique used to create it falls short of our stringent Floating Sheep standards. Indeed the map below submitted by Tom Koehler is a great example of the willingness to apply stringent GIS techniques to somewhat messy data [1] about an offbeat subject.

After combining a range of terms -- Satan, Satan Santa, Zombie, Bad Santa, Evil Santa, Devil and Lucifer -- Tom "created a surface raster of the results. I suppose there is some argument to be made over creating a contiguous surface from absolute data, but I figure the search hits aren't exactly absolute since they're searching a radius around that point." We suspect that the resulting map would look similar whether Tom used summed or absolute data but greatly appreciate the fact that he raised it as a possible issue/critique.

The resulting map of "more bad" clearly is shaped by major population centers; after all, population density is a key (albeit not the only) driver of geoweb material. Still some cities known for their "badness" such as Las Vegas, New Orleans and Miami come off relatively less bad than others, especially the Northeast. I guess Santa is looking forward to catching some time in the sun and warmth as well.

In any case, thanks to Tom for his map and he will be receiving his prizes in the mail soon.

[1] Is there any other kind of data?

January 03, 2012

Augmented realities and uneven geographies

In the "better late than never category" we offer the presentation that Mark and I gave last September at the iCS-OII symposium. The paper version is available as well is you email me.