Mapping racist Tweets: where post-election hate came from

After Barack Obama won America's 2012 Presidential election this week, a spate of vile racist tweets hit Twitter. One data company tried to figure out where that hate was coming from.

After the 2012 U.S. Presidential election, vocal racists took to Twitter.

It looks like the top state for the most racist Tweets is Alabama, closely followed by Mississippi.

The top ten states for post-election Tacist tweets - according to Floating Sheep's [interactive] GeoCommons map,  - are:


LQ of Racist Tweets


Alabama    8.1  
Mississippi    7.4  
Georgia    3.6  
North Dakota    3.5  
Utah    3.5  
Louisiana    3.3  
Tennessee    3.1  
Missouri    3.0  
West Virginia    2.8  
Minnesota    2.7


A hype-generating post by Jezebel (warning: offensive language) initially brought the Tweets to light. Geodata company Floating Sheep decided to combine their skills with Twitter's geolocation data and easy-to-use functionality of GeoCommons to make an interactive map of the racist Tweets.


As we can see in the top ten of their sampling, racist Tweeting is sadly not just a Southern phenomenon. The map shows that people suck in Canada and Mexico, too.

Floating Sheep cautions, "Keep in mind we are measuring tweets rather than users and so one individual could be responsible for many tweets."

To extrapolate, the anti-Obama racist Tweet map doesn't simply show the number of racist tweets.

The map uses a location quotient, which according to Floating Sheep, "indicates each state's share of election hate speech tweet relative to its total number of tweets."

Seven greyed-out states had no racist tweets. Interestingly, most of those seven seem to be states that have lower Twitter use than the rest.

Six of the grey states (Alaska, Idaho, S. Dakota, Wyoming, Montana, and Hawaii) had a very low number of tweets overall - although one (Rhode Island) had a significant number of tweets.

Details on racist Tweet map methodology

The geodata nerds at Floating Sheep used their very interesting DOLLY (Data On Local Life and You) Project ("Bringing Local Geodata to the People") to collect all the geocoded tweets from the last week (beginning November 1) with racist terms that also reference the election.

A total of 395 hate tweets were aggregated to the state level and compared to the total number of geocoded tweets coming out of that state in the same time period.

Using the examples of tweets chronicled by Jezebel blog post we collected tweets that contained the text "monkey" or "n****r" AND also contain the text "Obama" OR "reelected" OR "won".

A quick, and very unsettling, examination of the search results revealed that this indeed was a good match for our target of election-related hate speech. We end up with a total of 395 of some of the nastiest tweets you might possibly imagine.  

And given that we're talking about the Internet, that is really saying something.

GeoCommons is a GIS - geographic information system - that makes mapping items with geotagged data very easy.

Floating Sheep is wonderfully open about their process, detailing:

To be precise, we took a 0.05% sample of all geocoded tweets in November 2012 aggregated to the state level. The formula for this location quotient is:

(# of Hate Tweets in State / # of Hate Tweets in USA) 


(# of ALL Tweets in State / # of ALL Tweets in USA)

Floating Sheep's work is a mere snapshot, but it is a sobering one. 

I just hope that their little bit of data visualization can spark a bit more awareness raising, discussion and hopefully, change.

In the meantime, Alabama - maybe you just need a hug?

Or perhaps something stronger.

 Via Flowingdata - read also: Geography of Hate.