Twisting Twitter: The False 4.7% African-American Turnout for the 2010 Elections Stat

So, last night while I was cringing as the results of last night’s mid-term elections rolled in, I was also reading what folks were saying on Twitter.

I don’t quite remember what time I saw the following stat floating around, but I’d say it was about 11pm (Arizona time): “Only 4.7% of African-Americans voted in the election.” Sure enough, checking Twitter’s trending topics, “African-Americans” (or some variation) was near the top.

Many of the tweets repeating the stat would also go on to admonish “… but 100% of y’all gonna be in the clubs this weekend” and claim that, in contrast, 94.7% of African-Americans had checked out WSHH.

All of those ridiculous claims aside, the stat seemed funky to me. If there were still some ballots being counted across the state and the nation, how could anyone solidly assert that only 4.7% of African-Americans had shown up at the polls. Also, none of the tweets contained any links to a credible source for that stat.

So, I did a little searching this morning (using Google’s excellent Updates Search, which pulls from Twitter a lot more helpfully than Twitter’s own search currently can) and found this timeline of the 4.7% stat blowing up:

The original 4.7% of African-Americans blows up.

Here’s a link to the Google search that yielded the image above: http://j.mp/4point7

Here’s the first tweet, from Andre X. Douglas (@DRE_XD) of Wisconsin pointing out that 4.7% African-American voter turnout stat at about 8pm (10pm in Wisconsin) last night:

And here are Douglas’s follow up tweets, each within an hour of the initial 4.7% stat tweet:

So, the stat was originally quoted from someone’s status update (on Facebook, presumably) that was quoting early reporting from USA Today (a quick Google search couldn’t turn it up) and, if true, only applied to the turnout of African-American voters in a single district, Wisconsin’s District 1. Through all of the retweets and editing from others, it was extracted that only 4.7% of African-Americans voters nationwide.

I don’t think anyone can really blame Andre X. Douglas. He was expressing his own dismay at the stat and clarified it within his own Twitter timeline.

It’s just a good reminder of what an insane game of telephone the internet (not just Twitter or social media) can be.

—

UPDATE:

I decided to check out the still hot, if no longer trending, topic of “4.7% of African-Americans” on Twitter and stumbled upon a good read by Anna John (@DCntrc) of WAMU, American University radio in Washington D.C.:

http://dcentric.wamu.org/2010/11/why-was-african-americans-a-trending-topic-in-d-c-on-twitter/

John quotes the Washington Post as asserting that 10% of voter turnout was made up of African-Americans.

Saheli Datta (@sahelidatta) wisely notes in the comments that African-Americans making up 10% of total voter turnout last night across the nation actually translates to about 32-33% of registered African-American voters and around 25% of all African-Americans.

Datta:

Making up 10% of the voting electorate can’t meant that only 10% of registered African Americans voted. ~ 90 odd million people, so 10% of that is approximately 9 million voting African Americans. Voter participation in this midterm election was supposed to be 42%, so if 90 million = 42%, the total eligible voter pool was 214 million, and 13% of that means there are probably 27 – 28 million registered African Americans. 9 million out of 27 million is much closer to 32- 33% participation rate, not a measly 4.7%. That’s pretty comparable to the general rate of 42%, and I wouldn’t be surprised if the discrepancy is due to being disproportionately burdened by the factors that depress voting for the less than wealthy. It’s a working day, not a holiday, there’s no childcare, and polls are often hard to get to–all things we should fix. ( I got my statistics on the voting numbers from this AP article by Matthew Daly.).

According to Wikipedia In 2000 there were 36 million African Americans in the united states, so assuming growth was steady, there are about 40 million *now*. Many of those people are under the age of 18, so they can’t vote, so that means over all participation was AT LEAST 25%. That strikes me as pretty respectable–especially when you consider wide spread evidence that African Americans have been disproportionately and unfairly disenfranchised b/c of the discrepancy in felony status laws regarding the use of crack vs. the use of cocaine, not to mention other less clearcut kinds of unfairness in the criminal justice system.

In the above analysis of the over all participation rate (i.e. voters/eligible voters, not voters/*registered* voters) I’m assuming all African-Americans the census counts as such are citizens, which might be a more reasonable assumption than for any other ethnic group except I don’t think the census differentiates between century+ African-Americans and more recent immigrants. So the over all participation rate in the ‘indigenous’ African-American community might be much higher than 25%.