Is Social Media a good Predictor of Elections, not necessarily

There was an interesting post today by US Politics on Facebook (see below) about how Facebook fan counts were good predictors of 4 out of 5 Senate races and 3 out of 4 House races.  Intuitively this makes sense, but I wondered how it held up in some of the really close and heated races.  Here’s what I found…

In the California Governors race Meg Whitman crushed Jerry Brown on Facebook and lost big in the election.  Ditto for Sharon Angle, she had 108,740 fans to Harry Reid’s paltry 15,462 and she lost.  Even stranger, Lisa Murkowski, who lost in the primary, was a write in candidate (which is traditionally a guaranteed loss), appears to have won and she had half the number of Facebook fans.  Go figure on that one.  Facebook fan figures would have indicated landslides in all of these races.

There were other hot races where Facebook was a good predictor such as Scott v Sink in Florida, Rubio v Crist v Meek in Florida, and Boxer v Fiorina in California.

Twitter and Youtube views had similar patterns.  In “normal” races the followers and views seem to be a good predictor of the results.  But if the race has unusual candidates or if it is very close, toss that formula out the window.

One reason I suspect the stats may fall apart is that fans are fans and many will stay loyal no matter what happens.  In many of these races some of the candidates exhibit outright bizarre behavior.  I suspect many stayed loyal and others may not have bothered to “un-follow” on Twitter or “”un-like” on Facebook which would explain the numbers.  Once someone “likes” a page it is a pain to go back and “un-like” it because it is in tiny print at the bottom of the Page and you have to find it first – part of the Facebook Zuckering strategy, change is hard.

So here’s my take on this.  In a “normal” race (if there is such a thing in today’s politics) Facebook and social media trends are a good predictor.  The big take away here is that a candidate can screw up a great social media campaign with missteps in the traditional media.   Also a great social media campaign can be beat by a highly effective old school grass roots campaign.

Facebook post by US Politics on Facebook 74% of House candidates with the most Facebook fans won their race and 81% of Senate Candidates with most Facebook fans won their race.

Here’s some interesting data on some of the hot races.

It is interesting how it is all over the map.  Also note how some were behind on Facebook and way ahead on YouTube or Twitter.

W/L Candidate Facebook Fans Twitter Followers YouTube Views

California Governor’s Race

W Jerry Brown 98,984 1,100,473 1.2 million
L Meg Whitman 207,823 242,404 42,000

California Senate Race

W Barbara Boxer 41,767 310,032 249,043
L Carly Fiorina 22,845 23,917 1.1 million

Nevada Senate Race

W Harry Reid 15,462 6,314 795,642
L Sharon Angle 108,740 8,531 1.4 million

Florida Senate Race

W Marco Rubio 135,249 18,539 1.2 million
L Charlie Crist 29,640 7,467 145,558
L Kendrick Meek 24,141 6,055 100,153

Florida Governor Race

W Rick Scott 55,535 3,303 204,641
L Alex Sink 30,033 5,162 197,455

Alaska Senate Race

W? Lisa Murkowski 3,606 4,559 53,375
L? Joe Miller 8,214 5,618 251,174

Still not officially called as of this post

Tags: ,

1 Response to "Is Social Media a good Predictor of Elections, not necessarily"

  • kazi says:
Leave a Comment