Don’t Trust Polls

Why do we still using polling for debates anyway?

You’re probably already well aware that Rand Paul will not be in the included in the main stage debate tonight on Fox Business network. Whether you’re a libertarian greeting the news with disappointment or schadenfreude, a discussion on whether polls are appropriate measurements of debate inclusion.

Fox excluding Rand Paul from the main stage debate is big deal. Not only does it cost the campaign millions of dollars in exposure, being relegated to the the under card debate this late in primary season indicates to undecided voters that Rand’s campaign is on the way out or a second tier campaign. Understandably, Rand and his team are lashing out. The campaign can’t risk Horse Race effect undermining the campaign.

The Horse Race Effect often works like this: If a poll comes out showing high support for a candidate, voters and the media may experience a bandwagon effect that reinforces the candidate’s support. The opposite is also true. It’s the way the media tends to frame campaigns and voters respond to the increased coverage.

The Republican Primary
The Republican Primary as shown by the media

So the Rand Paul campaign was excluded and it hurts the campaign. “So what?” says the critic “Fox Business wants to include candidates that have a chance of winning and the Rand didn’t meet the criteria, end of story.”

Well sort of. To be included in the main stage debate, a candidate had to be in the top five in recent national polls or in the top six in one of the first two primary states (Iowa and New Hampshire). By Monday, Rand was within a fraction of a percentage point of some other candidates included on the stage. With many of these polls having a margin of error of +/- 5%, these polls may not indicate any meaningful difference between Rand Paul’s support and support for candidates like Jeb Bush, John Kaisich, and Chris Christie. Essentially, they’re all tied.

RCP average of polls: National, Iowa, New Hampshire

Furthermore, the strange release time of a Bloomberg Poll even has Politico raising questions:

Quite simply: Had the Register and Bloomberg released their poll, which was completed Sunday, on Monday at 5:59 p.m. instead of Wednesday at 6 a.m., Paul would likely have qualified for the main stage.

Finally, the entire premise of Fox’s debate criteria only makes sense if polling actually reflects support and likely voting outcomes.  Lately, there’s a lot evidence to suggest that polling no longer has the predictive power it once had. Likely voters are getting harder to reach and it’s resulting in more sampling errors and misweighted demographic groups. Polling powerhouse Gallup got out of the horse race business recently. Polling guru Nate Silver admits that some astoundingly bad election predictions over the last couple years exposes some deep flaws in polling.  For more on why we should be skeptical that polls, check out this article from Reason on polling. It’s a bit long, but it covers way more ground on the polling problem.

My point it, maybe it’s time to stop using polls as the sole criteria for debate inclusion.

I don’t want to overstate my case. Although Rand Paul’s campaign is putting on a brave face, they can’t dismiss the huge polling gap between him and candidates like Trump and Rubio.


Paul still could have participated in the undercard debate tonight with Carly Fiorina, Mike Huckabee, and Rick Santorum but chose not to. I was skeptical that Rand skipping the undercard debate was the right move, but it may be working out for him. A full week’s worth of media attention may have given him more viewership than the debate. Good for Rand Paul. Maybe the wave of positive coverage and a good ground game are just the boost the campaign needs to pull this thing off. But they’re not out of the woods yet.

So why should “big L” Libertarian like me care about this? Because we’re no strangers to the self-reinforcing impact of the horse race effect and the legitimizing power of the media. When Rand literally gave the the finger to the media today he said “They [the media] want to decide the ‘tiers’ of this race and name the winners and losers. I will not stand for this.” Libertarian Party candidates are excluded from most polls and face a constant media blackout. So yeah, we get it.

Gary Johnson’s Facebook Page put it this way:

Rand Paul finds himself arguing with a debate sponsor over polling percentages that are, in reality, within the margins of error. I can sympathize. Thanks to a percentage here or there, a legitimate candidate, who would offer a more libertarian view, is being kept off the stage. That’s unfortunate, but not surprising.

Don’t expect Rand Paul or any of his supporters to be quite as sympathetic to this problem when it’s time to include the Libertarian Party candidate in the general election debates.  The Libertarian Party is gearing up for our own much bigger debate battle this fall. Check out the Fair Debates Facebook page for more on that.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s