The Blog

Polls and Tequila: Better with a Few Grains of Salt

12:32 PM, Jul 29, 2008 • By GARY ANDRES
Widget tooltip
Single Page Print Larger Text Smaller Text Alerts

Confused about the array of contradictory polls released in the last few days? You're not alone. First there was the 8-point Obama advantage in Gallup tracking.

But wait: Then there was the 4-point McCain lead in another Gallup poll released the same day.

What gives?

Mark Blumenthal does a nice job of dissecting the "Berlin Bounce"--or lack thereof at Pollster.com.

This dizzying array of numbers requires a little perspective. First, keep in mind not everyone who answers a poll question holds opinions with the same degree of certainty. Moreover, some people may not have thought a lot about the race yet at all. This leads to survey volatility that's seen in these polls.

Second, there are significant differences among polls measuring the opinions of "adults," "registered voters" and "likely voters." Polls of those three groups will normally yield different results. Frank Newport of Gallup sums up how his organization sorts out the differences here.

Newport also notes that until recently, a higher percentage of "likely voters" in Gallup's 2008 model were Democrats--probably because of the heightened interest in the longer-than-usual primary campaign between Senators Obama and Clinton. But historically, a higher percentage of Republicans fall into the Gallup likely voter category. Newport writes this:

In earlier 2008 polls, more Democrats than Republicans were engaged in the campaign and considered likely voters. This is generally a rare occurrence given that Republicans have historically been more likely to qualify as likely voters under Gallup's model (a fact that has been borne out in the real world as Republicans are able to win elections despite facing deficits in party identification or pre-election standing among all national adults).

Therefore, McCain now doing better among "likely voters" compared to registered voters is in some respects a return to the historical norm.