[var rssXML]
The Times-Herald View Today's Print Edition

Matthew Towery

Blogs

Angela McRae

Tea with friends

Deberah Williams

Everyday Finesse

Lorrie Lynch

Who's News

USA Weekend Tween Tribune - News For Tweens
Click Here

Matthew Towery Columnist

Published Saturday, August 04, 2012

Pollsters may have fatal flaw

As one who polled the 2008 presidential race extensively, it dawned on me, looking at the various “swing state” surveys taken recently, that many pollsters might be making a significant error that results in President Barack Obama with a lead, when perhaps the lead in reality belongs to Mitt Romney.

When surveys are conducted, the people who are interviewed are randomly selected registered (in a good poll) likely voters who are willing to respond. But that just begins the process. Those responses must then be “weighted” or allocated based on some projected model of past or future voter turnout strength. In other words, those conducting the poll must adjust the numbers to reflect the percentage of men, women, Republicans, Democrats, Independents, Whites, African-Americans and other groups such as Hispanic-Latinos who will likely vote in the contest.

This is no easy business, and it is why polling is as much an art as it is a science.

But there are some ways to get to these determinations, such as polling a state and asking which party voters identify with separately and before the actual “ballot test” survey is taken. But enough of these geeky polling details.

I have in my gut a belief the polls we are seeing now under rate Romney’s strength.

For example, we saw a hugely enthusiastic young voter turnout for Obama in 2008. As a result, polls were weighted to give more voter strength to the category of 18- to 29-year-olds than in past years. In general, younger voters just do not turn out to vote.

This year, I am seeing evidence the enthusiastic and energetic vote of young Americans for Barack Obama has at the very least lost some of its mojo. That likely means they are not as excited about voting.

So, an adjustment in the “weighting” of a survey by a reasonable number of percentage points downward for that group could mean the difference in some of these critical “swing” states between Obama being up by one point and Romney being ahead by a percentage point or two.

Another group that likely needs some “tweaking” by pollsters this go-round are those who identify themselves as “independent” voters. In 2008, they helped decide the presidential race, as a significant percentage were sick of the Bush years and were desperate for any type of “change” they could believe in, so to speak.

This year, we have more voters than ever identifying themselves as independents. This is surely a result of their absolute frustration with a Congress that appears to be chasing its tail and producing nothing of value.

Independent voters seem to be leaning more toward Romney than they did toward John McCain in 2008. So, if one considers the fact this group of voters likely should be inched up in the process of producing the final poll, and the fact that Romney is doing better among this group, that too might move a survey that shows Obama with a one or two point lead in a given state and shift the more realistic results toward Romney leading by a few points.

As I said, polling is as much an art as it is a science. But my guess is that some of the states where we are seeing results with Obama leading are either showing too strong of a lead or perhaps have the wrong man in the lead.

Of course, polls don’t make campaigns. But they do sometimes create a sense of momentum or, as was the case with McCain, impending doom.

In the end, it is message, image and strategy that win the race -- and debates.

For Romney, the message is not clear, and the image is still fuzzy. But he may have a strategy that is going to work. While Obama has burned through his cash early on, Romney has been raising dollars and holding back on expenditures. Obama’s camp has spent a fortune trying to savage Romney early in the contest. But the Obama camp may run low on cash just as we enter the last and most important phase of the race.

If Romney can get his act together and outspend Obama in the end, Romney has a sporting chance of winning, no matter how the pollsters adjust their numbers.

................

(Matthew Towery heads the polling and political information firm InsiderAdvantage. His column is distributed by Morris News and Creators’ Syndicate.)

Comment On This Story

Times-Herald.com does not necessarily agree with the comments posted below. Responsibility of comments rests solely with the writer. Comments posted in ALL CAPS will be deleted.

Submission of a comment does not guarantee publication. Comments will be posted by a moderator after being scanned for abusive language, relevance, etc. See our Comments FAQ for more details.

Most Popular

  • Viewed
  • Emailed
  • Commented

© 2011 The Newnan Times-Herald Inc., Newnan, Georgia. Any unauthorized use, copying or mirroring is prohibited.