The Media Obsession with and Abuse of Polls

December 15, 2011 04:47


Polls have something to offer for everyone’s liking. Statistics can be massaged to make the point that the analyst wants. But there are different kinds of polls.

By Roger Aronoff

Before a single vote has been cast for the Republican presidential nomination, “the lead” has changed hands five or six times. At one point or another, Rick Perry led, Herman Cain led, Michele Bachmann led, Mitt Romney led, and the current leader is Newt Gingrich. But the obsession with all of these polls serves to take the place of real journalism and analysis, and turns this very serious process into another reality show. We should be very skeptical of most polls, and observant of how the media try to use them to shape our opinions, rather than reflect them. In addition, the liberal media are eager to have a role in picking the candidate who they look most forward to pummeling in a general election against President Obama.

For example, a look at what the polls said in the last three open nominating campaigns makes a strong case for the media to spend far less time trumpeting and analyzing them. Exactly four years ago this week, based on polls taken from December 14-16, 2007, just weeks before the Iowa Caucus and New Hampshire primary, here is what the polls were telling us:

First, for the Democrats, according to a Gallup poll, Hillary Clinton led Barack Obama by a margin of 45% to 27% in a national poll of Democrats and Independents who leaned Democratic. This is approximately the same lead that Newt Gingrich has today, according to the latest NBC/Wall Street Journal poll. Gingrich leads 40% to 23%, actually one point less than what Hillary led Obama by at this point. The latest Gallup poll shows a 37% to 22% lead for Gingrich.

One thing most of these polls have in common is that they interview more Democrats than Republicans, and claim to adjust by “weighting” to account for the difference. But for example, in the above mentioned NBC/WSJ poll, which shows both Romney and Gingrich in a statistical tie (within the margin of error) with Obama, of the 100 people polled, 42 voted for Obama last time, while only 32 voted for McCain. So right there, the results become skewed. The weighting process generally doesn’t really adjust sufficiently, though the pollsters would argue otherwise, especially when you consider that the largest number of those polled identify themselves as Independents.

In the Gallup poll from the same days in 2007, December 14-16, polling just the Republican and Republican-leaning voters, Rudy Giuliani led Mike Huckabee by 27%-16%, with John McCain in 3rd at 14%. Of course, the nominees ended up being Barack Obama and John McCain.

What about the previous open election, in 2004? George W. Bush was running for re-election, but there was an open process for a Democratic nominee. In a Gallup poll taken January 9-11 of 2004, before Iowa or New Hampshire, Howard Dean led with 25%, followed by former General Wesley Clark at 19%, and eventual nominee John Kerry had 9%. So clearly, in the last three examples in presidential primaries, polls taken shortly before the first votes were cast proved to be unreliable predictors of who would eventually get the nomination.

A USA Today Gallup Poll shows Gingrich and Romney winning 12 swing states over Obama.

Polls have something to offer for everyone’s liking. Statistics can be massaged to make the point that the analyst wants. But there are different kinds of polls. As I’ve shown, these presidential primary polls that are more popularity contests within political parties are not good predictors of what the voters are going to do. But there are polls, such as approval polls, that are measuring ideological commitments, in which the same questions are asked month after month, year after year, and trust is built up or eroded over longer periods of time. And even then there are often contradictions.

Within the same new CBS poll, we have findings that 75% say the country is on the wrong track while only 21% say it’s on the right track. But 57% believe the President is a strong leader. And once again, the numbers are stacked to favor the Democrats, as 32% of those polled were Democrats while 27% were Republicans. Obama’s job approval rating was 44%, with 46% disapproval, which as Bob Schieffer pointed out on the CBS Evening News the day they released this poll on December 9, “you have to go all the way back to Jimmy Carter to find a president whose approval rating was this low at this point in his presidency.” Scott Pelley, the news anchor, said “the weak economy is largely to blame.”

When Obama was asked by Steve Croft for the following Sunday’s “60 Minutes” about the economy, he replied, “we did all the right things to prevent a great depression and to get the economy growing again and to get job creation going again. But it hasn’t made up for the hole that was created in those six, nine, 12 months before my economic policies took effect.”

But the poll showed that just 33% of Americans approve of the way President Obama is handling the economy, while 60% disapprove. Only 28% believe that Obama has made “real progress fixing the economy” and only 37% felt he had prevented a deeper recession, while 49% felt that he hadn’t. 51% disapprove of “the President’s signature accomplishment in office, health care reform, while only 35% approve. Just 13% think it will help them—and 32% think it will hurt them.”

As to deserving to be re-elected president, 41% said yes, and 54% said no, while only 32% said they have a clear idea of what Obama wants to accomplish if re-elected, while 66% said they don’t.

But Obama regularly takes solace in the fact that his ratings are much higher than those of Congress, which he and his spokesmen, both in the White House and in the media, suggest is because it is a Republican controlled Congress. Republicans do control the House, but not the Senate, which as I pointed out in a recent AIM Report, has been the half of Congress that has really been derelict in its duties, mainly by not having passed a budget since April of 2009, but also for allowing more than a dozen jobs bills passed by the House to languish in the Senate.

Besides, Congress’s approval ratings one year ago, in December 2010, after being run by Democrats for four years, was at 13% with 83% disapproval, so that hasn’t changed much. But then there was no gridlock. With a year of near-gridlock behind them, the latest numbers are 11% approval and 82% disapproval. Part of that is the notion that people like their own Representative, but not Congress as a whole, which moves too slowly, and deliberatively, for most everyone’s liking.

The current polls may well prove to be a dead-on predictor of this year’s Republican nominating process, and Newt Gingrich might wrap it up by March. But if recent history is a guide, expect the unexpected in the coming months. New rules in place by the Republicans promise to make this race different than past races. The GOP has proportional representation primaries and caucuses until April 1st, after which each state can decide if it wants to offer winner-take-all or proportional. Proportional means that someone like Michele Bachmann or John Huntsman could have 50 or 100 delegates needed for the nomination, and thus they may decide to not drop out, but rather to take their delegates to the nominating convention in Tampa, Florida in August, to possibly have some bargaining power. If no one has enough delegates to win the nomination in the first round, then delegates are free, meaning they could switch candidates, or the convention could draft someone not in the mix, like Chris Christie, Jeb Bush or Paul Ryan. Time to buckle up and see where this ride takes us.

Roger Aronoff is the Editor of Accuracy in Media, and can be contacted at roger.aronoff@aim.org.



Help Make A Difference By Sharing These Articles On Facebook, Twitter And Elsewhere: