|
|
|
|
|
Learn From The Experts Articles
What Pollsters Faced in 2008 - A Perfect Storm
By Brad Bannon
Public opinion in 2008 was just about as stable as Britney Spears and a lot less predictable. The volatility of campaigns made the lives of pundits, prognosticators, politicians and pollsters very difficult.
The first problem they faced is that some pollsters have not adapted very well to the challenges of polling in an extraordinary set of primaries which are without precedent in modern American political history.
The events in this presidential primary cycle have created the perfect storm for polling in presidential primaries. Polling in any primary is difficult and the historic nature of this presidential primary campaign has made polling even harder than it usually is. Polling in a caucus state like Iowa is next to impossible.
On the Republican side, this is the first election in a long time when there had not been an obvious high profile heir apparent. For the Democrats, this is the first time that a woman and an African American have had a realistic chance of winning their party’s nomination.
This extraordinary set of circumstances has largely increased turnout in the Democratic presidential primaries. Women, blacks, independents, and young people have been voting in record numbers in this year’s Democratic primaries. The problem is that many pollsters base their turnout models on voting patterns in the 2000 and 2004 presidential primary elections and this year’s primaries are nothing like the primaries of the past.
The late polls overestimated Obama’s vote in New Hampshire but underestimated it in South Carolina. Women turned out in record numbers when Hillary Clinton won in New Hampshire and black voters turned out in record numbers when Barack Obama dominated South Carolina.
Pundits also deserve some of the blame for this mess.
In his book The Right Stuff about the early astronauts, writer Tom Wolfe described the bodies of crashed test pilots as burned beyond recognition. This year the same pundits have burned polls beyond recognition. They have twisted the data from polls into grotesque figures which are unrecognizable to the people who actually conduct the surveys.
In 2007, when Hillary Clinton was getting the support of half of the Democrats in national polls, the analysts pointed out that she had a 25 point lead over her opponents. What the analysts failed to note was that half of the Democrats were not voting for the Democratic candidate who had universal visibility in a race against opponents who had little or no name recognition.
I make my living as a pollster, but I also appear on television and radio shows as a pundit. What the media wants more than anything is predictions. They want tomorrow’s news today. As a pundit I find myself often tempted to say things that I would never say as a pollster.
I may be cynical but it strikes me that many pundits aren’t any more sophisticated using polls than they were 25 years ago, when I broke into the survey research business. I don’t know how many times, my mentors told me that polls were not moving pictures but snap shots. I don’t know how many times, my professors said that polls explained events but did not predict them.
It is not unusual for a third of the voters to be still undecided the weekend before a Tuesday primary. But the undecided voters at the tail end of primaries have been especially high this year. Since there is not a dominant well-known frontrunner on the Republican side, GOP voters waited to the very end so they could learn about and evaluate the candidates.
There were also large numbers of late undecided voters on the Democratic side. Many young white female voters want to vote for Barack Obama, but feel they should vote for Hillary Clinton. Some African American female voters want to vote for Clinton but feel they should vote for Obama. Because of these cross pressures, a lot of voters waited until the very end to make a final decision.
Despite the large number of late deciders, pollsters stop polling a couple of days before the primary so their media clients have the numbers before primary day. That gives the undecided voters plenty of time to make their decisions without pollsters recording their opinions.
That was probably the problem in New Hampshire. Everybody and their brother and sister were predicting an Obama victory on the heels of his big win in Iowa. But two nights before the primary, Senator Clinton had her now famous cry which may have convinced female voters at the last minute that the big boys were beating on their soul sister.
Also New Hampshire voters like many other voters are notoriously contrary and decided to vote for Senator Clinton after listening to the pundits predict that they were all set to vote for Senator Obama.
Early on, the pundits also made the mistake about the inevitability of Hillary Clinton’s nomination. Did analysts really think that her early lead in the polls was anything more than the advantage in name recognition that she had over Barack Obama and her other rivals at the beginning of the race? Did anyone really think she was going to win the African American vote once Barack Obama gained credibility and name recognition after the early primaries and caucuses?
The most meaningless and misleading numbers in any primary polls are the trial heats. Unfortunately the “head to heads” are the only things that the media reports and the only part of the surveys that people actually look at.
The trial heats at this stage of the 2008 campaign are just a reflection of the name recognition of the candidates. Voters have not focused on the race yet and when they do, the “head to heads” will change radically as primary voters know more about some of the candidates and come to know the other candidates for the first time.
A look at past presidential primary polls months before the nomination process even started, gives you a sense of the volatility in the trial heats. In August 2003, a New York Times survey showed Joe Lieberman leading the pack and John Kerry trailing in the dust.
In October 1991, Jerry Brown was the frontrunner with twice the support of some guy named Clinton. In January 1987, the only candidates in double digits were Gary Hart and Mario Cuomo while Michael Dukakis barely registered.
There are a number of problems with surveys this year. If pollsters had updated their turnout models and if pundits have been more cautious, we would all have been better off.
Brad Bannon is president of Bannon Communications Research which
polls for Democrats, labor unions and issue groups.
He also appeared as an analyst on many television
and radio shows. Brad can be reached at Click here to contact this Author
|
| |
Learn From The Experts Articles
|
| |
Special Interest For The Real Politico |
|
| |
|
High Stakes Politics ... Not Unlike High Stakes Poker |
| |
|
By David Rabinowe
Turn on the television almost any hour of the day and there are three things you are guaranteed to be able to watch, no matter what time it is ... |
|
| |
|
What Pollsters Faced in 2008 - A Perfect Storm |
| |
|
By Brad Bannon
Public opinion in 2008 was as stable as Britney Spears and a lot less predictable. The volatility of campaigns made the lives of politicians and pollsters more than difficult ... |
|
| |
|
Two Points of View |
| |
|
Anonymous Authors
Two humorous stories from opposing view points regarding our two major political ideologies, Republican and Democrat ... |
|
| |
|
Healing Post-Partisan Depression |
| |
|
By Donna Zajonc
Like hopeful new parents anticipating the birth of their first child, the nation anxiously awaited to see the face of the future ... |
|
| |
|
Politics Like Baseball But More Expensive |
| |
|
By David I. Rabinowe
The 2004 campaign season is over -- time for the “political hot stove” to heat up – and much like baseball - the biggest news involves $$$ ... |
|
| |
|
How a Coalition of Police Officers and Citizens Took on the Mayor of Minneapolis and Won |
| |
|
By Michael R. Shannon
At that time, in early November 2004, Mayor R. T. Rybak was working hard to pass a budget plan that cut eight police officers from ... |
|
| |
|
And The Winner IS - The Presidency 2008 |
| |
|
By Anonymous Author
Just for fun, let's take a look back at the odds makers assessment of the Presidential race in 2007 / 2008 and where they placed Obama's chances ... |
|
| |
|
Looking Back - It’s a Horserace …But Nobody Knows Who’s Running |
| |
|
By Tom Edmonds
Written After the 2006 Mid-Term Elections
With a presidential election on the distant horizon, everybody seems to want to predict what's going to happen ... |
|
| |
|
The Most Expensive Job Search Ever |
| |
|
By David Rabinowe, Esq.
Imagine for a moment, if someone told you there was an opening at your dream job -- the one you wanted all your life. It has everything ...
|
|
| |
|
The Lighter Side of Politics |
| |
|
Anonymous Humour
A cowboy named Bud was overseeing his herd in a remote mountainous pasture in California when suddenly a brand-new BMW ... |
|
| |
|
Beyond The Beltway - Issues In State And Local Politics |
| |
|
By Brad Bannon
For a number of reasons, state and local governments will be on their own trying to solve some of the most pressing problems facing the nation ... |
|
| |
|
All Politics Still Local Despite What Polls Say |
| |
|
By Michael Cohen
The most recent obsession in election analysis is to pre-certify that 2006 will be the first off-year national election since 1994 ... |
|
|
|
|
 |
|
 |
