Parsing the Polls: Why It's So Hard to Poll Iowa
The Iowa caucuses will again lead off the presidential selection process and -- especially on the Democratic side -- the candidates are treating it as the whole shebang, knowing that a win or a loss there could serve to kick-start or destroy their campaigns.
At the moment Sen. Hillary Rodham Clinton (N.Y.) holds a narrow lead on the Democratic side while former Gov. Mitt Romney -- fueled by millions in ad spending in the state -- has a consistent double-digit lead over his next contender.
Romney went up with ads months ago in Iowa and has worked the state harder than any other Republican candidate. The money spent as well as the grassroots campaigning have paid off as Romney has without question the strongest organization in the state. That organization was on display in August when Romney won the Iowa Straw Poll -- a key early test of support in Iowa. Clinton, for her part, started slow in Iowa and has managed to pick up steam thanks in large part to the organization know-how of her Iowa state director Teresa Vilmain.
These results have been born out in 11 polls -- six on the Republican side, five on the Democratic side --that were conducted in the last month alone. The polls were produced by American Research Group, University of Iowa, Strategic Vision, Rasmussen Reports, Insider Advantage and Selzer & Co. .
Here's a look at the results:
|U of I||10/17-24||13||13||6||36||11|
|U of I||10/17-24||5||29||20||27||7|
For all of the consistency that these polls show, questions linger about just how much stock to put in them. Pollsters and strategists frequently stress how difficult it is to poll Iowa (usually right after a poll is released that shows their guy or gal struggling). But, is it so hard? And, if so, why?
"To pollsters, nailing likely primary voters on political surveys can be challenging, but accurately identifying likely Iowa caucus attendees and which candidate they support (not to mention their second choice) takes the degree of difficulty of our business to a whole new level," said Neil Newhouse, a partner with the Republican polling firm Public Opinion Strategies.
Figuring out who is going to vote is always the most basic challenge for any pollster. Past results provide a guide but can never be taken as foolproof as turnout dynamics change from election to election.
This is especially true in Iowa's caucuses where an extremely small number of registered voters turn out to participate, voters can register the day of the caucus and turnout patterns fluctuate widely from caucus to caucus.
The challenge that anyone polling Iowa must face then is how to select an accurate sample of voters. Do you use the list of registered voters as your baseline? Or do you use the far narrower caucus list, which lists those that have participated in the most recent caucuses, to create your sample?
Each has its own problems.
As of Nov. 1, there were 1,912,197 registered voters in Iowa -- 600, 572 Democrats, 574,571 Republicans and 737,069 who affiliate with neither party. But, in the 2004 Iowa Democratic caucuses just 124,331 people participated.
So, if as a pollster you choose to use the registered voter list, it's uniquely possible that you are gathering opinions from people who won't even attend the caucuses. Such a method can "over exaggerate and overestimate the potential for turnout," said John Lapp, who ran Iowa for former Rep. Dick Gephardt (D-Mo.) in 2004.
The caucus list approach then would seem to make more sense. After all, given the low turnout patterns for caucuses the best predictor of who will come in 2008 is who came in 2004, 2000 and so on.
Not so fast. Several Democratic campaigns in 2004 adopted just that strategy -- planning for a voter universe of between 60,000 and 70,000 caucus goers based off the 59,404 who participated in the caucuses in 2000. The turnout -- 124, 331 -- more than doubled with nearly 50 percent attending the first caucus of their life.
"In general elections in Iowa about three in four registered voters turn up at the polls," writes J. Ann Selzer, the director of the widely-respected Iowa poll for the Des Moines Register. "This makes polling much safer, as an additional 100,000 showing up unexpectedly will not affect the outcome all that much. But at a caucus, this would be tantamount to a tsunami." (Selzer uses the registered voter file to develop her sample.)
In 2004 there was regular talk that pollsters using simply past caucus lists were missing the vast number of new caucus-goers -- especially young people -- who were going to turn out for former Gov. Howard Dean (Vt.). That prediction wound up being wrong as Dean placed third in the caucuses.
But, the possibility of a major turnout spike for a single candidate is again being floated -- this time for Sen. Barack Obama (Ill.) whose campaign has spent considerable time in Iowa trying to organize young people. "He does well with young voters and first-time caucus attendees and independents," writes Selzer. "These are not the sort who show up most plentifully on caucus night."
The truth of the matter is that polling the Iowa caucuses -- even more so than polling in other states -- is at least parts arts and science. Deciding who will vote and who they will vote for is a predictive exercise that forces a pollster to look backwards (at past voting trends etc.) and forward (at the growth potential of the electorate) simultaneously.
Looking for a port in this storm of information? The Fix conducted an informal survey of pollsters in Iowa and those who do considerable Iowa polling, asking them which Iowa poll they gave the most credence. The majority cited Selzer's Iowa Poll as the most reliable over time; it was Selzer, and Selzer alone, who had the top four finishers in Iowa in the correct order in her final poll before the 2004 caucuses.
The comments to this entry are closed.