Network News

X My Profile
View More Activity

The National Polls Don't Matter, Right?

With three early state returns under our belts, one thing is clear -- national opinion surveys don't mean squat.

Since early 2007, polls showed Hillary Rodham Clinton and Rudy Giuliani with comfortable leads, making them the respective frontrunners for the Democratic and Republican nominations.

And yet, Clinton and Giuliani have combined to win just one primary so far, defeated by candidates who often ignored national polls to make successful, state-specific appeals.

So surveys of the entire country are essentially useless at this stage in the 2008 nomination fight because of the state-by-state nature of the process.

Case closed, right?

Not so fast. Although the focus at the moment is on a handful of states with primaries and caucuses scheduled for the near term (South Carolina, Nevada and Florida), the real fight for candidates from both parties is on Feb. 5 when 22 states -- including delegate treasure troves like California, New York and Georgia -- will cast ballots.

These 22 states are scattered across the country, amounting to a de facto national primary, the first such national primary election in recent political history.

The sheer number of states holding primaries or caucuses on Feb. 5 means that no candidate will be able to run the sort of fully financed television and grassroots campaign that we've grown accustomed to in places like Iowa, New Hampshire and South Carolina. In those earliest of early states, candidates and their campaigns had months of time and millions of dollars to introduce themselves to voters and educate the electorate about their positions before any actual votes were cast.

Fast forward to Feb. 5. The top-tier candidates aren't even likely to visit every state voting on that day, and the cost of television ads in places like California, New York, New Jersey, Georgia, Arizona and Florida runs into the millions of dollars per week. Even Clinton and Barack Obama -- the two best financed candidates in the race -- don't have the resources to be on television in all the key Feb. 5 states at once.

That means that every campaign will have to try to tailor a message that can appeal to a broad swath of the voters set to cast ballots on Super Tuesday.

It's in assessing the broad political environment as well as voters' thoughts about the individual candidates where national polling shines, and it's the reason why these sorts of surveys will become increasingly relevant in the lead-up to Feb. 5.

"Because Super Tuesday involves so many states, the national polls begin to potentially reflect what is going on in that Feb. 5 primary," explained Neil Newhouse, a partner in the Republican polling firm Public Opinion Strategies.

Jon Lerner, another Republican pollster, said that while a number of states have "native sons" (Arizona, Arkansas, Tennessee, Utah, New York, New Jersey and Connecticut) there are also a large group that don't, and in that group national polling will matter.

"The national polls likely will be highly relevant regarding the general direction of those non-native-son states," Lerner said, quickly adding, however, that until Florida's primary on Jan. 29 the national stage won't be set and therefore national polling will be a less useful indicator of broad trends in the electorate.

Does that mean that whichever candidate enters Feb. 5 with the lead in national polls is the most likely to carry the day? No.

There is a big difference between using national polls to understand the political environment (the correct use of these sorts of surveys) and using them to predict the ultimate outcome. National polls will be increasingly vital in understanding the terrain on which the fights for the Democratic and Republican nominations will be fought, but the order of the horse race in any single national poll should not be taken as a predictor of the results in any single state.

"The national polls are telling us things but they are not designed to be predictive of individual states," said Jill Normington, a Democrat pollster with the firm Bennett, Petts & Normington. "National polls of 1,000 people will have about 20 respondents in most states -- far too few to make them worth analyzing at that level. "

John Anzalone, another unaffiliated Democratic pollster, was even more dismissive. "I never thought national primary polls were relevant," he said. "It is hard enough to do good polling in a single state primary universe and it becomes ridiculous to think that the polling we see accurately reflects the weighted national primary universe.

With the nomination fights for each party still very much up in the air and the Feb. 5 mega primary looming, scads of national surveys are sure to be released in the coming weeks that, more so than prior to Iowa and New Hampshire, can provide real clues about the issue environment and the resonance of candidates' messages. But don't try and extrapolate any head-to-head numbers as indicative of the likely order of finish in any one state. Remember: No matter how many states vote on a single day, this is still a state-by-state fight for the nomination.

By Chris Cillizza  |  January 17, 2008; 6:00 AM ET
Categories:  Eye on 2008  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Obama Aide: Clinton Is 'Prohibitive Favorite' in Nevada
Next: In Florida, Giuliani Hits Romney in New Ad

No comments have been posted to this entry.

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company