Related Topics

President Bush Job Approval

RCP Average
Send to a Friend | Print Article

No More '08 Polls!

By Jay Cost

Sometimes, I think this new 24-hour news cycle creates more problems than it solves. Case in point: last week’s FOX News poll on the 2008 presidential elections. Not content to inquire about any of the 504 elections we will have in 9 months, FOX preferred to ask about the one in 33 months. They even saw fit to ask what people thought about Oprah or Arnold as president. I was prepared to let this inane poll slide – to silently wonder how FOX could waste so much airtime pondering the electoral chances of a man constitutionally barred from the job, but could no longer find any room for Arrested Development. However, Gallup got in the act yesterday, producing its own 2008 poll for CNN and USA Today. With the gold standard in public opinion on this bandwagon, this much has become clear: we have to nip this nonsense in the bud.

So, here it is. No poll about the next presidential election is worth your attention until Summer, 2007 at the earliest. The reason for this is that their predictive power is literally nil. They might help Martha MacCallum or Daryn Kagan fill the seemingly endless hours of live “news” coverage, but if your interest is in what will occur in 2008, they are of no use to you. Political junkies, take heed. If Shepherd Smith teases a 2008 poll “you cannot miss”, feel free to change the channel. You can miss it.

There are two reasons for this. The first has to do with the nature of the American electorate. Their opinions tend to change as other factors change. In the case of electoral politics, as their attention, interest and knowledge increase, their candidate preferences change. For 2008, all of these causal factors are today at their low ebb. The average voter is paying no attention to the initial positioning for 2008, has no interest in it, and knows nothing about it. This is why 40% of the victory-starved Democratic electorate supports known losers. This is why 61% of the Republican electorate supports a candidate who has major ideological tension with them. What they give is not their answer; it is just their response.

In other words, these respondents are offering what professional students of political behavior call “non-opinions”. It is a strange, important phenomenon – one that too few professional poll interpreters appreciate. When you ask a question about which the respondent has given no thought, the respondent almost always responds with something. He almost never says, “I don’t know”, or “I have no opinion”. Instead, he tells you something. What he tells you is not his opinion. It is his non-opinion. From where does it come? In early presidential politics, his response is usually just the name that he hears most often from the pundit class.

These polls, therefore, are really nothing more than the conventional wisdom from the media echoed back to the media. This creates an interesting cycle – the pundit class speculates about Hillary, Rudy, etc; the public picks up on this speculation and, when asked what they think about 2008, because they do not really think anything yet, just regurgitate the speculation; this induces the pundit class to speculate more about Hillary, Rudy, etc.

There is no dishonesty on the part of the public here. The echo chamber exists in this way because people feel compelled to give an answer. It seems like the right thing to do when asked a question by a pollster. He wants your opinion, you should give him one. With virtually no knowledge on the subject, upon what else can a respondent draw except what he has picked up in dribs and drabs from the pundit class?

As we move closer to the next presidential election, the electorate will start to pay more attention and will start to consider the race more carefully. Non-opinions will become real opinions; it is only then that these polls will tell us something about the election. The poll results will also begin to change – as attention, interest and knowledge increase, opinions will change from what they “are” today. Thus, the polls today bear no reflection to the polls tomorrow. This makes today’s polls useless – unless you are interested in non-opinion dynamics.

The second reason is that the polling firms ask the wrong people. This problem manifests itself in two distinct ways. Consider the Gallup poll. Gallup reported the combined results of registered Democrats and Democratic “leaners”. The problem with this is that, in many states, being a “Democratic leaner” is not enough for you to vote in a primary. In Pennsylvania, for instance, you cannot go to your precinct at primary time and say, “I really like the Democrats, but I’m a registered independent.” They will not let you vote in the Democratic primary. Thus, the sample is skewed – by how much and in what way, it is impossible to say. Nevertheless, there is a skew. Furthermore, we know from experience that the national set of registered Democratic voters do not select the Democratic presidential candidate. That privilege is reserved for at most one third of the Democratic electorate. The latter two-thirds merely ratify the choice of the former. With the new primary schedules of both parties, it would be a wonder if either race remains unsettled by Super Tuesday. So, why on Earth is Gallup asking a national sample whom they wish to have serve as each party’s nominee? Even if the Democratic or Republican bases did possess something more than a set of non-opinions, this poll would be unable to gauge what those imply for the actual horse race.

These polls do nothing more than facilitate the endless, pointless blather that comes the media. It is a way for paid commentators who have nothing to say to fill the programming void. Sometimes, silence is golden. So is Arrested Development. They should put that back on, instead!

© 2000-2006 All Rights Reserved

Email Friend | Print | RSS | Add to | Add to Digg
Sponsored Links

Jay Cost
Author Archive