The Useful Idiocy of Polling

Polls are commonplace these days and are often reported on as if they are actual factual snapshots. Joe Carter provides some useful analysis on why polls are pretty much worthless:

Although opinion polls are often treated as if they were harmless detritus of the news-cycle, they are powerful tools for promoting overconfidence and slip-shod reasoning. Take, for instance, two of the worst types of polls—those that purportedly measure “favorability” and the “job approval rating” of politicians such as the president and members of Congress. Such polls might be useful if the general public were aware of the president and legislators’ duties, and if we could appeal to a single, objective standard to judge polls’ relevance and faithfulness to truth. But we don’t. Instead, polls create an illusion of assurance, allowing us to fool ourselves into thinking we have precisely quantified our vague qualitative judgments.

Be sure to read the whole thing.

What If The Polls Are Wrong?

Polling has been conducted at an unprecedented level during this election cycle. As of today, margins between the two presidential candidates are anywhere from two to sixteen points depending on which survey you read. But what if we’re heading for another election where the polls are totally wrong? I’m not referring to 2000 or 2004 where there were most famously problems with the exit polls. I’m talking about 1980.
Time Magazine conducted a lengthy analysis of the opinion polls following Ronald Reagan’s landslide victory over President Jimmy Carter. (Hat tip: The Corner) Two trends jump out from this analysis. First, there were estimates that many voters changed their minds in the last 48 hours. These voters broke towards Reagan because they were unsure about how the Iranian hostage crisis was being handled. As a result, they changed their minds late and affected the outcome of the election.
The other trend they noted then is one that certainly seems to fit today: pollsters oversampled Democrats. My guess is that’s exactly what’s happening now as pollsters have been guessing that turnout will be higher among Democrats but fail to take into account other factors that can affect how they vote.
Making matters worse is that current polls have had a tendency to overstate support for Senator Obama. This is because he draws a lot of younger voters but they tend to be notoriously unreliable in showing up at the polls on Election Day. Also, blue collar voters that make up substantial portions of the electorate in key states are very difficult to poll because they don’t respond to pollsters.
The bottom line is this: go out and vote today. Don’t let the pollsters or the media or anyone else tell you that this election is over. We could have a long night ahead of us and some very surprising results in the end.

The Media Bandwagon

One of the ongoing story lines in this election season is the preponderance of positive media stories for Senator Barack Obama. The latest example was Time’s Joe Klein gleefully speculating what an Obama presidency might look like. (Hat tip: Newsbusters)
Articles such as these should be seen by readers as what they really are: propaganda for Obama. To be more precise, the media are engaging in a technique known as the bandwagon:

Bandwagon is one of the most common techniques in both wartime and peacetime and plays an important part in modern advertising. Bandwagon is also one of the seven main propaganda techniques identified by the Institute for Propaganda Analysis in 1938. Bandwagon is an appeal to the subject to follow the crowd, to join in because others are doing so as well. Bandwagon propaganda is, essentially, trying to convince the subject that one side is the winning side, because more people have joined it. The subject is meant to believe that since so many people have joined, that victory is inevitable and defeat impossible. Since the average person always wants to be on the winning side, he or she is compelled to join in. However, in modern propaganda, bandwagon has taken a new twist. The subject is to be convinced by the propaganda that since everyone else is doing it, they will be left out if they do not. This is, effectively, the opposite of the other type of bandwagon, but usually provokes the same results. Subjects of bandwagon are compelled to join in because everyone else is doing so as well. When confronted with bandwagon propaganda, we should weigh the pros and cons of joining in independently from the amount of people who have already joined, and, as with most types of propaganda, we should seek more information.

(Emphasis mine)

Various media outlets will point to that unshakeable gospel of public opinion, the polls, as proof that their candidate, Senator Obama, has an insurmountable lead and Senator McCain should just go ahead and save everybody time and effort and concede the race now even though the voting doesn’t take place until November 4. Even Senator Obama can’t help himself from thinking ahead until after the election. (Hat tip: Hugh Hewitt) Perhaps that’s why he is charging admission to the media for his big victory party on Election night.
The fatal flaw in this logic is that polls are far from perfect. In fact, there is growing evidence that they are totally unreliable. Even though some media outlets will use an average of polls as a truer barometer of public opinion if each one of the polls that is figured into that average is, in and of itself, imperfect then the average of those polls is also imperfect. You can at least be certain of this much: journalists probably aren’t asking any questions to determine the validity of the poll results.
At least signs point to voters realizing by an overwhelming margin that journalists want Senator Obama to win. (Hat tip: Newsbusters) Maybe voters are a whole lot smarter than the media thinks.