They are, as the saying goes, works in progress. Polling organizations are constantly tinkering, making adjustments in how they do their work.
When I was in college, I studied opinion polling, and one of the things we studied was the methodology being used by the earliest pollsters. In the dark days of the Depression, they did their polling based on automobile registration and residential telephone service records. At that time, many people owned neither — and thus were ignored by the pollsters.
But their votes weren't ignored on Election Day, and — contrary to the polls' forecasts — President Herbert Hoover was defeated in his bid for re–election.
Pollsters have been learning from their experiences in the last 80 years, which is why it is news today if they are way off the mark.
From time to time in my life, I have heard supporters of underdog candidates refer to Harry Truman's "upset" victory over Tom Dewey in 1948. It is the most notorious example of polls being wrong so it is the holy grail of the apparently doomed. I have observed that the greater the underdog, the more likely it is that Truman's triumph will be mentioned.
The truth is, though, that polls are right more often than they are wrong. That's something else I have noticed over the years. The margin is seldom on the nose, but the winner is usually the one who was picked by the polls.
It's an evolving science. Until actual votes are counted, nothing is certain, but polls provide a glimpse into what the electorate is thinking.
And sometimes those polls can be positively prescient.
About this time four years ago, as Julian Borger wrote in The Guardian, Barack Obama and Hillary Clinton were shaping up to be the front runners in a Democratic field that included the party's most recent vice presidential nominee.
Borger's article didn't go so far as to predict that Obama would win not only the nomination but the election as well. It did, however, observe that polls at that time were suggesting Clinton would win the nomination "with ease" — which, of course, did not happen.
Consequently, anything the polls tell you must be taken with at least a grain of salt.
On the other hand, the findings of a poll — any poll — cannot be taken lightly.
Polls today are telling us that:
- Obama would have the advantage against any potential Republican challenger, and his widest margins, according to the latest ABC News/Washington Post poll, are against the women who are mentioned most frequently.
Perhaps that is an indicator that Americans are not yet ready to elect a woman president. Or perhaps it means the right woman for the job simply hasn't come along.
- Republicans are less satisfied with their choices for the nomination than they were four years ago.
But there's another interesting point in Borger's article.
Well, it wasn't really his point — it was a point that was made by unnamed Democratic strategists — but it may have a great deal of relevance to 2012.
In the 2006 midterms, you may recall, Democrats took narrow majorities in both chambers of Congress after being in the minority for more than a decade. Borger quoted a Democratic strategist as saying that winning one chamber would have been the ideal way to set up the party for the coming presidential election.
" 'One third is ideal,' one said. The wife of another top party fixer was heard saying her husband would rather the Democrats just won the House, in the interests of winning the big prize, the White House, in 2008."
To bolster the argument, Borger quoted David Corn, Washington editor of The Nation: "Had the Democrats won only the House, they could have positioned themselves as the insurgents of Washington, running against the sclerotic, corrupt Republicans in the Senate and the White House. But now they have the Senate and the House, they are 50–50 partners in governing."
Of course, the Democrats went on to win the 2008 election, anyway. But the same logic could apply to the Republicans of 2012.
Time will tell.