The accuracy of political polls

You know, sometimes things just don’t turn out as expected. If you don’t believe me, just ask Barack Obama.

All the polls taken right before last Tuesday’s Democratic Presidential Primary election in New Hampshire indicated that Mr. Obama would easily defeat top rival Hillary Clinton, perhaps even by a double-digit margin. What’s more, exit polls taken throughout the day right up until the polls closed seemed to confirm it. But then the actual results started coming in from the various precincts…

Clinton got off to an early lead, but the pundits and pollsters shrugged it off by saying that Obama’s stronghold precints in the western half of the state and the major college towns had yet to begin reporting their results. Cheria and I sat there in our den throughout the evening watching the coverage, expecting to eventually see Obama start to close the gap. It never happened.

When it was all said and done, Clinton had managed to maintain her steady lead until the very end, leaving her just as surprised at the win as everyone else. After all, in the days between Iowa and New Hampshire her campaign appeared to be floundering, and all day long there had been speculation that a major shakeup was in the works for the following day.

I think the accuracy of political polls really boils down to two questions:

1. Are the operators calling the right people?
2. Are they asking them the right questions?

It appears that in this case the answer to both questions was probably no. It’s actually pretty common for one or two pollsters to miss the mark on occasion with the rest of them “getting it right”, but on Tuesday everyone got it wrong, including the pollsters working for the campaigns themselves!

That’s why it is so important for voters to get out and cast their ballots even if the polls indicate a handy win or defeat for the candidate they are backing. You just never know when an expected blowout will turn out to be a squeaker, and just a vote or two really can make the difference.