Tuesday, October 21, 2008

So what's in a poll number?


Ironically, the national election have focused attention on "where does McCain stand now?" for purposes of polling. Everyone seems to have a poll, and they don't all agree. This begs the question - "Why do the polls differ?"

I assume most of these polls are done honestly and are not falsifying their numbers. If that is the case (assume that with me for now), then why do the polls differ? Although there is always issues of sampling (who is being asked), the biggest difference is what is the general pool being sampled - "likely voters" or "registered voters." In many elections only a small percentage of registered voters do vote, so these groups are not always the same, and often those elegible to vote were not registered. Does that hold true in such times for such an election?

As with the primaries, massive voter registration has occurred since 2004 or even the 2006 mid-terms, and many of these new voters are non-traditional voters, either disaffected voters in past cycles such as minorities or new voters such as young people. We, frankly, don't know if they will show up to the polls and actually vote. However, based upon their activity in the 08 primaries, their willingness to engage now is likely. However, my guess is that their participation is discounted by some polls as "unlikely voters" and therefore are not polled fully.

Even among polls that seek to include them, many of these people may be less able to be polled. For young people who list only cell phones, since that is all they use, they might not be fully captured. Likewise, for other disaffected voters, they might not be home to be polled, since some of these people may be holding down multiple jobs and are not at home at times when they can be polled. If this is true, then the polls may naturally favor more conservative voters, with more conservative and stable lives.

Traditionally, these same factors also affected the unlikely voters from actually voting or even registering. However, with so many states using early voting as a tool, the inability to get to the voting booth on the First Tuesday in November might not prevent them from voting now. The only question would be - do these voters make it a priority, did they register, and will they vote?

My guess is that they did, they did and they will!! Some of the polls are also making this assumption and this may demonstrate why some of the polls show as much as a 10% lead for Obama. But at the end of the day, all registered voters have to remember that ... truly, every vote counts. If you want to complain about the country, then put up or shut up - and make sure to vote

I think we will see a record turnout that exceeds even the primary totals. Hopefully voting districts are preparing for the "storm" so that the votes can be captured correctly!

No comments: