Polls: Still Meaningless

Just a few days ago, we argued in this column that pools mean little, especially so early in an election cycle (as we are well ahead of the 2012 Presidential election). With two recently released polls, one showing that President Barack Obama would lose to the kind of favorite son (he was born and raised here in Michigan) Mitt Romney while another right on its heels said the President would win, it seems appropriate to examine them in a bit greater detail.

The first poll had Romney winning 46 to 41 percent, with a margin of error of 4%. What this translates into is that he could win bigger, say, 50 to 37 percent, if the whole margin for error goes his way, or lose 45 to 42 percent if the error factor goes towards the President.

The second poll, which plays into Mr. Obama's favor, had him winning 47% of the vote to Governor Romney's 42. Yet the with the same 4% margin of error, that could mean a win of 51-38 for the President, or a win of 46-43 for Romney. Of course, neither projection really factors in how the remaining percentages would go. Yet that only exacerbates the problem. Winning margins could go wildly off course either way depending on how the rest of the voters, those not committing to either projected candidate in the polls in question, cast a ballot.

An interesting aside is that the polls agree that marginal candidates, such as Thaddeus McCotter and Herman Cain, are unknown by around 48% of the voting public.

So what have we here? Two polls which effectively tell us nothing about who would actually win the Presidency (if left up to Michigan) in 2012, with the caveat of something which common sense would have told us immediately if we would take a second to listen: nobody knows Livonia Congressman McCotter or businessman Cain.

Yet both poll results were big news locally. It had to be a slow day.