Tuesday, September 30, 2008

A Good Time To Have The Voting Begin

As we've noted here previously, a growing number of states are allowing so-called "early voting."

The early voting phenomenon will, we predict, soon become a very commented upon aspect of this year's presidential election. Already, voting has begun in eight states, including the battleground state of Ohio. In addition, voters in other states--including Virginia--are taking advantage of better publicized provisions of absentee voting laws to also vote early.

The result is that while the debates are still underway, and while the larger debate over our economy is in a state of flux, voters are going to the polls.

Given where public opinion polls are today--with Obama having a roughly 4-6 point edge over McCain--that's got to be good news for the Obama campaign. The Obama campaign is also using it's field organization to exploit early voting by getting to the polls some of those folks--senior citizens, African-Americans--who are sometimes more difficult to catch on election day itself.

The other thing about early voting is that it impacts the opinion polls. From now on out, pollsters are going to be capturing some people who have ALREADY voted. Furthermore, when the networks set up their exit polls on election day, they are going to be missing people who have ALREADY voted.

So, this year, we're going to get some pre-election exit polling, and on election "day" we're going to be missing a significant chunk of voters. It'll be interesting to see how this plays out.

It'll also be interesting to see if Obama gets an early lead in this very fluid election environment.

Here's one scenario that could cause more misinformation on election night than we had 8 years ago with Bush v. Gore: Let's say that in the roughly half of states allowing early voting, 20 percent of the electorate votes early. Let's also say that those voters skew 2-1 in favor of Obama. Then, on election day, the exit pollsters in a few of those key states pick up that McCain is leading amongst voters actually exiting a poll, but narrowly. That could turn out to be misleading if they haven't also factored in the early voting that favored Obama. Wouldn't it be nice if the exit polling data was so suspect that the networks just had to wait for the real returns?

No comments: