WHAT if ......

[Tigermike 2,602]

Personally I think there are much fewer undecided voters than the media and polsters seem to think.

MYTHS OF THE UNDECIDED

John Podhoretz

October 26, 2004 -- WHAT if there are no undecided voters left?  What if the polls we're seeing now are a precise mirror of where we're going to be on Election Day?

In polls seven days before the vote, the Bush and Kerry numbers add up to 95 percent to 98 percent. If that holds, up, the president is probably going to win narrowly.

And it may hold.

Right now, in two major polls, President Bush leads Sen. John Kerry outside the "margin of error" among likely voters. In one of them, the Gallup released yesterday afternoon, the Bush and Kerry vote totals 97 percent.

This is a very significant fact, because the presidential vote between the two major candidates almost never adds up to 100 percent. Even in a year without a serious third-party challenger, 1 percent goes to minor candidates like the Libertarian Party's Michael Badnarik. And this year Ralph Nader who will probably get 1 percent or so himself.

That leaves — get this — only 1 percent of the vote remaining undecided.

Under these conditions, using Gallup as the barometer, Bush will probably win. Why? Because his lead is outside the margin of error and more than a million Bush voters would have to switch over to Kerry to give the Massachusetts senator the victory.

But say the undecided number is really 2 percent, or even 4 percent. What happens then?

There's a great deal of controversy about how undecided voters make up their minds in the final week. For years, political professionals have told us that undecided voters "break for the challenger."

This bit of wisdom is based on a long-term trend in the Gallup poll. Since 1936, undecided voters in the last poll taken before a presidential election do seem to opt for the challenger in greater numbers.

But, as dalythoughts.com's Gerry Daly points out, the result changes significantly if you use the second-to-last Gallup as your reference point.

Only twice since 1936 have undecideds in the second-to-last Gallup poll broken for the challenger. And it was only significant once, in 1980, when the undecided swing to Ronald Reagan helped him win a landslide against incumbent Jimmy Carter.

In the other 10 presidential elections featuring incumbents in the past 68 years, the voters from the second-to-last Gallup either broke for the incumbent or broke evenly.

And let's be clear what "break" means. It's not that 100 percent of undecideds go one way — it means that 60 to 65 percent do. Using yesterday's Gallup poll as a benchmark, Kerry would need an even greater break than that to catch up to Bush — 70 to 75 percent of the undecided voters at least.

That huge margin for Kerry just doesn't seem likely right now.

Ah, but what about all the excitement over the newly registered?

Look at the numbers here. Some people are talking about a staggering increase of 10 million voters this year. Both parties have spent vast sums looking for these new voters and registering them, and there's reason to believe their efforts will basically cancel each other out.

But let's assign 55 percent of them to Kerry. That's 5.5 million voters. With those 5.5 million voters, surely then Kerry will win.

Um, no, he won't. Because Bush will get 4.5 million new votes. This would make Kerry's margin among new voters only 1 million votes — in an electorate of 115 million. That's not even a single percentage point increase. Kerry can't win that way.

Now here's why the tiny number of undecided voters may matter even less than these numbers indicate.

You're an undecided voter. Don't really feel hot either way. You get up next Tuesday, and you hear on the radio or see on TV that there's a record turnout — long lines at the polls. You may get discouraged right there, and decide not to show up the way you haven't shown up before.

Or you do go. And there are long, long, long lines. You might have to wait half an hour.

Do you? Or do you go home?

They won't all go home. But some will.

The perverse result of exceptionally high turnout is that the resulting problems will also keep hundreds of thousands of people from showing up.

Of course, elections are won at the state level, so maybe Kerry can use undecideds to take both Florida and Ohio, making it almost impossible for Bush to win.

But there won't be a million new voters in Ohio. There will be a few hundred thousand at most. And again, they'll go both to Republicans and Democrats.

Democrats who are assuming that undecided and new voters are going to carry the day for them are doing what craps players call "betting on the come." They are hoping for a result that the odds don't favor at this moment.

E-mail: podhoretz@nypost.com

http://www.nypost.com/postopinion/opedcolumnists/32697.htm