Time-tested polling methodology fails us sometimes. -- Polls & registration nos. don't reveal passion & motivation.

Several recent articles by Georgia's best have examined how traditional methods used by pollsters are being reexamined.

The PI, in a fascinating column on this topic on 10-03-04, wrote:

"Caller ID and answering machines allow wealthier homeowners to block calls from pollsters. This year, much has been made of the fact that pollsters are barred from calling cellphone numbers. And a growing number of young people use cellphones exclusively. "

And on 10-06-04 the PI noted:

"At least in the 2000 presidential race, last-minute registrants were less likely to vote in November than more practiced voters. For thousands, the passion that drove them to register evaporated by the time George W. Bush and Al Gore presented themselves on the ballot."

And Bill Shipp recently observed in one of his columns that:

"What about the polls? They're scary. In recent times, the polls, even some of the traditionally good ones, have been way off target, particularly in Georgia. The communications culture has changed. More people than ever shun telephone interviews. The opinions of blacks and other minorities are virtually impossible to gauge. They are increasingly suspicious of invisible interrogators."

Other articles I have read discuss how pollsters routinely exclude people who haven't voted in previous elections from the "likely" voters category, so newly registered voters (18-to-21 year olds excepted) don't make the cut. The reason for screening them out is simple: Statistics show that people who had the chance to vote and didn't, really aren't likely to vote.

In a 10-13-04 column in the Washington Post Terry M. Neal believes that time-tested polling methodology may have a blind spot for this year's newly registered voters. Why? They are highly charged, and such voters might swing the election.

Mr. Neal's column is entitled "What Poll and Registration Numbers Don't Reveal; Passion and Motivation to Vote Are Hard to Gauge."

Excerpts:

Anecdotal and quantitative evidence suggest that Democrats and independent groups that support Democrats have done a better job than Republicans at registering new voters in key battleground states. In a normal year, the difficulty in getting the newly registered to the polls might mitigate this advantage. But anti-Bush passions on the left are running exceedingly high, making it more likely that marginal voters -- people who rarely or never vote -- will turn out this year.

Aside from new voter registration, many Democrats and even some nonpartisans believe the polls are not accurately reflecting the intensity of passion felt by those on the left, many of whom will be motivated to vote for the first time out of anger at Bush and his policies.

Remember the Republican Revolution of 1994? Leading up to the midterm election that year, most pollsters and analysts expected GOP gains, but few predicted the ensuing blowout, in part because it was difficult to quantify through polls the emotions that were percolating among white male voters in particular that year.