Senior Fellow Karlyn Bowman

R-D-I. No, it's not the name of a new motor oil. It's one of the most important political indicators we have, and it is essential to understanding the underlying structure of any election.

Seventy-one years ago, in 1937, the Gallup Organization asked people this simple question for the first time: Do you regard yourself as a Republican, Democrat, Socialist or Independent in politics?

That year, 50% said they were Democrats, 33% Republicans, 15% Independents and 2% Socialists. The South was solidly Democrat, with 73% of its residents identifying with that party and 19% with the GOP. New England was more reliably Republican, with 39% calling themselves Republicans, compared with 34% Democrats. Nationally, 2% called themselves Socialists, but that rose to 6% in New England. When it became clear that socialism wasn't going to gain a toehold in America, Gallup dropped the category.

Now Gallup and other polling organizations ask variants of this same question: In politics today, do you consider yourself a Republican, a Democrat or an Independent? It has been asked hundreds of times in the past 70 years. From years of research, the surveyors know that most people who call themselves Independents actually lean toward one party. So they further inquire whether these respondents lean more toward the Democrats or the Republicans. In polling parlance, these people are called "leaners."

What do the answers to these questions tell us today? Until recently, these poll results were one of the main reasons election analysts felt that the Democrats had a "structural" advantage in this election.

Let's look at some of the numbers. In January this year, Gallup Organization reported that "the percentage of Americans who identified as Republicans in 2007 is the lowest of any of the 20 calendar years since 1988," the year Gallup started doing its interviewing primarily by phone. Based on 26,000 interviews in 2007, Gallup reported that an average of 50.6% of Americans identified with or leaned toward the Democratic Party and 39.6% identified or leaned Republican.

That's quite a change from results in 2003, before our last presidential election, when Gallup found perfect partisan parity. Forty-five point five percent identified or leaned to the Republican Party; 45.2% to the Democratic Party. Right before this year's Democratic National Convention, the Pew Research Center reported similar results from its polling, conducted between January and mid-August: "[T]he Democratic Party's advantage in party identification remains as large as it has been over the past two decades. . . . The Democrats have a 13-point lead in party affiliation (51% to 38%) among registered voters when independents who lean to either party are included." Four years ago, though, the Democrats' lead in Pew's polling was a slim three percentage points--47% to 44%.

Immediately after the GOP convention, Gallup found a sharp increase in the share of Americans who identified themselves as Republicans or leaned to that party. Forty-seven percent of those surveyed in its Gallup/ USA Today poll from Sept. 5 through Sept. 7 identified with the GOP, up from 39% before the convention. Forty-eight percent identified or leaned toward the Democratic Party, down from 53% before the GOP gathering.

Will the GOP gains last? Probably not. Gallup cautions that post-convention gains in partisan identification tend to be short-lived, averaging five percentage points (with leaners included). Pew's poll conducted Sept. 9 through Sept. 14 has the Democrats with a nine-point lead, down from 13 from the pre-convention compilation.

Real change in party identification comes slowly. The Republicans' image has improved, and Republicans are much more enthusiastic about the race than they were a month ago. But the party-identification measures still suggest the playing field favors the Democrats.

Karlyn Bowman is a senior fellow at AEI.