- The Washington Times - Monday, June 23, 2008

As they have in the past, independent voters hold the key to choosing the next president of the United States. Despite the recent decline in the number of Americans identifying with the GOP - and the boost on the Democratic side - neither party holds a majority when it comes to partisan attachment. And even if Democratic or Republican identification ever did surge over 50 percent, this partisan attachment advantage would be a crucial determinant in vote choice, but not a guarantee. There are always some Democrats who vote Republican and vice versa. For these reasons, independent voters play a pivotal role in picking the occupant of the White House.

But despite their importance, a great deal of confusion surrounds these Americans. Who are they? How and when do they decide to vote? What issues or candidate characteristics push their buttons? And what share of the overall electorate do they represent? These are just a few of the questions about independents that deserve answers. Between now and November, The American Survey will answer these questions. This week I will clear up several mysteries concerning independents and raise some cautions regarding the interpretation of surveys that report findings about this group of pivotal voters. “Independents” may be less of a “swing voter” group than you think.

Let’s begin with some definitions. Pollsters usually classify people as independents by asking, “Generally do you consider yourself a Democrat, a Republican or an independent?” Many surveys only present results using this three-part breakdown.

Looking at partisan affiliation today, roughly 35 to 40 percent of Americans are Democrats, 35 to 40 percent independents and 25-30 percent Republicans. These numbers change from poll to poll and month to month, but independent identification is about at parity with the Democrats today and Republicans hold about a 10-point deficit.

It wasn’t always this way. Independent identification grew from the 1950s to a peak at the end of the 1970s. It declined slightly during the 1980s and 1990s, but increased again in recent years to about where it was at its peak during the end of the Carter administration and the beginning of the Reagan administration.

But there’s an important distinction among so-called “independents.” Many survey analysts ask a couple of follow-up questions. They ask those who say they are “Republican” or “Democrat” if their partisan affiliation is “strong” or “weak,” and they ask independents whether they “lean” toward one party or the other.

These “leaners” are an important group. They represent a large chunk of the independents you read about in polls - in many cases as much as two-thirds of the group. But research has found these Americans are far from “independent.” For example, those who “lean Democrat” vote for that party almost as consistently as partisans. The same pattern is true among independents who “lean Republican” - they vote heavily for the GOP. For example, in 2004, according to the American National Election Study poll, 83 percent of independents who “leaned” Democrat voted for John Kerry for president, just shy of his share among Democratic partisans. A similarly high percentage of “lean Republican” independents voted for George W. Bush. Both are a far cry from the conventional view of independents as an unpredictable “swing” group.

Including these “leaners” in the independent camp lumps in a lot of Americans who vote like partisans. That’s why some surveys today even group these respondents in the Democratic or Republican camps when reporting partisan results. So, you might read a survey that reports the percentage of Republicans and Democrats “including leaners.” That means the pollster has taken a group of people who initially might say they are “independent” and included them in a partisan category. Doing so, at least based on recent voting behavior, makes some sense.

But while this allocation fixes one problem - by taking people who act more like partisans out of the independent camp - it creates another. The pool remaining “true independents” is then rather small - sometimes as low as 10 percent. Analysis of subgroups that small (for example, a random sample survey of 800 Americans might yield less than 40 “true” independent women) may not be accurate. So, lesson number one about independents and polls is to ask about the “leaners.” Lumping them together in one group makes the “swing” voter universe appear bigger than it really is.

Gary Andres, who served in the first Bush administration, is vice chairman of Dutko Worldwide.