As we approach the 2004 elections, the conventional political wisdom about the state of the electorate goes something like this: “This is an extremely polarized, evenly divided country with very few people in the middle.” [IMGCAP(1)]

Sometimes, however, conventional political wisdom is wrong, and I believe this is one of those times. Underestimating the opportunity that those voters in the middle present with respect to growing their respective bases could have dire political consequences for the parties this year.

I recently did an analysis of the past four national elections (1996, 1998, 2000 and 2002) based on nearly 60,000 exit poll interviews conducted by Voter News Service. Unlike most surveys in which voters are asked only their party identification, the VNS survey asks voters exiting the polls to reveal their actual voting behavior for each race in which they voted.

I categorized these respondents into four major voting groups: those who voted straight Democratic tickets, straight Republican tickets, straight other party tickets and split tickets. In the off years, the data covered races for governor, House and Senate. During on-year election years, presidential data was added.

My analysis found three key points. First, the country has been steadily moving toward Republicans. Over the past four elections, Democratic straight-party voters hovered around 36 and 40 percent of the vote depending on whether it was a presidential or an off-year election. In other words, there was no real movement.

Republican straight ticket voters, on the other hand, have increased from 34 percent in 1996 to nearly 42 percent in 2002. Although that number is likely to decrease slightly this year because Democrat performance historically tends to be stronger in presidential years, the data shows there has been a slow but sure drift toward the Republican Party. Today, based on voting behavior (not party identification) Republicans can fairly claim majority-party status.

Second, the data shows that for the most part, Republicans can thank ticket-splitters for their progress. The biggest jump to the Republican camp occurred in 2000 when George W. Bush, in contrast to the previous presidential election year, was able to move 5.4 percent of ticket-splitters, straight other party voters, and perhaps even some straight Democrats into the Republican column. While that may not seem like much, in a close election — which 2000 certainly was — those 5.4 million voters helped put Bush over the top in key states.

Finally, and most importantly, the exit polls clearly show that, despite conventional political wisdom, the middle is far from small. In fact, more than 20 percent of voters told interviewers they had split their tickets in the last two elections. Far from being a marginal group, the middle 21 percent translates into more than 21 million voters.

So who are these folks in the “big middle”? We know from other survey data that ticket-splitters in the most recent 2002 election were more likely to be found on the East Coast, in rural areas and in small towns. They tended to be moderate, younger, unmarried and Catholic.

But if we end our analysis at this level, comparing only straight-ticket and ticket-splitting voters, we miss another more subtle voting pattern at work. Closer analysis shows that the actual voting universe in play is even larger than the ticket-splitting middle. The universe of so-called swing voters is larger and more encompassing than merely the 20 percent who split their tickets. This group includes those who change their voting behavior dramatically from one election year to the next, such as those who transition from straight Democrat voting one year to ticket-splitting another. Including these kind of “mobile” voters and the ticket-splitters the big middle grows to as much as 28.9 percent.

Here’s an example of one these subgroups: women 60 years of age and older. In the 2000 election, they voted straight ticket 46 percent-37 percent in favor of the Democrats while 17 percent were ticket-splitters. In 2002, that same group voted for Republicans by a margin of 41 percent-38 percent with 20 percent ticket splitters. That’s a net shift of 12 points from Democratic straight tickets to Republican straight tickets. If these women vote this fall like they did in 2002, it’s bad news for likely Democratic nominee Sen. John Kerry (Mass.), but due to their propensity to shift parties the question remains: How will they vote? Groups such as this one are very much in play in 2004.

This kind of subtle but significant movement can occur in various subgroups and is difficult to detect in macrodata. Getting an accurate picture of the potential electorate will be crucial to both parties in the next election. Not only is the middle much larger than most pundits would have us believe, but it also exists in a shifting environment in which the movement of voters in smaller subgroups can impact a close race.

Clearly, the bases of both parties are polarized. That much is true. And in this coming presidential election, both sides are likely to turn out roughly the same percentage of party regulars. Due to the fact that Republicans have a slight edge in self-identified straight ticket voters, this does give them a slight advantage.

However, the election battle this fall won’t be won or lost in the base — right or left — but in the big, ticket-splitting middle. Both Bush and Kerry have solidified their prospective bases — Bush with his strong national defense and tax-cutting economic policies, Kerry with his Bush-bashing personal attacks and his unparalleled liberal voting record.

The question now is who is going to grab the middle? Bush is well positioned to appeal to this group of voters with his compassionate conservative agenda of health care, education, and homeland security policies that resonate with ticket-splitters and swing voters. In contrast, Kerry’s voting record, as the Senate’s most liberal legislator, may alienate him from the middle-of-the road voters he has to reach.

For both Republicans and Democrats, the key to victory is in not underestimating the size or volatility of the millions of Americans who inhabit the big middle. These voters are not looking for ideological dogma, partisan polarization or mudslinging — they want solutions to real problems. The candidate or party that does a better job at providing those solutions will ultimately win.

David Winston is president of The Winston Group, a Republican polling firm.