The 2000 presidential election produced two enduring stereotypes, one
accurate, one misleading. The first is that the two major political parties
are at absolute parity in the electorate as a whole. That remains accurate
today. The second is that America can be neatly divided into "red states" (Republican) and "blue states" (Democratic), with few opportunities for change. That's very misleading. In fact, between 13 and 20 states are likely to be in play in 2004, with a significant majority of those states being genuine toss-ups.
Many analysts read early polls and discuss campaign strategies. But it's equally important to understand the evidence from recent elections and from post-election demographic changes that determine the landscape on which candidates will compete.
Every objective sign at this point indicates another very close presidential election. Party identification numbers are dead even. The president's job approval rating continues to hover around 50 percent, which tends to be a pretty good measure of how the vote will go for an incumbent, especially as we get closer to November. And there's been a long-term decline in the percentage of the electorate that's truly up for grabs -- largely, I think, the result of the slow transformation of the two major parties from nonideological coalitions of regional parties whose voters often split tickets, to real national parties. That means swing voters, while likely to be decisive in the end, aren't numerous enough to give either candidate a big lead -- unless they go one way overwhelmingly.
Republicans are likely to benefit marginally from the decennial reapportionment of congressional seats after the 2000 elections, which gave red states a net gain of seven electoral votes over blue states.
Many factors will ultimately determine the "battleground" landscape, including candidates' strategies, the resources they and the parties devote to particular places, and how certain issues in the campaign "cut" in particular states. If one candidate or another obtains a big national majority, it will obviously take some current battleground states off the table while putting others into play. Geographical targeting is also a big factor: Al Gore's campaign largely gave up on Ohio, Arkansas, and Tennessee down the stretch in 2000; the Bush campaign, after a large feint, gave up on California and Illinois.
But demographic change is probably the least-noticed factor in shaping the battleground. Four years may not seem like a long time, but a lot has happened in a number of the 2000 battleground states that will help one party or the other. In some cases, demographics are driving the political balance of close states in opposite directions. And demographics probably matter more than ever, because small numbers of votes changing could "flip" so many states.
Taking all these factors into account, my guess is that there are 13 battleground states where neither presidential candidate has much of an advantage. They are Pennsylvania, West Virginia, Ohio, New Hampshire, Wisconsin, Minnesota, Iowa, Missouri, New Mexico, Arizona, Nevada, Oregon, and the ultimate 2000 battleground, Florida. Seven other states lean in one direction or the other, but they could become battleground states under the right circumstances. Illinois, Washington, and Michigan lean to John Kerry, while Colorado, Tennessee, Louisiana, and Arkansas lean to George W. Bush.
If you assign the "leaners" to the candidate they favor now, the battleground becomes clearer: Kerry has an electoral vote base of 200, while Bush has an electoral vote base of 196, with 142 electoral votes undecided, and 270 needed for victory. That means Kerry needs to get 70 of those 142 battleground votes, while Bush needs to get 74.
To get a sense of the dynamics at work in the battleground states, it's useful to take a closer look at three of them: Florida, which many Republicans think is leaning their way; Pennsylvania, which many Democrats think has shifted in their direction; and Ohio, which virtually everyone believes -- I am among them -- is the big prize among toss-up states.
In all three states, two particular demographic factors require some explanation. The first is the growth in minority voting age population (VAP) since 2000. With some exceptions, the size of this statistic helps Democrats, because roughly 90 percent of African-Americans and two-thirds of Hispanic voters tend to vote Democratic, and there's nothing about the dynamics of the presidential race so far that is likely to change those proportions. The second factor is the relative voting strength of fast-growing "exurban" or fringe city communities like Mesa, Ariz., or the outer suburbs of Atlanta, where, on average, Republicans enjoy about a 2-to-1 majority.
Everybody in politics understands the importance of minority voting patterns for Democrats. The importance of exurbs to Republicans is less well known, but equally striking. When people move to these areas, they tend to start voting Republican by significant margins. Their rapid growth throughout the 1990s and into the current decade has given the GOP the kind of concentrated geographical support areas that are highly susceptible to intensive get-out-the-vote efforts. An analysis of the big Republican sweep in Georgia in 2002 showed that exurban growth since the previous midterm election of 1998 was the single biggest factor in the outcome. It's no accident that Georgia was also the national pilot for the Republican Party's new "72 Hours" get-out-the-vote initiative, which will be very visible on Election Day 2004 in the evangelical churches and mega-malls of exurban areas all over the country.
In looking at Florida, Pennsylvania, and Ohio, I'll discuss underperformance and overperformance by the two 2000 candidates in key voter segments, and what that might mean for 2004, as well as how the issue landscape for this year might affect the electoral landscape.
The scorched earth of Florida. Though it was the ultimate battleground state in 2000, Florida's 2002 elections, featuring a landslide win by Gov. Jeb Bush and a solid performance by the GOP up and down the ballot, convinced many Republicans to firmly move the state into their column for 2004. But demographic trends in Florida since 2000 are not necessarily friendly for the governor's brother.
According to my estimates, the Hispanic voting-age population in Florida has
increased by an astounding 31 percent since 2000. Moreover, it appears that
most of that growth has been among groups that tend to vote strongly Democratic,
especially Dominicans in southern Florida and Puerto Ricans and Mexicans in
the Orlando area. Cuban-Americans gave Bush an estimated 81 percent of their
votes in 2000. But the lack of an issue like the Elian Gonzalez flap, coupled
with some disenchantment with the administration's policy toward Castro's Cuba,
makes it an uphill battle for Bush to match his 2000 performance. Moreover,
the turnout in the Cuban-American community may decline, based on 2004 demographic
realities.
Meanwhile, the African-American VAP in Florida has risen an even more surprising 13 percent since 2000 (non-Hispanic white VAP has gone up less than 2 percent). I think the 2000 exit poll numbers on African-American turnout in Florida were significantly inflated, so there's some room for growth that Democrats can try to exploit.
On the other side of the ledger, the rural vote in northern Florida is growing more conservative and Republican by the minute, and the exurban vote in Florida has grown by an estimated 10 percent since 2000. The fast-growing exurbs that tend to break 2-to-1 Republican are especially concentrated in the Orlando and Tampa metro areas, making the Interstate 4 corridor between the two -- also the home of a big chunk of the non-Cuban Hispanic population -- the obvious cockpit of the 2004 election.
It's worth noting that Gore did somewhat worse in Florida in 2000 than his national average among several key voter categories, according to exit polls and census data. He did 10 percent worse (33-66) among households with incomes over $100,000, and 7 percent worse (45-52) among voters with post-graduate educations. This may reflect the cultural conservatism of Sun Belt elites, but if Kerry can do better among these groups, he will do better statewide.
As for the issues landscape of 2004, Florida is a state with below-average unemployment and above-average growth, where the "job loss" issue may not cut very sharply. Kerry's ability to make income stagnation, high public debt, consumer debt, and the rising cost of college tuition, issues in the presidential race which may help him in states like Florida.
The Keystone State. If many Republicans assume the 2002 election results have made Florida a genuine red state, many Democrats are assuming that Pennsylvania will be a relatively easy target for 2004. Gore, after all, won the state by 5 percentage points in 2000, despite a poor performance among the culturally conservative non-college-educated white men who were supposedly the big target of his "people versus the powerful" convention message. Gore's strong performance in the Philadelphia suburbs was strengthened by Democrat Ed Rendell's comfortable win in the 2002 gubernatorial race.
But Pennsylvania's demographics are not necessarily cutting in favor of the Democratic Party. Minority population is relatively stagnant. Philadelphia (the powerhouse of both Gore's and Rendell's electoral base) and Pittsburgh are losing population steadily. And there's plenty of evidence that it will be hard for Democrats to improve on the massive minority and labor turnout effort of 2000, in which African-American turnout actually exceeded white turnout. Base turnout may have maxed out in Pennsylvania in 2000.
Pennsylvania's exurbs (which represent only 7 percent of the electorate) have grown since 2000, but not as rapidly as in the Sun Belt. Moreover, Republicans enjoy a significantly smaller electoral margin -- more like 10 percentage points than the 30 points or more common elsewhere -- in these areas.
The place to watch here is not southeast Pennsylvania -- dominated by Democratic Philadelphia and its Democratic-leaning suburbs -- but northeast Pennsylvania: Wilkes-Barre, Scranton, Allentown, and Bethlehem. That's where the state's swing voters are heavily concentrated, and where Kerry's ability to improve on Gore's performance among culturally conservative Catholic voters may be crucial.
The economy in these areas is changing. As technology and white-collar employment replaces the older, smokestack employment pattern, new jobs are being created while others are lost. Consequently, middle-class voters are uncertain who is at fault for their economic angst. In Wilkes Barre, the economy is stagnating, while small business has promoted an economic revival in the Lehigh Valley. The result is a concentration of middle-class centrist voters without a strong allegiance to either party.
Both there and in southwest Pennsylvania (where Democrats have done poorly in recent elections), many voters will have to decide between Bush's appeal on cultural issues and Kerry's appeal on the job loss issue -- and each candidate's ability to address the other's core issues, as well.
Buckeye bonanza. The conventional wisdom going into this election is that Ohio may prove decisive. My own analysis definitely suggests this is a state offering rich opportunities for John Kerry.
Yes, Ohio has been dominated by Republicans in recent statewide elections, but Clinton carried it twice, and Gore lost it by only 4 points after basically conceding it to Bush.
The narrowness of Gore's loss in Ohio in 2000 is made more striking by the numbers indicating a classic underperformance by the Democratic candidate.
Nationally, Gore won voters with a post-graduate education by 8 percentage points. In Ohio, he lost that same group -- which represented 15 percent of the state's electorate -- by 6 points. Nationally, Gore ran even with Bush among independent voters. But Bush ran 16 points ahead of Gore among Ohio independents, who represented one-quarter of the state's electorate. Nationally, Ralph Nader took 2.6 percent of the vote, and he won no more than 2 percent in Michigan and Pennsylvania. He got almost 3 percent in Ohio -- 114,000 votes, or about two-thirds of Bush's margin of victory. The don't-waste-your-vote message of the Gore campaign probably never got through in Ohio, and it was undermined by the concession of the state to Bush.
While labor turnout in Ohio was excellent, African-American turnout was not impressive, especially as compared to Michigan and Pennsylvania.
If Kerry targets Ohio, as he almost certainly will, then he should be able to eliminate all those discrepancies between Gore's national performance and his underwhelming numbers in Ohio.
On the Republican side, there has been significant exurban growth around Cincinnati, but dwarfing it in significance is the potential Democratic gain among voters statewide who earn between $30,000 and $50,000 per year, a demographic that Bush won by 5 points in Ohio in 2000. Nationally, Gore and Bush evenly divided voters in the $30,000 to $50,000 income range.
Ohio, with an above-average unemployment rate and an above-average dependence on embattled manufacturing sectors, is probably "ground zero" for the job loss issue in 2004. If Kerry manages to address it in a way that doesn't offer economically insecure voters a cultural reason to vote for Bush and that doesn't compromise him in high-growth battleground states in the Sun Belt, he is very well positioned to win Ohio, and the presidency.
A lot has changed. I could go on and on to look at other battleground states and their dynamics, but the three states we've analyzed provide several general lessons for the Democratic campaign of 2004:
Pay attention to how states look now, not just how they looked in 2000, much less how they look through the lens of old stereotypes.
Pay especial attention to the growth of Democratic-leaning Hispanic voting
populations, and to Republican-leaning exurbs. These dynamics could be crucial in states like Arizona, New Mexico, and Nevada that -- like Florida -- have both kinds of growth.
Focus on addressing specific failures of the Gore campaign in 2000, and don't take its specific successes for granted.
Think strategically about targeting states: A message aimed at truly competing in the South, for example, could not only make Arkansas, Louisiana, and Tennessee true battleground states, but could also help appeal to voters in northern Florida, and in culturally conservative pockets of voters in Pennsylvania, Ohio, and West Virginia. Conceding the entire South to Bush would, most likely, leave the Kerry campaign without a cushion, if one of the closely fought industrial states is lost. Moreover, if the South is uncontested, the Bush campaign will be able to redeploy media money and other resources elsewhere.
Above all, remember: The candidate with a message that can both energize the party base and appeal to a majority of swing voters will almost always win, on any battleground, in any year.
Blueprint Keywords: Extra Battlefield
