Follow the Votes

Driven by events both foreign and domestic, President Bush's job-approval rating gyrated wildly during his first three years in office. This year, as Bush's re-election campaign has gotten well under way, another set of numbers important to his political prospects has shifted far less dramatically. We've seen subtle, sometimes almost imperceptible changes in Bush's standing vis-a-vis his Democratic challenger, Sen. John Kerry of Massachusetts.

In early January, Bush had a double-digit lead over Kerry. But the incumbent's edge shrank, then vanished in the final days of that month. Bush ran behind Kerry by at least a bit in 11 of the next 18 polls, dead even in three others, and ahead in four.

Then, beginning in the first week of March, Bush was again consistently ahead for a time, before slipping once more into a trailing position. Bush pulled ahead once more in early April, but he fell back behind in early May and remained there until very recently. But now, almost the same number of national polls show him ahead as show him behind, indicating that the 2004 race for the White House is about even.

Since the third week of January, no major national poll has given either presidential candidate a double-digit lead. Indeed, most of this year's leads have been 5 percentage points or smaller.

State-level polling has found similar patterns. It's difficult to know with any real certainty how presidential candidates are doing within a given state, because state polls tend to be conducted much less frequently than national polls, and because press reports tend to lag.

Moreover, locally sponsored news media polls vary greatly in quality, from very reliable to truly shoddy. Some of the worst even use computers to interview voters: "Punch one for Bush, two for Kerry, three for Nader." That's a very dubious methodology, favored mostly by television stations with low budgets and even lower standards.

The best way to sort out the true battleground states this year is to start with the 2000 presidential election results. That election was, in effect, a 105 million-person sample survey of American voters.

For the most part, the states that were competitive in the contest between Texas Gov. George W. Bush and Vice President Al Gore are competitive today in the struggle between Bush and Kerry. The notable exceptions are Tennessee, which is no longer up for grabs; and Arizona, Colorado, and Louisiana, which weren't competitive in 2002 but now seem to be.

Although Gore lost his home state of Tennessee by a relatively slim margin, in the absence of a Tennessean on the ticket, the Democrats are widely expected to do worse there than they did four years ago. Meanwhile, the Democratic Party is taking a run at several states that Bush carried comfortably last time: Arizona, Colorado, Louisiana, and to a lesser extent, Virginia.

On the other side, the Bush campaign is throwing some money into two strong Gore states: Delaware and New Jersey. Otherwise, though, the list of competitive states suggests that we'll pretty much see a rerun of the last fight, albeit with considerably more money being spent -- and spent early -- by both sides.

The next step, in the words of Deep Throat to intrepid Washington Post reporter Bob Woodward, is "Follow the money." Where are the presidential campaigns advertising and at what levels? We can discount the "token" buys and the low level of cable television advertising that the candidates are airing virtually nationwide.

Then we need to answer another set of questions. Where are the Media Fund and MoveOn.org, a pair of pro-Kerry "527" organizations, advertising? Where is America Coming Together, the new Democratic group focused on getting out the vote, doing its organizing? Which states do the top strategists for each campaign and for the 527s think will be close?

Clearly, the Bush and Kerry campaigns are practicing some gamesmanship in choosing where to advertise at this stage. Democrats would love to force Republicans into spending a bunch of money on advertising in Arkansas and Virginia, as well as in West Virginia. Leading Democrats realize that Kerry is very unlikely to carry any of those states, but if Democrats can spend one dollar to make the Bush campaign spend three dollars in a state that's a must-win for the GOP, that's a good investment for the Democrats.

The same goes for big Republican talk about New Jersey, where both sides are already buying access to the southern half of the state through their Philadelphia-based advertising. To reach the northern half of New Jersey, the Bush campaign team would have to buy ads in New York City's media market -- something it is quite unlikely to do. But the Kerry campaign would be delighted to see that happen.

Putting the partisan jockeying aside, by my count, the Republicans start off with a rock-solid base of 15 states that they could not lose even if they tried. Those states will give Bush a total of 110 Electoral College votes: Alabama (9), Alaska (3), Idaho (4), Indiana (11), Kansas (6), Mississippi (6), Montana (3), Nebraska (5), North Dakota (3), Oklahoma (7), South Carolina (8), South Dakota (3), Texas (34), Utah (5), and Wyoming (3).

The Democrats begin with a smaller base of six solid states, plus the District of Columbia, for a total of 71 electoral votes: Connecticut (7), the District of Columbia (3), Hawaii (4), Maryland (10), Massachusetts (12), New York (31), and Rhode Island (4).

Next come the six "Likely Bush" states, with 64 electoral votes. These states might see a little bit of action, but there's very little chance that any of them will shift into the Democratic column: Arkansas (6), Georgia (15), Kentucky (8), Louisiana (9), North Carolina (15), and Tennessee (11). Combining the 15 solid Bush states with the six likely Bush states gives the president a fairly reliable base of 174 electoral votes.

The "Likely Kerry" states have a total of 94 electoral votes: California (55), Illinois (21), New Jersey (15), and Vermont (3). Adding these likely Kerry states to the Democrats' six solid states, plus D.C., gives the Massachusetts senator a base of 165 electoral votes -- or nine fewer than Bush.

Only four states fall into the "Leaning Bush" column. They total 37 electoral votes: Arizona (10), Colorado (9), Virginia (13), and West Virginia (5). In all likelihood, all four will stay in the Republican column, bringing Bush's count to 25 states with 211 electoral votes, 59 short of victory.

Six states, totaling 63 electoral votes, are in the "Leaning Kerry" column: Delaware (3), Maine (4), Michigan (17), Oregon (7), Pennsylvania (21), and Washington (11). If all of these stay in the Democratic column, as expected, Kerry will have 16 states, plus the District, for a total of 228 Electoral votes, 42 short of victory.

That leaves nine states, with a combined 99 Electoral College votes, in the toss-up column: Florida (27), Iowa (7), Minnesota (10), Missouri (11), Nevada (5), New Hampshire (4), New Mexico (5), Ohio (20), and Wisconsin (10). These nine states, five of which Bush carried in 2000, and four of which Gore won, are the collective ground zero of this campaign.

For those Americans living in the 29 states that aren't being targeted by either party, the campaign will largely consist of what they read in the morning newspaper or watch on the nightly news. Probably they'll also see a few scattered ads on cable television.

For this major slice of the country, watching the campaign will be the equivalent of sitting in the last row of the end zone in a very big stadium: Voters there will be able to see the campaign, but none of the big plays will be very clear, and there won't be much excitement.

Voters living in the up-for-grabs "purple" states, on the other hand -- particularly residents of the nine or so states that are closest to the edge -- will be sitting smack on the 50-yard line, getting splattered, but clearly having the best seats in the house.