Jack Rakove, in the SF Chronicle (Sept. 5, 2004):

Scholars use the term critical election to describe those contests that crystallize broad changes in the electorate, when voting blocs shift their loyalty from one party to another, new political alignments form, and one party achieves dominance over the other. There was a time, not so long ago, when Bush political adviser Karl Rove (who knows his history) was hoping that the 2004 election would prove to be critical in just this way.

It would finally cement the comfortable majority that Republicans have believed to be tantalizingly in reach since Ronald Reagan captured the presidency a quarter-century ago.

The other"Mission Accomplished" the banner on board the aircraft carrier Abraham Lincoln was subliminally conceived to herald was when President Bush prematurely celebrated the end of active fighting in Iraq. (And was it only chance, Michael Moore might wonder, that made a ship named Lincoln the site for this ultimate media event, linking the most recent Republican president with the party's first winner, even though the GOP's current homeland lies in the ex-Confederate states that Lincoln forced to remain in the Union?)

At this point in the campaign, Rove and other Republicans will be happy to snatch any victory they can, leaving a lasting realignment of the national electorate to another day and another candidate.

The 2004 election will still be a critical one -- but for two other reasons than those scholars invoke.

First, with less than two months to go, it seems likely to perpetuate the same bitter aftertaste that has soured our politics the past four years. That aftertaste is likely to linger longer in Washington than elsewhere in the country. An election in which the current administration narrowly retains power, or in which John Kerry captures the White House but Republicans retain control of Congress, is likely to exacerbate the pungent acrimony that permeates the national government.

There is a second, more obvious sense in which this election appears critical. Even if we recognize that no president can assure our security against the limitless threats we now confront, profound questions remain about the kind of security policy we ought to pursue in the post-Sept. 11 world we are doomed to inhabit indefinitely. This election is critical for the simple reason that it will determine who will manage the struggle against terror during the next four years.

Of course, a democratic election may not be the best means to determine which contender is better qualified to discharge this responsibility. The fact that each candidate can be portrayed in a heroic pose -- the president rallying the workers at Ground Zero, the senator on his swift boat in Vietnam -- contributes nothing to an informed choice. And if the campaign seriously addressed the likelihood that a nuclear device will be detonated in an American city in the next five years, we might all be too depressed to go to the polls anyway.

Perhaps it is better, then, that the campaign already seems to be developing its own bipolar dynamic. On the one hand, both candidates are trying to stake out basic, if rather obvious, positions on terrorism and national security. But out on the campaign trail, where their paths keep crossing, other seemingly conventional issues come to the fore: employment, the deficit, outsourcing, prescription drug benefits.

There are two basic explanations for this oscillation between the urgent and the mundane. One is that whatever color of the day homeland security czar Tom Ridge may choose, we cannot live our lives in a perpetual state of anxiety about terrorism. The desire for normalcy (that favorite word of Warren G. Harding, who succeeded a Democratic president, Woodrow Wilson, who cared too much about foreign policy) naturally reasserts itself. And so it should. Until the next catastrophe occurs and we reluctantly appoint another commission to examine our omissions, we should not be surprised that Americans remain concerned with those other forms of domestic security we traditionally know as employment, health, and education.

The second explanation, though, owes something to the peculiar way in which we elect our presidents. If our presidential elections were truly national, terrorism and the war in Iraq would be the decisive issues on which the campaign would be fought. But as we learned -- or at least were reminded four years ago -- we don't really have a national election, even for president. Instead, we have, in theory, 51 campaigns. But because the outcome in roughly two-thirds of the states is already a foregone conclusion, the contest will probably take place in a baker's dozen or so states.

At the moment, the critical states on which the election will pivot form an inverted U that circles solidly Republican Indiana and solidly Democratic Illinois. It begins in Ohio, swings north and west through Michigan and Minnesota, and heads south through Wisconsin into Iowa and Missouri. It seems inevitable that both campaigns will crisscross this region frenetically from now until the cows of Iowa come home in November. This is also the old Rust Belt, where the loss of manufacturing jobs remains a pressing issue that should give Kerry a further advantage.

In these closely divided states, issues of national security will doubtless have bite. But so will all the other concerns that the parties calculate are most likely either to turn out their supporters or sway undecided voters whose very existence leaves partisans in both parties puzzled.

Thus this campaign seems destined to have a surreal quality. The election is arguably as critical a vote as we have cast for decades -- at least since 1968, or even going back to the 1940s. ...