On the evening of October 29, 1991, meteorologists on the East Coast gazed into their computer monitors and observed three mighty weather systems, one a veritable hurricane, converging over the North Atlantic. No one had seen anything like it before. It was a natural event of such rarity and force that once experienced weatherman called it the Perfect Storm.

The equivalent of this phenomenon in American electoral history took place on the night of November 7, 2000. As the election returns for Congress showed the House and Senate moving towards parity between the two parties, the presidential race edged, eerily, towards an astonishing outcome: two candidates separated in the national popular tally by a few hundred thousand votes, an electoral college result that without the state of Florida produced no majority, and a popular-vote margin in Florida of under 2,000 ballots. It was the Perfect Tie.

Compared with the results of other general elections in American history, the contest in 2000 ended with the closest possible division of seats between the parties in the Senate (50 senators from each party), the fourth closest division of seats (by percentage) in the House, and the second closest presidential race by the electoral vote (excluding the 1800 election, when Thomas Jefferson and Aaron Burr tied). Observers of politics who emulate the meteorologists’ precision in measuring natural phenomena might consider a Closeness Index for American elections. If the differences in these three instances, expressed in percentage terms, are averaged together and subtracted from one, the 2000 election rates a remarkable score of 0.991. It is by far the closest election in American history.*

Political scientists in the middle of the last decade began to suggest that the general electoral condition of our era made a Perfect Tie conceivable. Even to think in such terms, it must be said, required a conceptual breakthrough in the study of elections. Historical patterns had led analysts to conclude that “normal” electoral alignments consisted of one party gaining a clear long-term advantage over the other, as the Democrats had managed to do for the two or three decades following the elections of 1932 and 1936. Electoral observers in the 1990s were therefore on the lookout for one of the parties to establish clear dominance, as some thought the Republican Party might have done after its vaunted “revolution” of 1994, and as other believed the Democrats might by on the verge of doing after Bill Clinton’s second “third way” victory in 1996.

But there was another way of viewing matters. Could not a situation of parity between the parties also be an alignment? After repeatedly observing in the middle of the 1990s that Democratic and Republican partisans were fairly evenly balanced in the electorate, a few analysts awoke to declare that this arrangement, far from being some kind of transitional phase, was the contemporary alignment, pure and simple. According to a well-known political scientist, Paul Allen Beck, this modern alignment consists of “a more even balance in electoral strength between the two major parties than the nation has ever seen, at least for a century.” Democrats are now about a third of the electorate (perhaps a bit more) and Republicans also about a third. The large remaining slice of the modern electoral pie contains the grouping of nonpartisan voters, also roughly a third of the electorate. This segment floats between the two parties and is easily attracted to independent candidates and third parties. Modern elections accordingly display a good deal of volatility from one contest to the next. Each outcome, to cite Beck again, reflects “election specific movements of a large nonpartisan electorate around a 50-50 ‘normal vote’ baseline.”

* * *

This does not mean that the floating vote is entirely without structure or direction. Although a portion of the vote in every election is predicated on responses to evanescent moods and an evaluation of personality and leadership traits, it is still possible to discern the outlines of what the floating voters appear to want, or more precisely, what they do not want. The floating vote, understood here to include most of the nonpartisans as well as a portion of weaker partisans, has made known its dislikes at least twice in the last decade: after 1992-93, when it reacted against a strong Big Government agenda and a cultural politics of permissiveness from the Left; and after 1994 and 1995, when it reacted against a strong anti-government agenda and a strident cultural-war rhetoric from the Right. Floating voters have therefore previously rejected the intellectual packages and world views that are embraced by the strong partisans of both parties. Successful presidential candidates have been those who have found a way to maintain the enthusiasm and support of their core supporters, while managing to appeal to a larger portion of this floating vote.

Now imagine an election in which almost all of the partisans stick by the candidate of their own party and in which the short-term factors pressure the floating voters equally in different and ultimately offsetting directions. Given the basic shape of the modern electorate, the result would approach parity. This is just about what happened in 2000. Party adherents voted overwhelmingly for “their” candidates, a result that probably had less to do with any extraordinary appeal of Gore and Bush among their own supporters than with their inability to generate much cross-partisan interest. Whatever the cause, partisan faithfulness in 2000 reached the highest level for any modern election and netted each candidate a roughly equal number of votes by partisans. Next, the net effect of short-term forces split the nonpartisan or floating voters down the middle. In some instances, these forces pulled and tugged the same people in different directions, causing cross-pressured voters, by far the greatest number of whom were women, to swing greatly back and forth between candidates.

It is also important, of course, to consider how people vote for the three major political offices at the state level: governor, the state Senate, and the state House. The results here are striking. For the state legislative chambers, Democrats in 2001 control 25 state lower houses to 24 for the Republicans, while Republicans control 24 state Senates to 21 for the Democrats. Four Senates are tied between the two parties. There is, however, one office today where one of the parties holds a clear advantage. Among the governors, 29 are Republican and 19 are Democrats. And therein lies a tale. It is surely no accident that the President, George W. Bush, not only came from the ranks of the Republican Governors, but also was nominated with their support and campaigned for the presidency, conspicuously, with their help.

At the beginning of the decade, the Republican Party was predominantly a presidential party. Republicans held the presidency (as they had since 1980), but both houses of Congress were in the hands of the Democrats. In the states, the Democrats had the edge in Governors (28D-20R) and the decided advantage among state Senates (34-11) and state Houses (40-9). By 1994, the contours of partisan control had begun to change. Democrats now held the presidency, while Republican captured the House and Senate. Republicans in 1994 gained their large advantage among governors (30R-19D) and pulled almost even among state Senates (25D-24R) and close among state Houses (26D-22R).

The elections of the 1990s therefore did indeed realign American politics. But the did so not by placing a single party in control, but by bringing one of the parties (the Republican) from an overall minority status into a position of parity. A large part of this has been the transformation of Southern politics, in which the change from Democratic to Republican voting that showed earlier in presidential races worked its way down to the lower levels of office holding.

A statistician might object that these numbers tell us which party won an office or institution, but not by how much. But if one were to construct an index, weighted by population, in order to show how well each party has done in each state for all six of the major elective offices (President, Senate, House, Governor, state Senate, state House) the results would show a nation that at the outset of the decade was more Democratic than Republican, but that in 1994 moved into and has since remained in a Perfect Tie.

* * *

The 2000 election was unusual, even unprecedented, in that all three of the elected national institutions—the presidency, the House, and the Senate—were considered to be “in play” for both parties during the campaign. No other general campaign in American history, certainly none in modern times, has offered this range of choice.

The election results confirmed that any of these combinations was possible. The outcome for each contest could just as easily have gone to the other party. The presidency was decided by a chad; the Senate ended with a tie; and the House left the two parties only a few seats apart. If there was any deviation from the most commonly expected results, it was in a reversal of the expectations for the Democrats’ performance in the elections for the two houses of Congress. The results in the House of Representatives were a big of a disappointment for Democrats. The 2000 election was the third contest since the Republican Revolution of 1994, and in each on the Democrats chipped away at the Republican majority. The Democrats gained two seats in 1996, six seats in 1998, but only one in 2000, falling short of their strong hopes of winning a majority. The Senate outcome moved in the other direction, with Democrats doing very well, picking up a net of four seats. This result overcame part of the huge Republican surge of 1994, when the Republicans gained eleven seats. Three Republican incumbents from that class—Spencer Abraham of Michigan, Rod Grams of Minnesota, and John Ashcroft of Missouri—went down to defeat.

Although a single partisan majority was attainable for all three institutions, voters showed little sign during the campaign that they were poised to give, in one fell swoop, a single national majority to either party. This separation of the presidential and congressional races resulted in part from a strategic choice by both candidates. Nether one, especially George W. Bush who was running from “outside” of Washington and who offered himself as a pragmatic candidate from the gubernatorial wing of his party, wanted to tie himself too closely to the congressional wing of his party. The separation was strikingly illustrated in the candidates’ convention speeches, in which neither man asked for a party majority in Congress or used his party label to rally a national majority. To be fair, the presidential candidates did not totally run away from their congressional running mates during the campaign. They offered modest support where it was possible. But no national, partisan-related theme was trumpeted simultaneously by the presidential and congressional candidates as it had been in 1992, when Bill Clinton urged voters to vote for him in order to end “divided government,” or in 1996, when Bob Dole’s campaign initially urged voters to vote for him in order to “finish the job” begun by the Republican Congress in 1994.

Political conditions in 2000 made it difficult to connect the presidential and congressional races, for the simple reason that no one could know for certain which party would win the majority in any of the contests. The striking point here is that foreknowledge of which party will control which of the institutions is now a major factor influencing voter behavior. This fact became really apparent in the light of an extraordinary change in voter attitude that began (or was first noticed) during the 1996 election, when a larger part of the electorate turned against the notion of entrusting all branches of government to the same political party. Instead of voters beginning with a natural bias of voting for all candidates for national office from the same party, which was the premise of the “party government” or “mandate democracy” theory made so popular by Woodrow Wilson, it now appears that a known majority in one institution might, on the contrary, provoke “repulse” voting—effectively denying that part control of another branch. Voters, or a significant portion of them, now seem to prefer divided government.

One of the few to take a stab at connecting the presidential and congressional voting in this way in 2000 was Bill Clinton, the master of repulsion politics. In one of his few foras onto the stump for Al Gore, he declared: “I think we’re going to win the House and the Senate. But if we don’t, someone needs to be doing what I’ve done for the last six years, which is to stop extremism in Washington, D.C., and you certainly only have one choice—Al Gore.” But any such outcome was nearly impossible for voters to count on in 2000, for with the presidential race in constant flux, no judgment about congressional control could reliably be made.

A political contest in which congressional elections are nationalized and strongly connected to the presidential race is more likely to occur when the Congress itself is prominent in the public mind. Congress certainly held such a position in 1996 when Newt Gingrich, locked in an ongoing personal duel with Bill Clinton, introduced a dramatic nationalizing and polarizing effect into congressional elections. Though Bill Clinton’s shadow still loomed over the 2000 election, Newt Gingrich’s was largely absent. Republicans as a collective entity in Congress managed to assume a far lower profile than in the elections of 1994, 1996, or 1998, and in the person of Speaker Dennis Hastert, they had found the perfect invisible man.

The Supreme Court, though not an elected branch, was as much “in play” in the 2000 elections as any of the elected institutions; and it, too, was a house divided. Although majorities on the Court vary from issue to issue, in recent years there have been fairly consistent conservative and liberal blocs, with a couple of justices swinging between them. Five to four votes, generally favoring the conservatives, have been a frequent occurance. Not surprisingly, therefore, the Court’s future majority became an important issue in 2000. As Al Gore pointed out in a speech during the last week of the campaign in Michigan: “The Supreme Court is at stake. There are going to be three, maybe four justices of the Supreme Court appointed by the next president of the United States. That means a majority on the court that will interpret our rights under the Constitution for next 30 to 40 years.” The Court, Gore was saying, would follow the results of the election. Never did he imagine that it would actually decide the results of the election.

* * *

One consequence of the election is unmistakable: there is no popular mandate of any kind. George W. Bush received nothing more than a key to his office. Neither party could make a plausible claim to be unambiguously in power. Spokespersons of the two parties were strangely subdued, and few on either side even attempted to argue for any kind of grand electoral victory. All they laid claim to were the races they had actually won—nothing less, yet also nothing more.

Americans awoke on November 8 to the realization that the nation had entered a period of unprecedented political equilibrium. Since 1994 the electorate had achieved political balance by distributing different majorities among the different national institutions. But until last November, leaders of both parties had some reason to think (or publicly claim) that, except for some temporary and inhibiting factor, their party owned the “real” advantage. In 2000 the party leaders acknowledged that for now, neither party is ahead.

Political leaders today are adjusting to this unusual situation of politics without a mandate. Neither party has been able to win an outright majority of the votes cast in a presidential election since 1998. Far from constituting a crisis, however, this situation may offer some new possibilities for governance. With politicians operating without the overhang of a mandate, they are free to consider patterns of conduct unrelated to an extra-constitutional source of authority. Depending on what techniques and tone President Bush adopts, he could reintroduce institutional models, both for the presidency and for dealing with Congress, that have a closer affinity to the spirit of constitutional government.

The politics of the perfect tie take place, paradoxically, under a situation of unitary government, with each branch in the hands of the Republicans. This situation represents, in one sense, a high watermark of Republican influence in the past half-century. But while unitary government describes the formal situation of national politics, it does not capture the reality. Republicans clearly have no reason to apologize for their precarious “control” of the government, but the truth is that the November results have nothing in common with other recent moments of achieving unitary government (1964, 1976, and 1992), when at a minimum a claim of a “party” victory could plausibly be entertained. In 2000, the Republicans took control of the government after losing seats in both houses of Congress and losing the popular vote for president.

How much the Republican Party will actually be judged by the public to be running the country over the next two years is an interesting question. Robert Samuelson best stated the conventional wisdom when he observed, “Bush and the Republicans will have enough power to be blamed for anything that goes wrong—but not enough to accomplish much.” It is traditional, in any case, for the president’s party to be held accountable in mid-term elections, regardless of whether or not it is in the majority. But there is reason now to wonder how much the traditional rules for parceling out party credit and blame in elections still hold in American politics, especially when so many voters have begun to question the idea of mandate democracy.

What do these development mean for President Bush’s ability to govern? At this point, three different scenarios or models seem possible.

One scenario foresees a period of intense ad open partisan warfare in which the Democrats, acting as the out party, find it in their interest to obstruct major governmental action, appealing to public opinion for vindication in the next election. The general idea is that if little by way of legislation is accomplished, it will rebound to the benefit of the out party. Obstruction would, of course, be concealed under claims that the president was not offering bipartisan proposals, a standard Democratic leadership would solely define. This strategy might also encourage questioning of the Bush presidency’s legitimacy, which in turn would lend further justification to an “outside” appeal to the people in a political referendum in 2002. This approach would parallel the tactics adopted by Andrew Jackson’s followers during John Quincy Adam’s presidency and by some Democrats in the aftermath of the disputed election of Rutherford B. Hayes in 1976.

A second scenario envisages a substantial degree of bipartisan cooperation and legislative action. President Bush would work largely with and through the established congressional leadership of the parties, negotiating in particular with the official leaders of the opposition, who would be authorized to speak for their party. Democrats would cooperate based on a calculation that voters are not in a mood to condone partisanship and would penalize either part if it were perceived as too obstructionist. Faced with this kind of uncertainty, Democrats might think that they have as much to gain as Republicans from a mutual record of legislative accomplishment.

A third scenario foresees “underpartisan” coalitions that accomplish modest legislative action through a series of tactical majorities combining Republicans and various Democrats, depending on the bill being debated. The president would deal here through individual members of Congress, especially certain sympathetic Democrats, rather than through the Democratic party leadership. This scenario reminds us that although everyone in Congress is a member of a party, in fact much of the action in Congress stems from individuals considering the interests of their own constituencies and their own particular electoral situation. Different coalitions can be forged on different issues, often taking place underneath official party leadership. Bill Clinton employed this strategy after 1996, often securing the backing of “pressured” Republicans from constituencies where the President had strong support.

* * *

Events like a perfect storm or a perfect tie possess a terrible beauty that can remind people of the tremendous power of nature and of fate. Just as the Perfect Storm swept aside the structures man built to shield himself from the environment, so did the Perfect Tie overwhelm the institutions that were designed, in the words of The Federalist Papers, to prevent “tumult and disorder” and to produce a legitimate presidential victor. One by one, parts of the electoral system teetered and gave way. Trouble began on election night with the collapse of the television networks’ vote projection system, which is relied on today, informally, to declare presidential winners. It spread next to disputes about the accuracy and reliability of election and balloting machinery in Florida, and by implication, the nation as a whole. It then finally moved on to confusion over Florida’s election laws and, more ominously, to controversy over the rightful powers of state judicial and political institutions.

The problems that began on November 7 seemed to most Americans to be at first the stuff of history books, drawn from a more primitive political era. Journalists found themselves searching for accounts of the last disputed presidential election, when a close national election result in 1976 led to charges of widespread electoral fraud and a revisiting of the votes in a few of the reconstruction governments in the Southern states, including Florida. Charges of a similar kind were raised in 2000, but most people realized that fraud in the literal sense was not really the source of problems in this election. Instead, it was the exceptional way the vote was distributed around the country, couple with an unexpected technical challenge that by chance occurred in the state of Florida, where the difference in vote totals between the two main candidates was smaller than the margin of error that the electoral mechanisms could handle. It was more than Florida had planned for, more perhaps than it might reasonably have been expected to plan for.

In the end, the nation’s political system held up. Or at least it help up in the sense that a president of the United States, George W. Bush, was finally chosen, albeit more than five weeks after election day. Civil unrest and violence were also avoided, owing in part, it would seem, to the good fortune of an election that that had taken place in times of plenty, and in an era lacking highly divisive issues. But the stress placed on the political system was evident. It took a highly controversial decision by the United States Supreme Court (determined by a five to four vote) overruling a highly controversial decision by the Florida Supreme Court (determined by a four to three vote) to assure George Bush his election. Bush assumed the presidency with the formal blessings of almost everyone in the political class, and with general support from the American people. But the very fact that issues of legitimacy were raised meant that his title to the office fell short of being universally acknowledged.

Even as the Hollywood movie The Perfect Storm devoted a good deal of attention to the rescue efforts that attended the calamity, so in this case citizens should ponder how different political institutions and actors responded to the extraordinary challenges they confronted after the election. These questions go well beyond what is usually covered in electoral studies and extend into realms that assess prudence and statesmanship. Here, alas, political science has developed no index to assist us.

* * *

*The next three closest elections were: 1876 with a score of 0.955, 1916 with a score of 0.939, and 1880 with a score of 0.921. The formula for creating the “closeness index” is: Index Value=1-((positive percent difference between the two largest parties in the House)+(positive percent difference between the two largest parties in the Senate)+(positive percent difference between the two largest parties in the electoral college)/3).