Karl Rove in a Corner

Karl Rove is at his most formidable when running close races, and his skills would be notable even if he used no extreme methods. But he does use them. His campaign history shows his willingness, when challenged, to employ savage tactics

It is the close races that establish the reputations of great political strategists, and few have ever been closer than the 2000 presidential election. From the tumult of the lengthy recount, the absentee-ballot dispute, the charges of voter fraud, and, ultimately, the Supreme Court decision, George W. Bush emerged victorious by a margin of 537 votes in Florida—enough to elevate him to the presidency, and his chief strategist, Karl Rove, to the status of legend.

But the 2000 election was not Rove's closest race. That had come earlier, and serves as a greater testament to his skill. In 1994 a group called the Business Council of Alabama appealed to Rove to help run a slate of Republican candidates for the state supreme court. This would not have seemed a plum assignment to most consultants. No Republican had been elected to that court in more than a century. But the council was hopeful, in large part because Rove had faced precisely this scenario in Texas several years before, and had managed to get elected, in rapid succession, a Republican chief justice and a number of associate justices, and was well on his way to turning an all-Democratic court all Republican. Rove took the job.

The most important candidate among the four he would run that year was a retired judge and Alabama institution by the name of Perry O. Hooper, of whom it is still fondly remarked that in the lean years before Rove arrived he practically constituted the state's Republican Party by himself. A courtly man with an ornery streak and a stately head of white hair, Hooper seemed typecast for the role of southern chief justice, a role he hoped to wrest from the popular Democratic incumbent, Ernest "Sonny" Hornsby.

At the time, judicial races in Alabama were customarily low-key affairs. "Campaigning" tended to entail little more than presenting one's qualifications at a meeting of the bar association, and because the state was so staunchly Democratic, sometimes not even that much was required. It was not uncommon for a judge to step down before the end of his term and handpick a successor, who then ran unopposed.

All that changed in 1994. Rove brought to Alabama a formula, honed in Texas, for winning judicial races. It involved demonizing Democrats as pawns of the plaintiffs' bar and stoking populist resentment with tales of outrageous verdicts. At Rove's behest, Hooper and his fellow Republican candidates focused relentlessly on a single case involving an Alabama doctor from the richest part of the state who had sued BMW after discovering that, prior to delivery, his new car had been damaged by acid rain and repainted, diminishing its value. After a trial revealed this practice to be widespread, a jury slapped the automaker with $4 million in punitive damages. "It was the poster-child case of outrageous verdicts," says Bill Smith, a political consultant who got his start working for Rove on these and other Alabama races. "Karl figured out the vocabulary on the BMW case and others like it that point out not just liberal behavior but outrageous decisions that make you mad as hell."

Throughout the summer the Republican candidates barnstormed the state, invoking the decision at every stop as an example of "jackpot justice" perpetuated by "wealthy personal-injury trial lawyers"—phrases developed by Rove that have since been widely adopted. To channel anger over such verdicts toward the incumbent Democratic justices, Rove highlighted their long-standing practice of soliciting campaign donations from trial lawyers—just as Republicans (which Rove did not say) solicit them from business interests. One particularly damaging ad run by the Hooper campaign was a fictionalized scene featuring a lawyer receiving an unwanted telephone solicitation from an unseen Chief Justice Hornsby, before whom, viewers were given to understand, the lawyer had a case pending. The ad, and the unseemly practices on which it was based, drew national attention from Tom Brokaw and NBC's Nightly News.

The attacks began to have the desired effect. Judicial races that no one had expected to be competitive suddenly narrowed, and media attention—especially to Hooper's race after the "dialing for dollars" ad—became widespread. Then Rove turned up the heat. "There was a whole barrage of negative attacks that came in the last two weeks of our campaign," says Joe Perkins, who managed Hornsby's campaign along with those of the other Democrats Rove was working against. "In our polling I sensed a movement and warned our clients."

Newspaper coverage on November 9, the morning after the election, focused on the Republican Fob James's upset of the Democratic Governor Jim Folsom. But another drama was rapidly unfolding. In the race for chief justice, which had been neck and neck the evening before, Hooper awoke to discover himself trailing by 698 votes. Throughout the day ballots trickled in from remote corners of the state, until at last an unofficial tally showed that Rove's client had lost—by 304 votes. Hornsby's campaign declared victory.

Rove had other plans, and immediately moved for a recount. "Karl called the next morning," says a former Rove staffer. "He said, 'We came real close. You guys did a great job. But now we really need to rally around Perry Hooper. We've got a real good shot at this, but we need to win over the people of Alabama.'" Rove explained how this was to be done. "Our role was to try to keep people motivated about Perry Hooper's election," the staffer continued, "and then to undermine the other side's support by casting them as liars, cheaters, stealers, immoral—all of that." (Rove did not respond to requests for an interview for this article.)

The campaign quickly obtained a restraining order to preserve the ballots. Then the tactical battle began. Rather than focus on a handful of Republican counties that might yield extra votes, Rove dispatched campaign staffers and hired investigators to every county to observe the counting and turn up evidence of fraud. In one county a probate judge was discovered to have erroneously excluded 100 votes for Hooper. Voting machines in two others had failed to count all the returns. Mindful of public opinion, according to staffers, the campaign spread tales of poll watchers threatened with arrest; probate judges locking themselves in their offices and refusing to admit campaign workers; votes being cast in absentia for comatose nursing-home patients; and Democrats caught in a cemetery writing down the names of the dead in order to put them on absentee ballots.

As the recount progressed, the margin continued to narrow. Three days after the election Hooper held a press conference to drive home the idea that the election was being stolen. He declared, "We have endured lies in this campaign, but I'll be damned if I will accept outright thievery." The recount stretched on, and Hooper's campaign continued to chip away at Hornsby's lead. By November 21 one tally had it at nine votes.

The race came down to a dispute over absentee ballots. Hornsby's campaign fought to include approximately 2,000 late-arriving ballots that had been excluded because they weren't notarized or witnessed, as required by law. Also mindful of public relations, the Hornsby campaign brought forward a man who claimed that the absentee ballot of his son, overseas in the military, was in danger of being disallowed. The matter wound up in court. "The last marching order we had from Karl," says a former employee, "was 'Make sure you continue to talk this up. The only way we're going to be successful is if the Alabama public continues to care about it.'"

Initially, things looked grim for Hooper. A circuit-court judge ruled that the absentee ballots should be counted, reasoning that voters' intent was the issue, and that by merely signing them, those who had cast them had "substantially complied" with the law. Hooper's lawyers appealed to a federal court. By Thanksgiving his campaign believed he was ahead—but also believed that the disputed absentee ballots, from heavily Democratic counties, would cost him the election. The campaign went so far as to sue every probate judge, circuit clerk, and sheriff in the state, alleging discrimination. Hooper continued to hold rallies throughout it all. On his behalf the business community bought ads in newspapers across the state that said, "They steal elections they don't like." Public opinion began tilting toward him.

The recount stretched into the following year. On Inauguration Day both candidates appeared for the ceremonies. By March the all-Democratic Alabama Supreme Court had ordered that the absentee ballots be counted. By April the matter was before the Eleventh Federal Circuit Court. The byzantine legal maneuvering continued for months. In mid-October a federal appeals-court judge finally ruled that the ballots could not be counted, and ordered the secretary of state to certify Hooper as the winner—only to have Hornsby's legal team appeal to the U.S. Supreme Court, which temporarily stayed the case. By now the recount had dragged on for almost a year.

When I went to visit Hooper, not long ago, we sat in the parlor of his Montgomery home as he described the denouement of Karl Rove's closest race. "On the afternoon of October the nineteenth," Hooper recalled, "I was in the back yard planting five hundred pink sweet Williams in my wife's garden, and she hollered out the back door, 'Your secretary just called—the Supreme Court just made a ruling that you're the chief justice of the Alabama Supreme Court!'" In the final tally he had prevailed by just 262 votes. Hooper smiled broadly and handed me a large photo of his swearing-in ceremony the next day. "That Karl Rove was a very impressive fellow," he said.

In the decade since, the recount and the court battle have faded into obscurity, save for one brief period, late in 2000, when they suddenly became relevant again. Almost as if to remind Al Gore's campaign of Rove's skill when faced with a recount, the case was revived in a flurry of legal briefs in the Supreme Court case of Bush v. Gore—including one filed by the State of Alabama on behalf of George W. Bush.

This summer, with the presidential race looking as if it would be every bit as close as the one in 2000, I spent several months examining the narrowest races in Karl Rove's career to better understand the tendencies and tactics of the man who will arguably have more influence than anyone else over how this election unfolds. Rove has already generated a remarkable body of literature, including several notable books and numerous magazine and newspaper articles. I spoke to many of Rove's former candidates and their opponents; to his past and present colleagues and the people who faced off against them; and to political insiders and journalists—primarily in Texas and Alabama, where Rove has done the majority of his campaign work. I learned much about Rove that hasn't made it into the public sphere.

One of the striking things about his record is how few close races Rove has been involved with—primarily because he usually wins in a walk. In the relatively rare instances when he is in a tight race, he tends to win that, too. Although Rove first rose to political prominence as a specialist in direct-mail fundraising (and worked on hundreds of races in that capacity), mail is only one facet of a campaign, and rarely the deciding factor. So I focused on races in which Rove was the primary strategist, and therefore in a position at least roughly analogous to the one he holds in this presidential race. The last strategist before Rove to win a Republican presidential election was his former colleague Lee Atwater, who by the time of the 1988 campaign had a career record of 28—4. To my knowledge, no one has calculated such a figure for Rove. As far as I can determine, in races he has run for statewide or national office or Congress, starting in 1986, Rove's career record is a truly impressive 34—7.

The mythologizing portrayals of a "boy genius" that characterized so much media coverage of Rove after 2000, and especially after the Republicans' triumphant sweep in the midterm elections, struck me as sorely out of date when I began this project. The Bush Administration was suffering through the worst of the fallout from the Abu Ghraib scandal, and the President's approval ratings were plummeting. Clearly, there are many differences between the circumstances in which Rove has been victorious in the past and those he faces now. But that is no reason to discount his record. By any standard he is an extremely talented political strategist whose skill at understanding how to run campaigns and motivate voters would be impressive even if he used no extreme tactics. But he does use them. Anyone who takes an honest look at his history will come away awed by Rove's power, when challenged, to draw on an animal ferocity that far exceeds the chest-thumping bravado common to professional political operatives. Having studied what happens when Karl Rove is cornered, I came away with two overriding impressions. One was a new appreciation for his mastery of campaigning. The other was astonishment at the degree to which, despite all that's been written about him, Rove's fiercest tendencies have been elided in national media coverage.

Democrats who want to feel sanguine about the coming election might well find comfort in the particulars of Rove's career. Several of his usual advantages are lacking this time around, conspicuously in geography. As a direct-mail consultant, Rove worked for races across the country, in blue states as well as red. The nature of that work mostly entailed identifying conservatives and motivating them to donate money—a fine skill for one in his current position as Bush's chief strategist, but not the equivalent of running a campaign. Rove compiled his stellar record in Texas and Alabama—and, of course, in the 2000 presidential election, even if his candidate lost the popular vote. During the period in which he rose to power, both states, deeply conservative, were transitioning from a firmly Democratic electorate to a firmly Republican one. A charge frequently levied against Rove by beleaguered Democratic consultants in Texas and Alabama is that he merely "surfed the wave" of the demographic change. This ignores his political talent. It's true, though, that for most of his career Rove has enjoyed a kind of home-field advantage, and in this election he does not.

A surprising number of Rove's former colleagues believe that his unprecedented success in Texas, where for years his candidates rarely faced serious challenges, has fostered what in the boxing world would be known as a "tomato-can" syndrome. Like a heavyweight champion who lets down his guard after beating up a series of hapless "tomato-can" opponents, Rove, they fear, may have been blinded to current national realities by hubris. "I think Karl's success in Texas is almost a hindrance," a veteran strategist who worked with him in that state told me. "The rest of the country doesn't emulate Texas in terms of voting behavior. But sometimes you see his southern roots in Texas and his experience in Alabama kind of overtake him, and he seems to think the United States is one big-ass Texas."

Several consultants pointed to the issue of gay marriage, which one described as a perfect Texas wedge issue because it would attract culturally conservative Democrats in the eastern part of the state—"the rednecks," as he put it—who are normally the key to winning statewide office. But he doubted that the issue would have the same effect in the less conservative battleground states that are expected to decide this election.

Rove is also riding on less of a decisive financial advantage than the one he normally enjoys. In their book Bush's Brain, James Moore and Wayne Slater explain how Rove's success as a fundraiser provided the impetus for his move into political consulting, and how, once established in that capacity, he consolidated his power by controlling candidates' access to major donors, usually ensuring that his clients were better funded than their opponents. This enabled him to engage in what amounted to asymmetric warfare against anyone who challenged his candidates. The authors recount an anecdote in which Priscilla Owen—then a Houston judge, later a controversial Bush appointee to the federal bench—approached a rich Republican donor whose job it was to vet candidates, and explained that she was thinking about running for the Texas Supreme Court. "Have you talked to Karl Rove?" he inquired. Taking the hint, she replied, "No, but I plan to." After Rove agreed to support her, she won handily, outspending her opponent. A similar imbalance applied in 2000, when Bush outspent Gore by a wide margin. But this year John Kerry's extraordinary and unexpected ability to raise money has largely closed the gap.

It will come as no surprise to anyone who has paid attention to the current campaign that Rove's most notable tendency in close races has been to go negative against his opponent, early and often. One of the first highlights of his career was the famously tight 1986 Texas governor's race, in which his candidate and mentor, the Republican oilman Bill Clements, sought to oust the Democratic incumbent Mark White. The race is legendary in Texas political lore for Rove's discovery that his office was bugged—news of which, coincidentally or not, distracted attention from an evening debate in which his candidate was expected to fare poorly. More pertinent to the current campaign is a strategy memo Rove wrote for his client prior to the race, which is now filed among Clements's papers in the Texas A&M University library. Quoting Napoleon, the memo says, "The whole art of war consists in a well-reasoned and extremely circumspect defensive, followed by rapid and audacious attack."

Though it is forever fashionable to denounce negative campaigning, every political expert understands that it can be extremely effective. Rove's career has borne this out perhaps better than any other modern political consultant's. But his very success leaves him precariously positioned if Bush stalls or founders. Once a negative course is set, it is nearly impossible to change; the perpetrator is usually stained for good. Furthermore, Rove's method is to plot out elaborate strategies well in advance of the campaign, and stick to them vigilantly. John Deardourff, Rove's media consultant for races in Texas and Alabama, says, "This rap Bush has of never changing his mind and never admitting a mistake—that's Karl! That's where it comes from." It is a tribute to Rove's strategic skill that he is so often right.

Throughout his career Rove has been able to stage-manage races to an extraordinary degree. This is possibly his least appreciated skill. The most revealing time in his career was 1994, when Rove fought more close races than in any other year, and managed to dictate the dynamic in every one of them. He pulled off highly unlikely upsets for Perry Hooper in Alabama (a race overwhelmingly about trial lawyer excesses) and George W. Bush in Texas (a race dominated by Bush's platform of welfare, juvenile-justice, tort, and public-school reform). However impressive, all but one of his races have been conducted at the state level, and thus have been comparatively insular affairs, unimpeded by the glare of the national media or a troublesome global issue like violence in Iraq—both of which could threaten Rove's ability to control this race.

In the rare instances when he has failed to set the terms of debate, Rove hasn't fared nearly so well. Four years ago, in a race to succeed Hooper, who was retiring as Alabama's chief justice, Rove lined up support from a majority of the state's important Republicans behind his candidate, an associate justice named Harold See. Like most of Rove's clients, See had an enormous financial advantage and ran a brutally negative campaign—but he was nonetheless trounced by Roy Moore, the "Ten Commandments" judge, who succeeded in making the race about religion. This loss may have helped Rove to recognize the power of religion as a political motivator: from the question of gay marriage to organizing churches for Bush, it features prominently in his playbook for the current election.

If there is any compelling reason to think that Rove may be out of his depth in this election, it is an odd lacuna in his storied career: no one I spoke with could recall his ever having to run an incumbent in a tough re-election race. This is partly a by-product of his dominance. Rove's power in Texas was such that he could essentially handpick his candidates, and once elected, they rarely lost. And he spent most of his career in the favorable terrain of the Deep South. One reason Rove was spared re-election fights is that as demographic changes swept across the South, and Republicans in Texas and Alabama began displacing Democrats, the likelihood that a Democrat could depose a sitting Republican became remote. Rove has long excelled at knocking off incumbents in tight races. Now, at last, he must defend one.

Despite all this, there are significant reasons to believe that Rove can pull it off this time. One is his prior experience in close races. Another is his preparedness and attention to detail, to which any discussion with a longtime Rove colleague invariably turns. "The thing that was most important to him was the mechanics: making certain that the campaign could block and tackle," recalls a staffer who worked for Rove's direct-mail firm in the 1980s and 1990s. Rove would typically begin a race by constructing seven-layer spreadsheets of the electoral history of a particular office, charting where votes for each candidate had originated and which groups had supplied them. In the 1980s these data led Rove to conclude that his candidates ought to target "ticket-splitters"—Texans who supported Ronald Reagan for President but voted Democratic in downballot races.

Rove's direct-mail experience had provided him with a nuanced understanding of precisely what motivates ticket-splitters. According to Karl Rove & Co. data on the 1994 Texas governor's race, Rove was aware, for instance, that households that received a single piece of mail turned out for Bush at a rate of 15.45 percent, and those that received three pieces at a rate of 50.83 percent. Turnout peaked at seven pieces (57.88 percent), after which enthusiasm for Bush presumably gave way to feelings of inundation, and support began to drop.

Rove's thirst for efficient advantage extended even to marketing. According to a former employee, rather than use costly dinners and Dallas Cowboys tickets to draw clients' attention, as other consultants did, Rove affixed antique stamps (though not valuable ones) to the weekly financial summaries he mailed to clients; he would send workers to estate sales to hunt out supplies.

When Rove arrived in Alabama, in 1994, his clients were initially puzzled as to why he was having them campaign in rural and less populated parts of the state rather than the urban areas they were accustomed to. It turned out that he had run an electoral regression analysis on each of the state's sixty-seven counties, and for efficiency's sake he put his four judicial candidates together on a bus trip to the counties with the highest percentage of ticket-splitters. "Karl got us focused on the fact that it was a matter of convincing Democratic voters who were already conservative to vote for Republican candidates," Mark Montiel, a candidate on the trip, explains, "because that was who best expressed their views."

Among Rove's other innovations was a savvy use of language, developed for speaking to the conservative base about judicial races. Candidates were to attack "liberal activist judges" and to present themselves as "people who will strictly interpret the law and not rewrite it from the bench." A former Rove staffer explained to me that the term "activist judges" motivates all sorts of people for very different reasons. If you're a religious conservative, he said, it means judges who established abortion rights or who interpret Massachusetts's equal-protection clause as applying to gays. If you're a business conservative, it means those who allow exorbitant jury awards. And in Alabama especially, the term conjures up those who forced integration. "The attraction of calling yourself a 'strict constructionist,'" as Rove's candidates did, this staffer explained, "is that you can attract business conservatives, social conservatives, and moderates who simply want a reasonable standard of justice."

As with direct mail, Rove was skilled at reaching specific voter segments with television commercials, buying air time only during programs that he believed would attract the audience he was trying to reach. In his Alabama races he was known particularly to withhold advertising from The Oprah Winfrey Show and similar afternoon programming—"trimming a media buy," as it is known in the trade. Bill Smith, who worked on a series of close races with Rove in Alabama, says, "There's a real overlap in what he specialized in professionally and what you need to do in a tight race." Whether he is seeking donors in a direct-mail fundraising campaign or manipulating a particular demographic sliver to win a close race, Rove's professional goal has been strikingly consistent: to reach the right people.

How Rove has conducted himself while winning campaigns is a subject of no small controversy in political circles. It is frequently said of him, in hushed tones when political folks are doing the talking, that he leaves a trail of damage in his wake—a reference to the substantial number of people who have been hurt, politically and personally, through their encounters with him. Rove's reputation for winning is eclipsed only by his reputation for ruthlessness, and examples abound of his apparent willingness to cross moral and ethical lines.

In the opening pages of Bush's Brain, Wayne Slater describes an encounter with Rove while covering the 2000 campaign for the Dallas Morning News. Slater had written an article for that day's paper detailing Rove's history of dirty tricks, including a 1973 conference he had organized for young Republicans on how to orchestrate them. Rove was furious. "You're trying to ruin me!" Slater recalls him shouting. The anecdote points up one of the paradoxes of Rove's career. Articles like Slater's are surprisingly few, yet as I interviewed people who knew Rove, they brought up examples of unscrupulous tactics—some of them breathtaking—as a matter of course.

A typical instance occurred in the hard-fought 1996 race for a seat on the Alabama Supreme Court between Rove's client, Harold See, then a University of Alabama law professor, and the Democratic incumbent, Kenneth Ingram. According to someone who worked for him, Rove, dissatisfied with the campaign's progress, had flyers printed up—absent any trace of who was behind them—viciously attacking See and his family. "We were trying to craft a message to reach some of the blue-collar, lower-middle-class people," the staffer says. "You'd roll it up, put a rubber band around it, and paperboy it at houses late at night. I was told, 'Do not hand it to anybody, do not tell anybody who you're with, and if you can, borrow a car that doesn't have your tags.' So I borrowed a buddy's car [and drove] down the middle of the street … I had Hefty bags stuffed full of these rolled-up pamphlets, and I'd cruise the designated neighborhoods, throwing these things out with both hands and literally driving with my knees." The ploy left Rove's opponent at a loss. Ingram's staff realized that it would be fruitless to try to persuade the public that the See campaign was attacking its own candidate in order "to create a backlash against the Democrat," as Joe Perkins, who worked for Ingram, put it to me. Presumably the public would believe that Democrats were spreading terrible rumors about See and his family. "They just beat you down to your knees," Ingram said of being on the receiving end of Rove's attacks. See won the race.

Some of Rove's darker tactics cut even closer to the bone. One constant throughout his career is the prevalence of whisper campaigns against opponents. The 2000 primary campaign, for example, featured a widely disseminated rumor that John McCain, tortured as a prisoner of war in Vietnam, had betrayed his country under interrogation and been rendered mentally unfit for office. More often a Rove campaign questions an opponent's sexual orientation. Bush's 1994 race against Ann Richards featured a rumor that she was a lesbian, along with a rare instance of such a tactic's making it into the public record—when a regional chairman of the Bush campaign allowed himself, perhaps inadvertently, to be quoted criticizing Richards for "appointing avowed homosexual activists" to state jobs.

Another example of Rove's methods involves a former ally of Rove's from Texas, John Weaver, who, coincidentally, managed McCain's bid in 2000. Many Republican operatives in Texas tell the story of another close race of sorts: a competition in the 1980s to become the dominant Republican consultant in Texas. In 1986 Weaver and Rove both worked on Bill Clements's successful campaign for governor, after which Weaver was named executive director of the state Republican Party. Both were emerging as leading consultants, but Weaver's star seemed to be rising faster. The details vary slightly according to which insider tells the story, but the main point is always the same: after Weaver went into business for himself and lured away one of Rove's top employees, Rove spread a rumor that Weaver had made a pass at a young man at a state Republican function. Weaver won't reply to the smear, but those close to him told me of their outrage at the nearly two-decades-old lie. Weaver was first made unwelcome in some Texas Republican circles, and eventually, following McCain's 2000 campaign, he left the Republican Party altogether. He has continued an active and successful career as a political consultant—in Texas and Alabama, among other states—and is currently working for McCain as a Democrat.

But no other example of Rove's extreme tactics that I encountered quite compares to what occurred during another 1994 judicial campaign in Alabama. In that year Harold See first ran for the supreme court, becoming the rare Rove client to lose a close race. His opponent, Mark Kennedy, an incumbent Democratic justice and, as George Wallace's son-in-law, a member in good standing of Alabama's first family of politics, was no stranger to hardball politics. "The Wallace family history and what they all went through, that's pretty rough politics," says Joe Perkins, who managed Kennedy's campaign. "But it was a whole new dimension with Rove."

This August, I had lunch with Kennedy near his office in Montgomery. I had hoped to discuss how it was that he had beaten one of the savviest political strategists in modern history, and I expected to hear more of the raucous campaign tales that are a staple of Alabama politics. Neither Kennedy nor our meeting was anything like what I had anticipated. A small man, impeccably dressed and well-mannered, Kennedy appeared to derive little satisfaction from having beaten Rove. In fact, he seemed shaken, even ten years later. He quietly explained how Rove's arrival had poisoned the judicial climate by putting politics above matters of law and justice—"collateral damage," he called it, from the win-at-all-costs attitude that now prevails in judicial races.

He talked about the viciousness of the "slash-and-burn" campaign, and how Rove appealed to the worst elements of human nature. "People vote in Alabama for two reasons," Kennedy told me. "Anger and fear. It's a state that votes against somebody rather than for them. Rove understood how to put his finger right on the trigger point." Kennedy seemed most bothered by the personal nature of the attacks, which, in addition to the usual anti-trial-lawyer litany, had included charges that he was mingling campaign funds with those of a nonprofit children's foundation he was involved with. In the end he eked out a victory by less than one percentage point.

Kennedy leaned forward and said, "After the race my wife, Peggy, was at the supermarket checkout line. She picked up a copy of Reader's Digest and nearly collapsed on her watermelon. She called me and said, 'Sit down. You're not going to believe this.'" Her husband was featured in an article on "America's worst judges." Kennedy attributed this to Rove's attacks.

When his term on the court ended, he chose not to run for re-election. I later learned another reason why. Kennedy had spent years on the bench as a juvenile and family-court judge, during which time he had developed a strong interest in aiding abused children. In the early 1980s he had helped to start the Children's Trust Fund of Alabama, and he later established the Corporate Foundation for Children, a private, nonprofit organization. At the time of the race he had just served a term as president of the National Committee to Prevent Child Abuse and Neglect. One of Rove's signature tactics is to attack an opponent on the very front that seems unassailable. Kennedy was no exception.

Some of Kennedy's campaign commercials touted his volunteer work, including one that showed him holding hands with children. "We were trying to counter the positives from that ad," a former Rove staffer told me, explaining that some within the See camp initiated a whisper campaign that Kennedy was a pedophile. "It was our standard practice to use the University of Alabama Law School to disseminate whisper-campaign information," the staffer went on. "That was a major device we used for the transmission of this stuff. The students at the law school are from all over the state, and that's one of the ways that Karl got the information out—he knew the law students would take it back to their home towns and it would get out." This would create the impression that the lie was in fact common knowledge across the state. "What Rove does," says Joe Perkins, "is try to make something so bad for a family that the candidate will not subject the family to the hardship. Mark is not your typical Alabama macho, beer-drinkin', tobacco-chewin', pickup-drivin' kind of guy. He is a small, well-groomed, well-educated family man, and what they tried to do was make him look like a homosexual pedophile. That was really, really hard to take."

Earlier this year the lone Democrat on the Alabama Supreme Court announced his retirement. There's an excellent chance that on Election Day the court will at last become entirely Republican.

Almost from the beginning Karl Rove has signaled that he expects a close 2004 election, and he has run George W. Bush's re-election effort accordingly. While John Kerry's campaign has made an extraordinary effort to gather moderate voters to his liberal base by stressing its candidate's decorated war record and centrist views, Rove—in contrast to 2000's invitingly gauzy message of "compassionate conservatism"—has returned to his traditional strength: motivating the base of conservative voters.

Bush's campaign has naturally focused on the battleground states, but Rove's strategy can be decoded by looking at the targets of emphasis within those states. They are predominantly solid Republican areas such as Pensacola, Florida, and Cincinnati, Ohio. Rove's gambit is to improve Bush's margins in places where the President fared well in the 2000 election, just enough—a few points higher among Catholics, evangelicals, Hispanics—to prevail once more. To achieve this he is following the lessons of tight races past, buying television time in solidly red Fargo, North Dakota, because the airwaves also reach the neighboring swing state of Minnesota, and in solidly blue Burlington, Vermont, so as to draw a few more voters to Bush in the battle for New Hampshire, next door.

Rather than soften Bush's appeal to reach moderates, Rove, as he has done throughout his career, is attempting to control the debate by expertly spotlighting issues sure to inspire his core constituency: the drive for a constitutional amendment to ban gay marriage, the pronouncements about love of country, the unremitting attack against anything in an opponent that seems impregnable. All these tactics stand out in Rove's most memorable past victories.

Privately, Rove has been challenged and even denounced for his approach. A common refrain I heard from Republican consultants a few months ago was that his approach is foolish, because for the sake of an ideologically intense campaign, Rove is ceding to the Democrats the moderates Kerry is pursuing. And, these consultants fear, it puts Bush in jeopardy of seeing outside events decide the race.

But an interesting thing happened as I worked on this piece. Early in the summer, as Bush was struggling, even Rove's allies professed to doubt his ability to control the dynamics of the race in view of an unrelenting stream of bad news from Iraq. Several insisted that he was in over his head—with an emphasis that seemed to go deeper than mere professional envy. Yet by August, when attacks by the anti-Kerry group Swift Boat Veterans for Truth were dominating the front pages, such comments had become rarer. Then they died away entirely.

If this year stays true to past form, the campaign will get nastier in the closing weeks, and without anyone's quite registering it, Rove will be right back in his element. He seems to understand—indeed, to count on—the media's unwillingness or inability, whether from squeamishness, laziness, or professional caution, ever to give a full estimate of him or his work. It is ultimately not just Rove's skill but his character that allows him to perform on an entirely different plane. Along with remarkable strategic skills, he has both an understanding of the media's unstated self-limitations and a willingness to fight in territory where conscience forbids most others.

Rove isn't bracing for a close race. He's depending on it.