Every presidential campaign has its share of hard-ball political tactics, but nothing is more discomforting than a smear campaign. The deeply personal, usually anonymous allegations that make up a smear campaign are aimed at a candidate's most precious asset: his reputation. The reason this blackest of the dark arts is likely to continue is simple: It often works. The premise of any smear campaign rests on a central truth of politics: Most of us will vote for a candidate we like and respect, even if we don't agree with him on every issue. But if you can cripple a voter's basic trust in a candidate, you can probably turn his vote. The idea is to find some piece of personal information that is tawdry enough to raise doubts, repelling a candidate's natural supporters. It's not necessary, however, for a smear to be true to be effective. The most effective smears are based on a kernel of truth and applied in a way that exploits a candidate's political weakness. Having run Senator John McCain's campaign for president, I can recount a textbook example of a smear made against McCain in South Carolina during the 2000 presidential primary. We had just swept into the state from New Hampshire, where we had racked up a shocking, 19-point win over the heavily favored George W. Bush. What followed was a primary campaign that would make history for its negativity. In South Carolina, Bush Republicans were facing an opponent who was popular for his straight talk and Vietnam war record. They knew that if McCain won in South Carolina, he would likely win the nomination. With few substantive differences between Bush and McCain, the campaign was bound to turn personal. The situation was ripe for a smear. It didn't take much research to turn up a seemingly innocuous fact about the McCains: John and his wife, Cindy, have an adopted daughter named Bridget. Cindy found Bridget at Mother Theresa's orphanage in Bangladesh, brought her to the United States for medical treatment, and the family ultimately adopted her. Bridget has dark skin. Anonymous opponents used "push polling" to suggest that McCain's Bangladeshi born daughter was his own, illegitimate black child. In push polling, a voter gets a call, ostensibly from a polling company, asking which candidate the voter supports. In this case, if the "pollster" determined that the person was a McCain supporter, he made statements designed to create doubt about the senator. Thus, the "pollsters" asked McCain supporters if they would be more or less likely to vote for McCain if they knew he had fathered an illegitimate child who was black. In the conservative, race-conscious South, that's not a minor charge. We had no idea who made the phone calls, who paid for them, or how many calls were made. Effective and anonymous: the perfect smear campaign. Some aspects of this smear were hardly so subtle. Bob Jones University professor Richard Hand sent an e-mail to "fellow South Carolinians" stating that McCain had "chosen to sire children without marriage." It didn't take long for mainstream media to carry the charge. CNN interviewed Hand and put him on the spot: "Professor, you say that this man had children out of wedlock. He did not have children out of wedlock." Hand replied, "Wait a minute, that's a universal negative. Can you prove that there aren't any?" Campaigns have various ways of dealing with smears. They can refute the lies, or they can ignore them and run the risk of the smear spreading. But "if you're responding, you're losing." Rebutting tawdry attacks focuses public attention on them, and prevents the campaign from talking issues. We chose to address the attacks by trying to get the media to focus on the dishonesty of the allegations and to find out who was making them. We also pledged to raise the level of debate by refusing to run any further negative ads -- a promise we kept, though it probably cost us the race. We never did find out who perpetrated these smears, but they worked: We lost South Carolina by a wide margin. The only way to stop the expected mud-slinging in 2004 is for both President Bush and Senator Kerry to publicly order their supporters not to go there. But if they do, their behavior would be the exception, not the rule. Richard H. Davis is president of the Reform Institute and a partner in Davis Manafort, a political consulting firm. He was a fellow at Harvard's Institute of Politics in 2002. He was campaign manager for John McCain in 2000 and has worked in every presidential campaign since 1980.