As they gather this week for their convention in Boston, Democrats allow themselves some hope. After all, the news for the White House has been bad for months now: chaos and casualties -- but no weapons of mass destruction -- in Iraq; the prison abuse scandal at Abu Ghraib; sharp rebukes from the Senate intelligence committee and the 9/11 commission; little or no credit from the public for economic recovery, according to recent surveys. A vice president snapping an obscenity at a U.S. senator, a president whose once-confident body language often now seems tense -- Democrats like the looks of this, too.

Yet the polls say the race remains close -- virtually tied, according to the latest surveys. Somehow Bush survives all the bad news. Perhaps those hopeful Democrats are kidding themselves? What's up with public opinion anyhow?

The easy explanation is the evenly divided electorate, as we learned so dramatically in 2000. This election looks as though it could be a rerun of 2000. Or so many commentators have argued. But political analysts are a little like generals, prone to re-fighting the last war when they ought to be figuring out how the next one will differ. And this one will be different from 2000; you can bet on it. Only twice since World War II has a presidential election closely resembled the one before it: in 1956, when Dwight D. Eisenhower skunked Adlai Stevenson for the second time, and in 1984, when Ronald Reagan dispatched Walter F. Mondale, just as he had trounced Jimmy Carter in 1980. In those two cases, both incumbents were popular, and each beat the same or a nearly identical opponent twice, during good and peaceful times.

Mark Mellman, Sen. John F. Kerry's pollster, uses historical comparisons to argue that President Bush is in trouble. Since 1964, Mellman notes, incumbents with approval ratings below 50 percent in the spring and summer of the year when they are running for reelection have always lost. This was the case in 1976, when Gerald Ford lost to Carter; in 1980, when Carter lost to Reagan; and in 1992, when this president's father lost to Bill Clinton.

By contrast, incumbents with approval ratings above 50 percent five and six months before the election always won: Lyndon Johnson, Richard Nixon, Reagan and Clinton. Depending on the poll, Bush's current approval rating is in the mid- to upper 40s.

Every incumbent who has won reelection in modern times had a double-digit lead over his opponent at this stage in the race, Mellman notes. A single-digit lead isn't enough. That's what the first Bush had over Clinton in early summer 1992. Carter held a small lead over Reagan in early summer 1980, and it survived until late October in many polls. Both those incumbents lost. So that puts Bush in company that he'd rather not keep.

Mellman is a Democrat promoting Kerry's candidacy. Would a Republican pollster see the history differently? I called Richard Wirthlin, Ronald Reagan's pollster and a principal strategist in Reagan's 1980 and 1984 campaigns. (His other credential: He won Outlook's first biennial Crystal Ball competition, coming the closest to predicting the results of the 1982 midterm elections.) Is Mellman right? I asked him.

Wirthlin thinks he is. "It's a pretty solid picture," he said. The problem for Bush is that a challenger enjoys natural advantages that tend always to erode an incumbent's early lead: "If a challenger runs an effective campaign, and that always has to be assumed, you've got to have a margin for the incumbent, because you almost always lose support as the challenger becomes better known, and is better able -- and I think this is often overlooked -- to pick and choose which issues can be driven to the disadvantage of the incumbent."

In other words, an incumbent is stuck with his record, but a challenger has a more open field to play on, and has more leeway to pick the issues that seem to suit him, and the times, best. This was just what Reagan did so effectively in 1980.

This year's race feels a lot like 1980. Then as now, a majority of the electorate disapproved of the incumbent's performance. Then as now, the party out of power was energized, and had raised a lot of money. Carter, like George W. Bush, initially won the presidency by a narrow margin, and he never built strong, lasting support in the country. Bush of course got a huge boost from popular support for his leadership after 9/11, and again from the invasion of Iraq, but in both cases, initially high approval ratings withered.

Carter remained close to Reagan in the polls all during the summer and fall of 1980. Indeed, he ran better in the horse-race polls than Bush is doing right now. But Reagan ultimately won an electoral landslide. Why?

Because the country really wanted an alternative to Carter. But initially, many Americans were nervous about the idea of a staunchly conservative movie actor as president. A brilliantly run campaign, aimed from the beginning at reassuring voters that Reagan would be a perfectly plausible -- and likable -- president, laid the groundwork for the one presidential debate that year, just eight days before the election. Reagan charmed the country that night. "There you go again," he said, brushing off Carter's attempts to depict him as an arch-conservative ogre. Reagan's question to the audience that night -- "Are you better off now than you were four years ago?" -- proved to be devastating. Most Americans felt they were not.

Well, that's the analysis of this old reporter, who has been covering presidential politics since 1968.

Today, according to the latest Washington Post poll, clear pluralities of voters think Kerry would handle domestic issues, including the most important economic ones, better than Bush. Asked which candidate can be called a "strong leader," nearly as many apply that term to Kerry as to Bush. On the question of who would do a better job fighting terrorism, Bush has the advantage -- 51 to 42 percent. But a clear majority also now thinks the war in Iraq was a mistake. And other recent surveys show that most Americans think the United States is "off on the wrong track."

Could we be seeing a kind of replay of 1980? "Every campaign is different," Wirthlin said. But, he added, "the parallels between our challenge [in '80] and what Kerry faces are quite striking." One key task, he said, was to lay out a positive vision for the country: "Running a more reassuring set of messages and images against a president who is under pressure and faces his own challenges is clearly advised."

Of course, events can intervene. "An uncontrollable event can change the shape of the last month of the campaign," Wirthlin said. This happened in 1980, when Carter's desperate effort to negotiate the release of American hostages in Iran came apart on the eve of the election.

But such last-minute surprises "almost always redound to the benefit of the challenger and not the incumbent," Wirthlin said. He spoke from experience. A Washington Post poll had Carter slightly ahead of Reagan just three days before the 1980 election, before the hostage negotiations collapsed. When the votes were counted, though, Reagan had won the popular vote handily, by a margin of 9.7 percent.

So Democrats gathering in Boston can cite some historical reasons to be hopeful this year. But of course, history is never doomed to repeat itself.

Author's e-mail:

Robert Kaiser is an associate editor of The Post. The first presidential campaign he covered was Nixon vs. Humphrey in 1968.

Precedential politics: Many commentators have called the 2004 election a replay of the deadlocked 2000 contest. But President Bush's latest poll numbers suggest a vulnerability that recalls Jimmy Carter's 1980 loss to his Republican challenger, California Gov. Ronald Reagan. Above, the candidates just before their one and only debate.