As the campaign races to the end, with George W. Bush holding a slight lead over John Kerry in most polls, it is interesting to go back and see how much of the political conventional wisdom (you read some of it in this space) has proved to be wrong.

  The Republicans will have a big money advantage, especially between the primaries and the conventions. Dead wrong. First Howard Dean, then John Kerry and the Democratic Party found huge numbers of unexpected money come in on the Internet. Energized Bush haters contributed again and again, and Kerry and the Democrats have had plenty of money to get their messages out.

 Senate Democrats have done a fine job of raising money, too. Huge amounts, mostly from very big contributors, financed the anti-Bush ads and organizing drives of 527 organizations. Three individuals -- Peter Lewis, George Soros and Steve Bing -- contributed more than $50 million to anti-Bush 527s. Thank goodness the McCain-Feingold law got the big money out of politics.

  It's the economy, stupid. During most of the 2003-04 campaign cycle, many commentators have said the election will come down to the state of the economy. It doesn't look like that today. The economic numbers -- unemployment, inflation, jobs -- have been quite good the last 14 months, similar to those when Bill Clinton was re-elected in 1996.

 Polls show the economy is a wash. Sure, John Kerry and other Democrats are talking about jobs in local areas that have had highly visible job losses. But they talk about Iraq much more. One possible reason: Voters who remember the 1930s Depression have always been exquisitely sensitive to even mild upticks in unemployment and downticks in the economy. But these voters were only 8 percent of the electorate in 2002, and that number is not going up.

  Voters have made up their minds about George W. Bush, and his numbers are not going to rise or fall much. This has been the line of many intelligent Democratic pollsters, and there is a lot of precedent to support it. Re-election races have tended to be referendums on the incumbent. But the numbers this year have told a different story. Bush's numbers against Kerry sunk during the spring, when the news media was running stories on Abu Ghraib and Iraq violence. Then Bush got a bigger bounce out of his convention than Kerry did during his convention. Bush's numbers sunk after the first debate, but rose after the next two, even though polls showed him doing no better than a draw.

 My theory: The news media, much of it heavily biased, has been a more effective Bush opponent than Kerry and the Democrats. That's why both Kerry and John Edwards in debates urged voters to remember what they've been seeing on television.

  Continuing violence in Iraq will sink Bush. That could still happen. But almost all the polls show most voters favoring Bush on Iraq as well as the wider war on terrorism.

  This election will be resolved by lawsuits. OK, the fact is that I don't know yet whether this will turn out to be true. It certainly could. But consider this about the 2000 Florida controversy: All the lawyering succeeded in moving only a few hundreds of votes. The problem was that after the first count only about 1,200 votes separated the candidates.

 The Gore campaign could not find a way to challenge Bush's 7,211-vote margin in New Hampshire, whose four electoral votes would have given Gore the election. It's very rare for a decisive state to be as close as Florida. This time, Democrats seem to be preparing to argue that challenged provisional ballots would wipe away Bush margins considerably larger than Florida's. But it's not clear that this will produce prolonged litigation.

  True conventional wisdom. We should remember that not all the political conventional wisdom has been wrong. This does seem to be a very close election, with voters and states standing pretty much where they did in 2000. Polls as this is written suggest Kerry might switch New Hampshire and Ohio from red to blue and that Bush might switch Wisconsin, Iowa and New Mexico the other way.

 Cultural attitudes still tend to determine presidential vote, and relatively few voters seem to be undecided. But even that conventional wisdom, which still seemed founded with the election more than a week off, may prove to be wrong -- as you may know if you are reading this article after the returns are in.