As we watch the Overtime Campaign of the 2000 election, the headache of reporting it continues. Election night, when the networks made erroneous projections about who had won, was probably the worst moment in the 50-year history of television coverage of politics. Newspapers that prematurely miscalled the election were humiliated, too. And now, as the networks scramble to fill time, we watch spin doctors and talk show hosts try to shape public opinion until one side or the other capitulates.

The public should be troubled, too. The press has become a player rather than an observer. Misleading coverage not only failed to report what had actually happened on election night but created and has sustained the suggestion of an election that was called back and might be stolen, when the truth is simply that a very narrow election is late in being decided.

The problem isn’t that this is a close race. What has happened is that an election that didn’t lend itself to being easily reported has exposed longstanding weaknesses in the modern political press. This embarrassing performance can still produce a good outcome if it leads to serious self-examination and appropriate changes.

Political reporting has given way to campaign reporting, a process that the changing technology of campaigning has encouraged.

With each election, campaigns become increasingly sophisticated at targeting voters in their homes. This, in turn, has made politics more individual and private and less a group activity occurring in public. Rather than follow the campaign into the private space, political journalism has become more focused on understanding the newly elaborate campaign machinery, like targeted e-mail messaging, push polls and focus groups.

The trouble is what’s missing. Reporters have lost touch with voters, with regional politics and with campaigns as national conversations. On election night, no network had adequate sources in Florida who could tell them that they were relying on questionable numbers. Campaign journalism increasingly tends to see voters as abstractions, through polls, or as targets of campaign manipulations, to be interviewed in artificial focus groups or panels assembled to observe debates. Even the polling is increasingly limited to tracking polls, which tell us only the horse race and rarely the underlying reasons why people feel as they do.

At the same time, financial cutbacks have left the press less knowledgeable and more vulnerable. The bad projections on election night were an accident waiting to happen. In 1988, as financial pressures grew, the networks consolidated their once separate exit poll operations into one. Then there was pressure on this single consortium, Voter News Service, to cut back on the number of precincts and sample voters it used to make its projections and divine its understanding of the electorate.

Further cuts by the networks squeezed down their own political teams and jettisoned experienced political reporters. Politics is considered deadly for ratings, and that has translated into less political coverage on the nightly news. Prime-time magazine-style programs devote virtually no time to political matters, governance or issues like national security or social welfare. These reductions in the basics of political and governmental coverage have not, however, been accompanied by a greater humility. The old network TV air of omniscience remains but now rings hollow.

Competition, too, has had its effect, overwhelming the standards of news organizations. It was not the Voter News Service that first called the election prematurely for George W. Bush, but Fox News, whose election desk was run by Mr. Bush’s first cousin. Then, even though their own polling consortium had not yet called it, other networks, fearful of being scooped, quickly followed. Newspapers, wary of having incomplete stories, went along.

Today, the hardest decision in journalism is not to go with a story. The time-tested adage “Get it first, but first get it right” is now trumped by “It’s already out there,” and there seems less and less penalty for being the first to be wrong.

Now, the Overtime Campaign has unexpectedly made politics a story with a mass audience again. Because the pattern of the modern press is to swoop in, set up camera stands and then try to fill the time and hold an audience, the press is oddly reactive, depending on news conferences and statements by campaign officials. The never-ending news cycle gives “reporters” little time to actually go out and report. So we see a tendency to jump on small, incremental events and treat them as if they were momentous. In between, time is filled with talk — from pundits, partisan spin doctors, and, occasionally, invited experts. Spin is treated as news, and this makes the press susceptible to making mischief rather than enhancing public understanding.

What might be done?

First, journalists should not forget the most important questions the public needs to have answered: what happened and why. The outcome of a race takes care of itself. Understanding the meaning of the race is the harder but more important job.

News organizations should either get out of the business of projecting elections or develop much firmer guidelines about how and when to do so. They should end the reliance on a single shared source of polling information and form alliances to develop competing research efforts, which would correct each other and help explain why people favor the candidates they do. There should be more in-depth polling and less tracking polling.

More fundamentally, the press should start viewing the election not as two campaigns but as a national conversation. It should stop interviewing voters in artificial focus groups and start knocking on doors. In short, reporters need to begin to try to understand the campaign that now goes on inside the private space. Only then can journalism begin to help us all understand the underlying factors in our elections and provide the context for societies to self- govern.