The decision by at least two of the TV networks to wait until New Hampshire's polls closed before flashing their projections of the presidential primary outcome on the nation's screens was the right one.
The public feels the news media meddle too much in politics as it is. Many resent the impression, when results are forecast before everyone has gone to the voting booth, that any further ballots don't matter. Network projections, based on surveys of voters exiting the polls, could have tampered with the results of the 1980 elections as the networks declared winners before Western state balloting had finished. Congress could decide to write restrictions on the timing and use of pre- and post-election polling. It's better that the networks set their own winner-projection guidelines to avoid the charge that they are intruding on the election process itself.
Exit surveys can be useful analytical tools. Balloting itself decides the winner and the loser, the apportioning of citizen votes. But the questions of who voted for whom and why are also important. How an election split along party , education, race, and income lines, and whether specific foreign or domestic policy issues influenced voting, can help decide the most crucial post-election question - what kind of mandate does the elected official have? The major network/newspaper polling operations make public their findings. This information helps to offset the advantage that politicians earlier had in financing their own private polls. It's similar to the effect of creating the nonpartisan Congressional Budget Office in economic affairs, which provides an alternative interpretation of economic data as a counterbalance to the White House's previous monopoly on data.
Washington State has banned exit polling within 1,000 feet of voting booths. Canada has adopted a kind of rolling blackout, by time zones, on projecting national election outcomes.
What the US networks will do about next fall's national elections, with irregular closing times across the continent possibly frustrating their impulse to broadcast their guess as early as possible about the outcome, has yet to be decided. It's unthinkable that, having developed the exit poll technique, the big network/newspaper polling combines would abandon it. But it's equally obvious that restraint - holding back state result projections until precincts close - must be practiced. An arbitrary time for projecting the presidential outcome may have to be set, although keeping to it in case of a landslide would surely strain credibility.
At base, elections are people voting. It's arguable whether the media orientation of today's politics dulls or whets the public's desire to cast ballots. But it's not arguable that a regard for the citizen's civic responsibility to vote should weigh heavily in the networks' election night coverage.