On Nov. 5, 1980, the day after 53.9 percent of all voting-age Americans cast their vote in the presidential election, a steady stream of mail and calls began to flow into the national office of the League of Women Voters. Most of the writers and callers complained that the media's projections of winners more than two and a half hours before polls closed on the West Coast distorted the political process and ran the risk of directly affecting -- rather than merely reporting -- the outcome.

Early projections on election night are not a new development. The Senate has twice passed legislation to make projections more difficult. The first measure, passed in 1972, provided for an Election Day of 12 hours with all polls closing simultaneously at 11 p.m. EST. The second measure, passed in 1974, prohibited disclosure of presidential election results prior to midnight EST. Both of these measures failed to pass the House.

Neither of these two proposals -- or proposals for Sunday elections, with or without 24-hour voting -- would fully eliminate the possibility of early projections. One of the techniques used to project elections -- exit polling -- has become so sophisticated that it is possible for projections of national election results to be based on early morning interviews of a few thousand exiting voters in sample precincts. Using exit polls, projections can be made, regardless of when the polls close, which day the election is held or whether or not election officials are required to withhold the results of an election.

Recently, the League, the Committee for the Study of the American Electorate and 30 other national organizations wrote to all major networks, wire services and broadcast groups asking them to refrain voluntarily from projecting winners in any given election before all polls are closed in the appropriate area. By this we mean that we do not object to the projecting of results in statewide races when polls are closed in that state; this practice can do no harm. We do object, however, to the practice of projecting statewide results if polls are still open in that state. Further, on presidential election night, the media should report the official results in the presidential race in such state and should not project presidential results when polls are still open.

Our request for voluntary restraint is based on our abiding belief in the freedom of the press and its role as a cornerstone of American democracy. We do not believe that any news agency should refrain from full and fair reporting of actual and final counts when they become available. But projections, no matter how they are derived, are not the same as actual results and can have the effect of influencing those results.

A commonly used argument against voluntary restraint is that we cannot isolate projections of victory from other Election Day 1980 happenings and prove conclusively that projections had an adverse impact on voter turnout. President Carter's concession speech at 6:45 p.m. PST certainly didn't encourage people to vote in the presidential race. But it is possible that Mr. Carter would not have conceded the election when he did had the networks not begun to project the Reagan victory so early.

The burden of proof should be on those who say early projections have no effect on election outcome or voter turnout. Indeed, it is difficult to argue persuasively that projections didn't have an impact on several 1980 contests that were decided by margins ranging from 25 to 800 votes.

To date, the response to our request for voluntary restrain has not been favorable. If the media do not agree to refrain voluntarily from projecting results, it may become necessary for Congress to weigh the value of votes lost due to disillusionment with the system against the costs of legislative solutions that attempt to address the problem.

Our call for voluntary restraint in no way diminishes our zeal for increased registration and facilitation of the elections process. We not only have to get people to register, we also must have them believe their vote counts.