Why do so many Americans not vote? Almost half of those eligible didn't vote in the 1980 presidential election; a solid majority didn't vote in the 1982 off-year elections. The latest study, conducted by ABC News in cooperation with the Kennedy School of Government at Harvard, finds, as others have, that most non-voters just don't think voting makes much difference. They tend to be younger, poorer and more mobile than those who vote regularly, and they are less informed and more cynical and negative about the political process than regular voters.

Turnout was somewhat higher in 1982 than in the previous off-year election--the first such uptick in two decades. That may reflect a more widespread positive attitude toward the political process, or it may have happened simply because more members of the baby boom generation have finally set down roots and gotten around to registering to vote. No one is sure. But turnout, as a percentage of those eligible, is still far below the levels of the early 1960s.

Can anything be done to increase turnout? State and local governments can make it easier to register. But that is likely to have only marginal effects. Some states allow registration on election day; that seems to have raised turnout, but only slightly. Holding elections on Sunday might actually depress turnout, the ABC-Harvard study suggests. Allowing absentee voting without requiring any special excuse raised turnout in 1982, when the Republicans conducted a massive absentee voter drive.

The ABC-Harvard study doesn't provide a conclusive answer to the question of whether turnout is depressed in states where polls are still open when television networks project the result in the presidential race. Some think that happened in 1980, although one factor there was the fact that President Carter conceded the election while polls were open on the West Coast. No one argues that the networks should have refused to air that statement, and it doesn't make any more sense to say that they should be prohibited from airing statistically valid projections.

A better idea is to hold the polls open across the nation over the same 24-hour period; a large majority in the survey react favorably to this proposal. The cost would be, by government standards, minimal. It is, unfortunately, probably too late to do this for 1984. It takes time to construct election machinery, and it must be done to zero-defect standards; state and local officials need plenty of lead time to hire more election workers and arrange to keep polling places open 24 hours. Congress has the power to act here, but it would probably be best for some states to lead the way. No great increase in turnout should be expected, but it would remove the suspicion some have that election results are affected by television coverage.