It's a side issue in American politics, but it's been a vexing one nevertheless: the fact that the television networks project election results from the 80 percent of the nation in the Eastern and Central time zones while the 20 percent in the Mountain, Pacific, Alaska and Hawaii time zones still have a chance to go -- or not go -- to the polls. A lot of Western state politicians were furious in 1980 when the networks projected Ronald Reagan the winner at 5:15 Pacific time and broadcast soon afterward Jimmy Carter's concession statement; they claimed that voters got out of the lines at the polls and that several Democratic congressmen were defeated in close races as a result.

No effect has been proven. But it violates many people's idea of fairness for a winner to be conclusively and reliably declared while some people still haven't had a chance to vote. The networks have for years refrained from projecting results in any state until most or all of its polls are closed. But in recent years they "characterized" the results of exit polls before the real polls closed. As you might expect, these "characterizations" have gotten closer and closer to outright violations of the networks' own undertaking not to project results early.

The result has been acrimony, with some politicians hinting threateningly that maybe the networks will have to be censored. Now both sides have taken a step away from confrontation and toward an intelligent solution. The networks agreed last fall -- ABC before election night, CBS and NBC afterward -- not to "characterize" results any more. Reps. Al Swift (D-Wash.) and Bill Thomas (R-Calif.) agreed Thursday that they'll hold hearings on proposals to set a uniform poll-closing time across the nation.

That may not be as impractical as it first appears, and more feasible than the proposal for 24-hour voting. Poll-closing times of, say, 10 p.m. in the East and 7 p.m. on the West Coast would mean changes of only a few hours in most states. There is some evidence that the number of hours polls are open has only a small effect on turnout, and abundant evidence that turnout is lightest in the late hours.

Mr. Swift and Mr. Thomas are right not to just jump in and support uniform poll closings; it's always dangerous to tinker with the fragile machinery of elections. For once, Congress is looking at this issue at the right time, three years before the presidential election; there's time to write a good law and make whatever practical adjustments are necessary. Is it foolishly optimistic to hope that this nettlesome issue could be settled this year?