HAS THERE EVER been an election in whose results someone has not spied the glimmerings of a political realignment? We suppose the answer is yes (someone will bring up the 1880 election or something), but we can't think of any time lately when some gaggle of analysts hasn't been clucking about long-range political change. Sometimes with fairly good reason. American political results, if you put them on a graph, veer upward and then downward in an almost crazy pattern, and if you draw a dotted line straight out from the latest squiggle, you'll get something that looks like a pretty interesting trend. The only trouble is, the voters at the next election usually don't follow your dotted line.
The latest version of realignment theory is a little more sophisticated. We are in the presence of, it is said, a split-level realignment. American voters have now more or less permanently settled on electing Republican presidents, the theory says, except -- always these exceptions -- for unusual circumstances. At the same time, the theory goes on, they -- or at least about half of them -- are ready, willing and able, again except for those unusual circumstances, to elect Democrats to congressional and state and local offices.
You can find a lot of support for this theory in the results of the last several elections. And that's one of the problems with realignment theories: they tend to explain better what's happened in the recent past than to forecast what's going to happen in the future. The most elegant and persuasive of realignment theories was Samuel Lubell's explanation of how, by the early 1950s, a once majority- Republican America had become majority-Democratic. Mr. Lubell explained masterfully the Roosevelt and Truman victories and said much that was useful about the politics that was to come. Lesser minds extrapolated from a set of Democratic victories and concluded, just 20 years ago, that the Republican Party was threatened with extinction. Since then its presidential candidates have outpolled the Democrats by 42 million votes.
So the split-level realignment theory provides a useful explanation of election results from 1968 to 1984. Whether it tells us much about the future is another thing. In an age when voters feel no moral obligation to support one political party regularly and when almost everyone is mechanically capable of splitting a ticket, realignment is inherently impermanent: we realign every time we vote. Since the ratification of the Constitution, we have had only 50 presidential and 99 congressional elections -- too few, statisticians will tell you, to be the basis for any statistically meaningful generalizations. To those who suggest that American politics has suddenly become more or less permanently realigned, we suggest the response that has always come naturally to political as well as other losers: wait till next year.