Election-night TV: analyzing the analyses

Election night is the political Super Bowl for network news organizations. They spare no expense covering the returns and attempting to win the loyalty of viewers. Unfortunately for the American public, the ''postgame'' analyses provided by the networks are frequently superficial, misleading, and erroneous. Consequently, the networks squander an unparalleled opportunity to help us understand better the intricacies of the electoral process.

The most controversial aspect of election-night coverage is the projection of results before polls close. But equally important is what the networks do thereafter - their attempts to explain why the elections turned out as they did. Election-night viewers should be forewarned that these explanations suffer from three major flaws.

First, network explanations for election results are often arbitrary. Network analysts are quick to offer opinions of why a candidate won. Such opinions are always plausible, because they do not contradict known facts. A convincing explanation, however, also needs supporting evidence. It is seldom provided.

In 1982 network commentators asserted whenever a Democrat won that it was because of voter dissatisfaction with Reaganomics. Republicans were said to have won because they distanced themselves from the President. This view was plausible, but supported by precious little evidence. And it assumed rather than demonstrated that no GOP candidates won because they backed the economic program of a popular President.

In 1984, any GOP House gains will undoubtedly be explained in terms of Reagan's coattails. This, too, will be plausible. But it may be more correct to attribute Republican gains to local issues, Democratic retirements, GOP edges in campaign spending, or other factors. As you watch election-night TV, ask yourself if the networks provide evidence for their assertions.

Second, many network analyses are simplistic, offering explanations for election returns which rely exclusively on one cause. Electoral behavior is more complicated and cannot be accounted for with a simple factor.

A good example of the oversimplification of electoral analyses was the explanation of why Democratic challenger Mark White won in the 1982 Texas race for governor. CBS attributed it to the conservatism of the state's traditional Democratic majority. NBC explained it as the opposition of women, blacks, and Hispanics to Bill Clements, on grounds he had insulted women by asserting that none were qualified to serve on the Public Utility Commission. Both factors probably had some effect on the vote, but each network mentioned only one cause for the outcome. Thus the viewers of NBC and CBS received very different impressions of why Mark White won; those who heard both explanations could only guess which was more sound.

Campaign spending is perennially misrepresented by simplistic analyses. In 1982 several prominent, big-spending millionaires went down to defeat, and analysts on all three networks concluded that money didn't make much difference in electoral outcomes. That is ridiculous: There is good evidence that big spending made several races far closer than they would have been otherwise. Spending money, as any candidate knows, is better than not spending it, although it does not guarantee victory.

This year, when the networks analyze campaign spending, don't think only about whether big spenders win or lose, but ask yourself how those same candidates might have fared in the absence of big spending. Over the years winners have consistently outspent losers.

Third, network explanations often confuse cause and effect, resulting in erroneous conclusions. In 1982, the networks concluded that Daniel Moynihan beat Florence Sullivan for a New York US Senate seat because he had outspent her 4 to 1 - a preposterous assertion. Florence Sullivan raised much less money than Moynihan because he was a popular incumbent: It was not that Moynihan's spending made him popular but that his popularity permitted him to spend more money.

Network commentators this year will certainly analyze such questions as the effect Geraldine Ferraro had on the Democratic ticket, the effect of the debates on the outcome, whether a realignment of our political parties is going on, and the importance of issues such as arms control and abortion. As you watch, consider any assertion a commentator may make as a plausible hypothesis, not as the definitive truth, unless you can answer yes to each of the following questions: Is any evidence presented to confirm the explanation? Have other assertions, equally plausible but competing, been ruled out on the basis of available evidence? Does an explanation point to more than one factor? And are causes and effects put in the most reasonable order?

We want to hear, did we miss an angle we should have covered? Should we come back to this topic? Or just give us a rating for this story. We want to hear from you.