Throughout history, leaders sought to predict the outcome of battles by examining chicken entrails or consulting oracles. Today, political campaigns -- our modern battlegrounds -- still require positive predictions to keep the money flowing and the troops stoked.
If they don't get a good omen from current polls, they can get it from a new cottage industry of election forecasters that includes economists, computer analysts and demographers.
In 1992, the Republicans found comfort when Yale economist Ray Fair predicted a Bush victory, while Democrats drew on another distinguished Yale economist, Edward Tufte, who predicted a Bush defeat using a different economic model.
This year, economic determinism is increasingly supplemented by demographic determinism -- using past election results to predict future ones. While Democrats have the polls on their side, Republicans have found a star demographer, John Morgan, who predicts a Dole win and a Republican gain of 15 to 18 seats in the House and three to four in the Senate. He is almost invariably introduced with the awe-inspiring comment that in 1994 he predicted Republicans would win 52 seats. And they did.
No one mentions that in 1992 he predicted with equal certainty a Bush victory. Mark Lubbers, who runs Anthem, a Fortune 500 health-care management company, remembers Morgan speaking to its top 100 executives back then: "He came to our headquarters in Indianapolis," he told me, "and assured us that the president would be comfortably re-elected."
This does not make Morgan a bad demographer. Indeed, Morgan is as brilliant a demographer as they come. But as Frank Luntz, the Republican polling and communications whiz kid, put it: "Within his own parameters, Morgan is correct. But it will take a good message and effective messengers to make his predictions come true."
Morgan acknowledged as much when I talked to him. "The performance of the candidate is not taken into account. My demographic projections assume that both candidates will run viable campaigns."
There's the rub. Who the candidate is, what his message is, what kind of campaign he runs -- all those variables come under the heading of "other things being equal." And that's what makes these projections barely more reliable than crystal-ball gazing.
For Republicans, demoralized by poll results and in-fighting over the last six months, it is mighty comforting to sit in a room with other party faithful and bask in Morgan's scientific-sounding predictions, dreaming of Nov. 5 and a clean Republican sweep.
Morgan will impress you with the facts: This is the first time since 1872 that the Republicans have a majority of Southern seats in the House, while in Florida the state senate went Republican for the first time since Reconstruction. He will also point out that since 1994, 206 elected officials have switched to the Republican Party nationwide.
He tells his audiences that a large number of open seats in the House and Senate are in the South and West, bastions of GOP strength. But he does not discuss the possibility that Republicans may go even further in the hole in the Northeast and Midwest. And he offers no explanation for public and internal surveys that show the GOP losing the generic vote for Congress by anywhere from 5 to 16 points.
It's no wonder Morgan is in such high demand among Republicans hungry for good news.
Demography may be a serious science, but politics is an art. Analyzing demographic trends to decide what age, gender or ethnic groups to target is one thing; using demography to predict the next election is another. As Bill Schneider, the CNN political analyst, put it: "The models work except when they don't work."
Schneider's law of election forecasting reminds us that there are no certainties -- either in life or in elections. But the higher the stakes, the greater our need to predict the outcome. And in high-stakes presidential campaigns, anything hopeful will do to keep the faithful humming. Did you know, for example, that Clinton is only the fifth left-handed president in American history, and none of the other four -- James Garfield, Harry Truman, Gerald Ford and George Bush -- has won re-election?
I have not recently consulted tea leaves or tarot cards, but Vivian, the Washington astrologer, has already announced to the press that it doesn't look good for Clinton.
"Double, double toil and trouble; fire burn and cauldron bubble," the witches told Macbeth when he went to consult them about his future. He interpreted their prediction to mean he was in no danger. GOP leaders are well-advised to approach "double-bubble" demographic predictions with a tad more skepticism.
[ Printer-friendly version ]