For most of us Americans, the 1980 election happened once, and that was enough. For Richard B. Wirthlin and Richard S. Beal, it happened many, many times. Their satisfaction with the final outcome was, at least in part, their pleasure at seeing the rehearsals reflected in the results.
Wirthlin is a PhD economist and a former head of the economics department at Brigham Young University, where Beal, a PhD in international relations, now teaches. More pertinent, they are, respectively, the president and the senior political analyst for Decision Making Information (DMI), the firm that was the source of the polling data for Ronald Reagan's campaign. Wirthlin doubled in brass as a member of Reagan's senior strategy board, and Beal worked with him in the Arlington headquarters, designing the sophisticated computer exercises that enabled them to "rehearse" the election many times during the fall.
Wirthlin, a lean, intense 49, and Beal, a roly-poly 34, are two of the brightest men I have met in politics. While their techniques are, in many respects, beyond my comprehension, they began showing me what they were doing back in September, with the understanding I would not write about it until the election was over. I do so now in the belief that you may be as intrigued as I was with the extent to which technology and human ingenuity have moved the art of campaigning beyond the methods of the old bosses.
The Political Information System (PINS) they devised was designed, in Beal's words, "to use polling data, not just to satisfy the information needs of the campaign, but to help the campaign decision-makers with their strategic judgements."
It combined current polling data with information on the state's historic voting patterns, and reasonable, subjective judgements on political trends in such a clear presentation on computer terminal screens that it was possible to get a scientific answer to such questions as these:
If the unions scare half their members about Reagan's labor record, should he step us his attacks on Carter or try to rebut their specific claims? Or, if John Anderson's vote begins to drop, should Reagan add a campaign stop in Connecticut, or can he afford to cancel one?
In senior staff meetings, these computer displays helped guide the allocation of the candidate's and surrogates' time, organizational resources and media dollars.
Two things were particularly impressive about the parts of the million-dollar operation they showed me during the fall. One was the discipline in their adherence to the basic theorem of Wirthlin's original campaign plan to give Reagan the best possible odds on winning 270 electoral votes, the minimum needed for victory. And the second was the flexibility in building into the computer designs a significant role for intuitive political judgments.
Combining these two principles, Wirthlin and Beal built the most successful model of the everchanging dynamics of a national election campaign yet designed, and used it to shape one of the most successful campaigns in American history.
With constant injection of fresh data from national-sample interviews and tracking surveys in 20 states, PINS showed in mid-October the Carter surge in Oregon and Washington, and cued the Reagan campaign to start, within two days, a stepped-up advertising effort to counteract it.
At a time in October when the press was reporting Reagan's campaign had stalled, PINS showed a significant firming of his support in key constitiuencies and blocked what Wirthlin called "some high-risk, off-the-wall decisions on what we had to do."
PINS also analyzed the shape of the election from Carter's point of view, correctly predicting in mid-October his forced abandonment of several southern battlegrounds in favor of a high-risk "big state" strategy. With that analysis in hand, Reagan was ready for the Carter push and stopped it cold.
In the closing stages of the campaign, Wirthlin and Beal used PINS to run simulations of the election every few days in order to maximize Reagan's chances of winning and to reduce the negative fallout from any "worst-case" developments.
Using current survey data (as corrected, within limits, by the subjective "feel" of campaign strategists), PINS would play out the election results on varying assumptions about the disposition of the undecided and Anderson vote and varying levels of turnout, then flash on the computer screens the simulated election results.Each "election" scenario could be processed through the PINS system in seven seconds, showing the result and its degree of probability.
When I last saw Wirthlin and Beal on the Friday before the real election, the two social scientists could hardly retain their academic composure.
That morning's PINS simulation had shown the very strong probability of a big Reagan win, a probability the public polls never quite reflected.
Reagan won the election through his own campaigning skills, with a lot of help from such intuitive politicians as Stuart Spencer, his de facto campaign manager. But for rival political strategists, looking ahead to future campaigns, the success of the Wirthlin-Beal PINS is enough to keep them on pins and needles for the next few years.