[Two articles follow. First, what the lead academic
prognosticators are forcasting (and they have historically had an
almost perfect success rate), and second, an article on exactly
how they do it. Neat stuff! --MS]
From:
http://www.washingtonpost.com/wp-dyn/articles/A9159-2000May25.html
By Robert G. Kaiser
Washington Post Staff Writer
Friday, May 26, 2000; Page A01
Is This Any Way to Pick A Winner?
Editor's Note: Though all are written as works of political
science and not aimed at a broad general audience, they are
comprehensible to lay readers, except perhaps in sections devoted
to statistical analysis. BEFORE THE VOTE includes the records of
all the forecasters mentioned in today's article.
So how exactly did Al Gore win the election of 2000? By making
the clever decision to run in the midst of an economic boom, and
by choosing to succeed a popular incumbent.
You didn't realize that Gore has won the election? A
technicality. According to half a dozen political scientists who
have honed and polished the art of election forecasting, the die
is all but cast. Today, with 165 days left before Americans go to
the polls, they are saying Gore will win 53 to 60 percent of the
vote cast for him and George W. Bush.
Although Bush's pollster finds fault with these forecasts, these
academic prognosticators have a startlingly good record
predicting election results months in advance. The fact that
opinion polls today give Bush a modest lead over Gore doesn't
faze them. Polls this early in the campaign "just have a
relatively low correlation with the fall vote," said Thomas Mann
of the Brookings Institution, who has written about election
forecasting. The forecasters have a better record, Mann added.
"It's not even going to be close," said Michael Lewis-Beck of the
University of Iowa, who foresees Gore winning 56.2 percent of the
two-party vote. Lewis-Beck's forecasting model is based on growth
of the gross domestic product from the fourth quarter of the
preelection year through the first quarter of election year, and
on poll findings on presidential approval and voter opinions on
which party's candidate will best promote peace and prosperity.
Should anyone listen to Lewis-Beck? Well, in 1996 he did miscall
the final results. His July forecast that year foresaw President
Clinton winning 54.8 percent of the two-party vote in November.
In fact, Clinton won 54.7 percent. Yipes. That was a better
prediction four months before the election than most commentators
and pollsters could make a few days before the voting. Indeed,
Lewis-Beck was much closer to the actual result than the national
exit poll taken on Election Day as voters left their polling
places, which overstated Clinton's vote by more than 3 percentage
points.
The leading academic forecasters share the belief that elections
reflect, first of all, underlying trends in the economy and
public opinion. Each has his own elaborate mathematical model,
but they share common ingredients: a measurement of the health of
the economy and poll findings on the public's political views.
Some leave it at that, some add other factors, such as how
Americans are feeling about their personal economic situations
(literally better than ever before, at this moment). And
different forecasting models use different measurements of the
economy and of public opinion, taken at different times from
about now through Labor Day.
All of them use elaborate higher mathematics to come up with
predictions of the share of the two-party vote that the candidate
of the incumbent party will win in November. (All agree that
third-party candidates have no palpable impact on their models,
and history bears them out.) Five of the best forecasters
(measured by their records) say that as of today, Gore can be
expected to win 53 to 60 percent of the two-party vote in
November. This means none of the forecasts predicts a really
close election. Most of these models have picked the winner
correctly in years since 1952 when the winner got 53 percent or
more of the vote.
If these models are right--and in fairness to their cautious
authors, none seems ready to bet his pension on his
prediction--Gore's biggest advantages are the popularity of the
president and the continuing economic boom. Clinton enjoys an
approval rating of about 60 percent in polls, a number that has
remained constant for many months. This suggests that by
November, many Americans may not agree with Bush and the
Republicans that it's time for a change in the party holding the
White House. While the longest economic expansion of the modern
era continues, the status quo retains considerable appeal.
Thomas M. Holbrook, a professor at the University of Wisconsin at
Milwaukee, uses as an economic indicator public responses to a
question asked regularly by one of America's oldest polls, the
University of Michigan's survey of consumers: ". . . Would you
say that you . . . are better off or worse off financially than
you were a year
