Nonetheless, had the electorate not been convinced that the election was over even before
the day to vote had arrived, the increased turn-out which would have occurred would have
given him the votes he needed in the key states where the election was decided to have
won.
XIII. Pretenses of Objectivity
There can be a number of difficulties with the polling process, but, with or without
conscious bias, these surveys can err in terms of their queries or samples. Often, while a
determination is made as to the leanings of those interviewed, such information is
generally glossed over in the reporting of results. A likely voter indicating they would
probably vote for Clinton were the election held that day might be identified or identifiable
as someone 'leaning' that way merely, or differentiated from more 'solid' Clinton support
because they happen to be Republican or independent and leaning toward such a vote.
Such characteristics could be crucial in understanding the potential vote options which
might occur.
Hypothetically, a survey reporting a 51% support for Clinton, within a margin of error
of four points, may actually be a report of 40% solidly for him, and 11% leaning that way.
It might also consist of 33 % Democrats leaning toward Clinton, with the other portion
actually being 12% independents and 6% Republicans likely to support him. In either
event, the glossing over of such information does little to add credibility to the poll and
may reflect a partisan predisposition by the pollster, the analyst, or the reporter.
Sampling can be just as precarious an exercise. In order to have any chance of accuracy
whatsoever, the sampling must be not only random and scientific, but scientifically
random. Not only would problems with selection skew the results, but the method of
aggregation can be a difficulty. Since over recent elections, black voters have cast ballots
at the level of 95% for Democrats seeking the White House, such calculation may be built
into the poll, assuming such numbers. Where the black vote changed, as it did in 1996, to
a 'mere' 85% Democrat, the impact on the numbers would be huge. If there were twenty
million black voters, such a change would reflect a shift of some two million votes into the
Republican column, enough to significantly alter the election.
Even if we give pollsters the benefit of the doubt, something increasingly difficult to
do, that cannot be done for the utilization of such survey data by the media. What gets
reported, when and how it is structured, the words or tone or even body language used, all
will impact on what the person hearing the report thinks. But this is a problem beyond
polling reports. Opponents of SDI always call it Star Wars, and reports on the economy in
1992 routinely described only 4% growth in GDP, while in 1996, reports had the economy
'surging' ahead at an annual GDP growth rate of 3%.
Part of the problem is that the media present themselves under the guise of
objectivity(Bennet 1988). Not only are they not objective, but they cannot be, nor do they
seem to make much effort to be. And for all of these reasons, polls are not objective,
either.
And yet, we are constantly bombarded with such 'objective, factual' reporting. Even for
a cynical public, there is something about numbers which may lead many to conclude their
reliability, especially when the patter that was consistently fed to the public during the
summer and fall of 1996 was of the character it was. That does not connote any
'conspiracy' of pollsters or anyone else to 'program' the electorate, but it does explain both
polling inaccuracy in this election as well as others, as well as the public's apparent
acquiescence in the 'inevitable' Dole loss. Compounded by the level of misinformation, the
results should hardly have been expected to have been much different than they turned
out.
XIV. Pavlov's Voters?
The Perot phenomenon itself was, while not exclusively, in large measure a media-
conjured event. In spite of his millions, had the press not trumped him up, had they been as
keen in scrutinizing him as they have been on some other matters, and should have been
regarding him, he would never have 'happened' at all, either in 1992 or 1996. In fact, the
refusal of the media to play him the same in 1996 as they had in the previous election
probably had as much to do with his lower polling and vote numbers as anything else. But
then, in 1996, there was a certain 'fear' that Perot might hurt Clinton more than he would
whomever the GOP ran.
We had a similar phenomenon in 1996. The media was a major factor in the strength of
Buchanan, as it had been for him in 1992. The evaluation of primaries and other delegate
selection processes is sometimes curious. Often, a candidate coming in second (or even
lower) is the 'winner' because they exceeded the commentator's expectations. The
spotlight on Buchanan did some harm to the Dole campaign also because Dole had to
spend money to win the nomination, and was left with no resources to counteract the
Clinton/Democrat/union blitz that ran throughout the spring and summer nationally and
locally around the country. For some considerable period of time, the Dole effort was
hamstrung during the primary process because of spending constraints.
And for a brief moment, the media seemed to be going to do the same thing with Steve
Forbes that they had with Buchanan. Perhaps, though, he and/or his ideas became too
viable, because the 'romance' ended quickly in scurrilous reporting, although the effort did
have some negative impact on the Dole campaign effort.
But undoubtedly the most insidious tactic used against Bob Dole was the matter of his
age. It was heralded as an issue, but then downplayed, though frequently referentially
implied in word and picture. Any such suggestion should have infuriated senior citizens
even more than the rest of us, but so, too, should have Clinton's tax on social security
benefits, instead of the fear-mongering campaign about nonexistent Republican attempts
to 'slash' Medicare and Social Security and let old people die in the streets. It is a
tremendous handicap to have to overcome such levels of disinformation, either for a
Republican candidate or for a voter, for whom the costs of information can be staggering
and stifling.
The pollsters reported such findings, too. Republicans were seen as mean-spirited, as
having slashed spending, etc., by the public. Where did they got that idea? A better
analogy than that of an electorate 'programmed' or 'conditioned,' although there is some
validity to such notions if only as the matter of information costs, might be the spirit
engendered in the citizenry by the surveys and their endless presentation. That was more
the reason for the decline in voting turn-out that the much touted antipathy for politics due
to concern over its control by monied special interests. It is not as disparaging of the
American voters to suggest that they might have been 'set up' to believe Dole could not
beat Clinton as it is to engage in any such effort, whether by plan, concerted effort, or
more likely on partisan bias of those in a position to pull it off. This does not require
conspiracy. It is part of the liberal mind-set that they know better than the rest of us what
we think, do, and say. But in spite of the spirit conjured up by the politicos, in large
measure built on the statistical data of pollsters, the best efforts of the pollster geist could
only narrowly win over Bush and Dole, and could not turn out the House.
XV. The Media-tion of George Bush
It is old wisdom that Presidential approval ratings tend to track in an inverse
relationship to the rate of unemployment. If unemployment rises, approval will decline and
vice versa. In the case of George Bush, there would seem to be a firm basis for such a
conclusion. As unemployment rates varied through 1991 and 1992, Bush's rating varied
just as would be anticipated.
It is not sufficient, however, to conclude causality from coincidence, although a strong
enough correlation does present considerable grounds. But there are obviously other
factors which can contribute to approval rate vacillations. While it is difficult to argue
absolute causality from these, it certainly creates an appearance that there may be some
causal connection arguable, and the occurrence of certain events at points of vacillation in
approval rating provides at least some food for thought, as for example, Bush ratings sky-
rocketing to near 90% during Desert Storm. It is also notable that some of the variation in
Bush approval rating apparently took place irrespective of changes in the unemployment
rate.
Continue
