ublic opinion surveys were
uncannily right in calling the results of the last three senatorial elections.
There were cases where they said the likely winners of the 10th, 11th and 12th
slots were too close to call, explaining this was the result of the built-in
margin of error for their surveys. So why do the accuracy of the surveys and
their predictive values continue to be under question?
Let’s leave aside the altogether different issue of whether
the election surveys are useful or desirable. They obviously are useful,
otherwise nobody would be paying good money for the services of the outfits
which do such surveys. Candidates need to have an accurate snapshot of the
voters’ pulse so they can design appropriate campaign strategies. Voters can get
a sense of who the candidates with winning chances are. Thus, the voters are
forced to sharpen the focus of their choice.
Some critics condemn this bandwagon effect of surveys. Voters
are made to believe their favorites, who might be the most worthy of the whole
lot, are clear losers from the very start. So they are prompted by the results
of the surveys to abandon their original preference. It’s a valid criticism, but
it runs counter to the people’s right to information.
There is something unacceptable in a situation where
candidates, their strategists, campaign contributors and public opinion makers
have access to voters’ preferences while the rest are denied similar
information. This would be the effect of proposals to ban publications of survey
results and the Supreme Court is right in ruling that such restrictions are
unconstitutional.
But there’s another line of criticism of surveys that we find
interesting. There is this line being peddled by the administration that surveys
are not capable of catching the preferences of those who rely on their local
officials for "voting guidance." In other words, the so-called command votes.
The administration’s line is that the surveys do not take
into account these votes. And come Election Day, the command votes will deliver
victory to the majority of the administration candidates despite what the
surveys say.
The surveys in the last three elections appear to refute the
influence of command votes. The same surveys, however, recognize the existence
of a section of voters who say they would vote for whoever are endorsed by their
local leaders. The latest surveys, in fact, show that the proportion of
respondents who are expected to follow the suggestions of local leaders has been
on the rise.
So how do we explain the apparent contradiction? Our take is that the surveys
that are conducted as Election Day nears already capture the sentiments of these
command voters. There are indeed command votes. But these are already reflected
in the surveys. This is the neatest and most simple answer to the conundrum.
