Judging surveys  

With political groups in Cebu reporting the results of so-called surveys favoring their respective candidates, the question of what to make of such conflicting results should be easy to answer.

Who made the survey? Do they have a credible track record? Who funded the survey? What are the sampling methods? What does the survey ask? If the answer to any of those is not clear, then take the results of the survey with a grain of salt.

Pre-election surveys aim to measure voters’ preference before election day. They are used to measure voter opinion, to know where the candidates are situated in the race, and to some extent help explain what voters think of the issues and the candidates. On the other hand, pre-election surveys are also criticized for promoting a horse-race coverage of elections, with news media emphasizing the candidates’ positions in the race instead of the issues that they stand for.

There were proposals long ago to ban not the conduct but the publication of election survey results. Proponents of the ban argue that the publication of the results of pre-election surveys might create a bandwagon effect in favor of candidates who are doing well in such polls. Some also fear that pre-election surveys might unduly confuse and influence the voters, which in turn may put into question the credibility of the election.

There may be basis for such fears. But the jury is still out there regarding the impact of pre-election surveys on actual voters’ behavior during election day. One study acknowledged the bandwagon effect but qualified the same by describing the effect as prevalent among people who may well have voted for the leading candidate anyway.

In truth, the matter on survey impact is a complex one. Election outcomes are still decided by many factors, among them effective campaigning, good timing, and the general sentiment of the populace. Election surveys also do not make projections as to which of the people interviewed are going to vote and how they will vote. That is a vital logistical factor which could influence the outcome of elections.

Likewise, anything can happen before election day that may knock the trend off and deliver an ugly surprise to the survey frontrunners. Hillary Clinton, for example, on account of the surveys, was predicted to win against Donald Trump in 2016. The election results surprised many, yet such did not create any major confusion or disruption of electoral democracy.

Fortunately, our own Supreme Court holds a more libertarian view when it comes to regulating surveys. It sees the matter as a free speech and press freedom issue. Restricting the conduct and publication of opinion polls during the campaign period amounts to a distrust of the capability of people to decide. McAllister and Studlar (1991) wrote that public opinion polls are too important to be left to politicians and policy makers.

In one case, this one about exit polls but which may imply also to pre-election polls, the Supreme Court concluded that “the interest of the state in reducing disruption is outweighed by the drastic abridgment of the constitutionally guaranteed rights of the media and the electorate.” Quite the contrary, according to the Supreme Court, instead of disrupting elections, exit polls --properly conducted and publicized-- can be vital tools for the holding of honest, orderly, peaceful, and credible elections.

But then there are fake surveys or surveys improperly conducted. To guard against such kind of surveys, we go back to the questions I earlier asked. Who made the survey? Do they have a credible track record? Who commissioned the survey? What are the sampling methods? What does the survey ask? How were the questions framed?

I can say that other than SWS and Pulse Asia, no other local or national survey firm can claim a credible track record. Even SWS and Pulse Asia make room for margin of errors in their survey reports. There are survey outfits managed by those in the academe who do surveys commissioned by local political groups, but the academics behind these surveys make it a point that the results are for internal use only and not to be published.

When a local politician starts brandishing survey results from an obscure “survey” firm, a non-commissioned survey at that, it makes you think twice of voting for that politician. The words “non-commissioned survey” already makes it suspect. Even a local survey, to be considered scientifically and statistically sound, is very expensive to execute. So what’s a foundation doing purportedly financing its own survey?