The conduct of election surveys and publication of the results are rooted in freedom of speech, said Dr. Mahar Mangahas, president and founder of the Social Weather Stations (SWS).
When a survey is done scientifically, even from a sample of of only 1,200 persons, it can give an accurate picture of voter preferences in the entire country.
Mangahas spoke in a forum on “Understanding Election Polls” held last week in the Cebu Cultural Center (CCC) organized by the University of the Philippines in Cebu.
He discussed the legal battles won by SWS to fight off attempts to ban exit polling and the publication of results of pre-poll surveys.
“We in SWS are non-partisan. We don’t take sides. We are warriors for democracy,” he said.
He said the Supreme Court has been very supportive and in May 2001 quickly struck down as “unconstitutional” a provision of the Fair Election Act that banned the publication of election surveys 15 days before national elections and 7 days before local elections.
For the May 2013 election, however, the SWS has no plans to conduct exit polls when voters step out of the precincts.
Chartering a jumbo jet
Mangahas said no media outlet has commissioned SWS to repeat its 2010 exit poll for national positions, possibly due to budget restraints.
Asked in an open forum how much it cost to commission a national survey, Mangahas wouldn’t pin down a figure.
Instead he compared it to “chartering a jumbo jet where you pay for the whole plane. Sometimes getting only a few seats is more affordable.”
Managhas clarified that the SWS surveys on voter preference “belong to SWS” and that results are posted on their website for everybody to appreciate.
“Nobody commissions that. That is information for all,” he said.
Mangahas said SWS actually does more surveys on other subjects like poverty, hunger, the peace process, and perceptions on corruption, as part of its mission “to promote meaningful development.” Political surveys are only part of their work.
But their value is in testing the agency’s “survey quality” and accuracy because a poll survey involves a full population count.
Face to face
He explained that SWS uses a sample size of 1,200 respondents or 300 each for Luzon, Visayas, Mindanao and Metro Manila in 240 geographical spots.
Random sampling is strictly followed such that people who “volunteer” to be surveyed are automatically rejected, said Managahas.
Trained researchers are sent on field to do face-to-face interviews and phrase questions in a fair way with no leading queries.
Survey locations are kept confidential to avoid having interested parties “polluting” the sample.
“We do not reveal where we go. If the location is found out, we cancel it,” Mangahas said. SWS also does not use mobile phones as random targets.
In scientific surveys, respondents are chosen by random sampling from the population.
Everyone must have a chance to be surveyed and be free to speak. However, to be practical a sample size is used.
Random sampling
Mangahas said SWS uses a standard cluster of five households in a barangay which is already a “very high quality survey.”
In a selected barangay, the survey may start, for example, from the barangay captian’s house or police station and take respondents in every third dwelling after that.
SWS simulates the voting experience, and gives respondents a paper ballot that requires shading answers just like the real ballot. .
“Don’t call it a scientific survey if it’s not random sampling. Don’t expect it to predict,” he said.
Every election, said Mangahas SWS gets lawsuits from disgruntled candidates.
“These are foolish, foolish suits. We have never lost one,” he said.
The politicians who get angry at SWS are often those who show up with weak ratings in the survey because the exposure means they risk losing campaign funding and volunteers, said Mangahas.
Mind conditioning
On the other hand, if candidates fare well in the monthly SWS campaign surveys, they stand to get more campaign contributions.
The SWS founder said it’s not true that political surveys during the campaign will unduly influence voters as part of “mind conditioning”.
“It’s not the surveys that determine who won, it’s the votes. The survey cannot judge the quality of the votes or the total count. Honesty is not determined in survey results. “Cheating is discovered in another way,” he added.
In the 2010 presidential race, the Comelec count based on certificates of canvass closely matched exit poll results of SWS with a difference of only .401 percent.
Aquino won 42.08 percent in the Comelec tally compared to 43.34 in the SWS exit poll.
Joseph Estrada came out 26.25 percent in the Comelec count compared to 26.38 in the SWS poll. /Eileen G. Mangubat with Correspondent Christine Emily Pantaleon