Surveys not useful in these times

I have written here after every election that the results bear out the surveys. The results of the 2013 midterm election for senators and the 2016 presidential election validated pollsters Social Weather Stations’ (SWS) and Pulse Asia’s forecasts.

In 2013 all the 12 senatorial candidates projected to win by SWS and Pulse Asia did win. In 2016 the two pollsters projected Davao City Mayor Rodrigo Duterte and Camarines Sur Representative Leni Robredo as the winning presidential and vice-presidential candidates. They even got the rankings of the candidates for the top two positions right.

Detractors of political polls cite the case of the American pollsters’ forecast of a Hillary Clinton victory over Donald Trump in the US presidential election in 2016 but it was Trump who was declared winner. Clinton did receive more votes than Trump as the polls projected. But the winner in the US is not based on who garners the most votes. The votes cast in the Electoral College determine the winner. A discussion of the US electoral system here would serve no purpose.

Other non-believers of surveys maintain that results of public opinion polls cannot be considered reliable indicators of the actual results of elections, pointing out that the ranks of the candidates in the polls do not jibe with the ranks of the actual election winners. But what the critics do not take cognizance of is that respondents are asked who they would vote for if the elections were held on the day of the interview, not on the day of the actual elections. A lot of things happen during the days between the day of the interview and the actual Election Day.

The changing preference of voters to a large extent is influenced by the events preceding the conduct of the surveys. The survey results reflect the sentiments of the people at the time the survey is conducted. SWS asks its respondents the question: “If the elections were held today, who would you likely vote as (position)?” Pulse Asia asks, “If the elections were held today, who would you vote as (position)?”

Voters’ attitude changes as circumstances change. Campaign strategists, like good military generals, adjust their strategies according to how the battle goes. We see that in the frequent changes made in the advertising messages of the candidates.

When Grace Poe ran for the Senate in 2013 as Grace Poe Llamanzares, her married name, she placed several notches below the 12th place in the polls. But when she dropped Llamanzares from her name and made known she was the daughter of Fernando Poe, Jr., her rank in the polls shot up to No. 5. When she ran for president in 2016, she emerged as topnotcher among the candidates in the early polls. When questions about her citizenship were raised, she fell behind Vice-President Jejomar Binay, who had been the expected winner before Poe’s name was floated.

Some doubters of the validity of survey results cite as evidence of faulty forecasting the case of two commissioned senatorial surveys of SWS conducted in the same period but presenting two different sets of probable winners. An SWS survey commissioned by Presidential Political Adviser Francis Tolentino showed Senator Cynthia Villar as the possible topnotcher in the 2019 elections. She was followed by Senator Grace Poe and Congresswoman Pia Cayetano, who were tied at 2nd place. Rounding the top five were Senator Koko Pimentel and former senator Lito Lapid, who shared the No. 3 slot.

Another SWS survey commissioned by the Lakas–CMD party forecast Senator Poe as the senatorial candidate in 2019 garnering the biggest number of votes. She was followed by Senator Villar then by Congresswoman Cayetano. In 4th spot was Senator Nancy Binay and No. 5 was former Senator Lito Lapid.

However, the Lakas-CMD-commissioned poll listed 39 candidates to choose from while the Tolentino-commissioned survey presented only 24 names. The more candidates to choose from the more widely scattered the votes cast would be. Votes that would have been cast for any of the top five in a list with 24 names went to any of the other 15 names in a list of 39 names.

Sometime in March a press release issued by the camp of former Special Assistant to the President Bong Go said that he (Go) received the approval of 47 percent of the respondents in a supposed SWS survey conducted in late February, placing him in solo third place in the ranking of senatorial candidates. The press release further said that the survey finding reflected a continuous improvement from his previous 5th to 6th place in January.

However, SWS President Mahar Mangahas said the survey was not theirs. The SWS posted on its website on March 7 a “Public Reminder” that it “again reminds the public that false reportage about SWS surveys is very common in the period leading up to an election. Everyone is enjoined to rely only on the SWS website (www.sws.org.ph) for its advisories, announcements, and survey reports released to the public.”

It noted that many surveys on preferences for national and local positions are being conducted by various groups in the run-up to the May 13 elections, that some of these groups are legitimate, while others are not.

Legitimate or not, survey results are now used as campaign materials. They are used to create a bandwagon effect. “Vote for the winnable candidate” is the exhortation of campaigners of leading candidates.

Social Weather Station and Pulse Asia have become mighty political forces. Their surveys have unduly influenced the outcome of national elections. Surveys have become significant factors in the political process. Instead of the campaign managers using survey results as basis in the formulation or revision of their campaign strategies, as they were originally meant for, they are now urging the voters to use the results as basis in the choice of their officials.

A religious sect’s endorsements are clear demonstration of the misuse of the poll results. Its endorsement of candidates is influenced by the survey results. It does not influence the voters’ decisions, as believed by politicians.

It has been my suspicion that the sect endorses only those candidates that the polls show to be likely winners. That suspicion has been proven to be correct in past elections. It endorsed early in the campaign period candidates who ranked steadily among the top five candidates favored by the survey respondents but delayed its endorsement of candidates ranked between 7th and 9th. By so doing, It sustains the myth that its endorsement assures victory.

I was in charge of the public opinion surveys at the Gallup Poll affiliate Robot Statistics, the first and at the time the only independent research firm in the country, in 1960. When the firm was commissioned to conduct a political survey, it was stipulated in the contract that the results of the survey would be for the private use of the client and were not to be released for public consumption. Robot released its election forecasts only after the polling places had been closed. The Daily Mirror (an afternoon daily) put out an Extra edition to announce the results of the survey after the close of polling booths, when the survey results could no longer be used to create a bandwagon effect.

That should also be the policy of SWS and Pulse Asia with regard to non-commissioned surveys (surveys they conduct on their own) — to announce the results of their surveys after the polling precincts have closed. I believe the heads of the two survey firms are both Political Science professors. They therefore should know that a large number of voters cast their vote like they are placing bets in a horse race — they vote for the odds on favorites.

Besides, survey results nowadays may not reflect the true sentiments of the respondents. As President Rodrigo Duterte has shown a disdain for criticism and opposition, as evidenced by the fate of Senators Leila de Lima and Antonio Trillanes,, Chief Justice Lourdes Sereno, media organizations Philippine Daily Inquirer and Rappler, and journalist Maria Ressa, survey respondents could also be afraid to say something not favorable to him, his policies, and his anointed senatorial candidates — Bong Go, Bato de la Rosa, and Francis Tolentino.

Interviews are conducted face-to-face. The respondent’s name and address are known to the interviewer. The interviewer’s true purpose may be dubious to the respondent. It should be noted that SWS has not conducted a survey on senatorial candidates since January or if it has, it has not released the results. I hope the other credible pollster Pulse Asia would follow suit, given that respondents’ answers may not be reflective of their true sentiments anyway.

Oscar P. Lagman, Jr. is a member of Manindigan! a cause-oriented group of businessmen, professionals, and academics.

oplagman@yahoo.com