2022 Philippine Elections

[OPINION] Surveys as oracles

Melba Padilla Maggay
[OPINION] Surveys as oracles
'It is worth asking if the surveys of Pulse Asia are accurate enough, or worse, willingly used to carry uninformed hordes of bandwagoners to the polls'

The past three decades have seen the surfacing of election surveys as a tool, not just for pulsing voting preferences, but for establishing trends that unduly lead the public towards the idea that the front-running candidates are the horses to bet on. 

This concern had been raised time and again based on the culture’s propensity to back whoever is llamado: “Huwag mong sayangin ang boto mo sa matatalo,” it is said. In western societies, opinion polls function merely as one empirical tool for sensing where the political wind is blowing. In this culture, what in truth is merely descriptive of a moving picture at any given moment is now seen as predictive, used as prophetic oracles by those who sponsor and publish them with a given aim in mind. 

Mahar Mangahas, founder of the Social Weather Stations, the poll firm that has had the longest track record for credible surveys, has found in their previous exit polls that surveys actually have had little influence on the people’s voting behavior. Nevertheless, in these days where you have a massive disinformation campaign not just in social media but among grassroots communities, fueled by the money and machinery of the supposed frontrunner, it is worth asking if the surveys of Pulse Asia are accurate enough, or worse, willingly used to carry uninformed hordes of bandwagoners to the polls.         

Must Read

[Newsstand] Dynamic race, static survey?

[Newsstand] Dynamic race, static survey?

Statisticians at the University of the Philippines have called out Ronald Holmes, president of Pulse Asia, on the soundness of its sampling and methodology. Holmes admitted that their sampling has been mostly among the D and E classes, which comprise 58% of the population. However, in both their February and the recent March survey, the sampling was not in accordance with the actual sizes of the country’s demographics. 

Class A and B, which stands at 1.1 million or 1%, was not represented at all and Class C, with 45.1 million or 41%, was reduced to a pro-ration representation of 8%. Then they added the 1% of the A-B class and the shaved 33% of the C segment to the D and E pro-ration. The proportional representation was thus: 92% from the D and E sectors, 8% from the C sector, and 0% from the A-B sector. 

Statisticians say that if the February survey was according to an accurate sampling based on demographic size, Marcos should have registered 50% and not 60%, and Robredo 23% and not 15%. In the latest March survey, where Robredo increased by 9% and Marcos decreased by 4%, the numbers should have been Marcos 46%, and Robredo 32%.

What these numbers tell us is that the supposed Marcos win is not a done deal; Robredo may yet come up from behind, as she did in the last vice presidential elections. Mangahas has cautioned that surveys are time-bound; half of the voters make up their minds only towards the runup, about two months or so. 

It should be recognized that surveys do not function in this culture in the same way that they do in the West. If the preponderance of data is among the urban poor, whose lack of resources is cynically taken advantage of by unscrupulous candidates, and who are rendered vulnerable to disinformation since access to accurate information is constrained by their circumstances, it is very likely that results will be skewed towards the motivating goal behind the survey.

In a culture where information, true or false, usually travels through a dense network of relationships, as evidenced by the “Marites” phenomenon, people are more likely persuaded by family and friends rather than by media. Opinion leaders in communities have a great deal of power among those who delegate their discernment to their judgment, since they are supposed to be more informed. Media and its chattering classes can only try to create the right atmosphere. But in these days of troll farms this gets trumped by the toxic belligerence of warring camps on social media.   

[OPINION] Surveys as oracles

The French sociologist Jacques Ellul, writing in the ’70s, has early warned that in a technologically-mediated social environment, we are actually living in what he calls “shadows.” What we know is no longer derived from real lived experience but from anonymous authorities in news rooms or, in these days, from the shadowy provenance of surveys and purveyors of slanted opinions disguised as “facts,” aided by those who have ways of manipulating algorithms in social media.

So far, the surveys of Pulse Asia do not cohere with the visual evidence and actual experience of teeming crowds in rallies organized by volunteers for Robredo and Pangilinan. We are witnessing the palpable rise of a new form of people power, a groundswell of popular support which tends to put to question the results of the surveys. It is also worth asking why the SWS to date has not found any sponsor for comparative surveys. I have been informed upon personal query that in the 2019 senatorial elections, no agency has been prepared to fund an exit poll by SWS as in the past. An exit poll is usually closer to the actual results of an election.   

In Singapore and Italy, publishing results of pre-election and exit polls is banned. Given the predilection of this culture to go where the crowd goes, it may be necessary to consider a similar move. Cultural factors such as this will need to be considered in later efforts at electoral reforms. – Rappler.com

Melba Padilla Maggay is President of the Institute for Studies in Asian Church and Culture.