When the polls get it wrong: Rethinking election surveys

MANILA, Philippines – In democratic societies, election surveys are treated like political weathervanes—offering a glimpse of how the wind might blow come election day. These surveys guide political strategists, shape public opinion, and even influence campaign financing.

Republic Act No. 9006, or the Fair Election Act, defines election surveys as measurements of voter opinions and perceptions regarding a candidate’s popularity, qualifications, platforms, or other matters of public interest during the campaign period—including voter preferences on candidates or key issues.

READ: Mother of an illegitimate child MUST sign the birth certificate

The law further requires that any person or entity who publishes a survey must disclose the following information:

1. The name of the person, candidate, party, or organization who commissioned or paid for the survey;
2. The name of the polling firm or organization that conducted the survey;
3. The period the survey was conducted, the methodology used, the number of respondents, the areas covered, and the specific questions asked;
4. The margin of error; and
5. Contact details for obtaining a written report of the survey.

Despite these transparency safeguards, recent election cycles—both locally and internationally—have exposed a growing problem: the surveys are increasingly off the mark.

America’s 2024 presidential election: A surprise landslide

In the 2024 US presidential elections, many polls showed a tight race between Donald Trump and Kamala Harris. Analysts even described it as “too close to call” or a statistical tie.

But the results told a different story. Trump won all seven swing states, securing 317 electoral votes versus Harris’ 226. He also garnered 77,302,580 votes, defeating Harris’ 75,017,613.

The last time a Republican candidate won both the Electoral College and the popular vote was more than 20 years ago.

The 2025 Philippine midterms: Polls vs. reality

Here at home in the Philippines, the just-concluded 2025 midterm elections have sparked controversy—not because of the results alone, but because of how far they strayed from pre-election survey projections.

In what may have been the final survey conducted (May 2–6) by SWS, the projected top 12 senatorial candidates were: Erwin Tulfo, Bong Go, Tito Sotto, Lito Lapid, Ben Tulfo, Ping Lacson, Abby Binay, Bato Dela Rosa, Camille Villar, Pia Cayetano, Bong Revilla, and Imee Marcos.

Just outside the “Magic 12” were Manny Pacquiao, Willie Revillame, Benhur Abalos, Bam Aquino, and Kiko Pangilinan. (https://www.inquirer.net/440421/erwin-tulfo-leads-final-sws-survey-on-senate-race/)

These rankings were consistent with other surveys in the weeks leading up to election day.

But after the official canvassing, Bam Aquino, Kiko Pangilinan, and Rodante Marcoleta placed 2nd, 5th, and 6th, respectively—surprising many. Meanwhile, high-ranking survey favorites like Bong Revilla, Ben Tulfo, and Abby Binay failed to enter the top 12.

What went wrong?

The disparity between survey predictions and actual results has reignited questions about the reliability of election surveys. Several theories have been proposed:

1. The surveys weren’t wrong—timing was

Most pre-election surveys (except SWS’s) were conducted about three weeks before May 12. Major developments in the final week—such as the Iglesia ni Cristo’s endorsements, Vice President Sara Duterte’s support for Imee Marcos and Camille Villar, and last-minute campaigning—may have significantly swayed voters.

On the other hand, candidates who felt safe due to high survey rankings may have relaxed in their campaign efforts while those ranked 12th to 15th doubled their efforts.

2. The surveys weren’t representative

Surveys rely on small, selected samples. If these don’t reflect changing voter demographics, results may be skewed. Bam Aquino’s strong performance, for instance, was credited to Millennial and Gen Z voters—who may have been underrepresented in survey samples.

3. The surveys were manipulated

Surveys can be commissioned to shape public perception. Candidates leading in polls may be seen as more viable, attracting more donors and support. While this tactic is known, it appears less likely in the 2025 elections, as all the different surveys were generally off the mark.

4. Respondents did not disclose who they were voting for

Survey respondents did not disclose who they were really voting for.

During the 2016 presidential elections, there were many instances where people who were supporters of certain candidates would verbally attack those who declared that they were supporting the other candidate, particularly those voting for former President Duterte.

Many friendships were lost during the last 2 presidential elections where friends unfriended friends on Facebook (now Meta) when they realized the other was voting for former President Duterte and then President Marcos, Jr.

When survey takers ask random people who they are voting for, the respondents are also hesitant to give their real answer if they feel that the people around them are voting for different candidates as they don’t want to be judged or ostracized.

5. “Bobotante” no more?

For years, critics lamented the “bobotante”—voters who allegedly sell their votes or follow trends blindly. But the 2025 results suggest a more discerning electorate. If voters were deliberately withholding preferences from surveyors, it may be a sign of sophistication, not ignorance.

The “bobotante” may be fading.

Impeachment drama on the horizon

As the election dust settles, a new political storm brews: the impeachment of the Vice President.

With the current Senate composition, will two-thirds vote to convict her? If they do, will they impose the additional penalty of lifetime disqualification from public office, thereby derailing any 2028 presidential bid ?

Will the Vice President resign if conviction seems inevitable, citing political persecution and an unfair trial? If she does, will Congress and the Senate still proceed to render judgment?

The 2025 elections may be over, but the political spectacle is far from over.

(The author, Atty. John Philip C. Siao, is a practicing lawyer and founding Partner of Tiongco Siao Bello & Associates Law Offices, an Arbitrator of the Construction Industry Arbitration Commission of the Philippines, and teaches law at the De La Salle University Tañada-Diokno School of Law. He may be contacted at [email protected]. The views expressed in this article belong to the author alone.)