The Tale of the Survey: Did SWS Get It Right in the Previous Elections?

SWS12

Less than two weeks before the election, the poll numbers are getting much more defined. Davao City Mayor Rodrigo Duterte’s survey ratings are beginning to narrate a dramatic story arc—that the presidency is destined for someone.

But before his rise in the survey, when Grace Poe was still the frontrunner, surveys were being criticized for being false. Well, you can accuse survey firms of being biased, unrepresentative, even propagandistic. But here’s one thing. Most—if not all—of the time, they get it right. At least, for SWS.

Social Weather Stations (SWS), founded in 1985 and one of the top survey firms especially during election season, has correctly “predicted” the winner of all presidential and vice-presidential elections since 1992.

1992 General Elections

Photo from wikipedia and Manila Bulletin.

Photo from Wikipedia and Manila Bulletin.

You heard that urban legend that Miriam Defensor Santiago should have won the 1992 elections? That could be true if election rigging really did happen. However, it must be known that in the last pre-election survey conducted by SWS from April 26 to May 4 of 1992, Fidel Ramos was already the frontrunner, with 18% rating over Defensor Santiago’s 16%. The percentage of undecided voters a week before the elections was at an alarming 34%, and the exclusive news article in the frontpage of The Manila Chronicle published two days before Election Day could have made people decide between Ramos and Santiago only. Actual election results gave Fidel Ramos 24% of the votes, the lowest plurality in Philippine electoral history. MDS got 20% of the votes.

1992 pres

For the vice-presidency, Erap Estrada won with 33% of the votes. He was also the frontrunner in the same pre-election survey with 19% survey rating. Fernan Marcelo was a distant second. Thirty-nine percent of the SWS respondents were undecided at the time of the last pre-election survey.

1992 vp

With 15.8% of the votes for vice president invalid, the 1992 vice presidential race ranks the highest percentage of invalid votes for any of the top two posts in the country in post-EDSA.

1998 General Elections

Photo from Inquirer.

Photo from Inquirer.

Then vice president Erap Estrada won the presidency by a landslide in 1998. He got 10.7 million votes or 40% of all the valid votes cast, just 4 percentage points behind his final pre-election survey rating of 36% conducted on May 2-4, 1998, a week before Election Day. Jose De Venecia, then Speaker of the House, was a far second, both 16% in the pre-election survey and at the actual election results.

1998 president

This would be Miriam Santiago’s second bid for the presidency, but she only placed a disappointing 7th in the race with 3% of the votes. She was a senator when she ran in 1998. Imelda Marcos, who was running for a second time, withdrew at the last minute.

The vice presidency was also an easy win for then-senator Gloria Macapagal Arroyo. Her final pre-election survey rating was 48%, just two percentage points less of the actual voting percentage she received. Then-senator Eduardo Angara was a distant second with 22% of the votes.

1998 vp

2004 General Elections

Photo from World News.

Photo from World News.

The year 2004 was a little tricky, and perhaps the election where SWS might have missed out on something. The final pre-election survey conducted from May 1 to 4 of 2004 showed incumbent president Gloria Macapagal Arroyo at a comfortable lead, with 37% survey rating over movie icon Fernando Poe Jr. with 30%. Twelve percent of the respondents were undecided.

Come Election Day, the results were closer than people thought. Arroyo got almost 12.9 million votes, or 40% of the valid votes cast. FPJ received almost 11.8 million votes, just three percentage points short of Arroyo’s. SWS, again, “predicted” it right, with their final pre-election survey ranking matching the ranking of the actual election results accurately. GMA would later apologize for the Hello Garcia Scandal that was connected to the alleged election rigging of 2004, and perhaps, if proven, SWS could have been wrong at the time.

2004 pres

For the vice presidency, both senators and broadcast media colleagues Noli De Castro and Loren Legarda were neck and neck at the final pre-election survey. De Castro had 43% while Legarda has 39%. Fourteen percent of the respondents were undecided. Election results put De Castro at 50% of the valid votes cast; Legarda was less than a million votes or three percentage points behind. Up to this day, Legarda believes that she should have won that one.

2004 vp

2010 General Elections

Photo from Twitter.

Photo from Twitter.

SWS aced it in 2010. Incumbent president Aquino had the most surprising narrative in 2010. His name was not even in the pre-election surveys in early 2009, but her mother’s death has made him the top ranking official of this country. His final pre-election survey rating was 42%, and it matched the actual election results. Fifteen million voters voted for him. Former president Erap Estrada, whose candidacy was also surprising as much as it was constitutionally questionable, was at second place with 26% votes. It was a six percentage point increase from his final pre-election survey rating of 19%.

2010 president

Villar, who could have been president had Cory not died, was a disappointing third with 15 percent of the votes, a four percentage point-decline from his 19% final pre-election survey rating.

Interestingly, the vice presidency was Mar Roxas’s to lose. In September 2009, he was the kind and sincere politician who took a step back to give way for what the country clamors who was Aquino. But alas, it was the vice presidency that got away. Roxas was the pre-election survey frontrunner until Binay’s attachment to the Cory legacy has given the latter a steady increase in the surveys. At the final pre-election survey, Binay and Roxas were statistically tied at 37%. Even the Iglesia ni Cristo endorsement did not save Roxas from the loss he did not see coming.

2010 vp

Actual election results put Binay and Roxas at 42% and 40% of the votes. Binay was less than 800 thousand votes ahead. Eight percent of the votes cast were invalid, calling Roxas to file a complaint.

SENATORS

SWS survey also gets the senators mostly right. Take a look at 2004. The final pre-election survey correctly matched with the list of senators who will be included in the Magic 12, only their rankings were different.

Slide1

Mar Roxas and his massive hit Mr. Palengke campaign ad secured himself at the top spot both at the final pre-election survey and at the actual election results. Bong Revilla and Nene Pimentel were also at second and third place both in the survey and actual election.

Photo from mati.gov.ph.

Photo from mati.gov.ph.

(Take note that the percentages vary, though. Mar Roxas’s final pre-election survey rating gave him 42% of the respondents’ votes; actual election results gave him 58% of the votes cast.)

In 2007, SWS got 10 of the 12 senators right. Indeed, 2004 VP loser Loren Legarda was reelected at the senate with 63% of the votes, a four percentage point-increase from her final pre-election survey rating.

Slide2

PGMA administration bets Ralph Recto and Juan Miguel Zubiri, both at the bottom half of the final pre-election survey, fell from the Magic Circle of 12 at the actual election. Recto, from 7th place, ranked a depressing 14th place, while Zubiri, who was initially proclaimed a winner, was later reassigned to the 13th spot.

Antonio Trillanes had the most surprising victory that time. SWS placed him at 15th to 16th place, but he managed to place 11th at the actual election with 38% of the votes. Koko Pimentel, from 14th in the SWS survey, placed 12th after a recount of votes.

In 2010, SWS could have never been more accurate with the final pre-election survey ranking. Except for Ralph Recto and Tito Sotto, they got the ranking right (in exact order). The final pre-election survey ratings of Bong Revilla, Jinggoy Estrada and Miriam Defensor Santiago also matched with the actual election voting percentage.

Slide3

And lastly, the most recent 2013 senatorial election had perhaps the most surprising result in a long time in terms of comparing it with the survey. Grace Poe, who only ranked fifth at the May 2-3 SWS survey, topped the election with 51% of the votes cast. The unstoppable Loren Legarda during the campaign season fell to second place. It is interesting to note that the names of the senators who belonged to the Magic 12 matched with the final pre-election SWS survey, although the order was quite different, especially with Cynthia Villar and JV Ejercito who ranked lower than expected.

Slide4

CONCLUSION

Of course, this fact check does not erase the doubt we have in surveys, and it shouldn’t. A number of studies have suggested that surveys could affect or condition the mindset of the viewers on who to vote for, and some methodologies of surveys could actually be problematic. Case in point is the recent SWS mobile survey that was criticized for its method.

Furthermore, reservations must be put in instantly believing commissioned surveys, as the results could possibly be manipulated to favor a candidate (I mean, we never know).

This fact check does not prove a lot a things, but this was written to tell you that despite its sometimes sneaky and unbelievable results, surveys could actually be accurate given a scientific methodology. In fact, they are far more reliable than Facebook and Twitter polls. Do not just dismiss the results because your belief or disbelief could be relative. You have your cognitive bias and unempirical, phenomenological experience that you falsely regard as universal truth. You think the results do not equate what you see, but have you really been offline, in the streets, outside your socioeconomic class? Believe it or not, there’s a world moving and working their ass off, and it does not adjust for you just because you’re in denial or sheltered.

Sources:

http://www.sws.org.ph/mahar.htm

https://www.sws.org.ph/erp-com.htm

https://www.sws.org.ph/index_med-indx.htm

http://www.sws.org.ph/pr070510.htm

https://www.sws.org.ph/pr20100509.htm

https://www.sws.org.ph/pr20130509.htm

Advertisements

Comments

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Baguhin )

Twitter picture

You are commenting using your Twitter account. Log Out / Baguhin )

Facebook photo

You are commenting using your Facebook account. Log Out / Baguhin )

Google+ photo

You are commenting using your Google+ account. Log Out / Baguhin )

Connecting to %s