Undervoting in 2016 VP election: Will it help Marcos’ case?

Michael Bueza
Undervoting in 2016 VP election: Will it help Marcos’ case?
Historically, the behavior of voters with regard to undervoting has been essentially the same in the last two presidential polls, PPCRV information technology director William Yu says

MANILA, Philippines – The camp of Senator Ferdinand “Bongbong” Marcos Jr announced they are set to file an electoral protest questioning the results of the 2016 vice presidential elections. (READ: Marcos camp to file electoral protest by the end of June)

Among the issues they raised were the huge number of undervotes in the VP race, which Camarines Sur Representative Maria Leonor “Leni” Robredo won by a slim margin.

Undervoting happens when the number of choices made by a voter for a particular contest is less than the minimum number allowed for that race, or when no choice is made for a single choice contest. For example, undervoting occurs when a voter chooses less than 12 names for senator on the ballot, or when a voter abstains in the choice for president, vice president, governor, representative, and other local executive posts.

During the congressional canvassing of votes for president and vice president in May, Robredo’s lawyers repeatedly argued that undervoting is common during elections, and does not necessarily translate to poll fraud.

What’s the real deal with these undervotes? Are they cause for concern or are they a non-issue?

Undervote statistics

For the 2016 elections, based on results from 96.14% of precincts received by the Commission on Elections (Comelec) Transparency Server, there were at least 2,918,905 undervotes in the VP election. This is 6.68% of the total voters who actually voted on May 9.

The VP race undervotes are more than twice the 1,249,197 undervotes in the presidential race, which correspond to only 2.86% of the total voter turnout.

The 10 provinces with the highest and lowest undervoting rates (total undervotes divided by voters who actually voted) in the 2016 VP election are:

Top 10 Bottom 10
Province Undervoting Rate Province Undervoting Rate
1. Lanao del Sur 21.29% 1. Metro Manila 1.50%
2. Tawi-Tawi 21.09% 2. Rizal 2.34%
3. Masbate 19.83% 3. Cavite 2.54%
4. Negros Oriental 18.10% 4. La Union 2.73%
5. Northern Samar 15.93% 5. Laguna 2.95%
6. Biliran 15.11% 6. Ilocos Norte 3.01%
7. Sulu 14.82% 7. Benguet 3.27%
8. Zamboanga del Norte 14.47% 8. Bulacan 3.32%
9. Siquijor 14.33% 9. Ilocos Sur 3.65%
10. Leyte 13.94% 10. Zambales 3.68%

Among the top 10 provinces, 3 are in the Autonomous Region in Muslim Mindanao (ARMM): Lanao del Sur, Tawi-Tawi, and Sulu. Robredo won in the first two provinces, while Marcos won in Sulu.

On the other hand, Metro Manila posted the lowest undervoting rate, at only 1.5%. It is followed by neighboring provinces Rizal and Cavite, then Laguna with the 5th lowest rate. Bulacan was at 8th.

Three of the 4 provinces in Marcos’ “home base”, the Ilocos Region – Ilocos Norte, Ilocos Sur, and La Union – are also in the bottom 10. (READ: How regions voted for Robredo, Marcos in 2016 VP race)

Compared to 2010

Historically, however, the behavior of voters with regard to undervoting has been essentially the same in the last two presidential polls, according to information technology director William Yu of the Parish Pastoral Council for Responsible Voting (PPCRV).

In 2010, Yu said, 4.94% of voters undervoted in the VP race, also more than twice the undervotes in the presidential race (2.28%).

Based on data from the 2010 Comelec transparency server, which was obtained by Rappler from PPCRV, and which reflects 90.2% of total precincts, these are the provinces with the highest decreases and increases in undervoting rates in 2016 compared to 2010:

Top 10 Decreases in Undervoting Rate (vs 2010) Top 10 Increases in Undervoting Rate (vs 2010)
Province % Point Difference Province % Point Difference
1. Sulu -17.81 1. Davao Oriental 3.52
2. Basilan -6.15 2. Capiz 2.09
3. Mountain Province -3.98 3. Zamboanga Sibugay 1.98
4. Abra -3.67 4. Compostela Valley 1.91
5. Ilocos Sur -3.61 5. Oriental Mindoro 1.87
6. Samar -3.32 6. Agusan del Norte 1.80
7. Biliran -3.27 7. Sultan Kudarat 1.78
8. Leyte -3.13 8. Marinduque 1.76
9. Tawi-Tawi -2.91 9. Palawan 1.72
10. Misamis Occidental -2.83 10. Batangas 1.44

Sulu posted the biggest decrease, from a 32.63% undervoting rate in 2010 down to 14.82% in 2016. On the other hand, undervoting increased the most in Davao Oriental, from 8.86% in 2010 to 12.38% in 2016.

Shown in the map below are the differences in undervoting rates per province in 2016 versus 2010. 

Notice that undervoting went down in Marcos’ and Robredo’s home regions – Ilocos and Bicol, respectively. It went down in 4 of 5 ARMM provinces, while it got worse in many provinces in Luzon and parts of Western Visayas and Southern Mindanao.

Meanwhile, Yu noted that undervoting rates rose slightly in Regions II and III, where Marcos led in the vote count. Yu argued this “might not support his case entirely” if the Marcos camp insists on undervoting issues.

Click on each province on the map below for more details.

First map: The darker the shade of blue, the bigger the under-voting rate in 2016. Second map: Provinces in green denote a decrease in under-voting in 2016 versus 2010, while provinces in red denote an increase.

 Under-voting rate (2016)

 % pt. difference (vs. 2010)

Looking at precinct-level data, a majority of clustered precincts (CPs) in 2016 (69,969, or 77.2% of all precincts in the partial, unofficial count) posted undervoting rates of less than 10%.

This is roughly the same as the 2010 VP election, when at least 76.91% of precincts (53,049 precincts in the Comelec Transparency Server) had undervoting rates of less than 10%.

In 2016, at least 14,601 precincts had 10% to 19% of voters undervoting, while only 191 precincts logged undervoting rates of over 50%.

‘Lack of depth’ in candidate pool

According to election lawyer Rona Caritos of the Legal Network for Truthful Elections (Lente), undervoting happens mainly because of “lack of depth” in candidate choices.

“You don’t always equate that to cheating. There are really many voters who undervote because they don’t like any of the choices,” Caritos said. “It’s not rare in our election.”

“It is a valid transaction,” added PPCRV’s Yu. “There are some voters who purposely undervote, who really choose not to vote [for someone in a certain post].”

This sentiment even led to a petition filed before the Supreme Court to include a “none of the above” (NOTA) option on the ballot. The SC dismissed this petition in January 2016.

How voters view the vice president’s role in government could also be another factor, said Caritos. “They may be more interested in the presidential race. They might also think na walang ginagawa ang VP (the VP is doing nothing), that it’s not an important position.”

To lessen the occurence, or even do away with undervoting, Caritos said Lente is looking into a “single ticket” system where voters would choose one ticket or tandem for president and vice president, instead of voting for them separately.

As for the distribution of undervotes, Caritos said that low undervoting in Ilocos and many areas in Marcos’ so-called “Solid North” is expected.

But undervoting in the Visayas was a “surprise”, said Caritos, given that it is the bailiwick of Liberal Party (LP) standard-bearer Manuel “Mar” Roxas II. The region should have also produced votes for Robredo.

Minimal voter complaints

For poll fraud allegations to prosper in relation to undervotes, Caritos argued that these should be backed up by reports of vote-counting machines (VCM) reading ballots wrongly.

One way to corroborate these is through complaints involving voter receipts, which should have reflected the voter’s choices on the ballot, as read by the VCM. The names of candidates that a voter shaded on the ballot and the candidates’ names printed on the voter receipt should have matched. If they didn’t, the voter should have complained to the Board of Election Inspectors (BEI) manning the polling precinct during elections.

Poll watchdogs like Lente were expecting such complaints on May 9, “but we did not receive a lot of it. In fact, what became an issue on election day were reports of VCMs breaking down,” said Caritos.

“So we’ll assume that the VCMs counted the ballots properly,” she added.

It would also be difficult to prove that cheating happened by changing election results during electronic transmission, Caritos continued. “Because you have election returns and Certificates of Canvass (COC) printed out. There are many checks [in the automated election system], if ever the votes would be changed.”


Caritos added that poll cheating allegations, if substantiated, could trigger a manual recount, at least in certain areas. For instance, “part of the procedure of the Comelec is, you must establish where the poll fraud possibly happened… Do that and then recount could really happen,” she argued.

This strategy is commonly observed in electoral protests for local races, she said.

In the case of undervotes, Caritos said, “You should prove that those undervotes were really intended for you, but weren’t counted by the VCM.”

She emphasized, however, that in the event that Marcos files an electoral protest, the procedure of the Presidential Electoral Tribunal (PET) could be different.

The PET has sole jurisdiction over protests in the results for the presidential and vice presidential elections. It previously handled the protest of Loren Legarda against former vice president Noli de Castro’s victory in 2004, as well as Mar Roxas’ electoral protest over then vice president Jejomar Binay’s win in 2010. – Rappler.com

Add a comment

Sort by

There are no comments yet. Add your comment to start the conversation.

Michael Bueza

Michael is a data curator under Rappler's Tech Team. He works on data about elections, governance, and the budget. He also follows the Philippine pro wrestling scene and the WWE. Michael is also part of the Laffler Talk podcast trio.