How trustworthy is the random manual audit (RMA)?

YouthVotePhilippines: With the technical glitches in yesterday’s first automated polls, the random manual audit (RMA) becomes more crucial in ensuring the credibility of election results.

The Commission on Elections has released Resolution 8837 on the conduct of the RMA after rejecting calls from various groups to conduct a parallel manual count last April 29.

In particular, while the resolution provides for the composition and duties of the RMA team (RMAT) and the general instructions of conducting the audit, the acceptable margin of variance between the Precinct Count Optical Scan (PCOS) count and the audit count remains uncertain.

“It’s crucial that we first determine what’s the acceptable tolerance for error between the automated and manual counts, before we conduct any form of audit,” said Jaime Garchitorena, YouthVotePhilippines information technology strategist.

“What if there is a variance but it’s not that significant to the results? We can’t decide which variance to accept after the fact. We should have clear rules on that even before voting begins,” he said.

Issue still about trust

Based on the proposal of the Parish Pastoral Council for Responsible Voting (PPCRV), the RMA will be done immediately upon the close of elections in five clustered precincts for each of the 229 congressional districts.

RMATs will be deployed in polling precincts randomly selected using a tambiolo with numbered balls. The audit will be conducted in 1,145 clustered precincts nationwide once the PCOS machines have been shut down.

The positions that will be manually counted for the audit include the president, vice president, district representative, governor and mayor.

The National Citizens’ Movement for Free Elections (Namfrel), however, earlier recommended that the RMA cover 5 percent of all PCOS machines per congressional district instead of just five clustered precincts. Originally, Comelec only planned for an audit of one machine per district.

It also suggested that “the audit process should include a review of the hash codes of the PCOS machines for comparison with source codes stored in Bangko Sentral (Central Bank).”

Despite Comelec’s non-accreditation, Namfrel still plans to conduct its own parallel count to ensure credible elections. It said in a statement, “Both the manual and automated counts should be made available immediately to the public for transparency purposes.”

In his column in the Philippine Daily Inquirer last February, Social Weather Stations president Mahar Mangahas wrote that former ambassador Henrietta de Villa of PPCRV had approached him to become a guest consultant, pro bono, in the RMA technical working group which met in November 2009.

Mangahas said the RMA process will have scientific meaning only if its individual audits are completely trustworthy.

“The question of how many PCOS machines to audit cannot even be addressed by statistical principles unless integrity is assured for each and every audit,” he said.

“If trust is the issue then it cuts both ways. To be objective about it, if you don’t trust the Comelec the same skepticism should apply to groups conducting the audit,” Garchitorena said.

“In more concrete terms, any person who offers to handle my ballot physically should be someone I trust completely,” he added.

To remove any suspicion of vested interest, Garchitorena suggested in a public call last month that all proponents of any manual count or audit be named and made to promise to decline any offer to sit in the next government at least for the next three years.

“Even the auditors should be kept in check,” he said. “How would it look if an auditor was found to have an unexplained sum of money in his account after proclaiming a winner in a controversial local election? In something as critical as this, it’s important to their credibility that they will not personally benefit from the activity.”

Cheating the audit improbable

Some groups and local candidates have expressed fears that excess ballots may be used for cheating the audit, but Comelec spokesperson James Jimenez said there is “no opportunity to insert unused ballots” into the machine.

The number of ballots per clustered precinct matches the number of voters in the precinct’s list.

“Kasama sa procedure natin ay ina-announce ng BEI (Board of Election Inspectors) na meron syang unused ballots, kasi gugupitin nya yun pagkatapos (It’s part of the procedure for the BEI to announce that there are unused ballots, because they will cut it in half after the voting period ends),” he said.

Audit rules state that in case the number of ballots that will be subjected to the audit exceeds the number of votes found in the Minutes of Voting and Counting certified by the BEI, the resolution requires for an investigation for “any possible reasons” of the discrepancy, such as spoiled or rejected ballots.

But in the event that no probable reason is identified, “all the ballots shall be returned in the ballot box and thoroughly mixed by the (Board of Election Inspectors) Chairman.” The BEI chair will then “randomly take out from the ballot box the number of ballots equal to the excess and place them in the RMA envelopes for excess ballots.”

“Without a clear acceptable level of variance, a situation where this happens can cause commotion if because of the random pick of ballots to be counted, the audit count differs by some margin with the PCOS count,” Garchitorena said.

Proclamation not affected by audit

With Comelec’s rejection of a parallel manual count, Namfrel has suggested for an audit of votes before the results are transmitted. Several groups have also expressed opposition to proclaiming candidates as winners, especially in local posts, before a manual tally confirms the PCOS count.

Jimenez, however, said while the audit results will not affect proclamation, they can be used as grounds to file an election protest. “Pwedeng gamitin yun bilang grounds to challenge proclamation (Audit results can be used as grounds to challenge proclamation),” he said. “Kasi malinaw naman, meron na kagad colorable doubt, na mukha talagang may nangyaring mali (Because then it would be clear that something went wrong and that would cause doubt).”

“But of course that will trigger yung sinasabi nating (what we call) discovery of root causes,” he said.

Comelec has been firm that results of the audit should not hinder the proclamation of winning candidates based on the PCOS count.

Garchitorena agreed with proclaiming without the audit results. He said that if audit or manual count results are given much weight without a clear provision on an acceptable discrepancy when compared to the PCOS count, “we’re setting up a no proclamation scenario.”

“Especially since no one is really pushing for clarifications on the constitutional engine that searches for a leader in the absence of duly elected officials or electoral power vacuum,” he said.

Garchitorena, however, cautioned the Comelec from comparing the ease of using the Automated Teller Machine (ATM) with the automated election system (AES) to illustrate its reliability.

“ATM machines still fail today. The key to accepting the ATM was the stability and trust in the banks which operate the machines,” he said. “People know that their money is safe and that if there is a discrepancy in their receipt and what they got, they can complain.”

Instead of saying the machine will run perfectly, Garchitorena said the Comelec should explain that the improvement of the AES comes with the constant use and adjustment of the PCOS machines, including successful testing.


Leave a Reply

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

You are commenting using your account. Log Out /  Change )

Facebook photo

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

Connecting to %s