Once more on the RMA
By Cean Angelo Reyes
Research Volunteer
(August 8, 2010) - Results of the Random Manual Audit (RMA) for the May 10 automated election have been released three months after it was begun. Media reports said the last mechanism of verifying the elections revealed that the last May 10 election was 99.6%, drawing praises from the Commission on Elections (Comelec). Before any final word is made on the RMA, it is important to make a run-through of its progress for the past three months.
Under RA 9369, an RMA is required to test if poll machines have counted the election results accurately. The RMA’s coverage was approved by Comelec: Five clustered precincts per legislative district, or a total of 1,145 precincts from 229 districts. These districts were to be chosen randomly through a “tambiolo” to be facilitated by Comelec's partner in the automated elections, the Parish Pastoral Council for Responsible Voting (PPCRV).
RMA is very important as people cannot verify the votes with the 76,000 precinct counter optical scan (PCOS) machines that changed the face of the elections. And the only way to assess the performance of the machines is by comparing it with the manual votes that could be read in the ballots.
RMA started on May 10 and the results should have come up a few days or a week after thereafter. According to Section 231 of BP 881:
“…the board of canvassers must complete their canvass within 36 hours in municipalities, 48 hours in cities and 72 hours in provinces.”
By May 14, all the poll personnel assigned to voting districts in Metro Manila that told this research volunteer that they already conducted RMA, adding they encountered no major problems. Having observed the RMA conducted in Caloocan, I can say that the manual counting could be done in less than six hours. In all, all the RMAs would have been finished in less than two weeks.
Unfortunately, this was not what happened. A few days after the elections PPCRV announced that it may take a month to finish the elections, mainly because of delays in sending the reports from different districts especially those far from Manila. It took three months before the results of RMA were revealed. And Comelec claims it was already done earlier but was delayed because of the discrepancies. Was this because of the unclear provisions on the RMA? And if this was the case, the report about these discrepancies should be revealed to the public as well and not just the results of RMA.
The long delay in the RMA has stirred speculations and uncertainties from the general public.
Meantime, the results of RMA show that the elections had a .4% margin of error. However, the error rate set by Comelec with Smartmatic was only at .005% and for IT experts an error rate of .07% is already unacceptable. So the question now is, how can Comelec be glad of the results when it didn't even pass the standards that they set? The 99.6% accuracy rate of the elections may seem impressive but it still isn't good enough based on the standards of Comelec and IT experts. There have also been unverified reports of two digit discrepancies in some areas that Comelec and PPCRV has to clear out.
The problem with RMA is that the process and the people are not ready to conduct it. There have been reports that BEIs who conducted RMA are not well aware of its process like the one recorded in Davao. A total of 350 barangays were excluded from the selection of RMA precincts because they were the least accessible. However, given that the RMA took almost three months to complete, then these barangays should have already been included given anyhow. Moreover, the PCOS machines in Hong Kong and Singapore were excluded from the RMA.
As of now, CenPEG’s IT experts are still analyzing the reports from the PPCRV in the spirit of credibility and accuracy. EU-CenPEG Project 3030