Who checks cheating done before the voting?

CHEATING: It is best that the possibility of automated cheating and of failure of elections in May 2010 is discussed this early, so the weaknesses of the computerized system are exposed and corrected.

We like to think that it is in this spirit that many wide-awake Filipinos have started to examine more closely the P7.2-billion system that the Commission on Elections and its technical partner, Smartmatic-TIM, will use for the first time in the national elections next year.

Reacting to our Postscript of Sept. 15 on the subject, there were more readers saying that failure of election was more remote a possibility than massive cheating.

But it seems that everybody is so engrossed with automation that the pre-voting stages are not getting the critical attention they deserve.

Who is working to prevent the usual pre-voting dirty tricks — the delisting of legitimate voters, scrambling of names and precincts, padding of registries, vote-buying, spreading of false and misleading information, etc.?

*      *      *

SPECIAL OPS: Reader Vic Endriga, an old election hand, said: “Election failure would be bloody and laborious to make happen. Maybe on a limited scale, there could be a failure in isolated areas because of intense fighting and maneuvering among local politicians, but on a nationwide scale, it is close to impossible.”

On possible cheating, he says: “With vote-counting automation, political operators will now require fewer people to undertake so-called ‘special operations.’ It will be difficult to prove cheating, because voters no longer need to write the names of the candidates or the party list groups that they support. All they need to do is mark or shade the oval opposite the names of their favored candidates.

“It is now easier and faster to prepare substitute ballots because there will no longer be handwritings to examine or check if the ballots were prepared by the same persons. The difficult aspect will be how to switch the ballots before they are counted or change the score before it is transmitted.

“Another option is to corrupt Smartmatic personnel, through the contacts of Comelec political operators, so that preparing of the machine-counted ballots or the transmitting of the results can be manipulated to fit the requirements of the favored candidate(s).”

*      *      *

REJOINDER: From Smartmatic-TIM comes a rejoinder to our Sept. 15 Postscript that discussed fears of possible cheating and a failure of elections. The firm’s public relations manager/spokesman Gene Gregorio said among other things:

1. On the clustering of poll precincts — Each cluster is designed based on the local environment, with the Precinct Count Optical Scan (PCOS) machine at the center and visible to all parties. A secrecy folder covers the ballot until it is fed into the machine. The vote is cast and the machine takes picture images of both sides (scanned).

2. On “time and motion” — The correct arithmetic for analysis involves a minimum of eight hours on 100-percent voter turnout, that is 1,000 voters in a clustered precinct with a minimum of 10 seats per precinct. The actual casting of votes (feeding ballot into machine) takes only a few seconds.

Assume 100-percent voter turnout (the most conservative position to take) on Election Day, and 50-percent machine utilization (this means each machine would be in actual ballot-scanning and vote-counting operation for only 5.5 hours out of the 11 voting hours).

On that assumption, Comelec did an informal time-and-motion study, and reported an average of eight seconds for a complete ballot-casting cycle. If we double this number and use 15 seconds to cast a single ballot, two ballots can be cast per minute, or 120 ballots per hour. In 11 hours, 1,320 ballots can be cast and counted per machine.

This gives Comelec a safety factor of 1.32 over the maximum number of voters per clustered precinct even assuming a 100-percent voter turnout.

It is erroneous to assume that only one voter at a time will be permitted to fill up a ballot in the precinct. Voters inside the precinct will prepare their ballots at the same time. The queue is only for feeding the ballots into the counting machine.

3. On automated system’s possibly not engendering trust among voters — We offer the ARMM pilot automated elections as proof that our people are techno-friendly. DRE (Direct Recording Electronic) machines were used there, and voters found them very easy and convenient to use.

In the 2010 national elections, PCOS (Optical Mark Reader-based) machines will be used. OMR is paper-based and is not a major departure from manual voting. Even voters who cannot read or write will be able to identify popular names (the candidates of their choice), without assistance.

4. On troubleshooting — There will be a technician per two PCOS machines — for a total of over 40,000 technicians. They will be supported by a 600-seat call center, apart from their municipal and provincial supervisors. Training is scheduled in the next few weeks.

5. On possible switching of memory cards — The cards, where election data will be saved, will only store information from that day’s election. The machines will not load any external info on the card. Thus, even if a potential election felon were to pass through the phalanx of physical security comprising poll watchers, as well as civil society and political parties’ representatives, his memory disk (which is non-standard and not commercially available in Philippine retail stores) will be rejected outright by the machine.

*      *      *

ePOSTSCRIPT: Read current and old POSTSCRIPTs at www.manilamail.com. E-mail feedback to [email protected]