PCOS did not improve voting — ex-poll official

Elections were automated for four aims: (1) improve accuracy of the vote count and result tabulation, (2) reduce, if not remove, cheating, (3) make the process more, not less, transparent, (4) speed up the process.

Did the precinct count optical scanner (PCOS), the automation technology supplied by Smartmatic of Venezuela, meet the objectives? “That’s what the Comelec should ask itself,” info-tech expert and former commissioner Gus Lagman says. But since the poll agency is skirting the P9-billion question (the PCOS lease-purchase price), he answers it for them. All this, in a paper to the Joint Congressional Oversight Committee about the May 2013 balloting.

*      *      *

Accuracy. No mock election conducted by Smartmatic ever yielded accuracy of 99.995 percent or better, Lagman says. The Automation Law of 2008 and the Comelec-Smartmatic contract of 2009 required that rate.

But the random manual audit after the 2010 national and local elections showed only 99.6 percent, despite very lenient methodologies. The July 2012 mock election in Congress showed only 97.215 percent, and Smartmatic already fudged it.

The 99.995 percent tolerates an error rate of only one mark per 20,000. The 99.6 percent translates to 80 errors per 20,000; the 97.215 percent, 557 errors per 20,000.

No accuracy and audit results were announced of the Comelec mock election last February in 20 precincts in ten areas across the land. After acknowledging various glitches, the agency abruptly cancelled the second mock election set for last week. Recall that in the 2010 balloting, there were errors in transmission (for canvassing) of even the final testing and sealing.

*      *      *

Preventing cheating. The 2010 cases of Glenn Chong of Biliran, Josie dela Cruz of Bulacan, and Grace Padaca of Isabela appear to disprove so, Lagman says. (Padaca incidentally is now with the Comelec, but is oddly mum about the PCOS.)

Compact-flash memory cards inside the PCOS contain computer commands for proper functioning. The CF cards found in the Cagayan de Oro city dump showed how easy it was to steal them.

The 60 PCOS units found in the house of a Smartmatic technician in Antipolo, Rizal, showed how easy it was to hijack the machine itself.

In 2010 the PCOS had an open port outside. Through that open port, a techie could connect a laptop and tamper with the software and CF cards inside.

Then and today stories abound about highest bidders for sure wins in manipulated results, Lagman says. “Forget about external hacking, but truly Smartmatic’s system is vulnerable to internal tampering.”

*      *      *

Transparency. Transparency is completely lost when precinct counting is automated, says Lagman. That’s why Germany, The Netherlands, Ireland, and some counties in the USA went back to manual precinct count.

Up to now voters in 2010 are unsure if their votes were counted right. There has yet been no review of the source code, ie., the human-readable commands that make the PCOS recognize, tally, and relay the votes. Smartmatic does not even own the source code; the Canadian developer, Dominion, merely lent but has taken back the license. These breach the Automation Law, which requires a source-code review, and lease-purchase straight from a developer instead of a middleman.

Worse, the Comelec has been very secretive, Lagman says. Its public website is incomplete, making impossible an accuracy-check of the canvassing.

“Secret balloting, public counting,” Lagman advocates. Automate the transmission and canvassing, but keep the precinct count manual.

*      *      *

Speed. Lagman purposely holds this for last. For, there’s no point in fastness if the process is inaccurate, fraudulent, and opaque.

The results in 2010 were fast, Lagman recalls. So fast that the Comelec announced partial results of the presidential contest as early as 6 p.m. of Election Day, one hour before the close of voting — a violation of poll rules.

There was one other aberration then, in the name of speed. The Comelec took it upon itself to canvass the presidential-vice presidential tallies. Only Congress could perform that role under the Constitution, but the poll body broke it.

At any rate, the public readily accepted the outcome because it was a landslide. And so the PCOS was hailed a success.

But wait, Lagman says, check the dates. Noynoy Aquino was proclaimed President on June 9, 2010, while Joseph Estrada was on May 29, 1998, in a completely manual balloting. From this, it would seem that the key to speedy results is not the PCOS, but a landslide victory.

Lagman sums up how the PCOS has affected the election process by far. It advanced the deadline to file candidacies, so names can be printed on ballots that have become long and expensive. It added warehousing (P400 million a year) and maintenance costs. Danger looms of automated cheating — by an insider; ballot secrecy is compromised. It needs an army of field technicians, and relies too much on a foreign vendor. It makes the country vulnerable to vendor problems, like the Smartmatic v. Dominion lawsuit. It heaps on the Comelec scorn, lost credibility, and lawsuits. It cost P9 billion. Transparency is forsaken.

All PCOS has to show for it is a shortening of the process by 12 to 24 hours.

*      *      *

Catch Sapol radio show, Saturdays, 8-10 a.m., DWIZ (882-AM).

E-mail: [email protected]