Roller-coaster poll automation
The May 13 elections went like a roller-coaster gone awry. It blasted off squeakily, accelerated as knots and bolts snapped, took a loop (bumping off riders), then halted in its tracks. Something was amiss: an error in the operating computers, absence of safety devices, power outage. Unruffled, the operator called it a perfect ride.
The winning candidates have been proclaimed and the latest poll exercise declared “perfect.” But the premature proclamation of the Senate winners and widespread technical failures beg explanation.
The Commission on Elections’ hyped 99.999-percent readiness was a dud. Aside from 1,000-plus PCOS machines malfunctioning, 18,000 of about 80,000 units failed to transmit election returns, potentially disenfranchising 10 million voters. Moving these precinct paraphernalia with the rewritable CF cards to canvassing centers ran the risk of card alterations and cloned machines replacing the real. The 12 million votes surging from just 1,000 precincts at the start of transmission was a statistical farce exposing Smartmatic’s canvassing and consolidation system program error—the same blunder seen in 2010.
Faulty modems, long power outages, ballot rejections and other technical woes had the Comelec cutting corners. Group canvass reports and certificates of canvass (COCs) were brought directly to the National Board of Canvassers, bypassing the ladderized transmission system. The proclamation of Senate winners kicked off with just 20 percent of the COCs. “Projected votes,” not actual results, will decide the winners, Comelec Chair Sixto Brillantes told the media. Did this explain the 60-30-10 pattern in the Senate outcome that manifested in the regional, provincial, and municipal levels? The Comelec stopped announcing the canvassing results at only 42 percent of the COCs. Otherwise, an IT analysis says, the votes would have exceeded the number of voters.
Discrepancies between PCOS and manual counts plagued the partial random manual audit. The results of the audit, touted to have been over hours after the elections, will instead be out end-June. The calibrated delay raises speculation that the real results will be unknown. Were the proclamations based on accurate count?
The disaster was bound to happen from the start. The Comelec made the same mistakes seen in 2010, on a bigger scale: Voter verifiability, source code review, valid digital signature, secured CF cards, and other minimum integrity mechanisms were not complied with. Was the low accuracy rating of the PCOS machines—97 percent, versus the required 99.995 percent, as found in the July 2012 Congress mock elections—rectified? Critics, and the Comelec Advisory Council, are now vindicated in opposing the poll body’s purchase of Smartmatic’s 80,000 PCOS machines. At the last minute, the Comelec dangled to the media the PCOS “source code,” but the drama unfolded way past the deadline for checking if it is the same program operating all the machines.
The right to vote was assailed by noncompliance with the election law and abuse of authority. The heart of suffrage is the voter’s right to see how his/her vote is counted and tallied accurately; transmitted results must correspond with the votes cast. This right was bulldozed by removing voter verifiability and by data inconsistencies in the Comelec website. Worse, voter intent and accurate counting were compromised by rewritable CF cards, invalid digital signatures, low accuracy rating, data discrepancies, all of which opened the system to fraud. A random postelection survey made by a think tank in 2010 showed most respondents in awe of the “speed” of the machines. Did they believe their votes were counted? “No” or “unsure” was the answer.
The intent to vote is not guaranteed. A ballot is rejected a few times, you lose your right. The midterm elections had no spare ballots in all precincts. How many voters lost their right this way, or through long queues, missing names, PCOS breakdowns, transmission failures? In 2010, 3-6 million voters were disenfranchised for reasons not of their making. In the PCOS system, the voter bows to the machine. But the law is clear: The technology must conform to “the actual conditions” in the country, including the voter’s IT literacy. The integrity of the vote and the election system has been replaced by the “infallible” PCOS.
The Comelec insists poll automation stops “human intervention.” But there was barefaced intervention when the Comelec and Smartmatic devalued the automated election system to its most fragile state. Trust the machine? Unlike here, where poll automation was rolled out and went full-blast pronto, public trust in other countries’ computerized elections was built through decades of rigorous experiment and pilot tests. In the end, many of them, like Germany and Ireland, gave up on full computerization for lack of transparency and electronic tampering.
The question remains: Why did the Comelec allow this travesty to happen? For political expediency? For an exchange? While the poll managers broadcast themselves as stringent in enforcing campaign rules, no credible response is given on allegations of bidding irregularities, unconscionable expenses like “intelligence funds” and unexplained property purchases.
The Automated Election System Watch has every reason to doubt the accuracy of the vote count and the ranking of winners and losers. But will Congress—which failed to assess the 2010 automated elections as mandated by law—and the Supreme Court bite, given the glaring questions?
The Comelec and its foreign technology supplier should be made to account for a flawed system that has all the makings of election failure. Unless these issues are addressed now, the 2016 presidential election will be imperiled and the question of who won and whether the votes were counted right will haunt us no end.
Bobby M. Tuazon is the former head of UP Manila’s political science program. He is the director for policy studies of the Diliman-based think tank CenPEG, which has done major researches and publications on the automated elections since 2008.
