COMING from a technology forum with journalists where I discussed, among others, trolling and fake news that dominated the campaign period in the 2016 elections and months afterwards, I received the following text message, “Re: VP Protest Case. Recount of ballots from 4-5 provinces will start end of November and will finish by December. Deliberations after. You can bet early next year, we have a new VP.” The sender said that the text message had been going around and wanted to know if it was true. He also wanted to know at what stage the vice-presidential election protest case really is.
The news source, predicting the future, and arousing readers’ concern, are among the telltale signs that would indicate that a news item may be fake. In the case of the text message in question, it’s difficult to determine the background of the news source but the way the news item was crafted, it appears that the news source may be an avid Marcos supporter. The contents of the news item and the way it was crafted had prompted recipients to express concern. And, the news source confidently predicts that the country will have a new vice president.Further verification showed that neither the early evening TV news programs nor social media and internet news sites carried the story. It was a bummer.
https://www.bongbongmarcos.com/protest-watch/ shows that the Presidential Electoral Tribunal (PET), the body that adjudicates presidential and vice-presidential protest cases, has yet to issue a retrieval order, the purpose of which is to place the ballot boxes in the PET’s custody. Will the retrieval order also cover the SD cards and/or the PCOS/VCMs?
It is not only in the Philippine elections that troll posts and fake news have been observed. Emily Bell of the Columbia Journalism Review (http://www.cjr.org/tow_center/) cites sources saying that fake news were generated out of the Philippines and that pro-Trump Facebook pages were generated from Macedonia. The troll posts and fake news must have generated quick revenues for the Filipinos and Macedonians involved. Did trolling and fake news influence the outcome of the elections? An interesting subject for academic study.
Back to the VP protest case, one wonders how the recount of the votes will be done. Will the PET use another PCOS/VCM for the recount? Or, will the recount be manually done using the paper ballots. Or, will the ballot images, stored in the SD cards, be used for a manual recount?
In past automated elections, the Commission on Elections relied on the ballot images in adjudicating protest cases. In one protest case, the comparison between the ballot digital images and the paper ballots revealed overvotes, which was the subject of the protest, such that if a PCOS machine was used for the recount the vote count results would favor the protestant. The protestant did not pursue the case.
In Maliksiv. COMELEC, G.R. No. 203302, March 12, 2013, the Supreme Court said: “The ballot images are electronically generated and written in the CF cards when the ballots are fed into the PCOS machine. The ballot images are the counterparts produced by electronic recording which accurately reproduce the original, and thus are the equivalent of the original. As pointed out by the COMELEC, the digital images of the physical ballots are electronically and instantaneously generated by the PCOS machines once the physical ballots are fed into and read by the machines. Hence, the ballot images are not secondary evidence. The official physical ballots and the ballot images in the CF cards are both original documents. The ballot images in the CF cards have the same evidentiary weight as the official physical ballots.”
But how reliable are the digital images of the ballots?
Finding high discrepancies between the manual count and the machine count in 234 PCOS machines, the 2013 Random Manual Audit Committee referred the 234 PCOS machines to the 2013 Technical Evaluation Committee (TEC) for root cause determination. The TEC found that digital ballot images in 11 of the 234 PCOS machines had digital lines on them. The TEC further found that in 7 out of the 11 PCOS machines, digital lines ran through voting ovals. Eleven of 234 is 4.7 percent. Close to 78,000 PCOS machines were deployed in the 2013 elections. Extrapolating, 3,666 PCOS machines appears to have generated digital lines on the digital ballot images. Extrapolating further, 2,333 PCOS machines had digital lines running through voting ovals. At an average turnout of 600 voters, a total of 1,399,000 ballots could have been potentially affected. For a presidential and vice-presidential contest, 1,399,000 votes is quite significant.
The source of the fake news appears to assume that the recount will be automated, predicting that the recount of ballots from four to fiveprovinces will be completed by December. But what if some ballots have been tampered with?
A manual recount is a tedious exercise and will most likely take longer. A revisor will have to carefully examine each mark on the ballot for the contest, the subject of the protest, to determine if the mark meets the threshold rule–a valid mark should cover more than 25 percent of the oval.
Let’s face IT. The technology used in the last three elections must be carefully reviewed and other technologies that may be considered for use in future elections must be studied in detail. The country anxiously awaits the outcome of the protest in the vice-presidential contest.