“Present enhancements on the Vote Counting Machines (VCMs) are still sorely lacking in basic safeguards to ensure accurate vote recording and counting.” This was the observation of Mr. Rick Bahague, IT expert and Convenor of Kontra Daya, a broad alliance of advocates of clean and honest elections, following the mock polls conducted by the COMELEC on February 13.
“After taking part in the mock polls in Brgy. Pag-asa Elementary School and Ramon Magsaysay Elementary School, we are still not assured if our votes are counted accurately by this VCM,” said Bahague.
Kontra Daya Monitors encountered the following behavior of the new VCMs during the mock polls:
1. Vote verification is still missing. To verify that each voter’s choices are correctly recorded by a VCM, it should reflect the ballot’s votes on the machine screen and print out a confirmation receipt. These were conspicuously absent from the VCMs during the mock polls. Mock voters, however, were now given an opportunity to add their vote to undervoted positions or positions where not enough candidates were selected.
2. The VCMs can now print out their system hash code (a unique identifier of any software program) as part of the initialization report that is printed out when a voting precinct opens. The printed hash, however, cannot be compared with the published hash of the source code available from the Comelec website. This is because only the hash of the archive or zipped source code is available online [*].
3. There were still inconsistencies with the ballots that the VMCs accepted and rejected. In some instances, a significant number of valid ballots were rejected. In Aklan, Konta Daya monitors reported that 11 ballots out of a total 66 were rejected. Comelec and Smartmatic technicians are now pointing to the ink or marker used for the voting as a possible culprit. Similar incidents were also reported in the mock polls in Metro Manila. If not addressed satisfactorily, ballot rejection could lead to massive disenfranchisement.
4. There were significant discrepancies found between the manual counting and automated counting of votes in the mock polls in Aklan.
5. Transmission of results to canvassing and consolidation servers, as seen in the mock polls, (specifically, the municipal, national, KBP and transparency canvassing servers) are still erratic.
“In the absence of a reliable vote verification mechanism, voters are forced to place their complete trust on yet unreliable VCMs to accurately record their votes. Not being able to compare the printed hash code on the initialization report to the actual hash of the certified VCM program would mean that voters can not verify if the correct program is loaded on the VCMs,” explains Mr. Bahague. “To add insult to injury, ballots continue to be rejected. Every rejected ballot constitutes a violation to the basic right to vote of all citizens.”
“With only less than three months to go before elections, it is troubling to note these glaring deficiencies in the new VCMs showcased during the mock polls. Since 2010, we have been demanding the same safeguards of genuine source code review, vote verification, hash code comparison, reliable vote counting, and reliable transmission from COMELEC so that their voting machines can be trusted. Six years hence, we are still faced with untrustworthy VCMs,”” said Bahague.
Up to now, the source code of the canvassing and consolidation servers have yet to be reviewed. “This puts into serious doubt the accuracy and veracity of the results of the upcoming national elections,” ends Mr. Bahague.##