THE Comelec Advisory Council (CAC) was convened only last June 13, 2017. It quickly went to work and, in such a short period, managed to organize and hold a technology fair where vendors and proponents of automated election systems (AES) were invited to showcase their election system solutions.
The CAC is mandated, among others, to “recommend (to the Commission on Elections) the most appropriate, secure, applicable and cost-effective technology to be applied in the AES, in whole or in part.” The technology fair held last July 26, 2017 allowed some members of the CAC and interested stakeholders to get acquainted with various automated election technology solutions.
A review of the policy declared in the Automated Election Law, or Republic Act 8436 as amended by Republic Act 9369, reveals that the chosen AES technology solution should have the following attributes:
1) Ensures the secrecy and sanctity of the ballot
2) Transparency and credibility of the (election) process
3) The results shall be:
c. Reflective of the genuine will of the people
4) Suitability of the technology
“Secret voting, public counting.” Security measures must be in place at all times to protect the secrecy and sanctity of the vote. It should be demonstrated that each ballot is kept secure and cannot be traced to a voter.
The PCOS/VCM used in the national and local elections of 2010, 2013, and 2016 has been criticized for absence of transparency. The recording of the vote is not publicly viewable. The counting of the vote and how the election returns is prepared is hidden from view. While the operations of the canvassing and consolidation system is open for observation by political party watchers and accredited observers, the canvassing and consolidation of votes is not publicly observable.
Transparency measures had been provided, including review of the source code and, for the 2016 elections, the printing of the voter verifiable paper audit trail enabled.
Review of the source code is an activity that can be done only by a few who have the skills to perform such activity. The 30-day period provided for public review of the source code for the 2010 elections was insufficient for experts to conduct a meaningful review. There was no source code review for the 2013 elections – the four-day period provided amounted to nothing! For the 2016 national and local elections, the uncustomized version of the source code was made available for review by interested parties and groups seven months before the elections. The customized version was made available for review three months before the elections. But the reviewers resorted to going through parts of the software code that they thought critical since the periods of review were insufficient.
The conduct of a publicly observable end-to-end test of the AES would be ideal. An end-to-end test would show how the voting results flow from the vote-counting machine to and through the canvassing and consolidation system hierarchy.
Any automated system performs tasks with speed. As observed in the 2016 elections, all 12 winning senators and 49 winning party lists were proclaimed 10 days after election day, while Rodrigo Roa Duterte and Maria Leonor Gerona-Robredo were proclaimed President and Vice-President, respectively, 21 days after the electoral exercise.
The conduct of the random manual audit, immediately done after each election, has been criticized, perhaps unfairly. But accuracy of the vote count determined at the end of the RMA in each of the past three elections is below the 99.995 percent accuracy requirement. Its conduct should be done with utmost transparency.
The most challenging requirement to quantify is “that the results be reflective of the genuine will of the people”. But this may be defined as a combination of all other requirements, including the assurance that each vote is counted and that none of the voters is disenfranchised.
Apart from demonstrated capability as stated in the Automated Election Law, the CAC should also define the attribute “suitability of the technology”. Perhaps, in order to define suitability of the technology, the voting environment must be examined – familiarity of voters with technology, ease of use, availability of other resources such as power and telecommunications facilities, and others.
Cost as criteria. Should the Commission on Elections deploy the same voting equipment in all voting precincts and posts? Or, should the commission use less costly voting equipment where there are fewer voters in a voting precinct and posts? A review of the distribution of the voting population shows that there are about 200 clustered voting precincts/posts with less than 100 registered voters and an additional 1,750 clustered voting precincts/posts with 200 voters or less. If the same voting technology is used in all clustered precincts/post, the cost per voter in a clustered precinct with 800 voters would certainly be lower than if a clustered precinct had only 100 voters or less.
Credibility of the election process and results can be gained if the election system possesses all the attributes defined in law and all processes are executed without a hitch.
Finally, the election system technology must be secure. Among other security features, the AES must allow digital signing by all members of the boards of election inspectors and the boards of canvassers of all election results to ensure authenticity of the election results.