Elections as an instrument of political control 171 have been certied 90 days earlier…. Why then did we bother with a source code review?” Indeed, what was the point of reviewing a software program that was not going to be used? Former Comelec IT director Ernie del Rosario also railed against the fact that the version of the software that Comelec actually installed in the precinct voting and counting machines was different from the version that had been earlier presented for review. This shell game was clearly the reason why Comelec issued no protocols or instructions on how the election ofcials assigned to the individual precincts could check and verify that the “hash code” – a precise identier that allows one to ascertain that one piece of software is an exact image of another – of the software installed in their precinct machines corresponded exactly with the hash code of the software that had earlier been reviewed. To be clear, since the software program earlier reviewed was different from the software program installed in the precinct-deployed machines, there was effectively no source code review – as required by law – and the integrity of the ballot appreciation and counting program in the precinct-deployed machines had to be in doubt. A malicious program could very well have been embedded in the software used and no one would be the wiser. Aggravating the absence of digital signatures and the inutility of the source code review, Comelec also refused to allow the precinct returns to be transmitted directly (and immediately) to a public website so that political parties and independent observers could verify if the results printed out by the machines at a precinct were exactly the same results transmitted to both the canvassing servers and the public website from that precinct. In addition to that crucial verication check, this would have allowed any observer to perform his own tally of the votes which could be compared to the Comelec tally. That Comelec would not even permit this simple procedure has to be a clear indication that they already anticipated – or knew – that there would be differences between the poll results that observers would see at the precincts and those that are transmitted to the canvassing servers for tallying. For the May 9, 2016 elections, a Certication by the government's Technical Evaluation Committee (TEC) – consisting of two representatives from the Department of Science and Technology and one from the Comelec IT department – that the automated election system being implemented by Comelec was acceptable and ready for use was supposed to have been issued on February 9, or 90 days prior to the poll date. This Certication would have conrmed that successful eld tests had been done, that the source code review had been completed, that the source code reviewed is the same as the ones installed in the machines deployed to the voting precincts, that credible mock elections had been held, that the precinct voting and counting machines were in perfect working order, and that the “Final Testing and Sealing” of the machines for deployment was done according to proper security protocols. Comelec could not produce such a TEC Certication until the citizens' watchdog coalition AES Watch insisted on April 28th (in a news conference) that the