The central question for Congress as a canvassing body is this. How do we know that the COCs are duly executed, if the electronic ERs on which they are based are not duly executed?
Here’s a method for ensuring a proper COC in the public canvass at Congress, under conditions where the precinct-level ERs were not digitally signed.
I presume that the printed ER is nonetheless manually signed by the BEI. At the very least there is a PCOS-printed ER distributed to candidates at the end of the election day.
COCs at the first level, say at the municipal level, are printed out by computer from electronic transmissions (by modem or by reading of CF cards). The proper person signing on the COC cannot properly attest that the COC is duly executed if there is a discrepancy between the printed ER and the electronically transmitted (but digitally unsigned) ER. How does that signatory know that there is no discrepancy? Ordinarily, he knows or he can safely presume there is no discrepancy because there was a digital signature on the ER indicating reliability of the electronically transmitted ER. Extraordinarily, he could publicly and manually compare the printed ER and the electronic ER. If he did this, then the COC is authentic on its face. (There may still be problems if the PCOS was “pre-loaded” or “post-loaded” with irregular ballots prior to printing the ER and its electronic transmission, and such problems may not even be discovered by a manual audit, but this is a different story.)
So there. The proper question to ask of the folks who deliver the COCs for further public canvass, such as at Congress, is this. What is the assurance that the very first step of generating and transmitting an ER is safe and reliable, given that the machine count is not done manually and publicly?
The problems identified above are based on the time-honored principle that voting is secret, but counting/canvassing is public. If counting/canvassing is public, it seems proper to presume that the result is correct because the public had the opportunity to object. If counting is not public, it is submitted that the only way to be assured that the end result is correct is to subject the machine to manual audit (something that the random manual audit is supposed to accomplish) AND to require that the printed ER be the same as the electronically-transmitted ER. This can still be done, if the results of the random manual audit are made public (preferably before any public canvass), and IF the certifying officer on a COC can attest to equivalence between all printed ERs and their electronic versions.
But here’s a nagging question: Why hasn’t Comelec yet released the results of the random manual audit?