28 February 2010

A 100 Percent Manual Audit of the AES?

From Felicito C. Payumo:

Lately, there have been enough people who, previously, were giving the COMELEC the benefit of doubt in making the Automated Election System (AES) “systems ready” before the May elections but are now clamoring that it switches to a “completely manual mode.” It is not just the readiness of the system but the acceptability to the public of the machine count that is being questioned. But a fully-manual system means going back to the 45 days of nerve-wracking waiting game and making our country again a laughing stock while the rest of the world watches our slow motion count. And wouldn’t that put to complete waste the P7.2 Billion rental payment to Smartmatic?

Do we have no other option but to make a leap of faith in the untried and distrusted AES or revert to the fully- manual system of counting and canvassing of votes? Are they our only two ineluctable choices?

While the law requires a Random Manual Audit (RMA) of one precinct per congressional district, everyone agrees that that sample size is woefully inadequate. Suggestions have been made to increase it from ¼ of 1 percent of the population to 7.5 percent - the bigger the sample the less margin of error- but the COMELEC has been non-committal. Equally important is the transparency of the manual audit by the BEI auditors using the traditional manual counting witnessed by official watchers of political parties and accredited observers. The results of both the PICOS machine count and manual count should be posted in the precincts. Needless to say, in case of discrepancy, the “root cause of the discrepancy” should first be determined as required by law, but the count that is most transparent –i.e. manual count- should be the one transmitted electronically. Common sense dictates that there be no proclamation of candidates before the manual audit. The argument that waiting for the manual count will cause delay betrays the “rush to proclaim” intent. The manual count, as everyone knows, takes at most 12 hours.

100 Percent Manual Audit- But there is a third and preferred option -a 100 percent parallel manual audit- this, being our first nationwide automated elections. All ballots counted by the 75,471 PICOS machines should be manually counted in the traditional fashion before the public at the precinct level. I can almost hear the COMELEC officials cry “why spend for the machines if we are to do 100 percent manual count anyway? It will mean extra cost and delay.” Even the technical experts among my friends think that a 100 percent manual audit is an overkill.

But they miss the point! Public acceptance of election results of the local contests is equally important as acceptance at the national level. A given sample size may be adequate to determine the accuracy of the count for national candidates but not at the local level. To illustrate, SWS or Pulse Asia, with a sample size of 1200 in their poll surveys, can give results with a + or – 3 percent margin of error for national candidates. But the same sample size - divided into 4 areas, namely, 400 for Metro-Manila, 400 for Luzon, 400 for Visayas and 400 for Mindanao – is too small to read the preferences for Governor, Congressman, and especially, Mayor with the same degree of accuracy. In like manner, a sample size for random manual audit good enough to determine the winners of national candidates, will certainly not be acceptable to determine the winners at the local level. A candidate for Mayor, especially, who has no trust in the machines, will not accept his loss when only 1 or 5 precincts have been manually audited in his municipality.

Hence, a 100 percent manual audit of all machine counts is not an exercise in superfluity. It is imperative that there be public acceptance of the election results from national down to local level. Public acceptance is indispensable in a democracy.

It also serves as the first block in confidence building for the AES. Should the results show a high congruence between the machine and manual counts, then the AES, subject only to normal safeguards will encounter less public resistance in the 2013 and subsequent elections. The funds spent for the machines, therefore, will not have been wasted, and automation, still “a consummation devoutly to be wished”, will be attained.

With respect to the extra cost of doing the manual audit, the same BEIs attending to the voters in their precincts can be used as auditors for other precincts. At 3 BEI-auditors per precinct, the 75,000 clustered precincts will need a total of 225,000 auditors. Paying P1000 per auditor will total only P225 million. It is a small amount to be paid for public acceptance of the election outcome.

After paying a staggering P7.2 Billion to a foreign company, Smartmatic, for mere rental of the PICOS machines, COMELEC should not quibble about a pittance to our own teachers and officials, (who will probably circulate the money in the country in a matter of days) in the interest of a credible election. But should the COMELEC refuse, the public has the right to ask, “Is there a sinister agenda that we do not know”?

Felicito C. Payumo

F.C.Payumo was a three-term Congressman of Bataan and former SBMA Chairman/Administrator. He is Chairman of the University of Nueva Caceres.

No comments: