Parallel validation run and expanded random manual audit are necessary
In light of numerous problems which arose when PCOS machines were tested at voting centers and Smartmatic’s subsequent recall of the compact flash (CF) memory cards for all PCOS machines, the National Citizens’ Movement for Free Elections (NAMFREL) re-emphasizes the need for stronger validation measures of the accuracy of the PCOS machines through more rigorous testing of the newly-configured CF cards, an expanded Random Manual Audit (RMA), and a parallel manual run.
With all CF cards now recalled, both COMELEC and Smartmatic must submit all recalled cards and their replacement cards, including those of the spare machines, to an audit process to determine the precise configuration or programming errors. All cards must be accounted for and all replacement cards must be fully tested. Each PCOS machine contains two CF cards.
At the same time, all PCOS machines must be re-tested before elections to check that automated ballot counts match manual counts. The machines are to be tested by May 7 before they are sealed just prior to elections.
We also propose that the Random Manual Audit be conducted immediately right after the close of elections in the precinct and that an additional step of verifying the hash codes of the PCOS machine’s source code be carried out to ensure that the source code used is one and the same as the one earlier tested and kept in the Bangko Sentral for safekeeping. As machine failure appears to be quite widespread, it is prudent to increase the number of PCOS machines subject to the random manual audit from just five machines per Congressional district to three to five percent of machines per district as we had originally proposed.
While the COMELEC dismissed the proposal of the IT community and business groups for a parallel manual run to validate the machine count, the widespread failure of the machines indicates that such a validation run is necessary. This validation run is relatively simple to conduct and will only require counting a limited of number of positions on all ballots to see whether manual counts match machine-printed Election Returns.
Given the glitches in the PCOS machines and the revelation that the AES system does not have time logs, NAMFREL also requests that the COMELEC make public a copy of their procedures on how it will handle evidence – both electronic and documentary – to support cases where electoral fraud is filed or should complaints arise.
Finally, all pre-election tests should include a full test of the canvassing software and programs to be used at municipal / city, provincial, and national boards of canvassers to ensure that summations of Election Returns to Statements of Votes Cast (SOVs) and Certificate of Canvass (COCs) are carried out accurately.
The failure of the machines for any reason is a serious matter given the public funds spent on the system thus far and its negative impact on our democracy. The COMELEC, contracting parties, and accredited watchdog groups must answer to the people regarding the failure of the machines or any part of the system due to poor project management. COMELEC must see to it that these errors are corrected before election day and that those responsible for these errors and others be held accountable.
NATIONAL CITIZENS’ MOVEMENT FOR FREE ELECTIONS (NAMFREL)
Room 601, DMG Center, 52 Libertad corner
Calbayog St., Mandaluyong City
Philippines
Telephone: +63 (2) 484-7590; 571-9943
Telefax: +63 (2) 470-4151
E-Mail [email protected]
Website: www.namfrel.com.ph