Philippines: Implementing E-Counting

Philippines Case Study

Project and Risk Management

Last updated on December 17, 2013

The COMELEC created a project management office (PMO) to manage the implementation of the different components of the AES. It included heads of different departments in the commission, including operations, administrative, human resources, legal, IT and voter education and planning, among others. The Executive Director headed the office.70 However, there was no concerted attempt to either define its structure or clarify its duties vis-à-vis the organizational set up and regular functions of the commission. While experienced in managing manual elections, members of the PMO lacked experience managing elections involving electronic technologies and could not anticipate the enormous challenges involved in such a task. The PMO did not establish regular meetings; formal reporting and communication process; or project controls, as it was more involved with day-to-day troubleshooting rather than quality control and risk management. Although the PMO developed a project management plan and timeline, it was not able to follow it, with deadlines adjusted as original targets were missed. Because of inefficiencies in the way the preparation was managed, the overall cost of the May 2010 national and local elections ballooned to PHP 16.5 billion from the allotted budget of PHP 11.3 billion.

Current and former election officials, parties, IT experts and civil society groups expressed concern that the COMELEC was unable to manage and oversee the vendor, Smartmatic, effectively. Several reasons have been cited, including the relative lack of IT expertise among the COMELEC and the shortened time frame, which required quick decisions and actions by Smartmatic, sometimes without following proper lines of authority. 

 

Voter Education and Public Relations

The COMELEC conducted a nationwide voter education campaign to inform the public about the new technology to help them become comfortable with it and instruct voters how to properly fill out the ballot. The campaign included broadcast and print media, instructional videos, billboards, flyers and a road show to demonstrate the PCOS machines and have people practice on it. Smartmatic provided voter education materials to the COMELEC, and COMELEC adapted these materials, as needed. 

The COMELEC’s voter education campaign was able to inform a significant percentage of voters. Public opinion research conducted by Social Weather Stations indicated an increase from a baseline October 2009 figure of 38 percent of voters who had either very much or substantial access to information about the new electronic system, to 67 percent just before the May 2010 elections. Given the limitations in resources and staffing, this is a significant achievement. There were also several areas for improvement that COMELEC staff, public opinion research and civil society cited. Most notably, voter education was not conducted in a strategic way and was not informed by public opinion research. As a result, the campaign did not sufficiently target those most in need of information and hardest to reach. Research during the elections showed that those with insufficient information were primarily elderly, rural and less-educated voters. In addition, some observers noted the campaign almost exclusively focused on the new technology at the expense of providing other important voter information.

The COMELEC also put a great deal of emphasis on public relations. Before elections, the public’s opinion of the COMELEC was very low. The COMELEC sought to improve this by being proactive and more open about emerging problems. Its policy was to work on a three-hour deadline to publicly address any problems and criticisms raised by others. COMELEC officials sought to build relations with key journalists, and staff attempted to answer all calls from the media. These public relations efforts contributed to a dramatic increase (approximately 30 percentage points) in public confidence in the COMELEC from before to after the elections. 

 

Equipment Delivery

Smartmatic was responsible for and had custody of the PCOS machines and accessories during their transport from the central warehouse to the hubs and polling centers. The delivery of PCOS machines and accessories was a tremendous challenge, given the short timeframe and geography of the Philippines. Smartmatic contracted three logistics forwarders to deliver equipment to the polling stations. Election observer groups criticized this bidding process for a lack of transparency, calling into question the “small size and limited access to networks” of the three companies, none of whom were in the top 10 in market share of freight shipping (by weight).71

The majority of the machines were delivered in the last few days before the elections, with some arriving on Election Day and a small number arriving after. Custody over the machines shifted to the relevant election officer when the PCOS machines and accessories were given to the BEIs. The guidelines further stated that in no case shall these machines and accessories be left in the polling places without any security. After the elections, BEIs were directed to give the PCOS machines to the technicians of the vendor, which shifted back the custody over the PCOS machines to the vendor. Regional election directors of the COMELEC indicated that this undermined their ability to supervise election preparations. They noted, for example, that they needed to secure the vendor’s approval to obtain backup PCOS machines and batteries in precincts that needed them. 

 

Software/Hardware Maintenance and Storage

Instead of an outright purchase, the COMELEC entered into a lease agreement with the vendor for the lease of the PCOS machines used in the 2010 elections, with an option to purchase. Of these, the commission initially bought only 920 units for electoral protest cases. For the remaining machines, the vendor assumed the task of storing and maintaining the machines after the elections. In March 2012, the COMELEC formally decided to exercise its option to purchase all remaining PCOS machines. In 2012, the COMELEC exercised its option to purchase the remaining machines.72 The Supreme Court eventually upheld the COMELEC on its position that it can exercise its option to purchase the remaining PCOS machines.

 

Final Sealing and Testing

A final sealing and testing was undertaken seven days before the election, when all PCOS machines had been deployed. During the sealing and testing, the COMELEC and the vendor discovered a problem with the compact flash card caused by the late modification in the ballot design without a corresponding reconfiguration of the software on the compact flash cards. As a result, the PCOS machines did not read the ballots properly. This caused a great deal of public uncertainty and calls to postpone the elections. The vendor and COMELEC had to take extraordinary measures to retrieve and replace some 76,000 compact flash cards with newly-configured cards just days before the elections.

After the arrival of the new, compact flash cards, testing and sealing were conducted in some polling stations. Most procedures occurred within two days of the elections. Election observer groups and some COMELEC officials interviewed noted there were a number of polling stations in which testing and sealing did not occur at all. In addition, the confusion and rush surrounding flash card replacement undermined chain-of-custody security procedures, which some pointed to as providing opportunities for tampering with the flash cards.

 

Election Day | Set-up, Security, Voting Process, Troubleshooting

The general instructions contain specific instructions for BEIs on: preparation for voting; manner of obtaining ballots; manner of voting; procedure in case of shortage of ballots; procedure in case of rejection of ballots by the PCOS machine; procedure for disposing unused ballots; procedure for the counting of ballots and transmission of results; the disposition of election returns; shutting down of the PCOS machine; and the disposal of PCOS, ballot boxes, keys, election returns and other documents.73 The COMELEC, however, was criticized for its failure to finalize and distribute the general instructions much earlier. 

To ensure integrity of the machine and the system, the general instructions outline steps for BEIs to follow before voting starts. These include initializing the automatic printing of a report showing zero votes for each candidate and including geographic information of the precinct. 

For the voting process, after authentication, voters were issued a secrecy folder and paper ballot, upon which they used a pen to shade an oval to mark each of their choices. After completing the ballot, the voter inserted it into the feeder slot of the PCOS machine. If the PCOS accepted the ballot, the machine

display flashed a confirmation message. Upon acceptance, the PCOS scanned the ballot and saved the image as a TIFF file in the compact flash card, along with data on how the PCOS read the ballot choices. The paper ballot dropped into a secure box under the scanner. After casting their ballots, voters returned to the BEI to have their finger marked with indelible ink (although observers noted that in many polling stations, voters were instead marked when they were handed their ballots). 

The PCOS machines returned a ballot out of the feeder slot if: the marks printed along the ballot did not match the assigned precinct; the ballot had already been accepted or rejected; or there were ambiguous marks. Voters had three more opportunities to correct and re-feed the ballot. After four total feeds, the ballot would be considered rejected, and the voter had to return the ballot to the BEI. Observers found that in most instances, ballots were accepted on the initial try. 

The issue of whether the AES provided a sufficient voter verified audit trail (VVAT) was debated. COMELEC officials contended that the ballot itself provided sufficient verification to the voter. However, several election observation groups and IT experts pointed out that voters were not able to verify how the PCOS machine interpreted their votes, which was the data transmitted as the official election results. 

Nearly 40 percent of BEIs surveyed in a Social Weather Stations survey had problems operating the machines; although, in most cases, the problems were not severe, and were eventually addressed. The most common problem reported was paper jamming during printing. Other problems reported during Election Day were inadequate real-time technical support for problems, such as running out of thermal paper; missing or drained batteries; and data transmission problems. 

To assist the BEIs, PCOS technicians were provided by the vendor, which claimed that over 48,000 technicians were recruited, trained and deployed for on-site support before and during Election Day. Call center agents were also mobilized during Election Day for monitoring the entire process and for remote support to field technicians. Election officers interviewed, however, complained that most of the PCOS technicians did not have the technical skills to assist them. Election observation groups and some IT experts interviewed also expressed strong concerns about the full level of access that the vendor-provided technicians had to the PCOS machines, particularly since most BEIs were completely reliant on technicians to resolve issues with the machines. 

One of the most significant problems on Election Day was long lines in the precincts.74  This may have led to disenfranchisement of voters who could not or decided not to wait in a long line. The lines were primarily caused by the need to cluster of precincts (i.e., the 250,000 precincts in 2007 were reduced to 80,000 precincts in 2010), wherein the number of voters per precinct was increased from 200 to a maximum of 1,000. The need to cluster precincts arose because the budget for the elections only provided for leasing approximately 80,000 PCOS machines. The long lines were compounded by the fact that the COMELEC did not increase the number of BEIs to handle the increase in number of voters per location, and the lack of training for BEIs on how to run the voting process efficiently. 


70 COMELEC M.R. No. 09-0612.

71 Final CenPeg Report, Project 30-30.

72 Election observation groups filed four different petitions challenging the COMELEC’s decision to purchase the PCOS machines. However, the Supreme Court eventually upheld the COMELEC’s decision.

73 Resolution No. 8786, Revised General Instructions for the BEIs on the Voting, Counting, and Transmission of Results in Connection with the 10 May 2010, National and Local Elections.

74 In a Social Weather Stations opinion poll, 71 percent of voters reported “very long lines” on Election Day.

NEXT:

Philippines: Election Day – Observation

 

Copyright 2024 © - National Democratic Institute - All rights reserved