Latest Issue
Special Issue: Democracy Beyond Border Edited by David Held and Robert Schütze, this special issue of Global Policy brings together a group of scholars to discuss and assess the extent to which democracy can crystallise beyond the nation state, and whether such developments are, in principle, desirable.

Automation Alone Cannot Counter Electoral Corruption

Pauline Eadie - 11th February 2013
Automation Alone Cannot Counter Electoral Corruption

In this column Pauline Eadie draws upon experiences in the Philippines to argue that electoral automation remains a contentious issue and that democracy in the digital age faces new and complex challenges.

Midterm elections will take place in the Philippines on 13 May 2013. Smartmatic supplied precinct count optical scanner (PCOS) machines will be used to electronically register votes. Automated voting was first introduced in the Philippines in 2010 in an attempt to counter cheating and accelerate the vote count and declaration of results. Concerns were voiced over the reliability of the automated system in 2010. One group raising such concerns was the People’s International Observer’s Mission (PIOM). The remit of the PIOM was to bear witness and report on the technical process of the election. The findings of the PIOM were widely disseminated in the local, national and international media. This author was a member of that mission.

Based on first hand observation this article will show that electoral automation has proven contentious for a variety of reasons in the Philippines and elsewhere. Mistakes were in evidence during the 2010 elections and the extent to which lessons have been learnt is debateable. Automation alone cannot counter electoral fraud. Those who may be inclined towards electoral cheating will just adapt their practices to electronic systems. Thus democracy in the digital age faces new and complex challenges.

The Shift to Automation in the Philippines

The manual process of voting and the recording of votes had been notoriously lengthy in the Philippines. Voters are faced with an array of choices on their ballot form given the complex mix of national and local government positions at stake. The archipelagic geography of the Philippines also hindered the gathering of votes. The system was difficult to police and protracted, it was common for national level posts to take weeks to announce. Votes could be lost, diverted or manipulated at many levels. One notorious way of manipulating results was “dagdag-bawas”, which means vote padding or vote shaving. Votes were simply removed from one or more candidates and reallocated elsewhere, this was most easily done by the reallocation of a zero here and there.

It took 18 years for national automated voting to roll out across the Philippines. Delays in the process tended to be legal rather than technological. The automated system was first used in the Autonomous Region of Muslim Mindanao (ARMM) in 2008 during elections for local governors and the ARMM Regional Legislative Assembly. Subsequently the Commission on Elections (COMELEC) Chairman Jose Melo signed a Php 7.2 billion contract with Smartmatic-TIM in July 2009 to provide equipment for the 2010 elections. This was despite a legal challenge from the lawyer Harry Roque who based his objections on technical irregularities during test runs of the voting equipment and irregularities in the procurement process that awarded Smartmatic and its local partner Total Information Management (TIM) the contract.

Questions marks remained over both the process of automated voting and the role of Smartmatic. Critics of automation cite Ireland, where automated voting was adopted in 2006 and then rejected on the basis of lack of trust and cost and Germany where the Federal Constitutional Court declared automated voting in the 2005 federal elections unconstitutional. More recently problems emerged in the 2012 United States’ elections when voting machines appeared to alter votes for Obama to Romney.

The Credibility of Smartmatic

Philippine critics argued that handing the functions of the election of to a foreign owned company was unconstitutional. Smartmatic’s owners are Venezuelan, and it had only ever managed national elections in Venezuela. Its management are allegedly linked to both the Chavez regime and the US government and “Venezuelan Ambassador to the USA Bernardo Alvarez Herrera is a relative of Smartmatic's wonder boys.” In 2006 the US Treasury investigated Smartmatic over its ties to Chavaz and the US State Department investigated the company over tax irregularities and the awarding of the US$91 million contract to Smartmatic for the 2004 Venezuelan national elections. Chavez won the 2004 election, a result that the political opposition deemed fraudulent. Allegations were fuelled by a lack of transparency over the voting process. Smartmatic was thus perceived as corrupt in some quarters.

On 4 May 2010 more than 72,000 memory cards for the PCOS machines were recalled by COMELEC. It was claimed that the cards were not recording the data correctly. Smartmatic Southeast Asia president Cesar Flores stated anomalies between the configuration of the memory cards and the layout of the ballot papers meant that only the votes of national level candidates were being recorded correctly. The memory cards don’t read names, just shaded ovals, and if the shaded oval does not match up to the correct candidate the vote goes to the wrong person.

The recall was greeted with horror by the national press and there was speculation that the elections would be delayed or that a manual count would be reintroduced. However the cards were returned within the timescale and the elections went ahead. It was this recall that is the focus of allegations of “Hocus PCOS.” Critics argue that the cards were electronically pre-programmed with the election results.

Polling Day May 2010

On Election Day the PIOM monitoring team arrived at voting precincts at around 6.30 am. Some people reported that their names were missing from voter lists so they could not vote. However early in the morning it appeared that voting was set to proceed without too many problems. The PCOS machines were in situ, the electricity was working, the machines took the ballot papers and recorded the votes and the actual process of voting, i.e. shading the oval markers of candidates was manageable. This meant that worries about lack of IT literacy amongst voters were unfounded.

However, as the morning wore on, it became evident that the election process was running into problems. By mid morning it became clear that the polling process was backing up as too many voters had been allocated to each PCOS machines. Up to 1000 in some cases. Some people waited up to eight hours to vote. There were some instances where machines jammed and COMELEC personnel were sparse. This meant that machines were out of action for long periods. Watchers from political parties were seen interfering with the machines when this happened.

The PIOM also saw mass evidence of vote buying and scant regard for the “secrecy and sanctity of the ballot” as defined under Article V, Section 2 of the Philippines constitution. We also noticed that on the next day when ballots were being canvassed locally 20% of the memory cards from the PCOS machines had not turned up by 3.00pm and whole PCOS machines went missing. There was also concern that the digital signature in the PCOS machines had been disabled. When the ballot paper went through the PCOS machine a congratulations message was flashed up but there was no way to confirm that the vote had been successfully recorded. However Smartmatic considered the elections to be a success. This was evidenced when they paid for two page spreads in the national press declaring just that. COMELEC have now seen fit to buy, rather than hire, PICOs machines in preparation for the May 2013 elections.

Conclusion

The 2010 elections were hailed as a success because the results were announced relatively quickly. The winning presidential candidate was not announced by congress until 9 June 2010 but early indications were of a clear landslide for Aquino and this proved to be the case. The margin of his victory, around five million votes assuaged fears that the wrong man was now president. The results mirrored pre-election surveys. Sample manual parallel counts also matched the results recorded by the PCOS machines. The failure rate of PCOS machines was recorded as 0.5 per cent.

However evidence clearly shows that vote buying, harassment and localised violence were in evidence. Organisation at the precincts was shambolic. Philippine elections have historically been characterised by cheating and the automated system was designed to remedy this. The automated system was introduced into a society that is widely perceived as corrupt. Therefore it is not unreasonable to suppose that those who would have cheated in a manual election will also finds ways to cheat in an automated one. Automation does nothing to alter vote buying, harassment and election related violence

These issues are also pertinent for Australia, who has been investigating electronic voting for a number of years, and for Indonesia who aims to be the second Asian country to introduce automated voting in the 2014 elections. Historically Indonesia has a comparable track of electoral corruption. Since 2010 Smartmatic has eyed Indonesia as a market for PCOS machines. Meanwhile watchdogs continue to report technological glitches in the run up to the May 2013 elections.