How the Philippines Can in Lead Election Automation

Posted on March 3rd, 2009. Written by Rico.

If the Philippine government ever gets down to automating elections in the Philippines, it has a chance to show the rest of the world how to do it. That’s because there’s already a lot of established “wisdom”, and the mistakes of other countries to learn from.

Problems in Other Countries

Forget about the cost of funding such a project (the COMELEC has asked for P11.3 billion). How can we make the automated system easy to use? Around two years ago in Florida, a poorly designed voting machine led to 13% not casting their vote for their desired representative. The touch-screen interface was inconsistent, leading to confusion and a phenomenon known as “banner blindness”.

In Finland, the lack of clear instructions (and once again, a poorly designed process) led to 232 voters (out of 12,234) not finishing the voting process. These voters failed to notice that they had to “validate” their votes after making their picks.

These numbers may not seem significant. But if you’re running a nationwide election, with millions of votes to process, these errors will increase in scale. Imagine the situation here in the Philippines, as election sore losers and naysayers have yet another reason to keep the new officials from simply getting down to business.

The Missing Ingredient: Designing Machines for the Voter

So why did these two systems, as well as countless others, fail? Based on my research, those managing automated elections always overlook a crucial factor: usability. No election machine designer seems to consider the user experience, or how the average user will react to their device. Or at least, consider the advice of an expert used to anticipating and accommodating these reactions.

The process of marking a paper ballot is straightforward, because it mimics an activity most Filipinos do—writing on a piece of paper. And even then, some people get it wrong; what more if you force people to vote in a way they’re not used to? The majority of election machines fail to take this reality into account properly.

How to Get Things Right

Most systems also lack the benefit of sustained and controlled testing. The real-world conditions of an actual election are hard to simulate through a controlled experiment. Since election systems serve a crucial role, it’s important to get things right. You need to first test them on a small scale, and resolve any issues encountered. Before you can even think of wide-scale implementation, all potential problems—and their solutions—must be clear.

Granted, other countries have implemented testing on a small scale (like in the case of Florida above). But there’s a noticeable lack of learning from the results of these tests. It seems those who implement automated election systems don’t learn from their mistakes.

Ready for 2010?

In short, by looking from the mistakes of other countries, Philippine election officials can learn how to run automated elections properly. Primarily by designing election machines that make sense to the average Filipino voter, and relying on the results of sustained testing for any necessary revisions to the system.

Unfortunately, since the need for testing is crucial—unless you consider chaos on election day acceptable—a properly automated election system may not be ready by 2010. Yet no matter how clear the benefits of automated elections are for Philippine politics, it’s more important to get things right the first time. That’s much better than coming out with a system that will provide election automation critics with more ammunition.