The politics of automating the elections

There have been several articles on the attempts to automate the election process.  I myself am not too optimistic about this and I believe that the “hidden” agendas of politicians will hamper the progress:

The challenge is not in the development of the system or in the technologies that it utilizes, but how well the system can stand up to the attacks from the non-techies with their own political agendas. Any non-techie can easily claim to be an expert and any non-techie can publish any techno-babble to fool the rest of the population.  In a country where politicians consider a fictional novel like “œThe Da Vinci Code” as a threat to religion and morality, I wouldn’t be surprised if people use other techno-garbage movies as an argument to show how weak or insecure technology is. 

Now what caught my eye, and which further substantiated by thesis, was a statement from Robert Verzola, who gives a myopic, and probably self-serving, view of the solution to a speedy and fair electoral system:  punish the cheats, not election automation, is the solution.  

(Also published in Technopinoy

PCIJ quotes Verzola:

Computerization gives a false sense of security that everything will be tamper-proof, he says. “œComputerizing the system will not stop them from committing fraud. Once they learn how to, cheats will make use of computers to cheat.”

Another problem with automation is that it will make cheating so much harder to detect and prosecute, says Verzola, pointing to the experience in the U.S. where there are “œbig issues about cheating precisely because the system is computerized.”

I will not debate on his last point because it is true, but Verzola fails to realize that computer automation will also make it harder to cheat, especially if (and I know this is a big IF) the following requirements are delivered as stipulated in HB 5352:

  • adequate security against unauthorized access
  • accuracy in recording and reading of votes as well as tabulation, consolidation/canvassing and transmission of results
  • error recovery in case of non-catastrophic failure of device
  • system integrity, to ensure physical stability and functioning of the vote recording and counting process
  • provision for voter-verified paper audit trail
  • system audit ability, which provides supporting documentation for verifying the correctness of reported election results
  • an election management system for preparing ballots and programs for use in the casting and counting of votes and to consolidate, report and display election results in the shortest possible time

I am not surprised that Verzola has another agenda. While quickly dismissing automation as the solution, he begins to promote his own “Halalang Marangal” (or “the Network of Citizens for Honest Elections and Truthful Statistics,” or NoCHEATS), which was formed because “because we cannot rely on NAMFREL anymore for an honest tally. NAMFREL’s national office lost its credibility after the 2004 elections.” (Source) The item which I found ridiculous was Verzola’s alternative solution: an SMS-based system. The process is quite crude and rudimentary—the volunteers send the precinct-level data via SMS to a centralized database. In fact, it appears that anyone can send the data, but more weight will be given to the trained volunteer. So, in other words, Verzola wants to create another “watchdog” group, and, paradoxically, present an automated tool to counter Namfrel’s manual process. What Verzola fails to realize is that the better approach is to eliminate the concept of a “watchdog” altogether. In fact, if the election automation technology is designed and applied correctly, it renders an independent watchdog group totally irrevelant because there is no way to tamper with the data end-to-end.

I believe there is no silver bullet solution for the whole election automation challenge. My personal wishlist is to first simplify the election process, mirroring the “electoral college” process of the US so instead of counting individual voters, you count, for example, precincts. Then you automate as much as possible the election process, eliminating any physical data handoff as much as possible. You should also make the entire election process transparent to the public—ideally the source code should be open-source so any aspiring geeks can setup their own election system to test the application—and the results should be in a website live, online, and real-time. Implement this, plus the requirements in HB 5352, plus all the necessary encryption tools, and it will be very difficult to cheat. Not impossible (as only history can judge the security of an application) but VERY difficult to cheat.

Then you can punish the cheaters…

For more technology news and gadget reviews, follow us on Facebook, Twitter and Instagram

  • what i dont get is how come presidential elections in the US takes only one day to find out the result and here in phil takes more than a month..? and the wried part is, the technology we have here in phil is almost the same as they are using in the US, wireless internet, 3G, PDAs, mobile, computers, etc.. weird, really weird..

  • silkshadow

    Those are consumer services. Unfortunately, we have a weak government that has a per cappa tax revenue less than some sub-saharn african countries. Even the VAT system, implemented to fix this problem, is now being abused so bad that I am dubious any business is returning to the government the amount that they have charged their customers.

    On the computerized voting system, heres my take on it. Nothing is foolproof. However, consider the way politicians cheat in elections right now: using strong-armed tactics, bribes or stuffing the ballot box. Will computerizing the system change any of this? I don’t think so. Like the VAT, charter change/cha cha and similar initiatives, it is only a superficial way to pretend to be fixing things. If we want progress in our country, we have to fix the underlying problem. That is the people we elect to represent us and the general weakness of our government.

    We all cry about corruption. Day after day it is always a topic of discussion (of course we are so immune to it now that we only make jokes instead of being mad). However, corruption cannot be fixed without a corruption-free and strong backbone to clean things up. In fact, nothing can be done without that. How do we get that is the problem. My first recomendation would be to institute a summary death sentance on any goverment or civil official convicted of a new all-encompassing corruption law. For judges, a death sentence coupled with a total stripping of all assets owned by any family member given to all parties who was ruled against by that judge. Ah yes, but the Catholic church will never allow that, will they? So what do we do…

  • ppdolina

    THe australians had developed an free & open sourced electronic election system. it was featured on a few years ago and from what i can tell it’s the best system i’ve seen to date.

    The aussies hires a professional software house to make an electronic system then they open source it to the public at large so hidden agendas and bugs are sorted out. a numbero of notable aussie academics found bugs. then a 3rd party auditing firm sorts it out again for professional measure then it is again shown to the public. they do this til the code is clean then they implement the solution.

    open source as it is most useful .:) too bad our leaders arent that well read.

  • bonjo

    Computerization for automation should not be feared. The problem is if the software that will be used will be done by a local company(I think which is the scenario) which have ties (and probably have) with the government thus levelling up the probability of infusing backdoors for cheating. This is the part where issues on cheat detection will be hard. I’m not defending Verzsola and maybe I guess due to his lack of wisdom on the field thus he was not able to address out (or he intentionally avoided to say) the repercussions if a non-ally could obtain this cheating privileges. They all lose their government power.

    IMO then “Open source” is only the real way and solution.

  • sniffles

    Ideally, election results should be available by 5pm of the election day. Voters who are 100m within the voting precints’ perimeters trying to beat the deadline should not make any significant change on the trend of the election. In fact the trend will be obvious by early afternoon of that day. This should be like a ball game, not a beauty pagent where the subjective judgements are tabulated and released later.Anyway, it’s not us who decides how the votes are counted, it’s the politicians who wants to make a beauty contest out of this ball game.

  • Excellent Article.

    Thanks for the hard work and the valuable information for sharing with us.

  • a

    philippine elections have not been historically clean, automated has to be monitored even more so, specifically comelec, smartmatic, and the gprs systems.