I Had To Use a Diebold Machine This Election

Posted by Gordon on Nov 8th, 2006
2006
Nov 8

In a sublime twist of irony I showed up to my precinct around 5 pm to find out that I was forced to use a Diebold voting machine. I say ironic because literally a couple hours before that I had blogged about the problems of voting machines. Here is the scenario as it unfolded. I showed up at the precinct around 5 pm there was a short line maybe 6-8 voters. After presenting my ID I was told that they were out of paper ballots. I had two options take a scantron ballot in chinese or go over and use the single Diebold machine in the corner. I laughed nervously and explained that I had a pricipled opposition to voting machines and that this put me in a quandry as to what to do. It was immediately apparent that the voting machine was not a favorite choice among voters. I thought about it for a few seconds and decided why not, at least I will get to have some empirical experience with a real machine. I think I was the first voter of the night to use the machine. The poll worker asked me to report back how it went and if I received any problems. Everyone was kind of curious how the experience would turn out.

The experience was fairly painless. Apparently there was a ticker tape that printed a physical record of my vote. Although I could only hear it running, there was no way as far as I could tell to physically view the paper trail output of my vote. I would say that this is a problem. I didn’t notice any particular issues with the touch screen not properly registering my voting intention. However I did notice what I consider a usability flaw with the process. As you are going through the ballot once you press an option there doesn’t appear to be any immediate way to change the vote. At the end of the voting process you are presented with an extensive review of your ballot choices. It is only at this point that you can go back and change a particular selection. I had not made any particular wrong selections, like voting for McGavick instead of Cantwell, etc. so I had no reason to try out the option to change a particular choice. However, from a usability perspective I would prefer the option to immediately undo a voting choice. What if I had made a wrong selection? The interface doesn’t seem to provide any feedback that you can change your choices until you are completely finished with the ballot. I could imagine this to be a problem for someone with fumbly fingers. If I had accidently made a wrong choice and found that I couldn’t change it immediatey I would be very frustrated and irritated. This is not a fatal usability bug since the voter can change his or her vote at the end but I do finding it lacking in the user feedback department.

Overall I wasn’t too displeased with the process but I still wouldn’t want to use a voter machine like this in the future. Even though there is a paper trail, it should remain visible to the user in the end. Perhaps display the ticker tape under glass or clear plastic so that I can see it being printed out. If we must use electronic voting systems then ideally, the system should provide me with a means to verify my vote after the election. One solution I can think of: Allow me to enter a unique password or user key that uniquely identifies my vote. Then register that vote in a secure database and allow me to look up my vote online or at the election office if I provide the key. The password can remain unique to me but with no voter identification so I can preserve my anonymity in the voting process. But provide me with a way to verify my vote and complain to the secretary of state in case there is an anomaly or mismatch. As far as passwords the system should force strong passwords upon the user and then require that a certain percentage of voters verify their votes against an official central database of votes before making the election results official. This would provide better security because it would require end users to police the results and if properly organized, an efficient means to audit the system and votes after the fact.

The basic problem now is that the entire process, paper ballot or otherwise, is a leap of faith for the end user. It doesn’t work this way in the banking world, or other systems that require strong security. Even if there is corruption the end user can be called upon to police the system and ensure integrity. However, providing a mechanism for individual voters to verify their votes seems like a good way to combat corruption and more importantly ensure accuracy. I am sure there are more issues to think through on this problem but I think the security premise of end user policing of the process seems like a viable hybrid solution.

And one last point. Why did my precinct run out of paper ballots? I don’t really have any reason to suspect intentional fraud. I will give the benefit of the doubt and assume that this election cycle the turnout was higher than expected therefore the precinct was not prepared for the number of voters. Which is a good thing. But I don’t understand why there has to be situation where a precinct would run out of ballots. Seems odd.

Hacking Democracy

Posted by Gordon on Nov 7th, 2006
2006
Nov 7

I’m Nothing

Posted by Gordon on Nov 1st, 2006
2006
Nov 1