Josephine Scott presented on usability of voting systems at MIUPA’s October 2004 chapter meeting. Tim Keirnan caught up with her in January, 2005, to hear her reflections on America’s 2004 election.
Interviewer: Timothy Keirnan
Guest: Josephine Scott
Tim: Two months have passed since the national elections. What can you tell us about the usability of our voting process in general or in specific that you’ve seen or heard about?
Josephine: Right now as I’m speaking to you there’s a show on discussing some of the things that we saw in the election this past year. We noted that there were many more direct recording electronic voting machines—the ones with touch screens—that were in use. Overall, there were concerns in certain areas but they didn’t seem to be as related to the technology as some might have thought they would have been.
On the other hand, Ohio just today announced that they are going solely with optical scan technology because of their experience in this election. So there were some issues. We heard of one case where people leaning on machines caused machines to record a vote improperly, so we’re working diligently to address those usability issues.
We’ve worked this year with the IEEE on voting standards and we all know that Whitney (current president of the Usability Professionals Association) is involved with—in fact, she chairs—the Human Factors and Privacy Subcommittee, a part of the Technical Guidelines Development Committee of the federal government’s new Election Assistance Commission. The committee is going to be advising the Commission on voting technology.
So the usability profession has input, we are certainly making inroads. We are introducing the idea that the voter—which is our user—is the most important part of a voting system design. The most encouraging thing is that the message is beginning to be heard.
Tim: I’m curious about what you said regarding Ohio. One of the things that concerned me last year was that they seemed to be leaping into electronic voting and I wasn’t so sure that was a good idea, because so many of those machines don’t produce a written, traceable record of the votes cast. But it sounds like—if they’re transitioning to optical scanning for the entire state—they must have thought it wasn’t working out as they’d hoped. What do you think about that?
Josephine: I think the Ohio Secretary of State learned something from this election. I know he had a great deal of investment in direct recording electronic voting machines and had great faith in them. And in fact, direct recording electronic voting machines are very usable, and they have a great advantage for the disabled community—like we talked about in my presentation last October—because they permit a great deal of independence. There are many pluses for the DREs. But they have had one fundamental flaw: they don’t provide a transparent vote count—one that is intuitively, obviously correct and accurate. And as a result, I think that caused a lot of discussion in advance of the election. Sure enough, on election day they did see some problems.
Also, there were a number of questions raised with regard to the use of technology in some precincts, the lack of technology in others, and problems in terms of the length of time that people had to wait to vote in some precincts. And the analysts drew some relationships between them and the makeup of that particular precinct in terms of demographics. So—I’m anticipating what Ohio’s secretary of state might have been thinking here—he, like other states including Michigan, made the decision that it might be wiser to make certain that the voting systems being used do have some sort of artifact that makes it possible to show that the vote has been counted accurately.
I can actually speak for Michigan and say that that is exactly why Michigan chose optical scan when the Help Americans Vote Act was passed and they made their recommendations. They knew from experience that the optical scan system was usable but at the same time provided them the reassurance they needed for voter confidence.
Ohio was the target of quite a lot of what I’ll call “discussion” (laughs) with regard to his conduct in the election. He promoted his choice of electronic voting machines and defended his conduct in the elections, but it actually came under a bit of fire. There were, according to reports, very long lines in some precincts. I can’t say exactly what kind of technologies they had but clearly when you have a line that might potentially stretch for four, five, six—I even heard ten—hours waiting time, you have a fundamental process issue, a fundamental usability issue that needs to be addressed. I think that he’s looking for a solution that will make sense in his state. And I think he settled on one that, based on our experiences here in Michigan, should work fine.
Tim: I must say I was delighted when I voted that there was optical scan there. I’ve always felt more comfortable with that voting technology in the states I’ve lived in—and I’ve lived in quite a few now—where, as a registered voter, I’ve been exposed to multiple ways of voting. I just feel confident voting by optical scan.
Josephine: It has the advantages of electronic systems in that it can recognize that the ballot has been over-voted or under-voted. It is easy to use for the precinct inspectors to tally their votes. So there are also tons of pluses with optical scan voting, and you still have that artifact, that piece of paper that says this is how the voter recorded the vote and how they intended it to be cast. It provides that backup. That’s been the fundamental issue about the touch screen types of machines.
Now, that said, I want you to know that the EAC did commission some folks to observe the polls, and I did some observation in a touch screen precinct in Westland and:
- There were no lines.
- Everything went very smoothly. The average vote happened in about six minutes’ time, but
- There were a couple of small things that were real human factors issues. They had a glare screen over the tops of their machines, so people had to bend down in awkward positions. But those are things we can advise them on; that’s something we can do something about. What we can’t do is re-create a vote that has been lost with other less reliable voting technologies.
So it was an interesting decision in Ohio from my perspective.
Tim: Let me ask you about the sample ballot drive. Last year UPA requested that people send you sample ballots from their precincts and I know I sent you one from mine.
Tim: They were very good about sending me a sample ballot. I was impressed with the staff in Freedom Township. I called on a Friday and got their answering machine, so I left a message saying “I’m Tim Keirnan, I’d like a sample ballot mailed to my home, here’s my phone number” and I forgot to leave my home address or spell my last name. And I wondered “Where is everyone on a Friday afternoon over there?” But Sunday evening when I was most likely to be home, the township clerk called me and she said “You left your telephone number and I wanted to confirm your address so I can mail this to you tomorrow.” I was really impressed and it arrived on Tuesday!
Josephine: (laughing): Good job, good for her! Number one, let me say something about Michigan’s system and the people who run it. I think it’s the finest in the nation. I think our clerks are terrific for the most part; there are tons of them compared with other states. We have a very decentralized system of about 2,500 people who are responsible for doing the kind of activities you’re talking about. That’s a lot of people, and then you get down to the number of precincts, inspectors, and so forth. It’s a really unique system and they do an amazingly great job. It’s my typical experience that that’s the service voters get when they ask for something just like you did. I suddenly realized I was going to need to vote absentee at the very last minute, and my clerk did a terrific job.
Tim: Well this is good news all around. I then sent my sample ballot on to you folks at UPA as you asked. Have people been sending you sample ballots?
Josephine: They did. It’s pretty much ended now, I don’t expect to receive any more at this point. I received a nice selection of different ballot systems, now I have a broad representation of different ballots. We did get what we were looking for: a baseline for ballot design moving forward into future elections.
Now it’s my job to take these ballots, scan them, and put them together in electronic format with lots of meta data so we can hand them off to the TGDC then to the EAC, and they can keep them for the future. We actually even have a relationship with a person who has done this for ballots internationally, so we may be able to work with her and create something even broader.
Tim: Considering the number of nations that permit voting, there must be elections going on year round. Would you like UPA members from countries besides the USA to send you sample ballots?
Josephine: Yes, we’re always checking into what other countries are doing. UPA UK’s Louise Ferguson has been hosting an electronic voting forum which has a very international flavor to it. We regularly hear from people about voting usability in the UK, in Italy, and some South American countries. Israel has an interesting voting system. India’s election system was very interesting. So, we are keeping our ears open for voting systems around the world because, frankly, usability in voting is so critical to democracy.
Tim: Would you like to tell us about UPA’s efforts in usable voting systems for 2005? What’s going on this year?
Josephine: This year is a year for us to organize and for work to continue. So Whitney continues to serve on the EAC’s TGDC. That’s probably where some of the most interesting and important efforts are going to come out of that body. NIST is continuing its work; it was through them that we did this little voting observation exercise I mentioned earlier.
Elections happen in waves. You do a lot of preparation in the off years, then interest really begins to peak in anticipation of the election and things get very busy. This I know from having been an election worker. And this pattern seems to be happening in our voting usability work as well. We’ll probably not see a lot of legislative activity until we get close to at least the mid-term elections. We can use this time to refine our efforts, formalize some things, and working towards project where we can have an effect on ballot design, on processes in the polling place, et cetera.
We encourage UPA members to become aware of what may be around them in their states here in the USA or elsewhere. One of our members is working with a ballot simplification committee that exists by law in their area. That’s certainly a strong area of interest for any usability professional. We encourage everyone to become aware, become informed, become involved—if not as a usability professional, then certainly as a citizen.
Tim: That leaves us in a good place at the end of this interview to ask where does somebody go in their country to help the usability in voting initiative?
Josephine: I would begin with the local executive body for elections in their government. There will be some type of committee, some type of bureau, agency, some sort of body responsible for executing the thousand of details involved with conducting an election.
Another thing you might do is contact any elected officials who have an interest in voting. In the USA, Representatives Rush Holt and John Conyers have been really involved in voting issues; let them know how you feel because that helps them move forward. And let your own district’s representative know your opinions about electronic voting and usability.
Systems continue to be designed without having the user of the system at the center of that design. If we work to promote the notion that voters need to be the center of the design process for voting systems throughout the world, I think —I hate to sound overly emotional about it—but I really think that is a fundamental concept for promoting democracy throughout the world: the idea that now voting systems can actually work for the voter.
You know, you no longer need a human being to count the votes but you need that voter to be confident that the vote has been cast properly. And that’s where we usability professionals come in: we verify that a voting system design is usable for its intended audience, and that there is a traceable record of each and every vote cast using the system to enable accurate counting and recounting.
Tim: Anything else you’d like to mention before we close this interview?
Josephine: Let’s see, yes there is.
- Voting system manufacturers, led by Diebold, are starting to offer new electronic models and retrofits on DREs that do in fact create a voting record.
- The states have asked congress to limit the authority of the EAC (which will be the body that defines usability standards for voting systems) arguing that voting traditionally has been conducted by the states. It is unclear how Congress will respond but that could impact any requirement that systems be usable.
- There is a Voting Workshop accepted for this year’s UPA conference. The goal of the workshop is to create detailed guidelines that can be used to evaluate local ballot and election materials.
Tim: Thanks for your time, Josephine, and for presenting to our chapter back in October.
Josephine: You’re welcome! UPA members should know that they can find information on our voting efforts on the home page of http://www.upassoc.org under Most Visited Links, or go directly to http://www.upassoc.org/upa_projects/voting_and_usability/index.html.