POLITICS
Can You Get a Fair Election
in Cook County?
By Stump Connolly
At the start of the new year, a blue ribbon commission headed by retired federal judge Abner Mikva issued a report on how Cook County handled its November 2006 election.
The report was not kind to the clerk’s office or Sequoia Voting Systems, the manufacturer that supplied the city and county electronic voting machines. It found that 44% of the 5000 precincts ran into problems transmitting results, that early testing of wireless reporting did not reveal interface issues with Sequoia’s software, and that the central tabulating software quickly bogged down when early voting results were input at the same time precincts were trying to report in.
The delays came to a dramatic head when Tony Peraica, the Republican candidate for Cook County board president, led a dramatic march on the election headquarters demanding faster results. With city elections coming up in a month, it seemed like a good time to pay a visit to County Clerk David Orr to ask the obvious question: Can you have a fair election in Cook County?
STUMP: The Mikva Commission did not paint a pretty picture of how the last election was handled in Cook County. How did you see it?
ORR: If you are looking at what makes an election good or bad, you look at what happens before Election Day; what happens on Election Day; and what happens in terms of getting election results and the accuracy of those results.
Election Day is like democracy. It’s messy. You’ve got 12,500 people who volunteer to be election judges. They get paid [up to $150 a day.] But these are people off the street, decent people, and sometimes they fight with each other. They get sick. Some of them fight with the voters. Some of them just show up having a bad day. In the real world, things are not always perfect – and I think you have to look at the last election in that context.
What you really want to know is: Did the voters have a good experience? Did their votes count? And were the vote totals accurate?
On every one of those things, the answer is yes, yes, yes.
STUMP: So you’re saying things went well?
ORR: Election Day wasn’t a perfect day. We had silly problems. Not enough pens in some precincts. Stuff like that. Our single biggest problem with the touch screens was that sometimes the cards would get stuck. It didn’t stop people from voting. It just caused a hassle. The judge had to come over and un-stick the machine. If we got complaints that touch screens were not working, we replaced them.
PETER McLENNON (Orr's policy analyst): Of the 9,000 pieces of equipment we had out there, we swapped out 22 machines.
ORR: Now, on election night, did the votes get counted, and was the count accurate? The votes were counted, and the count was accurate. But the reporting was slow.
That’s why I set up the Mikva panel. I wanted to get to the bottom of what happened. And I wanted to fix the problems. But don’t make this election into a disaster it wasn’t. Keep in mind the political climate. There was a hotly contested race for Cook County board president, and the media, particularly in Chicago, loves political controversy.
Where else in the country do you have drunken sailors, basically a bunch of drunks, frankly, coming downtown on election night to our vote counting headquarters and putting out all these outrageous statements? There never was, then or now, any proof that there was a conspiracy in the vote.
STUMP: But the Mikva Commission really went into the nuts and bolts of how electronic voting works. And my reading of the report is that there was a software interface problem . . . three pieces of software from different suppliers that didn’t work together.
ORR: But the results were never in question. The results were all calculated locally in the precincts by the judges without question. They had the tapes. The precinct workers did what they were supposed to. The judges did what they were supposed to. They used the “HAAT”, which is the card-activator, successfully. Our training worked. They just couldn’t connect to send those results downtown.
McLENNON: 90% of our judges did what they should and thought they had transmitted successfully, but there was this software problem. The user interface on the HAAT didn’t have a way to show judges in the precincts that the results were not received downtown. Essentially, there was no easy-to-read error message.
ORR: Every precinct had a scanner, a touch-screen (for blind voters), and this card-activator from Sequoia Voting Systems that did three critical things. Two of them it performed excellently:
1) The card-activator allowed us to give each precinct a unique ballot style. In a county as complex as Cook, different precincts have different candidates on the ballot so every card could be configured for a different polling location, and there were far fewer mistakes than usual as a result.
2) At the end of the day, the card-activator takes the electronic results and the paper trail results in each precinct, combines them and gives you, if you are an election watcher, a verifiable tape of the results.
3) The card-activator also is supposed to allow poll workers to wirelessly transmit their unofficial results – and I emphasize unofficial because the official results are on the tape – downtown, and this is where the main problem was. Only 56% of the precincts managed to connect.
STUMP: What about other technical problems?
ORR: A second key problem was that Sequoia’s WinEDS program [that tabulates all the results] was slower than it should have been, and our panel says it should be faster. And that’s important to us because Cook County is unique.
We have political demands on us that nobody else does. We are the second largest jurisdiction in the nation. We have probably the longest ballot in the nation, and, because we used cellular technology in the past to send in unofficial tallies, the media here expects that results will come in fast.
STUMP: The Mikva report also includes a timeline of all the problems you had with the software leading up to the November 2006 election. It cites all these little glitches that seem to demonstrate how unreliable these systems are.
February 27, 2006: “Early voting begins for the first time in Illinois history. . . . touch screens experience significant system problems and do not perform to expectations.” March 21, 2006: Primary Election Day “widespread ballot jams in optical scanners” July, 2006: “Sequoia does not perform a stress test.” August, 2006: “Sequoia retrofits optical scanners and retrofits HAAT’S.” September 7, 2006: “Sequoia discovers that they continue to have a problem with the HAAT’s corrupting memory packs. New controller chips are installed.” September 20, 2006 Test election in Cicero Township. “Judges . . . are unable to allow touch screen voting, or to consolidate and transmit results, because of a Sequoia programming error.”
What all this points to is the very nature of software development. There will always be new software and new upgrades. And it’s always going to have bugs, and pretty soon the bugs will get you. And it did get you. This time it was in the transmittal process, but next time it could just as easily come up somewhere else.
How susceptible are we here in Cook County to a massive failure not at stage four, but maybe at stage two next time?
ORR: What this points to is all the testing we did do prior to the election.
Nobody likes to go into an election with new stuff. If we had our way, we’d test everything in an election before we deploy it. But remember we were under a federal mandate. And there were unrealistic expectations because Congress passed a law way too late.
They didn’t get the E.A.C. [Federal Election Assistance Commission] going for almost a year and they didn’t have any phones for even a year beyond that. They were developing standards, but a lot of us county clerks couldn’t wait for those standards. We were negotiating with vendor companies -- in our case -- starting in 2004.
If you are asking how we cope with changing technology, I think it’s very important that election officials be on top of this process. That’s why we have asked the experts on our panel to stay on. We’ve asked Diamond Consultants [who produced an independent assessment of Sequoia’s performance] to stay in the picture. And we’re going to have our own technical project manager so we can conquer these problems.
Because we’re pretty smart here ourselves. We forced Sequoia to make several changes, hardware and software and a lot of the software upgrades between March and November were successful. We had the optical scanner redesigned and the card-activator was improved, and there were tremendous changes to the touch screen to make it work better for the disabled.
But because all these had to be certified, we didn’t get final certification of our system until the Friday before early voting started.
STUMP: The irony is that you are ahead of the game, nationally. You were among the first county clerks to demand a paper trail. You had a lot of input to the computer voting manufacturers –
ORR: And we still had a problem.
STUMP: But elections in America are changing very quickly; and it’s all software based, and nobody leaves their software alone. Even the players are changing. Sequoia was sold in the middle of your bidding process. Next time around, if it’s not the transmitter, it will be something else. Isn’t computer voting inherently fallible?
GAIL SIEGEL (Communications Director): We don’t believe it is. The software can’t just be changed willy-nilly. It has to go through re-certification. What inherently caused the problem last time around was the timeline.
ORR: We’re at the beginning of a whole new technological revolution in elections. And there’s a lot we don’t know. That doesn’t mean we’re not on top of things. We just pulled off an election with accurate results – despite slow returns.
But nationally, the question for a lot of smaller jurisdictions is, in the real world, how quickly can they integrate this sophisticated new technology into what are essentially local elections. They have to rely on testing by the federal government and the state governments, but the standards are slow in coming out. There are questions about these certification centers -- and the vendors paying for that -- and questions about whether there is too much privacy in the testing process. Up to this point, yes, the country has had to rely too much on the vendors.
In Cook County, it all goes back to the complexity of our ballot, offering ballot choices in different languages, assisting the disabled community – those are all good reasons to have computer-based elections.
But Congress moved too fast. They didn’t know what they were doing to some degree. They wanted to make some changes, but they didn’t listen to the clerks on how long this would take to implement. And federal oversight has been slow -- and barely there when it comes to these testing issues.
STUMP: Going outside Cook County for a minute, how do we know this paper trail works? In Cuyahoga County (Cleveland, Ohio), for instance, there’s a report that says when they checked the paper trail after the last election, they found 10% of the paper trail votes were blank or couldn’t be read.
ORR: My dilemma is that you’ve got an electorate that’s been led to believe none of this stuff works. Not by the David Dills of the world [Dill is a Stanford computer professor and founder of The Verified Voting Foundation], who have raised some legitimate questions, but by people who don’t follow the issue closely and media who just report electronic voting doesn’t work.
And you’ve got some election officials who simply don’t believe a paper trail makes a difference. So voters are filled with questions, like what happened to those 18,000 votes in Florida?
But it’s not the transition to electronic voting that’s causing us grief. Any election expert will tell you they had far more problems moving from the old mechanical machines to punch cards than we’ve encountered.
The real problem is that a presidential election was given to the loser in 2000 by The Supreme Court. It wasn’t even the hanging chads. There was a process, even though it would have been ugly and messy, to recount the Bush and Gore votes in Florida, but they never got to the bottom of who won the election because The Supreme Court wanted to step in. And that shouldn’t be forgotten.
A lot of the problems you read about in Ohio, or Florida, aren’t equipment issues. It’s people who didn’t get to register the way they should, or politicians finding ways to go around the law to help one party or another, and clearly they did. Every county clerk has been hurt because some people made decisions that were blatantly political.
Since 2000, elections are a fishbowl. All I ask is that people differentiate what errors occur in an election because we’re all humans and what errors happen because of electronic voting. Yes, we have to deal with changing software. But we can do that. Banks do. Homeland security does. The whole world relies on computers. We have to bring this burgeoning new industry forward with a certain amount of accountability.
STUMP: Cook County had the same problem with butterfly ballots as Florida did in 2000, isn’t that right?
ORR: That’s blatantly false. For one thing, we never had butterfly ballots. We used punch cards. And we had a problem in Cook County partly because -- this is how I would put it -- some right-wing guy in DuPage County (Sen. Pate Phillips) said you cannot use the new protections you just bought even though we are letting the Republican counties in the state use them.
McLENNON: We had new equipment in 2000, and the equipment was designed so if there was an under-vote or over-vote, the ballot could get kicked back and voters would have a second chance to review their ballot. But Pate Phillips championed a bill in the Senate that said voters in Cook County shouldn’t have any advantage that other voters didn’t have. And it passed. So we weren’t allowed to use the new machines.
ORR: But we did have a greater fall off than we wanted? Of course.
STUMP: “Fall off” is an interesting term. My understanding is that it’s the difference between how many people walk into the polling booth to vote and how many votes are actually counted. In 2000, that was 5% in Cook County, and 7% in the city of Chicago. Even last November the fall off was around 2%.
That’s a lot of votes to just disappear. Any time somebody loses by less than a 51-49% margin, isn’t the loser going to wonder what happened to those votes?
ORR: The normal standard for fall off is 2-3%. But you have to understand that, as hard as we work to make elections run smoother, individual voters sometimes make mistakes. There’s no system that can protect those people.
If you vote on a punch card and the chad doesn’t come lose, that’s the machine’s fault. It’s an error and it has to be dealt with. But there are other things that voters do. With punch cards, you have to put it in a certain way and some voters put it in backwards. There’s a certain amount of human error that occurs, and if you track it down, you’ll see it happens most often in areas of the city with maybe less education.
That’s why we went ahead with electronic voting. We know that if you use optical scanners, and to a lesser degree, touch screens, not only does the fall off rate drop, but it drops most in some of these minority areas.
STUMP: So the real question is what constitutes a fair election?
ORR: The national standard used to be 2-3% fall off. But you never really know.
I’m not concerned with legitimate fall off. There are voters who say I don’t like any of the top candidates, but I want to vote for my friend to be water commissioner.
We’ve had races in Illinois where 20 to 30 percent of the voters in some precincts didn’t vote the top of the ballot. So we have to be very careful about analyzing why people don’t vote.
With electronic voting, the expectation is that the fall off rate will go down even lower. And we know that not only does fall off go down, but you radically change the voting discrepancies caused by race, or class, or education.
Electronic voting provides for better elections, especially in big jurisdictions. We have to configure different ballot styles for about 5,000 precincts in Cook County – and you’ve seen the Chicago ballot so you know how long they can get.
With electronic voting, all of a sudden you don’t have the problem of a precinct getting the wrong ballot, or some judge directing you to the wrong machine.
Now think about printing those ballots in three languages, as we do in Cook County. All these options can be accessed electronically. Add in the value touch screens give the disabled, and electronic voting is clearly better than the old way.
Plus, and this is important for all the voters, they can see how they voted. They have a backup so they know if they make a mistake. Touch screens are even better at this than optical scanners. If you push Candidate Smith, a big check mark comes up next to Smith. Compare that to those little holes in the punch card. On the touch screen, there’s a review option that shows you what races you skipped. And there’s a paper trail. So there’s really a triple check.
Because of the technology, we can provide far more protections for people. And there’s no question it makes it harder to cheat.
STUMP: Do you think that’s important in Cook County?
ORR: It’s important everywhere. One of the reasons we have all these redundancies is to make it really hard to cheat. This notion that elections somehow in the past were so much fairer is absurd, particularly if you are from Chicago. Paper ballots were much more unreliable. If you didn’t like the vote, you just go in a backroom and change it.
And if you ask the voters -- which we did in a survey of 1,000 early voters last November -- 98.5% said they were very pleased with the new machines.
STUMP: Let me clarify something I didn’t understand. The Mikva Commission said it didn’t look into the accuracy of the paper trail because it wasn’t disputed. Has anyone checked to see whether the electronic tallies last time were accurate?
ORR: Yes. By state law, we are required to audit 5% of the precincts in every election. They are chosen randomly, after the polls close, and we go back to see that the paper trail matches the electronic tally. It’s very time-consuming, but we do it because we believe in it. Someone holds up the paper tally and reads off the results in each race, then someone else reads the electronic results.
STUMP: Did it match? One for one, in every precinct, in every race?
ORR: Yes.
If you would like to be an election judge in the upcoming election, Chicago residents should CLICK HERE. Cook county residents outside Chicago should CLICK HERE.
