How Secure are Electronic Voting Machines?
Election Day 2012 is just around the corner, and many Americans will be casting their ballots on electronic voting machines. But how reliable are these devices? Michael Alvarez, professor of political science at Caltech, discusses the technologies at your polling station.
IRA FLATOW, HOST:
This is SCIENCE FRIDAY. I'm Ira Flatow. Next up: How will you be voting this year? You will be voting, won't you? Electronic, paper, absentee? If your polling booth is in the path of Hurricane Sandy, will you still be able to vote?
Here to look at some of the technologies you might see at your polling station, or perhaps in place of your polling station, is Michael Alvarez, professor of political science at the California Institute of Technology, Caltech in Pasadena. He's also co-director of the Caltech-MIT Voting Technology Project.
Welcome to SCIENCE FRIDAY.
MICHAEL ALVAREZ: Thank you for having me on your show.
FLATOW: Hey, you're very welcome. You know, we've been talking about, what, since about the year 2000, electronic voting, I would guess?
ALVAREZ: Yes, we have. So our project, the Caltech-MIT Voting Technology Project, was actually spawned in the aftermath of the 2000 presidential election, in particular the problems that were seen in Florida with those voting machines, especially the problematic punch-card machines that were then used in Florida.
FLATOW: Ah, hanging chads. I thought I'd never say that again.
ALVAREZ: You can always stop now.
ALVAREZ: We don't want to hear any more about hanging chads.
FLATOW: Well, give me - let's talk about evolution of the voting - of the electronic voting machine. How would - how has it evolved over time?
ALVAREZ: Well, it's like any other, you know, process in our lives, where the, you know, electronic revolution, the computer revolution has really kind of hit it hard and produced all sorts of innovations. You know, the punch-cards themselves were seen many decades ago as a solution, as an innovation to the way in which Americans and others vote. It was introduced largely in urban areas in the U.S. to facilitate the counting of ballots, the rapid counting of ballots in complicated urban areas where they have lots of things on the ballot and lots of ballots to count, like Los Angeles County, where I happen to live.
ALVAREZ: What's happened over the last 12 years has been that those types of voting machines - the lever machines and, in many places, the paper ballots themselves - have been replaced with either an electronic voting machine - like we call them DREs, the kind of ATM-style voting devices - or optical scan ballots, which are the ballots that people complete by sort of filling in the circles or kind of inking in a little line to mark their ballot.
So they've evolved considerably. And the good news - we just did a report that we released recently, and the good news is that the work that we've done in Caltech and MIT indicates that these new voting machines that people are going to be using in this election are generally cleaning the process up in the sense that they're producing more accurate voting. They're producing more reliable voting for folks.
FLATOW: Hmm. 1-800-989-8255 is our number, if you're going to be - you'd like to talk with us about voting, electronic voting, maybe your experience with it.
When you say, Michael, that you've cleaned it up a little bit and they're producing accurate votes, can you be more specific on what you mean by that?
ALVAREZ: Yes, certainly. So when, you know, back in the 2000 election and earlier prior to that, when we used punch-card ballots or we used lever machines or even optical scan systems, when people dropped those votes in the ballot box, they didn't have any technology that, in most places they used, to check those ballots for common errors. And now what'll happen when people show up to vote in person on Election Day - or in early voting, in-person early voting - is there will be some technology in place to either scan their optical scan ballot, or the DRE machine will be programmed to check for simple mistakes like over-votes and under-votes.
And what we found is as states and counties have moved to replace the punch-cards, the lever machines and the old technologies with the newer technologies especially that allowed voters to check their ballots for mistakes, that there seem to be fewer mistakes.
ALVAREZ: And that's a very good improvement. Now, the flipside of this, of course, as I think you mentioned in the intro, is that many people will be voting by mail.
ALVAREZ: And when you vote by mail, you don't have the opportunity to use those technologies to check your ballots for mistakes. And you also have to be careful that you make sure that you return your ballot with enough time so that it actually gets back to the election officials so that they can include it in the tabulation.
FLATOW: You know, that's such a cumbersome process, and we were talking about it today. In a real case, I was talking about someone who just voted absentee in Pennsylvania. And the ballot is - has to be even folded the right way.
If you don't fold it the right way and get it back in the envelope, if there's a stray mark of pen on place, the ballot is so old, it asks you to mark it in fountain pen if you have it.
FLATOW: You know? It almost seems like it's made to discourage you or to make it harder for you to vote.
ALVAREZ: Well, you know, I think that, really, the bottom line is that, you know, these ballots that we cast by mail, they need to be read by, you know, high-speed ballot reading machines or high-speed scanners, essentially, that are reading these ballots. And those have to, you know, be formatted, laid out and marked in very careful ways. And so, I think that voting by mail is incredibly convenient. Voters like it. Election officials like it because it takes people out of the polling place on Election Day and reduces lines.
But voters have to be extremely careful. They need to follow those instructions. They need to make sure that they check their ballot carefully so that they don't make common mistakes. And they definitely need to make sure that it's either postmarked on time or they drop it off using whatever appropriate procedures exist in their state or county.
FLATOW: Now, it, you know, the perfect solution and if we could make it perfect, I guess, would be to be able to vote by your cell phone, your iPad, your tablet, your Android, whatever. Fill out the ballot right there and put little checks in there and press the send button. And then you wouldn't have to worry about that, all those different problems except the security problem.
ALVAREZ: Well, that's the bottom line, is that, you know, in some countries, in Estonia, for example, other nations are experimenting with various types of remote Internet voting. You know, you can imagine that those systems exist and some countries use them, and we certainly are studying those in the work that we do here at Caltech and MIT. But in the United States, we have a very different problem than in many other countries. We have much more complicated elections.
And we really do have, seen, you know, the rise of concerns about the security of electronic voting systems. And those security concerns have really kept the U.S. from moving down that path towards online voting. I mean, it may be off in the future at some point, but there are many security issues that really need to be resolved before we'll see that.
FLATOW: All right. Now I have a suggestion for you. I don't know if anybody's thought of this, but I thought of one of the ways to solve everybody's problem with the voting system. Let's say you could have the facsimile of a paper ballot. In order words, on your phone or your tablet, you have, you know, a - what looks like a paper ballot and you can check it off and do all of those things. And then you take that thing over to the scanner, and it's like you go, when you're checking into an airline, you know, it scans your boarding pass.
Why can't it just scan your telephone or your iPad to see you voted? And then they would never leave your hand, and it would go right through the machine?
ALVAREZ: Well, you know what, it's interesting. And that that's a very good idea. And that idea sounds a lot like one that was discussed in the report that we wrote 12 years ago.
ALVAREZ: We released that report in 2001, and the idea there was to develop systems that's, sort of, like what you're talking about, where a voter would have an opportunity to get a ballot in some way, shape or form. It couldn't even be electronic. They could deal with it however they wanted. They could mark it up electronically or print it out and...
ALVAREZ: ...do something with it. And then take that to a polling location, authenticate themselves and cast the ballot. The only downside to, in particular, to the idea that, you know, you've just advanced is that we need to be very careful that those ballots are kept secret and they're kept private so that, you know, voter's don't have coercion or get, you know, all sorts of other problems associated with potential election fraud.
FLATOW: Mm-hmm. 1-800-989-8255. Let's go to Jason in Reston, Virginia. Hi, Jason.
JASON: Hi, Ira. Thanks for taking my call.
FLATOW: You're welcome.
JASON: I just had a question about the integrity of the software inside of the machine. Who actually verifies that it hasn't been tampered with or that, you know, that we can absolutely trust the software and that it hasn't been designed by partisan people?
ALVAREZ: Well, that's a good question. And what happens in states and counties when they acquire a voting system is it does go through a process of testing and certification. And then most - in most places prior to the, you know, any particular election, there's another round of testing and certification to try to ensure that the machines and the software inside them are working in the way that they should work. And, you know, we live in a country where private enterprise has been involved in the production of and acquisition and selling of voting systems.
And so, you know, there are these concerns that arise from time to time. And, you know, again, I think that, you know, there's a fair bit of trust that we have in the integrity of our election officials, in particular, that they have, you know, gone and tested and verified the integrity of these systems. You know, by and large, you know, our elevation officials do a wonderful job, I think, given the resource that they have to test and certify these systems.
FLATOW: Let's go to Chris in Cape Cod. Hi, Chris.
FLATOW: Hi there.
CHRIS: I'm calling because I'm a computer professional. I work with computers in my daily job. And I would never trust an electronic voting system. I think that everyone should be able to participate in a recount. No special skills should be required to assist in a recount or understand what's going on. I think that we can afford paper ballots, and that, you know, electronic tabulation is fine. But everyone's ballot should be on a piece of paper for someone to inspect at a future time. And anything short of that will basically be compromising our democracy.
FLATOW: Chris, when you vote at your polling station, do you fill out a paper ballot?
CHRIS: I do.
FLATOW: Is it the kind that gets scanned in the machine that you take it over to?
CHRIS: It is.
FLATOW: So don't - wait, don't you have that paper ballot as a record (unintelligible)?
CHRIS: Absolutely. No. My town is fine. But there's all kinds of Dibold machines all over the country and lots of places where - are advocating touch screens. And I might add that a lot of very partisan companies own and control this technology. So, you know, there's obviously no proof of any sort of tampering. But there's no way to absolutely prove that there isn't, would be my point.
FLATOW: Michael, how do you respond to that?
ALVAREZ: Well, you know, I think he's got a very good point. And, you know, for a long time, we have advocated that the electronic voting machines, these DRE-style voting machines ought to be equipped with some form of voter verified paper audit trail. And many states have adopted that. And again, that's a system where there are actually two recordings of the vote. There is what the - there's the electronic recording and there's a paper recording that the voter also can look at and verify. And so we have an independent way of trying to verify the integrity and accuracy of what's been recorded by the electronic voting device.
The other thing that we've been working on, and a number of people working on this outside of our project as well, have been to really stiffen up the ways in which election officials engage in post-election ballot auditing. So after the election, how they go forward with trying to verify the accuracy of the tabulation on the electronic side, but also how they try to verify the tabulation of the paper-based ballots. And so, in combination, we think that we can get to a place where, you know, we can have a high degree of confidence in the integrity of the outcome.
FLATOW: This is SCIENCE FRIDAY from NPR. I'm Ira Flatow, talking with Michael Alvarez, professor of political science at Caltech. What are - the chances are this might be a very close election on election night? Why do we see these bottlenecks in counting the vote? Are those strictly paper ballots or is there something going wrong with some of the electronic tallies going on there?
ALVAREZ: Well, I think that the - a lot of times when we see kind of late election results or results that don't kind of come in as quickly as some people might expect, that's because of a lot of the ballots that are brought in, either by voters who are dropping off absentee ballots, it's the tabulation of late absentee ballots. And it's also the election officials who are trying to deal with many provisional ballots. And so there will be states like Ohio, for example, where typically there have been a large number of provisional ballots being cast.
And those all have to be examined after the election. They need to check the voter - with the voter registration roll to see if those are indeed eligible voters, and then those get included in the tabulation. So it tends to be the result of a lot of ballots that are coming in outside the typical pathway that votes are being cast.
FLATOW: Let's go to Skip in Tampa Bay, Florida. Hi, Skip.
SKIP: Good afternoon.
FLATOW: Hi there. Go ahead. Skip...
SKIP: I want to make a comment on the DRE machines that were mentioned earlier in the broadcast. And 25,000 of those machines were taken out of service in the state of Florida as they want to scan ballots instead of - from touch screens. And essentially, the source code in those is always secret. They have thousands and thousands of (unintelligible). Some of those machines were sold to West Virginia, some to South Carolina, and they're in service today, and perhaps, Erie County, Ohio. And they were found to be defective in the state of Florida.
FLATOW: The same machines without being fixed, you're saying?
SKIP: Well, they said they'd probably refurbish them, but they were broken from the start. So the first time electronics came out in 2001 and it continued to support under votes all the way along, which is where you go to vote on a race but don't vote on it. And essentially, racked up thousands and thousands of under votes that weren't put into the machine to begin with.
FLATOW: All right. Let me get a comment because we're running out of time. Michael, any comment on that?
ALVAREZ: You know, I don't know anything about the specifics of those. So I can't really comment.
FLATOW: OK. One last question before we go. Here in the Northeast, there are several states that are still dealing with the mess from Sandy. Could this have a huge impact on voters and voting machines?
ALVAREZ: Well, it's certainly going to have an impact on anybody who's been displaced. And, you know, I think that the - as far as I can tell, the election officials throughout the states affected by Sandy are scrambling right now to try to ensure that they can facilitate the process for, in particular, those voters who've been displaced. It's likely to slow things down. I mean, I wouldn't be surprised to see, you know, kind of slow returns for many of the states in the Northeast.
And, you know, again, the election officials are scrambling to figure out what to do. And we're just going to have to wait and see what some of the effects are as we get into Tuesday and Wednesday.
FLATOW: Well, a week from now on SCIENCE FRIDAY, we'll know what the effects are and the results and maybe we'll be talking about it. I want to thank you, Michael, for taking time to be with us today.
ALVAREZ: Oh, thanks for having me on your show.
FLATOW: You're welcome. Michael Alvarez, professor of political science at Caltech, co director of the Caltech/MIT Voting Technology Program. Transcript provided by NPR, Copyright NPR.