Do we have time to get paper trails on electronic voting machines before November?

We’ve had several threads on this topic, but here’s an update.

“How They Could Steal the Election This Time,” by Ronnie Dugger, in The Nation, August 16, 2004 (http://www.thenation.com/doc.mhtml?i=20040816&s=dugger, printer-friendly version at http://www.thenation.com/docprint.mhtml?i=20040816&s=dugger):

So, I propose two issues for debate (which have not, as yet, been debated in exactly these terms):

  1. How serious is the threat that these electronic voting machines will not count or report our votes honestly?

  2. If that threat is serious, can we do anything about it between now and November? Is there enough time?

The threat is very serious. Even if you ignore the… curious results and practices we’ve seen with e-voting so far, the big target for hackers, tamperers, and other assorted low lifes comes this fall. And simply knowing the election was rigged might not be enough. What do you do when every Diebold machine registers a million votes for Kevin Smith? Throw out the election?

Every district with electronic voting machines will have paper ballots as well for absentee votes and power failure issues. We can ditch the machines and use those for now. Maybe by next election cycle the electronic units will be ready, but I doubt it.

I flubbed a couple of links in that last paragraph. Let me try again:

While the problems associated with electronic voting machines do raise some new challenges, the lack of a paper trail isn’t one of them.

For most of my adult life, I’ve voted in places that had the metal monsters, the machines with the little levers next to each candidate’s name. You’d pull down the levers for the candidates of your choice, and when you used the big handle to open the curtain, the votes were tabulated by the mechanical device. Or were they? I’m quite sure I remember instances of machines malfunctioning, and failing properly to register all of the votes cast.

There were certainly no paper ballots to be recounted. A “recount” consisted of double-checking the counter wheels inside the back of the machine, and adding up the results from all of the machines again.

In fact, I’ve never cast a paper ballot in any election. The lack of a paper trail is hardly a new issue.

But at least that was a trail. Something verifiable, something in a clunky mechanical machine where any tampering would be fairly obvious. Spotting tampering in these new electronic voter machines would be much harder, maybe even impossible.

From that same article in The Nation::

As a computer programmer, I can think of half a dozen easy ways to set it up so that one candidate would be much more likely to win (and subtly - how about one vote in 20 going to the candidate of choice? In this close election, that would be enough, and pretty closely within the margin of error in the polls.) It would be easy to include a ‘recoding’ of the machine in the shutdown sequence or when it turned midnight or at some other cue, so that the guilty software would never be detected - it would be gone.

This is not a conspiracy that would take large numbers of people - the primary argument against most conspiracy theories. Two correctly placed people could do it - one programmer, one person who allows that programmer to put his/her version of the code in at the right time. And this is just at the manufacturing point - lord knows what a hacker could do!

It seems pretty simple to add a printer to what is basically a computer, and write code to send each vote to the printer. Not exactly cutting-edge technology, and it sounds like it could be implemented extremely quickly (if the system already has general printer functions).

But, then all the hacker/rogue programmer would have to do is make sure his cheating code executes before the code that prints the record. Then the paper trail and any recount will fail to catch it.

The idea is to have the machine print a paper ballot right after the voter makes his/her final entry – so the voter can see that the paper ballot accurately reflects his/her votes. From the same Nation article:

It occurs to me that some patriotic hacker ought to pull just such a stunt: register a million votes for Kevin Smith, just to expose the flaws in the system.

Or maybe rig it so that Ralph Nader wins, say, Georgia outright.

IF the system can be hacked (and I am not smart enough to know whether this is so), then it seems to me the most patriotic think a concerned and computer-savvy person could do would be to commit such a blatant hack.

The danger is not that some outsider will hack the system but that the machines (which are stand-alone units, I believe, not directly connected to the Internet) will arrive at the precincts prehacked – rigged by their manufacturers, or by some groups who have an inside track with the manufacturers. In other words, rigged by Republican sympathizers to yield Republican victories. Nobody appears to be afraid of such machines being rigged to favor Democrats or Greens or Libertarians, because none of them are positioned to do the rigging. The Repubicans are. Diebold is a major GOP contributor. In fact, all the voting-machine companies appear to have a Republican bias.

On that note – see the bolded paragraph in the excerpt from Dugger’s article in the OP. That might be what actually happened in Georgia in 2000: Cleland was expected to win re-election to the Senate, and Barnes as governor, by comfortable margins; and it is possible they would have done so, had not somebody stolen some of the machines and reverse-engineered them and gotten into a position to rig the vote totals.

Sorry, Georgia in 2002, not 2000.

This is a rather timely issue for me. I’m sitting here in Vegas at Defcon 12, sitting by the poolside with hundreds of geeks who are proud to be hackers and general computer security experts.

From some of the stuff I’ve seen here this weekend, there is no way I would trust an electronic voting system. In fact, they don’t trust anything electric even here - you have to register in cash, and none of the vendors accept credit cards. Someone hacked the air conditioning system before.

Anyway, the seminars I’ve been to so far have been very interesting, but mostly they reminded me about exactly how unsecure systems are - from both ends (outside users getting in, and inside administrators managing data). No way in hell I would trust that system when even a paper system leaves so many questions.

As a side note, I’ve been hearing a lot on various “civil disobediance” plots for the Republican Convention. I’m somewhat curious now to see it all happen, and how much of a reaction it will get. I don’t see how any system can be 100% secure, and it would have to be, because even a tiny hole would be total disaster.

There would be two types of hacks:

  1. A serious effort that would be similar to the plot of Office Space; a tiny change in the system to make a small change over a large amount of time. This would not be detected until it was possibly irreversable. This is the most dangerous.

  2. A grandstanding attention whore stunt, as the aforementioned votes for Kevin Smith. This would be immediately detected, and the main result would be a smaller disaster that I hope someone has a contingency plan for. O_o

People will cheat if they can. The voting machine type doesn’t matter. For most of the history of US elections, fraud has gone on in one form or another…by all the major political parties. We’ve always managed to muddle through. People have stuffed ballot boxes, used dead people to cast votes, etc etc. Electronic voting machines are no different. Sure, there is potential for abuse. But there is no more potential for abuse than any other system. The ‘scary’ thing for people is they dont understand how the machines work…seems like magic. No paper trail…oh my, how will we recount?!?

The reality is the systems use secure and encrypted databases. Could they be ‘hacked’? Sure they could, though its my understanding that the voting machines aren’t connected directly to the internet, but supposedly through secure point to point connections to a central repository. The physical machines though, they surely could be hacked. Guess what? So could those old clunky voting machines. ‘Hacking’ just means bypassing normal operations, and I guarentee you that you can bypass the normal operation of an analogue machine if you really wanted too.

Ok, so you add in a few lines of code that adds an extra vote for every 20 to a certain candidate…and you are a god programmer and able to do this without a trace (it IS possible). What voter registration numbers are you using for those field in your data base recording of the vote, AvhHines?? Random? Or do you have a list of ‘valid’ voter registration numbers you can use, perhaps those of dead people? How is this different than the ‘regular’ voter fraud that political party machines have used in the past?? Actually, you’d be somewhat MORE vulnerable than they are, because if you tried to insert those numbers for your 1 in 20 it would be trace-able by looking at the code. Again, sure you COULD do it, but then you can cheat with nearly anything.

They are blowing sunshine up your skirt. If it were that easy, it wouldn’t be the election to worry about…money would be the prime goal. I’m in IT and security, have been for 15 years and even own my own IT company…I know what can and can’t be done. If these guru’s could hack and get away with credit card info or other financial data that easily, they’d all be millionares…hell billionares. They aren’t. The fact that the hotel is stupid enough to play up to them and make all transactions cash…well, there are a lot of gullable people out there. Most people and places that get their credit card info ‘hacked’ do so because they were careless…and a lot of times its a PHYSICAL security breach, not some fancy hacking routine.

Again, this isn’t to say that its impossible to ‘hack’ the election, though I’d need to see exactly how the machines were configured as far as their infrastructure goes. At a guess and from what I remember from talking to a friend, the machines aren’t directly connected to anything that can be directly accessed from the outside (i.e. they don’t have an internet connection). Now, the central repositorys perhaps could be hacked (i.e. where the various districts voting records are tabulated electronically from the machines) but guess what? So could the central repositories for the old fashioned records (remember, ‘hacking’ doesn’t have to mean fancy electronic wizardry…it just means circumventing normal operations).

There are certainly issues with electronic voting we need to look into. There needs to be a ‘neutral’ organization with access to the code for the voting machines as well as the tabulation machines and the data bases from the machines. There should be a way to do a check sum on the data, in case the physical machine itself was hacked (the most likely scenerio IMO). The links should be secure, encrypted point to point PRIVATE connections…no connections at all to the internet or any other pubic connection. I have no idea how many of these things HAVE been implemented (most would be my guess), but those and a few more SHOULD be implemented. There doesn’t need to be a paper trail if these things are done…it would be redundant and unnecessary.

-XT

The real challenge isn’t doing the crime - it is not getting caught.

It is a simple matter to get the credit info and account access. The TRICKY part is doing something with it without leaving a nice trail right to yourself. Similarly, it would be rather simple to commit fraud. Anyone can do that. The hard part is to get away with it, and that is why it isn’t practiced almost constantly.

However, the national presidential election is worth much more than stealing a few thousand dollars on pilfered credit cards.

The difference should be obvious, xtisme. Stuffing ballot boxes, in order to make a big enough difference to win a national election, has to be done on a large scale. Can you have one or two people fly to every precinct in the country (or even every counting location) and somehow invisibly slip in enough false ballots to win without being noticed? Sounds like Santa Claus delivering presents to every child in the world on Xmas Eve - it can’t be done. Therefore it must be done by lots of people, and the possibility of detection is very high. There are, by definition, physical traces, and opportunities to be detected. And, in case you’re not aware, many safeguards have been put into place to prevent the kinds of voter fraud that used to take place - people can vote only in a single location, and are basically signed in and checked off as they do so. I’m not saying cheating can’t take place in a paper ballot system, but I am saying it would have to be on a large scale to make a national difference, and that it would have to involve a lot of people in on the conspiracy.

Electronic voting machine fraud, as I mentioned, could be done by two people, for every location in the nation in which those machines were used. Three, if you want to have someone looking up the voter registration numbers of the recently (within the past four years) deceased. I don’t know if any kind of id of that nature, that connects an invidual with their vote, is stored within the systems.

Which do you think is more likely to succeed - a conspiracy with lots of physical evidence, that takes hundreds if not thousands of people to implement, or a conspiracry with no physical evidence that takes three?

And yes, I am a god programmer. Bow down before me, mortal :smiley:

Sorry to disappoint, but from the reports i’ve read (e.g. in RISKS) a lot of them use Microsoft Access.

Are the machines connected to a public network? If so, then the people doing this are making a mistake, even if they are using multilayered firewalls. It would allow a hole in their security with no real benifite. If not, I’d like to know how you would circumvent this, unless you were in on it and on the inside. You simply can’t hack a secure private point to point encrypted link from the outside, unless you are part of the telco providing the service…and even then you’d need some pretty high powered equipment to break the encryption if the people setting it up had half a clue. And if you were on the inside, there would be easier ways to hack the system. However, same goes for analogue machines…if you have people on the inside it becomes easy to hack a lot of machines…no need for ‘Santa Clause’. :slight_smile:

As to whether or not a voter ID is one of the fields in the vote data base, I assume its in there somewhere. I seriously doubt that the data base record consists of a single vote field containing the candidate voted for. Somewhere there is a record of WHO voted, even if its not key linked to the actual vote…and that means that if you are going to monkey with the code to allow for a vote in 20 for your candidate you are going to need to also put in a voter ID code. Again, this is an assumption, though I have a friend that worked on an electronic voter systems a few years ago, and at least her code worked that way.

And I’m a god like IT engineer. Everyone I’ve ever met in IT was an arrogant son of a bitch…and with good cause. Millions, even billions of dollars rest on our decisions, and we are all so young (well, I was when I first started anyway). :smiley:

-XT

I’m thinking in terms of being on the inside, rather than hacking. Electronic voting machines are not performing a very sophisticated function, and it would be easy enough to have the “right” programmer in the right place at the right time.

No, x, they wouldn’t do it by adding votes but by altering them. Every 20th vote for Kerry is recorded as a vote for Bush. Much easier to get away with. Might already have been done (Georgia in 2002).

Check out this thread: http://boards.straightdope.com/sdmb/showthread.php?t=264462