Well, today's the day that the U.S. votes, yet again. Today, I woke up, ate breakfast, and voted. It was practically the first thing I did this morning. It's that important. If you haven't done so, you should as well.

We've had electronic voting machines in our county for a while now; I think I've voted three times on them. I actually didn't think to look at the brand of voting machines that we have, but I do know that today and during the last election (where Californians said no to Schwarzy's propositions), paper receipts are printed out when you vote. You have the opportunity to confirm them, as well, before casting it as your official paper receipt. (I did take my camera, but didn't end up taking any pictures. I should've snapped a couple of the voting machines themselves.)

The system seems to work well; I didn't have any hiccups during my vote, and I didn't observe any irregularities either. I did overhear a poll worker say that some voters were confused that you actually print the paper receipt first, and then confirm it before finally casting your ballot. That seems to be the best way of doing it, since you can actually make sure that your paper receipt records your vote correctly. (I didn't try going back to the voting screens after printing a paper receipt, so I don't know what gets printed in order to invalidate that partial paper receipt.)

But do we really need electronic voting?

Electronic voting, at worst, seems to be a solution in search of a problem. At best, it's a system that, in theory, would work better, but in practice is just creating all sorts of different problems. What are we really trying to fix by changing over to an electronic voting system? What is wrong with traditional punch-card voting? (I realize that there are other voting methods, like optical voting systems which, I think, are kind of like "scantron" ballots. But let's just put that issue aside.)

There are two answers I can come up with: counting accuracy, and counting speed.


Counting Accuracy

The 2000 elections were the first that really pushed the former "problem" to the surface. Because the presidential race between Bush and Gore was so tight, recounts and invalidations that changed the number of votes by only a few hundred actually mattered. Issues of butterfly ballots, which confused voters, and hanging chad, which caused incorrect ballot counting, were the primary issues that people had with paper voting.

But these two issues can easily be solved without electronic voting systems. 1) Eliminate the butterfly ballot style entirely: have sample ballots and voter information packets list candidates down one page, and then down the other. Make sure that all candidates for a specific office are on a single page, creating blank space where necessary in order to eliminate confusion on where to punch to vote. 2) Instruct both voters, poll workers, and election officials to check for any irregularities in the punch-card ballot that could cause a misvote or a miscount -- i.e.: it's fairly trivial to simply run your hand down the back of a punch-card ballot to eliminate any pieces that didn't fully get punched off.

Granted, it would be better if we didn't need to worry about problems like hanging chad in a punch-card ballot, but consider the following: if electronic voting machines are to replace traditional voting machines, they need to be unequivocally better at preventing voting inaccuracies. And that's really hard to achieve.

With electronic voting, at least in the current system, the paper receipt is an afterthought. Electronic voting machines were first introduced by themselves, and only when voters and constituents complained did states start to require paper receipts. The point is: even if there are irregularities in the process of electronic voting, they have to be detected first before the paper receipts are even looked at.

In light of the fact that creating a voting virus is not only possible but undetectable on current Diebold voting machines, this is especially troublesome. Consider an election in a toss-up race based on polls done prior to the election. If a voting virus was planted that gave just a few more votes to a certain candidate in a few districts, there wouldn't be any reason to necessarily even consult the paper trails created by the electronic voting machines. Even if there were a required recount for races that were decided by a small number of votes, there still would be cases where deliberate tampering with electronic voting machines would escape scrutiny. And since there's no way to detect tampering electronically, the paper receipts that may have revealed the problem wouldn't be looked at.

The point is, the only way to ensure that there are no irregularities is to make the paper receipt an integral part of the voting system. That is, the paper receipts should always be counted, and the electronic votes should always be counted, and the results from both should then be compared -- any discrepancies should be thoroughly investigated and accounted for before releasing the official vote count. But if you do that, why not simply use paper ballots anyway? With electronic voting, you're just adding another layer to the system which is completely unnecessary.

The Princeton study on creating a voting virus shows how easy it is to game the electronic voting system, something not easily done with paper ballots. Having a specific hole for a candidate pre-punched, for example, is surely going to catch the eye of many voters; the fact that the ballot is a physical object and that voting creates a physical change to the ballot itself practically guarantees that there's no way to game the system, save for voter disenfranchisement -- something which, by the way, can happen even with electronic voting machines.

Here's another way in which electronic voting causes problems: consider the ironic situation in Texas' 22nd congressional district. Tom DeLay got indicted because of his link to corrupt lobbyist Jack Abramoff. But not before he got his name on the ballot. He tried to "move" to Virginia to get the Texas courts to strike his name from the ballot, but they didn't buy it. So DeLay's name is still on the ballot, even though he's not running. That sucks for the replacement Republican candidate, Shelley Sekula-Gibbs.

With electronic voting, the cards are stacked even higher against Sekula-Gibbs. She's a write-in candidate, which means that voters using the electronic voting machines have to type out her name, letter by letter, using a trackball. Not only is it a painfully slow process which will frustrate voters, but there's also no hyphen character, providing an additional point of confusion.

Using a paper ballot, it's simply a matter of writing the candidate's name down, something which can be done in a few seconds, and which virtually everyone knows how to do. The trackball system makes it a few minutes, and that's if we're thinking of a voter that has good eyesight and good hand-eye coordination. How many 80-year old Republicans with arthritis are going to persevere in entering Sekula-Gibbs' name on the ballot? It's virtually a shoe-in for the Democratic candidate!


Counting Speed

This is the other "advantage" to electronic voting machines that I can come up with. But even with paper ballots, tentative results of the day's voting can almost always be released the next day. Exit polls and polls prior to election day already give a good overview of which way various races are probably going to go, and for the tossups you can just wait for the tentative counts.

Do we really need to know the results of elections any faster? This is a dubious advantage, at best, and one, in my opinion, that doesn't deserve much thought beyond two short paragraphs.


Other Problems

There are other problems with electronic voting machines. Who makes them? Who oversees the design? Who ensures a design that has the security of the votes of the American people in mind?

Right now, the answers to these questions are worrying. Diebold is one of the main manufacturers of voting machines, and it's a heavy contributor to the Republican party. Even if it was a heavy contributor to the Democratic party, I wouldn't feel comfortable with them making the systems that lie at the foundation of our nation's society.

Not only that, but Diebold is actively refusing to divulge the source code to members of the public, and even security researchers are having trouble accessing the source code to look for vulnerabilities. The company even has the gall to bully cable TV networks for airing documentaries that even suggest we have a problem with the electronic voting machines that we already use.

Shouldn't any member of the public be able to look at the foundation of our democracy and be able to check, by themselves, whether it's reliable, secure, and valid? With paper receipts, nothing is left to the imagination: if you want to know how the voting system works, just go vote! You see the ballots, you know what happens to the ballot when you vote for a certain candidate or proposition, and you know where that paper ballot goes. With electronic voting, there is an ominous black box that you touch or press buttons on, and magically a vote is counted. How do you know that your vote was recorded correctly? How do you know that the paper receipt printed out actually matches the vote that was placed by the machine on the voting card?

And why isn't our federal government requiring voting machine manufacturers to divulge the source code to anyone who wants it? The transparency in creating voting machines is a necessity if we can really trust the foundation of our democracy.


Back to Paper!

With electronic voting machines, we seem to have put into place a system that is non-transparent, fraught with problems (ones that are happening today), and easily hackable, three things that couldn't be said for the previous paper-based system. And yet we trudge on, we stay the course.

I'm not deluded enough to think that the nation will revert back to a paper voting system. But it still seems to be the right choice. Don't get me wrong: I'm all for pushing the boundaries of science and technology, but only when it offers obvious benefits, not just drawbacks.