Voting in this Country is a Mess – Let’s Fix it Already.

I recently read a couple of really interesting articles about two big voting/election issues the United States is currently facing – the system we use to elect our politicians, and the process we use to do it. And while the status quo falls short on the trust and transparency we should all demand, there are some pretty good alternatives. As always, changing is the tough part.

First is a piece in Mother Jones titled “The Verdict Is In: Our Voting System Is a Loser“. This interview with William Poundstone, author of the forthcoming Gaming the Vote: Why Elections Aren’t Fair (and What We Can Do About It), discusses our current election system and some interesting alternatives. It turns out that plurality voting (one vote for one candidate), the most commonly used system in the U.S. and Canada, is one of the least effective systems. It’s highly vulnerable to the effects of spoiler candidates (see Ralph Nader in 2000 and Ross Perot in 1992), and ends up reinforcing the broken two-party system.

There are more alternative types of voting than I ever knew existed, but one of the best appears to be one called range voting. In this method, each voter rates every candidate on a scale (1-5, etc.), and the results are either summed or averaged to arrive at the winner. If you’ve ever rated a movie on NetFlix or a video on YouTube, you’re already familiar with the idea. The only real downside to range voting is that you can’t tally the results without the assistance of a computer. Which brings us to electronic voting machines…

Article two was published in the New York Times Magazine by Clive Thompson and is titled “Can You Count on Voting Machines?“. Thompson recounts the many problems of electronic voting systems and how they simply cannot be trusted. Ignoring the possibility for active voting fraud with these machines, they are just too unreliable and buggy to be depended upon. Even the current ones with a paper audit trail can’t be trusted, as their thermal printers run out of paper and jam too easily, creating the possibility of a paper trail that can’t be trusted.

That’s basically what any voting method, electronic or paper, comes down to – trust. For a purely electronic voting system, that means getting all of the voting machine vendors out of the process and replacing their blackbox software with an open, transparent solution that can be tested and audited by anyone. Personally, my favorite solution would be a paper ballot read by at least three different optical scanners, each based on verified open source software. This would give you multiple levels of assurance in the final vote counts:

1. The software would be audited by multiple parties of all political persuasions since it could be downloaded and tested by anyone. The certified version of the program could then be digitally signed before being distributed to polling places. The optical scanner could then check the running program against the official signature on every scan of a page, making it very difficult to slip in a hacked version of the software without raising a lot of red flags.
2. Running a piece of paper through two extra scanners would add a minimal amount of time to the process, but would easily identify units that were counting incorrectly.
3. There is always the official paper ballot, which could be used to “replay” a vote count whenever necessary.

Now, I’m sure there are some flaws in my plan above, but I think this gives us a trustworthy process that everyone can be satisfied with. What do you think?