Vote Tech

Dear Reader,

The nature of the print publishing industry demands that I write this column some time before you read it. The first copies go on sale two weeks after our deadline, and, depending on where you live in the world, you could be seeing this issue one month or even two months after these words reach layout. Print publishing lives on because it has many admirable qualities, but low latency is not one of those benefits. This introduction is my graceful way of apologizing that what I'm thinking about now is probably not what you're thinking about when you read this. I'm thinking about the election in the US, which is happening the very day I write this column. You already know who won, and you are happily free from having to think about it, but maybe you should.

A few days ago, the security vendor Cylance released a video showing how you can hack a Sequoia AVC Edge voting machine using a PCMCIA card [1]. You can change the votes (and even the names of the candidates) recorded on the machine. The Sequoia AVC Edge is used in assorted precincts in 13 states across the US. The community of activists who follow the electronic voting industry were not impressed with this demo, stating that this hack of the AVC Edge, which will also work on several other touchscreen-style voting machines, has been known for years.

At least the PCMCIA card hack requires physical access. Another voting machine known as the AVS WINVote was in the news a year ago when it was revealed that the WINVote system featured an option for wireless system administration using weak, hard-coded passwords for the admin account and relying on long-discredited WEP encryption. These WINVote systems use a version of Windows XP that hasn't received a security update in years. Experts estimated the system would take about 10 minutes to hack. The State of Virginia published a scathing report on the WINVote system, and, fortunately, all of these systems were taken offline in time for the 2016 election – but many years too late.

Why do these backward and dangerous voting machines still exists long after the experts have pronounced them inadequate? The answer is in the time divide that separates you from me. Here I am at the brink of the election. Election coverage dominates the US news, and even outside the US, the election is receiving significant attention. There you are after the election – after a year of constant and persistently annoying election coverage – and people are just dying for another topic. No one really wants to talk about insecure and ridiculously obsolete voting machines.

Why isn't anyone addressing this issue in a systematic way? Many outside of the US (and in the US for that matter) might not be aware of how decentralized the voting process really is. The headlines refer to a national election, but voters actually choose electors at the state level. The laws for the election are largely managed by the states, and the actual voting and vote tallying happens at the county or municipal level. Decisions about voting hardware might be made by the state, county, or city depending on the jurisdiction.

Buying new voting machines would actually cost a lot of money. Most counties don't have the money for it, and most counties don't have fair vote activists on hand to show up at committee meetings and demand new hardware.

But the problem has a reached a new urgency. This is the first election when a major-party candidate has tried to claim the election is "rigged." He is most certainly incorrect. (Despite the problems with these voting machines, the decentralization of the US vote means it would be extremely difficult for a group to systematically hack a national election without detection.) However, the accusation is destabilizing for national unity, and, frankly, the continued existence of clearly deficient voting machines plays into the hands of conspiracy theorists who thrive in a climate of distrust.

The best solution is to fix the whole mess. It is time for a uniform policy across the US establishing minimum standards for:

• security of voting machines
• uniformity of the voting process, including a verifiable, hard-copy record of the vote result
• an end to the alarming voter suppression practices that have gained popularity in some jurisdictions
• a non-partisan, scientific methodology for drawing legislative district lines that avoids the partisan gerrymandering tactics that both the major parties seem willing to use when given the opportunity.

Many organizations are already working on these problems. Check out the work of TrustTheVote.org [2] and its sponsor, the Open Source Election Technology Institute [3], for more on creating a fair and verifiable election process using open technologies. The Verified Voting site [4] provides information on voting machines used throughout the US.

Several groups are working on the problem of unfair legislative boundaries through gerrymandering [5], and a group of researchers recently unveiled a technique for determining gerrymandering in a legislative map using high-performance computing [6].

How will uniformity arise from a diverse system with 50 states and thousands of counties? Ultimately, the only way these voting problems will get fixed is if the people who are in a position to fix them will keep their focus after the election – and be willing to put the country's business before whatever personal advantage they get from letting it all stay broken.