Electile Dysfunction

It would seem that election crazy is happening in the US again. I’ve no idea what will happen tomorrow, but whatever the result some people are going to be unhappy - either because their candidate lost, or because they feel the election was stolen in some way. Who knows, maybe those pesky Russians will weigh in again like they did in 2016? 😂

I’ve no idea who Mick Foley, the author of the above Tweet, is, but I suspect he doesn’t wear red baseball caps with acronyms on them.

What’s fascinating, though, is the logic that lies behind the Tweet. Whilst it’s not explicitly stated, the implication is clear. If you deny an election result, you deny ‘democracy’.

It rather depends on precisely why one is said to be ‘denying’ a result. Indeed, we might even ask what, exactly, does it even mean to ‘deny’ an election result?

Mick Foley is far from being a lone voice. It seems to be the official position of the current US President that voting Republican is the same as voting for the destruction of democracy. Mind you, a significant fraction of the US population would probably agree with the description of the current President as their Commander-in-Thief.

I’m going to go way out on a speculative limb here and suggest that the vast majority who ‘deny’ the 2020 election result (or, for that matter, the 2016 result) do so because they think the result was unfair in some way, that the process was flawed.

It seems rather strange to describe people who want a process to be improved, to be made more trustworthy, as being in favour of removing that process altogether.

I watched the US 2020 elections with interest. There were, for me at any rate, a significant number of alarm bells going off in the run up and on the night itself. Obviously, I’m one of those people who wish to destroy democracy because I’m not at all wholly convinced there wasn’t widespread manipulation and fraud going on.

Whilst I wouldn’t describe myself as an election ‘expert’, I do have some passing familiarity with the issues involved in securing an election. In my time at an industrial research lab, I worked as a cryptography specialist as part of a wider security research team. One of the projects I worked on was an EU-funded project on electronic voting involving teams of researchers from across Europe.

In my view, for example, anyone who thinks an election can be properly secured via mass mail-in ballots is either duplicitous, or a moron (or possibly both).

You can’t design a good crypto algorithm without being also good at cryptanalysis. In order to design and build good security you need to know what you’re securing things against. You need to be able to anticipate the kind of attacks that might be made.

In terms of an election, you would start off with a list of properties you might want to achieve. In broad terms we’d start off with a list of high-level objectives

• Integrity : the vote cast for one candidate cannot be assigned to another candidate

• Authentication : only those who have a legitimate right to vote can cast a vote, and only one vote

• Privacy : the specific voting choice must be secret

• Transparency : the process must be auditable so that the objectives can be demonstrated to have been achieved by an independent observer

Then you might start tabling suggestions as to how these things can be achieved. So, for example, one suggestion might be a postal vote. You’d then look at how you would achieve the required objectives using this methodology. How could we ensure that a postal vote, when received, went to the candidate voted for? How can we ensure that this person has submitted only one postal ballot? How can we ensure that the identity of the voter is not associated with his or her choice of candidate? How can we audit this process to demonstrate these things?

You’d then start to look at all of the things that could go wrong. You also put yourself in the position of someone who wants to subvert the election and then figure out how you would do it. You hope you can think of all of them but, realistically, you’re going to have to accept that someone will probably think of some really clever way to fuck things up that you didn’t.

Having thought of umpteen things that could go wrong, or things that could be maliciously done, you then adapt your proposed method to combat those things. Having done that, you start again and think of ways to subvert this adapted, improved, methodology, or ways in which this adapted method could go wrong.

You keep iterating until you have a system that would be decently robust - as far as you can tell. By ‘decently robust’ I mean that the goal is to design a system that may not be perfectly secure, but one that would require a very special set of circumstances for (innocent) failure, or that would require prohibitive resources to maliciously subvert.

A very easy way to ensure a reasonable level of authentication, for example, is to require voter ID. It’s not foolproof, by any means, but it raises the bar for those attempting subversion. It’s such an easy thing to implement, especially because of the near ubiquity of other forms of ID (passports, driving licenses, etc).

It’s not a methodology for voter suppression, it’s a methodology for illegitimate voter suppression.

Anyone calling for relaxation of the requirements of authentication in an election process must be viewed with extreme suspicion. These people could be described as a potential true threat to democracy.

It is a balancing act. Implementing security is often subject to the law of diminishing returns. We can make a certain system very secure, for example, but in so doing, also make it almost unusable for the users of that system. When you’re trying to design good security, you have to keep an eye on all these things.

The ‘argument’, such as it is, for not requiring voter ID is based on ‘usability’ issues. Requiring voter ID makes an election too ‘difficult’ for some people, it is argued. But is an election a serious thing or not? We don’t let people drive without the requisite ID because of the serious consequences of not doing. Do we not view elections as serious enough to warrant such basic level protection?

Looking at the US (admittedly from the outside) I see very worrying trends in the acceptance of postal ballots (these should only be allowed for a very small number of people with legitimate reasons, and with a heightened authentication process attached), and calls for the relaxation of authentication requirements, for example.

If these trends continue, I fear the US will suffer from electile dysfunction for many years to come.

---