Selfie Slow-Down

I have already blogged about one American judicial decision on the constitutionality of a “ballot selfie” ban, which has since been upheld on appeal by the Court of Appeals for the 1st Circuit. And I have also written about the history of the secret ballot, which in my view explains why measures to protect ballot secrecy ― including bans on something that might at first glance appear quite innocuous, like a selfie showing for whom a person has voted ― are actually more important than they seem. Another American decision issued last week, this one by the Court of Appeals for the 6th Circuit, provides some additional food for thought on this issue.

Much of the discussion in Judge Sutton’s majority opinion in Crookston v Johnson is procedural. The case came up as an application for a preliminary injunction preventing the enforcement of Michigan’s prohibition on “exposing marked ballots to others”, (1) and Judge Sutton concludes that it is simply too late to grant one now in anticipation of the elections to be held on November 8. The people who will be running the election have already been trained and have received specific guidance on photography at the polling stations. Changing the rules at this point would create unnecessary confusion. So Judge Sutton does not rule on the merits of the case, which will be assessed later, assuming the applicant still cares. (This situation is reminiscent of the Canadian cases about election debates, which are invariably brought on an emergency basis when the debates are set up, and invariably abandoned before a full merits hearing once the election has taken place.)

But Judge Sutton does make some comments that bear on the merits of the dispute, and, although preliminary, these comments strike me as quite sensible and interesting. One observation is that

many Michigan voting stalls … are simply tall desks, placed next to each other, with three short dividers shielding the writing surface from view. In this setting, posing for a ballot selfie could compromise the secrecy of another’s ballot, distract other voters, and force a poll worker to intervene. (4)

My memory of Canadian voting stalls is a bit hazy ― I skipped the last election because I couldn’t tell which of the parties was worst ― but something like that might be true of them too. And indeed, even if it is not in any given case, it is worth thinking about whether our voting arrangements must actually be planned so as to cater to the “needs” of people wishing to snap a selfie.

Another practical point is that allowing ballot selfies could create a “risk of delay” at the polling stations, “as ballot-selfie takers try to capture the marked ballot and face in one frame—all while trying to catch the perfect smile”. (5) In a brief concurrence focusing entirely on the issue of delay, Judge Guy makes the additional point that “with digital photography, if you don’t like the way you look in the first one, you take another and so on ad infinitum.” (7) He wonders, too, whether “the allowance of taking a selfie also include use of the ubiquitous selfie stick”. (7)

And then, there are the issues that I have already discussed here ― whether the absence of evidence of ballot selfies’ harm shows that there is no reason for banning them or, on the contrary, demonstrates the effectiveness of the bans as a prophylactic measure. Judge Sutton clearly thinks that the latter is the case. Moreover, “[t]he links between [voter corruption and intimidation] and the prohibition on ballot exposure are not some historical accident; they are ‘common sense'”. (5, quoting US Supreme Court precedent.) Chief Judge Cole, dissenting, takes the contrary view, as have other American courts that have addressed selfie bans.

For own part, without expressing an opinion as to which of these views is correct as a matter of U.S. law, I have more sympathy for Judge Sutton’s. While I have been dwelling on the importance of evidence in constitutional adjudication for some time now, and critical of restricting rights on the basis of assumptions no later than yesterday, the evidence is actually there, albeit that it is mostly historical. Moreover, a court should be able to pronounce on the issue of delay without waiting for an “experiment” to take place. Common sense can be an unreliable guide to adjudication, but ― absent evidence to the contrary ― courts should be able to rely on it sometimes.

Prohibitions of ballot selfies might seem counter-intuitive or even quaint. In the United States, they run counter to the very strong tradition of virtually untrammelled freedom of expression. While I sometimes wish that Canadians took more inspiration from that tradition than they do (for example when it comes to the criminalization of “hate speech”), this is one instance where a more even-handed weighing of competing interests might be in order. Judges Sutton and Guy provide a useful reminder of what some of these interests are.

Show ‘Em

Earlier this week, an American court issued a decision on a topic that is all but certain to come up for discussion in the weeks after October 19: the ballot selfie, and the attempts ban it. Judge Barbadoro of the U.S. District Court in New Hampshire declared unconstitutional that state’s law that made it an offence to show one’s completed ballot to another person “with the intention of letting it be known how [one] is about to vote or … has voted,” including by means of “taking a digital image or photograph of [the] marked ballot and distributing or sharing the image via social media or by any other means.” Over at the Election Law Blog, Rick Hasen comments on the decision, hoping that it will be reversed on appeal.

For my part, as usual, I express no views on the propriety of this outcome under American law. However, because the issue has already come up in Canada (though without as yet resulting in a court judgment, so far as I can tell), and is very likely to come up again, I think it worthwhile to briefly summarize the court’s reasoning, and highlight a number of very interesting questions that it gives rise to.

Judge Barbadoro’s decision begins with a review of the history of the secret ballot ― or, as he calls it, the “Australian” ballot ― in the United States. As in the United Kingdom (about which I blogged here) and in Canada, elections prior to the introduction of the secret ballot were a brutal business. Corruption and intimidation, even violence, were commonplace. The use of the secret ballot, as well as legislation targeting such manipulation of the voters more directly, helped mostly (although not entirely, the judge says) eradicate it. Indeed, there is no evidence of voters being threatened or bought in New Hampshire recently, and the supporters of the ballot selfie prohibition did not cite any such cases, beyond one dubious anecdote, in defending the law.

Applying the analytical framework developed by the U.S. Supreme Court for freedom of expression cases, Judge Barbadoro then asks whether the prohibition is a “content-based” restriction on speech and, having concluded that it is, whether it is “narrowly tailored” to serve a “compelling government interest.” The judge concludes that the law meets neither of these conditions. While preventing corruption at the ballot box is a compelling interest in theory, the government must also “demonstrate” that it is invoking it to address an actual problem. Since there is no evidence of corruption or intimidation actually going on, the government has failed to do so:

even though small cameras capable of taking photographic images of ballots have been available for decades and cell phones equipped with digital cameras have been in use for nearly 15 years, the [State] has failed to identify a single instance anywhere in the United States in which a credible claim has been made that digital or photographic images of completed ballots have been used to facilitate vote buying or voter coercion. (32)

Judge Barbadoro adds that even if he had found the government’s interests compelling, he would still have held that the prohibition on ballot selfies is not “narrowly tailored.” Instead of banning people from sharing images of their ballots regardless of their motivations for doing so, it should focus on the use of such pictures to enable corruption and intimidation. In any case, “[t]he few who might be drawn into efforts to buy or coerce their votes are highly unlikely to broadcast their intentions via social media given the criminal nature of the schemes in which they have become involved.” (38)

The most important question judge Barbadoro’s conclusion elicits concerns the role of the courts in cases where they are dealing with prophylactic legislation, which aims not to address existing problems but to prevent problems from happening in the first place. By their nature, such laws are harder to justify by reference to evidence. And it stands to reason that that could open to the door to governmental abuse. Restricting constitutional rights “just in case” is a disturbing idea. Yet must the government wait for problems to arise before it can do anything about them? Especially when the problems at issue are not something inherently vague and uncertain, like the unfairness of the electoral process purportedly caused by the absence of campaign spending restrictions, but actual criminality, like bribery and intimidation. And all the more so when there is a history, albeit a somewhat remote history, of such problems actually happening.

The interpretation of this history is another big issue raised by Judge Barbadoro’s decision. What are we to make of the decline and virtual disappearance of voter bribery and intimidation in the wake of the adoption of the secret ballot and the enactment of the laws against such practices? How do we disentangle the effects of these laws from those of the secret ballot itself? Is the very success of these techniques a reason for letting ballot secrecy fall by the technological wayside? And then, there are questions about much more recent history ― specifically that of the ballot selfie and the innovations that enable it. Judge Barbadoro asserts that cellphone cameras have been around for 15 years, but how widespread were they then? The Oxford English Dictionary’s lexicographers have tracked down an example of the word “selfie” being used back in 2002, but the explosive spread of the phenomenon is surely more recent, and the idea of the “ballot selfie” might be more recent still. Is it, then, really the case that any problems that this phenomenon might generate ought to have become manifest, as Judge Barbadoro suggests? (Incidentally, while his opinion provides a wealth of citations to materials on the history of the ballot, this technological history is merely asserted.)

Finally, I think it is worth asking whether outright corruption or threats are the only reasons to worry about ballot selfies. In fairness to Judge Barbadoro, they were the only reasons asserted by the New Hampshire prohibition’s defenders. And the judge is probably right that people involved in corruption or subject to intimidation are unlikely to post their ballot selfies on social media. They’ll just show them to the persons paying or threatening them, and go undetected. But should we not also be concerned about the more diffused social pressure that can be brought to bear on people who let ― or those who fail to let ― others know how they voted? Should we not worry about people being pressured to vote, or to vote in a particular way, and to prove that they have done so, not by a specific manipulator, but by their online peers? People involved in “shaming” a person who didn’t vote to their satisfaction might not be committing an offence, and the line between legitimate and immoral behaviour in this realm is probably too blurry to lend itself to legal implementation. In this respect, the prophylactic prohibition on ballot selfies might actually be necessary.

As I said that in the post linked to above, I believe that the secret ballot “was one of the greatest inventions of a century that did not lack for them, and there is no reason to give it up.” And I am inclined to further believe that even coercive enforcement of ballot secrecy is justified. My hunch is that Canadian courts would agree, though I haven’t thought the matter through in detail. (I’ll try to do that before October 19.) In the state whose motto is “live free or die,” however, you’re now entitled not just to tell people how you voted, but to show them, too.

UPDATE: Prof. Hasen now has an op-ed for Reuters, further explaining his views. His most compelling argument, in my view, is the following:

the effectiveness of the selfie ban and the continued occasional prosecutions for vote buying, especially for absentee ballots, show that where there can be verification of how someone voted, this is a real — not theoretical — problem.

Keeping Secrets

I wrote, a while ago now, about the electoral practices of Georgian England, including the brazen, and fantastically expensive, corruption which elections involved. This weekend, the BBC published a fascinating story by Alasdair Gill, looking at a change in the electoral rules that happened during the Victorian age ― in 1872, to be precise ― and which helped put an end to the corruption, as well as the violence, which also used to be an inextricable part of a British election campaign: the introduction of the secret ballot.

We now take “the secrecy of the polling booth” for granted, as well as the fact that elections do not involve brawls, injury, or even death. But as Mr. Gill reminds us, “[u]ntil the Ballot Act was passed in 1872, voters had declared their candidate openly, leading to drunken fights, intimidation and bribery.” Each elector’s vote was recorded in a “poll book” ― and if he had not done as he had been told, those doing the telling would know:

[d]eciding who to vote for was previously a dilemma not just for you but for your boss at the local factory or your landlord, both of whom could see you out of your job or your house.

And for those who could not be threatened, there was always bribery of varying degrees of subtlety, including booze by the pint, quart, and gallon. For instance,

[i]n Gloucester during parliamentary elections in 1857 and 1859 the Tory party agent gave out food and drink to local supporters and lavished funds on Tory voters who acted as messengers, flag-bearers and bandsmen.

On other occasions, votes were simply bought, with supply and demand determining the price. On one occasion cited by Mr. Gill, a vote went for “about £14 – £1,200 in today’s money.”

And there was violence, ranging from “voters being kidnapped, plied with alcohol and then marched to the polls with a candidate’s name in their ear,” to “gangs” that “smashed windows around the venue of the hustings and waved poles topped with the severed heads of sheep,” to actual murder.

Incredible as it may seem now, there were quite a few people who liked things to be that way. According to a professor quoted by Mr. Gill, some thought “that it was only right for people to vote in an open fashion because that was the responsible thing to do.” Others simply “liked the parties and the bribes.” And, of course, many politicians were reluctant to let go an electoral system that had worked well for them, although it would have been interesting if Mr. Gill had discussed the effects of the Reform Acts of 1832 and, especially, of 1868 on the politicians’ views. Both of these statutes expanded the franchise, the latter quite dramatically, and the expenses on booze and thuggery which might have been sustainable with relatively few voters to bribe or intimidate might have become unsustainable as the country moved towards universal suffrage, which may well have helped overcome the resistance of some incumbent MPs.

* * *

In my post on 18th-century elections, I pointed out that it is sometimes “enough to enjoy the similarities and the differences between the past and the present for their own sake, as curiosities,” without attempting to draw lessons from them. Here, however,  there is actually a pretty obvious lesson to be learned.

It is that rules on ballot secrecy, including provisions, such as par. 164(2)(b) of the Canada Elections Act, which provides that “no elector shall … show his or her ballot, when marked, so as to allow the name of the candidate for whom the elector has voted to be known,” exist for a very good reason. When people are prosecuted for infringing such rules ― even when they are just proud and excited voters tweeting their pride and excitement for the world to see ― they are not the victims of a brutish Leviathan enforcing obsolete rules that serve no useful purpose. These rules protect us from the abuses that become possible when ballots are not secret ― abuses that are not the figments of some incumbent-protecting imagination, but a well-documented feature of our past.

Remarkably, Steven Fielding, another academic quoted by Mr. Gill, believes that the secret ballot is not that important anymore. “You might think,” muses, “that if people are voting secretly, then they are voting for their own selfish interests.” In his view, the secret ballot was introduced because “you couldn’t police [public voting] properly to protect people from intimidation, but you could do that now.”

I suspect, however, that Mr. Fielding is wrong both counts. He is certainly wrong to claim that the secret ballot causes people to vote selfishly ― simply because, as for example Jason Brennan recently pointed out over at Bleeding Heart Libertarians,

we have overwhelming empirical evidence, drawn from hundreds of studies, that they don’t vote their self-interest. Instead, they vote altruistically, for what they perceive to be in the national interest.

And Mr. Fielding is, it seems to me, much too optimistic to believe that we could police the intimidation that would result from our votes being made public. Intimidation can be subtle, and ― almost by definition ― victims will often be reluctant to complain. Add to this the pressures of this era of online shaming, and you have a recipe for abuses that will perhaps take a somewhat different form, but a no less disturbing magnitude, than those of the Victorian age.

“Maybe we don’t need the secret ballot any more,” wonders Mr. Fielding. Actually, we do. It was one of the greatest inventions of a century that did not lack for them, and there is no reason to give it up.

NOTE: The situation in Canada, before the secret ballot was introduced here in 1875, was no better than in the UK. I don’t discuss it here because I’ll probably do at least one separate post on the history of Canadian voting arrangements, based on an excellent account available at the Elections Canada website.