The Naked Voter

Election day in Upper Canada

Imagine my surprise when on reading Montesquieu’s Spirit of the Laws (1748), I discovered that the good Baron had some unkind words about the secret ballot. He thought it was a bad idea, in part because “the lesser people must be enlightened by the principal people and subdued by the gravity of certain eminent men” (Pt. I, Bk. II, ch. 2). As with so much in Montesquieu’s book, this passage seemed to me rather cryptic. All I could gather was that he believed the “better sort” ought to be kept apprised of what their social inferiors were thinking, so that they could correct the inevitable political errors of the lower classes.

I was only a little further illuminated upon chasing down Montesquieu’s footnote reference to Cicero, De Legibus, 3.15.34, where the great Roman laments that non latebra danda populo, in qua bonis ignorantibus, quod quisque sentiret, tabella vitiosum occultaret suffragium (“the people should not have been provided with a hiding-place, where they could conceal a mischievous vote by means of the ballot, and keep the better sort in ignorance of their real opinions”). Here again we find the idea that elites should be kept apprised about what the mob is up to. But there is also the added idea that the secret ballot acted as some kind of a nefarious cloaking device, like a bank robber’s mask or a ninja’s pyjamas, serving to disguise the mob’s evil plans. Whereas Montesquieu’s main concern seems to have been the ignorance of the mob, Cicero’s fear is of the mob’s vicious intentions.

The secret ballot is considered to be such a fundamental part of a citizen’s democratic rights in countries like the US and Canada that we neither stop to think it might possibly have drawbacks, nor that it is a relatively recent institution that was by no means an obvious good at the time it was instituted. Montesquieu’s passage spurred me to do some reading and thinking on the subject.

Some History.

In ancient Rome, the secret ballot was introduced into elections for magistrates by the lex Gabinia tabellaria in 139 BC, some 370 years after the founding of the Republic (in truth, there was voting in popular assemblies even in the earlier time of the kings). The sources indicate at least two reasons for its introduction.

First, it was thought that secret balloting would prevent the massive electoral bribery that had become a basic characteristic of Roman politics. If Titius was offered 10 denarii to vote for Seius, in a secret ballot Titius could pocket the 10 denariii and vote for Gaius instead, without Seius being the wiser. This would obviously make it pointless for Seius to offer such a bribe in the first place. Under a secret ballot regime, bribes would have no effect, and so none would be offered.

However, I suspect the introduction of secret balloting had less to do with bribery than with voter intimidation, especially given the historical timing. Electoral bribery never seemed to be a big concern among the Romans and was even considered acceptable within certain limits of tastefulness. On the other hand, the 130s BC was a time of increasingly ugly politics at Rome, with mob agitation for agrarian reform and land redistribution on the political agenda. This decade saw the rise of militant tribunes like Tiberius Gracchus and battles between rival gangs of hired political thugs (for example, Gracchus was himself beaten to death in one such battle). In much the same way as secret balloting made bribery pointless, it was also hoped that it would reduce some of the violence and intimidation associated with elections.

Although we may never know for sure, there may have been yet another reason Rome introduced the secret ballot, which becomes clear if we look at the example of Great Britain. The secret ballot was introduced in Great Britain by the Ballot Act of 1872. It seems to have been passed not to combat electoral violence and intimidation, nor to prevent bribery per se. Rather, it gained acceptance among legislators (who were obviously themselves politicians) largely because it would cut down the cost of campaigning. In other words, those who passed it might not have had reservations about the effects of bribery on the political system or the electorate, but they were rather more worried about how expensive it was getting to bribe people. You see, the franchise extended throughout the nineteenth century in Britain after successive reform acts. Each such extension meant an increase in the number of potential voters who would require bribing. At some point, politicians decided that enough was enough.

We can easily imagine a similar incentive working in the Roman case. As Rome grew, more citizens required bribing, so running for office was becoming increasingly expensive. The secret ballot offered a way out of a sort of bribery arms race: aspiring politicians didn’t necessarily want to hand out bribes, but as long as their opponents were doing it, they had to as well. The secret ballot, at least in theory, would have put a stop to this.

The secret ballot was introduced in Canada in 1874, and in the US in various stages between 1884 and 1892 (though West Virginia still allows people to cast open ballots if they so choose). Thus, it is not such an ancient institution, though it is more ancient than women’s suffrage.

My reading on the topic led me to a very enlightening paper by Geoffrey Brennan and Philip Pettit entitled “Unveiling the Vote” (British Journal of Political Science 20 (1990), 311-333). I cannot do full justice to their argument here, but I will give a rough outline of it.

The Preference Ideal.

Brennan and Pettit note a difference between two ideals of what voting should be: the preference ideal and the judgment ideal. According to the former, people vote based on their preferences, all things considered. These considerations can include anything, even considerations of self-interest. By contrast, in the judgment ideal, people vote based only on preferences between outcomes based only on considerations of public interest, or what their preference is between outcomes in light of their conception of the common good.

It may sound rather unusual to refer to the preference ideal as an ideal. After all, is having people voting based on narrow self-interest really something to which we should aspire? On the other hand, it is an appealing idea to economists and other utilitarian types, because they would say that it is the only way of coherently aggregating overall well-being through a democratic process. If the goal is the overall maximization of people’s satisfaction, then we should be adding up what will satisfy the greatest number of people, as measured by what people actually want. If people are assumed to vote for their preferred outcome or policy, based on their interests, and if the majority of voters (or at least the plurality, depending on the process in place) have their preferences implemented, then the majority have their preferences satisfied, and overall well-being is maximized. That, economists and utilitarians say, is the only conception of the common good that makes coherent sense.

Of course, this reasoning relies on at least two problematic assumptions. First, it assumes that people are reliable judges of what their interests really are. What I prefer at the moment of voting may be based on mistaken or incomplete information, cognitive bias, or other forms of irrationality. To this, the hard-nosed economist or utilitarian might reply that the way people vote is the most eloquent revealer of their preferences. We really have no other way of measuring “true” preferences (if there even is such a thing). Furthermore, although some people here and there may be mistaken about their preferences, it would be remarkable if a majority were. (This last is a very dubious line of argument, but I will leave it as it stands).

Second, it assumes that once the majority will is revealed, the government-elect will actually implement the policies that the people voted for. Unfortunately, there is little reason to believe that any government will do what it said it would do during an electoral campaign. This is a problem inherent in any democratic system. And on occasion it may not really be a problem as such, for a government elected on a certain platform may take office only to discover that the platform was not really feasible after all. Under such circumstances it would be foolish for the government to pursue what they now know to be bad policy, simply to satisfy a still-ignorant public. However, I would argue that they would then have a duty to educate that public about why it is a bad policy after all (which would also presumably mean admitting they were mistaken, not something governments are good at doing). It would certainly be immoral of them to cynically get themselves elected on the basis of a policy they know is bad but popular.

The Judgment Ideal.

In any case, those who, like myself, are sympathetic to civic republicanism will not be impressed with a democratic process that aspires only to satisfying the occurrent, self-interested, and possibly unreflective preferences that happen to be held by the greatest number of people on voting day. We would instead prefer something like the judgment ideal, where citizens vote on the basis of opinions formed in the course of public deliberation on what is in the interest of the common good.

The difference between the preference and judgment ideals can be seen in light of their differing attitudes towards the issue of low voter turnout. For adherents of the preference ideal, low voter turnout is a problem because voting will not adequately track the preferences of the citizenry if too few people are showing up to register their preferences. Only the preferences of those who vote can be aggregated, and if these are too few or too unrepresentative, the resulting outcome may not be optimal from a utilitarian viewpoint.

Low voter turnout is less problematic for adherents of the judgment ideal. Insofar as those not showing up to vote are also those who are less well-informed and less public-spirited, such low quality voters may only add noise to the process and pollute or distort the general will. Such citizens should be encouraged to become more responsible and informed. But if they remain unresponsive to such encouragement, then perhaps they ought to be actively discouraged from voting. In such cases, low voter turnout is something rather to be celebrated than lamented. Low voter turnout is only a problem if large numbers of high quality voters are not voting, as this may be a sign of democratic system failure. My impression is that both of these phenomena ― the apathy of the ignorant and the frustration of the well-informed ― are at work in low voter turnout in Canada, a nation with probably the most archaic and least representative electoral system of any Western democracy.

(I should warn readers that the argument in the preceding paragraph is not to be found in Brennan’s and Pettit’s paper, but rather represents my own editorializing. I am, however, gratified to learn that a forthcoming book by a philosopher at Brown University argues a similar point. The book I refer to is Jason Brennan’s The Ethics of Voting, to be published by Princeton University Press. To my knowledge, he is not related to the Geoffrey Brennan aforementioned.)

As nice as it sounds in theory, the judgment ideal relies on some premises that are not self-evident. For one thing, it would only be ideal if it could be taken for granted that voters’ judgments actually track public interest (where “public interest” means something more than the largest aggregation of voter preferences.) There are a couple of problems with this assumption.

First, it assumes that there is such a thing as the “public interest” that can be clearly demarcated from mere private interest. What if “public interest” merely means that which is in the private interest of the greatest number of people? If this were the case, then we would arrive back at the preference ideal. What if there is no such thing as an interest belonging to everyone? Economists, with some justification, might say that the judgment ideal’s notion of “public interest” seems a little mysterious.

Second, let us grant that there is such a thing as a truly public interest, based on disinterested and objective considerations (these being themselves problematic and slippery concepts). What if honest deliberation on what is in the public interest points to multiple and divergent policy recommendations? Perhaps this will happen because deliberations are based on incompatible fundamental values (not mere preferences based on desires) held by different segments of the population. If a society is not homogenous with respect to its fundamental values, then it is unlikely to be homogenous with respect to what it believes is in the public interest. And again, it takes some pretty deep metaphysical examination to get a handle on what exactly separates desire-based preferences from values in such a way that the latter deserve priority. Most economists are skeptical that there is any such fundamental difference.

Third, it assumes that people are reliable judges of what is in the public interest. Empirically speaking, this is not the most warranted of assumptions (indeed, people are often not even the best judges of what is in their private interest, let alone the public interest). This can be partly due to simple lack of information on the part of the voter. But there are often various biases at work too. For instance, there is time inconsistency and hyperbolic discounting: we tend to discount benefits and losses more deeply the further they are projected into the future, which leads us into making bad spending decisions. Also, there is a tendency to rationalize one’s own private preference and elevate it to the status of a “public interest”. For example, for those who work for General Motors, there is a tendency to believe that what is good for GM is good for America (or Canada, as the case may be). Most people would say that the majority of corporate lobbyists advocate for private rather than truly public interests. The world would be a much simpler place if such lobbyists could be written off as selfish and corrupt monsters. Unfortunately, my personal experience of lobbyists has led me to realize that the majority of them probably believe (or have come to believe) that their lobbying is in the public interest, even when in reality it is to the public detriment.

However, there is this to be said in favour of the judgment ideal over the preference ideal: when I enter the voting booth and cast my ballot for a preferred outcome, should my vote be decisive, my preferred outcome will be forced on everyone, including those who voted against it. In this sense, bad voting is like an economic externality; gratification of my preferences may impose unwanted costs on others. Thus, I have an ethical duty to at least keep the well-being of others in mind as a consideration in casting my vote. I do not have a right to try to impose my selfishness on others through the use of my vote. My self-interest should be pursued with a due consciousness of the rights of others. Perhaps this can give us some minimal conception of what it means for there to be such a thing as a distinctly common good which the vote is supposed to forward. Since the preference ideal does not have room for even such a minimal conception, it should be rejected.

The Secret Ballot as an Instrument of Corruption.

The general drift of the argument put forward by civic republicans like Brennan and Pettit is that it is possible, through greater openness, dialogue, and political discourse, along with being required to justify one’s voting decisions publicly, for people’s judgments to be shifted more in line with what is in the public interest. This objective could be served by eliminating the secret ballot.

The secret ballot reduces the pressure on citizens to vote in a responsible manner. They can engage in the public discourse or not, but when they finally vote they are answerable to no one. Thus, they can vote according to their private interest rather than for the public good, and nobody will be in a position to call them on it, or to hold them to account for their irresponsible citizenship. If, on the other hand, they were required to openly announce who they were voting for before an assembly of fellow citizens, they might give their vote more consideration, or at least think about how they might make it pass the test of public scrutiny.

There is an irony in the institution of the secret ballot. Remember that secret balloting was intended to make voters unaccountable to those who would try to bribe or intimidate them, thereby reducing the incentive of the latter to offer bribes or intimidation. It achieved this by making the voter unanswerable to bribers or intimidators. The problem, however, is that this also had the effect of making voters unanswerable to their fellow citizens and ― if it’s not improper to speak of such a thing ― to the public conscience.

In reality, I’m not even sure secret balloting was successful in preventing electoral bribery. I would argue that instead it has forced bribery to take on a different form. Instead of saying to the voter, “Here is a hundred dollars; vote fore me,” the candidate for office now says, “If you don’t vote for me, this policy that is in your interest won’t be implemented.” In short, voters are now bribed with policies rather than cold cash. This ensures that the voter receives his bribe after he has voted in the required way rather than beforehand. And since the bribe no longer comes directly out of the politician’s own pocket, the voter is in effect being bribed with his own money. Even worse, voter corruption is now hidden behind the veil of the secret ballot.

This form of public corruption has probably been aided and abetted by the preference ideal of voting spoken of earlier: economists and their ilk assume that voting is not supposed to be anything more than a procedure to discover what the greatest number of people believe to be in their own self-interest. Voting one’s self-interest is considered the norm rather than a sign of terminal corruption. If it is commonly believed that people are supposed to vote for their own interests, then voters do not even realize there is anything wrong with what they do.

It turns out Montesquieu’s and Cicero’s misgivings about the secret ballot may not have been so misplaced after all. Maybe secret balloting truly is little more than a convenient way for the citizen to mask his viciousness ― not necessarily from his betters, but from his fellow citizens more generally.

Stripping the Voter Bare.

If you are like me, there is probably something in you which rebels at the thought of eliminating the secret ballot. After all, for generations now we have been taught that this innovation was a hard-fought victory for democracy. So it is worth asking: What would happen if the secret ballot were eliminated? Would there be a reversion to massive electoral bribery or intimidation, as in the bad old days?

It could be argued that the conditions that once made the secret ballad necessary or desirable no longer exist. For example, in republican Rome, there was no publicly-funded police force in the modern sense that could maintain order at the polls, and there was no public prosecutor either. Thus, Rome didn’t have the sort of checks on voter intimidation that we do today.

Also, we in many of the Western democracies live in what can be called “extended republics”, extended in both territory and population. In earlier times, in communities where citizens knew and depended directly on each other, there could be serious consequences to voting one’s conscience. It was easier to intimidate voters at the polls, and in places like Upper Canada, political parties stationed thugs at the polls to ensure that people voted the way they were supposed to. It was feasible for them to do this, since in the absence of universal suffrage there were fewer voters to intimidate. Today, with our much larger agglomerations of people, the thugs are vastly outnumbered, and there aren’t enough of them able to be present in sufficient numbers at each polling station. There are now too many voters to effectively bribe or intimidate.

Even in modern despotic regimes that hold periodic “elections” (as most of them in fact do), direct bribery and intimidation are no longer the preferred methods of control. It’s much easier (and cheaper) to simply stuff the ballot boxes after the fact, or report falsified election results, which is why such regimes have no problem with allowing a secret ballot, since the voting means so little anyway.

If the secret ballot is compatible with despotism, then maybe it is not the cornerstone of democracy we assume it to be.