Secret Ballots

by Harry on May 15, 2006

This is something that came up as an aside in a dissertation defence the other day (congratulations, A!). The dissertation was about privacy, and a brief comment was made that secret ballots might protect a voter’s right to privacy. I was surprised that I already had a half-thought-out but very strong dissent from this idea, so I thought I’d articulate it here and see what you think. There are some practical arguments in favour of having secret ballots in representative democratic elections for governmental positions; most obviously the argument that secret ballots obscure the information needed to perfect a market in votes; so that the vote remains effectively inalienable. But is there a right to vote secretly?; that is, if other measures could effectively prevent the emergence of a market in votes, or government retaliation against individual voters, would voters have a complaint if ballots were public? The privacy thought it that the interest in privacy, or something like that, protects this information; it protects us from others having the information about how we voted, just as it protects us from others having information about various other details of our thoughts and personal life.

I don’t think so.

The pragmatic reasons count so strongly in favour of secret ballots that the reason for making them public would have to be very powerful indeed to override them. But I think there is a reason, albeit not a strong enough one. When we vote in governmental elections, we exert (or attempt to exert) coercive power over other people. A standard view about what is required for coercion to be legitimate is that it must be justifiable to the person who is coerced; the coercer must have reasons which she can offer to the coerced person and which he can, if he is reasonable, at least understand and see that they have some force. These reasons might concern the benefits that the act of coercion will produce for both parties or for third parties or, if the coercer is sufficiently worse off than the coerced, for the coercer.

But there seems something odd in thinking that it is legitimate for me to be able to force someone to do something only if I can offer her reasons that she would be able to understand and evaluate if she were reasonable, but also to hold that I have a right to refrain from even letting her know that I am forcing her (or actively authorising others to force her). I can see that when the balance of forces are such that the coerced can (and might well) retaliate against the coercer, because she is unreasonable or because, though reasonable, she is simply self-interested, the coercer has a claim to remain anonymous, but this just seems like one of the practical reasons for keeping the information secret. In well-ordered society, not only would there be no need for secrecy, but wouldn’t secrecy seem sneaky, cowardly, and an instance of bad citizenship?

Full dislosure: I voted Labour in 83 and 87, couldn’t figure out how to vote absentee in 92 or 97 or 05, and voted and campaigned for a LibDem in ’01. Almost always voted Labour or left of Labour in local elections. Can’t vote in the US, where I have campaigned only for leftist third parties, but I did not prevent a Kerry sign being placed in my front yard in 2004, even though it made me feel physically ill every time I looked at it.

{ 1 trackback }

Ascription is an Anathema to any Enthusiasm » Blog Archive » Inalienable
05.16.06 at 9:03 am



Robin 05.15.06 at 11:32 am

There was a long debate in the 19th century as a result of Mill’s thoughts on the issue. From Mill’s “Considerations on Representative Government”:

In any political election, even by universal suffrage (and still more obviously in the case of a restricted suffrage), the voter is under an absolute moral obligation to consider the interest of the public, not his private advantage, and give his vote, to the best of his judgment, exactly as he would be bound to do if he were the sole voter, and the election depended upon him alone. This being admitted, it is at least a prima facie consequence that the duty of voting, like any other public duty, should be performed under the eye and criticism of the public; every one of whom has not only an interest in its performance, but a good title to consider himself wronged if it is performed otherwise than honestly and carefully. Undoubtedly neither this nor any other maxim of political morality is absolutely inviolable; it may be overruled by still more cogent considerations.

I think that Geoff Brennan and Philip Petit recently also made an argument along these lines–publicity as a way of forcing justifications for political preferences, at least forcing deliberation–for an open vote.


Tom Hurka 05.15.06 at 11:43 am

Interesting post and interesting topic. But isn’t the tendency of the “justifiability to the coerced” argument to favour making public voting mandatory, i.e. a legal requirement? And wouldn’t that, quite apart from any bad effects re the possibility of buying votes, make voter turnout much lower?

I suspect the reason has something to do with a kind of privacy. Many people, though obviously not those who put political signs on their lawns, want their relations with those they encounter but aren’t close to, such as many of their neighbours, not to be coloured by politics, so their different political attachments can’t affect or disrupt those relations. And this desire is important enough to them that having to make their political views public when voting would dissuade them from voting.

May there not be some legitimate interest here, i.e. in having social relationships which won’t inevitably be infected by political attachments because those attachments can be kept private? And isn’t that part of what’s at stake in privacy about religion, i.e. you can want your religion not to be publicly known because you want your relations with some people to be independent, and safely independent, of all religious issues? It’s not privacy in the most straightforward sense, and I don’t know how much weight I’d want to put on it, but it does seem to me a kind of legitimate interest, i.e. an interest in the possibilty of social relations that are insulated from highly contentious and divisive issues.


Chris Bertram 05.15.06 at 11:53 am

Robin is right about Brennan and Pettit, the reference is

Brennan, G. and Pettit, P. (1990), Unveiling the Vote, British Journal of Political Science, 20, 311-333.


KCinDC 05.15.06 at 12:02 pm

… give his vote, to the best of his judgment, exactly as he would be bound to do if he were the sole voter, and the election depended upon him alone.

So I’m obliged to vote for the one person who I think would best represent the public interest — a person who is highly unlikely to be one of those who appear on the ballot and so almost certainly must be a write-in? Fortunately that can be overruled by the “more cogent consideration” that other voters do actually exist.


Peter Levine 05.15.06 at 12:03 pm

In the US, we’re ignoring the argument that a secret ballot is necessary to prevent a market for votes. Almost half the states allow people to vote by mail (“absentee”) without giving any reason. Oregon runs its elections entirely by mail. If you can fill out a ballot in your home, then you can sell your vote. However, it’s also possible to conceal your vote, and that can protect you against retaliation–not only by the government, but also by a spouse, an employer, or a fellow member of your congregation. Even granting the Mill/Brighouse argument that voting affects others and therefore should be made public, the risk of retaliation and coercion seems fairly intractable.


y 05.15.06 at 12:07 pm

It’s not just a question of selling votes, or of government coercion. It’s a question of getting fired from your job, or kicked out of college, or having your husband beat you up, or your customers boycott you, because you voted for a Democrat. Didn’t you fight all of this out in England in the 19th century, when voters might well have been clients or tenants of one candidate, or of his sponsor?

Start living in the real world. The power of an individual voter is very small, and the power others have over that voter is very large. How can you think to change that, even in principle, short of eliminating every form of power that one person might have over another?–in which case, there would be no voting, either.


KCinDC 05.15.06 at 12:25 pm

Peter Levine, the coercers can force you to vote by mail, so I don’t think there’s a difference between the issues of bribery and coercion there. I am worried about the increasing use of vote-by-mail, and I imagine we’ll have an election scandal resulting from it before long.


saurabh 05.15.06 at 12:29 pm

A standard view about what is required for coercion to be legitimate is that it must be justifiable to the person who is coerced…

This is the Rousseaun social contract, I take it. I’m not sure why that legitimacy requires accountability from each individual voter, rather than merely saying that entering into the bargain of democracy implies acceptance of coercion. If votes (and voters) are equal, how does it matter whether a majority is composed of a panel of hicks and sodomites, as opposed to Ivy League professors? Is that information really germane to the functioning of a democracy? And can it actually outweigh the real dangers of voting markets or, as y points out, retaliation against voters?


harry b 05.15.06 at 12:31 pm

I thought Peter was saying that voting by mail involved greater risks to secrecy than when you have to go to the poll booth. Britain has already had a sort of scandal with mail in ballots; there are going to be lots of them, I imagine.


paul 05.15.06 at 12:41 pm

I’m definitely unlettered here, but can someone explain how the right to privacy doesn’t become a mere “practical consideration” in almost all cases, given this argument? As a member of the community whose resources you have a call on, and whose stability you might affect, I can make a pretty good argument that I have an interest in your medical records, your financial affairs, probably your sexual behavior, and (given the current NSA kerfuffle) whom you have the temerity to call on the phone. There are obvious practical reasons why I shouldn’t have access to any of that information, but if all possibility of retaliation or sanction for things I found in those records were wiped away, would your right to privacy be gone as well?


abb1 05.15.06 at 12:42 pm

You, liberal pussies. The only right way to vote is by raising your sword on the central square; anything else is an abomination.


Richard Bellamy 05.15.06 at 12:56 pm

I was in the unusual circumstance of taking my 4 year old daughter into a voting booth last month for a School Board election.

There were 4 people running for 3 seats, so I read her two names and told her that if she chose 2 names, I would promise to vote for at least one of them.

I read her the names, and my little girl — bless her heart — then screamed at the top of her lungs: “I WANT TO VOTE FOR KAREN SMITH!!!”

And she was immediately shushed by everyone outside of the voting booth.

This led me to the inexorable conclusion that the Secret Ballot has involved into YOUR right to not know who I voted for.


loren king 05.15.06 at 12:56 pm

Another pragmatic consideration that weighs in favour of private ballots: public balloting could create perverse informational and reputational cascades. But that doesn’t much speak to the question of whether we have a right to privacy at the ballot box.


djw 05.15.06 at 1:13 pm

I’m inclined to agree with Paul; the vast majority (although perhaps not all) of the goods we group under the heading of privacy rights are, in the final analysis, attempts to best negotiate the freedoms of different members of society. I’m fine with extending the logic of civil rights in employment, rentals, etc. to a host of other categories, but it can be quite intrusive to enforce, as we know. Privacy about voting provides a less intrusive way of providing the positive outcome of non-discrimination based on voting records, with less substantive incursions into citizen’s freedom. Same with sexual history, religious affiliation, etc, etc.


djw 05.15.06 at 1:17 pm

That said, as a deliberative democracy-sympathizer, I think a norm of privacy for it’s own sake about voting isn’t a good thing at all, and people who don’t have reason to expect to incur real costs through disclosure should be encouraged to do so, complete with reasons. But this is for enhancing the quality of democratic discourse reasons, and not concerns about coercion. I’m pretty sure I think your line of thinking about coercion is wrong-headed, but I’ll have to think about precisely why a bit more.


Jason Kuznicki 05.15.06 at 1:22 pm

“When we vote… we exert… coercive power over other people. A standard view about what is required for coercion to be legitimate is that it must be justifiable to the person who is coerced; the coercer must have reasons which she can offer to the coerced person…”

I’m inclined to think that it’s the responsibility of our elected representatives both to do the coercing and to explain why they are doing it. Explaining the reasons behind an act of coercion is a nearly impossible task when, as a voter, the communicative symbols at your disposal are merely a set of names.


abb1 05.15.06 at 1:24 pm

What about those more nuanced voting systems – plurality voting, Yes/No/Maybe – in this case coercion is less obvious, it seems more like building consensus than coercing, no?


DC 05.15.06 at 1:34 pm

There might be a comparable case for transparency regarding personal wealth and income. Money is an inherently social, not private, thing, which entails power over others.

Is there any good reason why people should have the right not to disclose their financial resources?

The best I can think of is that the poor might fear stigma. So maybe it should be cumpulsory above a certain middling level of income?


RickD 05.15.06 at 1:36 pm

The right to a secret ballot, as with all privacy rights, and indeed all rights, has come into existence as a result of the evolution of the form of government, and not merely as a thought exercise on the parts of the people conducting elections throughout history. Indeed, it’s widely believed in many fields that anonymity allows for a greater degree of candor than the public assignment of every choice a person makes to the person making that choice.

From that standpoint, I would argue the right to privacy is more fundamental than any obligation owed to the fellow members of a state. To approach things from the “social contract” standpoint, I am considerably more amenable to a social contract that allows me to make choices while retaining anonymity than to be required to go on public record for every vote I make.

As other commenters have noted, there can be tremendous public influences on the vote, once it’s required to be public. The next logical step after a public vote is the creation of binding loyalty oaths. It is easy to imagine social organizations whose membership is contigent upon perfect voting records. This kind of change to the governmental structure would lock in an undesirable degree of inertia into a democratic system that is better of if change need not be so painfully won.

As for the comments about votes being bought: not only could votes be bought (which can happen already with promises of tax cuts, etc.), votes could be coerced or extorted.


marcel 05.15.06 at 1:41 pm

1) There was a story during the 2004 presidential campaign about a woman in the South, I think in Alabama, losing her job for having a Kerry sticker on her car. Not protected by a union, tenure or civil service, she could be fired w/o cause, and was. Her boss did not want to see a Kerry bumper sticker in the company lot.

2) What about posting a comment anonymously or under a pseudonym to blog, say CT. Anyone think my name is actually Marcel? Raise your hand. I have a bridge in Brooklyn that I might be able to interest you in.


M. Townes 05.15.06 at 2:50 pm

Claims for a “right” to the secret ballot most often derive from the right to vote, as the former is believed necessary to the latter. When the secret ballot was introduced in the mid-19th century, it was not viewed as a right, but as a mechanism against coercion (in the U.K.) and bribery (in the U.S.). Arguments for its adoption focused on the instrumental value of such a mechanism. Arguments against – and here I am thinking of Mill – focused on the function of the vote and the importance of open discourse to democratic society. Mill especially feared that the secret ballot would lead voters to think of their vote as a personal, selfish thing, rather than a means to common good. A lot of things have changed since then, and I think his argument has more traction now than it did in 1868. For those interested, I would greatly appreciate feedback on my paper on this very topic; it’s available at


M. Townes 05.15.06 at 2:51 pm


Colin Danby 05.15.06 at 2:53 pm

The anti-retaliation case for voting secrecy strikes me as overwhelming, but there’s still a good argument that secrecy undermines responsiblity to others. I might be less inclined to vote selfishly, let’s say against school bonds, if I knew that my neighbor with kids would know. Sartre wrote that “The polling booth standing in the lobby of a school or town hall is the symbol of all the acts of betrayal that the individual may commit against the group he belongs to.”

This is related to the case Harry builds above by extending the scope of “coercion,” but given that a certain level of coercion seems unavoidable in any sort of state power, it may be better to disaggregate terms.

Sartre, “Elections: A Trap for Fools” Life/Situations Pantheon Books, 1977. Here’s an (imperfectly scanned) e-version:


gmoke 05.15.06 at 3:35 pm

Some interesting thoughts on this subject might arise from reading Jose Saramago’s _Seeing_, about an election in which over 80% of the ballots are left blank by the voters. Halfway through it and it is both stylistically and literarilly ingenious.


Dan Kervick 05.15.06 at 3:53 pm

I’m inclined to agree with Y here. Presuamably a democracy is a society in which – at the least – all of the society’s members are equally empowered to participate in the work of government, and to influence public decisions. Differences in individual power are already a deviation from democracy. But what kind of democracy can a society expect to have if the already powerful are further enpowered to dominate the weak by punishing them for their votes? Not even an imperfect one. One might say that the society can rely on the law to prevent these sorts of punishments, i.e. laws against job and housing discrimination, physical coercion, etc. But a society’s laws are the products of those who have the power to make them. An aspiring democracy constantly needs to take measures to protect itself from the tireless efforts of the few to draw power away from the many. One of these measures is for the majority to use their fragile democratic power to preserve the secrecy of their own balloting and shield themselves from the encroachments of anti-democratic usurpers.

I propose that in understanding the powerful reasons for secret ballots we not think so much in terms of whether a person has a right to privacy in voting, but whether a society that aspires to be democratic has an interest in maintaining privacy in voting.

As for this:

A standard view about what is required for coercion to be legitimate is that it must be justifiable to the person who is coerced; the coercer must have reasons which she can offer to the coerced person and which he can, if he is reasonable, at least understand and see that they have some force.

The view might be standard, but it strikes me as both a weakly motivated and relatively empty principle, the latter because almost any motivation for an action one can think of is such that a reasonable person would “see that it has some force.” In any case, there is no clear connection between this principle and public balloting. Harry says:

But there seems something odd in thinking that it is legitimate for me to be able to force someone to do something only if I can offer her reasons that she would be able to understand and evaluate if she were reasonable, but also to hold that I have a right to refrain from even letting her know that I am forcing her (or actively authorising others to force her).

This doesn’t seem so odd to me. The justification for the proposed standard for legitimate coercion, insofar as I can grok it, seems to have something to do with grounding the highest principles of practical rationality in he ideal standards of rationality prevailing in certain ideal, possible worlds. But the right to protect the secrecy of one’s ballot can be seen as an ordinary, practical right grounded in the conditions that prevail in worlds very close to the actual world in which we act, and the realization that in such worlds people are not ideally rational, and the means of coercion must be suited for the practical ends toward which that coercion is designed.

The passage from Mill that Robin cites strikes me as doubly confused. For one thing, the supposed absolute moral obligation he descries is incompatible with his utilitarianism. For another, contra his assertion, he fails to make even a prima facie case that the existence of such an absolute individual obligation argues in favor of public voting.

Other problems with Mill’s approach, it seems to me, lie in the conception of absolute obligation generally, and also the utilitarian account of absolute obligation specifically. But this isn’t the place for those criticisms.


Alex 05.15.06 at 4:57 pm

Surely the argument against an open ballot is that the balance of power between the individual voter and the government is immensely unfair? If it was publicly known that the ballot was open, it would only need a very small degree of real repression to exercise a deterrent effect on anyone who wanted to be a real opposition candidate, just because of the risk they might lose.

The ability to lose and survive is a key criterion of democracy – any opposition has some chance of failure, so a guarantee of nonrevenge is required to have an opposition, and an opposition is required to have a nontyrannous government.

It’s just a special case of the general principle that majority rule is constrained by the constitutional right to opposition, or in other words the obligation to respect the rights of the minority.


Tracy W 05.15.06 at 5:32 pm

A standard view about what is required for coercion to be legitimate is that it must be justifiable to the person who is coerced; the coercer must have reasons which she can offer to the coerced person and which he can, if he is reasonable, at least understand and see that they have some force.

How does revealing your vote meet this requirement? I know some people who have voted in ways I think it is absolutely nuts. The mere fact of telling someone who you voted for does not meet this “obligation”. Knowing that more than 5% of the population voted for Winston Peters in the last election does absolutely nothing to convince me of the reasonableness of electing Winston Peters. If this standard view requires justification of voting, it requires more than just publication of voting, it requires also an explanation of why you voted for whoever you did.


Sean Aas 05.15.06 at 7:28 pm

Interesting post. One line of response might take off from the notion of *what* needs to be justified when coercive power is exercised by the casting of a vote. Is it the act itself, or (so to speak) the maxim licensing it?

In particular, couldn’t we say that making public one’s general political orientation (precisely as you did, Harry), is sufficient to meet the additional justificatory demands on coercive acts? This kind of response seems to fit well with the Kantian/Rawlsian ideas that support the requirment that such acts be justified to those coerced.


digamma 05.15.06 at 7:48 pm

I did not prevent a Kerry sign being placed in my front yard in 2004, even though it made me feel physically ill every time I looked at it.

Uh oh. I think Atrios has found his Wanker of the Day.


harry b 05.15.06 at 7:59 pm

What is that supposed to mean?


harry b 05.15.06 at 8:09 pm

It didn;t occur to me, but should have, that there might be (fundamental) privacy interests even though they are outweighed, or at least matched, by obligations to disclose (which is what you are suggesting, right, tom?). Some people here are invoking secrecy as a kind of contribution to a public good, which seems right to me, but not a voter-centered consideration; rather another non-right-involving reason for secrecy.

y and dan; I need to think more about what it is we are doign when we are doing ideal theory; its never clear to me, but its too hard for me to think through in this context right now.

tracy — no, of course, saying how you voted falls far short of offering a justification. But doesn’t it seem odd to say I owe someone a justification of how I voted, but not the facts about how I voted? Necessary, not sufficient. Hence my self-revelation that digamma either pours scorn on or thinks atrios would pour scorn on; I’m implying that I might well have voted that way if I were allowed to vote, but offering a little further information that goes beyond that, in our electoral system, not very informative revelation.


Belle Waring 05.15.06 at 8:14 pm

re: 20, is it simply inconceiveable that someone might be named marcel? it is surely among the top 100 names for french boys. and some people post under their own names! yours, celestine.


Jim Johnson 05.15.06 at 9:45 pm

Mill was simply inconsistent in advocating for the pubic ballot while alos being worried about the threat of social factors on liberty. In my view he is correct on liberty wrong on public voting – as too are Brennan and Pettit!

Tom Schelling rightly points out that the secret ballot prevents me from proving how I voted and so eliminates your ability to try to buy my vote. (Hence Harry is correct to see it as a way of undermining markets for votes – which would require something that much more closely approximates symetrical informaiton.)

That means, I think that the case for the secret ballot is grounded in eaulity not privacy. So, considering social choice views on voting, the secret ballot is the real world embodiment of anaoymity or (more weakly) non-dictatorship. It is among the equality conditions that makes the casting and counting of votes into an election.


abb1 05.16.06 at 4:27 am

Seems to me that the concern about ‘markets for votes’ is highly speculative (in the contemporary US context, at least); intimidation is much more probable.

However, the same problem exist at the collective level: vote for our party candidate because our party controls the government and thus our party candidate will be able to bring more dough to your district or, if you elect the opposite party candidate – risk the retaliation, forget about that new highway/hospital/keeping that military base.

Secret ballot doesn’t help here, it’s all legal; I understand this is exactly how Tom Daschle lost N.Dakota in 2004.

Thus (if we try a li’le ‘reductio ad absurdum’ here) to follow the logic of ‘secret ballot’, the election results in the districts and states need to be kept secret as well, elected representatives need to be anonymous. Quite ‘absurdum’, ain’t it?


Rob 05.16.06 at 4:48 am

An interesting case for comparison, although obviously not anything like totally analogous, might be secrecy for the votes of legislators. The argument for secrecy for votes in elections rests in part on the various and undoubted associational costs of publicity for such votes, yet there are clear associational costs of publicity for the votes of legislators, most obviously related to party discipline. Clearly the idea of secret votes in a legislature would be ridiculous, because it would prevent constituents from holding their representatives to any kind of account. Holding legislators to account matters because they exercise significant power over citizens as legislators: by making law, they coerce citizens. Equally, voting is the exercise of a significant power: it aims and in aggregate is successful at the granting of the power to legislate to specific individuals with specific records. It does not seem prima facie unreasonable that if legislators ought to be in a position such that they can be held to account for their acts as legislators, then those who aim at placing them in that position should be as well. The analogy isn’t perfect, obviously – the associational costs of electors votes are probably higher than those of legislators, and the burden they individually impose on others is undoubtedly lower – but it doesn’t seem totally wide of the mark.


lurker 05.16.06 at 4:59 am

I just can’t believe this! 33 comments and 12 hrs. on a topic titled ‘Secret Ballots’ and not a single Soviet Russia joke. What has academia come to?

In Soviet Russia you are given a sealed envelope to drop in the ballot box.


chris y 05.16.06 at 5:51 am

What about posting a comment anonymously or under a pseudonym to blog, say CT. Anyone think my name is actually Marcel? Raise your hand. I have a bridge in Brooklyn that I might be able to interest you in.

What Dr Waring said at 32. I am forced to conclude either that Marcel lives in China, Egypt or Iran, or that he is studying with Ivan Tribble. Otherwise this comment is barking mad. All the principals at CT post under their own name. The overwhelming majority of commenters user their own names (often informal) or initials. Whatever does he think will happen to him if he comments openly at CT?

If I vote publicly, it is conceivable that my boss will refuse me promotion, that the fascist nutter down the street will put a brick through my window or even that I will have a blazing row with my wife. In some countries the risks are a lot more serious and immediate. But commenting on a blog? Perleeze. Does Marcel write to the papers under an assumed name?

Christopher Young
123 Newbould Lane
S10 2PL


Cranky Observer 05.16.06 at 7:16 am

> I might be less inclined to vote
> selfishly, let’s say against school
> bonds, if I knew that my neighbor
> with kids would know.

Someone better at game theory than me might want to try to connect this to the concept of the 2nd price auction.



Mike Otsuka 05.16.06 at 5:07 pm

So Harry, is this an accurate summary of your position?: Your claim is that in a well-ordered society, we’d each be obliged, all things considered, to publicize our reasons for voting the way we voted, and it follows from this obligation that we’d each be obliged to reveal how we voted.

Am I right in thinking that you haven’t provided an rationale for revealing how one voted in the absence of a publicizing of one’s reason for voting the way one has?

Another question: What if detailed exit polling provided an extremely accurate picture of why people voted the way they did? Suppose that the exit polling is so accurate that one wouldn’t learn any more from having access to each person’s justification except who voted which way. So the information from the exit polling wouldn’t reveal how your next-door-neighbor voted, though it would tell you which proportion of the population voted for what reason. Have we any right, in a well-ordered society, to know, not just which proportion of the population voted which way for what reason, but also how our next-door-neighbor and any other particular individual voted? I doubt it.


harry b 05.16.06 at 5:49 pm

mike: answer to question 1 is yes, you’re right. Any reason I have given for publicity hinges on the obligation to offer justification.

My answer to 2: I need to think about it. My initial reaction is that there still is a reason, but I can’t articulate it exactly. I’ll think some more.


Mike Otsuka 05.17.06 at 1:44 am

A standard view about what is required for coercion to be legitimate is that it must be justifiable to the person who is coerced; the coercer must have reasons which she can offer to the coerced person and which he can, if he is reasonable, at least understand and see that they have some force.

Even if detailed exit polling doesn’t satisfy this requirement, it appears that the following would:

Each person in our well-ordered society attaches an unsigned, anonymous statement to his ballot explaining why he voted the way he did. (Since we’re assuming a well-ordered society, we assume that these statements are not dishonest.) All such statements, and the ballots to which they’re attached, are made available for public scrutiny.

Here, (i) everyone would know why each vote was cast the way it was cast, but (ii) nobody would know who cast which vote (i.e., ballots would be secret).

So I don’t think Harry’s requirement rules out secret ballots.


abb1 05.17.06 at 4:29 am

In 1972 Nixon won 60% of the vote and a few years later no one would admit voting for him. Isn’t there some kind of ‘personal responsibility’ argument, as an incentive to take voting seriously? That’s assuming a well-ordered society, of course.


ajay 05.17.06 at 5:59 am

Are you really attempting to coerce when you vote in a governmental election? Don’t forget that an MP’s a representative, not a delegate. You aren’t voting (as in a referendum) for a single initiative, for or against – that’s definitely coercive; you’re voting for a person who you think will be the best to represent your area in parliament. It’s a big hire/fire decision – the only people being coerced are the losing candidates, who don’t get the job. The MP isn’t expected to convey your instructions to parliament, he’s meant to do what he thinks is best.


abb1 05.17.06 at 6:12 am

Nah, in effect you vote for a party with a platform, even in the US.


Paul Gowder 05.17.06 at 9:34 am

I’m not sure I understand how secret ballots impair the (discursive? for lack of a better word off the top of my head. justificatory?) virtues of the process. As a practical matter, individual voters could never be interrogated en masse as to their reasons for voting any particular way anyway, because of the sheer number of participants. Sure, I suppose individual voters could be interrogated by their being-coerced friends about their vote (“hey jerk, why’d you vote for the tax increase initiative?”). However, a secret ballot doesn’t meaningfully impair this informal process of justification, because those friends could just ask how they voted. (In current world, they can refuse to answer, but in public ballot world, they could just refuse to justify. Either way, the ultimate choice to provide a justification is the voters.)

Moreover, parties and platforms seem to be a good practical stand-in for the justification of individual votes. In order to get votes, they have to make some public justification of their platforms, however non-ideal. Sure, a full deliberative system would probably be better at providing justifications, but excluding that, I don’t really see how public ballots would be better than private ballots in the context of the current system.

What am I missing here?


micah 05.17.06 at 9:36 am

Since we’re assuming a well-ordered society, we assume that these statements are not dishonest.

Not to complicate things too much, but why assume that the conditions of a well-ordered society includes a sincerity requirement for public justification? I actually think a strong sincerity requirement is justified, and have argued for this, but I’m not sure the demand for public justification requires it. Suppose A wants to justify policy P to B, and A is justified in thinking that B is (reasonably) committed to P based on reason R1. Now assume that A is (reasonably) committed to P based on R2. What’s wrong with A giving R1 as her public justification to B? Obviously, this gets much more complicated in Mike’s example above, where the justification is intended for a mass audience. But if we assume that there might be multiple reasonable grounds for justifying a political outcome, why must people be sincere in giving their actual reasons? I’ve tried to work this out elsewhere, but I think the answer has to do with the deliberative/epistemic benefits of sincere public justification, related to the sorts of cascading effects mentioned in comment 13 above (which I think cut the other direction from Loren’s suggestion).


Paul Gowder 05.17.06 at 11:22 am

Micah — you say you have argued for a strong sincerity requirement, but it’s not clear from your comment whether you did so because of cascading effects, etc., or some other reason. Could you clarify? I’ve been wondering in the back of my head about sincerity requirements too.

One possible response to the formulation you give is, I guess, to define the problem out of existence. If A is justified in thinking that B is reasonably committed to P based on reason R1, then A must believe that R1 is a reasonable reason to support P. (Otherwise, A could not believe that B’s committment is reasonable.) In that case, on one notion of sincerity, A is being perfectly sincere when A gives R1 to B, because A is giving one of the reasons A thinks is reasonable.

To invalidate A’s conduct, one would have to come up with a justification for a requirement of sincerity that requires giving all of one’s reasons. Is there one? If the goal, stated broadly/loosely, is to respect the autonomy of the person entitled to claim reasons, how does a withholding of some reasons impair it?


micah 05.17.06 at 11:35 am

Paul, I’m not sure it’s possible to dissolve the problem the way you suggest. A may think B has reasonable grounds for supporting P (at least from within B’s perspective), but A might think those grounds are defective, or perhaps just not the best (or most reasonable) justification for P. If she were candid, A might say: “Look B, given what you believe, you’re committed to P. I don’t happen to agree with what you believe. In fact, I think your background views are wrong. But I agree with you that P is the right policy. It’s just that I have different reasons for supporting P.” A is pointing towards an incompletely theorized agreement on P. But on my hypothesis, A doesn’t say any of this. She just argues from R1 to convince B to support P. I think there are reasons to worry about that case (aside from the obvoius concerns about manipulation), at least from within a developed theory of public reason.


Cranky Observer 05.17.06 at 12:23 pm

> A standard view about what is required for
> coercion to be legitimate is that it must
> be justifiable to the person who is coerced;

I think the hidden assumption here is that the coercor is the individual voter, rather than the polity as a whole ala Hobbes. If one takes the latter view, then the requirement to persuade and/or publish one’s vote disappears.

Even if one believes that voters must persude others, there would then be a need to explain why voters cannot choose to use game-theory type tactics of misdirection in their persuasive efforts.



Christopher Ball 05.17.06 at 1:35 pm

I think #44 is closer to the mark. An individual’s vote is not an immediate coercive act, since it is a vote for a representative who would presumably deliberate with others before making laws, which would coerce others. And #46 gets at the real issue. The more problematic issue is whether publicity of votes would be an effective means to coerce the voter into disclosing his or her reasons. Knowing that I voted for Smith does not tell you why I voted for Smith. To be effective, voters would have to be compelled to explain why they voted as they did. What we are really debating is not a right to privacy but a duty to explain.

However, what is interesting is that in the US, there is no obligation for the legislators to have their votes recorded in most cases. Roll call votes are not mandatory in Congress; voice votes are the norm. A first step for justified coercion would be to require roll call votes except for unanimous consent.


Tom Doyle 05.17.06 at 1:47 pm

Upon this point a page of history is worth a volume of logic.

Justice Oliver Wendell Holmes, New York Trust Co. v. Eisner, 256 U.S. 345 (1921)


Your post raises interesting questions related to the secret ballot, most of which had never occurred to me. I checked at what I have found to be an excellent resource on voting and elections, the Administration and Conduct of Elections web, which offers a huge amount of information on election theory and practice around the world. The following summarizes the history, purpose, and present status of the secret ballot:

History of Corruption and Its Regulation

Today, universal suffrage has been achieved in most parts of the world. The election process has evolved from elections by a handful of privileged elite men to one where leaders are chosen based on a system of one-person-one-vote. The elite’s control over the government has weakened as the world has moved toward democracy by the expansion of suffrage, increased public political participation, and more competitive elections. Since some elites could not survive the competitive elections, they have turned to using election fraud as a means to decrease the degree of competition and sustain their power. This fraud can include bribery, undue influence, ballot stuffing, physical intimidation, and dissemination of false information. As methods and the degree of fraud grew, governments found it necessary to regulate fraud and maintain elections free and fair. Governments had various reasons to hold free and fair elections, for instance, to gain national or international respect and legitimacy.38

Significant attempts to regulate electoral fraud emerged in the late 19th century. France introduced the concept of the secret ballot in 1789; however, Australia was the first country to actually employ it in the late 1850s.39 40 The purpose of the secret ballot was to discourage bribery and intimidation. It was intended to reduce the likelihood that others would know for whom the bribed and/or intimidated citizen had voted. The United Kingdom adopted the secret ballot in 1872, and most countries in Western Europe followed in the late 19th century.41


Rob 05.17.06 at 4:30 pm


I think Michael Otsuka may have a fairly conclusive reply to the argument Harry presented, but #44 is obviously wrong because of what #45 says. Legislators aren’t pure delegates, and even if they were, they would still have publicly expressed views on a large number of issues.


Rob 05.17.06 at 4:31 pm

Also, what the hell? You can’t find out how members of Congress voted?


abb1 05.18.06 at 2:15 am

I can’t understand how anyone could argue that voting is not about coercion. Virtually everything in politicis is about coercion, monopoly on violence; governing is a form of coercion. It’s often obscured by rhetoric but the fact remains.

I don’t think you necessarily owe me an explanation why you voted this way or another, other than the obvious “I find candidate (or party) A more appealing, more in line with my worldview”; nevertheless, I think you could make a good case against secret ballot in a “well-ordered society”.

Here’s what Wiki says:

One response to the issue of political privacy as such is that any political privacy outside the expression of voting is inappropriate. Bernard Crick, for instance, a notable advocate of politics as the necessarily-public form of ethics, defined politics itself as “resolution of moral conflict, or ethics, in public”. From this point of view, one may have no right to political privacy, and even such measures as a secret ballot may be denied based on the grounds that political differences between neighbors or friends must be open, in order to avoid over-empowering some central authority.

Fear of such a central authority, that argument goes, is the problem, and knowing there is no political privacy, people become maximally motivated to ensure that the central authority has no arbitrary power. Junius said that “the subject who is truly loyal to the Chief Magistrate will neither advise nor submit to arbitrary measures.” Knowing that such measures might be directed against oneself may be the strongest possible motivation to avoid advising them against others.


abb1 05.18.06 at 2:29 am

And I think Nixon’s 1972 election with his ‘great silent majority’ is a case in point here.


Chris 05.18.06 at 10:45 am

Very very interesting post. Especially as not many people think of electing a government as a form of coercion by proxy.

Though may I ask, given your voting record, whether you still wish to see the Labour manifesto of 1983 implemented?


Christopher Ball 05.18.06 at 11:04 am

Re #55,

I can’t understand how anyone could argue that voting is not about coercion. Virtually everything in politicis is about coercion, monopoly on violence; governing is a form of coercion

The aggregation of individuals acts of voting determines who controls the coercive apparatus of the state but it seems a stretch to say that the act of voting is in and of itself a coercive act and therefore requires justification. Consider it another way: to receive any form of government aid required the coercive act of revenue gathering by the government, therefore aid recipients are engaging in coercive acts indirectly and so they have a duty to explain why they are receiving aid (not a duty to qualify per regulations but have the fact of their recipient status made public in order to enable other citizens to have them justify their receipt of aid). To me the distance between immediate coercion and voting or aid receipt is too far to compel justification.

While we may get into semantics over what is politics v. say civic or public action, much in politics is persuasive rather than coercive and even at the organizational level, voluntary rather than compulsory mechanisms can be put into place to govern.


harry b 05.18.06 at 11:38 am

chris; at the time I voted Labour as a lesser evil. I’m sure if I re-read the manifesto now I’d support more of it than I did then, though, having mellowed….


Mike Otsuka 05.18.06 at 2:01 pm

not many people think of electing a government as a form of coercion by proxy

The ‘not in my name’ anti-Iraq-war slogan in Britain suggests that, in at least some contexts, many people do.


Tom Doyle 05.18.06 at 9:45 pm

While the secret ballot was originally conceived “not……as a right, but as a mechanism against coercion (in the U.K.) and bribery (in the U.S.)” its status has changed. That elections must be by secret ballot is provided in many international human rights agreements, certainly all that deal with elections. The vast majority of states a parties to at least one such treaty, most a parties to several.

Administration and Cost of Elections Project/ Media and Elections

International Law

The Universal Declaration of Human Rights adopted by the UN in 1948 is the definitive statement of principle on human rights. … Article 21, guarantees the right to take part in periodic secret [ballot]elections

The UDHR imposes obligations upon all members of the international community. But, as a declaration, it is only what is termed customary international law. With the adoption of the International Covenant on Civil and Political Rights in 1966, these same provisions were amplified and given the force of binding and enforceable law over all those states that ratified….Article 25 of the ICCPR states in part:

Every citizen shall have the right and the opportunity, without any of the distinctions mentioned in Article 2 [distinctions of any kind such as race, colour, sex, language, political or other opinion, national or social origin, property, birth or other status] and without unreasonable restrictions:

…(b) To vote and to be elected at genuine periodic elections which shall be by universal and equal suffrage and shall be held by secret ballot, guaranteeing the free expression of the will of the electors.


The main regional human rights treaties – the European Convention on Human Rights, the American Convention on Human Rights and the African Charter on Human and Peoples’ Rights – contain a similar combination of guarantees of the right to freedom of expression and information and the right to political participation without discrimination. [All have boilerplate secret ballot election provisions essentially the same as UDHR and ICCPR]

There are many more examples. The nations of the world seem to approach unanimous consensus on this point. I don’t know why things developed as they did.

m. townes:

I read your excellent article. You seem to know your way around election history. What do you make of this?


jack 05.19.06 at 12:21 am

The idea of secret voting would not suffice the motive behind it.Most of the people don’t think it would be likely to vote through proxy.
There are better ways to vote!

Comments on this entry are closed.