We are all aware voting is largely biased. Of course, it would be attractive to reduce that bias as much as possible because greater policy-based voting allows laws and regulations to better reflect the interests of society.

So is it potentially practical to make the identity of political candidates completely anonymous? In such a way that, we would not know their gender, race, age, religious affiliation, and other physical characteristics or irrelevant characteristics. Hell, part of hiding some of these characteristics require cloaking their voice, their name (because most names are gendered and have some racial/ethnic connotation), et cetera.

I understand this doesn't prevent much party-voting (voting purely for party and not for the stature of a candidates policies), and I am aware one can guess ones race/gender based on their platform - but it would be silly to expect any deterrent to be 100% effective. It seems the only question here is practicality, and that includes the legality of this.

edit: This does not mean we hide the political history of candidates, or any other genuinely relevant piece of information for that matter

  • Comments deleted. Comments should be used to provide constructive criticism on the phrasing of the question. Please don't use comments for debating the subject matter of the question. For more information on what comments should and should not be used for, please review the help article on the commenting privilege.
    – Philipp
    Commented Feb 18, 2019 at 18:08
  • Strictly speaking in countries where the administrative power is held by the party with the majority of the votes or approved by the majority of the parliament, the position of prime minister is decided after the vote. In practice this rarely happens since the head of the party is typically assumed to be the person who is going to occupy that position. But take the case of Theresa May which took office after David Cameron resigned without any general election (but with a party leadership election). So voluntary anonymity is feasible in many countries (some parties have more than one leader).
    – armatita
    Commented Feb 20, 2019 at 12:42

9 Answers 9


No, this isn't at all practical. You're removing virtually everything a voter could possibly use to decide who to support. If you don't know a candidate's identity, all you're left with is what policies they claim to support. But you can't even trust those, because you don't have any way to compare it with things they've previously done. There'd be no reason to even attempt to do what your constituents wanted, because it's not like anyone could hold it against you in a future election.


... it would be attractive to reduce that bias as much as possible because greater policy-based voting allows laws and regulations to better reflect the interests of society.

Theoretically, this can be achieved by reducing the role of elected officials to the minimum with some version of direct democracy. Typically, it's implemented on a regional scale, but there're some experiments on the national level, as well. It's likely that your proposed system of blind elections shares implementation problems with direct democracy, so you might want to look through an earlier discussion on the topic: Why are binding referendums proposed directly by citizens so rare?.

If we go to an even smaller scale, candidate anonymity (blind recruitment) was and is frequently implemented to eliminate bias on a workplace level, with varying results (see The Benefits And Shortcomings Of Blind Hiring In The Recruitment Process).

Implementing the system with blind elections on a larger scale isn't practical for many reasons, to name a few:

  1. Candidates will out themselves.

    Candidates will publish their complete personal information as soon as they find out that they will benefit from that. If it's illegal (which is extremely undemocratic), they will leak that information one way or another.

    Seriously, this point should be enough to render the whole system impractical. If your electorate has a strong racial/ethnic/religious/gender bias the system won't fix that, as the candidate on the right side of this bias will be eager to publish their own and their opponents' personal information.

  2. The system forces candidates to have an unrealistic/self-contradicting set of policy proposals.

    With a politician's character out of the picture, every election will turn into a betting competition of increasingly populist policy proposals. A sane, level-headed candidate will be easily outplayed by a populist that will propose lowering taxes and expanding government services at the same time.

  3. No one agrees on what kind of information is genuinely relevant in the context of national elections.

    It's unclear what exactly should be published in the political history of a candidate. Nationally elected officials usually have to deal with extremely broad issues and it becomes hard to argue that some pieces of information are irrelevant. For example, should we publish the candidate's job history, marital history, health evaluation, age, appearances in court or tax returns? If not, why do you think this information should be hidden from the public view?

    As a result, the published report will either be too detailed (effectively deanonymizing each and every candidate) or not detailed enough (undermining the public access to the information).

    Also, the anonymity by itself will either significantly restrict investigative journalism or benefit candidates with the least amount of information available (the faceless, nameless candidate is less likely to be identified).

  4. It becomes impossible to research the candidate's stance on the issues outside of their published proposal.

    The candidate is likely to avoid the issues he/she is uncomfortable with. Without any public debates and campaign events, it becomes impossible to question the candidate on his/her positions.

    Many issues are not a part of national discourse. Voters who care about local or specialized issues will be left with no way to research candidates' positions on them.

  • Comments are not for extended discussion; this conversation about various other ways to inform voters about the relevant information of anonymized candidates has been moved to chat.
    – Philipp
    Commented Feb 19, 2019 at 15:25
  • 1
    +1, and excellent breakdown. That said: point 2, and to a lesser extent aspects of 3 and 4, can be eliminated with proper use of modern technology: every candidate generates a public-private key pair. Votes are signed by the private key and are permanently publicly recorded (for example, on a blockchain), so people can look up vote history by just knowing the public key. Voters will naturally trust candidates with actual voting records matching desired policy more than candidates with populist claims but empty or contrary voting records.
    – mtraceur
    Commented Feb 20, 2019 at 4:16

Candidates partially compete through their networks. Hillary Clinton spent years developing the support that carried her past Bernie Sanders. Donald Trump's path was different, but he too relied on his established brand. It seems unlikely that the existing candidates, who are accustomed to competing in this way, will give way to this new system that works differently. And by definition, these are the candidates with established networks.

Another issue is whether it is practical to hide identity this way. Look at the show The Masked Singer. In that show, contestants hide their identity, as you suggest. But people speculate, often correctly, on who they are. For example, I know that the Alien was correctly identified before her identity was revealed. As was the Raven before her. How do I know? I read those predictions before the shows and saw them proved right in the after-show reports. Why would it be more successful in politics than in a reality show?

What happens if someone simply reveals that they are running? What happens if a supporter does? An opponent? I can easily see a court case that establishes that it is a candidate's right to reveal identity, overturning any rule against it.

How would this work with the existing ballot system? Currently candidates ask to be on the ballot (in the United States). But under this system, you couldn't ask that. You couldn't go to voters and tell them who they are and ask for signatures.

What happens to experience? Successful presidents were often successful governors or senators or generals. But under this system, we wouldn't know the candidate's experience. And we can't check the candidate's past positions or truthfulness or, well, anything. You aren't just hiding gender and race. You're hiding their past identity.

If you don't hide that identity, then you're not hiding their race or gender either. E.g. in 1980, a retired actor who was formerly the head of the screen actors guild is running for president after almost winning the nomination in 1976 and being governor of California for eight years. How would people not recognize Ronald Reagan? And of course, he was known to be a white male.

More examples:

In 2016, a billionaire businessperson who has also appeared on reality television is running. Against a former first spouse.

In 2012, a multi-millionaire financier who had been governor of Massachusetts is running.

In 2008, a former community organizer from Chicago who is now Senator from Illinois. Is running against a former Marine famous for being tortured while held as a prisoner of war, who is also a current Senator.

In 2004, a Senator from Massachusetts who had previously been a decorated military officer is running.

Does anyone seriously think that these people would win their first office without these aspects of their biography? Not to mention that politicians have been known to exaggerate their biographies. How would you check?

If it worked, it might remove bias. But it would also remove discernment. Unless everyone did this, all the time, candidates couldn't give enough of their background to be considered. And I don't think it is really feasible for everyone to do this to eliminate bias. How would people date?

An easier way to achieve the same objective would be sortition. Just draft the right proportions of each group. It's a mechanical process, so it can only be biased if the process is. And it would save the problem of anonymity. You still lose discernment, but you gain representativeness.


Its possible when the basis of approval is something that can be demonstrated separately from personal information.

For example, some orchestras interview candidates who are behind a curtain, to judge them only by their musical competence. They found that without it, unconscious bias does happen a lot.

Similar might work for objects or documents people create, or problems they solve,or factual data that doesn't identify them.

If probably doesn't work in politics generally, because the skill you want include .matters that are inherently unable to be tested other than by linking with the person and their history - how honest they've been, how well they executed past roles, positive and negative features in their background, possible past controversies, personality, and so on.


The problem with that is trust. Politicians usually have a reputation for either following through or not following through with their proposals.

With this, anyone could run if they list good enough sounding proposals, whether they intend to follow them or not. Voters have no way of knowing who has and does not have a good reputation. You'd be better of just implementing the proposals of whoever wins some other way, instead of having some anonymous person be the leader.


This is practical and already in practice in some contexts, for example moderator elections here on StackExchange and administrator elections on Wikipedia.

As these sources become increasingly used in AI and other ways of influencing decision-makers' decisions, such elections may become more consequential in the future; they arguably already are more consequential than some formal-governmental elections.


It probably would be practical, since if we look at it in a certain way it's not so far from the system we have in the US now.

In a sense our candidates are all anonymous. Which is to say that the carefully groomed personas and cleanly scrubbed backgrounds of US candidates are virtual plastic masks designed to agree with public sentiment -- the masks often look nice enough, but God only knows what's really under those masks... and if ever the public learns, it's often too late.

Similarly, candidates have good names and say good things, but the public doesn't know if they mean what they say, nor if those names shouldn't eventually become infamous.

Perhaps the really good candidates are not anonymous, and are who they appear to be. But the worst ones are never who they seem to be and are always, (in that sense), anonymous, or struggling to remain so. And since we mix up those two sets, so that everyone has the face of, or wears the mask of, a good candidate, it's difficult for the public to tell them apart, just as if they were a group wearing Guy Fawkes masks...

Note: I'm not saying candidates' masks are secret to everyone, it's more of a partial sufficient anonymity that exists so long as most of the general public can't quite figure it out.


A strategy which goes a long way to achieve this is already in use in some European countries for elections to the European Parliament.

This is the so-called "closed list" system where electors only vote for a party, and not for individual candidates. The number of representatives for each party is based on the total number of votes cast (either nationally or subdivided regionally) and the representatives themselves are selected only by the parties.

This is not completely anonymous, because the party will usually publish its list of candidates in selection order (i.e. the appropriate number from the top of the list will be chosen) but it certainly minimizes the personal influence of candidates who are not close to the top of the party's list.

This "closed list" system is used for EU elections in France, Germany, Greece, Portugal and Spain, and a variant of it (with various complications for internal political and historical reasons) was used in the UK also.

There are less extreme versions of this idea, where voters can indicate preference for a particular candidate in the party's list, but the number of candidates elected from the list is still based on the proportion of votes for each party as a whole.

Of course both of these system effectively eliminate the concept of "independent" (non-party) candidates. An independent candidate can set up a one-candidate "party," but the system is strongly biased against such candidates ever being elected.


I think the reflex is to figure that there are a lot of downsides to this sort of system, in that candidate's orientation to a position may be subjectively quantified incorrectly to steal more votes, or that the candidate's character may be too much of a liability to outweigh the popularity of their positions. Casting a real vote for a blind candidate can be binding in an overly final way. However...

(The rest of this answer is going to sound like an ad) I think that using a candidate-matching application like isidewith.com might shore-up some of these problems. I say this because, although it will tell you which candidate's policies are most aligned with your own, using the app doesn't constitute a binding vote. When you're using it, you fill-out each of your policy preferences in a manner the same as the OP is describing, and the app matches you with the best fit candidate same as the OP is describing. But once the results are revealed to you, it's not too late for you to pass on any candidate that does not meet your criteria based on character or background. There is also no risk that a machine will forcibly cast a binding and blind vote on your behalf based on dishonest policy matching. The major flaw that prevents this from being a more perfect match to the OP's question is that we don't have, and never will have, universal participation in a candidate-matching program. Currently, it is by each voter's discretion, and I think low-information voters may also be the kind who are less likely to seek-out the advice of candidate-matching software.

A related possibility would be for a voter to fill-out policy positions up-front in the voting booth, which will fill-out a pre-selected candidates based on their responses, at which time it is within the voter's discretion to change these selections. This offers a high degree of convenience, however, it would still be troubling that the voting machine would be overly guiding you to select specific candidates. This is a bit reminiscent of past incidents of touch screen voting machines checking the opposite of the candidate you clicked on, or selecting the opposite candidate when you opt for a straight party vote.

Based on user581844's input, another way you could apply candidate-matching software would involve not just stating what your positions are, but choosing a weight for how much you care about that particular position. This would probably preserve more fidelity to the voter's preferences. Downsides would be -- further obscuring the transparency behind the math, it caters to single-issue voters (which they may like, but I think single-issue voting often erodes the quality of voting), and the time factor of filling-out twice as many selections (and the risk that it could create longer lines at the polls, if this software is used at the polling site). Nevertheless, this is a valid alternative possibility.

  • 1
    I love this idea for many reasons, however, it does have an additional huge issue. Mathematically, it makes sense. It's perfectly ideal to go with the most politically similar candidate. The issue, though, is that even the most politically similar candidate may be undesired because one or more of their policies are vehemently vile to you. Say the one thing they support, such as abortion or heresy laws, is what keeps you from voting for them. This is why voting is a bit more nuanced than plain old value overlap
    – user581844
    Commented Feb 18, 2019 at 20:20
  • @user581844 what prevents you from being allowed to define dealbreakers? There's no reason a similar method can't start by excluding candidates that support or don't support your most important policies and then look for the one that represents your values the most closely
    – Thymine
    Commented Feb 20, 2019 at 10:24
  • @Thymine, I suppose there is no reason. It seems if we tweak our system just enough, and are just technical enough, a system like this is totally practical. Unfortunately I don't see it happening for a long time
    – user581844
    Commented Feb 20, 2019 at 19:26

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .