You say don't try to use game theory to figure out how to best "make a difference" but admit you will have virtually no influence in this election and instead just vote for the person you like best, among the candidates listed on the ballot. But why not continue with this logic, and "write-in" the person in the world you like best? Why not write them in even if write-ins aren't officially allowed in this election? Why not skip the official elections and make up your own polling place to vote at? Why not just declare your vote for them in a blog post?
"With 300 million available candidates, many of whom are not nincompoops, why does America keep electing nincompoops to political office?"
The whole argument breaks down rapidly if you believe this point is false because an overwhelming majority of people are, in fact, nincompoops. Even more so if you believe anti-nincompoopish behavior is exceedingly likely to bear a high social cost (including being 'unelectable').
"Then some combination of the party structure, and the media telling complicit voters who voters are likely to vote for, is exerting on the order of 14-15 bits of power over the Presidency; while the voters only exert 3-4 bits."
I don't buy this. The vast majority of random people would lose an election race against Hillary/Giuliani/etc even if the party structure and media supported them. So I would say many of those 14-15 bits are actually forced moves caused by good estimates of voter preference. Am I missing something? I'm having trouble thinking of a standard of nincompoophood according to which the average presidential candidate is a nincompoop but the average voter is not.
"I'm glad I didn't do the "sensible" thing. Less blood on my hands."
I tend to think that what matters is whether or not the blood is still in the body of the person it belongs to and not so much whose hands it's on when it's out.
"Stop trying to be clever, just don't vote for nincompoops."
Thank you! Strategic voting makes me want to stab people in the face, particularly when they turn around after the election and moan about idiots getting in.
Steven - you are missing a small something. The issue of whether or not Joe Bloggs would or wouldn't beat the Front Runners if it came down to it is immaterial. We're not given that choice, it's made for us long before. It's not just that we only have 1-2 bits of freedom, it's that they're the final 1-2 bits.
Robin - continue even further. Why not just ignore the entire thing and get drunk instead? The difference between your suggestions and Eliezer's is that voting (within the system) has a nonzero weight (within the system). Yours have precisely zero weight. That's where the line is drawn. The key word in your comment is 'virtually'. No one believes their vote will be the casting vote, but people vote.
I must say though Eliezer, your final sentence begs the question 'Why vote at all?' Spend that half hour lobbying the winner. In fact, stop wasting time reading a blog website and get out there and promote your Wonderful Candidate!
"The issue of whether or not Joe Bloggs would or wouldn't beat the Front Runners if it came down to it is immaterial. We're not given that choice, it's made for us long before."
The parties/media don't support Bloggs because he wouldn't win if they did. Bloggs wouldn't win because the people don't want him as a president. So Bloggs isn't an option because people don't want him as a president. How is that not influence?
Since it's likely that who the next few Presidents are will strongly affect the probability of existential disaster, and because of the stupidly huge utility of existential risk reduction, the expected utility of a strategic vote for an electable candidate could be surprisingly high. Given the uncertainty of the effect of a protest vote, and the uncertainty of the US's existence in a few election cycles (which is greater the more likely you believe a near-term existential disaster or Singularity to be), it could be higher than the expected utility of a protest vote.
I wonder how to decide what candidate is best on existential risk. I would guess military policy is the biggest component. Energy policy might be second. MNT research funding also matters, but candidates aren't likely to have clear distinct positions on this, and I'm not sure whether more or less funding is good.
Ben: re what Steven is saying, think of polls. The candidates are strongly influenced by the voters before the election.
Steven - you're making way too many assumptions here. No Joe Bloggs wouldn't get voted in, but that's not what the original post is talking about, nor yesterday's reference to Colbert. Nor is it my point. The point is that there are a number of practically insurmountable obstacles/safeguards in between Joe Bloggs and the voter even getting a say, and for most people this doesn't even mean anything. They get a list of names of Ordained Front Runners from on high, and debate them without a thought as to where they came from.
"The parties/media don't supp...
As I understand Robin's point, he is saying that if we don't consider the expected utility of our vote, but only consider the goodness of the candidate, then whether or not we vote within or apart from the system is irrelevant. The only reason for not voting on a blog post is that one doesn't suppose that this has a high expected utility.
Of course, since no one has provided any statistics, it is not at all evident that voting for a major party candidate has a higher expected utility than a vote for a third party, especially since the fact that one's candid...
I'm a bit puzzled why it is rational to bother to vote and irrational to purchase lottery tickets, based on the mathematics. Surely there is a clear, obvious unbiased Bayesian explanation somewhere, but I seem to have misplaced it. . .
How could we distinguish between:
"1. You should vote for the less evil of the top mainstream candidates, because your vote is unlikely to make a critical difference if you vote for a candidate that most people don't vote for. 2.You should stay home, because your vote is unlikely to make a critical difference."
Altruists should vote on the basis of expected utility. Depending on the state and the election (primary or general), as well as information from polls and political prediction markets, you may be able to determine that you have much more th...
One interesting alternative to voting for those who are unable to participate in a tightly contested New Hampshire primary might be to help pay for the creation of a randomly selected nationally representative paid (providing full salary replacement if necessary to ensure high participation rates) private 'jury' to study the candidates and issues and ultimately vote its preferences.
Initially this would have no legal force, but it could reveal systematic differences between the decisions of the actual electorate and a representative sample treated for great...
Carl, the random jury approach makes a lot of sense, but has been proposed for a long time without generating much interest. That fact says something important.
Robin,
I know, and I have read your essay on the subject (along with others encountered prior to that, and earlier study of Athenian democracy). Government-established juries have been used to come up with electoral reform proposals to be subjected to referenda in the Canadian provinces of British Columbia and Ontario in just the last few years, bypassing the vested interests of officeholders.
Also, the political/legal barriers to replacing elections with juries are immense (e.g. expressive voting, tradition, etc), but do not apply to privately funded juries...
In California, I'm a registered independent who chose to vote in the Democratic primary. However, I do not need to get in a car and drive to some cold, ill-lit polling place. I vote from the relative comfort of home via the "absentee" ballot. BTW, in 2000 it was obvious that Bush would be a national disaster -- it was obvious in 2004 that Bush was a disaster on a planetary scale. What went wrong in the body politic that the people brought on themselves a postmodern replay of the plagues of Egypt?
Eliezer:
"Besides, if you're going to apply game theory to the situation in a shortsighted local fashion - not taking into account others thinking similarly, and not taking into account the incentives you create for later elections based on what potential future candidates see you doing today - if, I say, you think in such a strictly local fashion and call it "rational", then why vote at all, when your single vote is exceedingly unlikely to determine the winner?"
You mention:
Kip : The Newcombe problem only needs about 30 seconds thought: as soon as you've postulated reversed causality, any reasoning based on the premise 'there's 1m€ in box B at the moment of decision' breaks down on the meaninglessness of the notion 'at the moment of decision' under reversed causality. Are all 'philosophical paradoxes' so trite ? At least I suppose while people are 'exercising their thinking' over such trivialities they're not doing us serious harm by working on self-improving AI.
Voting has a tiny probability to affect anything, but when it does, it affects a proportionally huge number of people. So while voting is irrational for egoists, it seems like a good deal for altruists/utilitarians because the 1/N in the probability and the N in the effect cancel out (except insofar as they have other unusually worthwhile ways to spend their time and insofar as it doesn't matter who wins).
If you'd like someone to try the random jury approach, you need to think about how to turn it into good TV.
I'm certainly never going to vote for a substandard candidate again. I was fooled by Bush in 2000, and now I have blood on my hands. Well, my state went to Gore, but conceptually... No, I'm definitely going to show up for the caucuses, and try to become a delegate, to avoid this sort of outcome, and if the duopoly choice is between two bloodthirsty fascists who want to sell my children into slavery, one on the left, and one on the right, I'm going for a third party, and after that, probably for ex patriation while it is still an option.
Since I don't accept being part of a majority that dominates a minority, I only consent to vote in a situation where my vote is for the minority, and therefore cannot possibly influence the outcome. This is mathematically identical to staying home, except staying home is more pleasant. So, I'd rather stay home.
For those who believe in majority rule, I still don't understand why you vote, since your vote cannot make any difference. There is no such thing as a deciding vote in a large election, since the error present in the system even for a fair election i...
TGGP,
A waste of time for whom? With what values, level of ability, wage rate, citizenship, residency, etc?
Your one vote may be a statistical waste, but campaigning, and especially campaigning in the primary, can exert influence on hundreds or thousands of votes. That's a huge force multiplier. Ron Paul himself is basically making headway solely on the strength of his campaigning supporters - in defiance of media and party.
Robin, TGGP: Though who you vote for is of course confidential, whether or not you vote is a matter of public record. So, if you ever become famous enough that advocating a candidate isn't a waste of your time, you can expect your record to end up on The Smoking Gun (as happened to musician Lenny Kravitz). Also, while obtaining a voting record now requires a written request and usually a small fee, there's no reason why the information couldn't eventually become freely available on the Internet and thence a common tool for judging the character of potential employees.
"...which would create a chance to elect an exceptionally competent governor."
Competence is only an issue when we agree on the proper societal function of the candidate/office. No one ever asks: "I wonder how competent John Gotti was." If a Hitler comes to power, do we want him to be competent or incompetent in implementing his policies?
Given that we're talking about a national election here, I can somewhat sympathize with those who fail to vote because the expected utility of the vote is either zero or vanishingly small. But does the same logic apply to local elections, especially those in rural areas or small precincts where individual votes suddenly carry a much stronger weight?
Since I was first eligible to vote, in partisan races, I have always voted for Libertarian candidates. My reasoning has been that my vote won't be decisive and symbolic votes "against" one candidate are always reported as votes "for" the recipient, so the symbolism is entirely in the individual voter's head. This means the only symbolic vote that anyone can read afterward is one for a candidate who clearly represents your point of view. I'm a libertarian (once I was a Libertarian.) The extra vote I give to Libertarian candidates is ...
Forget voting. Here's how to make a big difference in society: at least once a month, do something amazingly kind for a perfect stranger. My preference is leaving $100 tips for waitresses or hotel maids, because I'm basically lazy.
Also, raise your kids with kindness.
Practice showing courage in challenging situations.
Don't instigate a lawsuit unless it's reaaaaaally important.
What's great about America is not democracy, but the sense we have that we can travel almost anywhere here and other people will smile with us, do business with us, and not hate us. There are still many places and people within America for which and whom this is not true (or not true enough). But let's keep working toward that idea with our daily actions. No amount of voting will solve that problem.
You seriously have to wonder how they manage to pull this sort of thing without a deliberate conspiracy. But I stand by my probability estimate: someone would have blabbed by now.
...Besides, if you're going to apply game theory to the situation in a shortsighted local fashion - not taking into account others thinking similarly, and not taking into account the incentives you create for later elections based on what potential future candidates see you doing today - if, I say, you think in such a strictly local fashion and call it "rational", then why vote at all, when your single vote is exceedingly unlikely to determine the winner?
Consider these two clever-sounding game-theoretical arguments side by side:
- You should vote f
Hmmm.... the fewer people that vote, the more influential each voter is.
Everyone else, stay home! ;)
How much of the apparent nimcompoopery of presidents is actually a result of their choices being constrained while in office?
If some fraction of the problem is a result of the desire to be re-elected, should we be pushing for only letting them have one term?
Is voting at all some variant of the dust-speck problem?
If some fraction of the problem is a result of the desire to be re-elected, should we be pushing for only letting them have one term?
You want to free up your presidents from the corrupting influence of democracy?
Maybe. If I were being serious about this, I'd look at whether Presidents were significantly different in their second terms compared to their first, and whether the difference (if any) seemed generally an improvement or a deterioration. I'd also look at how term limits play out generally.
Maybe the South Park episode "Vote or Die" is relevant...
But Eliezer, don't you know? Voting is always between a giant douche and a turd sandwich. Nearly every election since the beginning of time has been between some douche and some turd. They are the only people who suck up enough to make it that far in politics.
...Amazingly, the media collectively exerted such tremendous power, in nearly perfect coordination, without deliberate intention (conspiracies are generally much less necessary than believed). They genuinely thought, I think, that they were reporting the news rather than making it. Did it even occur to them that the entire business was self-referential? Did anyone write about that aspect? With a coordinated action, the media could have chosen any not-completely-pathetic candidate to report as the "front-runner", and their reporting would thereb
Consider these two clever-sounding game-theoretical arguments side by side:
You should vote for the less evil of the top mainstream candidates, because your vote is unlikely to make a critical difference if you vote for a candidate that most people don't vote for. You should stay home, because your vote is unlikely to make a critical difference. It's hard to see who should accept argument #1 but refuse to accept argument #2.
Yes! This is the argument I've been making in government class for the past few weeks. This disjunction shows that the type of reaso...
But there's this thing called the Internet now
I'm
I find it pretty hilarious in retrospect that Hillary Clinton and Rudy Giuliani were listed in this essay as the obvious choices, and Obama as a wacky protest vote.
Warning: I haven't checked if someone else has already made this point.
If you plan on attempting to influence anyone's vote other than your own, it would be hard in practice to convince other people that they should vote for X if you yourself didn't signal intent to vote for X, and the easiest way to do that is to actually intend to vote for X. In other words, a good reason to vote for X is so you don't have to lie when trying to convince other people (e.g. your friends in swing states) to vote for X (the less-wrong-lizard). In other other words, being abl...
The idea that those nominated for national office are usefully categorized as nincompoops is extremely low probability. Most of us, even the more popular among us, even Robin and Eliezer and Anna themselves, could not get nominated even for the senate. Or if they could, by the time they had done what they needed to do to get nominated, Eliezer2008 who wrote this post would measure them as nincompoops.
Among other issues, the electorate is going to filter for some mix of popular opinions. If you study for years in order to move your policy opinions way ...
...Amazingly, the media collectively exerted such tremendous power, in nearly perfect coordination, without deliberate intention (conspiracies are generally much less necessary than believed). They genuinely thought, I think, that they were reporting the news rather than making it. Did it even occur to them that the entire business was self-referential? Did anyone write about that aspect? With a coordinated action, the media could have chosen any not-completely-pathetic candidate to report as the "front-runner", and their reporting would thereby ha
This felt a bit too much like naive purity ethics back in 2008, and it looks even worse in the light of the current situation in the USA.
As for a substantive criticism:
...Consider these two clever-sounding game-theoretical arguments side by side:
- You should vote for the less evil of the top mainstream candidates, because your vote is unlikely to make a critical difference if you vote for a candidate that most people don't vote for.
- You should stay home, because your vote is unlikely to make a critical difference.
It's hard to see who should accept argument #1 bu
Followup to: The Two-Party Swindle, The American System and Misleading Labels
If evolutionary psychology could be simplified down to one sentence (which it can't), it would be: "Our instincts are adaptations that increased fitness in the ancestral environment, and we go on feeling that way regardless of whether it increases fitness today." Sex with condoms, tastes for sugar and fat, etc.
In the ancestral environment, there was no such thing as voter confidentiality. If you backed a power faction in your hunter-gatherer band, everyone knew which side you'd picked. The penalty for choosing the losing side could easily be death.
Our emotions are shaped to steer us through hunter-gatherer life, not life in the modern world. It should be no surprise, then, that when people choose political sides, they feel drawn to the faction that seems stronger, better positioned to win. Even when voting is confidential. Just as people enjoy sex, even when using contraception.
(George Orwell had a few words to say in "Raffles and Miss Blandish" about where the admiration of power can go. The danger, not of lusting for power, but just of feeling drawn to it.)
In a recent special election for California governor, the usual lock of the party structure broke down - they neglected to block that special case, and so you could get in with 65 signatures and $3500. As a result there were 135 candidates.
With 135 candidates, one might have thought there would be an opportunity for some genuine voter choice - a lower barrier to entry, which would create a chance to elect an exceptionally competent governor. However, the media immediately swung into action and decided that only a tiny fraction of these candidates would be allowed to get any publicity. Which ones? Why, the ones who already had name recognition! Those, after all, were the candidates who were likely to win, so those were the ones which the media reported on.
Amazingly, the media collectively exerted such tremendous power, in nearly perfect coordination, without deliberate intention (conspiracies are generally much less necessary than believed). They genuinely thought, I think, that they were reporting the news rather than making it. Did it even occur to them that the entire business was self-referential? Did anyone write about that aspect? With a coordinated action, the media could have chosen any not-completely-pathetic candidate to report as the "front-runner", and their reporting would thereby have been correct.
The technical term for this is Keynesian beauty contest, wherein everyone tries to vote for whoever they think most people will vote for.
If Arnold Schwarzenegger (4,206,284 votes) had been as unable to get publicity as Logan Clements (274 votes), perhaps because the media believed (in uncoordinated unison) that no action-movie hero could be taken seriously as a candidate, then Arnold Schwarzenegger would not have been a "serious candidate".
In effect, Arnold Schwarzenegger was appointed Governor of California by the media. The case is notable because usually it's the party structure that excludes candidates, and the party structure's power has a formal basis that does not require voter complicity. The power of the media to appoint Arnold Schwarzenegger governor derived strictly from voters following what someone told them was the trend. If the voters had ignored the media telling them who the front-runner was, and decided their initial pick of "serious candidates" based on, say, the answers to a questionnaire, then the media would have had no power.
Yes, this is presently around as likely as the Sun rising in the west and illuminating a Moon of green cheese. But there's this thing called the Internet now, which humanity is still figuring out how to use, and there may be another change or two on the way. Twenty years ago, if the media had decided not to report on Ron Paul, that would have been that.
Someone is bound to say, at this point, "But if you vote for a candidate with no chance of winning, you're throwing your vote away!"
To which the economist replies, "But you can't always jump from a Nash equilibrium to a Pareto optimum," meaning roughly, "Unless everyone else has that same idea at the same time, you'll still be throwing your vote away," or in other words, "You can make fun all you like, but if you don't vote for a lizard, the wrong lizard really might get in."
In which case, the lizards know they can rely on your vote going to one of them, and they have no incentive to treat you kindly. Most of the benefits of democracy may be from the lizards being scared enough of voters to not misbehave really really badly, rather than from the "right lizard" winning a voter-fight.
Besides, picking the better lizard is harder than it looks. In 2000, the comic Melonpool showed a character pondering, "Bush or Gore... Bush or Gore... it's like flipping a two-headed coin." Well, how were they supposed to know? In 2000, based on history, it seemed to me that the Republicans were generally less interventionist and therefore less harmful than the Democrats, so I pondered whether to vote for Bush to prevent Gore from getting in. Yet it seemed to me that the barriers to keep out third parties were a raw power grab, and that I was therefore obliged to vote for third parties wherever possible, to penalize the Republicrats for getting grabby. And so I voted Libertarian, though I don't consider myself one (at least not with a big "L"). I'm glad I didn't do the "sensible" thing. Less blood on my hands.
If we could go back in time and change our votes, and see alternate histories laid out side-by-side, it might make sense to vote for the less evil of two lizards. But in a state of ignorance - voting for candidates that abandon their stated principles like they discard used toilet paper - then it is harder to compare lizards than those enthusiastically cheering for team colors might think.
Are people who vote for Ron Paul in the Republican primary wasting their votes? I'm not asking, mind you, whether you approve of Ron Paul as a candidate. I'm asking you whether the Ron Paul voters are taking an effectless action if Ron Paul doesn't win. Ron Paul is showing what an candidate can do with the Internet, despite the party structure and the media. A competent outsider considering a presidential run in 2012 is much more likely to take a shot at it now. What exactly does a vote for Hilliani accomplish, besides telling the lizards to keep doing whatever it is they're doing?
Make them work for your vote. Vote for more of the same this year, for whatever clever-sounding reason, and next election, they'll give you more of the same. Refuse to vote for nincompoops and maybe they'll try offering you a less nincompoopy candidate, or non-nincompoops will be more likely to run for office when they see they have a chance.
Besides, if you're going to apply game theory to the situation in a shortsighted local fashion - not taking into account others thinking similarly, and not taking into account the incentives you create for later elections based on what potential future candidates see you doing today - if, I say, you think in such a strictly local fashion and call it "rational", then why vote at all, when your single vote is exceedingly unlikely to determine the winner?
Consider these two clever-sounding game-theoretical arguments side by side:
It's hard to see who should accept argument #1 but refuse to accept argument #2.
I'm not going to go into the notion of collective action, Prisoner's Dilemma, Newcomblike problems, etcetera, because the last time I tried to write about this, I accidentally started to write a book. But whatever meaning you attach to voting - especially any notions of good citizenship - it's hard to see why you should vote for a lizard if you bother to vote at all.
There is an interaction here, a confluence of folly, between the evolutionary psychology of politics as a football game, and the evolutionary psychology of trying to side with the winner. The media - I am not the first to observe this - report on politics as though it is a horse race. Good feelings about a candidate are generated, not by looking over voting records, but by the media reporting excitedly: "He's pulling ahead in the third stretch!" What the media thinks we should know about candidates is that such-and-such candidate appeals to such-and-such voting faction. Since this is practically all the media report on, it feeds nothing but the instinct to get yourself on the winning side.
And then there's the lovely concept of "electability": Trying to vote for a candidate that you think other people will vote for, because you want your own color to win at any cost. You have to admire the spectacle of the media breathlessly reporting on which voting factions think that candidate X is the most "electable". Is anyone even counting the levels of recursion here?
Or consider it from yet another perspective:
There are roughly 300 million people in the United States, of whom only one can be President at any given time.
With 300 million available candidates, many of whom are not nincompoops, why does America keep electing nincompoops to political office?
Sending a message to select 1 out of 300 million possibilities requires 29 bits. So if you vote in only the general election for the Presidency, then some mysterious force narrows the election down to 2 out of 300 million possibilities - exerting 28 bits of decision power - and then you, or rather the entire voting population, exert 1 more bit of decision power. If you vote in a primary election, you may send another 2 or 3 bits worth of message.
Where do the other 25 bits of decision power come from?
You may object: "Wait a minute, not everyone in the United States is 35 years old and a natural-born citizen, so it's not 300 million possibilities."
I reply, "How do you know that a 34-year-old cannot be President?"
And you cry, "What? It's in the Constitution!"
Well, there you go: Since around half the population is under the age of 35, at least one bit of the missing decision power is exerted by 55 delegates in Philadelphia in 1787. Though the "natural-born citizen" clause comes from a letter sent by John Jay to George Washington, a suggestion that was adopted without debate by the Philadelphia Convention.
I am not necessarily advising that you go outside the box on this one. Sometimes the box is there for a good reason. But you should at least be conscious of the box's existence and origin.
Likewise, not everyone would want to be President. (But see the hidden box: In principle the option exists of enforcing Presidential service, like jury duty.) How many people would run for President if they had a serious chance at winning? Let's pretend the number is only 150,000. That accounts for another 10 bits.
Then some combination of the party structure, and the media telling complicit voters who voters are likely to vote for, is exerting on the order of 14-15 bits of power over the Presidency; while the voters only exert 3-4 bits. And actually the situation is worse than this, because the media and party structure get to move first. They can eliminate nearly all the variance along any particular dimension. So that by the time you get to choose one of four "serious" "front-running" candidates, that is, the ones approved by both the party structure and the media, you're choosing between 90.8% nincompoop and 90.6% nincompoop.
I seriously think the best thing you can do about the situation, as a voter, is stop trying to be clever. Don't try to vote for someone you don't really like, because you think your vote is more likely to make a difference that way. Don't fret about "electability". Don't try to predict and outwit other voters. Don't treat it as a horse race. Don't worry about "wasting your vote" - it always sends a message, you may as well make it a true message.
Remember that this is not the ancestral environment, and that you won't die if you aren't on the winning side. Remember that the threat that voters as a class hold against politicians as a class is more important to democracy than your fights with other voters. Forget all the "game theory" that doesn't take future incentives into account; real game theory is further-sighted, and besides, if you're going to look at it that way, you might as well stay home. When you try to be clever, you usually end up playing the Politicians' game.
Clear your mind of distractions...
And stop voting for nincompoops.
If you vote for nincompoops, for whatever clever-sounding reason, don't be surprised that out of 300 million people you get nincompoops in office.
The arguments are long, but the voting strategy they imply is simple: Stop trying to be clever, just don't vote for nincompoops.
Oh - and if you're going to vote at all, vote in the primary. That's where most of your remaining bits and remaining variance have a chance to be exerted. It's a pretty good bet that a Republicrat will be elected. The primary is your only chance to choose between Hilliani and Opaula (or whatever your poison).
If anyone tells you that voting in a party's primary commits you to voting for that party in the general election, or that a political party owns the primary and you're stealing something from them, then laugh in their faces. They've taken nearly all the decision bits, moved first in the game, and now they think they can convince you not to exercise the bits you have left?
To boil it all down to an emotional argument that isn't necessarily wrong:
Why drive out to your polling place and stand in line for half an hour or more - when your vote isn't very likely to singlehandedly determine the Presidency - and then vote for someone you don't even want?