Government cannot exist without the tacit consent of the populace. This consent is maintained by keeping people in ignorance of their real power. Voting is not an expression of power, but an admission of powerlessness, since it cannot do otherwise than reaffirm the government's supposed legitimacy.
--Fred Woodworth

Reason I
Votes held almost everywhere in the U.S. today are winner-take-all affairs. What this means is that whichever side gets the majority of the vote wins and wins completely. The losing side gets nothing, even if it is a significant chunk of the population (over forty-nine percent, for instance). This leads to a tyranny of the majority wherein those with unpopular views (or not-quite-popular-enough views) get virtually ignored with respect to the result of the vote.

It also means that if one side (even the 'bad side', as it were) manages to rally a majority of the public to their cause on flimsy grounds (say by using emotive euphemisms or lying), that result will go through regardless. Thus voting results (1) often fail to represent a large amount of the population, and (2) are won through popularity contests rather than because one side is innately better. From (2), we see that the debate about a vote (and thus its result) is likely to revolve around shallow issues which affect popularity rather than establish an even forum to give all sides a fair and balanced airing so that voters can choose among them responsibly (still as a majority, of course). Voting in the U.S. today rides on the shallow tyranny of the majority, and so I choose not to vote¹.

Reason II:
Vote fraud renders the entire voting process a joke, and the chance of vote fraud in government elections in the U.S. right now is significant. Specifically, changing over to vulnerable electronic voting systems with untested, trade-secret source code (programmed by corporations with political connections) has left the possibility wide open for unprecedented vote fraud. For examples from recent news, check out the stories of easily tampered Diebold voting systems (with no paper trail to audit), easily tampered net voting systems (developed by the Pentagon, shredded by independent security analysis), and evidence of a willingness on the part of political parties to illegally exploit technological security flaws to benefit the party.

Additionally, related to Reason III below, any instance of fraud which is not discovered and corrected leaves the defrauded voters with the mistaken assumption that they are submitting themselves to the result of a 'fair vote' - which is not necessarily to say a fair vote, but one which follows the expected rules, such as one vote per person and every valid vote counted - and so there is a significant chance that those who vote are not just accepting the authority of the voting system, but are unintentionally accepting the usurped authority granted to the fraudulent results. Voting results are easily tampered with, and so I choose not to vote.

Reason III:
Among a group of friends trying to decide where to eat dinner, an informal vote is used when no consensus can be reached. The assumption behind such a vote is that those who lose are agreeing to go along with the decision regardless; they are implicitly submitting to the authority of the vote. Unfortunately, it follows that those with unpopular tastes lose out, over and over, to those with the more popular menu choices.

Now, among a group of friends, the vote winners will often take up the suggestion of the loser once in a while to keep things less unfair. However, in government elections there is no such hope for those who lose. If a person registers their vote, if they contribute their opinion, and it just does not fit the popular one, then their vote - their opinion - is inefficacious, worthless. At the same time, by participating in the system, the person has at least symbolically granted it legitimacy for himself or herself, agreeing to go along with the decision regardless.

To register a vote is to say that one accepts the authority of the vote-system - not just any outcome, but every outcome, every possible outcome. I do not wish to accept that authority, and so I choose not to vote.

Corollary To Reason III:
People who vote are less likely to attempt deeper change in the system. Even if an individual person's vote does have a significant effect on the results, the change to the system coming through voting is likely to be relatively small reform within the system. Those who seek larger changes to the system itself are not likely to find results through voting; yet by voting, they may feel they have made progress when no real progress has been made (with respect to change at the level of the system itself).

More generally, if a person's vote does not have a significant effect, voting gives a false sense of having done something, of having made progress. People come home from the voting booths feeling they have taken action toward their desired goals. Yet if, as will be argued below, an individual's vote has no significant effect, and yet people think it does, then significant change is less likely to occur because people feel less need to act in other ways if they feel they have already 'done their part' merely by registering a vote.

Put another way, people who submit to the voting system are likely to blame bad results on other voters, rather than the underlying system. This transference of blame discourages reform at the level of the system itself.

Reason IV:
At the lower, local levels of government, where small numbers of voters are involved, a single vote may occasionally - thought still rarely - make the difference between one outcome and another. The higher a person goes, however, the less chance that any one person's vote (say, the vote of someone who would not have voted otherwise) would make any difference at all. Certainly at the level of national elections in the U.S. today, the chance of one person's vote making the difference in any given state is so close to zero that the difference between it and zero is negligible.

I invite historical statistics to the contrary, keeping in mind, e.g., the distinction between one electoral vote and one citizen vote. Even if there is such a case, I would be interested in whether there was any evidence that any single additional vote would not have gone to the side that was already in the lead rather than the side one vote behind; which is to say, even if there have been rare historical instances of one-vote leads, there also needs to be reason to think that the non-voters would have voted for the losing side rather than splitting evenly or voting for the winner (which they would be likely to do if they were representative of the voting population at large). Without such evidence, even historical instances of one-vote leads (which are rare enough to work for the argument in general anyway) would not undercut this argument.

Given the above, there is little reason for a person to place their own vote. Much more effective in practice (assuming the outcome of the vote mattered, I should add) would be to take the time to campaign for one's cause, which is to say convince a number of other people who would have not voted (or voted differently) to vote in the desired way. The idea here is not that the small number of people so swayed would make much more difference than an individual vote, but that they would likely start off a chain reaction that leads to a much more significant effect.

For example, if one convinces twenty people to vote for cause Y (and none of them were already going to do so anyway), then perhaps two of those twenty will go on to campaign in a similar manner, bringing in another twenty each. Two people from each of those new groups of twenty (so four new people) might then go on to recruit twenty votes each, and so on in a process that repeats itself for a while. If a person could reach the critical mass required to start such a chain reaction, the chance of affecting the outcome of the vote is drastically increased. Whether a successful vote (that is, one in the campaigner's favor) leads to significant worthwhile change is up in the air - see Corollary To Reason III - but it is reasonable to assume that it could at least have some positive effect. Campaigning may change the outcome of a vote, but registering a vote one's self has no significant effect, and so I choose not to vote.

Corollary To Reason IV:
Even campaigns may not be effective. For every person who campaigns for cause Y, there is a good chance another person will campaign for cause Z, where Y and Z are mutually exclusive. All else being equal, this effect cancels out the efficacy of even campaigning for others to vote, since for every people swayed to any side, the same number will on average be swayed to the opposing side.

Follow-up: Emergence
However, all else is not necessarily equal, so campaigning will sometimes be effective. For example, if those campaigning for cause Z have already swayed all the voters who are likely to be swayed by their campaigning, while many more people would still be open to the arguments of those campaigning for cause Y, then campaigning for cause Y could actually change the results of the vote.

However, there is still the question of whether the results of a vote can ever actually lead to more than moderate change either way. I submit that this is unlikely, and related to Reason III's corollary above, even if such change can occur as a result of voting, I suspect it is less likely when people are involved in voting. That is because if they continually see small changes (or for that matter, think they 'have a say in' those small changes simply by voting, even if the vote for change fails), they are less likely to be pushed over the edge necessary to seek larger changes through any means, vote or otherwise. And certainly votes that take place - not to mention those that actually win - are much more likely to involve small changes than big (since stable politics is inherently conservative, and so less radical changes are much easier to push through).

There is a much more important point to be addressed, though. At the very best, campaigning to win a vote is likely to involve changing a voter's mind on that one issue (or for that one candidate). What would be better, I think, is an attempt to change the way people think in a more general way. What I mean is that convincing a majority to vote for party X in 2004 using arguments based mainly around the 2004 vote means that when 2008 (or 2032) comes around, voters may not stick with party X. That is because they have been persuaded that in 2004 party X best represents their values, rather than being persuaded to change their values themselves to those of party X (assuming party X's values are the best ones to the campaigner in question). The same point can be generalized beyond party politics, of course.

So especially if one is seeking a radical change to the larger system (as opposed to minor or moderate reform within the system), campaigning for a given vote-result will pay off only with short-term results, not long-term ones. If a person really wants change, it seems much more effective in the long-run to worry less about campaigning others to vote their way at some particular vote (or even run of votes, given the two-party lock in the U.S. right now) and more about campaigning others to think their way, to share their values. This might in turn result in the swayed people voting the desired way in any particular vote.

However, if a person's values are even more general - say simply to think critically - then even though the result might end up with the swayed people voting against the original campaigner's opinion in any particular vote, the campaigner can at least rest easy in the knowledge that other people are more likely to vote critically and intelligently than they would have otherwise. This, in turn, seems to be the most reasonable goal for any campaigner who is humble enough to recognize the fallibility of their own opinions and would rather have the critical input of others - even if they disagree - than to have other people merely switch to their side blindly.

When you change the way a person thinks, you change the rules acting at the lowest level (individual) which dictate the results at the higher levels (society, humanity). Rather than lobbying for change from the top-down (that is, trying to sway individual votes at the societal level), I suggest it is much more worthwhile to aim for a bottom-up strategy (that is, trying to sway a critical mass of individuals to play the game differently) and count on changes to continually seep upward and naturally emerge at the top levels.

The most important vote of all is how we choose to live our own lives.
--Strange Loops