Now that Bernie Sanders has endorsed Hillary Clinton, there will be some debate on the left as to what to do. Inevitably, there will be those who are disappointed in Sanders, or who consider themselves “Bernie or Bust” camp and refuse to contemplate the idea of voting for Clinton. A serious question among leftists will arise, as it does every four years, about whether third-party candidates are principled alternatives or “spoiler” candidates.

There are good reasons why socialists and other lefties intensely dislike Hillary Clinton. Their values conflict with hers in major ways. They are disturbed by her close alignment with the financial industry, dangerous hawkishness, willingness to compromise key progressive values, reflexive defenses of Israel, support for the death penalty, and aversion to transparency. When the Democratic Party runs a candidate like Hillary Clinton, it can seem to a leftist as if there is little meaningful difference between the two parties. Republicans are the party of war, big business, and the surveillance state, whereas the Democrats are the party of drone strikes, Wall Street, and government watchlists.

At the same time, we do know that there are differences between the parties. However corporate-friendly and warlike moderate Democrats may be, Republicans are almost always more so. Appointments to the Supreme Court and the lower federal courts do actually matter; they can mean the difference between the preservation and the destruction of fundamental rights.

Socialists are therefore faced with a dilemma: do they follow Bernie’s lead, suck it up and vote for Hillary, or do they vote third-party?

In a recent essay, John Halle and Noam Chomsky explain the most sensible strategy for approaching elections, the “Lesser Evil Voting” (LEV) method. It’s not difficult: because the significance of your vote depends on where you live, where you vote should guide how you vote. In safely Democratic or Republican states, vote for the losing third-party candidate of your choice, or do not vote at all. In competitive swing states, where the votes of leftists might mean the difference between the Democrat and the Republican winning, vote for the “lesser evil” Democrat, in this case Hillary Clinton. The strategy is based on the simple principle that one should make choices based on the predictable consequences of one’s actions. If leftists’ refusal to vote for the Democrat could actually install the Republican in office, then leftists should vote for the Democrat. If there’s no way it could make a difference, then they should do as they please.

People on the left are understandably weary of the whole “Nader voters gave us Bush” argument. And it’s often used as a cudgel to convince leftists to sacrifice their values. But it is an extremely important case study. Purely in terms of consequences (as opposed to culpability), it is the case that if 500 Florida Nader voters had changed their minds in 2000, multiple thousands of Iraqi children might be alive instead of dead. Climate change policy might have been very different under a quasi-environmentalist than a former Texas oil executive. Of course, these are still mights. But when you’re voting, you go based on probabilities. And the probability is that Republicans like George W. Bush will inflict more damage on the world than Al Gore would have. It was therefore foolish for Nader to campaign in swing states, and for swing state voters to cast ballots for him. They gained nothing but personal satisfaction. 

Several objections can be made to this argument. First, one could challenge the claim that in the particular case of Clinton versus Trump, Clinton is even the lesser evil to begin with. Some, citing Clinton’s record of deadly, irresponsible hawkishness and interventionism, may argue that Clinton’s foreign policy actually offers the more serious menace to global peace. The factual point is worth debating. But if one is to argue that it should change the strategy, one must argue that Trump is, on the whole, more in keeping with left-wing values. Few leftists are likely to seriously accept this. No matter how false Clinton’s protestations of progressivism may be, it is impossible to treat a racist, totalitarian, climate change-denying billionaire as the greater friend to the international working class. Anyone who is “not afraid of Trump” is probably not afraid because they are neither a Muslim immigrant or an undocumented Mexican worker. For those groups, Trump represents by far the greater threat. Thus to be blasé about the possibility of a Trump presidency is to openly gamble with the lives of the vulnerable. Hillary Clinton’s presidency will be predictably awful. Trump’s could be better, or could be much, much worse. It’s precisely the uncertainty about Trump’s presidency that is so worrying, the risk is one that the country cannot afford to take.

A second objection reflects a common belief people have about voting: it suggests that one should always vote in accordance with the candidate that most reflects one’s values. It’s a very powerful inclination, felt deeply by those who believe that one must stick by what one believes, no matter the cost. But this position excessively romanticizes the act of voting itself, and treats casting a ballot as more significant than it actually is. It sees elections as being a key way in which one expresses one’s personal moral identity, instead of simply being a means to a possibly marginally better set of outcomes. What matters is not who you vote for, what matters is what happens in the world as a result. 

To adopt this conception is to maintain the idea that by voting for something one has fully endorsed it, and that by not voting for it one has been absolved of responsibility for it. This supposition cannot be justified; by abstaining from participating in a decision one can nevertheless be fully responsible for it, if one’s refusal is what has caused the decision to occur. When it comes to voting, to consider anything other than consequences is to buy into the American idea that voting is some crucial mystical enactment of our civic identities.

Voting for a lesser evil is often seen as sacrificing “principle” for the sake of “pragmatism.” But actually, it’s not sacrificing principle at all. It’s a very principled decision to think in terms of moral consequences. So long as you don’t consider voting as an important part of your identity (and why would it be?), you don’t compromise anything whatsoever through the exercise of strategic decision-making. Voting lesser-evil is morally acceptable not because Hillary is good (she’s horrendous), but because voting doesn’t have any moral content outside of its direct consequences.

A third criticism of lesser-evilism suggests it passively accepts the bipartisan status quo. According to this perspective, mainstream Democrats are constantly attempting to cajole leftists into voting for their “lesser evil” candidates, and so long as leftists keep obliging, they will forever be sacrificing any possibility of realizing their own agenda. If leftists are willing to fall in line behind any Democrat, however loathsome, they will eliminate their own ability to pressure the party for meaningful change. Leftists therefore need to threaten to stay home, or to vote third-party, so that centrist Democrats are forced to make concessions.

This objection accepts the position that voting should be strategic. But it is mistaken, in that it views “voting third-party” as necessarily advancing left-wing political goals. Here’s the important thing to remember about American elections: you either win them or you lose them. If Jill Stein gets 3% of the vote, she does not get to control 3% of the Executive Branch. She gets to control precisely the same amount as she does now: none of it. Unless there is a plausible world in which a third-party candidate could win the electoral college, no number of socialists voting for a third-party candidate will produce a useful electoral outcome. There are simply not enough socialists. Voting for a third-party presidential candidate must therefore either (1) be purely symbolic or (2) increase the likelihood of achieving left-wing outcomes even while losing.

Pure symbolism is a poor reason on its own. It’s hard to defend risking the well-being of every Muslim American for the symbolic value of a slightly-less-of-a-landslide loss for the Green Party. The only remaining theory, then, is that voting third-party helps left-wing political goals in some other, non-symbolic way. But it’s not clear how it does so.

One critic of Halle and Chomsky says that to vote Democratic is to avoid the “demanding process of building an autonomous working class movement” and to accept the moderates’ constant insistence that “the time is not right for revolution.” It’s certainly correct that voting Democratic doesn’t make a revolution. But that’s not because of the particular choice of vote. It’s because American presidential elections aren’t the arenas in which revolutions occur. They’re little more than time-consuming media spectacles, which people should pay far less attention to for far shorter periods of time. By all means, build a movement. But as should be obvious, it’s not going to come in the form of “refusing to vote for Hillary Clinton.” It’s very strange to see “revolutionaries” who believe that voting third-party in a swing state will somehow help to catalyze a social revolution. 

The other strategic theory is that by threatening to withhold one’s vote, one brings pressure on the Democrats to become more progressive. But voting third-party in general elections does not exact concessions from the mainstream Democratic Party. We saw that plainly in 2000. Ralph Nader’s run neither dragged the Democratic party leftward, nor caused Democratic centrists to believe they needed to be more accommodating of their leftist brothers and sisters. Instead, it simply enraged them, and resulted in major backlash. If a third-party challenge, or widespread left-wing refusal to participate, did create a Trump presidency, it is unlikely that Democrats would respond by becoming more generous to their progressive wing in the future. As in 2000, the charge will simply do damage to the left’s ability to build support.

In the case of this particular election, one can even imagine this strategy backfiring even further, and causing the Democratic Party to become more conservative: If Clinton seems to be losing votes on the left, she might simply spend more energy going after Republican voters. Courting conservatives has always been a core element of Clintonism (she already has the endorsement of seemingly half of the Bush Administration), and there is no reason to expect that a Clinton looking for votes would move left rather than right.

There are some ways in which electoral politics can be effective for leftists to advance their agenda. Primary challengers can bring pressure on a party, as the Tea Party ably demonstrated to the Republicans. And Bernie Sanders has had great success in using the leverage granted by primary victories to secure concessions from the party (though these are mostly in the form of non-binding promises, which Clinton could simply ignore after her inauguration). Elections can therefore provide strategic opportunities. But it’s hard to construct a theory for how general election voting will help the left, and particularly how refusing to vote for Clinton in Ohio will do that. 

Rather, the left should stop trying to assert itself with presidential voting choices, and start trying to build political muscle at the state and local level. Energy spent on participating in the national electoral extravaganza imposes a significant cost on one’s ability to build meaningful political opposition. The left should devote the minimum of time necessary to exercise the LEV choice, then immediately return to pursuing its actual goals. Much more effort should expended in developing political organizations, building left media, engaging in strategic forms of protest, and running for office in winnable races.

It’s no secret that I am not particularly fond of the Clintons. I have, after all, written a book about Bill Clinton called Superpredator (now available in both paperback and Kindle editions). I might, therefore, be expected to be sympathetic to the argument against voting for Hillary Clinton. But I’m not in the least. That’s because I don’t understand anyone who sees voting as the route to revolution. It’s precisely because of my left-wing politics that I don’t see voting as anything more than a very limited, purely strategic choice, which should be made solely on the basis of how likely it is to advance my favored political causes. It will be easier to implement my progressive values under a Clinton than under a hare-brained, unpredictable thuggish racist who admires Saddam Hussein. And that’s all that matters, as far as voting is concerned.

Of course, it’s not yet certain whether any of this will matter. It’s unclear how many people are actually in the “Bernie or Bust” camp at this point, and the whole thing may well erode now that the man himself is campaigning alongside Clinton. Plus, Donald Trump seems to have decided he doesn’t feel like being president, and to be engaged in an intentional effort to sabotage his own chances. If the electoral map keeps looking like Nate Silver predicts it will (though God knows he’s not terribly reliable), then nobody need worry about the handful of intransigents who feel that casting a lesser-evil ballot in a presidential race is some kind of betrayal of the revolution. But if Trump rebounds, and there are reasons why he may well do, then the responsible way to vote is obvious. Wherever it doesn’t matter (which is most places), vote however you like. Wherever it might possibly matter, just vote for Clinton. But do so without the illusion that how you vote matters especially much. The revolution is to be made elsewhere.