One of my more controversial views has long been that voting in general elections is pointless, perhaps even immoral. The only scenario in which a single vote can determine the outcome is one where exactly one vote separates two governing coalitions. Since that has never happened, and in fact has never come remotely close to happening, it is almost always pointless to vote.

If such a situation were somehow to arise, I would personally decide which coalition governs the country. That would be an extraordinary responsibility. I do not believe I am qualified to bear it. More than that, I doubt anyone is. No individual knows enough about the likely consequences of competing political programs to make that choice in a genuinely just and informed way. Reaching such a judgment would require years of study, and probably the coordinated efforts of multiple experts attempting to forecast the future as rigorously as possible. As an ordinary private citizen, I plainly lack that capacity. For that reason, I have never voted in a Swedish parliamentary election.

Whenever I have explained this view to friends, family, colleagues, or acquaintances, the response has usually been skeptical. Some simply insist that voting is a civic duty. Others argue that one forfeits the right to complain about society if one refuses to participate in elections. Most puzzling of all, however, is that everyone seems eager to persuade me to vote, including people whose preferred parties I would never support. A voter on the left apparently believes that a vote for Sweden’s centre-right Moderate Party is better than no vote at all, and vice versa. I have always found this difficult to understand. If you support the Left Party, you presumably think it is better for the Left Party to have more political influence rather than less. If so, then my abstaining from voting for the Moderates should be preferable to my casting a vote for them. I have pointed out this inconsistency many times. But it has never made much of an impression.

One common objection, however, has always struck me as more serious: if everyone thought like you, democracy would collapse. That is obviously true. My reply has always been that almost no one thinks like me, so democracy is under no threat. For most of my adult life, I found that response satisfactory. Recently, I have begun to doubt it.

My change of mind began with the famous philosophical thought experiment known as Newcomb’s Problem. Imagine two boxes in front of you. One is transparent and contains one thousand dollars. The other is opaque. It may be empty, or it may contain one million dollars. What is inside the opaque box has been determined by a computer that knows everything relevant about you: your psychology, your preferences, your patterns of reasoning.

You are given a choice. You may either take both boxes, or take only the opaque one. The computer made its decision before you arrived. If it predicted that you would take both boxes, it left the opaque box empty. If it predicted that you would take only the opaque box, it placed one million dollars inside. If you take both boxes, you receive one thousand dollars. If you take only the opaque box, you receive one million.

Philosophers typically distinguish between two ways of thinking about this problem. The first is causal decision theory. On this view, you should take both boxes. The computer’s decision has already been made, and your present choice cannot affect it. Either the opaque box contains one million dollars or it does not. In either case, taking both boxes leaves you better off.

The second is evidential decision theory. On this view, you should act in the way that is characteristic of people who succeed in situations like this one. The people who walk away richest are those who take only the opaque box. Therefore, that is what you should do.

Causal decision theory has always struck me as the more intuitive position. But I have come to appreciate that evidential decision theory captures something important about how people actually reason. We often make decisions by imitating successful people, especially in domains where we ourselves hope to succeed. Perhaps that way of thinking is not as irrational as it first appears. At the very least, I no longer find it obviously mistaken.

And if evidential decision theory is right, its implications are stranger than they first appear.

The reason has to do with cosmology. The universe is very large. According to some interpretations of modern cosmology, it may even be infinite. If that is true, then somewhere in the universe there are beings who are effectively identical to me. They share my psychology, my preferences, and my way of reasoning. When I make a choice, I therefore gain evidence about what they would choose in the same circumstances. If I prefer vanilla ice cream to chocolate, so do they. If I vote for the Moderate Party, so do they.

A related idea was recently discussed by Will MacAskill on the 80,000 Hours Podcast. He suggested that evidential decision theory may help explain why sufficiently advanced artificial intelligences could cooperate on the provision of moral public goods. If one AI recognizes that its own decision is evidence of what other structurally similar AIs will decide, then cooperation may become rational even without direct coordination. Each agent treats its own choice as informative about the choices of its counterparts.

The same logic applies here. My vote is not merely my vote. It is evidence of how countless sufficiently similar minds would act in comparable circumstances.

Indeed, the universe may be large enough that entire civilizations exist composed of people almost exactly like me. They would share my dispositions and reason in much the same way I do. In a causal sense, I have no influence over what they do. But from an evidential perspective, I can affect their behavior by changing my own.

That means the familiar objection to my abstention was more powerful than I realized. When people said, “If everyone thought like you, there would be no democracy,” I was too quick to dismiss them. In an important sense, that is exactly my situation. There are places where no one votes because everyone reasons as I do. Those places are populated by beings who are, in the relevant sense, me.

If I choose to vote, I may help ensure that they do the same.

That is why, for the first time in my life, I will vote this September.