In the climactic scene of Star Trek II: The Wrath of Khan, a dying Spock explains to Captain Kirk why he sacrificed himself to save the Enterprise. "The needs of the many," he famously quips, "outweigh the needs of the few, or of the one." Many libertarians hate that scene. Spock, you commie! Don’t you know that interpersonal utility comparisons are impossible? You’re from the future! This, I think, displays a fundamental confusion; Spock is not making an interpersonal utility comparison. Spock is dealing entirely with his own value scale. He is saying that, to him, the lives of his friends and crewmates are more important than his own. As such, he is not behaving "governmentally" at all; it becomes a noble act, not a tyrannical one.
The distinction, of course, is one of ownership. Spock is (at least presumably) a self-owner, who can elect to use or dispose of himself in any way he sees fit. For him to sacrifice himself in this manner is not in any way a violation of anyone’s rights. In contrast, we would view the same scene as monstrous if Spock were to order Chekhov to make the sacrifice instead — well, okay, maybe not if it were Chekhov, but like Scotty or McCoy or somebody we care about. You get it. But what’s the difference, really? In both cases, one man dies, and everybody else lives. Why is one noble and the other vile?
In economics, we have the term pareto superior to describe an outcome in which at least one involved party benefits, and in which no involved parties lose. The flipside sitation is called pareto inferior; that describes an outcome in which we cannot say for certain that anyone benefits, or in which we cannot say for certain that nobody loses. Spock sacrificing himself, perhaps counterintuitively, is pareto superior; we know that Spock prefers this outcome because he himself chose it. In fact, the majority of purely voluntary interactions are pareto superior (at least ex ante) quite by definition; since nobody ever voluntarily chooses something other than his choice, pareto inferior outcomes occur only through error, confusion (deliberate or otherwise), or coercion.
Which brings us to poor Chekhov. If Spock orders Chekhov to go to his death for the greater good, the situation becomes pareto inferior. Suddenly, there’s one involved party — Chekhov — who may not be better off. It’s simple enough just to say that, hey, Chekhov would have died anyhow, so no big deal; is that really the truth of the situation, though? Perhaps Chekhov was working on a brilliant plan that could have saved the Enterprise with nobody dying, but his plan was interrupted when Spock glommed onto him and shoved him into the radioactive death room. Or maybe Chekhov was a horrible misanthrope, and he’d have preferred to spend his last few minutes of life cackling madly and groping Uhura rather than saving the ship. We could certainly argue that his priorities are out of whack in this case, but we couldn’t really say that he himself was not made worse off by compelling him to do something other than what he would have chosen.
The other tine of this fork is that most ephemeral of concepts: utility. "Utility" is a term economists use to describe the overall value of a thing; if somebody benefits from an event, we say that he has "gained utility." Speaking theoretically, a more preferable outcome is said to be of "higher utility" than a less preferable outcome. When individuals make decisions, they implicitly rank their preferences in order of expected utility, and act to fulfil that end they believe will generate the most utility. Obviously, nobody acts in quite so mechanistic a fashion — we don’t carry around little pocket whiteboards and constantly update the rankings — but the principle is the same.
The catch is that utility is an ordinal, not a cardinal, quantity. What that means is that, while it’s possible to identify a clear preference — for example, I could say that I like baseball more than football — there is no way to assign any exact value to this preference. I can’t say that I like baseball exactly 4.381 times as much as I like football. There’s no way to take any such measurement. All we can know for sure is that I like it more, or possibly some vague guidepost like "a lot more" or "a little more." This lack of cardinality throws a wrench in the typical utilitarian dream of "social utility" — it’s impossible to say whether satisfying your third-most-valued end or my third-most-valued end adds the most utility overall.
Which brings us back to where we started. The only way we can conclusively say that an outcome increased overall social utility is when that outcome was pareto superior. If we can say that nobody was left worse off, and we know that at least one person was left better off, then it must follow that overall utility was increased (relative to what it would have been otherwise, not necessarily relative to what it was before). If the possibility exists that one person’s utility increased and another person’s utility decreased, however, then we can never know.
This means that, if one wishes to maximize overall utility, one should never favor government action of any sort, because government is the one institution in society that is positively incapable of generating pareto superior results. Why is this so? Because any time the government acts, at least one person is being coerced. Such is the nature of government. The money it spends it does not earn in the market; it takes it by force. The rules it makes are not voluntary contracts people can agree to or not; they are enforced at gunpoint. No matter what government does, somebody is being forced to go along with it instead of what he’d otherwise prefer to do. A modern state is sophisticated enough, of course, to obfuscate this simple fact; the costs are spread around and the benefits concentrated, making it less obvious, but no less true.
This is why the idea of relying upon government to optimize social welfare is, to borrow an expression from noted socialist ratbag Jeremy Bentham, nonsense upon stilts. Of all the institutions in all of society, government is the only one that positively can’t ever be said to have increased social utility. If you were getting together with your neighbors to hire somebody to keep everybody’s lawn maintained, would you hire the one guy without a lawnmower? Your blind friend might be a great guy, but do you ask for his advice when coördinating your living room set? If you need a candidate who can beat Donald Trump, do you pick the least electable person in the world? Wait, scratch that last one.
Spock had the right of it: if you want to help other people, you do it with your own property. That way you can be sure that everybody gains. The rest of the Enterprise crew understood this, too; they put themselves on the line for Spock in Star Trek III: The Search for Spock. And let us not forget Captain Kirk’s immortal rejoinder when Spock questioned the wisdom of such an act: "the needs of the one outweigh the needs of the many." Only when the many all participate voluntarily does that make sense.