Whether or not a situation is just depends on whether or not the people involved get what they deserve. Justice is, obviously, a subject of great moral concern–it is of vital moral importance, for example, that undeserved suffering be minimized. The question I am interested in concerns how this moral importance works. There are (at least) two conceptions of the moral importance of justice in the literature.
The first conceives of justice as a limiting condition on how objects of value may be pursued. This view could be described as “deontological,” and we might think of it as the “orthodox” view. The defining characteristic of this view is how it deals with cases in which the best outcome can be achieved only in an unjust manner. Prominent examples include: the Organ Harvest case, according to which the only way to save the lives of five very productive, beloved people is to slaughter an innocent person in order to harvest his organs for transplant; the Punish-the-Innocent case, according to which the only way to make everyone happy is to frame and unfairly punish an innocent person while letting the guilty party go free; Free Rider cases, in which a person takes advantage of the system in a way that makes himself better off and which does not cause an offsetting harm to others, but which is unfair because the effects of everyone engaging in this behavior would be disastrous; etc.
In these cases, the fact that the gains in utility, welfare, value, or whatever, can be had only at the expense of justice seems to be sufficient to render the relevant actions morally impermissible. Although it would be “better” in a clear sense to perform the organ harvest, or punish the innocent, or free ride, these actions are morally forbidden because they are unfair. The injustice involved rules them out.
This conception can be found throughout the history of moral and political philosophy from at least the modern period forward. Kant’s Formula of Humanity, according to which we must treat humanity always as an end in itself and never as a mere means, is a clear example. Rawls’s theory of justice, according to which “each person is to have an equal right to the most extensive scheme of equal basic liberties compatible with a similar scheme of liberties for others,” regardless of the consequences, is also an instance of this conception. As different as Nozick’s political philosophy is from that of Rawls, his principles of justice in acquisition and transfer are also instances of this conception.
The second conception sees justice as an object of value in itself. On this view, states of affairs consisting of a person getting what she deserves are good, and those consisting of a person failing to get what she deserves are bad, and this is only indirectly related to the moral permissibility of actions that bring such situations about. Ross’s insistence that, in addition to pleasure and virtue, the apportionment of pleasure to the virtuous is intrinsically good is an instance of this conception of justice. This conception can also be found in Feldman’s various desert-adjusted versions of utilitarianism and hedonism, according to which the value of an episode of pleasure is determined by its intensity, duration, and the degree to which it was deserved; Portmore’s recent attempts to “consequentialize” various moral theories (that have traditionally thought to be essentially deontological); and, interestingly, in Kant’s discussion of the value of happiness in the opening pages of the Groundwork, in which he says that happiness is valuable only if it is associated with a (deserving) good will, and that the highest, most complete form of goodness obtains when the perfectly good will enjoys perfect happiness.
The first conception has the advantage of being the traditional view, and it is intuitively attractive–even intuitively obvious. But, assuming that we have a fundamental obligation to bring about all-things-considered better states of affairs rather than worse ones, and given that this obligation is not closely related to an obligation to do what would be fair, it is theoretically disunified. It contains a larger number of fundamental moral obligations, and is therefore more difficult to integrate into an overall theory of moral permissibility and obligation, than its main competitor. Such a theory would have to emphasize the competing duties to do what would be best for everyone and to do what would be fair, and introduce ways of weighing these considerations against one another. And it is not clear how this could be done in a consistent, principled, non-arbitrary way. (For example, as it is traditionally interpreted, Kant’s categorical imperative seems to entail that the impact of our actions on the welfare of others is morally irrelevant; this is implausible.)
This second conception has the advantage of being simpler and more theoretically unified than the first. It is also easier to integrate into the sort of overall ethical system I prefer (in the manner of Feldman and Portmore). But it seems to me that its biggest selling point is that there are possible cases in which utility and justice conflict, but in which the injustice is so insignificant and the gain in utility is large enough that it would be permissible–or even obligatory–to perform the unjust act.
For example, suppose that a group of friends has gone out to dinner and is now considering how to divvy up the bill. Suppose that everyone is in the same ballpark—it’s not as though everybody had the lobster and washed it down with hundred-year-old champagne while one guy had ice water with his salad, but it’s also not as though everyone had the same thing. Not everyone’s meal costs the same. The most fair way to settle up would be for everyone to pay exactly the price of his or her food and drink plus tip. That way everyone would pay exactly what he or she owes and no more and no less. But it is possible, if there were enough people, and if there were enough different items on the bill, and if the bill were disorganized enough, that this would not be the best way to settle it. It is possible that the all-things-considered best way to settle the bill would be to divide it evenly so that everyone pays the same amount, even though this would not be the most fair. And I don’t mean just the easiest way; I mean the best way. I mean that the easiness of doing it this way could outweigh the unfairness involved, and that this is why it is okay to settle the bill this way. And I think it would be permissible to divide the bill this way even if not everyone agreed. I think you could be outvoted on a decision like this, although this would not be true if the unfairness outweighed the easiness.
This makes it seem like the second conception of justice is the right one; that justice is an object of value in itself, and that although it ordinarily outweighs considerations of utility, it does not automatically trump them. But I also think it’s telling that the injustice involved in the above example is highly trivial. Minor but nontrivial injustices generally outweigh significant potential gains in utility.