We generally believe that people should only be punished for events and actions they have control over. To take a trivial but often-used example, someone who deliberately steps on your toes is deserving of moral censure, while someone who is pushed and stands on your foot accidentally has done nothing morally wrong. More seriously, an actor who fires what he thinks is a prop gun that kills somebody may be blameless whereas a person who knowingly shoots a loaded gun at someone may be guilty of murder.

However, in conflict with this, there are also situations where outcomes are determined by chance yet one outcome is judged more severely than another. Compare a drunk driver who loses control and kills someone standing by the side of the road with another drunk driver who lost control at the same place but hits nobody. The fact that the driver kills someone is taken into account in the moral judgments both of bystanders and of the driver, although it was not the driver's intention to kill and the same behaviour in other situations might cause no injury. In another much-discussed scenario, if I shoot at someone and hit them I might be charged with murder; if I miss, only with attempted murder. In these cases, people make the choice to drive while under the influence of alcohol, or choose to shoot at someone, and it is this act which is morally wrong. Yet people are punished and morally censured not merely for the decisions they make but for the consequences, even when these consequences have a large amount of luck involved.


Moral luck is the term for an action or event which is governed partly by chance but is legitimately subject to a moral judgment. If a person is praised or blamed for an event which is party due to luck, then moral luck is involved. The term moral luck was coined by the philosopher Bernard Williams in the late 1970s and the debate began with a paper by Williams and a response by Thomas Nagel.

The contrary idea, that moral judgment should only be applied in situations where people have freedom in their actions and then only to the extent that they are able to determine the outcome, is commonly called the Control Principle; it is a basic principle of much moral philosophy and goes back to Adam Smith and Immanuel Kant. If moral luck exists i.e. if there are situations where blameworthiness depends on chance, then this challenges the way we usually think about morality and specifically threatens the Control Principle; however if moral luck does not exist it is hard to justify many existing practices and intuitions about morality.

Williams used another example, which takes us away from the concept of crime, by looking at the artist Paul Gauguin. Gauguin abandoned his wife and family to go live in the South Pacific; this desertion would normally be considered immoral. However, while in the South Pacific, Gauguin produced a series of paintings which were highly influential on modern art and are still greatly acclaimed as things of beauty. Williams suggests that Gauguin's actions could be morally justified in retrospect by the greatness of his art; however, if Gauguin had not gone on to become a good painter, then his actions in abandoning his family would be immoral. There was no way for Gauguin to know what the consequence of his action would be. For Williams, this is troubling, because it seems that the chance after-effects of one's actions can determine their moral value (this is perhaps not a simple example to discuss, because it seems to challenge moral value with aesthetic value, suggesting one can overpower the other).

Types of moral luck

Nagel attempted to clarify Williams's arguments by defining four kinds of luck and asking which are used as the basis of moral judgments:

1. Resultant Luck. This is the sort of luck in all the examples described above, the drunk driver, the shooter and Gauguin. It is where chance or luck determines the effects of an action. Resultant moral luck is involved if we judge people according to the results rather than the intention.

2. Circumstantial Luck. This is based on the environment and the situation in which you find yourself. Some people will be put in situations where they make bad moral decisions (the usual example is those Germans in Nazi Germany who cooperated with Hitler), whereas others might have emigrated and avoided the choice. Circumstantial moral luck would be a factor if we judge those who collaborated with Hitler differently from Germans who had emigrated but who would have collaborated.

3. Constitutive Luck. This is the element of chance that governs the make-up of our personality. Someone with a volatile personality might get in a rage and hit somebody, while someone else with a calmer temperament would let it pass. Constitutive moral luck is a factor if we judge the volatile person who hits out differently to the calm person who keeps their cool.

4. Causal Luck. This relates to ideas about free will and determinism; if our actions are determined by external forces then any moral judgment is based on an element of chance. Even if we accept a moderate version of non-determinism in which prior circumstances affect the probabilities of actions rather than determining them entirely, then there is still an element of causal moral luck in the judging of actions. Only if you consider human actions as entirely undetermined are you free of this. However, this fourth category is sometimes considered as already covered by the previous three.

For Nagel, all four types of luck seemed to play a role in moral judgments in certain circumstances. In the following section, I outline a few of the ways philosophers have reacted to the apparent existence of moral luck.

Consequences and responses

It has been suggested by a few people, e.g. Margaret Urban Walker, that moral luck is not a problem, i.e. applying moral responsibility to matters of chance is totally legitimate. Walker argued that our responsibilities cover a wider area than those things we have control over and therefore moral judgments go beyond what we have the power to change; virtues such as dependability recognise that we sometimes have a moral responsibility to be there for people even if we cannot help them. However, most philosophers have tried to find a way to eliminate or limit the scope of moral luck rather than denying it is a problem.

The Control Principle appears central to moral philosophy, and yet when faced with the above instances of the role of luck, few people have argued that all the above types of luck and chance circumstances should be ignored (e.g. that attempted murder and actual murder should be judged the same). Most writers on the topic have approached the issue in other ways, with strategies such as: to question the definition of luck; to differentiate between various types of moral judgment thus allowing consequences to be considered in some types of judgment; or to find other reasons for differential punishments. In many cases these arguments require restricting moral judgment to only some types of luck.

The notion of luck involved in moral luck is often far from what we normally consider luck (the popular definition of luck encompasses things such as lottery draws.) Some philosophers have debated whether the wider categories described by Nagel can be called "luck" at all. For instance, Nicholas Rescher claimed that identity must precede luck; he suggested that luck involves an unplanned event with a significant good or bad result and this does not not describe the formation of personality. Thus he rejected constitutive luck as a factor in moral judgments.

It can also be argued that there is a distinction between different sorts of moral judgment, between those which judge intention and those which judge consequences, which can be seen in the many different ways that philosophers talk about morality (based on consequences, duty, virtues, or character, and other ways). Thus by some standards it is correct to judge people differently on whether or not their actions result in a death, while other moral standards depend only on the decision or action itself. Bernard Williams went further, saying that we should give up the narrow conventional notion of morality in favour of a more general concern about how lives should be lived, which he termed the "ethical"; this would allow him to subsume consequences and actions into a general ethical theory. Such approaches potentially avoid the problem of moral luck by rejecting the Control Principle.

Finally, even if we want to deny moral luck, there may be sound reasons in practice for punishing people depending on the chance outcome of their actions. One justification is what Andrew Latus called the "epistemic argument": that people's intentions can best be judged by seeing the consequences of their actions. It is hard to tell how serious somebody is about wanting to kill somebody until you see the corpse lying dead on the floor, and if we have greater certainty about the murderer's motives he should be punished more severely. This can also be applied (perhaps with greater validity) to circumstantial luck: it is almost impossible to know how somebody would have acted in Nazi Germany, so how can we possibly punish them for what they might have done? There is another practical argument for differential punishment advanced by Henning Jenson. He argued that everybody makes bad decisions and takes risks, and it is impractical to punish everyone irrespective of consequences; punishing only those acts which have consequences has "restorative value" and acts to satisfy the regret of the driver and the feelings of bystanders.

The problem of moral luck is a major one, and the above discussion only touches on a small part of the debate. In an ideal world, where everything happened exactly as we willed, there would perhaps be no space for moral luck. However in the real world, where actions often have unexpected consequences, we find our notions of moral value to be constantly challenged. It seems that we must sometimes praise or blame people for things outwith their control; the questions are to what extent, and what sense of human agency and moral responsibility can we preserve?


  • Andrew Latus. "Moral Luck". Internet Encyclopedia of Philosophy. 2001.
  • Dana S. Nelkin. "Moral Luck". Stanford Encyclopedia of Philosophy. 2004.
  • Govind. "Moral Luck". Govind's web page. 1998.