On Philosophy

August 3, 2006

Being Risk Adverse

Filed under: Ethics,General Philosophy,Self — Peter @ 12:38 am

As I mentioned last time it is often thought that we should use the expected value of our actions to determine if they are rational. Specifically we weigh each outcome of an action by its likelihood, and if the total is positive we consider the action to be a good choice. However, there are some cases where the expected value does not seem to reflect the actual decision making process (as judged by the choices made). For example if there is a high probability of some small reward and a very low probability of a great loss many people will consider this an bad choice to make, even if the expected value is positive. Similarly, when faced with two bets, one which has a high possibility of a small reward and the other which has a very low possibility of a large reward people will often choose the bet that is “more of a sure thing” even when the other bet has a greater expected value. Are these choices rational, and thus reveal that expected value doesn’t fully capture rational action, or is being risk adverse irrational?

One possible solution is to say that we should weight each outcome not by its probability but by some other factor, say its probability squared, which reduces the weight of unlikely outcomes. This would explain why people often prefer more of a “sure thing”, but then this makes the case where people avoid the very low possibly of a large loss seem even less of a rational decision. Perhaps then we can explain our behavior by assuming that people mentally discount favorable outcomes and overemphasize negative ones. We might even be able to tell an evolutionary story about this, arguing that a group whose members take the riskier choices will have a smaller outcome due to the occasional unlucky result, although the remaining members will be well off. On the other hand a group that “plays it safe” will have more of its members survive, even though on average they will be worse off than the other group, and thus in the long run will come to dominate the gene pool. This definitely provides a concrete explanation for why people are risk adverse, and to different degrees, but it also effectively denies that there is any rationality in the individual’s choice to be risk adverse, simply an innate psychological drive.

I don’t think that we should so quick to abandon rationality though. Let us set aside expected value for the moment and consider how we justify our choices in general. It has been proposed that our choices on matters ethical are guided by a desire to be able to defend our actions to other people (specifically by Scanlon). A slight modification to this idea is that we seek to act in ways that we can justify to ourselves (echoes of Korsgaard), specifically to our future selves. There is not just one future self to consider however, if an action has multiple possible outcomes we seek to be able to defend our choice to all of our future selves, equally. This is why that many of us won’t be willing to take a gamble that has the possibility of a large loss, even if there is a high probability of a small reward, because we can’t justify our choice to the future self that loses. We can justify some gambles however, especially when the loss is comparable to the rewards, or the loss is negligible, and the degree to which we feel that we can justify our gambles determines how risk adverse our choices are. Expected value then is what we use to choose between the risks that we can justify to our future selves, but considerations of justification come first, and thus explains why not all choices seem to be guided by expected value.

I admit that in some ways this might be seen as a radical stance to take about human behavior, that it is motivated more fundamentally by a desire to act in ways that can be defended, if only to ourselves, than by what will, on average, result in the most rewards. It is true that the actions of some people might seem to be motivated solely by self-interest, but I think that for these people the principle of self-interest is the only way that they can defend their actions to themselves, and that not all people are motivated in this way.


Blog at WordPress.com.