Monday, April 07, 2008

What Is 'Self Interest'?

I still have a hard time listening to economists talk about anything because they always seem to start with an assumption that, to me, makes absolutely no sense.

They say that we are all ‘self-interested’ (or ‘selfish’, in a sense). And that all of economics is built on the behavior of individuals who are assumed to be self-interested. Somehow, these self-interested individuals, in the right economic and political candidate, can find incentive to act in self-interested ways that also benefit others. This is the economic ‘invisible hand’ of Adam Smith. The truth of this assumption is borne out by the success of economics – by the way that economists are able to explain and predict behavior.

My objection is not that the assumption that humans are ‘self-interested’ is false. It’s that the assumption is ambiguous and confused. Economists who use this assumption are saying a bunch of contradictory things, and they pick out the parts of the contradiction that serves their purpose at the moment. When their purpose changes, so does their concept of ‘self-interest’.

In an earlier post I distinguished between two types of self-interest; interest of the self, and interest in the self. All of my desires are mine. They are ‘interests of the self’ because my desires belong to me. They are encoded into my brain which is the only brain (given current technology) directly connected to my muscles. Of course I always act on my ‘interests of the self’ – because my ‘interests of the self’ are the only interests connected to my muscles in the right way.

When we talk about interests of the south, we have no difficulty explaining how a person might be altruistic or concerned with the welfare of others. An agent can simply have an interest in a state in which other people are better off, or an interest in avoiding a state in which other people are worse off. These are still ‘interests of the self’ in the sense that they are interests of the agent. However, they are not ‘interests in the self’. Instead, they are interests in the well-being of others.

When economists speak about self-interests, and claim that it creates some special problem for altruism, they need to be talking about ‘interests in the self’. These are a special subset of potential ‘interests of the self’ – they are ‘interests of the self’ that also take the self as an object. More specifically, they are ‘interests of the self’ in providing the self with a benefit of some sort.

If it were the case that all human interests are ‘interests in the self’, then this would be a problem for altruism, and we may need to do some fancy theorizing to explain how, if all human interests are ‘interests in the self’, we could possibly drive people to behave in ways that benefit others. However, we have no reason to accept this assumption that all human interests are ‘interests in the self’ is true. So, we have no particular reason to think that there is a problem here in need of solving, let alone reason to examine whether various theories have solved it.

Evolution has a number of reasons to favor ‘interests of the self’ over ‘interests in the self’. The latter simply require a lot more work to program into a living system. My favorite example concerns the relationship between antelope and lions. The antelope does not run from the lion because he is afraid of being killed and eaten. It runs from the lians because he is afraid of lions. It is just so much easier to program into an antelope brain, ‘if you see a lion, or anything that might be a lion, run’ than it is to program ‘if you see something that might eat you, run’.

Indeed, if we were going to program a robot to survive in a hostile climate, the vast majority of that programming will have nothing to do with ‘self-interest’. The robot will be programmed to avoid going off of a cliff without spending an iota of processing capability calculating the effects of falling off the cliff. The latter would require the use of resources that the agent can better allocate to other items.

The agent is going to eat the plants that taste good to it, without regard as to why some plants taste better than others. The agent is going to engage in sex, not because it has a desire to procreate and some complex set of beliefs that link sex to the fact of procreation. The agent is going to avoid the fire because the fire makes it feel uncomfortable, not because it has made some complex calculation saying that the source of pain might also produce some other harmful side effects that are best avoided.

This is not to say that we do not contain any interests in the self. There are certainly interests of the self that take the self as an object. The desire to avoid pain is most strongly realized as a desire to avoid my pain. Hunger, thirst, and the desire for sex are realized in the form of a desire ‘that I eat’ or ‘that I drink’ or ‘that I have sex’

Even ‘interests of the self’ that take the well-being of others as their object can be selfish in some important way. For example, the desire for the well-being of one’s children is often realized as a partially selfish desire for the well-being of one’s own children; or, more precisely, a desire that my children are healthy and happy.

Still, we are dealing with a set of ‘desires of the self’ that includes desires for the well-being of other people. Within the context of desires of the self there altruism is not a problem. Altruism consists of desires of the self in the well-being of others. Just as we have some desires of the self that are desires in the well-being of ourselves, we can have desires of the self that are desires in the well-being of others.

We have reason to promote these desires. We not only have desires-in-the-self reason to promote desires in others to do no harm to self and to provide benefit to self, we also have desires-of-the-self-in-the-well-being-of-others reason to promote desires in others do no harm to the people we care about and to provide benefits to them when they are in need.

That is to say, to the degree that desires are malleable, we have reason to promote desires-of-the-self in the well-being of others, since we (and those we care about) are the ‘others’ who benefit when we promote ‘desires-of-the-self’ in the well-being of others.

One special problem with the idea of desires-in-the-self has to do with what counts as a benefit. What does it really mean to be selfish?

This problem comes about because the only form of value that actually exists in the real world is that of relationships between states of affairs and desires. “The self” can be made better off only to the degree that “the self” can be put in a state that better fulfills the most and strongest desires of the agent.

This invites us to ask, “What are the most and strongest desires of the agent?”

Let’s assume that an agent wants to be somebody who takes care of sick children. He invests a great deal of time and effort into making himself somebody who is better able to take care of sick children. To the degree that he does so, he is making himself into a person that best fulfills the more and stronger of that agent’s desires. This desire ‘that I be the best pediatrician available’ is not only a desire of the self, it is a desire in the self. Yet, even this type of ‘interest in the self’ desire is fully compatible with altruism.

So, where do we get this idea that humans are confined to a set of interests that are incompatible with altruism . . . so much so that economists (and others) think that there is a huge problem with respect to altruism that we need to solve?

Just as it is not necessarily the case that all interests in the self are bad, it is also not the case that all interests of the self in others are good. A person who has a desire to kill others, or a desire to see them suffer, also has an interest of the self in other people. It just happens to be an interest in thwarting the desires of others, rather than an interest in fulfilling the desires of others.

So, the problem goes both ways. We try to distinguish between good and evil in terms of ‘desires of the self in the self’ and ‘desires of the self in others’. But there are a lot of desires of the self in the self that can be good, and desires of the self in others that can be bad. This is just not a good distinction to use when trying to distinguish between what is good (what we have reason to promote) and what is bad (what we have reason to inhibit).

It is true that humans do not have the strength of ‘desire that the desires of others are fulfilled’ or even ‘desires that tend to fulfill the desires of others’ that we have reason to cause them to have. It is true that we have reason to promote more and stronger desires that tend to fulfill the desires of others. However, the problem is not one that can be reduced to a type of ‘self interest’ that is incompatible with altruism. The problem, instead, is one that can be reduced to the fact that people have desires that tend to thwart other desires – self-regarding interests and other-regarding interests that we have reason that do harm.

No comments: