The issue of altruism has long tickled our collective fancy. Our civilization admires acts of altruism, even when we as individuals rarely consider them in our best personal interest. Hence they remain exotic to our behavior. Those two facts are congruent; if it were commonplace it would not be admired.

So the question is: what causes altruism?

A completely "pure" sort of altruism, in the sense that the act you perform to help someone has no conceivable benefit to you, seems plain absurd. We just don't do anything for no reason, everything we do is motivated by something, a root cause of some kind. And if we have no relation to that cause, then it's absurd to undertake this action. As absurd as it would be to do anything that isn't motivated.

I suspect the reason why we idolize this notion is precisely because it makes no sense. It's a fantasy, and people like to indulge in fantasies that give you something for nothing.

The biological perspective

So apparently Darwin had trouble explaining altruism in terms of his theory of evolution, whose slogan is survival of the fittest. If an individual organism is fitter than another, why should the fitter help the less fit, who is also his competitor, when his self interest is better served by letting the competitor perish? It turns out that Darwin is not the last biologist, and others have since added to his work.

William Hamilton suggested that survival of the fittest was the right slogan, but with the wrong meaning. It wasn't exactly individuals that were competing, it was genes. And it wasn't so much about survival of individual as it was about survival of genes, which means the individual could be sacrificed as long as the gene survived (reproduction). Now, Hamilton wasn't a crackpot outlier, Richard Dawkins's The Selfish Gene apparently explains all about this, and other biologists have bought into this view as well.

So if selection is driven by genes, we're playing a different game. Hamilton explained altruism in terms of kin selection. That is, an individual will share 50% of its genes with its parents and siblings. Subsequently, it will share 25% with its grandparents and cousins, and so on. Therefore, the genetic material found in the individual's kin is going to be a lot more valuable than that in a completely unrelated individual.

Hamilton's rule is an equation that captures this insight. The equation (or inequation, to be precise) describes the balance in an altruistic act.

C < rB,

C : cost to the individual
r : genetic relatedness
B : benefit to the other individual

So if my cost is 0.25 (likelihood of survival when doing this), the relatedness is 0.5 (a sibling or parent) and the benefit is 1.0 (save the individual's life), then 0.25 < 0.5 x 1.0 and I'm likely to do this.

Notice that is r is zero, ie. this person is a stranger, then I will not lift a finger.

Hamilton's rule has never been proven, but it formalizes an intuition that we already have when we perform acts of altruism: there is an equation of some kind. It *is* a calculation we perform when deciding whether or not to help someone. When people are asked for a favor, the first thing they always say is "what is the favor?". We need to know, because we'll factor that into our calculation.

And the reason we calculate is obvious: it costs something. If you could help without giving up anything then you'd have no reason to refuse. But the only person we know who has that opportunity is God (and even he doesn't do it).

It also explains why acts of altruism that incur a high cost to the individual, with a low benefit, are not common.

Altruism is egoism

It's not really important that Hamilton's rule applies at a genetic level while our calculations are cognitive. What's important is that there is an equation that relates our cost against another's benefit, but not without some sort of relatedness factor that accounts for our benefit out of the deal. And this is very obvious, because if C < B is all that describes the relation, then every time we have a chance to help someone, and it costs us less than it benefits the other person, we would do it. Everyone would constantly go around helping other people and altruism would be as unique as oxygen.

The implication here is that there is no "pure" altruism. Everything we do is motivated by some kind of self interest. And the self interest is captured in the relatedness factor one way or another. In Hamilton's rule this describes the kin relationship between individuals, which works quite well in our world too. Family members help each other all the time.

But it can be any other reason. Even if you derive no material benefit from helping, you might still do it because you like this person and it matters to you that they get help (self interest), or because you expect them to reciprocate when the time comes (self interest), because you don't like to see people suffer (compassion, which is a personal need you have to help, again self interest), because you believe in a certain cause (which could be plain self interest, eg. global warming), because you care about society (again, self interest, what kind of society do I want to live in?), or for the rather more vague reason that it makes you feel good (self interest). Obviously, several factors can be present at once, which complicates the calculation.

Wikipedia has this sober debunker:

According to psychological egoism, while people can exhibit altruistic behavior, they cannot have altruistic motivations. Psychological egoists would say that while they might very well spend their lives benefitting others with no material benefit (or a material net loss) to themselves, their most basic motive for doing so is always to further their own interests. For example, it would be alleged that the foundational motive behind a person acting this way is to advance their own psychological well-being ("good feelings").

Not surprisingly, people are more eager to act when the equation is tilted heavily to the right. If you can give someone 10 bucks that will cause a benefit of 100 bucks (because you have 10x the buying power), that seems like a great deal. And while we don't get something for nothing, we like great deals like this. But again, if we had the opportunity to give out 10 bucks to benefit someone 100, and there was no self interest involved, nothing would stop us from going broke doing this.

A more meaningful definition altruism

People like to throw around slogans and one popular one that gets around is this one.

The true measure of a man is how he treats someone who can do him absolutely no good.
- Samuel Johnson

That's pretty much the canonical explanation of altruism the way we commonly use the term. Or is it? Well, according to Hamilton, in order to have altruism it has to be an act that costs something. And "how he treats someone" does not imply that. I can treat someone well without incurring cost by any measure of cost that I can come up with. How much does it cost to smile at someone as opposed to not doing it?

Of course, we can take this further and assume that Johnson was talking about acts of altruism. First question: how do we define cost? Is someone who is more reluctant to do it incurring more costs (in terms of overcoming their own skepticism) than someone who's naturally inclined to help people? For that matter, cost may be expressed in terms of risk. Conservative people would determine the cost higher ("this will probably cost more than we think") than people who are willing to take risks ("I'm sure it's nothing to worry about").

Does greater cost deserve more credit? What about cost-to-benefit ratio, should I get more credit for doing something if I'm not getting that great deal? What if I actually incur greater cost than the benefit is, am I good or bad? If I accomplish something small at great cost, that's a big waste, because I could have potentially helped someone else to greater benefit. It seems obvious, at least, that benefit-to-self-benefit is a clear case, if I'm doing something to help myself and only incidentally helping someone else, I won't get much credit.

All of these questions certainly come up when we assign credit for acts of altruism. But the pressing question is: is there some definition of altruism that agrees with our intuitive one? In other words, are there some acts that are "sufficiently selfless" to qualify as altruistic, to not be tainted by self interest? If we did that, we would move away from this misconception of selflessness and derive altruism as a personality characteristic of a person who appreciates certain values we like: personal freedom, free choice, compassion etc. And I think that is exactly what Johnson meant, and what we mean.