Incidentally, this would also explain why people falling in love tend to be intensly happy at first. In order to get and keep a mate, you need to be ready to take risks. Also, if happiness is correlated with resources, then being happy signals having lots of resources, increasing your prospective mate's chances of accepting you.
We can actually find at least three levels of granularity at which evolution might optimize the level of risk aversion that organisms exhibit. Take a species of organism which will end up in various kinds of environments. Some of these environments are rich in resources, and in those it pays off to take a lot of risks. Some of the environments are poor in resources, and in those it's more important to hang on to what you already have. If an organism was only found in one kind of an environment, then over time it would become genetically predisposed towards only exhibiting the level of risk aversion most optimal to that environment. But suppose that the organism was found in a varying range of environments. An individual member of the species might spend its whole life in a resource-rich environment, and its offspring in a resource-poor environment. In this case, evolution wouldn't have the time to tailor the whole species to match the environment.
In such a scenario, genes for high and low risk aversion both would be found in the population. When members of low risk aversion ended up in a resource-rich environment, they would gain a fitness advantage, only to lose it when they ended up in a resource-poor environment. Because environments vary, both as a function of time and location, neither gene variant would be strictly superior and therefore both would be maintained in the population.
But this first level of granularity is still rather coarse-grained. The risk aversion of any particular organism (and at least 50% of its offspring) is determined for life, and there is no way of revising it for exceptional circumstances. Now, evolution has a trick for situations like these. An organism can have an inborn capability for several different approaches, one of which is locked in based on the environment the organism spends its early life in. Members of the species could have a genetic predisposition towards a certain level of risk aversion, but have that be modified by the amount of resources available in the first environment encountered.
Yet even this second level of granularity determines an organism's level of risk aversion for its whole life. Even in a relatively stable environment, the organism might have good or bad luck at times. Therefore it makes sense to have part of the risk aversion level be determined by a cognitive process that runs throughout the organism's life. When it notices it's doing well, it is inspired to strive for even higher levels of accomplishment. When it notices it's doing badly, it is discouraged.
All three levels of granularity seem to be present in humans. We know that happiness has a major genetic component, and some people are just born happier than others. On the other hand, childhood experiences seem to have a major impact on how optimistically one ends up looking at the world. And of course, we all experience periodic ups and downs in our happiness throughout life.
I was previously talking with Will about the degree to which people's happiness might affect their tendency to lean towards negative or positive utilitarianism. We came to the conclusion that people who are naturally happy might favor positive utilitarianism, while naturally unhappy people might favor negative utilitarianism. If this theory of happiness is true, then that makes perfect sense: risk aversion and a desire to avoid pain corresponds to negative utilitarianism, and willingness to tolerate pain corresponds to positive utilitarianism.
Note that most Western humans have a far greater access to resources than our ancestors did, so we are likely all far more risk-averse than would be optimal given the environment.
