Sunday, May 24, 2009

Richard Dawkins is wrong

I've been reading Richard Dawkins's The Selfish Gene, and I've come across a passage that betrays a very common misconception in microeconomics. Chapter 8, "Battle of the Generations," talks about the evolutionary pressures promoting kin altruism, or behaviors that lead animals to aid others that are likely to carry the same genes as they do (i.e. their kin.) Dawkins talks about favoritism, or a genetic tendency which leads an animal to favor one of its offspring over another. He has just pointed out that some animals neglect or kill runts in order to devote more resources to other members of a litter. There are situations in which sacrificing one offspring will ultimately lead to more surviving grandchildren, which is in a sense the ultimate goal of reproduction. After bringing up the case of runts, Dawkins continues:

"We can make some more general predictions about how a mother's tendency to invest in a child might be affected by his age. If she has a straight choice between saving the life of one child or saving the life of another, and if the one she does not save is bound to die, she should prefer the older one. This is because she stands to lose a higher proportion of her life's parental investment if he dies than if his little brother dies. Perhaps a better way to put this is that if she saves the little brother she will still have to invest some costly resources in him just to get him up to the age of the big brother." -- The Selfish Gene, p. 125
It should be noted that "Parental Investment" or P.I. is a hypothetical resource used in abstract discussions of of the behavior of parent animals. It refers to all the resources an animal has at its disposal to aid the survival of a member of it species, be it a child, another relative, or itself.

The whole discussion relies on talk of what a mother "should" do, which is a shorthand for describing which behavior will lead to more of her genes living on in the bodies of her offspring. There is, of course, no moral judgment made anywhere in this discussion. It is merely a description of the mechanism of natural selection.


The problem with Dawkins's explanation lies within the last two sentences:
"This is because she stands to lose a higher proportion of her life's parental investment if he dies than if his little brother dies. Perhaps a better way to put this is that if she saves the little brother she will still have to invest some costly resources in him just to get him up to the age of the big brother." These two statements are not equivalent, and the first one is flatly incorrect.

The amount of P.I. previously invested is
what economists refer to as "sunk costs." There is nothing a mother animal can do to retrieve the calories, nutrients, time and other resources she has expended in raising her children up the day on which we find her.
In making investment decisions, people often consider the investments they have already made in the past, which is irrational.

If you've already invested $100,000 in a factory, but no level of output from that factory will allow you to meet your break-even point, you are best served by shutting down the factory. You surely want to recoup your investment on the factory, but if you will lose money by producing things at there, then opening the factory is not the way to go about it.
Dawkins is talking about genes, which are perfectly rational in the sense that they do not make conscious decisions. Genes exist in the population based on how their expression affected the reproduction of previous generations. Therefore past investment is not a factor in the "decision" made by the genes in his scenario. The tableau constructed in the quoted paragraph presents two "choices":

A.) Save the older child, leading to the death of the younger.
B.) Save the younger child, leading to the death of the older.

The prior levels of investment of P.I. have absolutely no bearing on whether a gene happens to promote option A or option B. Therefore past investment is not a consideration. To illustrate this further, let's instead assume that the older child is not the offspring of the animal in question but instead its sibling, which also shares 50 percent of its genes with the decision-maker. Let's assume, though, that that sibling has entailed zero previous investment by the decision-maker, but that it will be under the care of the decision-maker from now on. In this case, it is still the right choice, genetically speaking, to save the older child, for the reason Dawkins explained in the final sentence of the paragraph.

What matters is the future investment necessary to see the child pass on its genes, half of which are the same genes as those in the adult decision-maker's body. The decision that leads to that outcome with the least expenditure from the decision-maker going forward is the one which will be adaptive, all else equal.Under my scenario, saving your brother is more adaptive than saving one's son, if the brother is significantly older than the son.

UPDATE: Having read further in Dawkins, he points out precisely this fallacy in the next chapter. He continues to make statements about the waste of past investment, but he does say explicitly that, strictly speaking, past investment doesn't effect future decisions. In fact, he published a paper on this theme in response to the work by R.L. Trivers, who did the pioneering work on this subject and apparently made this mistake.

No comments: