There's much more to Hamilton's rule than haplodiploidy
Updated 27 October 2010: Evidence for kin selection's role in the evolution of eusociality in insects includes ancestral-state reconstruction showing that all eusocial species are descended from ancestors with monogamous queens, which increased within-colony relatedness. See this earlier post.
A recent paper in Nature, challenging an obsolete explanation for the evolution of eusociality (characterized by nonreproductive individuals, like worker bees), may be misinterpreted as evidence against Hamilton's rule.
Hamilton's rule, you may recall, is that alleles (gene variants) for altruistic behavior (increasing another's fitness in ways that reduce one's own) will spread if, on average:
c < b r
that is, if the fitness cost to the altruist is less than the fitness benefit to the recipient, times the extent to which the recipient is more likely (relative to the population with which they compete) to have the same altruism allele. Often, this increased likelihood of sharing the same gene is due to genetic relatedness.
The haplodiploid genetic system used by ants and bees means that a worker can be more likely to share alleles with her sisters (r=0.75, relative to unrelated individuals) than with her own offspring (r=0.5). It was once suggested that this could explain why worker ants care for their sisters (the queen's daughters) rather than having offspring of their own. But then someone pointed out that Hamilton's r between a sister and a brother is only 0.25. So, if the queen has equal numbers of sons and daughters, there's no reason for workers to favor siblings over their own offspring. The central point in this new paper, that haplodiploidy doesn't necessarily lead to eusociality, is therefore both consistent with Hamilton's rule and old news.
I'm not sure how long we've known this, but The Selfish Gene, published more than 30 years ago, mentions the low Hamilton's r of 0.25 for brothers and focuses, not on any link between haplodiploidy and eusociality, but rather on conflict between workers and queen over the sex ratio of the latter's offspring. In particular, Dawkins discussed work by Trivers and Hare, who used Hamilton's rule to predict a 3:1 female:male ratio in most ants but a 1:1 ratio in slave-making species, predictions that were confirmed by field observations.
Carl ZImmer's discussion of this paper in the New York Times includes comments from Andy Gardner, a leading evolutionary theorist:
"This is a really terrible article." One problem Dr. Gardner points to is the Harvard team's claim that the past 40 years of research on inclusive fitness has yielded nothing but "hypothetical explanations."
"This claim is just patently wrong," Dr. Gardner said. He points to the question of how many sons and daughters mothers produce among the many insights inclusive fitness has brought.
Jerry Coyne (author of Why Evolution is True) didn't like the paper either.
Nor did Richard Dawkins.
Or Jeremy Joder.