A student recently brought up an interesting Human Nature article by Catherine Salmon comparing middle-born children with firsts and lasts. She found that middle-born college students were less oriented toward family, and more oriented toward peers.
This small study suggests a partial solution to a puzzle in birth-order studies. I follow Frank Sulloway's argument that birth order has a big effect. He argues that siblings are in a Darwinian competition for parental attention. First-borns get first choice, and they usually choose to be parent-oriented. The middles get next choice, which might be to pick up a secondary parental interest, or another niche not occupied by the first. Last-borns are "born to rebel" because by their turn all the good parentlike niches have been taken; they are open to experience and are most likely to try new things.
I have to take seriously the challenge that Judith Harris raises in The Nurture Assumption that birth order does not have a big effect because kids are more peer-oriented than parent-oriented anyway.
Both Sulloway and Harris have empirical support for their positions, especially Sulloway's massive historical studies. The way I had reconciled their contradictory findings is to say that parents have a strong direct effect on children when they are little, which varies by birth order. Peers have a stronger influence during adolescence. Parents get the last word, though, because they have an indirect influence on which peers their children hang out with -- an effect that varies by birth order.
Salmon's finding adds another piece to the puzzle. If middle-borns show a stronger peer orientation, this would mitigate the parent effects that Sulloway found, and bolster the peer effects that Harris found -- for some kids.