EconLog | Bryan Caplan | The Decline of the Rabbit Strategy:Point well taken about the small sample size of mothers with 8+ children, but surely the GSS provides variances for that data so that we could check statistical significance. (I don't know how to use the GSS interface efficiently, and I'm not going to take the time to learn now.)From chapter 5 of the first draft of Selfish Reasons to Have More Kids:The facts at the end are from the General Social Survey. If you look carefully at the thirty-something moms, you might notice that compared to married moms, those who never married are equally likely to have 7 kids, and two-and-half times as likely to have eight or more. However, families with more than six kids are so rare (1% of the married moms and 1.6% of never-married moms) that these ratios mean next to nothing. Could availability bias plus the extreme behavior of the tail of the distribution account for the popular stereotype of the welfare mom with an army of kids?
... You don't hear the [breeding like a] rabbit analogy much anymore. In part, it sounds faintly racist. But the main reason we stopped comparing people to rabbits is that - at least in developed countries - we rarely get the chance. People today have too much foresight to breed like rabbits. Sure, many babies remain the fruit of impulsive, unprotected sex. Yet these babies rarely have a lot of siblings. If you look at thirty-something American moms who never married, 45% have just one child, and 26% have two; married moms in their thirties, in contrast, are much more likely to have two kids (41%) rather than one (22%). Almost no one nowadays sticks with a rabbit strategy after they have one or two unwanted children.
Another point: doing some literal back-of-the-envelope calculations, I figure that even if only 1% of mothers have 8 or more kids, they would still be responsible for about 4.5% of total births. Put another way, if only 1 in 100 mothers you meet have had 8 or more kids, 1 in 22 people you meet have 7 or more siblings.
This may be affecting our cognitive biases as well.