The following post grew out of conversations on the fringes of the Thursday meeting. Since it has little to do with the main event (Japan-China), I'm putting it up here for your perusal. This looks like a subject ripe for a Freakonomics makeover.
Speaking of books, Take a look at the Lou Dobbs 1 video clip on the Daily Show page at the Comedy Central website. Jon Stewart asks Lou Dobbs why the American middle class isn't revolting, when the Hungarian prime minister says he lied, the people take to the streets... That's obviously The J-Curve argument. He's a quick study. Ian Bremmer should get royalties from that show.
Traditionally, infanticide and simple neglect were the two main factors that kept down the female population. (War, homicide and, more generally, Jackass mindset kept the male population in check.) They were society's collective response to population pressures, i.e. scarcity of resources per capita. Boys were preferred presumably because males had more value as manual labor (and thus represented a better form of investment for their parents).
The recent excess of male offspring in Hong Kong and elsewhere in newly emerging economies appears to be a more recent, largely technology-driven phenomenon, and likely uses abortion as the means. (Undercounting females is widely believed to be significant in China under its one-child policy.) . Technological advances allow us to identify gender at ever earlier points in gestation. Thus, rapidly developing societies are seeing their birth rates decline precipitously, even while they have not been able to adjust their traditional patriarchal values. These societies have seized on the new technologies to identify the gender of their fetuses and disproportionately abort females, to make sure that they can fulfill the culturally-ingrained need for male offspring (which these societies have not had the time to outgrow).
The technology has come to India as well, and there does not seem to be a strong taboo against abortion in Hindu society. Members of the emergent middle class (as well as significant segments of the less well off) are already making choices. But how much of the overall Indian disparity between genders is attributable to traditional, post-natal methods of population control and how much to abortion, I have no idea.
Of course there is a downside to being behind the curve on shifting demand for labor (in a generally gender-free direction) and partners (favoring the increasingly scarce women). Seeing a future oversupply of marriageable males, the rational couple will see that a daughter looks more and more like a better investment. I have looked only at the South Korean figures in the CIA Factbook, but I believe they show that this is manifesting itself as a measurable correction in the market for gender choice in offspring.