Thursday, April 03, 2008


odd man in?

I caught this very, very interesting story on NPR yesterday: "Male Birth Rate Among Asian Americans Studied".

Here's the short version: Columbia Univ. economists Douglas Almond and Lena Edlund studied the 2000 Census and found that, for Chinese, Korean and Asian Indian American parents, those with two daughters (as their first two children) were 50% more likely to have a son for their third child. This simply isn't naturally possible, suggesting that there is some kind of sex selection going on though the researchers have been very careful not to draw conclusions as to what form said selection takes since they didn't collect data on that part of the question.

Here's the personal anecdote: I've known at least two Asian American families growing up where the parents had four daughters. Not that I ever asked the parents but you just assume, in those cases, they're trying for that son and finally gave up.

Here's getting back to Almond/Edlund's study. I took a look at the summary version of their research (warning: you need to be logged into a university system to access) which appears in the Proceedings of the National Academy of Sciences. Some findings worth sharing:

  • The son-biased ratio (SBR) is apparently new - there was no evidence of such a bias in the 1990 Census. This suggests one of two things (at least): sex selection was less popular then (unlikely) or sex selection technology is more popular now (more likely).

  • The 1.5:1 SBR at third parity (i.e. after two daughters) is actually higher than what it is in India (1:39:1) though still lower than China in 1990 (2:25 : 1) though the restrictions put on by the one-child rule in China goes a long, long way to explaining how ridiculously out-of-whack their SBR tends towards. However, the fact that it's higher here compared to India (and I think South Korea as well) could also bolster the argument that sex selection methods and technology are more readily available/accessible in the U.S. than elsewhere.

  • The study compares Asian Americans to Whites (Whites show no statistically significant bias one gender or another) though I'd be curious to see what it looks like compared with African American and Latino American populations. I would still expect that the Asian American SBR to be higher, regardless, 1) because I think patriarchal preference for sons is more entrenched and 2) I think sex selection is less ethically challenging for Asian Americans compared to say, Catholic-raised Latinos (this presumes abortion is one primary method of sex selection).

  • "Male bias...was true irrespective of the mother's citizenship status" - which suggests that it may not just be immigrant families, but also American-raised families who exhibit a SBR.

  • "SBR were found despite the absence of many of the factors advanced to rationalize son bias in [Asia] such as China's one-child policy, high dowry payments (India), patrilocal marriage patterns (all three countires) or reliance on children for old age support and physical security." In other words, the SBR in the U.S. makes even less sense here than it does elsewhere in the world. But then again, it's not like patriarchy has ever required much rationalization to assert itself.

  • It should be noted that in families where a first-born son is present, there's no subsequent SBR with future children, meaning that one could read this finding as suggesting that what Asian Americans really want is at least one son but past that, they're fine with daughters. But they really want that son.

  • Small aside, but the study claims that Chinese, Koreans and Indians, collectively, make up less than 2% of the overall U.S. population. Accepting that the study is leaving out the second biggest Asian ethnic group (Filipinos), I think their math is wrong. Check it yourself: those three populations - not even accounting for people of mixed-Chinese/Korean/Indian descent - would still be over 2% of the total U.S. population, according to Census 2000 figures.

    So what does this all mean?

    For starters, let's just ask the unspoken question here: are Asian Americans more likely to use abortion as a means for sex selection? Given the wide availability and affordability of prenatal sex testing and abortion (compared to ineffective or more expensive means of sex selection), it's a rational economic argument that, if you were going to sex select, testing + abortion would be the way to go.

    I was looking at the CDC's abortion surveillance stats but given that 1) I'm not a quant guy and 2) it's past midnight, I'm not sure if these tell us anything meaningful since, 1) Asian American women are aggregated with Native Americans and others under the always-popular "Other" category so it's impossible to parse the numbers down just for Asian Americans, let alone just Chinese/Korean/Indian. 2) Table 14 suggests that Other women (presumably including Asian Americans) over the age of 30 are more likely to pursue an abortion relative to White and Blacks in the same respective age group but since Table 14 measures overall abortions rather than per capita, I'm not sure one can read the chart as suggesting that the abortion rates are actually higher within those populations vis a vis others. Maybe someone who is more stat-trained and/or awake than I can crunch that.

    Even taking the abortion angle off the table - and I noticed a pro-life group is already using the study to suggest that Asian Americans are abortion-crazy - the study confirms something that most of "us" already knew: Asian American society has a patriarchal slant (pun intended), at least when it comes to prioritizing sons. If someone else has a counter-read to this conclusion, I'd be curious to hear what it is.

    Lingering questions: First of all, I find it interesting that among my circle of Asian American couples, everyone wanted daughters - not necessarily exclusively, but most certainly at least one, if not the only child they'd ever have. That was certainly true for Sharon and I and we feel pretty lucky we ended up with a daughter (albeit a daughter who doesn't always eat her veggies and is in the middle of a worrisome princess phase but that's another story). But I was struck by how common this seemed to be with other people we know. This could suggest that, given a generation or so, the SBR might fade within Asian America, at least among 2nd and 3rd generation APIs (something, quite notably, the study doesn't parse but perhaps they didn't have access to the necessary data to do so).

    Second, and this is a bit of an aside, but the study notes that the biological norm for male/female is 1.05 sons born for every daughter. I'm no evolutionary scholar (but if you are, chime in!) but wouldn't it make sense for that to go the other way? Wouldn't, from an evolutionary point of view, having a higher ratio of women being born be more advantageous, especially since most women only bear one child at a time, with a long gestation period? I don't see the benefit in producing an excess of men when it's really women you need to propagate the species.

    eliaday said...

    i just have to say that i can't help but feel like this one finding doesn't really prove anything. if there was merit to the idea that chinese, korean, and indian parents in the US were using technology to give them boys instead of girls, shouldn't this show up in first born and second born children, why just third born?

    i also want to clarify that these families are 50% more likely than their white counterparts to have boys; not just 50% more likely.

    Poppa Large said...


    Thanks for clearing up the percentage issue - this is why I'm not a quant guy!

    The study does show a slight SBR with second children but it's most pronounced with a situation where the first two children are girls.

    As for why they wouldn't use that technology on the front end, if I recall - and I may be getting this wrong - among first-born, I think the ratio is normative and that makes sense. Outside of China, in areas where having more than one child isn't met with punitive measures, the concern isn't having your oldest be a son. The concern is having sons, period. Therefore, if sex selection is being deployed, it doesn't necessarily hold that such technology/techniques would be used for the first child. But I think the reason you see a spike with third children is that, at that point, intervention to have a son is more relevant.

    In other words, most parents are fine flipping the coin for the first kid. Second kid too...but once you're past the nuclear household threshold, the data suggests that a higher % of families are willing to take extra steps to insure a son.

    eliaday said...

    yeah, so i just was able to read the summary...

    the results are based on 2000 census data; they weren't looking at births in 2000 or anything. in the summary piece it talks about determining the sex of a baby in blood typing as being available in 2005; ie this couldn't have been done for kids that were already on the census radar in 2000.


    Jennifer said...

    Actually, I can answer why the biological gender ratio norm is the way it is. Males are more likely to die (not too much more, but enough). When the offspring age, the ratio comes closer to 50:50. I don't know why boys are more likely to die. I think I heard that's it's because they are more reckless. They also die in the fight for a female and other resources.

    Another thing that I think you'll find interesting is a theory that the professor that taught me this had. He's an entomologist who specialized in sex ratios in social insects. During times of hardship, a wasp is more likely to have as many females as possible, because it's better to have a weak female than a weak male because a weak female is much more likely to find a mate than a weak male. He believes that there's evidence of this in humans. Is it possible that these families had more resources when they got to that third child? However, that alone shouldn't account for a 50% increase. Or does it? Maybe this is just a viable an explaination as sex selection abortion.

    Anonymous said...

    Had a son the first go around but if we didn't already contemplated using PGD for sex selection for the second. My wife is a Korean-adoptee raised by non-minorites and I'm half-Native American. We just preferred sons and have the money to do so.

    Lisa Y. said...

    I agree with Jennifer on why there are more male babies born. Males tend to be physically weaker than females in early childhood. Also, historically men have always had a shorter life span since they take on more dangerous roles in society: hunting, going off to war, stealing their neighbor's wife, etc..

    Anyway back to your point, were the parents in this study first generation or second generation and later? I think it makes a big difference.