Flickr

  • www.flickr.com
    Abiola_Lapite's photos More of Abiola_Lapite's photos

« Battlestar Galactica: Best US Show of 2005 | Main | I Don't Get It »

December 28, 2005

Comments

radek

"*A problem Japan is alone amongst its East Asian neighbors in not having: for this, I blame the much greater sway of Confucian thought in mainland East Asia"

I think the reason lies someplace else, as historically female infanticide was as widespread in Japan as elsewhere in East Asia - the change in Japan is a pretty recent one.
The evidence for this is indirect but still suggestive - in a world without modern contraception the two main forms of birth control are getting married later in life (shorter life bearing lifespan) and infanticide (particularly of girls). So if you know the fertility rate, the sex ratio and the average age at marriage you can estimate how much (female) infanticide is going on by figuring out what the fertility "should be". Prior to 20th century all three of those variables are essentially the same for Japan and the Continent (of course the data's noisy here but there's no reason that it's biased one way or another). I'd guess that sometime around the beginning of the Meiji was when Japan switched its birth control to the "get married later" method (which is what essentially the article is talking about above).

Abiola

"historically female infanticide was as widespread in Japan as elsewhere in East Asia - the change in Japan is a pretty recent one"

I won't dispute that plain old infanticide was as prevalent in Japan as elsewhere in the region, but knowing what I do about the place of women in Japanese society before the Meiji restoration, I'm not at all as ready to accept that females have always been singled out as much for abandonment in Japan as in Korea or China. Top-ranking samurai's wives aside, Japanese women have always been much freer of restrictions on their sexuality and general autonomy than their counterparts in the mainland, going back not just to the Heian court in which women like Sei Shonagon and Lady Murasaki wrote openly about love affairs involving both single and married women, but all the way back to the first embassy from the kingdom of Wa sent by the priestess queen Himiko around 300 A.D., something unimaginable in China even at that early date.

"Prior to 20th century all three of those variables are essentially the same for Japan and the Continent (of course the data's noisy here but there's no reason that it's biased one way or another)."

It's precisely because of what I do know from the histories of the three countries that I don't accept this assumption as being a valid one to make. Comparing the treatment of women in Choson Korea to that in Tokugawa Japan is like looking at night and day - the latter country was no female paradise by any means, but in some respects it was actually a better place to be a woman than Europe, while Choson women lacked the right to initiate divorce (and when they were divorced by their husbands or even widowed, the right to remarry) or to inherit property,

http://web.archive.org/web/20030628084815/http://www.hf.uio.no/east/om-east/ansatte/ansatte/Ansettes+tillegg/Koreas+samf.+og+polit.10.htm

and even being seen by an unmarried man peeping into the family compound was enough to bring ruinous consequences on their heads; even on those rare occasions on which they were allowed to leave the home, they were expected to wear shrouds to conceal them from the gaze of strange men. Even (or rather, *especially*) at the highest levels of society, Choson's attitude towards women had a lot more in common with today's Saudia Arabia than it did with the Japan of the Bakufu.

"I'd guess that sometime around the beginning of the Meiji was when Japan switched its birth control to the "get married later" method"

I don't buy this at all - contraception is just as freely available in Korea, and the statistical evidence suggests that Korean women make even more use of it than their Japanese counterparts, and yet the sex-skew in births in Korea has continously gone *up* over the past few decades even as women have been marrying ever later in life. What is more, the mere fact of later marriage shouldn't in itself skew the sex ratio, just reduce the total number of children born per woman. The only explanation for the difference between the two societies must lie in differing cultural attitudes towards daughters, and the historical evidence leads me to blame Korea's much more fervent embrace of Confucianism (a difference which is also strongly manifested in the two country's *very* different attitudes towards female chastity).

PS: For a good overview of the historical place of women in Korean society, take a look at the following article and the book mentioned in it.

http://times.hankooki.com/lpage/culture/200512/kt2005121619374810980.htm

It's rather revealing about just how worse off Korean women have had it that widows were only granted the right to remarry in *1989*!

radek

Maybe I misunderstood your original post but your asterixed argument seemed to be that the difference in the sex ratios between mainland and Japan was due to the difference in the influence of Confucianism. Presumably today this works through the channel of selective abortion and historically through abortion/ infanticide of daughters.

Or did you mean that Confucianism explains the difference in status of women?


"I'm not at all as ready to accept that females have always been singled out as much for abandonment in Japan as in Korea or China...Japanese women have always been much freer of restrictions on their sexuality and general autonomy than their counterparts in the mainland..."

You could still have high rate (or same between two societies) of abandonment/female infanticide along with a higher status of women in society, if both the mother and father would prefer a male offspring to a female offspring.

"I don't buy this at all - contraception is just as freely available in Korea, and the statistical evidence suggests that Korean women make even more use of it than their Japanese counterparts"

Yeah, today - but since you brought up Confucianism as an explanation I was thinking more of historical and social forces which shaped today's observed cultural norms - and anyway, use of contraception cannot explain the skewness in the sex ratio.

"What is more, the mere fact of later marriage shouldn't in itself skew the sex ratio, just reduce the total number of children born per woman"

But that's the point. Contraception aside, there are two* ways of lowering fertility; infanticide or getting married later. The first skews the sex ratio, the second doesn't. So you can't infer the frequency of infanticide from observing fertility - you need to also know either the sex ratio or at least age at first marriage. Pre 19th century China and Japan had different fertility rates (due to different levels of income/technology) but from what we know about the sex ratios and age at first marriage in each, the estimated rates of female infanticide are pretty similar (at least way closer to each other than to Europe). The number (adjusted for population size) of "missing women" seems to be roughly equivalent (again, historically).

This in itself doesn't say anything about the status of women in each society. It does suggest however that link between the freedom enjoyed by adult women and infanticide as birth control is very weak. The mother in the household may have the same desire to limit number of children, and the same prejudice towards boys as the father.

I don't dispute at all that there was a significant difference between the status and well being of women in each of those societies. I just don't think that Confucianism has much to do with the skewed sex ratio - historically the two (3) societies were roughly "Confucian" to the same degree as they are today. But in terms of birth control practices were very similar. Something changed in Japan around the late 19th century. It didn't change in China and Korea.

Personally I'd go with "modernization (in Japan, but not China) plus some serious cultural lag/hysterisis/path dependence (in Korea but not Japan)".

Having said all this, and having discounted the role of religion (or whatever Confucianism is) in explaing the difference between China/Korea and Japan, I am left wondering if the difference between population control practices in historical Europe and Asia CAN be attributed by Christianity. I think though I'm not sure that infanticide was a lot more prevelant in Roman times than later, and even today East Asians have a pretty cavalier attitude towards abortion so that sort of fits, but I'm sceptical of myself on this one...

* Of course what I mean is that these are the two most common forms of population control in a pre-contraception world. There are others. Apparantly the French were quite good at the "rhythm method" (same age at first marriage and male/female ratio as rest of Europe, but overall lower fertility). Also time between children matters - in so far as this data is harder to get, I've seen folks proxy this by either the life expectancy of the married women (more frequent childbirths = higher chance you'll die in child birth) or by looking at the family structure (outsource the child rearing to grandma and get busy making the next kid) - but it doesn't seem to play a huge role.

Abiola

"Presumably today this works through the channel of selective abortion and historically through abortion/ infanticide of daughters."

Yes, that is just what I'm saying, and the reason I'm noting that you can also see this difference in attitudes at work in the way women's sexual behavior is policed is to show that there's nothing the least bit implausible about it.

"You could still have high rate (or same between two societies) of abandonment/female infanticide along with a higher status of women in society, if both the mother and father would prefer a male offspring to a female offspring."

Er, no, you couldn't: the extent to which both mothers and fathers prefer sons to daughters *is* a manifestation of the lower status of females in a society, not some mere epiphenomenon. You don't so readily throw away babies of one gender if you think their lives are worth anything nearly as much as those of the other.

"So you can't infer the frequency of infanticide from observing fertility - you need to also know either the sex ratio or at least age at first marriage."

Even if you knew the age at first marriage, how would that tell you much of anything under a regime of non-shrinking population size? Besides, the picture of feudal East Asian civil servants carefully collating population statistics by age and sex for illiterate peasants who had but single names is one I find extremely farfetched - I very much doubt that rulers' interest in their populations in the pre-modern era extended far beyond knowing how much rice could be demanded in tax from how many households.

"from what we know about the sex ratios and age at first marriage in each, the estimated rates of female infanticide are pretty similar (at least way closer to each other than to Europe). The number (adjusted for population size) of "missing women" seems to be roughly equivalent (again, historically)."

Your final statement here would only follow from the one preceding it if one also knew the rate of male infanticide in the societies in question, and that it was lower than that for females. Somehow I doubt such statistics were meticulously collected by the officials of Qing, Choson and the Bakufu.

"I just don't think that Confucianism has much to do with the skewed sex ratio - historically the two (3) societies were roughly "Confucian" to the same degree as they are today."

And I am saying that not only can you can see this difference directly at work in the much greater readiness of Korean women to abort female fetuses (it certainly isn't a greater aversion to abortion in general which prevents Japanese women from doing the same), but that you also see it at work in many other facets of the lives of women in the two societies, a difference no knowledgeable historian would dispute goes back quite a long way; the only aspects of Confucianism Japan ever wholeheartedly adopted were those that served to buttress the ruling samurai elite.

"Personally I'd go with "modernization (in Japan, but not China) plus some serious cultural lag/hysterisis/path dependence (in Korea but not Japan)"."

And how would that explain the increasing intensity of sex-selected abortion in today's Korea, despite the readiness of Koreans to adopt so many other aspects of Japanese and Western culture? Korea *is* currently an exporter of "cultural modernity" to many parts of the world that don't have its sex-ratio problem, after all, and besides it's a funny sort of time lag which increases with ever more wealth and international exposure ...

"Having said all this, and having discounted the role of religion (or whatever Confucianism is) in explaing the difference between China/Korea and Japan"

How can you so casually dismiss Confucianism as a factor if you haven't even investigated it thoroughly enough to know for certain that it isn't a religion?

"even today East Asians have a pretty cavalier attitude towards abortion so that sort of fits, but I'm sceptical of myself on this one..."

Again, a greater readiness to use abortion as the premier means of birth control does not necessarily translate into ratio of live births heavily skewed against girls - witness Japan.

I am thoroughly convinced that the reason why Koreans, Chinese and other East Asians go in for sex-selective abortion while Japanese families don't has everything to do with Confucianism, not some purported cultural time-lag which is in any case largely illusory - Japan long had a much more liberal attitude to matters of sex and religion than the West, so any lag isn't on their side in this respect. Confucianism is why women have been so late to get their rights in Korea (and why even now so said rights lag behind those enjoyed by women in Japan), it is why Korean women have to go in for hymen restoration surgery to find good husbands, and it is why families who don't want to have large families are so ready to abort a fetus as soon as they learn it's a girl, while today's Japanese families are actually something of an exception in the modern world in having something of a stated preference for girls (a preference not so strong that they seem willing to act upon it, however).

By the way, the following article from the Journal of Population Research would seem to agree with me in blaming Confucianism for China's skewed sex ratio.

http://www.findarticles.com/p/articles/mi_m0PCG/is_1_21/ai_n6155263/print

It really isn't a remarkable claim to anyone who appreciates the importance of sons in the Confucian system in maintaining the bloodline and venerating ancestors. One point worth noting which shows why your "cultural time lag" hypothesis can't be taken seriously is that according to the article above, Taiwan, which also currently suffers from a skewed sex-ratio, enjoyed a more normal balance throughout the 1915-1940 period when it was a Japanese colonial possession. If it were a mere case of cultural lag which accounted for the difference, one would think 50 continuous years of Japanese rule during which selective infanticide and abortion were strictly forbidden would have left its mark on the Taiwanese, who are otherwise even more enamored of all things Japanese than the Koreans are, and openly so too. Taiwan has been exposed to modern trends since 1895 at the very latest, so the only explanation that makes sense for the strange imbalance in Taiwan's birth statistics is that long-standing Chinese confucianist cultural preferences were free to reassert themselves once the Japanese left, and that these preferences were subsequently reinforced by the flight of Nationalists from the mainland.

PS: Yet more people who lay the blame on the same culprit I do.

http://iussp2005.princeton.edu/abstractViewer.aspx?submissionId=50261
http://iussp2005.princeton.edu/abstractViewer.aspx?submissionId=50793
http://iussp2005.princeton.edu/abstractViewer.aspx?submissionId=50404
(look at the actual contents of the last two papers to see what I'm referring to)

Presumably, not all of the authors above are clueless on the subject.

Nathan

Excellent blog, my friend. I surfed in after looking up the search term "Noam Chomsky is a hypocrite" and read your position on his tax sheltered trust.

Your attention to detail and interest in hard data is a refreshing read - even though I might disagree with some of your sentiments.

Nice work.

Jing

Abiola, I'm afraid your statistical information for male/female birth rates is very out of date. Yes, South Korea's ratio was even higher than China's at one point, but I distinctly remember reading other surveys post 1993 that have shown that the trend is in decline. The peak ratio was somewhere around 116:100 males to females but it has fallen significantly from then since.

As for the explanation of lopsided gender balances in China, I have heard of other explanations before. One possibility is the high prevalence of Hepatitis B within the general population of China the vast overwhelming majority of whom do not even realize they are carriers. Apparently female Hepatitis B carriers were shown in case studies to give birth more often to boys rather than girls. This coupled with gender preference in birth (abortion/infanticide) is what is possibly causing China's skewed birth rate. However, if the example of South Korea proves to be apt, this problem is ultimately self-correcting when the imbalance will reach a peak and then continue to decline and return to normalcy.

In regards to Taiwan, do you have any evidence as to the prevalence of birth control during the colonial period? Mainland China also almost certainly enjoyed a normal gender balance circa 1895-1945 but that of course has nothing to do with the Japanese.

I think you are too quick to dismiss time lag and too ready to name Confucianism as the responsible factor when looking at birth ratios. Mostly I think this is due to you using out-of-date data. I unfortunately cannot find offhand the same source which cited a high of 116:100 male/female birth ratio but the CIA factbook lists South Korea's present ratio at 108:100, closer to natures own 105:100 figure. I know it is problematic to use data from two different sources but it is the best I can manage now.

BTW, does anyone have access to historical Japanese census data? It is quite possible that Japan went through a similar phase of skewed birth rates in the early half of the 20th century.

Abiola

Nathan,

Thanks for the compliment. I'm glad you find what I write interesting.

"Abiola, I'm afraid your statistical information for male/female birth rates is very out of date."

Jing,

Really? Is the data being used by all those researchers whose abstracts I linked to on Princeton's website - all dating from 2005 - also out of date? Until you can show me some hard, well-sourced evidence to the contrary, or explain why researchers publishing this year should use obsolete data, I say you're fooling yourself here.

"Apparently female Hepatitis B carriers were shown in case studies to give birth more often to boys rather than girls. This coupled with gender preference in birth (abortion/infanticide) is what is possibly causing China's skewed birth rate."

This is a popular theory amongst those who don't want to open their eyes to the facts, but it is a manifestly false one. Hepatitis B rates in Korea were *dropping* simultaneously with the rise in the sex-skew of childbirths, and it isn't Hepatitis B which makes the overwhelming majority of children abandoned after birth in China female. Abortions are not recorded as miscarriages in countries like Korea and Taiwan, and it is a well-established fact that in both cases, the great majority of babies which are aborted are *female* - it's trivial to tell with such late term fetuses - while Hepatitis B could not possibly explain why the sex-skew increases with the order of children born into families in Confucian-influenced mainland Asia, getting worse as one goes from second to third to fourth child, etc.

"Mainland China also almost certainly enjoyed a normal gender balance circa 1895-1945"

If you'd bothered to read the material I linked to in my response to Radek you'd know that this is an utter falsehood: China's gender balance during the period in question was highly skewed thanks mostly to female infanticide, and it only got worse with the age cohort of children looked at because girls were less well fed and less well treated for illnesses. Only with the advent of the Communist Party did this tendency go into (temporary) decline.

"the CIA factbook lists South Korea's present ratio at 108:100"

The CIA World Factbook hardly stands comparison with a publication like the Journal of Population Research as a quality data source on this particular subject.

"BTW, does anyone have access to historical Japanese census data? It is quite possible that Japan went through a similar phase of skewed birth rates in the early half of the 20th century."

If you can read Japanese, all you have to do is go to the following website:

http://www.stat.go.jp/

As for your claim about Japan having skewed birth rates, there's zero truth to it whatsoever. Japan has been collecting good quality statistics on such matters since 1890, and there has not been a single era since then in which this phenomenon has occurred (which isn't surprising, seeing as they also repressed it in Taiwan): this is a mainland Asian problem.

Now, to address any possible claims that this difference between Japan and the mainland is a modern phenomenon, I refer to the following paper, which works with data from the period between 1671 and 1871:

http://www.indiana.edu/~pirt/PIRT_WP/wp97-2.html

["Did sex selective practices exist in early modern Japan as is observed in many developing countries in Asia today? I developed a set of hypotheses based on the long-term and short-term perspective people may have used, regarding the sex preference under the stem family tradition in early modern Japan. However, we did not observe much evidence to support any sex preference. From the long-term perspective, the strong male preference (especially the first son preference) is generally assumed in the patriarchical tradition, but we did not observe much evidence to support such preference."]

Japan has never been a paradise of female equality, nor is it close to being one today, but it just isn't true that female children were historically regarded as so much disposable garbage as seems to have long been the case elsewhere in East Asia, and this difference I attribute to the same sources even Chinese population specialists do - the stronger grip of Confucianism in mainland Asia.

PS: I've managed to dig up some statistics to suggest that Korea's selective abortion of girls *has* declined somewhat from its peak; the thing is, though, that it appears to have since leveled off at a still highly abnormal 109-110:100 ratio of boys to girls, with 140+ boys being born as third or later children for every 100 girls in the same group.

http://www.ipss.go.jp/webj-ad/WebJournal.files/population/2003_6/21.Chang.pdf

Of course, this in itself says nothing about childhood mortality rates, which have also been higher for girls than boys in Korea, right into the 1990s - a phenomenon which very much goes against nature, given the greater resilience of females at every age (which is why the 105:100 normal ratio exists to begin with). If as many as 100 females to 115 males make it to adulthood even amongst the youngest Korean cohorts, I should be very much surprised.

Jing

[Insulting, cheeky rubbish deleted. That should teach you a lesson about the wisdom of slinging such crap at someone else on his own property. It's my fault for taking a retard like you seriously in the first place. - A.L.]

Abiola

You're a worthless idiot. Consider yourself banned - don't you ever dare try to stain my blog with your presence again in the future; I don't have time to waste on foul-mouthed little shit-for-brains who can't deal with the fact that their cultures just might be less than perfect - you are one hell of a presumptious moron not to even realize that the only reason Singapore's sex ratio returned to anywhere near normal after soaring in favor of boys in the 1980s is because Lee Kuan Yew's institution of draconian punishments; but then I suppose any facts which might confuse cowardly, mannerless, gutless, egotistical little punks like yourself who would wet themselves if we ever met face-to-face, can safely be ignored ...

gene berman

radek:

French good at the "rhythm" method? Here, I'd simply assumed that what they were really good at was the "French" method!

radek

"Er, no, you couldn't: the extent to which both mothers and fathers prefer sons to daughters *is* a manifestation of the lower status of females in a society, not some mere epiphenomenon."

This is true, but we're comparing two different societies here - Japan and mainland East Asia. Women in Japan have had (have) higher status than women in China/Korea. But in both places their status was still lower than that of men, which is what matters. If infanticide was a means of controlling fertiliy then in both societies when faced with the decision of which gender to "choose" for their offspring, the payoff would be higher for boys. If the status of women were to rise (or be higher than) one would expect the frequency of female infanticide to decrease. But I don't think the data bear this out very strongly. To put it in jargon, the elasticity of female infanticide with respect to women's status is fairly low. You need to increase the status a lot to lower the infanticide to any appreciable degree.

"Even if you knew the age at first marriage, how would that tell you much of anything under a regime of non-shrinking population size?"

If actual fertility (which you observe) is a function of age at first marriage (which you also know) and of frequency of infanticide (which you don't know) you just compare it to what fertility would be if infanticide was zero and solve out for the frequency of infanticide. You'd really need to know the sex ratio too (of marriage age population, not at birth) but you could come up with a reasonable estimation (by positing some dynamic process)based on extrapolation from one or two observations. I don't think the fact that a population is shrinking, growing or hanging constant has anything to do with this since what we care about is just fert, not fert-mort.

"Besides, the picture of feudal East Asian civil servants carefully collating population statistics by age and sex for illiterate peasants who had but single names is one I find extremely farfetched"

Well, yes, which is what makes the job of any historian (cultural, economic, demographic or otherwise) a pain in the ass and fun as well. But some records do exist and while one needs to be cautious about potential biases it would be silly to ignore this information.

"And how would that explain the increasing intensity of sex-selected abortion in today's Korea, despite the readiness of Koreans to adopt so many other aspects of Japanese and Western culture? "

Yeah, Korea is the weird one here, which is why I hand-waved with the statement about *serious* pathdependence/lag. Honestly I don't have that much confidence in my thesis either, I was just trying to provide an alternative explanation, because I don't think Confucianism is it either (see below).

"How can you so casually dismiss Confucianism as a factor if you haven't even investigated it thoroughly enough to know for certain that it isn't a religion?"

Man. Red herring. Big glowing neon oily salt water fish. That was said very tongue in cheek.

"Again, a greater readiness to use abortion as the premier means of birth control does not necessarily translate into ratio of live births heavily skewed against girls - witness Japan."

Yes, but historically infanticide was used in Japan to skew the ratio, unlike abortion today, which is my point.
At the moment the best I can do is this:

http://wwwlib.umi.com/dissertations/preview_page/1380703/14

It's only the first 24 pages of the article (gotta pay for the whole thing) but the table on page 8 gives the population for Japan and the implicit sex ratios of males to females:

1762 - 114:100
1804 - 110:100
1828 - 109:100
1846 - 106:100

Note that this is for population as a whole, not at birth, so it should be closer to 100:100 rather than the 105:100 one would expect at birth. This does show that I might've been wrong about the Meiji thing - it could've started evening out earlier.
At any rate, given this skewed ratio and in the absence of alternative explanations, female infanticide seems to be the culprit - the article does note the role of infanticide in controlling fertility. In fact I don't think the notion that infanticide was practiced in Japan as well as mainland to be at all controversial, the only question is was it gender selective in Japan like in the mainland (unlike abortion today).

"the rate of male infanticide ...lower than that for females...I doubt such statistics were meticulously collected"

Again, there's no detailed records but there are some. Male infanticide did occur in both mainland Asia and Japan, but all evidence says that at much lower rates - the skewed sex ratios in both are themselves evidence of this.

"the following article from the Journal of Population Research would seem to agree with me in blaming Confucianism for China's skewed sex ratio"

They DO state (and I apologize for the length of the quotation - but it supports your argument, not mine):
"The main reason was a strong Confucian value system that honoured almost all males over nearly all females. Important religious and ancestral rituals were reserved for males. It was believed, and still is today, that the family lineage can be continued only through sons; this attitude is especially strong in Taiwan and southeast parts of the Chinese mainland (Poston et al. 2000; Yang and Chen 2003). Men but not women owned property. A girl was owned by her father and a woman was controlled by her husband and his family. At marriage in her late teens, a daughter was lost to her natal family as she married out to a different village and thereafter worked for and cared for her husband's relatives. A daughter was generally seen as costing more to her parents than they would ever get back from her. Therefore daughters, at least some daughters, were treated as expendable."

However
1. The first statement, regarding Confucianism is merely an assertion, especially in light of the fact that
2. the rest of the paragraph describes the institutional/cultural framework of historical Europe as well. Hell, it's a pretty accurate description of most pre-modern societies.
3. They then go on to provide some contradictory evidence themselves by noting that (again, sorry for long quote):
"Some countries discriminate strongly against girls and women in society, culture, economy and politics, yet this discrimination does not lead to statistically detectable shortages of girls through sex-selective abortion, female infanticide, or maltreatment of daughters. Japan is a good example (I'D ADD "TODAY"). Further, most of the world's Muslim countries have normal sex ratios of children. However, the South Asian Muslim countries have unusually high population sex ratios as do India and Nepal suggesting that the whole cultural area of South-central Asia is inimical to the normal survival chances of females. Why is it that some cultures such as most Muslim societies and Japan, while displaying undeniable and massive discrimination against girls and women, do not go so far as to cause the untimely deaths of girls or engage in sex-selective abortion of female foetuses? This author has seen no research on this question, which is certainly worthy of study."

So you have the non-Confucian historical Japan, India, Nepal and South East Asian Muslim countries as counter examples. Which at best suggests that Confucianism is sufficient but not necessary.

The bigger problem I have with this "argument from Confucius" is that it is almost always called upon to explain any difference between China/Korea and Europe or Japan or any other part of the world (lots of responsibility to lay on one guy's shoulders). It's like explaining economic phenomenon by an appeal to "change of tastes". Now of course it may very well be correct, tastes do change, and Confucianism has had important influence, but it's also a bit of a lazy man's answer, particularly in light of fact that you have non-Confucian societies which exhibit the same phenomenon.

So while I wouldn't reject Confucianism completely, I want something more substantial to go along with it. As an aside, this is sort of parallel to the same problem I have with folks invoking Weber and the Protestant Ethic, since all these explanations essentially boil down to "them folks is just special and that's all there is to it".

"Taiwan, which also currently suffers from a skewed sex-ratio, enjoyed a more normal balance throughout the 1915-1940 period when it was a Japanese colonial possession"

Hmmm, yes, this does seem to be a point in support of the Confucian thing.

"http://iussp2005.princeton.edu/abstractViewer.aspx?submissionId=50793"

This one's mostly about the impact of the "one child" policy or the "one son, two child" policy and it very thoroughly documents its impact. They say:

"However, prenatal sex selection has been practiced not only in Mainland China and Taiwan and Korea, where the Confucian culture remains strong, but also in other nations where political systems and
population policies are significantly different. This suggests that the motivation for this discrimination against girls is indeed mostly cultural (Coale and Banister, 1994; Gu and Roy,
1995; Goodkind, 1996, 1999)."

So again Confucianism gets mentioned as an assertion, but then it gets pointed out that the same thing happens in non-Confucian societies. I don't know if the citations are to works which support the view that Confucianism is the culprit or merely document that female infanticide happens elsewhere. So they say it's *cultural* but apparantly compatible with many different cultures. Which is almost saying nothing at all.

"http://iussp2005.princeton.edu/abstractViewer.aspx?submissionId=50404"

This one's mostly about the likely long term effects of the one child policy and a pretty good summary about the history of birth control policy in China since early 20th century. Mao changing his mind back and forth (reminds me of his classic failure to appreciate the concept of diminishing marginal product: "With every mouth to be fed there will be born along with it two hands to feed it"). But not much on Confucius except:

"One reason is that China (along with Taiwan and South Korea) has a Confucian
patriarchal tradition where son preference is strong and pervasive (Arnold and Liu, 1986;
Gu and Roy, 1995; Kim, 1997; Park and Cho, 1995; Poston et al, 1997)."

Again, I'd have to look at the references to see if they support the idea that Confucius is behind it or just document that son-preference exists in other places.
But this is again mostly in nature of an ad-hoc, slap on, explanation. The meat of both articles is with more tangible stuff about how many women are missing and how many Chinese guys won't find a wife and what will that do to China.

In absence of a comparison to other places this doesn't really tell much in what the underlying cause is.

So...
We have Confucian countries which practice(d) selective abortion/female infanticide, we have non-Confucian countries which have also done so, and we have non-Confucian countries which did not. If there was a Confucian country which did not practice it I'd say you're totally wrong (the low sample size is a problem here)(we do have non-Confucian but patrilineal, patrieverything else societies which do not practice it however). As it is I'm willing to concede that it might play a role but it's definetly not the whole story.

Which is why Japan is an interesting an example since it switched it's behavior sometime in the 19th century - that fact can also shed light on other factors that matter.

"Presumably, not all of the authors above are clueless on the subject"

I never said anyone was clueless. In fact all three of these articles, and the one above are really good and worth reading/

Some interesting facts in'em:

1. Raising the minimum legal age for marriage raised number of marriages in China producing a fertility spurt - because local (rural) administrators had previously insisted on even higher ages. This is like the ironic economic occurance where a regulator imposes a non-binding price ceiling and all prices jump right to that ceiling.

2. Abortion and infanticide seem to be complements not substitutes. In areas where previously abortion was not available and families practiced infanticide, after availability of abortion, abortions increased but infanticide did not decline.

3. Pre 1979 the campaign to lower population growth through lower fertility was mostly just through propaganda. In this sense at least, it would seem that Mao was less oppresive than China's later rulers.

Abiola

"Women in Japan have had (have) higher status than women in China/Korea. But in both places their status was still lower than that of men, which is what matters."

Not at all - what matters is the *degree* to which this status is lower. If the mere fact of lower female status were all there were to the issue, skewed sex-ratios would have been recorded for essentially *all* societies which have ever kept such records.

"But I don't think the data bear this out very strongly. To put it in jargon, the elasticity of female infanticide with respect to women's status is fairly low. You need to increase the status a lot to lower the infanticide to any appreciable degree."

But this is the very point at issue, and one on which I completely disagree with you. The evidence that I've seen thus far *does* bear out that the relatively higher status of women in Tokugawa Japan translated into a lower propensity to throw away female infants or to maltreat them during childhood.

"Yes, but historically infanticide was used in Japan to skew the ratio, unlike abortion today, which is my point."

This is a mere assertion on your part, and one which I see absolutely no evidence for, not even from the thesis page you link to. The mere fact that more men were numbered than women tells us nothing about the relative rates of female infanticide, as it could just as easily be that fewer women are listed than men purely on account of their being of less interest for tax purposes to census-takers, their higher mortality rate due to childbirth practices, or some other such reason; in any case, the same figures show that between 1762 and 1804 the female population increased by 65 thousand, even as the male population *shrank* by 358 thousand, hardly what we'd expect in a society rife with female infanticide, and keep in mind that Japan was at peace throughout the period in question. The study by Saeko Kikuzawa I pointed to doesn't support the notion that there was a historical bias against female children either.

"I don't think the fact that a population is shrinking, growing or hanging constant has anything to do with this since what we care about is just fert, not fert-mort."

It does matter, for the simple reason that if all we know is the number of households or some such figure, an expanding or shrinking population coupled with a bias towards men marrying women substantially younger than themselves would give rise to a distorted picture, as it would mean lots more young people of one gender remaining members of their parental households than those of the other sex.

"Man. Red herring. Big glowing neon oily salt water fish. That was said very tongue in cheek."

Whatever. I responded to what you actually wrote, not what you're claiming in retrospect to have actually meant. Certainly, your refusal to acknowledge the very direct and well attested relationship between Confucian notions of proper familial relationships and the relative value of women in East Asian societies does not suggest an intimate familiarity with it as a system of thought.

"Yes, but historically infanticide was used in Japan to skew the ratio, unlike abortion today, which is my point."

That there was a substantial amount of infanticide and child abandonment in pre-Meiji Japan is a point I've never disputed, but what's at issue here is whether Japan historically had as strong a bias in favor of male children as to render female children the primary victims of such mistreatment as was true in Korea and China, and for this claim I simply don't see good evidence from you to back it up, nor am I willing to go along with the default assumption that Japan had to have been similar to Korea and China in this respect simply because it neighbors them. What is more, Saeko Kikuzawa offers an easily understood mechanism whereby a difference in attitudes towards female newborns could have arisen, and it just so happens to be in the form of a Japanese divergence from orthodox Confucian thinking about family inheritance; better yet, she even backs it up with several references which should put to rest your claim that Confucianism's role is a mere "assertion" being resorted to by scholars too lazy to dig up or carry out any real research on the issue:

["It is repeatedly documented that son preference is particularly prevalent in countries with strong Confucian traditions (Goodkind, 1996; Cleland, Verrall, and Vessen, 1983; Cho, Arnold, and Kwon, 1982; Freedman and Coombs, 1974). In traditional China, in a patrilocal joint family structure, all married sons ideally lived under the same roof with their parents. One of the primary obligations of sons was to provide male offspring so that the ancestral line might be continued and so that there would always be someone to burn incense for the deceased (Parish, 1978). Note that the family system in China (joint family system) is also similar to Japan (stem family system) in its patrilocal tradition but different from Japan in the pattern of inheritance. In the joint family system, all the sons inherit the households, while any daughters cannot. In the stem family system, generally the first son inherits the household, but if the family lacks sons, the first girl can inherit the household. In this respect, sex preference may be more acute in China than in Japan."]

See all the bibliographical refs in the above quotation?

"The bigger problem I have with this "argument from Confucius" is that it is almost always called upon to explain any difference between China/Korea and Europe or Japan or any other part of the world (lots of responsibility to lay on one guy's shoulders)."

Your dismissal of Confucianism as the main contender for this difference in attitudes on the ground of it supposedly being "a lazy man's answer" is what is most dubious, not the fact that a man whose ideas laid the framework for the organization of entire societies for millenia should be regarded as supremely important in explaining how those societies differ from others; doing so is no more "lazy" than it would be for someone looking at the nature of Western culture to call upon Aristotlean thought and Pauline Christianity in seeking to understand traditional American and European attitudes towards sex, family life and the just social order. If someone said to me that it was a "lazy man's answer" to blame the extreme distaste felt towards homosexuality by many Westerners on the Judeo-Christian tradition, I'd find him or her hardly any easier to take seriously than your own unwillingness to admit the importance of a less rigorous adherence to Confucian values in determining Japan's better balanced sex ratio.

"So again Confucianism gets mentioned as an assertion, but then it gets pointed out that the same thing happens in non-Confucian societies."

In other words, the only condition under which you'd accept that Confucianism made the difference between China/Korea and Japan is if it were the *only* possible cause of skewed sex-ratios? Would you be equally sanguine about dismissing State Shintoism as a factor in WW2 kamikaze attacks simply because one could note that the 9/11 attackers were Muslim, or to claim that Nazi ideology was not the cause of German aggression in World War 2, simply because it didn't motivate the Finnish or Japan war efforts? This is a really illogical argument for you to be advancing - there's no law saying any phenomenon must arise in every society for a single reason shared between all of them.

"So you have the non-Confucian historical Japan, India, Nepal and South East Asian Muslim countries as counter examples. Which at best suggests that Confucianism is sufficient but not necessary."

The thing is, I'm not talking about South Asia, which shares a lot less in common culturally with East Asia than the countries of the latter region do with each other, and of the nations you mention, only Japan happens to be in the set I'm concerned with. The point at issue is why Japan hasn't recorded the same skew in births as other traditionally Sinocentric countries have, and based on everything I've learnt about the history of ideas in the region, it's clear to me that the explanation lies in the stronger grip of Confucian thought in the mainland.

"As it is I'm willing to concede that it might play a role but it's definetly not the whole story."

It most certainly is as far as East Asia is concerned; rank sexism alone hasn't been enough to spur the emergence of a similar demographic bias in Africa, South America or the Middle East (in many parts of which the tools required are common and widely affordable), all of which regions have attitudes towards women which aren't by any means progressive, so what else can it be other than a value system which says girls are nothing but a burden?

"Which is why Japan is an interesting an example since it switched it's behavior sometime in the 19th century"

Again, Saeko Kikuzawa's study provides at least one argument that no such skew existed even before the 19th century, and the stats on the thesis page you linked to don't constitute at all convincing evidence for your assertion. If you want to maintain that Japanese attitudes towards women underwent some sort of transformation towards the close of the Edo period, you're also going to have to explain where such ideas could have originated, and how they could have percolated into a Japan which had been sealed off from the West for some 200 years by that point. It certainly couldn't have been via Korea or China ...

pinkelephant

Your argument is just superb and impressive.
I really enjoyed it.Thanks!!

radek

At this point the whole argument boils down to:

1. Was gender selective infanticide practiced in pre-Meiji Japan (as opposed to more balanced infanticide)?

and

2. If not, is the difference due to the stronger influence of Confucianism on the mainland compared to Japan?

(If the answer to 1 is in fact yes then I assume you'd agree that Confucius should go back on the shelf, at least temporarily).

The articles you cited earlier all document the fact that gender selective abortion/infanticide are/were preveland in China and Korea and also note in passing that these are Confucian countries. This does not constitute proof in itself. On the other hand, you're entirely correct in pointing out that the fact that selective abortion/infanticide occurs(ed) in non-Confucian societies like India does not constitute a refutation of the role of Confucianism since the same phenomenon can arise in different societies for different reasons.

Even if Kikuzawa's article (which now I have read carefully having merely skimmed it earlier) shows that the answer to 1) is false the answer to 2) is still open. As it is this is what I get from the article (btw, if this is what sociologists are doing these days then I'm quite impressed):

- From the data she has she finds no role for parental "choice" in affecting child and infant female mortality differently than male mortality. This is weak support for a "no" answer to 1).

- Reason it is weak is because, as she herself notes, of the way that the temple recorded the deaths with a 1 year lag which might mean very well that infanticides would never be reported (this is her "possible explanation number 2") - this would be true if selective infanticide occurs, but once you're born you get treated the same whatever gender you are (so no effect on relative mortality) - and there's some evidence to support the view that this is the way it works - see her citation of Goodkind.

(As an aside, it's strange that after questioning the accuracy of historical data in general and all that, you seem to have no qualms taking this dataset at face value).

However even if we were to accept the "weak no" it still doesn't give answer to 2. In fact Kikuzawa herself in various points mentions the shared Confucian heritage of Japan and China (via Korea) and refers to Japan as "a Confucian country". Now this might be somewhat inaccurate in the sense of degree or whatever but still... what she does emphasize is the difference in the inheritance law/custom between China and Japan - that it was possible for daughters to inherit, given no sons were around in Japan but not in China. So to her the difference between the two lies in the joint-family vs. stem-family systems.
So fundementally it seems its the institutional framwork.

Now of course the difference between the two systems could be due to the diverging degree of Confucian influence but I don't think, with my limited knowledge of Confucianism, that there's an obvious direct connection, given that both systems are essentially patriarchical and patrilineal - the devil's in the details I guess and a slight change in the structure of the patriarchy can have big consequences.

However as I said, I'm not entirely convinced that there was no selective infanticide in Japan - though now I'm less sure that there was and I'm guessing it happened to a lesser degree than in China - partly because as the author admits the sex ratios during this time still ended up pretty screwed and something had to be doing this screwing up. In Kikuzawa's regression the only variable that could be the culprit is the interaction between illlness and child care which would mean that women (girls) died more frequently than boys because child care was conducive to spread of disease. It makes sense. But somehow I doubt that it could explain all the deviation of the sex ratio from what it should be (with good enough data one could decompose it and look at the residual). Which would mean that there's another culprit out there and selective female infanticide is still a prime suspect here.

Also note her comment on the link between status of women and female mortality:
"(The result) suggests that the increase in economic opportunities for women was not the major reason for the decrease in their mortality"
which means, as I speculated earlier, that the link between status of women or their economic value and female infanticide/mortality is weak.

At the end of the day I'm willing to say that historically the frequency of female infanticide was probably lower in Japan than in the mainland, that this was most likely due to the nature of the family and inheritance system and the difference in these systems may or may have not been due to Confucianism, and that this past has echoes which persist to this day. In fact I'd much rather focus on the family system as the explanation, because it is something tangible and direct rather than Confucianism (or "culture" in general) which is always a more ethereal concept and hence a more sketchy explanation.

(I know some people who know way more about this topic than I do - in terms of data and all that - and I'll ask'em when I get back from vacation but that'll be more than a week from now, by which time this argument will be old news)

gene berman

radek:

In constructing your scenario of likely explanation, you seem to favor the particular family and inheritance system in Japan as more responsible for the differential (lower) female mortality vis-a-vis the other societies under consideration.

Regardless of what the actual truth of the matter may be, this snippet is an example of a common logical fallacy, like explaining night and day by saying "day is caused by the sun coming out." True or not, it's no explanation in the sense one is sought.
Rather, "family and inheritance structure" are phenomena of the same type as manifest differences in mortality begging explanation: likely results of human behavior shaped by the ideas held by those humans. And, unless there is some more impressive known differential between the ideas of the compared cultures, it seems that "degree of Confucian influence" stands as the best surmise, if only by default. Note that this does not settle the matter--it's just the best offered so far (or likely to be offered, given remoteness of the time considered, reliability of accounts and data, etc.) "Cause" of such phenomenon might be pursued further: to those predisposing certain societies (and not others) to adoption of particular social systems, etc.


radek

gene, that is Kikuzawa's explanation. Also, ultimately everything but land and resources are endogenous (which is why so many economic historians are obsessed with land labor ratios and the like) so your critique applies as much to Confucianism as it does to the form of family structure - why did people in one society adopt a particular set of belief and in another a different set of beliefs? There might be an element of chance involved but these beliefs had to offer something, solve some problem, work for someone, else they would not have persisted.
However generally human behavior "on the ground", with all its propensity toward habit and rule-following, is a lot more flexible then social institutions, whether you're talking about the structure of the family or the underlying ideological foundation. So it makes sense to take the stem-family vs. joint-family structure as exogenous at least in the "short run" (which here could mean several generations and a hundred years or more)

gene berman

radek:

What you say is true enough; but, while it is true that "As long as it's possible that another cause exists to explain the phenomenon in question than the one to which we subscribe presently, our present explanation cannot be regarded as final" it's also true that that's the generally-recognized state of all scientific investigation and especially of those which proceed (or pretend to proceed) empirically.
And, inasmuch as the original question had to do with whatever difference might account for differential male/female ratios (and not with what might account for differential conformances with Confucianism), your interjection amounts to "There may be other explanations having nothing to do with Confucianism and some which, while related, are antecedant and causative of both Confucianism and the phenomenon we seek to explain; the former seem unlikely but I don't think we can relinquish investigation of the latter." Again, it's a quibble: something not denied in Abiola's original thesis and distinctly a separate subject, related or not.

Consider that you discovered that, in the different areas under discussion, the male/female ratios were directly related to data concerning winter temperatures and the abilities of various population segments to heat their homes to a specific degree. One might be tempted to equate one with the other--to account for the ratio inequity on the basis of physical data, to "explain" the practices of the people as due to the climactic conditions. But that would be hasty. A comparison with other populations subject to similar climactic conditions but not exhibiting the same gender imbalance would suggest that it not climate per se which was causative of the particular pop stats but rather the MEANING attached by those people to the conditions encountered. Again, we'd be back to culture and the degree to which the culture had been influenced by the ideas of the man Confucius (and regardless of contributions of unidentified others to the set of ideas and behaviors called Confucian).

To the same (or very similar) stimuli, men react differently. Snowshoes for some, skis for others; carts and travois. Production vs. predation. It's the ideas that count. And some idea systems are summed and expressed as associated with specific individuals (Confucianism, Marxism, Christianity), whether or not "ragged around the edges."

The comments to this entry are closed.

Notes for Readers