For a few years now, there have been periodic mentions in the press of a study on the potential threat posed by China’s impending surplus of young, unmarried males. In these stories, they have been described as a “geopolitical time bomb”, “bachelor bomb” and a threat to democracy. The source of this theory is a paper written in 2002 by Andrea Den Boer and Valerie Hudson entitled “A Surplus of Men, A Deficit of Peace: Security and Sex Ratios in Asia’s Largest States”, later expanded to book form in Bare Branches: The Security Implications of Asia’s Surplus Male Population, published in 2004. The study always seemed suspect to me, and after reading some of the supporting literature, I have a few ideas why.
Hudson and Den Boer define “surplus males” differently from “single men”, though many news articles on the topic use these interchangeably. The term “bare branches” refers to the Chinese word guanggun (光棍 ), or men without marriage prospects. Surplus males are the unintended consequence of son preference, as there aren’t enough women to go around for all of these favorite sons to marry. A “single man”, in contrast, has prospects for dating, marriage and reproduction. Bare branches are simply left out of the marriage market because supply doesn’t meet demand. Den Boer and Hudson estimate the number of surplus males by counting the difference between the number of men and women aged 15-34. By this reckoning, according to the US Census International Database, China had nearly 15 million surplus males in 1990. Using the sex ratio at birth, they estimate this will reach 30 million by 2020 – possibly 40 million if more generous birth ratios are substituted. the same Census tables, I calculate that current the number of surplus males is 14 million and change.*
Den Boer then go on to argue that surplus males, according to overwhelming literature, tend to be more aggressive, violent and criminal. I don’t really have a problem with arguing that unmarried single men in their twenties are often a bunch of hotheads. But they cite a study by Allan Mazur and company stating that testosterone levels, very much correlated with aggression, are higher in unmarried and divorced men. What they do not mention, however, is that Mazur’s study focuses on a reciprocal model for testosterone rather than the widely used basal model. In other words, the idea that testosterone levels are affected by being divorced or unmarried, as opposed to a man’s naturally high testosterone level affecting his chances of being divorced or married. Mazur et al. say that while the reciprocal theory better explains short-term T increases in Air Force veterans after divorce, the basal theory better explains overall likelihood of divorce. Moreover, they point out that one would expect a concentration of high-T men in lower classes, but in fact they are evenly distributed, pointing to an “invisible” stream of prosocial high-T men.
Basically, there’s a chicken-and-the-egg problem (do men lack a partner because they have high testosterone that makes them antisocial, or do men become antisocial because they have no partner?), as well as evidence that high testosterone does not necessarily make one antisocial, aggressive or violent. Like I said, I personally find it believable that young unmarried men are more likely to make trouble, but Den Boer and Hudson peg their argument on an evolutionary psychology model that seems to raise doubts.
Then there is the argument that when bare branches congregate its bad news, for which they cite David Courtwright’s Violent Land:
Men who congregate with men tend to be more sensitive about status and reputation. Even if they are not intoxicated with drink or enraged by insult, they instinctively test one another, probing for signs of weakness. . . . disreputable, lower-class males . . . exercised much greater influence in bachelor communities like bunkhouses and mining camps. They both tempted and punished, for to fail to emulate their vices was to fail, in their own terms, to be a man.
While this certainly describes, say, a frat party, it seems awfully reductionist, as does the emphasis on testosterone. Men are reduced to hormone driven robots of sorts. Den Boer and Hudson do not address factors such as family, economics, political events, moral codes, social policies or organizations, or indeed anything else that might affect the behavior of these men. I can’t help but be reminded of the dehumanizing labeling of child offenders as “superpredators” in the 1990s, or jokes about how a woman president would start a nuclear war during her time of the month.
They go on to say “It is possible that this intrasocietal violence may have intersocietal consequences as well.” This is where Den Boer and Hudson make the leap from an increase in crime due to surplus males to the supposition that it could increase the chances of war or other major forms of conflict. They give no citations, no footnotes and no supporting literature for this assertion. Instead, they then qualify their statement:
It is important to note that we are not claiming that the presence of significant numbers of bare branches causes violence; violence can be found in all societies, regardless of sex ratio. Indeed, to give but one example, the sex ratio of Rwanda in 1994 was normal. Rather the opportunity for such violence to emerge and become relatively large-scale is heightened by socially prevalent selection for bare branches. We see this factor as having an amplifying or aggravating effect. To use a natural metaphor, the presence of dry, bare branches cannot cause are in and of itself, but when the sparks begin to fly, those bare branches provide kindling sufficient to turn the sparks into a fire larger and more dangerous than otherwise.
While it is a lovely metaphor, this hardly seems a supportable claim. If they are simply stating that violence is “heightened” by the presence of bare branches, it begs the question: heightened compared to what? Remember that according to U.S. Census data, China has had about 15 million surplus males for at least the past 20 years. Shouldn’t we be seeing an increase in violence that can be demonstrably linked, or at least correlated to the male surplus population right now? There are the reports of increased mass demonstrations that are heard so often, but these have been increasing while the male surplus population has held steady, and there are plenty of other reasons for these protests (many of which involved women and older people), such as growing inequality, the rise of the Internet and cell phones, etc. etc. Plus the statistic itself isn’t at all reliable, as there are conflicting definitions of a “mass incident”. Den Boer and Hudson point to reports of high crime rates among migrant workers in China, who they believe share a large overlap with bare branches. But during this time the number of bare branches has remained steady. Other factors such as discrimination, lack of a social safety net, poor economic opportunities and the like can easily explain this. Do we even need an argument that testosterone is involved, and can this even be demonstrated?
Moreover, Den Boer and Hudson calculate it simply by subtracting the male population of 15-34 years of age from the corresponding female population. But there are other factors affecting the gender ratio. For one, prostitutes. Maureen Fan reports in the Washington Post that estimates of Chinese prostitutes range from 1 to 10 million. One estimate from 2001 mentioned by the Kinsey Institute says 3 million, which doesn’t seem unlikely. It doesn’t seem a stretch to assume that a majority of female sex workers will be in the 18-34 range, or that they are off the marriage market. Consider also that concubinage is not completely unknown at the moment and just a few years ago it was estimated that Hong Kong men alone have half a million children by mainland mistresses. So slap on another million, and you find China has been dealing with a surplus of nearly 20 million bare branches for at least a decade. Yet the effects and management of these surplus males through the years is not addressed at all in Den Boer and Hudson’s paper. Considering that China is hardly becoming instable right now due to violence, despite this surplus, one would assume that a) there is a tipping point at which the ratio becomes too imbalanced, which DB/H never suggest or b) there are some social or cultural factors at work that are mitigating the violence caused by surplus males. But DB/H explicitly state that the only way to reduce the violence is to reduce the number of surplus males, something that is not happening.
DB/H proceed to give three “suggestive” historical examples, one of which is the Nien Rebellion. A series of natural disasters led to widespread poverty and starvation in Huai-pei, which increased the level of infanticide and resulted in a sex ratio of 129 men to every 100 women, and as many as 25% of men were unable to marry at all. DB/H cite James L. Watson’s work on bachelor subcultures to point out that “unmarried men have little face to preserve because they do not command much respect in the community… these “bare sticks” had nothing to lose except their reputations for violence.” They claim that right now China is recreating “the vast army of bare sticks that plagued it during the nineteenth century”. But they quote Watson as also saying “most bare branches in his study were semiliterate and were third, fourth, or fifth sons whose families were too poor to offer them an inheritance. In many cases, these noninheriting sons were “pushed out” (tuei chu) of their fathers’ houses in their teens, and came to live in bachelor houses with groups of other unmarried youths. In their early twenties, they would move out of the bachelor house and in with a collective of men—a dormitory of workers, a monastery or religious brotherhood, or the local militia. In each case, they would spend much of their leisure time learning and practicing the martial arts.”
There are significant differences between this case study and modern China. First of all, the Nien Rebellion and the gender imbalance itself, by DB/H’s own description, were primarily the result of massive disasters and poverty. Whether the bare branches intensified the conflict seems unprovable – if you are married with a family you cannot feed, are you any less likely to join a rebel army in the hopes of providing for them? More importantly, in contemporary China, there are few 3rd, 4th or 5th sons. Most bare branches will be only or second children, and with China’s growing elderly population, it seems less likely they will be pushed out of the house to prove themselves. Rather, it seems quite likely they will be kept around the house to care for their aging parents, grandparents, aunts and uncles. If they migrate to work, they will still have parents dependent on them for support. Watson illustrates a clear role for social factors in the likelihood that these men will turn to violence – social factors that DB/H consider irrelevant later on.
And finally, monasteries and brotherhoods hardly play a role today like they did in the mid-nineteenth century. In later detailing some of the history of these groups, DB/H opine: “Given the high sex ratios of its society, perhaps the grave suspicion with which the current Chinese government views movements such as the Falun Gong is not entirely unfounded in light of this history.” The Chinese government probably does have groups like the Heaven and Earth Society in mind when they worry about Falun Gong, but it should be pointed out that Falun Gong has overwhelmingly appealed to the elderly, not angry young men. In this case it would be China’s growing elderly population that would be the source of trouble, not bare branches.
The most problematic assertion in DB/H’s paper is when they suggest that the presence of bare branches drives the development of authoritarian societies. Perhaps one could argue a surplus female population does too, since Hitler rose to power while Germany had around 2 million extra women. But here Den Boer and Hudson quote the work of Christian Mesquida and Neil Weiner: “Choice of political system made by the members of a population is somewhat restricted by the age composition of its male population.” However, they also quote Mesquida and Weiner’s conclusion:
“Our analyses of interstate and intrastate episodes of collective aggression since the 1960s indicate the existence of a consistent correlation between the ratio of males 15 to 29 years of age per 100 males 30 years of age and older, and the level of coalitional aggression as measured by the number of reported conflict related deaths.”
Mesquida and Weiner do indeed name a tipping point for major conflict, and allow for mitigating circumstances – unlike Den Boer and Hudson:
“Populations with ratios of young males exceeding 60 per I00 males 30+ are predicted to move toward a state of internal or external conflict, unless there exist particular mitigating circumstances such as an extremely rapid increase in resource availability or new possibilities to migrate to more productive environments.”
If one looks at U.S. Census data comparing the ratio of 15-34 year old males to 35+ year old males (weighing the ratio even more in Den Boer and Hudson’s favor than Mesquida and Weiner’s original age groupings) in China from 2000-2020, it is clearly steadily decreasing as China’s population grays. In 1996, the ratio of younger men to older was 104:100, which by Mesquida and Weiner’s scale ought to have provoked a major conflict. In 2005, the ratio was closer to 75:100. In 2020? 65:100. By Mesquida and Weiner’s argument, the chances of China’s sex ratio causing problems in the future are less, not more, than the past decade. Den Boer and Hudson’s references to Robert Wright’s The Moral Animal and Laura Betzig’s study of the link between despotism and polygyny (men hoarding wives and mistresses), since they infer that these works argue that gender imbalances are only governable by authoritarian regimes, when both seem only to suggest that there is a correlation between imbalanced sex ratios and authoritarianism – not that only one can deal with a preponderous of men.
Den Boer and Hudson’s prescribe “there is only one short-term strategy for dealing with a serious bare-branch problem: Reduce their numbers. There are several traditional ways to do so: Fight them, encourage their self-destruction, or export them.”
The assumption here is that non-spousal family obligations, non-violent forms of single male social engagement, increased economic opportunity and other social support mechanisms count for nothing. Surplus males cannot be made into productive members of society, and therefore must be eliminated. It seems grossly reductionist but is a natural conclusion given all the ignored factors previously mentioned.
Similar criticisms have also been raised before in the Chronicle of Higher Education.
One case study that Den Boer and Hudson don’t explore at all is the Chinese bachelor subculture that existed in nineteenth century America. According to Thinkquest, “In 1860, the sex ratio of males to females was already 19:1. In 1890, the ratio widened to 27:1. For more than half a century, the Chinese lived in, essentially, a bachelor society where the old men always outnumbered the young.” Chinese immigrants did indeed represent a disproportionate number of arrests and criminal activity, but there were such aggravating factors such as the racist violence perpetrated against them. Some have argued that Chinese were still less prone to violence despite mob violence against them. Moreover, there is a great deal of evidence that “”public officials and the general public” accepted [Chinese on Chinese] “interpersonal violence … [as] to a large extent a private matter””. Den Boer and Hudson could argue that this is an example of a government (though I wonder if they would characterize California officials as “authoritarian”) allowing bare branches to destroy one another. But one could just as easily suggest that with proper law enforcement and a lack of discrimination, these immigrants would not have been anywhere near as violent. But if one accepts that these men are unstoppably driven to be violent, as Den Boer and Hudson argue, then there wouldn’t be any reason to help them.
* It should be noted that I consulted the China Statistical Yearbook as well, but found that according to their, the 18-34 population actually has a tiny surplus of females, not males. Also perplexing was that, when totaled, the number of currently married women exceeded the number of married males by several million. To my knowledge there are nowhere near that many Chinese women married to foreigners (who remain Chinese citizens counted in the census). Anybody know why?
Dave,
Good post and this is a subject that has fascinated me for many years. I too wonder at the biological determinism in the 2004 study and I also think that the parallels between the 19th century rebellions and present day are far from exact.
That said, I do think that the guang’gun phenomenon should be considered, both as a partial explanation for the social instability of the late-Qing as well, as well as something to watch for the future.
I’ve mentioned him before, but anyone interested in the issue of guang’gun in Qing history, should check out the works of Matthew Sommer (Sex and Law in the Qing among others). Sommer’s work looks closely at the legal and social implications of the guang’gun in the mid- to late-Qing.
I would argue that while the outbreak of rebellions in the 19th century (Nian among them) cannot be blamed entirely on the presence of large groups of unmarried or underemployed males, it was this group that provided a ready pool of recruits once those rebellions started to roll.
I would agree with your assertion that China today is far more stable than in 1860, principally because, as you said, the sort of widespread destruction and economic dislocations (Taiping Rebellion, foreign invasion, floods, silting of the Grand Canal) associated with the mid-19th century have no close parallels in the present day. Furthermore, in the 19th century, underemployment was as much a factor as being unmarried. Today’s go-go China siphons many single and underemployed males and redeploys them in the booming cities.
That said, the stability of rural China is hardly something to take for granted, as the number of mass incidents suggests.
I actually like the kindling metaphor, but I’ve a better one: There’s no guarantee a person with high blood pressure will have a heart attack, but it is more likely, especially if the problem is allowed to continue without treatment and monitoring. A large number of unmarried/economically frustrated males doesn’t mean unrest will necessarily occur, but when I’m studying rebellions or social instability, it’s one of the things (among many) that I’m looking for in my research/diagnosis.
Anyway, it’s a great post. Thanks for the thoughtful review and critique on this important subject and Happy Halloween.
@J: I agree the phenomenon should be considered, and I like your high blood pressure analogy. It’s not that I think the kindling metaphor is wrong; its that I think it loses alot of punch because of the sort of argument that Den Boer and Hudson make, not that the topic isn’t worth looking at. As I point out at the end, Chinese bachelor subculture in the U.S. would be a good area to explore, especially since some studies indicate that Chinese violence in California was often over women.
There have been surplus males for a while now, and somehow they are being managed without being destroyed or exported, which are the sole options DB/H offer. As for “mass incidents”, I really recommend you check out the link to ESWN where he details how those statistics come about. I’d add the additional caveat that I’m extremely suspicious that “mass incidents” went unreported in previous decades, either to the central government itself or to the general public.
Finally, the most damning part of my critique, for me, is that they mischaracterize Mesquida and Weiner (although Weiner endorsed their study). My reading of M&W’s work predicts that China is less likely to have internal or external conflict because of the ratio of younger to older males. If I’m correct, then this is a case of bad scholarship in order to reach a predetermined and rather sensational conclusion.
Dave,
Agreed that “mass incidents” should hardly be codeword for a society on the verge of collapse. But I do believe that rural society in China is far from stable. Whether its getting worse or better is a more debateable topic.
And I agree with you that sensationalizing the gender imbalance obscures the complexity of the situation on the ground. This is, however, a situation that bears watching in the future especially for unpleasant side effects beyond “simple” male aggression (prostitution, trafficking in women, domestic violence.)
@J: Prostitution is an interesting one. An increase in bare branches could mean an increase in the demand for prostitutes. But not necessarily, especially since in China married men frequent prostitutes. It’d be worth finding some numbers breaking down the typical Chinese john. Again, it might be that social factors explain the demand for prostitution far better than the number of unmarried men.
But let’s say that an increase in unmarried men does lead to increased demand for prostitutes. This would create a bad feedback loop, since unmarried men are less likely to marry a woman with experiences as a prostitute, either because she can’t get married while working, is unwilling to marry an unemployed man whose prospects are even dimmer than her own, or because they view her as damaged goods. This would mean that the number of available brides would decrease and thereby increase the number of bachelors. As I point out above, a head count reveals the current surplus of men to be 14 million, but when you factor in unavailability of brides due to prostitution, the number of bachelors could reach as high as 24 million (using the most generous estimate).
Again, however, social factors could play a role. Increasing bride shortages could provoke the government to begin a massive campaign to “reclaim” prostitutes, perhaps through some sort of rehabilitation/relocation program. It would prevent an increase in bachelors beyond the fundamental imbalance in gender ratios.
Transnational trafficking, on the other hand, is something I think would certainly increase. An interesting byproduct of this could be that Chinese racial attitudes would become much more visible.
There’s tons of interesting things to talk about when it comes to this issue. It’s a pity that Den Boer and Hudson skipped most of it to make a facile and weak argument about democracy.
Great analytical post.
There is the prospect that the gender imbalance will be leveled by the fact that many Chinese in cities won’t get married due to the fact that they’ll be too busy working 60 hours a week or playing computer games. There is also the possible rise in cancer rates (and other life-threatening diseases), brought on by the pollution in many parts of China, decreasing some of that population. And if those don’t come about, there will probably be a much larger gay community.
@Matthew: Whether they are working 60 hours a week or not depends on unemployment. DB/H believe that the overlap between bare branches and the indigent will be substantial, if not virtually identical. But I don’t think they make a good case for it in their article. They imply that the marriage market will select for the wealthiest men, leaving the poor men in the surplus population. But they never bother to document that this indeed the case today with the current surplus.
As for the gay community, I don’t quite see surplus males = more gay guys. But its possible that being as being unmarried is a more common occurrence, being gay (and not marrying a woman) may become more socially acceptable. There’s so many different directions the society can go in response to the imbalance.
China does have much better economic prospects overall than it did in the 19th and early 20th century, but humans are human and men are men.
There are more and more confrontations between local younger men and foreigners in the presence of a local girl and it isn’t some nationalist thing. Chinese girls, or any girls for that matter will turn down the lowest common denominator of man.
And these bare branches aren’t just poor laborers, they are typically the least educated, the least physically and mentally capable and they probably don’t come off as “good looking” either.
That being said, even in SH and BJ all you know what any group of migrant laborers are saying to local girls in the presence of foreign men and it isn’t the typical “hey baby” cat calls.
The shortage is magnified by the fact that so many Chinese girls are marrying foreign men (even thoooose Japanese) and tens of thousands of baby girls being adopted by foreigners every year.
The Chinese gov’t is aware of this problem and you can some attempts to defuse the situation by sending legions of Chinese laborers to the work sites of Chinese companies around the world.
There are also diplomatic problems as SE Asian girls working in China are being kidnapped for marriage to the sons of local bosses.
There may not be a united uprising, but sporadic attacks against foreign men or Chinese girls known to associate with foreign men.
An increase in homosexuality is a reality as well. Men gotta have sex, and just like prison if there are no women they will go after each other for companionship and satisfaction.
One point that doesn’t seem to have come up in the discussion here so far is that gender imbalances in movements of population around the country are likely to have far more impact on localised concentrations of (young) surplus males than any imbalance in the birth rate. Concentrations of excess men in the cities in China is largely the result of migrants from the countryside seeking work (not that there aren’t women migrants too; but I think male migrants are the massive majority); while many rural areas are left with a huge surplus of women.
I don’t buy that the number of prostitutes will increase the scale of the surplus male ‘problem’. That problem is seen to arise from sexual frustration, not from an inability to marry (although it is measured by the number of males deemed unable to marry); and a prostitute can (usually) keep far more men reasonably sexually satisfied than a wife does. Moreover, I’d guess that most prostitutes are eventually able to “put their past behind them” and find a husband.
Homosexuality? Yep, that always increases in single-gender environments. (I thought you might have been alluding to that obliquely with the phrase “non-violent forms of single male social engagement”!) Anecdotal evidence suggests that the rates for male homosexuality (at least amongst the urban middle class) are already extraordinarily high, even among married men. And I’d imagine it’s pretty common in the construction camps as well.
@Froog: I don’t know if surplus males will result in more prostitutes. But what I am saying for certain is that an increase in prostitutes will result in more surplus males, because it will reduce the bride supply.
I doubt that, Dave. Their entry into the bride market might be deferred, but I think it’s rarely precluded forever.
And my point was that the alleged social problem regarding ‘surplus males’ is reckoned to arise not from their mere numbers, but from their sexual frustration. Widespread prostitution might increase the surplus number slightly, but it’s doing wonders to relieve the frustration.
@Froog: “And my point was that the alleged social problem regarding ‘surplus males’ is reckoned to arise not from their mere numbers, but from their sexual frustration.”
That may well be the case, but that is absolutely not what Den Boer and Hudson argue in their paper, though others do. Den Boer and Hudson speak only of marriage, and make no claim that prostitution will ease any of the pressure. Walker’s “The Geopolitics of Sexual Frustration” makes this case, I believe, but not these two scholars. They suggest that marriage has some ineffable quality that reduces testosterone and hence aggression, keeping men off the streets. They make no suggestion whatsoever that regular sex from non-spouses will affect the bare branches problem at all.
As for prostitutes rejoining the marriage market, I’m sure many will, and it will be more likely as gender ratios become increasingly imbalanced. But my point was that while they’re working, it seems pretty unlikely.
But yeah, many will be “reclaimed”, and one of my points in the comments above was that you may even see Chinese society (and/or the government) change its attitude towards prostitution (i.e. become less tolerant of it) as the lack of brides is increasingly felt.
@Froog: just to reiterate in case my last comment wasn’t very clear, while prostitutes may re-enter the bride market, during the lag between when they enter and exit the prostitution trade the number of available brides go down. This could mean periods of months or years in which the adjusted surplus of men is increased substantially.
Those factory towns in Eastern China staffed almost entirely by women who aren’t allowed to leave the factory grounds or have any opportunity to meet any males outside the gates would also be a contributing factor as well. Sure, they’ll leave at some point and meet a guy. But again, it could create periods of artificial shortages in the marriage market. Distribution of males and females is a key element as well, since simply counting the numbers of each does you no good if all the women are in Guangdong sewing buttons and all the men are in Shanghai building office towers. Conversely, some areas of China may not have a gender imbalance in either direction and wonder what the fuss is about.
I would like to point out that even if more unmarried men lead to intrasocial instability, whether this will lead to intersocial violence is not clear. At least if you consider inter-state organized violence, the decision makers of such endeavors are typically not unmarried men, but rather the political elite, who probably have wives plus mistresses a plenty. This takes a lot of the “yellow peril” sensationalism out of the whole topic.
The discrepancy in the number of married males and females is very curious.
I wonder if government stats fail to re-classify people after the death of a spouse? Women tend to live longer than men, so that ‘surplus wives’ figure might be largely composed of widows.
And if that figure comes from voluntary poll samples rather than comprehensive government statistics, it might just be a case of separated, divorced, or widowed women (and maybe unmarried women too) being more likely to describe themselves as still ‘married’. It’s an old adage that a single woman will often fib that she has a partner; whereas a man is more likely to do the reverse.
Is polygamy ever sanctioned under Chinese law? I’d never heard of it, but it does seem plausible that it might be allowed for certain ethnic or religious groups.
@Froog: It’s possible that the online version of the CSY2006 that I used is wrong. If you go there, the table of contents isn’t linked to anything, though the pages are on the server. Not exactly a pro job.
Anyway, the way the 15-34 pop is split 50-50 male/female, I think they’ve been very creative to meet some in-house goals and led to some unintended silly numbers.
I say, let’s go global. Not enough CHINESE girls? where is the problem? Isn’t it about 50/50 globally??
I am sure I am not nearly as sophisticated and thoroughly informed as most of you.
Just thought though, maybe if we can look past the nationalities and boundaries, things could be a lot simpler in this world, or maybe cultural understanding draws its line at “I have been somewhere.”
great blog…one factor not consider is the role of the Chinese mother and Chinese grandmother in guiding and influencing her son and grandson……
Remember the hand that rocks the cradle rules the world….especially true in the Chinese world…..vastly different than American mothers and their place of low honor and low influence in their male children’s lives and their grandsons. The Chinese mother son relationship is absolutely pivotal and future defining regardless of educational and economical status.