China’s Gender Imbalance and its Economic Performance

Jane GOLLEY

Centre for China in the World
Australian National University

 

Rod TYERS

Business School
University of Western Australia, and
Research School of Economics
Australian National University

August 2012

Key words:
China, growth, demography, sex ratio, imbalance

 

JEL codes:
D58, J11, J13, J16, J21

 

Abstract

Chinese GDP growth faces rising challenges that include the slowdown and eventual contraction of its labour force, a complication of which is its rising sex ratio at birth.  The undesirable consequences of the resulting gender imbalance include excessive savings as families with boys compete to match their sons with scarce girls, trafficking in women and rising disaffection and crime amongst the low-skill unmarried male population.  These consequences are reviewed and analysed using a dynamic model of both economic and demographic behaviour.  The results show that the proportion of unmatched low-skill males of reproductive age could be as high as one in four by 2030.  Policies to rebalance the sex ratio at birth will take decades to reduce the sex ratio at reproductive age and any associated allowance for higher fertility would slow growth in real per capita income.  Yet the results suggest that the beneficial effects of reduced male disaffection and crime could outweigh the losses from reduced saving and higher population.

 

1. Introduction

According to the United Nation’s (UN) Population Prospects 2010 revision, China’s sex ratio at birth (SRB) reached 120 (male births per 100 female births) in 2005-10, compared with a world average of 107.  This earned China first place on the global ranking of imbalanced sex ratios at birth, a rank it has held since the mid-1980s when the SRB first moved into the “abnormal range” (see Figure 1). Sub-national figures are even more alarming, with the 1% inter-census survey in 2005 revealing six provinces that recorded sex ratios over 130 for the 1-4 age group, and nine provinces that exceeded 160 for second-order births (Zhu et al., 2009).  Compounding China’s gender imbalances are its rates of excess female child mortality (EFCM), which are measured by comparing the normal ratio of male to female child mortality (of between 120 and 130 for infants between birth and age one) with the observed ratios.  In 2005 these were “severely abnormal” at 80 for children in the first year of life and 84 for children in their second year (Li, 2007).  While women in China, as elsewhere, have higher life expectancies than men, the combination of high SRBs and high EFCM has left China with the world’s highest sex ratio in the total population, at 108 and rising in 2005-10, compared with the global average of 101 (United Nations, 2010).

These gender imbalances have resulted in substantial numbers of “missing women”, a term coined by Amartya Sen in the late 1980s, based on his own research indicating that more than 100 million women were missing worldwide[1].  Klasen and Wink (2003) review the range of estimation techniques used since Sen, and show that the number of women missing in absolute terms has continued to increase worldwide, while the percentage in terms of women alive has fallen.  However, in China there has been further deterioration in both absolute and percentage terms, with an estimated 41 million missing women in 2000 or 6.7 per cent of the female population (very similar to those of Li, 2007 and Bulte et al., 2011).

According to demographer Zeng Yi (2007), if China’s SRB remains constant and current fertility policy is unchanged, the proportion of excess men aged 20-49 compared to women of the same age will increase from 6% in 2010 to 11% in 2030, peaking at over 14% in 2050. Li and Jiang (2005) project that if SRBs remain at 2000 levels, by 2030 the population size will be 84.2 per cent of what it would be under normal SRB patterns (in the range of 104-107).  That is, 15.8% of the population would disappear because of the number of women missing, and the proportion of excess males would reach 20.7 per cent!  As with most phenomena relating to China, the issue of rising gender imbalances is taking place on an unprecedented scale, with consequences that extend into all realms of economic, social, and political life.

China’s rising gender imbalances are occurring in tandem with the slowdown of China’s labour force growth, an issue that has received much attention from economists in recent years (Golley and Tyers 2012a,b, Cai, 2012).  By itself, this slowdown brings forward the Lewis “turning point” beyond which the rural to urban transformation proceeds more slowly, constraining future growth.[2]  Rising gender imbalances compound this slowdown since, for given fertility levels, the declining share of women further reduces the growth rates of the population and hence the labour force. This is just one of a number of undesirable consequences of gender mismatch, others of which centre on the rising proportion of unmatched males, including excessive saving as families with boys compete to match their sons with scarce girls (Wei and Zhang 2009, 2010), trafficking in women and increased rates of crime (Edlund et al. 2007, Hvistendahl 2011).

In this paper we review prior work on the causes and consequences of China’s high SRB and offer a prospective quantification of its effects via simulations using a dynamic model of both economic and demographic behaviour.  The baseline simulations show that the number of unmatched males could achieve extreme proportions by 2030, with the proportion of unmatched low-skill males of reproductive age being as high as one in four.  The results suggest that an associated crime-induced productivity slowdown would outweigh the increased savings of families with boys and so further retard growth, contributing to the yet-avoidable possibility that China could fall into a “middle income trap” (Easterly 2000, World Bank 2010, Eichengreen et al. 2011).  The economic impacts of policies to rebalance the sex ratio at birth will take time, and not all of them will necessarily be positive. Yet the results suggest that the beneficial effects of reduced male disaffection and crime could outweigh the losses from reduced saving and higher population.

The section to follow reviews the growing literature on China’s rising SRB and the economic implications of gender imbalances more generally.  Section 3 then summarises the simulation model we use, while Section 4 examines the demographic consequences of the gender imbalance and its dependence on trends in the SRB.  Section 5 associates the gender mismatch with both aggregate household savings and the productivity effects of low-income male disaffection and crime.  Conclusions are offered in Section 6.

 

2. China’s rising gender imbalances: causes, consequences and prospects

Guilmoto (2009) examines the causes of sex imbalances at birth in a range of Asian countries, including China and extending to Azerbaijan, Armenia, Albania and Georgia.[3] Beginning with the observation that son preference is widespread (not only in Asia, but just about everywhere), he identifies three further motivations for deliberate sex selection: “fertility squeeze” (brought about by parents wanting or needing to limit the number of births), “ability” to limit those births (from traditional methods for dealing with unwanted girls through to high tech methods including ultrasound gender identification) and “readiness”, which includes the social and legal circumstances that allow parents to take advantage of the birth limiting options available to them.

China clearly checks all the boxes, with a long-standing cultural preference for sons, the introduction of the one-child policy in the early 1980s and the widespread use of Ultrasound B technology to detect gender from the mid-1980s onwards (Ebenstein, Li and Meng, 2010). For Li (2007), the fundamental cause of China’s gender imbalance is essentially cultural, tracing to the patrilineal system of the Han Chinese.[4] As he points out, there are numerous socioeconomic factors that help to sustain the traditional preference for sons, including the need for parents to have sons who will support them in retirement, and unequal rights and opportunities for women in terms of education and employment.[5] This point has been acknowledged by Wang Xia, head of China’s National Population and Family Planning Commission (NPFPC) who stated in 2012 that “The bias against females in economic, social and cultural fields is still the root cause of the current gender imbalance”.[6]

While some research has pointed to the under-reporting of female births in China (which would give a false impression of a large number of ‘missing’ girls who are in fact just ‘hidden’), the bulk of evidence indicates that sex selective abortion is primarily responsible for the rising SRB in recent years and that this rise is indeed very real (Cai and Lavely, 2007, Li, 2007, Ebenstein, 2008 and Zhu et al. 2009).  This view appears to have been accepted by the Chinese government, with Wang Xia announcing in 2012 that “the authorities will crack down further on illegal prenatal gender tests and selective abortions, which are believed to be the primary causes of the gender imbalance”.[7]

Although strict family planning policies are not a prerequisite for fertility squeeze, they are surely a contributing factor.[8]  Based on the 2000 census, Gu et al. (2007) show that China has implemented one-child policies, 1.5 child policies, two-child policies and three-child policies in different parts of the country, giving rise to SRB levels of 112, 125, 109 and 198, respectively.  This indicates that two-child policies bring SRBs down to almost normal levels, while the 1.5-child policy, which allows for a second child in rural areas if the first one is a girl, has been even more detrimental to China’s SRB than the one-child policy itself.

While Chinese officials are reticent to attribute rising gender imbalances to their family planning policies, they are well aware that the associated problems are looming large. In 2004, following an announcement by the then head of China’s NPFPC, Zhang Weiqing, that males of marriage age would outnumber females by 30 million by 2020, Li Weixiong, vice-chairman of the Population, Resources and Environment committee of the National Committee of the Chinese People’s Political Consultative Conference (CPPCC) stated that “Such serious gender disproportion poses a major threat to the healthy, harmonious and sustainable growth of the nation’s population and would trigger such crimes and social problems as mercenary marriage, abduction of women and prostitution.” [9] Similarly, Hudson and den Boer explicitly link rising gender imbalances and domestic social conflict in their 2004 book entitled “Bare Branches: The Security Implications of Asia’s Surplus Male Population” (“bare branches” being the Chinese term used to describe unmarried men without children) and in another 2008 article in which they argue that “a society with a masculinized young adult population, such as China’s, is likely to respond to significant economic hardship with severe domestic instability and crime”. Edlund et al. (2007, 2010) show that a 0.01 increase in the sex ratio caused a three per cent increase in property and violent crimes in China between 1988 and 2004, indicating that the rise in surplus males may account for up to one-sixth of the overall rise in crime during this period.  While this does not validate all of Hudson and den Boer’s claims (some of which are alarmist and speculative), it does indicate that the domestic costs of China’s gender imbalances could be substantial.

There may also be numerous indirect costs associated with having a growing number of unmarried men in Chinese society.  Anthropological studies have shown that men in societies with large numbers of surplus men will engage in non-productive and risky “wife-seeking” behaviour, sacrificing their own productivity and paternal investments that would have raised their children’s future productivity as well (Henrich et al., 2012). In addition to these productivity losses, unmarried men also suffer from poorer physical and psychological health, with one recent study showing that unmarried Chinese men are 11% less likely to describe themselves as being in good health than married men (Ebenstein and Sharygin, 2009, Zhou et al. 2011).

Korenman and Neumark (1991) point out that if marriage really does make men more productive, as their analysis of US data suggests, then “changes in marital status composition potentially can affect the productivity of the labor force” (p. 283).[10]  However, it is not obvious how these changes will play out in aggregate in the Chinese context. Edlund et al’s (2010) evidence that higher sex ratios in China are associated with higher educational attainment and wages for men on average, and lower educational attainment for women on average, supports the notion that men may invest more heavily in characteristics to make them competitive in the marriage market, while women may choose to invest less.  They also show, however, that higher sex ratios increase the variance of men’s labour market outcomes, which they put down to a “hetereogenous incentive effect where men with low initial endowments choose to engage in crime” (p.17).  Thus it seems clear that the impact of rising sex ratios varies significantly across different members of society, both within and between the sexes. However, it is unclear what the aggregate impact will be in terms of productivity, wages, educational attainment, and other factors that impact on long-term growth performance.

Savings is one factor for which the aggregate impact is clear, at least according to a recent string of papers that links China’s rising sex ratio to its rising saving rates. Wei and Zhang (2009) show that the rise in China’s sex ratio explains half of the increase in Chinese household savings as a share of disposable income (from 16 per cent in 1990 to 30 per cent in 2007), consistent with Du and Wei’s (2010) theory of “male excess savings”, which arises in the context of a steady increase in the ratio of men to women as men seek to be competitive in the marriage market. Wei and Zhang (2010) also show that the sex ratio imbalance stimulates entrepreneurial activities as men are driven to increase their wealth, providing evidence that men in regions with relatively high gender imbalances are more willing to accept dangerous and unpleasant jobs and supply more hours of work.  This boosts private sector growth, suggesting that not only household, but also corporate savings are higher as a consequence.  Du and Wei (2012) describe further how the increase in savings and the expansion of the supply of male workers both put downward pressure on the real exchange rate (assuming that non-tradables are more labour intensive in China and that tradable prices are determined in world markets).  This mechanism feeds into Du and Wei’s (2010) overlapping generation model, which they use to demonstrate the link between China’s gender imbalances and the current account surpluses and deficits of China and the United States respectively, and which leads them to conclude that although “The sex ratio imbalance is not the sole reason for global imbalances, it could be one of the significant, and yet thus far unrecognised, factors” (page 2).

Given the potentially gargantuan impacts that China’s gender imbalances could have on the domestic and global economies, it is worth considering whether they are likely to be self-correcting, through evolutionary, economic, social or other means.  Evolutionary biologists have proposed an “adaptive sex ratio adjustment hypothesis”, in which healthy, well-nourished and high-status mothers are more likely to give birth to sons, whereas unhealthy, poorly nourished and low-status mothers are more likely to give birth to daughters. Song (2012) finds evidence of this hypothesis in the abrupt decline in China’s sex ratio at birth during the Great Leap Forward famine.  Rob Brooks (2012) explains how this hypothesis can bring about a more balanced sex ratio in societies where hypergamy[11] is common, as high-status (rich) mothers give birth to more sons who match up with the relatively large number of daughters born to low-status (poor) mothers.  However, the ready availability of ultrasound technology throughout China, combined with strong son preference, suggests that this kind of balance is unlikely to emerge naturally, short of another serious famine, which is hardly a solution worth counting on. Most critically, the concentration of unmatched males in lower socioeconomic groups may have serious social consequences, an idea we explore further below.

Becker (2007) places his trust in the market, arguing that “As children become adults in cohorts with a high ratio of boys, the advantage of girls and women increases since they are scarcer” so it is men and not women at a disadvantage as the value of women rises and value of men falls.  Qian (2008) finds some evidence of this in her analysis that shows that a rising sex ratio leads to higher female income and higher survival rates among girls.  She estimates that if annual rural household income could be augmented by 10% and given entirely to the household’s women, the fraction of girls would rise by 1.3 percentage points.  Becker goes on to defend sex selection of births because the market will correct for it in the long run, essentially advocating a laissez faire approach to China’s gender problem.[12] While other scholars recognise a range of socioeconomic forces that tend to level out sex ratios over time, Murphy et al. (2011) show that these are having very little impact in rural China to date because of female disadvantages in land rights, wages and education. [13]  In other words, the long-run market solution to gender imbalances has not emerged just yet.

The Chinese government evidently has less faith in the market than Becker, and has recently identified reducing the SRB as a national priority, aiming for 115 newborn males for every 100 females by 2015.[14] Numerous efforts to improve women’s rights and promote gender equality with a view to lowering the SRB and improving girl survival rates have already been undertaken, including the “Care for Girls” campaign, first introduced in the Chaohu Experimental Zone in 2000, and then in the entire country from the beginning of 2006 (Li, 2007).  Officials have credited the fall in the official SRB between 2009 and 2011, from 119.5 to 117.8, to their recent crackdowns on illegal prenatal gender tests and selective abortions, while acknowledging that enhanced efforts to promote equal opportunities and the social status of females are fundamental solutions to the problem.[15] These points notwithstanding, it remains to be seen whether, and how, the 2015 SRB target will be achieved.

One option being discussed, by academics at least, is to relax the current fertility policy. Zeng (2007) argues that the economic and social costs of maintaining China’s current fertility policy (which he refers to as a 1.5-child policy) are too high, with the consequences of the sex ratio imbalance being prominent among those costs. Compared with the worst-case scenario of a constant SRB and “current policy unchanged”, his preferred policy of a “two-children with late childbearing soft-landing”[16] would see the SRB normalise by 2030, with the number of excess men peaking in 2040 and returning to normal by 2050. While the official stance regarding current fertility policy indicates that such a policy relaxation is not yet supported,[17] these results are indicative of the potential impact that alternative fertility policies could have on China’s gender problem.

Finally, it is worth noting that large variations in the broader sex ratios of human populations are not uncommon in history, with low values occurring when disproportionate numbers of males have been lost in wars and other disasters.  Extraordinarily high values have also occurred when humans have expanded into new territories, including in Australia and California during the gold rushes of the 19th Century.[18]  In these cases, gender imbalances were righted by subsequent female immigration from regions with much larger populations.  This is unlikely to provide a solution for China, however, given the sheer size of the problem.

3. Modelling the impact of gender imbalances on economic performance

Given ongoing debates regarding both the causes and consequences of China’s rising gender imbalances, in combination with uncertainty about the pace at which nature and/or nurture might reduce these imbalances, it is clearly impossible to predict China’s gender structure in the future, let alone what the economic impacts of that structure will be. Instead, we turn to a modelling exercise aims to shed light on some, although admittedly not all, of the issues introduced in Section 2.

The economic model we use is a development of GTAP-Dynamic, the standard version of which has single households in each region and therefore no demographic structure.  The approach adopted here follows Tyers and Shi (2007), in that a complete demographic sub-model is integrated within a dynamic numerical model of the global economy. The model enables us to project changes since 1997 in population and labour force levels, gender imbalances, dependency ratios, real GDP and wage levels and the levels of real per capita income in the countries represented through to 2030.

Each region represented in the model includes four age groups, two genders, and two skill categories, for a total of 16 population groups in each of 18 regions, two of which are Mainland China and India.  The four age groups are the dependent young, adults of fertile and working age, older working adults, and the mostly retired over 60s.  The skill subdivision is between households that provide unskilled (production) labour and those that provide skilled (professional) labour.[19]  Each age-gender-skill group is a homogeneous subpopulation with group-specific birth and death rates and rates of both immigration and emigration.  If the group spans T years, the survival rate to the next age group is the fraction 1/T of its population, after group-specific deaths have been removed and its population has been adjusted for net migration.

The final age group (60+) has duration equal to measured life expectancy at 60, which varies across genders and regions.  So the narrowly demographic parameters are birth rates, sex ratios at birth, age-gender specific death, immigration and emigration rates, and life expectancies at 60.[20]  The birth rates, sex ratios at birth, life expectancy at 60, and age-specific death rates all trend through time, approaching exogenous targets asymptotically.

The economic component of the model considers each region to contribute seven industries: agriculture, light manufacturing, heavy manufacturing, metals, energy, minerals and services.  To reflect compositional differences between regions and failures of the “law of one price”, these products are differentiated by region of origin, meaning that the “light manufactures” produced in one region are not the same as those produced in others.  Consumers and firms purchasing intermediates substitute imperfectly between manufactures from different regions.[21]  Regions have open capital accounts with empirically-based investment biases and trade distortions.

To capture the full effects of demographic change, the multiple age, gender, and skill groups are modelled as separate households.  These 16 groups differ in their shares of regional disposable income, consumption preferences, saving rates, and their labour supply behaviour.  While the consumption-savings choice therefore differs for each age-gender group, it is dependent for all on current group-specific real per capita disposable income and the real lending rate.[22]

As in other dynamic models of the global economy, the main endogenous driver of simulated economic growth is physical capital accumulation.[23]  Technical change is introduced in the form of exogenous productivity growth that is sector and factor specific.  The overall rate of economic growth proves quite sensitive to these exogenous shocks since the larger these are for a particular region the larger is that region’s marginal product of capital.  The region therefore enjoys higher levels of investment and hence higher marginal products of labour and real wages.  This causes a double boost to its per capita real income growth rate.  For China the continuing productivity shocks are at levels that are high compared with the other regions.

The model has the Solow-Swan property, shared with all neoclassical dynamic models that embody diminishing returns to factor use, that an increase in the growth rate of the population raises the growth rate of real GDP but reduces the level of real per capita income.  What distinguishes it from its simpler progenitors is its multiple households, the endogeneity of age specific saving rates, its multiregional structure and its open capital accounts, which allow collective regional households to hold claims on capital abroad.  The demographic behaviour allows regional average saving rates to respond to changes in age distributions.  As a young population ages, the proportion of its population in the saving age groups rises and so therefore does its average saving rate.

In older populations, further aging raises the proportion of non-working aged, and so its average saving rate tends to fall.  The Chinese 60+ age group is unusual, however, in that it has had low labour force participation but high state-financed retirement incomes.  The pensions, combined with low consumption expenditures due to extended family sharing, lead to high initial retiree saving rates.  The result is flat age-specific saving behaviour and hence not much decline in China’s average saving rate as its projected population ages.  In the analysis that follows, however, we impose exogenous shocks to China’s saving behaviour that are in the range suggested by Wei and Zhang (2009).

 

4. The demographic significance of China’s gender imbalances

There have been a number of valuable, though purely demographic, analyses of China’s gender imbalances, including those by Li and Jiang (2005) and Zeng (2007) cited above.  In this section we cover some of the same ground but with two differences.  First we are more pessimistic than the official view of China’s fertility and its future population growth, assuming that its total fertility rate will decline to 1.2 by 2030.  We choose this as a baseline fertility projection because we are now convinced by the analysis of Zhao (2011) and Zhao and Chen (2011) that even the United Nation’s (2010) medium fertility variant for China is too high.[24]

Second, the model’s disaggregation of the overall population between households dependent on skilled (professional) and unskilled (production) occupations enables us to move beyond the standard measure of the share of unmatched males as the difference between the male and female populations, expressed as a proportion of the male population.  In particular, we calculate four alternative shares: 1) the share of unmatched boys 0-14, which is a signal to parents that there will be competition for partners for their sons, 2) the share of unmatched low-skill males of reproductive age, if marriages are restricted to partnerships between people of the same socioeconomic status (i.e., no hypergamy), 3) the share of unmatched low-skill males if women from low-skill families choose to marry up (i.e., hypergamy), ensuring that there are no unmatched skilled men, and 4) the share of unmatched low-skill males if sufficient numbers of low-skill women choose to be second partners of high-skill men so as to allow these men to have 1.5 partners on average.  While the fourth share offers a socially uncomfortable prospect, there is ample anecdotal evidence for women from low-income households preferring second partnerships with rich men over exclusive marriages to poor men.  The number is clearly larger than unity, though our choice of 1.5 is arbitrary and illustrative.

Alternative SRB Scenarios

Recent studies suggest that China’s SRB has already begun to decline in some provinces, while the rate of increase has slowed in others (das Gupta et al. 2009).  That the growth of the SRB is slowing is also supported by population census and 1% population sample survey data (Figure 2).  We therefore choose a baseline scenario that has the SRB converging upon a level not far above that already reached, as shown in Figure 3. Our comparatively pessimistic baseline path for Chinese fertility, combined with the SRB path shown in Figure 3, yields the paths for population components shown in Figure 4.  Note that the total population peaks in the current decade and then declines, as do both the total male and female populations.  The skilled population rises continuously, while the unskilled population begins declining sooner and falls more sharply than the total population.  Under this scenario then, the effects of population slow down and decline are already in train.[25]

Scenario 2 has policy actions (regarding fertility and gender discrimination in particular), combined with endogenous preference changes, bringing about a return of the SRB to 106 by 2030, also shown in Figure 3.  We also consider two counterfactual scenarios. In Scenario 3, the SRB remains constant at 106 throughout the period to indicate how different China’s demography and economy might have looked had the gender problem not arisen. Scenario 4 makes this same assumption for the SRB, but in addition assumes an average fertility rate for China that is higher than in the baseline. The purpose of this scenario is to explore the additional effect of using some relaxation of fertility restrictions as part of a policy arsenal to reduce the SRB.  It is unclear how much China’s fertility restrictions actually constrain birth choices so we make the arbitrary choice to allow a return to the United Nations’ medium fertility variant in this fourth scenario, which sees an average fertility rate of 1.8 in 2030 (up from our baseline level of 1.2).

The key demographic outcomes for 2030 based on these four scenarios are summarised in Table 1.  The baseline results reveal the extent of China’s gender problem implied by a “no policy change” stance: 27.7 million (or 11.6%) of males between the ages of 15 and 39 would be unmatched, or ‘excess’, by 2030. This percentage is very similar to the 11% estimate of Zeng (2007), even though his estimate is based on the 20-49 age group.  In the worst-case (and least palatable) scenario in which skilled men take 1.5 women, there would be 47.8 million unmatched unskilled males, indicating that the social stability and crime-related problems described by Hudson and den Boer (2004, 2008) and Edlund et al. (2010) could be very substantial indeed.

The large proportion of low-skilled Chinese men of reproductive age that is projected to be unmatched is placed in an East Asian context in Figure 6.  This figure shows our baseline projection of the number of such unmatched Chinese men, which grows to between near 30 and near 50 million by 2030, and compares it with the total number of women of reproductive age from low-skill households in East and Southeast Asia, including the Koreas, Indochina and Southeast Asia but excluding Japan.  This comparison motivates the scale of the mismatch problem.  By 2030 there will be less than 100 million women to make potential matches with the 30-50 million low-skill Chinese men, but these women are in no sense in surplus.  They are all fully matched in their own regions, if not in short supply themselves, and they would have no incentive to relinquish marriage in their own countries for marriage to low-skill Chinese men.  The clear conclusion is that, while trafficking of Asian women into China is likely to occur at the margin, female immigration cannot address the scale of the mismatch.

Considering the long lead times required for demographic changes to affect overall populations, it is not surprising that the demographic outcomes of Scenario 2, being identical in all other respects, are similar to those in the baseline. However, there is a considerable reduction in the gender imbalance by 2030, at least for children, with the sex ratio of 0-14 year olds reaching 111.5 compared with a baseline of 120.1, implying 10.3% of excess males in this age group, compared with 16.7% in the baseline.

Since the SRB is lower in the two counterfactual simulations than the actual data for the base year show, base year age-specific sex ratios are well above balance, and indeed above the assumed SRB of 106.  Consequently, in these simulations there are initial declines through time in age-specific sex ratios as the effect of the lower SRB flows through the age groups.  It is clear from Table 1 that the number of unmatched low-skill males by 2030 is significantly lower in this counterfactual.  The number of such unmatched males remains high, however, if skilled males take 1.5 women, since this creates unmatched low-skill males even if the SRB were always balanced.  In the higher-fertility variant, births are boosted additionally (if only slightly) by the increased number of women to bear children.  The result is higher population and labour force growth.  As we will see from the analysis in the next section, this is a key source of economic impact.

 

5. The economic implications of China’s gender imbalances

Overall, the economic effects of the lower SRB trend reflected in Scenario 2 compared with the baseline are very small since a decline in the SRB alone supplies very little change in the levels and skill shares of the labour force by 2030. However, as discussed in Section 2, gender imbalances impact on saving rates and on productivity as it is affected by the disaffection of low-skill males and related crime. We incorporate these impacts in the simulations that follow.

Simulation 1: A declining sex ratio with saving effects

Our first application is to the Du and Wei (2010) hypothesis that China’s household saving rate has risen at least partially because of competition amongst families of unmatched males.  Since this competition is likely to take place between families of male children, we show the baseline shares of unmatched males aged 0-14 in comparison with the recent paths of household and corporate saving as shares of GDP in Figure 6.  From these results it is clear, first, that the saving shares of GDP and the unmatched shares of male children have similar upward trends, offering some tentative support for the Du and Wei hypothesis.  Second, the levels of the unmatched shares are high; suggesting that, in 2010, between a sixth and a fifth of low-skill male children can expect to remain unmatched.  By 2030, these shares will have risen to between a fifth and a quarter, amounting to between 19 and 27 million boys.[26]

The empirical work by Wei and Zhang (2010) suggests that changes in the sex ratio explain 30-60 per cent of the change in the household saving rate.  From Figure 6, household saving as a share of GDP rose by nine percentage points between 1997 and 2010.  A liberal interpretation of the Wei-Zhang results would suggest that 45 per cent of this, or four percentage points, was due to the changing sex ratio.  Coincidentally, our simulated unmatched share of boys (0-14) also rose by four percentage points in this period.  This does not suggest a unit-elasticity for all China’s saving, however, since household saving is only 60 per cent of the total.  With this adjustment we deduce an elasticity of 0.6, indicating that a one percentage point rise in the unmatched share of 0-14 males yields a 0.6 per cent rise in the share of GDP committed to saving.[27]

We use this information in Simulation 1, which is based on the decline in China’s SRB from 120 to 106, as shown in Figure 3 for Scenario 2.  As shown in Section 3, this lower SRB causes the unmatched share of boys to fall and we now assume that this induces a decline in saving. The decline in savings manifests itself in two ways.  In the short run, reduced savings cause a rise in consumption expenditure on home goods (which are assumed to be more inelastic than foreign goods). Other things equal, this tends to boost GDP in the short run.  However, in the medium to long run, reduced savings lead to reductions in home investment and foreign assets from which income accrues later, both of which reduce GDP.

Since the shocks to saving are proportional to the difference between the two SRB scenarios in Figure 3, they occur gradually, allowing the long-run investment effect to dominate from the outset.  Yet these net effects are small, as indicated in Figure 7.  The changes in real investment, real GDP and real per capita income are shown to amount to no more than a percentage point by 2030. At the same time, the rate of change by 2030 is very considerable, suggesting that the saving effect could become much more important in the years beyond 2030. This raises the unfortunate possibility that rectifying China’s gender problem could actually have negative economic impacts in the future.

Following Du and Wei (2012), it is also possible to see from this simulation the effects of these shocks on China’s real exchange rate.  The net effect is small, however. Reduced saving tends to appreciate the real exchange rate, as Du and Wei indicate, but with a magnitude of only half a per cent by 2030.

 

 

Simulation 2: A declining SRB with saving and crime-related productivity effects

Characterising the implications of disaffected low-skill men for aggregate economic performance is heroic at best given the non-complementary nature of social research on this issue for China to date.  We embark on this with the objective of illustrating the potential scale of impacts rather than to offer a definitive forecast.  Our reasoning is conservative throughout and it runs as follows.

As described in Section 2, Edlund et al. (2007, 2010) use Chinese statistics to draw a clear link between the sex ratio and crime. They control for other causes and deduce that one per cent rise in the sex ratio is sufficient to cause property and violent crime to increase by three per cent.  The next link in the argument is an estimate of the scale of the productivity effects of crime in China, but this is unavailable to our knowledge.  Instead, we note from studies of the economic implications of crime in the US that its cost has been measured at around five per cent of GDP (Harwood et al. 2009).  We then speculate that violent and property crime in China could impact its economy in approximately the same proportion.  This yields a link between the sex ratio and the productivity impact of associated crime.

Combining the Edlund et al. relationship with a GDP shortfall by five per cent due to property and violent crime, it follows that a fall in the sex ratio by one percentage point would increase real GDP by 0.03*0.05=0.0015=0.15 per cent.  This implies that a fall in the sex ratio by 10 percentage points would increase real GDP by only 1.5 per cent.  However, crime is primarily an urban phenomenon that impairs the services sector more than others.  Combine this with the fact that the first urban destination of most low-skill men is the construction sector, in which the majority of investment expenditure occurs, and the effects of crime are leveraged noticeably.  Through these admittedly speculative steps we apply the following sequence: a rise of one per cent in the sex ratio of 15-39 year olds yields a three per cent rise in the crime burden.  Since the services sector supplies roughly half of GDP in China, if crime reduces GDP by five per cent, but its effects are concentrated in services, it must reduce services GDP by roughly 10 per cent, and so the effect of crime originating with unmatched males deducts three per cent from about a tenth of services GDP.  Thus an increase of one per cent in the sex ratio of 15-39 year olds thereby reduces total factor productivity in both services and capital goods production by 0.3 per cent.[28]

While we have leveraged what little supporting information is available we believe the resulting elasticity of -0.3 to be conservative, considering the first argument of Rodrik (1998), that “social conflict” is typically responsible for the much larger growth collapses observed in the more poorly performing developing countries, and second, the observations of Easterly (2001) on the causes of growth stagnation at middle income. The widening of China’s income distribution is just one factor that makes more fragile the political threads that hold its growth performance together (Tyers 2012), to which 30-50 million disaffected males of reproductive age would add.

In our second simulation, we assume the same decline in China’s SRB to 106 by 2030.  At the same time, this decline is linked to consumption behaviour so as to reduce the average saving rate as described previously.  Now we add the links between the sex ratio of the 15-39 age group and total factor productivity in the services and capital goods sector.  These have the effect of boosting productivity as the sex ratio declines toward balance.  The results are illustrated in Figure 8.

The story here is that the crime-related productivity shocks are shown to outweigh the slowing effects of lower saving and so they lead to stronger growth in the latter stages as the sex ratio of 15-35 year olds falls.  Investment is affected most because low-skill men work in the construction industry, the largest single component of the capital goods sector.  GDP and real wages rise by less than investment because the crime shocks are restricted to the services and the capital goods industries and so productivity in agriculture and manufacturing is unchanged.  Real per capita income rises least because it depends not only on home output (real GDP) but, increasingly in the later stages, also on income from assets abroad, which are considerable but made smaller by the lower saving rate in this simulation.

Finally, the effects on the real exchange rate are here mixed.  The saving and crime shocks are partially offsetting.  Reduced saving tends to appreciate the real exchange rate, as Du and Wei indicate, but improved services productivity from reduced crime affecting the largely non-traded sectors tends to depreciate it.  The trajectory is flatter than in Figure 8 and it turns negative near the end as the productivity effects outweigh the saving effects.  Overall, the effects on China’s simulated economic performance across more than two decades are still smaller than a single year’s growth.

 

 

Simulations 3 and 4: Low-SRB Counterfactuals

Here we construct two counterfactual simulations based on Scenarios 3 and 4, both with a constant Chinese SRB at 106, the first with baseline fertility and the second with higher fertility, as described in Section 4.  The lower SRB is assumed to reduce savings and crime, and therefore to increase productivity in same manner as Simulations 1 and 2.  The cumulative per cent departures from our baseline simulation are as indicated in Figures 9 and 10 respectively.

Here again, a struggle is evident between the counterfactual’s lower saving and its higher productivity.  The higher productivity wins in terms of investment and real GDP, which is larger in 2030 by a twentieth.  But the reduced saving retards the growth of real per capita income, the net effects on which are negligible.  When we introduce higher fertility in the second counterfactual simulation the differences are due to the resulting higher population and labour force indicated in Table 1.  From Figure 10 we see the expected (Solow-Swan) combination of reduced real wages and real per capita income higher real GDP.  The increased workers and the reduced crime cause additional real GDP growth, sufficient to raise its level by one-tenth in 2030.  The increased population, however, weighs on wages and real per capita income, resulting in comparatively small real wage increases and a net decline in real per capita income.

Reduced saving enters this mix by tending to reduce the growth in investment, mostly in the middle of the period, but this is more than offset by crime-related productivity improvements in the capital goods and services sectors, so investment growth is enhanced on net.  The path of additional investment tends to tail off at the end because the convergence of the sex ratio on 106 in both the counterfactual and he baseline sees growth in the productivity shocks die away at the end.  As this happens, additional accumulated capital lowers home capital returns and hence tends to attract less investment.  Since services are skill intensive compared with the other industries, stronger services growth in the counterfactual simulation raises the relative demand for skilled workers, which is why the gains in real skilled wages exceed those in real unskilled wages.

6. Conclusions

One of the constraints to continued rapid Chinese GDP growth is the slowdown and eventual contraction of labour force, which has been accelerated by China’s one-child policy.  A complicating associated factor arises due to the Chinese traditional preference for at least one son, which, via selective abortion, has caused a rise in the SRB.  This has reduced the share of women of reproductive age and so further slowed population and labour force growth. While slower population growth in itself has impacts positively on per capita income growth, gender imbalances have other widely noted undesirable consequences, including excessive saving as families with boys compete to match their sons with scarce girls, trafficking in women and rising disaffection and crime amongst the low-skill male population.

Results from our dynamic modelling suggest that, depending on the tendency for low-skill women to “marry up” and to choose second partnership with wealthy men over marriage to poor ones, by 2030 the number of low-skill men of reproductive age that is unmatched will be between 27 and 50 million, and still growing.  Though this is sure to result in rising trafficking of women from other East Asian countries, the number is simply too large for female immigration to allow a return to balance.

Even if policy changes can succeed in reducing the SRB to normal levels, we expect that the purely demographic implications would be small, at least by 2030.  The decline in saving by families observing improved marriage prospects for their boys would slow investment and income growth, but there would be fewer disaffected men of reproductive age and so less crime.  In our preliminary quantification of these effects, the reduced crime offers more growth benefits than are lost due to the reduced saving, although given some of the heroic assumptions made to reach this point, we concede that the opposite outcome could occur: rectifying China’s gender problem could actually have negative economic impacts in the future.

This does not amount to conceding that China’s gender problem should not be rectified, however. The pursuit policy reforms to achieve gender balance is further supported in that, as other political tensions rise through time, it is possible that the high SRB and the large number of unmatched and disaffected low-skill males could constitute the marginal pathology that tips China into Rodrik’s political malaise and significantly slower per capita growth.  Yet such policies will not come without cost.  If a further relaxation of the “one child” policy is needed to help restore gender balance, for example, more rapid population growth will slow the growth of real per capita incomes and hence the growth in the welfare of individual Chinese.

 


[2] The timing of China’s Lewis turning point is a subject of controversy, as suggested by the contrasts between the views expressed by: Cai (2010), Garnaut (2010) and Golley and Meng (2011), which offer just a sampling of a substantial literature.  There is, however, little doubt that the turning point is on its way, even if there is little agreement as to whether recent real wage rises suggest its presence.

[3] See also Attané and Guilmoto (2005).

[4] As evidence, he shows that, by 2000, SRBs had reached abnormally high levels in provinces with “strong traditional cultures”, including, Anhui, Henan, Hubei, Hunan, Shaanxi, Fujian and Guangdong, and were only normal in provinces with large ethnic populations, including Tibet, Xinjiang and Qinghai.

[5] See also Murphy et al. (2011), who show that patrilineal, patriarchal and patrilocal social networks and family systems have been a major determinant of the extent of son preference in China.

[6] See ‘Officials vow China will correct gender imbalance’ at http://news.xinhuanet.com/english/china/2012-05/24/c_131608451.htm

[7] See ‘Officials vow China will correct gender imbalance’ at http://news.xinhuanet.com/english/china/2012-05/24/c_131608451.htm

[8] Interestingly, in his report prepared for the United Nations Population Fund, Li (2007) makes no mention of the one-child policy in his discussion on the causes of China’s high SRB. For a cynical interpretation of why this is the case, see Hvistendahl’s (2011) fascinating book.

[9] http://www.china.org.cn/english/government/94926.htm http://news.xinhuanet.com/english/china/2012-05/24/c_131608451.htm .  Rodrik’s (1998) claim that social conflict has been a major determinant of growth collapses and stagnation in many countries since 1975, and The World Bank’s World Development Report 2011 on “Conflict, security and development,” both attest to the potential costs this could inflict on the Chinese economy.

[10] On the debate concerning whether married men earn more because they are more productive, see Becker (1981, 1985), Angrist (2002) and Antonovics and Town (2004).

[11]  Hypergamy exists when girls from lower socioeconomic groups marry boys from higher ones. Evolutionary psychologists indicate that  females have evolved a preference for higher status males because they offer their prospective children both “better” genes and greater resources. Men, on the other hand, tend to invest less in their children and therefore have less reason to prefer mates with high social status.  Moreover, some may choose to “marry-down” to ensure that their mates have a stronger incentive to remain faithful.

[12] Qian’s far less contentious policy conclusion is that “One way to reduce excess female mortality and to increase overall education investment in children is to increase the relative earnings of adult women.” (p.1281).

[13] These factors include the weakening of male-based traditions and gender discrimination (Guilmoto, 2009), and higher education and increased labour force opportunities for women associated with urbanisation and industrialisation (das Gupta et al., 2009).

[14] See ‘Officials vow China will correct gender imbalance’ at http://news.xinhuanet.com/english/china/2012-05/24/c_131608451.htm

[15] See ‘Officials vow China will correct gender imbalance’ at http://news.xinhuanet.com/english/china/2012-05/24/c_131608451.htm

[16] This preferred policy would involve a smooth transition period through to 2014, when all couples in China would be allowed to have a second child with appropriate spacing, reaching a soft-landing around 2035, by which time everyone would be able to choose their family size and timing freely.

[17] Chinese Vice Premier Li Keqiang announced in 2010 that China will continue to “coordinate its national family planning policy, stabilizing an appropriately low birth rate and improving the quality of its population” (cited in the The China Daily 2010).

[18] See Hudson and den Boer (2004) for a fascinating discussion of gender imbalances in history.

[19] The subdivision between production and professional labor accords with ILO’s occupation-based classification and is consistent with the labor division adopted in the GTAP Database. Mothers in families providing production labor are assumed to produce children who will grow up to also provide production labor, while the children of mothers in professional families are correspondingly assumed to become professional workers

[20] Immigration and emigration are also age and gender specific.  The model represents a full matrix of global migration flows for each age and gender group, which are sensitive to real wage differences and quantitative restrictions.  See Tyers and Bain (2006) for further details.

[21] Consumption and production behavior correspond to the standard GTAP Dynamic structures.

[22] There is, however, no endogeneity of saving rates to life expectancy, as suggested by Bloom and Canning (2005), nor to the sex ratio of the population, as suggested by Du and Wei (2012).  Death rates, and hence life expectancies, follow largely exogenous paths, as does the SRB.

[23] The transformation of workers from unskilled to skilled is another endogenous force in this model, with transformation rates differing by age and gender and depending on the real skilled wage premium.  Its role is limited in the experiments conducted here, however, and so it is not given emphasis. See Tyers and Bain (2006) for further details.

[24] See Golley (2012) and Golley and Tyers (2012b) for further explanation.

[25] A more detailed analysis of the implications of the baseline projection, for dependency and demographic dividends in particular, is offered by Golley and Tyers (2012a,b).

[26] Figure 6 also demonstrates that the most extraordinary thing about China’s saving is the proportion due to corporations, not households, indicating that the Du-Wei link may be overstated.  Strong corporate savings are most likely an industrial reform problem, since China’s corporate saving consists primarily of retained earnings by oligopolistic state owned corporations that are both lightly taxed and lightly burdened by dividend payments to the public and so move profits directly to investment at an extraordinary rate (Kuijs 2006, Tyers and Lu 2008).  While we do not model corporate savings explicitly here, recall from Section 2 that Wei and Zhang (2010) do provide a link between rising gender imbalances and corporate savings, as well as household savings.

[27] More precisely, since about half of China’s GDP is saved, for each per cent rise in the unmatched share of boys, a negative shock of this magnitude is applied generically to the consumption equations in the model, so that saving increases with the unmatched share of boys.

[28] Given our earlier reasoning it would be more germane to link the sex ratio of low-skill adults of reproductive age, or the proportion of low-skill males that are unmatched to crime behaviour.  We seek to be conservative here, however, being cognizant of the Edlund et al. results that use the overall sex ratio.

References

Angrist, Josh, 2002. “How do sex ratios affect marriage and labor markets? Evidence from America’s second generation”, The Quarterly Journal of Economics 117(3): 997-1038.

Antonovics, Kate and Robert Town, 2004. “Are all the good men married? Uncovering sources of the marital wage premium”, American Economic Review, Vol 94(2): 317-21.

Attané, Isabelle and Christophe Z. Guilmoto (2005) Watering the Neighbour’s Garden: the growing female deficit in Asia, OECD, Paris.

Becker, Gary S. (1981) A treatise of the family, Cambridge, Mass, Harvard University Press.

Becker, Gary S. (1985), “Human capital, effort and the sexual division of labour”, Journal of Labour Economics, 3(1-2): S33-S58.

Becker, Gary (2007), “Is sex selection of births undesirable”, downloaded from http://www.becker-posner-blog.com/2007/02/is-sex-selection-of-births-undesirable-becker.html in May 2012.

Bloom D.E. and D. Canning (2005), “The effect of improvements in health and longevity on optimal retirement and saving”, NBER Working Paper No.10919, Cambridge, MA: National Bureau of Economic Research, April.

Brooks, Rob (2012), “Little China girls: how history’s worst famine shifted the sex ratio” http://theconversation.edu.au/little-china-girls-how-historys-worst-famine-shifted-the-sex-ratio-6312, downloaded on May 20 2012.

Bulte, Erwin, Nico Heerink and Xiaobo Zhang, 2011. “China’s One-Child Policy and the mystery of missing women: ethnic minorities and male-biased sex ratios”, Oxford Bulletin of Economics and Statistics, forthcoming.

Cai, Yong and William Lavely (2007), “Child sex Ratios and their spatial variation” in Zhao, Z.; Guo, F (eds) Transition and Challenge: China’s Population at the Beginning of the 21st Century, Oxford University Press, Oxford

Cai, F. (2010), “Demographic transition, demographic dividend and Lewis turning point in China”, China Economic Journal, 3(2): 107-119, September.

Cai, Fang (2012), “The coming demographic impact on China’s growth: the age factor in the middle-income trap”, Asian Economic Papers Volume 11(1): 95-111. Cárdenas, Mauricio, 2007. “Economic growth in Colombia: a reversal of “fortune””?, Working Paper Series, February, No. 36.

Das Gupta, Monica Woojin Chung, Li Shuzhuo (2009), “Evidence for an incipient decline in the numbers of missing girls in China and India”, Population and Development Review (35(2): 401-416.

Du, Qingyuan and Shang-jin Wei (2010), “Sex Ratios, Savings Rates and Current Account Imbalances”, unpublished version, Columbia Business School, December 6.

Du, Qingyuan and Shang-jin Wei, 2012 “A Darwinian Perspective on “Exchange rate undervaluation”, http://www.nber.org/confer/2011/ease11/Du_Wei.pdf

Easterly, W. (2001), “The lost decades: developing countries’ stagnation in spite of policy reform 1980-1998”, Washington DC: World Bank, February.

Ebenstein, Avraham, Hongbin Li and Lingsheng Meng (2010), “The impact of ultrasound on the status of women in China, downloaded from http://ebookbrowse.com/ebenstein-li-meng-ultrasound-december-2010-pdf-d55208326

Eichengreen, B., D. Park and K. Shin (2011). “When Fast Growing Economies Slow Down: International Evidence and Implications for China,” NBER Working Paper, 16919, Canbridge MA.

Edlund, L., H. Li, J. Yi and J. Zhang (2007), “Sex ratios and crime: evidence from China’s one-child policy”, IZA Discussion Paper 3214, Bonn: Institute for the Study of Labor.

Edlund, Lena, Hongbin Li, Junjian Yi and Junsen Zhang (2010). “Marriage and Crime: evidence from China’s rising sex ratios” http://empac.ucsd.edu/assets/001/501502.pdf

Garnaut, Ross (2010), “Macroeconomic implications of the turning point”, China Economic Journal, 3(2): 181-190, September.

Golley, Jane (2012) “Uncertain numbers, uncertain outcomes”, chapter in Barmé, Geremie (editor), Red Rising, Red Eclipse, The Australian Centre on China in the World, ANU, Canberra.

Golley, Jane and Meng, Xin (2011), “Has China run out of surplus labour?”, Chinese Economic Review,

Golley, Jane and Rod Tyers (2012a), “Demographic dividends, dependencies and economic growth in China and India”, Asian Economic Papers, 41(3) in press.

Golley, Jane and Rod Tyers (2012b), “Population pessimism and economic optimism in the Asian Giants”, The World Economy, in press.

Gu, Baocheng, Wang Feng, Guo Zhigang and Zhang Erli (2007), “China’s local and national fertility policies at the end of the twentieth century”, Population and Development Review 33(1): 129-147/

Guilmoto, Christopher (2009), “The sex ratio transition in Asia”, Population and Development Review 35(3): 519-49, September.

Henrich, Joseph, Robert Boyd and Peter Richerson, 2012. “The puzzle of monogamous marriage”, Philosophical Transactions of the Royal Society 367:657-669.

Hudson, Valerie M. and Andrea M. den Boer (2004), Bare Branches: The Security Implications of Asia’s Surplus Male Population, Cambridge, MA: MIT Press.

Hudson, Valerie M. and Andrea M. den Boer (2008), ‘China’s Security, China’s Demographics: Aging, Masculinization, and Fertility Policy’, The Brown Journal of World Affairs XIV, no. 2 (Spring/Summer): 185-200.

Hvistendahl, Mara (2011), Unnatural Selection: Choosing Boys over Girls and the Consequences of a World Full of Men, New York: Public Affairs Press, 314pp.

Klasen, Stefan and Claudia Wink (2003), “Missing women: revisiting the debate”, Feminist Economics 9
(2-3): 263-299.

Korenman, Sanders and David Neumark, 1991. “Does marriage really make men more productive”, The Journal of Human Resources 26(2): 282-307.

Kuijs, L. (2006), “How will China’s saving-investment balance evolve?” World Bank Policy Research Working Paper 3958, Beijing, July.

Li, Shuozhuo, 2007. “Imbalanced sex ratio at birth and comprehensive interventions in China”, paper presented at the 4th Asia Pacific Conference on Reproduction and Sex Health and Rights, Hyderabad, India, October 29-31.

Murphy, Rachel, Ran Tao and Xi Lu, 2011. “Son Preferences in Rural China: patrilineal families and socioeconomic change”, Population and Development Review 37(4): 665-90.

Qian, Nancy, 2008. “Missing women and the price of tea in China” the effect of sex-specific earnings on sex imbalance” Quarterly Journal of Economics 123(3): 1251-85.

Rodrik, Dani (1998) “Where did all the growth go? External shocks, social conflict, and growth collapses” Manuscript, Harvard University, and NBER Working Paper 6350.

Song, Shige (2012), “Does famine influence the sex ratio at birth? Evidence from the 1959-61 Great Leap Forward Famine in China”, Proceedings of The Royal Society, downloaded from http://rspb.royalsocietypublishing.org/content/early/2012/03/16/rspb.2012.0320.full.pdf+html

Tyers, R. (2012), “The rise and robustness of economic freedoms in China”, in Moore, G. (ed), The Open Society and its Enemies in East Asia, Nordic Institute of Asian Studies Press, forthcoming, accepted January 2012.

Tyers, R. and I. Bain. (2006), “The global implications of freer skilled migration”; Working Papers in Economics and Econometrics No. 468, Australian National University, Canberra, July, revised September 2007.

Tyers, R. and F. Lu (2008), “Competition policy, corporate saving and China’s current account surplus”, Working Papers in Economics and Econometrics No. 496, College of Business and Economics, Australian National University, July.

Wei, Shang-Jin and Xiaobo Zhang, 2009, “The competitive saving motive: evidence from rising sex ratios and saving rates in China” NBER Working Paper 15093.

Wei, Shang-Jin and Xiaobo Zhang, 2010. “The sexual foundations of economic growth: evidence from China, working paper.

Wei, Xing Zhu, Li Lu and Therese Hesketh, 2009. “China’s excess males, sex selective abortion, and one child policy: analysis of data from 2005 national inter-census survey”, BMJ 2009(338) b1211

World Bank (2010), “Escaping the middle-income trap”, Chapter 3 in East Asia and Pacific Economic Update 2010 Volume 2: Robust Recovery, Rising Risks, November.

World Bank (2011), World Bank Development Report (Conflict, security and development), Washington DC.

Zeng, Yi (2007), “Options for Fertility Policy transition in China”, Population and Development Review 33(2): 215-46.

Zhao, Zhongwei (2011), ‘Reflections on China’s recent population statistics and current demography situation’ in Chinese Cross Currents, July: 44-57.

Zhao, Zhongwei and Chen Wei (2011), ‘China’s far below replacement fertility and its long-term impact: Comments on the preliminary results of the 2010 Census’, Demographic Research, Vol. 25, Article 26: 819-36.

Zhu, Wei Xing, Li Lu and Therese Hesketh, 2009. “China’s excess males, sex selective abortion, and one child policy: analysis of data from 2005 national intercensus survey”, BMJ 2009(338) b1211.

 

Figure 1: Long-term trends in China’s Sex Ratio at Birth

Source: United Nations (2010)

 

 

Figure 2: Recent trends in China’s Sex Ratio at Birth

Source: Li (2007), based on China population censuses in 1982,1990 and 2000 and 1% population sample surveys in 1987,1995 and 2005.

Figure 3: Exogenous Projection of China’s Sex Ratio at Birth

         Source: Population census data (Li 2007) and exogenous projections described in the text

Figure 4: Chinese Population Structure and its Change Through Time

Source: Baseline simulation of the model described in the text.

Figure 5: Unmatched Chinese Males 15-39 in East and SE Asian Contexta

(Millions)

a This compares the number of unmatched mainland Chinese production worker males with the number of women in the same age group that are in production worker families in Taiwan, Hong Kong, Indonesia, Korea, Indochina and other Southeast Asia.

Source: Baseline (Chinese SRB and fertility) simulation of the model described in the text. 

Figure 6: Chinese Saving and Unmatched 0-14 Male Shares, %

Source: Saving rates are derived from official macroeconomic statistics, including “flow of funds” data.  The unmatched male shares are from a retrospective simulation under baseline assumptions about the SRB and fertility, as described in the text.

Figure 7: Effects of Reducing the SRB – Saving and Labour Structure Changes Onlya

a These are cumulative % deviations from the baseline simulation with high SRB.

Source: Simulations of the model described in the text.

 

Figure 8: Effects of Reducing the SRB, including Saving, Labour Structure and Crime-Related Productivity Changesa

a These are cumulative % deviations from the baseline simulation.

Source: Simulations of the model described in the text.

 Figure 9: Departures from the Baseline of a Counterfactual with Constant SRB=106 and Associated Differences in Saving, Labour Structure and Crime-Related Productivitya

a These are cumulative % deviations from the baseline simulation.

Source: Simulations of the model described in the text.

 Figure 10: Departures from the Baseline of a Counterfactual with Constant SRB=106, Higher Fertility and Associated Differences in Saving, Labour Structure and Crime-Related Productivitya

a These are cumulative % deviations from the baseline simulation.  These results parallel those in Figure 10, except that they also include higher fertility, with the TFR remaining at 1.8 through 2030.  The differences between the two figures are therefore due only to this fertility change.

Source: Simulations of the model described in the text.

Table 1:  Demographic Outcomes of China’s gender imbalances

(Levels for 2030)

 

Baseline Scenario 2 Scenario 3 Scenario 4
Unmatched males, 0-14 millions 18.7 11.2 8.4 8.4
Unmatched % of males 0-14 16.7 10.3 6.0 6.0
Sex ratio 0-14, % 120.1 111.5 106.4 106.4
Unmatched males 15-39, millions 27.7 23.5 15.2 16.3
Unmatched % males 15-39 11.6 9.9 6.5 6.4
Sex ratio 15-39 113.1 111.0 106.9 106.8
Unmatched unskilled males 15-39 (millions), if
class segmented     18.1 14.4 7.2 8.3
women marry up     27.7 23.5 15.2 16.3
skilled take 1.5 women     47.8 43.4 34.9 37.0
Skilled workforce 116 115.5 115 120
Unskilled workforce 488 487.7 487 520

Note: Scenario 2 is lower SRB, Scenario 3 is counterfactual SRB of 106 with baseline fertility, Scenario 4 is counterfactual SRB of 106 with higher fertility.
Source: Simulations of the model discussed in the text.

 

Funding for the research described in this paper is from Australian Research Council Discovery Grant No. DP0557885 and it also draws on the resources of the Australian Centre on China in the World at the Australian National University.