For any serious attempt to assess China’s future outlook, an examination of the country’s population prospects is not only advisable but absolutely indispensable. There are two reasons for this.
First: of all areas of inquiry of interest to us at this seminar about China’s future, it is perhaps China’s demographic future that is *least* uncertain over the coming generation. The reason, quite simply, is that the overwhelming majority of the people who will be living in China in (say) the year 2040 are already alive, living there today. Population projections are far from error-free, but if we are trying to peer ahead a couple of decades they are most assuredly more reliable and empirically grounded than corresponding projections of economic change, much less political change or technological change.1
Second: demographics and demographic change actually matter—to economic performance, social development, and in some measure as well arguably to such things as military potential, political stability, and international security. This is not to invoke the “demography is destiny” claim, often attributed to the 19th Century French polymath Auguste Comte. A less florid, more immediately defensible reformulation of that aphorism would be that “demographics slowly but unforgivingly alters the realm of the possible”. In the following pages we will try to show just how the realm of the possible is being reshaped in China by impending demographic changes over the decades immediately ahead.
China’s Current and Future Population: What We Know and How We Know It
Before presenting the demographic projections underpinning this paper, we are obliged to address two basic questions about China’s demographic outlook: what do we know, and how do we know it? Answering these requires us to discuss data limitations today, and the intrinsic limitations of demographic projections for tomorrow.
Consider first the limits of current Chinese population data. Vastly more population information is available for China today than was the case for most of the Maoist era, when a virtual statistical blackout prevailed; China today also has trained and groomed a large cadre of top-rate demographers and population economists who work in the nation’s universities, state-sponsored think tanks, and government. On the other hand, China has not yet achieved complete or near-complete vital registration, meaning that analysts must rely mainly on reconstructions of trends from censuses and “mini-censuses”—and these counts are far from error-free.
Many errors in China’s population data are essentially politically induced—the data are deformed by mass misreporting due to ordinary people’s attempts to avoid the harsh consequences of Beijing’s various population control policies (using that term broadly). With regard to Chinese household registration data (which are derived from a separate demographic system run by the Ministry of Public Security), the 2010 Census indicated that at least 13 million Chinese citizens lacked legal identity papers because they were born “out of hukou”, i.e., outside the locality that the state mandates to be their residence2 (more on the hukou system shortly). But that guesstimate is based on official assumptions about China’s true population totals—and China’s vital statistics, census returns, and sample population surveys have undercounted the nation’s actual numbers for decades due to Beijing’s heinous “One Child Policy” and the familial incentives it established for birth concealment. The United Nations Population Division (UNPD) currently suggests that the 2010 China Census estimate missed the mark by about 30 million, even after its own internal undercount adjustments, and that it may have failed to enumerate well over a quarter of all female children under 15 years of age.3 From the 1982 China census onward, population totals and sex ratios for given birth years from one census to the next have proved unstable for babies, children and youth. These errors due to politicization of demographic rhythms of life may at perhaps now that Beijing appears to be scrapping its anti-natal campaign—but they are embedded in the data we use for projections to 2040.
As for projections themselves—these are no more reliable than the baseline data they use and the assumptions they input about future trends in fertility, mortality and migration. International migration is negligible for China in relation to its enormous population, and the assumption is this will continue to be true—lucky that, since demographers have no really defensible method for projecting international migration trends into the future. Demographic techniques for projecting survival trends for the currently living are fairly good—thanks to actuarial mathematics, after all, the life insurance industry has not gone out of business—but catastrophes of Biblical proportion do take place from time to time, and Providence has already visited a number of them on post-Liberation China. As a matter of simple population mathematics, however, fertility trends dominate longer-term population projections, and since there is no reliable method for projecting future fertility levels, assumptions are critical. China clearly undercounts births (with a reported total fertility rate of 1.18 births per woman per lifetime) but no one knows exactly by just how much. The consensus, for better or worse, is that the actual rate in recent years has been around 1.6, or about 30 percent below the level required for long term population stability in the absence of in-migration—but consensuses are not always correct.
China 2040: The Population Projections
Like sausages and law, the making of demographic projections may not look so pretty when seen up close. Irrespective: the fact of the matter is that population projections for China are likely to be *less* problematic than for many other countries or regions of the world. China is a low-migration, low-mortality, low-fertility society. This means there is, so to speak, relatively little “turnover” in the population from one year to the next. To go by the projections of the UNPD or the US Census Bureau, almost four-fifths of China’s projected 2040 population would be 22 or older then—meaning they have already been born at this writing. This brute fact far outweighs many of the smaller uncertainties highlighted above, and in a sense rescues us from them.
Nonetheless we need to know the assumptions built into the China population projections. Consider UNPD’s, which we will mainly use in this paper. (UNPD’s assumptions, by the way, are fairly close to those of the US Census Bureau, and for that matter also the China National Bureau of Statistics.) For the 2015-2040 period, UNPD assumes negligible net-outmigration from China of 0.2% per year—a rounding error, essentially. With respect to mortality, UNPD estimates overall male plus female life expectancy at birth was a bit over 76 years in 2010/15, and that it would rise to 80 by 2040/45 (with detailed “life tables” offering survival probabilities for males and females of every age over the interim). As for fertility: UNPD “medium variant” fertility projections envision a gradual rise in China’s TFRs from 1.6 to just over 1.7, meaning that childbearing in China would still be almost 20 percent below the level required for long term population stability around 2040. While this assumption about China’s fertility trajectory is highly debatable--this is a “known unknown”—the fact of the matter is this assumption has relatively little influence on our overall assessment of the implications of coming demographic trends. These assumptions, for example, would only affect the small share of the 2040 labor force as yet unborn (those then in their late teens or very early 20s)—and even for this cohort the impact of errant assumptions would be marginal.
Perhaps the clearest and simplest way to see what these changes would portend is to superimpose the projected population structure of China 2040 on the estimated population structure of China 2015. [SEE FIGURE 1]
Overall, total numbers in 2015 and 2040 would be quite similar: somewhere around 1.4 billion. But this is only a coincidence. Due to steep and prolonged sub-replacement childbearing (China’s net reproduction ratio is widely believed to have dropped below 1.0 in the early 1990s), China’s population would be on track to peak in about a decade (circa 2028 or 2029), and to shrink at an accelerating tempo thereafter: whereas China is thought to be growing by around 5 million a year nowadays, by these projections it would be shrinking by about 4 million a year in 2040. (In these UNPD projections, incidentally, India edges out China as the world’s most populous country just before the year 2025.)
Although China’s population totals are similar in 2015 and 2040, a fundamental transformation of China’s population structure is manifest in Figure 1—a change so dramatic we might even call it a leap into the demographic unknown. To be sure: there are esoterica in this tableau that would naturally catch a demographer’s eye—the population bulge for “the class of 1987”, for example, which is itself an “echo” of the upsurge in births in the early 1960s after the end of the famine unleashed by Mao’s catastrophic “Great Leap Forward”. But the main story on display is the extraordinary redistribution of China’s population—upward, toward the top of the so called “population pyramid”.
Two broad differences between China 2015 and projected China 2040 stand out. First: the overall population under 50 years of age is larger in the China 2015 than in China 2040—and for certain cohorts, such as those in their mid-twenties or early forties, China 2015 is dramatically larger than China 2040. Second: the overall population over 50 years of age is far larger in notional China 2040 than actual China 2015—over half again as large, in fact—and for many age cohorts, including septuagenarians, octogenarians, and nonagenarians, China 2040’s population is vastly larger than China 2015’s. In fact, the China of 2040 in Figure 1 would contain a quarter billion more people over 50 more than the China of 2015—while the ranks of those under 50 would be diminished by almost the same amount.
Most people understand intuitively that steep sub-replacement fertility levels eventually lead to depopulation (absent compensatory immigration). Less appreciated but no less avoidable is the relationship between low fertility and population aging: very small families make grey societies. In these projections, two generations of pronounced sub-replacement fertility would bring China to a place where none have gone before (at least so far). By UNPD medium variant projections, median age in China 2040 would be 47 years—higher than the median age for any country or territory on the planet as of 2015, according to UNDP estimates.
There are other aspects of Figure 1 that a discerning observer may notice—among them, the surfeit of males over females for the cohorts born during the decades of One Child Policy population control. And additional, potentially quite significant, population changes are underway that cannot be detected by a simple “national headcount approach”. We now examine these various issues.
Manpower and Labor Availability
By 2040, in these projections, China would have experienced almost half a century of sub-replacement fertility—and for most of the decades in question the nation’s fertility level would have been far below replacement. Thus it should come as no surprise that the working age population is thought to have peaked just before 2015—and from 2015 to 2040, is projected to shrink at ever greater speed.
By ancient convention, demographers talk of the “working age groups” as ages 15 through 64, and we shall do so as well in this paper. We know this formulation is arbitrary and also a bit archaic: nowhere is everyone between 15 and 64 in the workforce; growing numbers of teens and twentysomethings are out of the workforce because they enrolled in the training they need or want in order to join it; and in the real world ever increasing numbers of people 65 and older happen to be earning pay, in China and elsewhere. Yet as a first approximation the 15-64 cohort may not be a bad one for China’s working ages—and in any case population decline is in the cards between now and 2040 for most of the subgroups within this broad category too.
China’s past trends and future outlook are presented in Figure 2, which details estimated and projected changes in China’s “adult” (age 15+) population by broad age groups from 1970 to 2040. [SEE FIGURE 2]
Between the fateful December 1978 Plenum of the 12 CCP Congress (where Deng Xiaoping pointed China on a historic new economic direction) and 2010, China’s working age population grew by about 80 percent, swelling roughly from 560 million to one billion. Thus over that period, overall manpower availability rose by an average of 1.8% per annum, and total national work hours may have risen more rapidly as underemployed labor was absorbed in both the cities and the countryside. But between 2010 and 2015, manpower growth was roughly zero—reaching its projected (all time historical?) peak around 2014. Thereafter, China’s working age population is projected to commence a long decline—dropping by well over 100 million by 2040, to around 880 million, at which point it would be shrinking at a rate of 1% a year.
In terms of simple economic “growth accounting”, increased labor inputs did not account for all of China’s spectacular economic growth during the 1978-2010 period, or even for most of it—but it did account for a hardly trivial fraction of that boom.4 The coming reversal of the delta for manpower change in the years immediately ahead means that, from the standpoint of the very simplest sort of growth accounting, the Chinese economy will be facing increasingly unfavorable headwinds simply due to manpower decline, everything else being equal.
But everything else will not be equal. We already know, for example, that the composition of the working age population is irrevocably set to change, and in ways that would seem inauspicious for economic growth. Over the coming generation, the pool of young manpower is on track to shrink sharply, with only the pool of older manpower expanding. This is not the way economic planners would have designed things. The youth labor group (ages 15-29) in modern societies always has the highest educational attainment, is the most IT and tech savvy, and tends to be the most flexible (all the more so in China since most people in this age group have not yet started to form families). Between 2015 and 2040, the 15-29 group is projected to shrink in size by 75 million, or roughly a quarter, and to shrink as a share of total manpower from a little less than a third to just over a quarter. The 30-49 group, for its part, might be regarded as a part of the life cycle in which entrepreneurship and inventiveness comes to fruition: Benjamin F. Jones’ international findings on “the age of great discovery” are particularly intriguing in this regard. [SEE FIGURE 3] Without getting too deterministic about this, we may entertain the conjecture that Thirtysomethings and Forty somethings add a “secret sauce” to the workforce and the economy. Too bad for China’s outlook if so: between 2015 and 2040, this group is projected also projected to shrink in size by a quarter, by well over 100 million men and women, and to drop from 43% to 37% of total manpower. It is only the 50-64 cohort that can be expected to grow over the generation ahead: the least educated and health healthy contingent in the labor force (although of arguably also the most experienced)—its share jumps from about 25% of total manpower to about 35% over the years under consideration, but even projected numbers for this group start to fall before 2040.
Beijing’s economic policymakers have some options in responding to this unfavorable impending change. Improving education of the workforce is one option—but the 50-64s of 2040 are already out of school, and China’s inverted population pyramid makes the task of increasing overall educational attainment through schooling much slower than would be the case for a youthful population. Raising the capital-labor ratio is another theoretical option (this is what Ronald D. Lee and others call “the second demographic dividend”5), but China’s gross domestic capital formation ratio today is already bizarrely, perhaps unsustainably, high. Some have argued that China can muddle through this problem by raising labor force participation rates for the working age manpower pool.6 But it is far from obvious this will be feasible. China’s everyday labor statistics are notoriously poor. The most reliable numbers available come perhaps from the 2010 census. If we go by those figures, China’s working age manpower may not be as ferociously mobilized as say Kim Il Sung’s North Korea7, but its LFPRs for both men and women are comparable to or higher than most OECD countries, and the same is true for work rates (employment to population ratios).8 Some might hope there would be room for coaxing additional labor out of China’s underemployed adults, especially those in the countryside—but Cai Fang, perhaps China’s most eminent population economist, has argued that China already reached the Lewis-model “turning point” a decade ago.9 Some countries and places—Singapore, Gulf states, and even Western Europe among them—have attempted to redress labor shortages through international migration: but as (currently) the world’s largest country, China has a scale problem; attracting 100 million plus workers voluntarily and through economic incentives over the coming generation is simply inconceivable in the world as we know it. There is the option, however, of internal migration: of ramping up productivity for the existing, dwindling manpower pool by moving peasants to more remunerative work in the cities. Beijing has seized on this option and is actively promoting it through its ongoing “National New-Type Urbanization Plan”, also widely known as the “urbanization drive”.10 There is promise in this strategy—but as we shall see in a moment, it is not exactly an unalloyed cure.
Despite the prospect of overall population decline in the era ahead, China will be experiencing a very particular type of population explosion: an explosive increase in its number of senior citizens 65 years of age and older. Between 2015 and 2040, in these UNDP projections, China’s 65+ population would jump by almost 150%: from 135 million to almost 340 million. That is a long-term growth rate of 3.7 percent a year: a breathtaking tempo of growth for any major population group for decades on end, and one that perforce should be expected to shape the nation’s economic, social, and perhaps even political outlook. By 2040, if things go well, China will be a “super aged society”11 with 22% of its people 65 or older (21% being the conventional threshold for defining “super aged”). (The only scenarios under which China does not become super aged are catastrophic ones.) By the criteria median age and share of population 65+, China would in fact be more aged than the USA—meaning the USA of 2040 (at least, by Census Bureau projections). How China copes with its coming senior tsunami and the attendant impending old age burden is a critical question for China’s future.
Over the past generation there has been considerable research on the role of the “demographic dividend” in spurring economic development.12 This work holds that the fertility transition offers a once in history chance to accelerate development by raising the share of a nation’s working age population—not only increasing the availability of laborers more rapidly than total population, but also propitiously influencing savings and investment possibilities through this shift in population structure. Most population economists today attribute some of China’s spectacular success over the past four decades to this “demographic dividend.”13
For better or worse, China’s demographic dividend has already been cashed. Between 1978 and 2010, China’s 15-64 group shot up from 58% of total population to an amazing 74%. Now it is on its way back down, and by 2040, in our UNPD projections, it will be back to 62%--where it was in 1982—and still heading south. The dependency ratio of 2040 and 1982 may be identical, but their portent is very different, since almost all of the non-working age population then were children and in 2040 the great majority will be elderly adults.
To be fair: in 2040, on current trajectories, the 65+ population in China will be the healthiest and best educated cohort of senior that have ever inhabited the Chinese mainland. Among other things, this means they may be *less* economically dependent those before them, more capable of making do financially on their own. But that is a relative comparison. China’s seniors in 2040 will also be China’s least educated adult grouping. By the projections of the Wittgenstein Centre, for example, in 2040 nearly half (46 percent) China’s seniors would have a primary school education or less—i.e. 6 years or less, with 5 percent of them having no education at all. (The corresponding share in 2040 for the 20-39 group would be 13 percent.) Over three fourths of China 2040’s seniors would have no more than lower secondary education—i.e. 9 years of school or less. [SEE FIGURES 4 AND 514] Thus paradoxically it is the seniors—China’s most physically fragile contingent—who would be the group most likely to be obliged to engaged in physical labor if attempting to support themselves economically. Professor Wang Feng of UC Irvine and Fudan University and colleagues calculated that as of 2010 Chinese seniors earned less than 40 percent of the resources that were sustaining them at age 65; just 20% at age 70; and maybe 10% at age 75.15
Recall as well that the very fastest growing contingent of seniors in China as elsewhere are the oldest old, men and women 80+ years of age. This group is on track almost to triple as a share of China’s population, from 1.7% in 2015 to 4.9% in 2040. The risk of dementia and Alzheimer’s increases very rapidly after age 80; until and unless humanity finds the silver bullet for this terrible affliction, the burdens implicit in an Alzheimer’s explosion also have to be taken into account with China’s senior tsunami. In addition, a steadily growing share of China’s seniors and oldest old are living by themselves in one-person households (a trend not particular to China, but to the contrary witnessed worldwide). The especially rapid growth of China’s live-alone senior population can only make for additional vulnerability and risk on the years ahead.
Note that the greying of China promises to be a highly varied process geographically, with one of the most dramatic cleavages separating urban and rural China. Projections in collaboration with colleagues by Professor Zeng Yi of Peking University and Duke, one of China’s foremost demographers, suggest that rural China is already far greyer than urban China thanks to rural-to-urban labor migration—and that the gap is only set to widen in the decades ahead [SEE FIGURE 6].16 Zeng et al anticipate a China in 2040 where something like a third of the rural population would be 65 or older—twice the ratio for urban areas. By way of comparison—the very greyest spot on earth today (2015) according to UNPD is Japan, with 26% of the population 65+. As the famous aphorism notes, Japan got rich before it got old; one does not have to be a Sino-pessimist to recognize that rural China is getting ready to do things the other way around. In fact, China’s very greyest regions in the future are most likely to be its poorest, least educated, and least healthy as well; they will no doubt benefit from remittances (from migrant working age children living off in urban China), but only to a degree. Even with better education, health, and capital investment in 2040, seniors in China are set to be dependent on support from resources other than their own earned income.
The question inescapably rises: who will provide for China’s immense population of future seniors? A first response would be: current government policies will almost certainly *not* do so—or at least will not do so comprehensively and adequately. Beijing has been dithering about nation-wide public pension and old-age health care guarantees for over two decades now, and while a number of important steps have been taken, the situation is what might charitably be called a work in progress. As of 2017, for example, less than 65 percent of China’s working age population was covered by any pension schema, and only 35 percent of urban migrants were covered; in rural areas, for their part, the pension schema offered a “basic benefit” of 70 RMB per month for qualifying retirees—just over $10 US per month at today’s exchange rates.17 Even so: China’s real existing pension and health system is severely underfinanced, due in part to overpromises to special constituencies such as urban residents and state-owned enterprises (SOE) employees; according to IMF calculations, the implicit debt (net present value of unfunded liabilities) in China’s current health and pension programs amounts to about 100 percent of the country’s GDP.18 China’s pension liabilities can be reduced through practical and feasible reforms—re-examining vested benefits for urban groups and SOEs, and raising retirement ages from the Stalin-normed levels of 60 for men and 55 for women that were set in the early Maoist era—but that still begs the question of coverage and support levels for the grey needy, of whom there will likely be vast numbers. And this budgetary problem stands in addition to China’s other notorious looming debt challenges.
If public policy will not fill the gap, what will? Personal savings are one answer—and uncertainty about future government old-age guarantees may help in part to explain China’s strikingly high private savings ratio.19 But generally speaking, savings and need are inversely correlated in China as elsewhere, meaning China’s vulnerable aged of tomorrow cannot count on their own savings and assets for old age security.
China’s historic mechanism for assuring care and income security for seniors was called the family. In the event, the family mechanism will no doubt be relied upon to provide for frail and failing elders in the coming generation, too. Just how well it will acquit itself in this task is another question. Two generations of sub-replacement fertility will have taken their toll on the family unit in China 2040 (and on the extended family as well, about which more later). The son—or rather, the daughter-in-law to which he is attached—has been the notional caregiver and provider for aging parents under Chinese norms since at least the consolidation of the Chinese empire under the Qin dynasty.
What happens though when there is no living son? We are about to find out, and bigtime: back of the envelope calculations suggest that the proportion of Chinese women 60 years of age with no male child may have risen from 7 percent in the early 1990s to 30 percent or more for post-2025 China. Dutiful daughters may of course step in, but their loyalty attention and resources may be all too frequently divided, inadequately, between two sets of aging parents.
All of this, however, presupposes that two and a half millennia of Confucian values will inform the behavior of adult children toward their elderly parents in the generation to come. That means taking the near-universal continuation of filial piety for granted. Such devotion might have been easier when the elders were scarce and the children were plentiful; tomorrow those tables will be turned. Beijing has already begun to lay down markers here, criminalizing non-support of parents and even non-visiting in 2013.20 Why, we may wonder, do authorities feel such laws to be necessary?
We can assume China will be considerably richer in 2040 than today—ceteris paribus that would mean more resources for elderly support. Human beings tend to cope—work from Ronald D. Lee’s National Transfer Account project has suggested, intergenerational household resource transfers may be quite effective in dealing with population aging in much of East Asia, including perhaps China. Some Pollyannas have even suggested that the sheer scale of need for China’s rising cohort of seniors will help the nation undertake much needed reforms, such as a shift to consumption oriented growth. Suffice it to observe that darker scenarios are also possible—including the prospect in rural areas of something like a pervasive, slow motion humanitarian tragedy, met with Darwinian solutions. It is also not impossible that these “optimistic” and “pessimistic” scenarios could unfold at the same time within the same country.
Gender Imbalance and “Marriage Squeeze”
China’s eerie and biologically unnatural One Child Policy era upsurge in sex ratios at birth raises another demographic dilemma for China, this one seemingly out of a science fiction novel: the prospect of a Chinese future with a major and persistent shortage of brides.
In normal human societies of any size, there is an abiding regularity to the sex ratio at birth (SRB) across history, countries, and ethnicities: typically running in the range of 103-105 baby boys per 100 baby girls. From the 1982 China population census to this writing, however, weirdly high SRBs have been reported in official Chinese population data. By the 2010 Census, the reported SRB was about 120. In certain provinces, the SRB has reportedly exceeded 130—and in not a few localities, it has exceeded 150. These numbers point to mass female feticide in the context of reliable prenatal gender determination technology and unconditional abortion. China is by no means the only spot on the planet where this is taking place (elsewhere I have written about “the global war against baby girls”21), but it is arguably the largest and most brutal battlefront in that campaign.
Given the persistent undercounting of China’s babies in the One Child Policy era, we cannot tell directly from official data just how severe China’s true SRB distortion has been. It is left to demographers to reconstruct actual trends—and their assumptions about the degree of undercounting can differ; likewise in projections about the future. UNPD today estimates that China’s SRB peaked in the 2000s at 117, is currently around 115, and will be 111 around 2040; the Census Bureau says SRB peaked at 118 in 2005, is 113 in 2018, and will be 107 in 2040. (Such differences of course have a multiplier effect on the dimensions of the “marriage squeeze” one envisions for the brides- and grooms-to-be several generations hence, but do not much affect estimates for the dimensions of the “marriage squeeze” facing China in 2040.)
Up to now, family formation in China has been influenced by what we might call a “universal marriage norm,” an ethos strongly informed by the Confucian metaphysical imperative of continuing the family lineage through the male issue. Towards the end of the 20th century, that norm was translated into reality: in those years, all but 4 or 5 percent of men and an even smaller fraction of women in their late 40s had been married. Now the arithmetic of gender imbalance means all this must change. It will change even faster, and more acutely, if the universal marriage norm erodes. In almost all the rest of East Asia, that norm has already fallen into considerable disrepair; demographer Gavin Jones in fact talks of a “flight from marriage” in the region,22 with ever greater proportions of young men and women postponing marriage or forgoing it altogether. If young women in China follow the example of their peers in Japan, Taiwan, Hong Kong, and elsewhere, the prospects for the marriageability of China’s prospective future bridegrooms will dim all the more.
In this paper I project one outlook for China’s coming “marriage squeeze,” utilizing Prof. Zeng Yi’s invaluable PROFAMY software for this task.23 [SEE FIGURE 7] (I regard this as a “middle” variant set of assumptions about fertility, SRBs, “flight from marriage,” and the rest: more dramatic scenarios could be imagined.) . In this scenario, 20% of Chinese men in their early 40s are never-married by 2030—up from just 4% in 2000. This would be a nationwide average, though: needless to say, the odds of being unable to marry would be higher for men who were rural, poor, and/or poorly educated. Such projections suggest a China 2040 with tens of millions of essentially unmarriageable men (although my method also includes some men who would be voluntarily never-married).
A reality check on my projections comes from UNPD and US Census Bureau projections for China 2040. In UNPD medium variant, there would be 23 million more men than women in the 25-44 cohort, and 29 million more men than women if the more appropriate comparison is men 30-49 for women 25-44. In the Census projections, the corresponding surfeit of marriage-age males would be 22 million and 30 million, respectively, or 13%-17% of the total male reference population. Since we are dealing with stocks and flows in calculating the prospective marriage squeeze, these figures are not so inconsistent with my own projections.
What will it mean for China to have a growing internal army of unmarriageable young men—a contingent predominantly poor and poorly skilled? Counterintuitively, there may be some positive economic spillovers: some research suggests male competition for brides has already promoted something like a savings race.24 Economists and public intellectuals like Gary Becker and Judge Richard Posner, further, have mused that the scarcity of females in China would eventually have the beneficial effect of increasing their “value.”25 To date, alas, “rising value of women with Chinese characteristics” has meant kidnapping, sex trafficking and other violations of human rights.
Moreover, there is no obvious policy solution in sight for the coming marriage squeeze. Places like Hong Kong and Taiwan have dealt with their own bride shortages by “importing from abroad”—but China has a scale problem. It also would be in a different segment of the marriage market, requiring (huge numbers of) willing brides for its relatively impoverished hinterlands. Zeng Yi and others have suggested that establishment of solid national pension and health care guarantees would reduce “son hunger” in China, especially in the countryside—but as already mentioned, nothing like this is yet in place in rural areas, and it would take another generation for such policies to affect the marriage squeeze once implemented.
Does China’s coming “bare branches” problem portend social or political instability? The question occasions continuing, sometimes heated, debate. Professor Valerie Hudson, author of the “bare branches” thesis,26 famously argues that a surplus of men tends to make for domestic and international tensions.27 On the other side, Professor Wang Feng and others have pointed out that a serious surplus of marriage-age men has been the norm rather than the exception throughout Chinese history due to the heinous but time-honored practice of mass female infanticide and killing of girls,28 suggesting that Chinese customs and institutions had long adapted to this demographic anomaly.
Wang Feng et al. have gotten those historical particulars right—what those mean for the future, however, is another question. In the generation ahead, China may well be on the rise—but an increasingly powerful and affluent nation will be inhabited by growing numbers of presumably frustrated young men who find their life chances worsening in a most personal and bitter fashion. Their expectations will be shaped not by ancient Chinese history, but by marriage prospects within living memory. Will this make for millions of stories of quiet personal desperation or something more collective and convulsive: for anomie or fury? It is too early to tell. At the very least, however, we should regard China’s future marriage squeeze as a potential wild card—possibly an important one.
Domestic Migration, Urbanization, and the Hukou System
Integral to the structural transformation of the Chinese economy during its extended spate of exceptionally rapid growth has been a reallocation of labor out of agriculture and into industry and services, and a corresponding movement of population out of the countryside and into the cities. Between 1978 and 2015, the population of what China officially defines as urban areas has grown by almost 600 million (roughly 200 million more than total national population growth) and the official urbanization ratio has more than tripled, catapulting from 18% to almost 56%.29 The UNPD envisions a further increase of China’s urban population of over 300 million between 2015 and 2040, at which point China would be over three fourths (76 percent) urban.30 Chinese leadership is counting on urbanization as an engine of economic growth for the Chinese future and is attempting to accelerate the rise of cities through the aforementioned far-reaching “urbanization drive.” Authorities in Beijing are right to regard cities as engines of growth—a corpus of economic research corroborates that judgment.31 But “urbanization with Chinese characteristics” involves a population problem that does not show up in conventional “headcount” statistics. It relates to China’s peculiar institution, the hukou system.
Hukou is a system for household registration and personal identification that traces far back into imperial China, but whose modern import derives from Mao’s weaponizing it as a tool of totalitarian control. From the 1950s onward, the Chinese Ministry of Public Security has supervised hukou and designated the official place of residence of every Chinese citizen. It is illegal to live outside of ones authorized hukou—although temporary hukou can be approved in certain circumstances (for example, if one has found a job in a city or a different province).
Although Beijing relaxed its stringent controls on domestic migration in the early 1980s, to date the Chinese government has proved extremely reluctant to “update” workers’ or migrants’ hukou in accordance with their new place of residence. Thus an enormous “floating population” of out-of-hukou migrants has emerged with the rise of Chinese cities and the attendant upsurge in urban demand for labor. As of 2010, China officially numbers its floating population at around 220 million; about one fifth of all working age men and women were out-of-hukou then, and both the totals and proportions would be higher today.
Most of China’s floating population today is comprised of migrant peasants working in urban areas. (Their hukou identity papers ascribe that class status, by the way.) With very few exceptions, these men and women—and for that matter as well, the migrants who left their home towns and cities for other urban jobs—work in places where they are at best second-class citizens, at worst de facto illegal aliens. In their “temporary” residence, as a rule, they have no right to local services (health care, education etc.). They have no right to bring their (non-working) family members with them. Their compensation is lower than for in-hukou counterparts of the same education and skill levels. And in legal or other disputes involving authorities, they are virtually sure to lose.
Urban China desperately needs migrants—for reasons demographic as well as economic. Fertility levels are extremely low in Chinese urban areas, and extraordinarily low in China’s bigger cities. In recent years total fertility rates in places like Beijing and Shanghai have sometimes been below 1.00—one birth per woman per lifetime. Overall, prolonged extraordinary sub-replacement fertility means that these places can no longer sustain their overall population totals, much less their labor forces, without a constant inflow of new migrants.
Thus the contradiction: cities (and economic planners) need new urban migrants—but those very same migrants must be treated as inferior beings by the logic of the current hukou population control system. The contradiction is highlighted in Figure 8, which depicts the population structure by hukou status of China’s big cities (shi) as of 2010. [SEE FIGURE 8] In all, over 40 percent of the big city population that year was accounted for by migrants—and migrants comprised the outright majority of all big city residents in their late teens, twenties, and early thirties. Remember: the data in Figure 7 are now 8 years out of date—undoubtedly migrants form a majority of even more working age groups in China’s larger cities today.32
Even within the Beijing dictatorship itself, the patent injustice of hukou-based exploitation of the new caste of migrant workers is widely recognized. Official and semiofficial discussion of hukou “reform” (meaning abolition) has been in the air for two decades and more. And some adjustments are underway: the official “urbanization drive” talks of granting local hukou to 13 million workers a year through at least 2020. But since urban population will be growing rapidly at the same time, such measures would merely more or less cap the size of the urban out-of-hukou contingent. If everything goes according to plan, in fact, 15% of China’s total population will be “temporary urban residents” in 2020: over 200 million people. To judge by current indications, Chinese leadership to maintain the hukou system indefinitely.
But why? There are at least two obvious answers. First: under current arrangements, migrant workers are cash cows for the cites and townships in which they toil; vesting them with the same rights to services as in-hukou urbanites would throw public finances into disarray for municipalities across the country. The central government could fix this problem through budgetary consolidation relatively easily—but this would also be expensive, and Beijing does not want to assume these costs. Second: the hukou system still seems to be viewed by this police state as an indispensable instrument of control. Social and political stability in urban areas is a paramount concern for Chinese leadership (in part for historical reasons: dynasties fall when the capital and the major cities fall), and the hukou system helps assure public order in cities.33 Recall that just a decade ago, during the global financial crisis of 2008/09, an untold number of million migrant workers (perhaps 20 million or more) were sent back home when export demand slumped; this mass resettlement was enforced via hukou. What would have happened if those unemployed masses had stayed in place, milling about in the cities? Chinese authorities didn’t want to find out. That experience—and more recent exercises of hukou power for mass ejection of migrants (in Bejing for example)—have presumably demonstrated the utility of the hukou and reinforced the regime’s determination to keep it in place.
In the hukou system we see a political problem in demographic form. It is akin to “influx control” under the old South Africa apartheid regimen; thanks to hukou, urban centers now look a bit like “Soweto with Chinese characteristics.” We know what happened in Soweto. Until and unless the hukou system is genuinely reformed, China may have comparable tinderboxes in every one of its big cities. For this reason, migration and urbanization should be regarded as another “wild card” in China’s future—and one whose risk of being cast could be considerably higher than the “marriage squeeze” card.
The Coming Revolution in Family Structure
One immensely important and utterly unstoppable demographic change now underway in China has attracted curiously little attention form Chinese policymakers and their think-tank advisers. This is the coming revolution in Chinese family structure. While the Chinse Academy of Social Sciences, DRC, and other organizations have provided voluminous analyses on coming labor force trends, the implications of population aging, urbanization, and migration, and even SRB gender imbalance, there has been so far as I can tell no research yet on mapping out the dimensions or examining the implications of the now unavoidable atrophy of the extended family, or the equally unavoidable rise of a “new family type” within China. Perhaps this is because such work would take us beyond “the headcount approach;” the Chinese government, like other modern states, collects demographic data on individuals and households, not kinship networks.
Changes in childbearing and survival patterns cannot help but change nuclear family and extended family patterns too—and dramatic changes in childbearing and survival lead to dramatic changes in family patterns. (For simplicity’s sake we will be discussing only consanguineous family here—but that is not a bad first approximation for family in China today.) Generally speaking, improved survival increases the number of living family members and kin at any point across one’s life course, while declining fertility. But the distribution of births (the so called “parity progression ratio” or PPR) also matters: a society could have an average of two births per woman if half of all women had four children and half had none—and this would look very different from one where all women had exactly two children. China’s tremendous improvements in life expectancy since the early 1950s greatly increased the number of living kin for grown men and women over the past three generations. But total potential living kin depend on birth patterns and of course fertility has plummeted in modern China.
Of particular interest in this regard is the number of only children in China, today and tomorrow. The rise of the only child radically transforms not only the structure of the nuclear family but also extended kinship networks. Both official Chinese and independently conducted reconstructions of PPRs indicate that almost all women of childbearing age got married and almost all had at least one child—but from around 1993 on fewer than half of first births led to further births for the country as a whole. In 1990, by these reckonings, about one in six Chinese births would end up only children; by 2000 the fraction would jump to two in five or perhaps even higher.
Calculations for the proportion of only children are highly sensitive to under-reporting of births, so it is possible that they overstate the nationwide proportions of only children in China. That said, there is no doubt whatever that only children have comprised a majority of newborns in urban China for decades. Whereas the national fertility level fell below replacement in the early 1990s, it appears to have dropped below replacement in urban China in the early 1970s, years before the One Child Policy; by 1982, in the early days of that population control drive, total fertility rates in urban areas were already down to an estimated 1.4 births per woman per lifetime.34 By 1984 over half of all urban births (townships plus big cities) may have been only children—and in the big cities the ratio may have been closer to 70 percent. In the biggest metropolitan areas, the share of only children might be even higher. By collaborative estimates of the China National Statistics Bureau and the East West Center in Hawaii, four fifths of babies born in Shanghai were only children as early as 1990, with ratios for Beijing only slightly lower.
Today only children form a majority of urban China’s (legal hukou) population under 35 years of age—and a supermajority of the under-35 population in the country’s big cities. This means we are starting to see the rise of a new family type in China: only children begotten by only children—boys and girls with no siblings, cousins, uncles, or aunts, only ancestors and (perhaps eventually) descendants. For this new family type, the traditional extended family has essentially collapsed. This new family type is now beginning to account for a sizeable fraction of urban China’s (officially authorized resident) children—very possibly, an outright majority in the country’s economic and political nerve centers (Shanghai and Beijing) and in other cities of size as well. But even in places where the emerging new family type does not dominate, in the rising generation who will be the parents of 2040, the extended family and its kinship networks are being dramatically compressed by long-term sub-replacement fertility.
The family unit matters everywhere, but it has assumed a particularly prominent institutional and even spiritual role in Chinese tradition. For millennia guanxi networks—comprised principally though not exclusively of fellow clansmen—have helped provide financial and human security for the Chinese population; they have been integral to getting business done at the micro level, and at the macro level have improved national economic performance by reducing transaction costs and risk. What will happen to economic performance in China as its guanxi networks come under extraordinary new demographic pressure? We are about to find out. There are of course functional substitutes of sorts for family-based guanxi networks: deep personal friendships among unrelated individuals would be one; impersonal spheres of “social trust” now witnessed in China’s fascinating “fintech revolution” would be another. But it is far from clear that these substitutes are complete substitutes, much less perfect ones.
The ongoing family revolution in China might possibly also have implications for political cohesion and national security. Remember the tragic Sichuan earthquake a decade ago, in which thousands of schoolchildren perished. Many of them were only children; their deaths brought a permanent end to untold family lineages. In those localities and across China there was a spasm of social rage as people learned that the earthquake knocked down cheap and shoddily constructed schools, even though nearby CCP and government buildings survived the tremors. The tragedy took on an electrifying import across China, one magnified by its consonance with the age-old Chinese theme of unjust rulers losing the mandate of heaven. This disaster was thus also a public relations disaster, and forced the regime into contrition mode, requiring the unusual spectacle of conspicuous and repeated public apologies by Chinese officialdom, all the way to the very top.
Consider what this domestic tragedy may portend for a future international confrontation or conflict involving serious Chinese casualties. Many of the soldiers in the PLA will presumably be only children. Major losses would mean the end of a great many family lines. If China suffers setbacks in international military operations, or if the Chinese public deems these losses to be the result of an illegitimate use of power, what sort of explosion of social rage might Beijing face? (No less pertinent: how might regime calculations about the possible risk of social rage due to military losses condition China’s defense strategy and tactics in the years ahead?)
Demographic factors suggest, among other things, that the coming generation will not see a repeat performance of the phenomenal economic rise that China enjoyed over the past generation (or generation and a half). To be sure: despite the demographic constraints outlined above, China is certainly capable of generating creditable rates of economic growth for the foreseeable future. But demographic realities (among other forces) are likely to bring an end to China’s era of “heroic economic growth,” possibly sooner rather than later. Utilizing a human capital-based model, Stanford’s Scott Rozelle and colleagues have ventured that the Chinese economy would grow by an average of 3% per annum over the coming 20 years.35 In my own current work, a simple human-resources-plus-business-climate model comes up with results more or less consistent with Rozelle et al: about 2.5% per annum GDP growth for the 2025-40 period when GDP is measured in PPP terms (with somewhat higher rates for exchange-rate based GDP).36 We should recognize that such projections are lower than most prevailing estimates—but they highlight the headwinds the Chinese economy faces when demographic trends are taken into consideration.
Demographic constraints could also complicate Beijing’s quest to mobilize political power and/or apply it abroad. At this writing the Chinese regime seems to be behaving in an increasingly ambitious and assertive fashion: the era of “hide your strength, bide your time” appears to be over. Demographic stresses could reduce social cohesion, or even contribute to social or political instability. Note the verb “could”: this is by no means a certainty. But it is a possibility that would be unwise to ignore. Dynasties in China always end. When and how the current regime will end, and whether demographic forces will play any appreciable role in its demise, will only be known in the fullness of time. In any event, over the next generation that regime must cope not only with the “marriage squeeze” and the “floating population” problem but with an upending of an extended family system that is as old as Chinese civilization itself. It is still difficult for us to imagine what China will look like, much less how things will work, with the rise of the “new family type.” We cannot yet dismiss out of hand the conjecture that this development could prove to be an existential, civilization-challenging event.
Finally: there is the question of China’s long-term fertility trends, and what the government may do to affect these. This question is not too pertinent to the demographic outlook for the year 2040, as we have already seen, but it is highly relevant to speculation about China 2050 and beyond.
It is important to recognize that the regime still holds the national birth rate to be a matter of state, not of parental choice. The adjustments to the population policy in recent years did not vitiate the regime’s claim for itself to the right to set national fertility levels—it merely raised the birth quotas the government would permit.
Now there are hints that Beijing may be toying with a population policy U-turn, a 180-degree shift to a pro-natal population policy. Some have noted, for example, this year’s new Lunar New Year postage stamps, featuring a cartoon of five happy pigs—two parents and their three children. Little signals like this are sometimes leading indicators for new political campaigns.
Absent government pressure, China’s “natural” fertility trajectory might well be further decline: after all, fertility levels today are decidedly lower than China’s in neighboring Hong Kong, Taiwan, South Korea, and Japan. They are lower in authoritarian Singapore too, despite that nation-state’s attempts to encourage births through three decades of pro-natal measures.
Could Beijing succeed where Singapore has failed? Police state power may be effective in forcing births down—but could it also force births up? In the 1960s, Communist Romania banned abortion without notice and doubled the nation’s birth rate the following year—but that was a one-off, and birth rates gradually returned to the abortion-era levels.
Beijing may have more sophisticated, and intrusive, tools at hand for any future pro-natal campaign. “Social media credit ratings” through fintech could be one of these: far-reaching financial penalties for those evidencing unpatriotic tendencies, including childlessness. Think of it as “market totalitarianism.” To date pro-natal policies around the world have met with at best limited success. But then again none yet have experimented with “market totalitarianism.”
Nicholas Eberstadt holds the Henry Wendt chair in political economy at the American Enterprise Institute and is a senior adviser to the National Bureau of Asian Research.
1 Population projections lose their reliability when demographers begin to makes guesses about how many babies the currently unborn are going to bear. There is no reliable basis for such projections today—and as long as human births are a matter of future individual volition, there presumably never can be. If we try to look more than a generation ahead for any population, we are thus getting into the realm of science fiction.
2 Yong Cai, “China’s New Demographic Reality: Learning from the 2010 Census,” Population and Development Review 39 no. 3, 371–396 (September 2013), http://doi.org/10.1111/j.1728- 4457.2013.00608.x.
3 Wei Chen “Evaluation of the Completeness of Birth Registration in China Using Analytical Methods and Multiple Sources of Data” Preliminary Draft from United Nations Expert Group Meeting on “Methodology and lessons learned to evaluate the completeness and quality of vital statistics data from civil registration” New York, (November 2016), http://www.un.org/en/development/desa/population/events/pdf/expert/26/notes/Chen_2016_Birth%20Registration%20in%20China.pdf.
4 Very roughly speaking, the raw labor input simplicitur may have contributed close to one quarter of China’s measured economic growth over those decades.
5 Ronald Lee and Andrew Mason, “Fertility, Human Capital, and Economic Growth over the Demographic Transition,” European Journal of Population. 26 no. 2:159-182 (2010),. doi:10.1007/s10680-009-9186-x.
6 Baozhen Luo, “China Will Get Rich Before It Grows Old,” Foreign Affairs, October 30, 2015, accessed September 24, 2018, https://www.foreignaffairs.com/articles/china/2015-04-20/china-will-get-rich-it-grows-old.
7 Nicholas Eberstadt and Judith Banister, The Population of North Korea (Berkeley: Inst. of East Asian Studies, Univ. of Calif., Center for Korean Studies, 1992).
8 OECD, “LFS by Sex and Age - Indicators,” Health Status, accessed September 24, 2018, https://stats.oecd.org/Index.aspx?DataSetCode=LFS_SEXAGE_I_R.
9 Fang Cai and Yang Du “Wage increases, wage convergence, and the Lewis turning point in China,” China Economic Review, 22 no. 4, p 601-610, (2011), https://econpapers.repec.org/article/eeechieco/v_3a22_3ay_3a2011_3ai_3a4_3ap_3a601-610.htm.
10 Formally, the “National New-Type Urbanization Plan (2014-20)”, http://www.gov.cn/zhengce/2014-03/16/content_2640075.htm . Cf. “China Unveils Urbanization Plan”, The Wall Street Journal, March 16, 2014, https://www.wsj.com/articles/china-unveils-long-awaited-urbanization-plan-1395024223; See also the World Bank’s and Development Research Center of the State Council, the People’s Republic of China related background study, “Urban China : Toward Efficient, Inclusive, and Sustainable Urbanization,” July, 2014, https://openknowledge.worldbank.org/handle/10986/18865.
11 Martina Miskolczi – Kornélia Cséfalvaiová, “Process of Population Ageing and its Dynamic,” presented at The 7th International Days of Statistics and Economics, Prague, September 19-21, 2013,
13 Nancy Birdsall et al., The East Asian Miracle: Economic Growth and Public Policy (Washington DC, 1993), accessed September 24, 2018, http://documents.worldbank.org/curated/en/975081468244550798/Main-report.
14 Note that Wittgenstein Centre relies on its own population projections in these estimates, not UNPD’s.
15 Feng Wang, Ke Shen, Yong Cai, Ding Li, Xiwei Wu. “Economic boom, population aging, and policy shift: What’s ahead for China?.” Fudan Working Group on Comparative Aging Societies, June 8, 2013.
16 Yi Zeng, Kenneth C. Land, Danan Gu, Zhenglian Wang, “ProFamy: The Extended Cohort-Component Method for Household and Living Arrangement Projections” in Household and Living Arrangement Projections from The Springer Series on Demographic Methods and Population Analysis, vol 36, (New York: SPRINGER, 2013), https://link.springer.com/chapter/10.1007%2F978-90-481-8906-9_2.
17 The World Bank, “China Economic Update - December 2017: Growth Resilience and Reform Momentum,” World Bank, accessed September 24, 2018, https://www.worldbank.org/en/country/china/publication/china-economic-update-december-2017.
18 International Monetary Fund “Fiscal Monitor: Capitalizing on Good Times,” April 2018, https://www.imf.org/en/Publications/FM/Issues/2018/04/06/fiscal-monitor-april-2018.
19 Marcos Chamon, Kai Liu, and Eswar Prasad, “Income Uncertainty and Household Savings in China,” Journal of Development Economics, 105th ser. (July 27, 2013), https://pdfs.semanticscholar.org/85c2/eb00498e554f69b691cef3c1b8660b3ba3ed.pdf.
20 Celia Hatton “New China law says children ‘must visit parents,” BBC, July 1, 2013, https://www.bbc.com/news/world-asia-china-23124345.
21 Nicholas Eberstadt, “The Global War against Baby Girls,” The New Atlantis, 2011, accessed September 24, 2018, https://www.thenewatlantis.com/publications/the-global-war-against-baby-girls.
22 Gavin W Jones, “The Flight From Marriage” in South-East and East Asia. Journal of Comparative Family Studies 36, no. 1 (2005): 93-119. http://www.jstor.org/stable/41603982.
23 Yi Zeng, Kenneth C. Land, Danan Gu, Zhenglian Wang, “ProFamy: The Extended Cohort-Component Method for Household and Living Arrangement Projections” in Household and Living Arrangement Projections from The Springer Series on Demographic Methods and Population Analysis, vol 36, (New York: SPRINGER, 2013), https://link.springer.com/chapter/10.1007%2F978-90-481-8906-9_2
24 Shang-Jin Wei, and Xiaobo Zhang, “The Competitive Saving Motive: Evidence from Rising Sex Ratios and Savings Rates in China,” NBER Working Paper. No. w15093, (June 2009). https://ssrn.com/abstract=1422971
25 Gary Stanley. Becker and Richard A. Posner, Uncommon Sense: Economic Insights, from Marriage to Terrorism (Chicago, IL: Univ. of Chicago Press, 2010).
26 Valerie M. Hudson and Andrea M. den Boer, Bare Branches The Security Implications of Asia’s Surplus Male Population (MIT Press, 2005).
27 Valerie M. Hudson, and Andrea M. Den Boer. “A Surplus of Men, a Deficit of Peace: Security and Sex Ratios in Asia’s Largest States.” International Security 26, no. 4 (2002): 5-38. http://www.jstor.org/stable/3092100.
28 James Lee and Feng Wang, One Quarter of Humanity Malthusian Mythology and Chinese Realities, 1700-2000 (Harvard University Press, 2001).
29 National Bureau of Statistics of China, China Statistical Yearbook 2017 (China Statistics Press, 2018), http://www.stats.gov.cn/tjsj/ndsj/2017/indexeh.htm.
30 UNDP, “World Urbanization Prospects - Population Division,” United Nations, accessed September 24, 2018, https://population.un.org/wup/.
31 Edward L. Glaeser, Triumph of the City: How Our Greatest Invention Makes Us Richer, Smarter, Greener, Healthier, and Happier (New York: Penguin Press, 2012), https://www.penguinrandomhouse.com/books/303439/triumph-of-the-city-by-edward-glaeser/9780143120544/
32 These are data from China’s 2010 census as officially reported. We know that UNPD believes the nation-wide figures involved an undercount of about 30 million—but since most of that undercount is thought to have been rural, this should not distract us here.
33 Jeremy Wallace, Cities and Stability, (Oxford University Press, 2014), http://www.oupcanada.com/catalog/9780199378982.html .
34 Guo, Z., Wu, Z., Schimmele, C., & Li, S, “The Effect of Urbanization on China’s Fertility,” Population Research and Policy Review, 31(3), 417-434 (2012), http://paa2012.princeton.edu/papers/120365 .
35 Li, Hongbin, Prashant Loyalka, Scott Rozelle, and Binzhen Wu. “Human Capital and China’s Future Growth.” The Journal of Economic Perspectives 31, no. 1 (2017): 25-47. https://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.31.1.25 .
36 Nicholas Eberstadt’s ongoing research for the Office of Net Assessment/Office of the Secretary of Defense contributed to this essay.