The world is experiencing a period of unprecedented demographic change. For the first time in human history, marked disparities in age structures exist across the globe. Around 40 percent of the world's population lives in countries with significant numbers of elderly citizens. In contrast, the majority of the world's people live in developing countries with very large numbers of young people as a proportion of the total population. Yet, demographically, most of the world's states with young populations are aging, and many are doing so quickly. This first-of-its kind systematic theoretical and empirical examination of how these demographic transitions influence the likelihood of interstate conflict shows that countries with a large number of young people as a proportion of the total population are the most prone to international conflict, whereas states with the oldest populations are the most peaceful. Although societal aging is likely to serve as a force for enhanced stability in most, and perhaps all, regions of the world over the long term, the road to a “demographic peace” is likely to be bumpy in many parts of the world in the short to medium term.


The world is going through a period of unprecedented demographic change. For the first time, marked disparities in age structures now exist across the globe. Around 40 percent of the world's population live in aging countries with significant numbers of elderly citizens. This group mostly includes the economically advanced countries, but also a few developing countries—most notably, China. In contrast, the majority of the world's people live in developing countries with very large numbers of young people as a proportion of the total population. As Somini Sengupta aptly concludes, “At no point in recorded history has our world been so demographically lopsided, with old people concentrated in rich countries and the young in not-so-rich countries.”1 And yet, most of the world's young states are themselves on the path to becoming much older, and many are doing so quickly.

Demography is not destiny, but it is an extremely powerful force. Economists, demographers, sociologists, geographers, political scientists, and others have written hundreds of studies illuminating how demographic change influences issues such as economic growth and decline, the welfare of women, the state of the environment, the likelihood of terrorism and civil conflict, the platforms of political parties, and migration patterns. But how do demographic transitions influence the likelihood of conflict between states?

Scholars of security and conflict have not yet provided a satisfactory answer to this question. Analysts have long recognized how population size contributes to a state's economic and military power, and some studies have posited a link between demographic changes and international conflict. All of these studies, however, analyze theoretically only some of the key elements of ongoing demographic transitions, and none undertake an empirical analysis of how these shifts influence the prospects for international conflict.2

The aim of this article is to help fill this gap by developing and testing a new general theory that we call “the demographic transition theory of war.” In so doing, we provide answers to two core questions. First, are young societies the most prone to international conflict? And second, are states with the oldest populations the most peaceful?

Regarding the first question, a number of studies have examined how so-called youth bulges contribute to violence and conflict within states, revealing that having a high ratio of people aged 15–24 among the adult population increases the likelihood of domestic terrorism and civil war.3 Yet no study has thus far examined, either theoretically or empirically, how youth bulges alter the probability of military conflict between states. Ours is the first to do so. Our empirical analysis finds that countries with a large number of young people as a proportion of the total population are more likely to engage in interstate conflict than are countries with fewer young people.

As for the second question, some scholars have theorized that old societies are likely to be more peaceful.4 Yet, the theoretical reasons why we would expect societal aging to lower the likelihood of conflict have not been fully delineated, and no scholar has evaluated this proposition empirically. Our empirical analysis shows that old countries are the least likely to engage in international conflict.

The remainder of the article is organized as follows. The first section reviews the essential features of how demographic change occurs and shows the dramatic variation in age structures that currently exists around the world. In the second section, we delineate the underlying theoretical reasons why countries with youth bulges are more conflict prone. We posit that the demographic characteristics of such societies cause both their motives (or “willingness”) and capacity (or “opportunity”) to engage in international conflict to be high.5 In the third section, we review the theoretical reasons why countries with a large proportion of old people are likely to be the most peaceful. Here, we maintain that the demographic characteristics of these societies cause their motives and capacity for engaging in international conflict to be low. The fourth section undertakes a statistical analysis of the relationship between country-level demographic variables and the probability of interstate conflict after World War II. The fifth section uses survey data to examine whether older individuals tend to support war to a lesser extent than do younger individuals. The sixth section lays out the implications of our analysis for the current major powers: the present hegemon (United States), the next two largest military powers (China and Russia), and core U.S. allies in Europe and Asia.

The Demographic Transition and Today's Demographic Diversity

Demographers have outlined a multistage process of population change called the “demographic transition model” (see figure 1). The first stage of the demographic transition—the one in which all countries had remained for millennia until around two centuries ago—displayed three key characteristics:

Figure 1.

The Demographic Transition Model

Figure 1.

The Demographic Transition Model

(1) very short lifespans (life expectancy was around twenty-seven years on average in all centuries prior to 1800); (2) very high birth rates (approximately six births per woman on average prior to 1800); and (3) relatively stable population sizes (as high death rates more or less offset high fertility levels, with occasional short-term spikes in mortality caused by pandemics, famines, or wars).6

In the early nineteenth century, a small number of so-called forerunner countries in the Northern Hemisphere, mostly in Western Europe, entered the second stage of the demographic transition: mortality rates declined, thereby increasing the average life expectancy. Rudimentary advances in technology and knowledge, not medical breakthroughs, primarily accounted for this change: people developed a better understanding of how disease was transmitted; women learned better child-bearing and rearing practices; water quality and sanitation improved, as did food storage; new farming practices such as crop rotation increased food production; and so on.

The reduction in mortality was especially notable for babies and infants. Before the nineteenth century, only around half of newborns survived past age 5. And as people started to realize that more of their children were surviving, they began having fewer of them. This decline in the fertility rate represents the key feature of the third stage of the demographic transition.

During the second and third phases of the demographic transition, the populations in these forerunner countries began to increase, because, for the first time in human history, birth rates were higher than death rates. Eventually, these societies entered the fourth stage of the demographic transition. This stage was marked by a significant decline in the fertility rate to around the replacement level (2.1 children) or even below it. With high levels of life expectancy and low levels of fertility, countries in this fourth stage have populations that are relatively stable in size and in which a growing proportion of the population is elderly.

For the first time, the world today possesses a significant and ever growing number of countries that are aging rapidly given their increasingly large proportion of elderly people (see figure 2). In the fifth stage of the demographic transition, the very oldest societies have populations so skewed in favor of the old relative to the young that their populations are shrinking. Japan is the archetype of this kind of society: since 2010, its population has declined by almost 1.5 million; looking ahead, the United Nations projects that Japan's population will decline by around 35 percent in this century (see table 1). Among the current major powers, only the population of the United States is expected to grow over the course of this century, by around 44 percent (see table 1).

Figure 2.

Proportion of Population Older than 60

Figure 2.

Proportion of Population Older than 60

Table 1.

Population Projections for the Current Major Powers, 2010 to 2100 (millions)

Population 2010Projected 2050 PopulationProjected 2100 PopulationProjected Population Change from 2010 to 2100Projected Percentage Change from 2010 to 2100
China 1,360 1,348 1,004 −356 −26 
Germany 83 75 63 −20 −24 
Japan 127 107 83 −44 −35 
United States 312 389 450 138 44 
Russia 144 129 117 −27 −19 
Population 2010Projected 2050 PopulationProjected 2100 PopulationProjected Population Change from 2010 to 2100Projected Percentage Change from 2010 to 2100
China 1,360 1,348 1,004 −356 −26 
Germany 83 75 63 −20 −24 
Japan 127 107 83 −44 −35 
United States 312 389 450 138 44 
Russia 144 129 117 −27 −19 

SOURCE: United Nations, “World Population Prospects, 2015” (New York: United Nations, 2015), https://esa.un.org/unpd/wpp/.

NOTE: All figures use medium variant.

In the Northern forerunner countries, the demographic transition described above began in the nineteenth century, took roughly two centuries to complete, and occurred largely in response to changes from within these societies. Throughout most of Asia, Africa, and the Americas, in contrast, the demographic transition began only after World War II, has been occurring in a matter of a few decades, and began largely in response to actions that occurred outside these regions of the world—namely, medical advancements and technologies (including vaccines, immunization programs, oral hydration therapy, and antibiotics) that were created mostly in other parts of the world and then imported into countries in these regions.7

Significantly, the gap between the decline in mortality and the decline in fertility was far shorter in the Northern forerunner countries (5–10 years on average) than it has been among most developing countries (30–40 years on average). In the forerunner countries, writes David Reher, “mothers who participated in the initial decline of mortality were also the ones initiating fertility limitation later in life.” In most developing countries, in contrast, “fertility control essentially skipped a generation. The women who first began controlling their fertility were probably those girls who had grown up in crowded households with decreasing living standards and who went on to experience lower than expected childhood mortality during the early part of their own reproductive lives.“8

The relatively large gap between declining mortality and declining fertility in most developing countries had two straightforward and dramatic repercussions: (1) population growth exploded: it took all of human history to reach a population of 1.6 billion at the beginning of the twentieth century, but in little more than a century (in 2011) the world's population reached 7 billion; and (2) the age structure of these developing societies generally became extremely skewed toward the young; for the first time, a very large number of countries had youth bulges (see figure 3).

Figure 3.

Proportion of Population Younger than 30

Figure 3.

Proportion of Population Younger than 30

In sum, the demographic transition has played out in different ways in different parts of the world, with the result being a dramatic divergence in age structure. Over time, however, the number of old societies will progressively increase. As Ronald Lee writes, “Dramatic population ageing is the inevitable final stage of the global demographic transition, part and parcel of low fertility and long life.”9 Indeed, the process of aging in many developing countries is likely to be far more rapid than it was in the forerunner countries. As Reher observes, “The pace of the demographic transition among the forerunners was a leisurely one, giving rise to a gradual process of ageing.” In contrast, for many developing countries that “initiated their own transformations more recently …the breakneck pace of the demographic transition” means that their experience “of explosive population growth with extremely young populations” is likely to be followed by “one of rapidly ageing populations and diminishing supplies of labor in very rapid succession. This is not a matter of opinion or an educated guess, but rather the most likely scenario for the future of much of the world.”10

In short, a great many countries now have extremely young populations, but over time the general trend is toward an increasing number of old societies (see figure 2). The next two sections offer theoretical examinations of how these ongoing demographic changes are likely to influence the prospects for international conflict.

Why Youth Bulges Are Likely to Enhance International Conflict

There are three theoretical reasons why we expect youth bulges to increase the likelihood of international conflict. First, a youthful population enhances the capacity of states to engage in military conflict. Second, grievances and an augmented capacity for radicalization in youth-bulge societies increase their willingness for conflict. Third, youth bulges increase the probability of domestic conflict, which can heighten the likelihood of international hostilities.


Young states are likely to have an enhanced capacity for military conflict for two main reasons. The first concerns the low per unit costs of military personnel in young societies; the second pertains to the influence of the “working-age” bulge on the long-term economic capacity of these societies.

“per unit soldier cost” mechanism. Scholars emphasize that youth bulges augment the military capacity for domestic conflict because a surplus of young men significantly lowers the costs of recruiting rebel soldiers and replacing fallen ones. As youth bulges develop, the pool of potential recruits grows, making it easier for rebel leaders to form and maintain an army.11

This logic is generalizable to all states and to all forms of conflict. In other words, this “per unit soldier cost” mechanism affects the ability of all states, not just those with youth bulges, to engage in conflict. For young societies, a surplus of youth makes the recruitment and replacement of all soldiers easier, which gives youth-bulge states a greater capacity to engage in all forms of hostility. For old societies, the relative scarcity of young adults means that per unit soldier costs are high, which ceteris paribus reduces the capacity of old states to engage in conflict.

Regarding this last point, states at advanced stages of the demographic transition are prone to spend more of their military budgets on personnel and less in other areas, including weapons development and procurement. As the fertility rate of these states declines, so does the supply of military-age labor; as a result, the per unit cost of each soldier increases (assuming constant demand). The militaries of these states therefore will have to pay more to attract and keep the best soldiers in vital areas of operation—especially soldiers in high-technology fields, who usually have the best employment options and can command high salaries in the private sector.

This dynamic is most pronounced among voluntary militaries, which have to offer good wages to attract recruits. Yet, even conscripted armies have incentives not to let the gap between civilian and military wages become too large, lest dissatisfaction in the ranks becomes a problem. Between 1960 and 2000, the percentage of individuals in the prime military-age range of 15–24 declined in three regions of the world: Europe, North America, and eastern Asia (including Australia and New Zealand). For governments in all three regions, the amount of money spent per soldier increased, often dramatically. The North Atlantic Treaty Organization (NATO) powers in Europe and North America averaged a 58 percent increase, in real terms, in the unit costs of their military personnel between 1975 and 1999. Japan, Australia, and New Zealand experienced even more dramatic increases in costs per soldier, with the amounts rising by more than 150 percent from 1981 to 1995.12

The evidence shows that states with aging populations and shrinking military-age cohorts have generally devoted increasing percentages of their defense budgets to personnel costs. This was true, for example, for all of the members of NATO except France, Norway, and Portugal from the 1970s to the late 1990s.13 Unless military budgets are expanding, increasing personnel costs as a percentage of these budgets will necessitate cuts in other areas, which will diminish the capacity to project power, all other things being equal. This reduction, in turn, will shrink the opportunity to initiate international hostilities.

In sum, low per unit soldier costs in societies with an excess of youth increase military capacity for all forms of conflict, whereas the high per unit solider costs in older societies reduce it.

the influence of a working-age bulge. Swift population growth during the middle stages of the demographic transition creates a temporary working-age bulge in which those aged 25 to 64 constitute the majority of the population. This bulge, in turn, creates a highly favorable environment for enhanced economic growth. (States, of course, still need to have the appropriate policies and institutions in place to take advantage of these favorable demographic conditions.) During this period, there are relatively few young and old dependents demanding resources from families and governments; household savings rates tend to increase, which facilitates investment as interest rates decline; and family caregivers (usually women) are freer to enter the workplace, because they tend to have fewer children in an era of shrinking fertility. Analysts frequently refer to this onetime working-age bulge as the “demographic bonus” or “demographic dividend.”14 According to one study, differences in the sizes of states' working-age populations accounted for 40 percent of the international variation in gross domestic product (GDP) from 1980 to 2010.15 Another study estimates that demographic dividends accounted for roughly one-third of the growth by the “Asian Tigers” (Hong Kong, Singapore, South Korea, and Taiwan) in the second half of the twentieth century.16

The enhanced economic conditions associated with the working-age bulge are relatively short lived, however, usually lasting between 30 and 50 years.17 As the demographic transition enters its later stages and fertility rates fall while life expectancy rises, the workforce bulge shrinks. As these workers begin to retire, the percentage of elderly people increases. Eventually, more people will leave the workforce than enter it, and the working-age bulge will transition into an elderly one. Growth tends to slow in countries that no longer have an expanding workforce.18 Moreover, advances in productivity will likely slow in aging societies. And as seniors spend down their savings, national savings rates are likely to decrease, contributing to the crowding out of investments as interest rates rise. The elderly also are likely to be more conservative with their investments than are younger people, resulting in lower levels of funding for entrepreneurship.19

In short, the youngest states are very poor. But as states move through the demographic transition and become “somewhat young,” they receive a onetime demographic boost to their economic growth that enhances their capacity to purchase and generate the military technologies and equipment needed to project power. With this change comes an increased likelihood of international conflict.


Scholars who analyze the domestic consequences of youth bulges argue that young people in countries with young populations often feel aggrieved because, unlike their older cohorts, they desire independence but lack the political, social, and economic means to achieve it.20 Labor markets have difficulty absorbing a rapidly growing number of job seekers in youth-bulge eras. Therefore, young adults in these periods are likely to experience particularly high levels of unemployment.21 As a result, they are likely to become frustrated with the economic and political status quo. Bleak economic opportunities also lower the opportunity costs for individuals to join rebellious groups that challenge their government, thereby increasing the odds that more will do so.

Analysts of youth bulges also argue that young adults tend to be relatively more susceptible to radicalization and violent ideologies. They want to be integral members of a group or a community and to have a purpose for their lives. As Saul Levine notes, “Many young people are susceptible to the overtures of groups that offer clarity, simplicity, communality, and an overriding belief in the righteousness of a cause.”22 Too frequently, movements that possess these characteristics are extremist and violent, thereby increasing the probability of armed conflict. Moreover, because political leaders are likely to be aware of the potential attractiveness of radical ideologies in young societies, they may appeal to these ideas during mobilization campaigns.23 The more that elites use extremist ideas to form and rally their political base, the more likely it is that policies will be extremist as well.

These features of youth-bulge societies have implications not just for domestic conflict, but for the likelihood of international hostilities. Enhanced grievances and an augmented susceptibility to radicalization are ultimately fuel for youth-bulge societies to engage in all forms of conflict.


For three reasons, the increased likelihood of domestic conflict in youth-bulge societies is linked to an enhanced risk of international conflict. First, by raising the probability of rebellion and civil conflict, youth bulges increase the likelihood of revolutions as well: the more rebellions that occur, the greater the number that will succeed. Scholars such as Jack Goldstone have long recognized the close connection between youthful societies and successful rebellion.24 And in a recent statistical analysis, Tongfi Kim and Jennifer Dabbs Sciubba found that “a higher youth ratio significantly increases the likelihood of irregular removal of leaders in a state.”25

Revolutions undermine the prospects for international peace. As Stephen Walt documents, revolutionary regimes tend to be a source of international hostilities because they result in destabilizing shifts in the balance of power, increase the level of uncertainty and misinformation among states' leaders, and bring to power highly ideological leaders who may consider war with others as inevitable and advantageous.26

Second, governments have strong incentives to defuse domestic hostilities and armed rebellions that emerge as a result of youth bulges, and some of the options for doing so may increase the likelihood of international war. State leaders could stoke nationalist passions or provoke an international crisis in an effort to diminish domestic discontent and division.27 Governments could also try to minimize the civil threat created by a surplus of young adults by expanding the size of their militaries to better control youthful populations. Historically, this tactic has often increased the probability of international hostilities by increasing the power at states' disposal or by fueling rivalries with other states.28 A prominent example is Iran in the 1980s: as a Central Intelligence Agency report concludes, the Iranian government “turned the youth bulge to the task of eliminating perceived ‘enemies of the state.‘” And although the Iran-Iraq War was not “initiated with the idea of absorbing the youth bulge, [it] accomplished that end. The wholesale conscription of young adults in Iran… accomplished the absorption of youth into the national fabric.”29

Third, civil conflicts can create incentives for international interventions, including those that use force. Foreign powers may decide to intervene in other states' civil conflicts to prevent the violence from spreading, to help empower preferred groups, to exploit opportunities created by the weakening of the contested state, or to offer humanitarian relief.30

Since the end of World War II, a number of intra-state armed conflicts have become “internationalized” (in which one or more states contribute troops to the warring sides in the civil conflict); this trend has only intensified in the twenty-first century.31 According to the Uppsala Conflict Database, 292 intra-state armed conflicts with a minimum of 25 battlefield casualties were internationalized from 1946 to 2012.32 Such foreign interventions can result in international conflict if hostilities escalate between the intervening troops and the target state's forces, or between rival foreign countries intervening in the same intra-state dispute. According to the Correlates of War project, from 1823 to 2000, there were forty-two civil or nonstate wars that developed into interstate or extra-state wars that involved at least 1,000 battle-related deaths among the participants.33 Many more internationalized intra-state conflicts have resulted in international hostilities that fell short of this definition of full-scale war.


The discussion above reveals that countries with younger populations have the following characteristics: (1) higher availability of military personnel, (2) augmented motives for conflict, and (3) an enhanced risk of domestic conflicts being externalized. The youngest societies, however, will not yet have accrued the long-term economic benefits of having a large pool of working-age adults; and because these societies are poor, they will have a highly constrained capacity to develop or purchase military technology and equipment, especially the kinds needed to project power beyond their countries' borders. It is only later, when countries progress further through the demographic transition to become somewhat young societies, that they will be able to economically benefit from having a working-age bulge.

Taken together, these arguments lead to the following nonlinear expectation: although all young societies will be relatively prone to conflict, it is the somewhat young societies that have shifted toward the middle portion of the demographic transition that will be especially likely to engage in international conflict.

Why Old Societies Are Likely to Be the Most Peaceful

In this section, we explore the theoretical reasons why we would expect old societies to be the most peaceful. We first examine how the demographic characteristics of old societies are expected to influence the capacity of these societies for international conflict. We then analyze demography's influence on their motives regarding war.


Old societies at the advanced stages of the demographic transition are likely to have a reduced capacity for conflict for two core reasons. The first reason pertains to the effects of societal aging on the ability of these states to secure funds for military spending, what we term the “crowding out” effect. The second concerns a dynamic explored in the previous section: the high per unit costs of military personnel in old societies.

crowding-out effect. Old societies are likely to experience significant strain on their finances, and thus the ability to secure funds for military spending, even before their working-age bulges disappear. Substantial increases in a country's number of senior citizens as a percentage of its population tax public resources as governments face growing pressure to increase spending on their welfare. (By 2000, almost every country in the world had in place benefits programs of some kind to aid the elderly.)34 As a society ages, governments are not only likely to be responsible for a rising number of seniors, but their expenditures per senior will tend to increase. Studies have shown that the elderly consume three to five times the health-care resources of younger people.35 The strains on a state's finances will become even more intense when its working-age bulge disappears, with the likely result being a slowing of economic growth.36

Senior citizens make up substantial proportions of many of the populations of today's societies, and the associated costs to governments have increased commensurately. By the beginning of the twenty-first century, individuals aged 60 or older constituted at least 9 percent of the total population in three regions of the world: Europe (20 percent), East Asia (12 percent), and Latin America and the Caribbean (9 percent). As these populations grew, so too did governmental spending on their welfare.37 In Latin America, average social-welfare spending on the elderly increased by roughly 3.5 percent of GDP between 1980 and 2000.38 From 1970 to 2000, Japan's expenditures in this area increased by 10.4 percent of GDP and Taiwan's by 6.6 percent. Between 1980 and 2000, South Korea's expenditures on social security increased by 10.2 percent of GDP.39

Significant increases in social-welfare expenditures have a crowding-out effect: the more that governments spend on elderly care, the less money they can spend in other areas, including the military. From 1960 to 1995, for example, governmental expenditures in leading states of the Organisation for Economic Co-operation and Development increased by 4.7 percent of GDP. This increase, though, was not nearly sufficient to cover rising public expenditures (more than 10 percent of GDP) for health care and pensions, much of which were dedicated to seniors. The result was less money available for other areas, including the military. During that period, military spending fell by 1.4 percent of GDP.40 Similar declines in military spending occurred in Latin American and East Asian countries from the early 1970s to the mid-1990s.41

high per unit soldier costs. Earlier, we noted that the relative scarcity of young adults in old societies because of low levels of fertility leads to high per unit soldier costs, which ceteris paribus is likely to reduce the likelihood of conflict: when both the pool of potential military recruits shrinks and the amount of money invested in each individual soldier rises as per unit labor costs grow, leaders' aversion to casualties is likely to intensify as they seek to conserve a valuable resource.42 There is, however, an important caveat to this argument. States can protect soldiers while still projecting force if they adequately substitute technology for personnel. States, whether democratic or authoritarian, tend to take precisely this step as their wealth increases.43 This strategy could allow states to protect a valuable resource and minimize casualties while maintaining combat power.

It can be argued, therefore, that if such states can shift from labor-intensive to technology-intensive warfare, they may not face this demographic constraint on initiating conflict. There are three rejoinders to this argument, however. First, even if states are able to substitute capital for labor while maintaining or even enhancing combat power, military missions such as counterin-surgency, nation-building, and militarized humanitarian operations will continue to be labor intensive.44 Second, even for non-labor-intensive operations, some military personnel will be at risk. Third, many states do not have sufficient economic resources to be able to substitute enough capital for labor to maintain a given level of combat power. For these countries (e.g., developing countries or developed ones that operate in eras of budget austerity), rising unit labor costs will consume an increasing percentage of a state's military budget without enhancing its power-projection abilities. In this scenario, a government's aversion to military casualties is likely to grow.


Old societies are likely to have lower motives for engaging in conflict for two reasons. The first, what we call the “family cost mechanism,” concerns the increased public aversion to casualties as family sizes shrink. The second, what we term the “societal preferences mechanism,” refers to the greater aversion to war among older citizens.

family-cost mechanism. According to the “family-cost” mechanism, societies characterized by smaller family sizes are likely to be more averse to casualties—and thus to warfare—than are societies with larger families. The notion that shrinking fertility rates are linked to growing casualty aversion is rooted in a central finding of numerous multistate surveys that began in the 1970s, the purpose of which was to ascertain people's primary motivations for having children.45 To our knowledge, we are the first to link these data to claims about casualty sensitivity. A key finding of the surveys, as Rodolfo Bulatao summarizes, is that “parents anticipate different rewards and costs from children, and that these change over the family-building cycle.”46 In these surveys, the children born early in the birth order tend to confer more emotional and psychological benefits on their parents (e.g., to have someone to love and care for, increased joy and self-esteem created by having a family, and knowledge that someone will carry on the family name). At some point in the family-building process, these surveys indicate that parents' principal benefits for having more children shift from emotional/psychological benefits to more material ones, such as receiving help with farming or having caregivers as parents age. The surveys from the 1970s show that this shift occurred after the fifth child.47 Surveys from the early 2000s indicate that it occurred after the second child.48

The finding that parents around the world and across multiple generations tend to attribute different benefits to different children in the birth order potentially has critical effects on attitudes toward child deaths and casualty sensitivity. There are substitutes for the material benefits that children bring to parents. There is, however, no substitute for many of the emotional and psychological benefits created by children who come early in the birth order.49 As Bulatao concludes, “The benefits of children [to parents] are substantially greater earlier in the [family-building] process.”50

If the benefits of children born early in the birth order are particularly high because they have no substitute, then the loss of children in low-fertility households is likely to be especially great.51 Parents mourn not only the loss of the child (which is pain that is almost certain to be similar no matter how large a family), but also the material, emotional, and psychological benefits the child creates, which vary by birth order and family size.52 These relationships have obvious implications for the public's support for war: societies generally characterized by smaller families are likely to be more casualty sensitive and thus more opposed to war than are societies characterized by larger families.53 Moreover, this transition to greater casualty sensitivity should occur when the primary benefits of additional children switch from substitutable material advantages to nonsubstitutable emotional and psychological ones.

Numerous studies show that societies over the last fifty years have become more sensitive to casualties as their fertility levels have fallen.54 This evidence is particularly robust for democracies, where public opinion data are the most reliable and individuals have the most freedom to express their views to help shape public policy. Significantly, similar evidence is present even in some societies with a high threat environment—most notably, Israel. Yagil Levy documents that casualty aversion and the “politicalization of bereavement” in Israel in the 1980s resulted in a growing number of protests against conflict.55 The protests forced the Israeli government to institute new procedures for investigating and trying Israeli military commanders when Israeli soldiers are killed. They also led to greater tolerance for killing Palestinian civilians if that meant fewer casualties for Israeli citizens. Israeli officers have acknowledged that the “military [has] internalized the public expectation of casualty-free wars,” and key political leaders have asserted that Israeli citizens' casualty sensitivity has been a major factor shaping their military choices.56 The dramatic drop in Israeli fertility rates in the 1970s (from more than five children per woman in 1970 to fewer than three children in 1980) likely contributed to this increased casualty aversion.

Other countries have also tailored their military policy and decisions based on their perceptions of the public's level of casualty aversion. John Gentry documents that, since the 1970s, the United States has “embraced the goal of ‘force protection,’ which in extreme forms effectively places the protection of troops and commanders' careers above mission accomplishment.”57 In a statistical analysis using data from 1900 to 2005, Benjamin Valentino, Paul Huth, and Sarah Croco found that casualty sensitivity in democracies pushes these states to adopt four military strategies designed to minimize combat deaths: increasing military spending in wartime, forming more powerful alliances during wars, utilizing battlefield tactics (e.g., high reliance on airpower) that minimize military fatalities, and fighting on battlefields not contiguous to their territory.58

societal preferences mechanism. States at the tail end of the demographic transition have older-age cohorts constituting an ever increasing percentage of a country's total population. Based on current trends, the United Nations projects that the number of people in the world older than 60 will more than triple by 2100: from 0.9 billion in 2015, to 2.1 billion in 2050, and then reaching 3.2 billion in 2100. And as a proportion of the world's total population, the number of people older than 60 will grow from around 12 percent now to slightly less than 30 percent in 2100.59

As older-age cohorts grow as a percentage of a state's population, their political power, especially in democracies, is likely to increase—especially because, as noted, older voters are much more likely to go to the polls than are younger people. According to the “societal preferences” mechanism, if older people tend to be more opposed to war, then their increasing political power would create an enhanced disposition to peace and greater barriers to conflict.

A key reason why older people might be more averse to conflict involves the previously mentioned crowding-out effect. Just as it is true that the more that governments spend on elderly care, the less money is available to spend on the military and all other purposes, the same is true in reverse: the costs of war or aggressive foreign policies will leave less money available for the elderly and all other purposes, which could force politicians to reduce social-welfare spending on seniors. The elderly thus have reasons to believe that wars or aggressive foreign policies are likely to place governmental expenditures that are made for their benefit in jeopardy.60


Ultimately, there are a series of overlapping theoretical reasons why we would expect the demographic characteristics of old societies to make them less likely to engage in international conflict: the effect of their demographic structure on their motives, military capacity, and long-term economic capacity all point in the same direction in this regard. This provides the basis for a strong theoretical expectation: states with the oldest populations will be the most peaceful.

Demography and International Conflict: A Cross-National Analysis

In this section, we use regression analysis to evaluate the proposition that old societies are the most peaceful and that young societies—especially those that are somewhat young—are the most prone to international conflict. We do so using an undirected dyad framework, the full details of which are described in the online appendix.61 In all models, the dependent variable is the onset of militarized interstate disputes. We restrict our analysis to militarized interstate disputes that involve the use of force.

We separately consider the effect of four different demographic variables: (1) the median age of the population; (2) the youth-bulge ratio (the ratio of people aged 15–24 to the total adult population); (3) the fertility rate (the average number of births expected per woman over her lifetime); and (4) the estimated life expectancy at birth. Data for each of these estimates were obtained from the United Nations Population Division. In each case, we model the relationship using polynomial functions to account for the possibility of nonlinear relationships to conflict onset.

Because rich countries tend to become old countries, we include GDP per capita (the standard measure of wealth) in our analyses. Each of our models also includes the following control variables that are commonly used in dyadic models of interstate conflict: contiguity, distance between capital cities, common alliance membership, level of bilateral trade, shared democracy, capability ratio, and time elapsed since the last conflict.


Figures 47 show how the probability of conflict varies as a function of changes in the level of each of these demographic variables, while all of the other variables are held constant at their median levels. To convey the uncertainty surrounding these estimated probabilities, we chose to present the results in the form of “spaghetti plots.”62 The uncertainty around the estimate can be judged simply by the extent to which the lines overlap: points in the graph where the lines are clustered more closely together indicate higher levels of confidence in the relationship between these demographic variables and conflict, whereas in the areas where the lines drift further apart (usually at the extreme ends of the distributions of each variable where there are few observations), our levels of confidence are much lower.63

Figure 4.

Median-Age Model

Figure 4.

Median-Age Model

Figure 5.

Youth-Bulge Model

Figure 5.

Youth-Bulge Model

Figure 6.

Fertility Model

Figure 6.

Fertility Model

Figure 7.

Life Expectancy Model

Figure 7.

Life Expectancy Model

To clearly demonstrate the average effect of these demographic variables, we provide an additional line (shown in white) that shows the mean of the predicted probabilities at each level of these variables. We include a histogram below each graph to show how the variable is distributed among the data used in our analysis.

Figure 4 shows the results for the model of median age—the indicator that is the best overall measure of a country's age structure. These results indicate that states with the oldest populations are the most peaceful and that young societies—especially those that are somewhat young—are the most conflict prone. The probability of conflict peaks when the median age (or, to be more precise, the lower of the median ages of the two members of the dyad) is 20 years, before beginning to decline. By the time median age reaches 30 years, the probability of conflict falls to about one-quarter of its peak value.

As for the models of youth bulges, figure 5 shows that those societies with the smallest proportions of young people—that is, the oldest societies—are the most peaceful. We reverse the x-axis to place those states with lower ratios of youth to the total population on the right-hand side of the figure.

Unlike the analysis in figure 4, the analysis shown in figure 5 does not suggest that somewhat young societies are the most conflict prone; instead, it appears that the probability of conflict is relatively constant among all young societies, but then begins to drop sharply once the proportion of people aged 15–24 falls below about 30 percent of the total population. By the time the proportion of young people reaches 20 percent of the adult population, the probability of conflict is less than one-quarter of its value when the young constituted 30 percent or more of the adult population.

The results for the fertility model shown in figure 6 suggest that as states' fertility rates begin to decline, the probability of interstate conflict at first begins to increase. (Once again, we reverse the x-axis to ensure that societies with lower fertility levels—those in the later stages of the demographic transition—appear on the right-hand side of the figure.) The probability of conflict reaches a peak when the fertility rate of the more fertile state in the dyad reaches around 5.6 births per woman. As the fertility rate declines beyond this level, the probability of conflict also starts to fall. By the time the fertility rate reaches two births per woman, the probability of conflict drops to about one-third of its peak level. Notice that these fertility numbers closely correspond with the transition point, as identified by the Value of Children surveys from the 1970s cited earlier, when the primary benefits of additional children shift from substitutable material benefits to nonsubstitutable emotional and psychological ones. In the 1970s, this transition point occurred after the birth of a family's fifth child and has declined since.

Finally, the results for the life expectancy model in figure 7 indicate a similar inverse U-shaped relationship. All else being equal, the probability of conflict increases over the lower ranges of life expectancy before reaching a peak once life expectancy reaches 58 years. Further increases in life expectancy are then associated with a significant decline in the probability of conflict. By the time life expectancy reaches 75 years, the probability of conflict falls to about one-third of its peak value.

To conclude, the four analyses above indicate that states with the oldest populations are the most peaceful. In turn, they show that young societies are the most prone to conflict. And regarding youth, three of the four models indicate that it is the countries that are somewhat young that are the most dangerous.

Age and Attitudes toward Conflict

In this section, we examine individual-level data to analyze whether the elderly are less supportive of war than younger individuals. Decades of results from U.S. surveys indicate that this is indeed the case. A 2006 Pew analysis of surveys from the early 1960s to the mid-2000s concludes that “nearly four decades of survey data show…there is a generation gap over U.S. military interventions,” with elderly Americans exhibiting “the greatest wariness about using military force. This was evident during the war in Vietnam and remains the case today.“64 In an analysis of 200 surveys of eleven different conflicts conducted by the United States military from 1964 to 2006, Val Burris similarly concludes that there is a ”relative lack of enthusiasm for war among elderly persons.“65 And in their examination of U.S. casualty sensitivity, Christopher Gelpi, Peter Feaver, and Jason Reifler find that ”age is …a significant predictor of casualty tolerance; older respondents [in surveys] are less willing to tolerate casualties than younger ones.“66 Moreover, age differences remain even after wars conclude: a Gallup survey found that ”older Americans are more likely than younger ones to say the wars in Iraq, Afghanistan, and Vietnam were a mistake.“67

All of these survey data showing the elderly tend to be “doves” lend credibility to the general notion that age and attitudes toward war are linked in the direction we posit. In addition, the consistency of these U.S. survey results across different time periods strongly suggests that “cohort” differences between generations are not a factor. In other words, despite a dearth of longitudinal studies charting changes in individual attitudes toward conflict over time, a substantial body of indirect evidence suggests that Americans become pacific when they become old.

Significantly, surveys of older people in countries other than the United States reveal similar results. For example, scholars have found that in the United Kingdom older individuals were less likely than young people to support intervention in the Iraq and Afghanistan missions.68 The same finding emerges with respect to German support for intervention in Iraq.69 And in France, older people are significantly more likely to identify themselves as “antimilitarist” than are younger people.70

The consistency of these findings across so many different surveys and countries is significant. Nevertheless, isolating the effects of age can be challenging, because every war is unique, involving different objectives, different cues to the public from political elites, different media coverage, and other distinctive features. In particular, it can be hard to assess whether the extent of the age gap might be dependent on key features of the wars in question.

To probe this issue further, we reanalyzed data from a study conducted by Deborah Jordan Brooks and Benjamin Valentino that examined gender and support for war in the United States using a survey-based experiment they conducted through YouGov/Polling Point.71 We added age as an independent variable to their experiment on the stakes of war (which randomly assigns an otherwise identical conflict over oil or humanitarian goals in two fictional countries to more than 600 respondents). After reading a fictional newspaper article about the conflict, survey participants answered a series of questions about the scenario and standard demographic questions.72

Under these more controlled conditions, our reanalysis of the Brooks and Valentino data confirms that elderly Americans are less likely to support war than their younger counterparts; moreover, our analysis demonstrates that this relationship is not dependent on the type of conflict in question.73 We obtain that same result with and without control variables.74 In addition, our finding holds whether we construct age as a continuous variable or whether we compare those 65 years and older with those aged 30–44 and 18–29. In short, this examination shows that the elderly support war to a significantly lesser degree than younger individuals, even when we can control for a variety of factors that are inherently noncontrollable in regular surveys, including the stakes of a conflict.

Implications for the Current Major Powers

Our analysis of the effects of societal aging has implications for the current major powers: the present hegemon (United States), the next two largest military powers (China and Russia), and core U.S. allies in Europe and Asia. Although all these countries are aging, the speed of their aging processes and their current levels of aging vary significantly; this variation will have important consequences for security policy in the decades ahead.

Of course, demography does not determine the fate of a country and its policies. Leaders can try to disregard the incentives and pressures created by aging, but the costs for doing so will increase the older a population becomes.75


The scope and pace of China's aging process and their significance for the country's security policy are arguably greater than for any of the other current major powers. China is about to become extremely old, extremely quickly. By 2040, China's projected median age (47 years) will exceed that of the world's oldest state at present (Japan, which in 2015 had a median age of 46.3). To put this in perspective, it was only in 2000 that any country had ever reached a median age above 40 (Japan at 41.2 in 2000).76

demographic mechanisms and china's security policy. China's military power is already being affected by population aging in ways consistent with our analysis. First, aging in China is contributing to slowing economic growth. As early as the mid-2000s, China experienced costly labor shortages that in large part could be traced to population aging and reduction in the number of those aged 15 to 35.77 China's working-age population has peaked and will decline precipitously in the decades ahead. As a point of comparison, another large country in East Asia that leveraged a working-age bulge to create a sustained period of export-led growth was Japan, whose economic prospects plunged once its working-age population began to contract dramatically starting in the 1990s. China's current working-age population profile is comparable to Japan's in the early 1990s and is projected to follow the same pattern (see figure 8).

Figure 8.

Working-Age Populations in China and Japan

Figure 8.

Working-Age Populations in China and Japan

Second, the Chinese government's social-welfare spending on the elderly has increased substantially in recent decades and is poised to balloon in the decades ahead, likely creating significant crowding-out dynamics. Between 1998 and 2005, governmental spending on social welfare relief and pensions (most of which went to seniors) increased by 1.7 percent of GDP.78 This increase was close to what China spent annually on the military, which has been roughly 2 percent of GDP throughout the twenty-first century.79 China's public expenditures on the elderly are predicted to rise steeply in the years ahead, increasing by nearly 8 percent of GDP between 2010 and 2040.80

Third, according to data released by China's government, rising personnel costs have played a major role in the country's expanding military budget since the end of the Cold War.81 Even though the military budget has grown significantly over the last twenty years, and even though China has reduced the size of its military in the same period by roughly a million soldiers,82 the percentage of its military budget dedicated to personnel costs has remained nearly the same at close to one-third of expenditures, resulting in mounting per unit labor costs.83 China today compared to the preceding decades is thus spending much more money on far fewer soldiers. As a result, China's military capabilities, in both absolute terms and those relative to the United States, are not growing nearly as much as China's overall increases in military spending would seem to indicate.84

These increasing personnel costs have resulted largely from efforts to keep military wages on par with those in the civilian economy, which have grown significantly in the post–Cold War period.85 China's military, according to the Economist, is “experiencing higher inflation in defense than in the economy as a whole. In the past it has relied for strength on the sheer number of lower-paid recruits to its armed forces. Now it must form a more professional and technically adept force. That means paying salaries to compete with those of skilled workers in civilian life, where wages have recently far outstripped inflation.” Consequently, “a big cost advantage that China enjoyed is now being eroded.“86 These inflationary pressures are likely to continue to escalate as the size of its military-age cohort is cut in half over the next three decades.

Although the standard projections for China's demographic future are dire, they may even be too optimistic. In an effort to boost its fertility rate and reduce the extent of its aging problem, China in 2015 ended its one-child policy in favor of a policy that allowed and even encouraged parents to have up to two children. This change, however, has not had the intended results.87 Findings from a 2015 census released by China's government in October 2016 revealed that the country's fertility rate was 1.05, making it the lowest in the world.88 This number is half a child (or 32 percent) lower than the figure provided by the United Nations. If the 2015 census data are accurate and this fertility level is sustained, then the extent and pace of China's aging—in terms of increasing median age and shrinking working-age and military-age cohorts— will be considerably higher than currently predicted. This would be notable given that China already appears to be aging faster than any other state in history.

the world's first old rising power. China is in a fundamentally different demographic position from all other previous rising powers; the frequent comparisons between its ascent and that of past rising states (e.g., Germany in the first half of the twentieth century) are thus inaccurate. Unlike these states, China is ascending at a time when its population is rapidly aging, which will hamstring it both economically and militarily. Breaking out of the so-called middle-income trap is extremely hard, and the relatively small number of countries that have done so in the post–World War II period have been comparatively young.89

Moreover, in the past, the leading state and the rising state have had similar demographic profiles. China, in contrast, will be rapidly aging just as it approaches economic parity with the United States, which is aging at a much slower rate. China's rapid aging—in both absolute terms and especially relative to the United States—will result in arguably the swiftest change of demographic fortunes in history.

Because China will become old before it can become a global peer of the United States, the likelihood of a dangerous power transition involving China and the United States is much lower than analysts often assert.90 Nevertheless, there is still cause for concern given the effects of closing window-of-opportunity dynamics. As the costs of China's rapidly aging population begin to surge, China's leaders may feel that they confront a shrinking demographic window of opportunity to achieve international objectives that are likely to require the use or threat of force.91 This dynamic may help to explain China's recent move away from its “peaceful rise” grand strategy to a far more assertive posture—most notably, in the South China Sea.


Although the United States is growing old, the scope and pace of its aging are much less severe than they are in the other current major powers, particularly compared to its main global competitors: Russia and especially China.92 The United States' most significant demographic advantage is that its working-age population will continue to expand in coming decades, while those of Russia and China are forecasted to precipitously decline. From 2010 to 2050, the number of those aged 15 to 64 in the United States is expected to grow by 13 percent, while Russia's working-age population is predicted to decline by 23 percent and China's by 18 percent.

Arguably, the most important effect of the United States' comparatively youthful demographic profile is that it will provide a major boost for the continuation of U.S. dominance in global power projection capacity.93 Whereas the United States' growing workforce should significantly aid economic development, the rapid contraction of the workforces in Russia and China over the next few decades will be a powerful brake on economic growth in both countries.

Not only is the United States' workforce expanding, but so too is its prime military-age cohort of 18- to 23-year-olds—a group forecasted to expand by 9 percent from 2010 to 2050. In China and Russia, in contrast, this cohort is predicted to decline by 50 and 39 percent, respectively. These diverging demographic trends are likely to give the United States a comparative advantage in keeping down military personnel costs, allowing it to devote a relatively high percentage of its military budget to weapons development and procurement. Indeed, the United States' personnel costs as a percentage of its military budget are already significantly lower than those in countries with older demographic profiles. For example, in 2016 the United States was one of three NATO members that spent below 36 percent of its military budget on personnel expenditures (Britain and Norway were the other two). This percentage averaged 58 for NATO's other twenty-five members.94

Although the United States possesses important demographic advantages compared to other great powers, in the coming decades it is still likely to experience some costs created by its aging population, including slowing economic growth, crowding-out dynamics, and increasing military personnel costs. These costs could constrain U.S. military capabilities in the future. Still, because the costs associated with aging are likely to be significantly higher for its main global competitors, demography will be a powerful force for the continuation of U.S. military superiority.


Based on numerous indices, Russia has a severe aging problem. As expected, the graying of Russia's population has constrained the country's ability to generate military power. During the 2000s, Moscow spent, for example, considerably more on military retirees for much of the decade than it did on either weapons procurement or military research and development.95 Moreover, Russia's military-age cohort in the 2000s was so small that the country did not have enough conscripts to meet its declared needs.96 As a result, Russia has had to rely increasingly on contract soldiers, which will significantly increase costs as the government is forced to compete with the civilian labor market (a contract soldier's starting pay is ten times higher than that of a conscript).97 One analysis estimates that this competition could increase the military's total wage fund by 40 percent.98 In the 2000s, Russia spent nearly three times as much money on military personnel than it did on weapons, which was the third-highest ratio among the great powers (only Germany and France possessed greater ratios).99

Despite its severe aging problem, Russia has undertaken a number of assertive foreign policy actions since 2010, which seems to run counter to our analysis. Beginning in that year, Moscow committed to a decadelong, large-scale increase in military spending that has significantly improved the country's arsenal. And recently, its leaders have displayed a growing willingness to use force, albeit in situations with a low risk of suffering significant casualties. Examples include seizing Crimea from Ukraine in 2014, sending soldiers into eastern Ukraine beginning in the same year to support rebel groups that were fighting against the government in Kiev, and launching an air campaign in Syria in 2015 against rebel groups.

There are three potential ways to understand Russia's recent assertiveness despite its aging population. The first is to take a closer look at the demographic data for the country, which indicate that Russia is in a unique position compared to the other major powers. Russia's median age (39.6) is comparable to those of many other economically developed countries, such as the United States (38.1) and the United Kingdom (40.5). Its median age is held down, however, by its unusually low life expectancy (70.5), which is an outlier among economically developed countries (virtually all of which have life expectancies in the low 80s). Consequently, the crowding-out effect has been relatively less acute for Russia than it has been for all other developed countries: the burden of expenditures for seniors is lower in Russia because they live, on average, a decade or more less than their counterparts in all of the other major powers.

A second possible way to understand Russia's recent military assertiveness is to recognize the probabilistic nature of our analysis: population aging does not provide an absolute barrier to conflict; it simply reduces its likelihood. Political leaders can try to disregard the incentives and pressures created by aging. Russia's government, for example, is currently cutting social welfare spending on the elderly in order to increase military expenditures.100 Only time will tell if this policy is sustainable, or if it will create a backlash that forces the government to spend more on “canes” and less on “guns.”101

The third potential way to understand Russia's recent military assertiveness is based on closing window-of-opportunity dynamics.102 The scale of population aging in Russia has not yet reached its most extreme levels, but analysts predict that it will soon. Nicholas Eberstadt summarizes the scope of Russia's aging problem as follows: “There is a profound and fundamental difference between the depopulation underway in Russia today and the depopulation facing… affluent Western nations. Germany, Japan, and Italy commonly confront the prospect of population decline in the context of robust and steadily improving levels of public health. The Russian Federation, by contrast, has been seized by an extended mortality crisis—an affliction of historic and truly tragic dimensions.” The result, Eberstadt continues, is that “Russia today is in the grip of an eerie, far-reaching and in some respects historically unprecedented population crisis.”103

Russia's leaders seem to recognize that their country's aging problem is likely to soon become even more severe. In his first state-of-the-union address in 2000, for example, President Vladimir Putin warned that if current demographic trends continued, Russia faced “the threat of becoming a senile nation.”104 In 2006, he declared that demography was “Russia's most acute problem” given the severity of the challenges associated with population aging.105 In an article published during his 2012 presidential campaign, Putin indicated that Russia's demographic decline, had critical geopolitical consequences: “In a global sense we are facing the risk of turning into an ‘empty space’ whose fate will not be decided by us.” Putin vowed to take measures to reverse Russia's population decline, and his government has supported pro-natalist policies throughout his presidency and premiership.106 If Russia's leaders understand the severity of their country's aging problem and the constraints it is likely to create, they may seek to achieve at least some revisionist international objectives before these constraints become even more powerful.


The United States' key European and Asian allies are rapidly growing old. (Japan has the highest median age in the world, followed by Germany and Italy.107) The older these allies become, the harder it will likely be for them to fund their defense budgets and for their leaders to be willing to project force abroad to advance shared interests. Military spending in these countries is already very low, while their sensitivity to casualties in war and aversion to conflict are generally high.108

In recent years, U.S. officials have expressed growing frustration over the unwillingness of European powers to shoulder a higher share of the burdens to advance common goals. NATO countries' defense expenditures—both relative to that of the United States and as a percentage of GDP—are now at historically low levels. For most of the Cold War, the United States accounted for 50 percent of total NATO military spending. That figure was 72 percent in 2017.109 The consistent major disparity in military spending between the United States and its NATO allies led Secretary of Defense Robert Gates in 2011 to refer to NATO as a “two-tiered” alliance.110 This bifurcation, Gates asserted, was causing substantial resentment in the United States. More recently, Donald Trump accused NATO members of “ripping off the United States… Either they have to pay up for past deficiencies or they have to get out. And if it breaks up NATO, it breaks up NATO.”111 As the aging crisis in the United States' European and Asian allies intensifies, this spending gap is likely to increase, as will the level of resentment.

Low levels of defense spending by U.S. allies have already damaged these states' military effectiveness. In the 2011 campaign to topple Muammar Qaddafi's government in Libya, the United States' European allies lacked the weaponry, as well as the reconnaissance, intelligence, heavy airlift, and refueling equipment, necessary to defeat a minor power. As Gates put it, “The mightiest military alliance in history is only eleven weeks into an operation against a poorly armed regime in a sparsely populated country—yet many allies are beginning to run short of munitions, requiring the U.S., once more, to make up the difference.”112 Without major investments in weapons and military equipment, NATO's European powers will be hard-pressed to project force beyond their borders for a sustained period of time against weak states such as Libya, let alone to defend against a great power such as Russia. The severity of the aging problem in the United States' long-standing allies, though, decreases the probability of such investments occurring.

The effects of aging in the United States and its core allies point to a paradox for U.S. international interests: while aging in the United States is likely to increase the appeal of retrenchment and “coming home,” the graying of its allies will tend to increase the need for sustaining its current “deep engagement” grand strategy.113 On the one hand, in a time of growing budget constraints caused by the effects of an aging population, calls for the United States to pull back from the world and deflect the costs of balancing threats to local actors (a tactic known as “buck passing”) are likely to become more compelling to the American people, because they mesh with the need to reduce governmental spending.114 On the other hand, the severity of the aging problem now occurring among U.S. allies makes pulling back less viable, because it reduces the likelihood that these states will be effective “buck catchers,” meaning that they are less likely to possess the capacity and will to balance emerging threats. These conflicting tendencies will add significantly to the challenges of effective U.S. foreign policy making in coming decades.


Age structures are rapidly shifting across the world in unprecedented ways. This article shows that these ongoing demographic transitions matter significantly for international conflict. We have provided a systematic theoretical and empirical examination of two core questions: (1) are young societies especially prone to international conflict? and (2) are states with the oldest populations the most peaceful? The answer to both questions is yes. Four key takeaways follow from this analysis.

First, scholars who study security and conflict now need to account for demographic change in their analyses. In particular, they should add median age to the list of control variables that are typically used in empirical analyses of international conflict in a way that properly accounts for its relationship to conflict. They should also reexamine core existing findings from the conflict literature to confirm that these findings still hold once median age is included.

Second, demographic transitions are likely to have a progressively stronger stabilizing effect over time. Policymakers and analysts frequently decry the effects of societal aging at the domestic level, given the many difficult policy choices that accompany it. Societal aging has a largely unrecognized international benefit, however. According to our analysis, the likelihood of interstate conflict drops sharply (to around one-quarter to one-third below the peak values) once countries reach the following thresholds: (1) a median age above 30 years, (2) a youth-bulge ratio below 20 percent of the adult population, (3) a fertility level below two births per woman, and (4) a life expectancy above 75 years. Although we do not know the precise pace of global aging, we do know that in just a few decades the world's population structure will be far older than it is today.115 Over the long term, we can expect that societal aging will serve as a force for enhanced stability in most, and perhaps all, regions of the world.

Third, the road to a “demographic peace” is likely to be bumpy in many parts of the world in the short and medium term. At present, a majority of the world's people live in poor countries that are very young; thus, demographic change is currently a force for international instability in much of the developing world. In addition, the two main potential security challengers to the current international order—Russia and China—are about to become extremely old, extremely quickly. The world has never experienced a situation in which countries with very powerful militaries undergo a rapid, dramatic aging process, and there are reasons to worry that security dynamics could get worse before they get better regarding these two countries. The scope of Russia's demographic aging will cause it to fade more quickly and more significantly than many former great powers. Additionally, all other states that rose toward the top of the international hierarchy did so when they were young; China is not simply the first rising power to lack a youthful demographic profile, but it will be one of the very oldest societies on Earth in just a few decades. Although ongoing demographic changes in China and Russia should over the long term reduce their capacity and willingness to pose security challenges, in the short and medium terms, one or both countries may feel that they have a rapidly closing demographic window of opportunity and will seek to lash out before they become truly old.

The fourth takeaway flows directly from the third and concerns U.S. grand strategy. The scholarly debate about the merits of the United States continuing a “deep engagement” grand strategy has now spilled over into the public realm; for the first time in seven decades, there is a serious prospect that the United States may significantly pull back from the world. As one of us has argued elsewhere, there is a strong case that the United States should maintain its strategy of deep engagement.116 Factoring in the influence of demographic change makes that case even more compelling in the short and medium terms, though not necessarily forever. It is true that China and Russia may feel the need to lash out militarily before they become enfeebled by societal aging. It is also true that much of the developing world, including the Middle East, is likely to be more conflict prone in the next few decades given the currently youthful profiles of countries in this region. But at the same time, societal aging is likely to act as a long-term force for enhanced stability in Asia, Europe, and the Middle East, which could make an active U.S. effort to foster stability in these three regions less necessary. Moreover, even though the United States' relatively youthful profile compared to those of the other major powers will make it easier to maintain its dramatic lead in global power projection capacity in the next few decades, the country is aging in absolute terms; as a result, it will have increasing difficulty maintaining the necessary capacity over the long term to perform a stabilizing role in the security realm. A few decades from now, ongoing demographic transitions may have reduced both the United States' capacity and need to maintain a globally engaged grand strategy, perhaps significantly. Someday, therefore, demographic change may point toward the United States pulling back from the world, but not yet.


The authors thank the anonymous reviewers, as well as participants in seminars sponsored by the U.S. Military's Joint Staff J7 and the Applied Physics Laboratory at Johns Hopkins University; the Russian Presidential Academy of National Economy and Public Administration (Moscow); the Sasakawa Peace Foundation and Woodrow Wilson International Center for Scholars (Tokyo); and the Air, Army, and Naval War Colleges. The authors are also grateful to Jeffrey Becker, Duncan Brown, Richard Cincotta, John Doyle, Mark Duckenfield, Nicholas Eberstadt, Jack Goldstone, Richard Jackson, Robert Keohane, Joshua Kertzer, Jonathan Markowitz, Keisuke Nakashima, Jon Pevehouse, Stuart Reid, Bruce Russett, Benjamin Valentino, and Toshi Yoshihara for their help and advice at various stages of this project, and to Pirzada Ahmad, Zuo Ming Koh, Gardiner Kreglow, Priya Krishna, and Keshav Poddar for outstanding research assistance.



Somini Sengupta, “The World Has a Problem: Too Many Young People,” New York Times, March 6, 2016, http://www.nytimes.com/2016/03/06/sunday-review/the-world-has-a-problem-too-many-young-people.html.


Studies of the link between demography and the probability of conflict include Mark L. Haas, “A Geriatric Peace? The Future of U.S. Power in a World of Aging Populations,” International Security, Vol. 32, No. 1 (Summer 2007), pp. 112–147, doi.org/10.1162/isec.2007.32.1.112; Richard Jackson and Neil Howe, with Rebecca Strauss and Keisuke Nakashima, The Graying of the Great Powers: Demography and Geopolitics in the 21st Century (Washington, D.C.: Center for Strategic and International Studies, 2008); Valerie M. Hudson and Andrea M. den Boer, Bare Branches: The Security Implications of Asia's Surplus Male Population (Cambridge, Mass.: MIT Press, 2004); Valerie M. Hudson and Andrea den Boer, “A Surplus of Men, A Deficit of Peace: Security and Sex Ratios in Asia's Largest States,” International Security, Vol. 26, No. 4 (Spring 2002), pp. 5–38, doi.org/10.1162/016228802753696753; Monica Duffy Toft, “Death by Demography: 1979 as a Turning Point in the Disintegration of the Soviet Union,” International Area Studies Review, Vol. 17, No. 2 (June 2014), pp. 184–204, doi.org/10.1177/2233865914535597; Ronald R. Krebs and Jack S. Levy, “Demographic Change and the Sources of International Conflict,” in Myron Weiner and Sharon Stanton Russell, eds., Demography and National Security (New York: Berghahn, 2001), pp. 62–105; Susan Yoshihara and Douglas A. Sylva, eds., Population Decline and the Remaking of Great Power Politics (Washington, D.C.: Potomac, 2012); Jack A. Goldstone, “Flash Points and Tipping Points: Security Implications of Global Population Changes, 2005–2025,” paper prepared for the Mackinder Forum, Minster Lovell, Oxfordshire, United Kingdom, March 14–15, 2006; and Nazli Choucri, Population Dynamics and International Violence: Propositions, Insights, and Evidence (Lexington, Mass.: Lexington Books, 1974).


See, for example, Henrik Urdal, “A Clash of Generations? Youth Bulges and Political Violence,” International Studies Quarterly, Vol. 50, No. 3 (September 2006), pp. 609–610, doi.org/10.1111/j.1468-2478.2006.00416.x; and Richard P. Cincotta, Robert Engelman, and Daniele Anastasion, The Security Demographic: Population and Civil Conflict after the Cold War (Washington, D.C.: Population Action International, 2003), pp. 44–48.


The most extensive current treatment is Haas, “A Geriatric Peace?”


The “opportunity” and “willingness” terminology and framework are from Benjamin A. Most and Harvey Starr, Inquiry, Logic, and International Politics (Columbia: University of South Carolina Press, 1989).


Ronald Lee, “The Demographic Transition: Three Centuries of Fundamental Change,” Journal of Economic Perspectives, Vol. 17, No. 4 (Fall 2003), p. 168, doi.org/10.1257/089533003772034943.


See, for example, John C. Caldwell, “Routes to Low Mortality in Poor Countries,” Population and Development Review, Vol. 12, No. 2 (June 1986), pp. 171–220, doi.org/10.2307/1973108.


David S. Reher, “The Demographic Transition Revisited as a Global Process,” Population, Space, and Place, Vol. 10, No. 1 (January/February 2004), p. 27, doi.org/10.1002/psp. 313.


Lee, “The Demographic Transition,” p. 187. It is possible that states at the end of the demographic transition could experience major increases in fertility levels that would move them back to an earlier stage. This development would represent a significant departure from historical trends, however. States, moreover, have dedicated considerable resources to boosting fertility, but these policies, for the most part, have been ineffective. There are also good demographic, sociological, and economic reasons to expect that governments' pro-natalist polices will continue to be largely unsuccessful. For details on this last point, see Wolfgang Lutz, Vegard Skirbekk, and Maria Rita Testa, “The Low Fertility Trap Hypothesis: Forces that May Lead to Further Postponement and Fewer Births in Europe,“ Vienna Yearbook of Population Research, Vol. 4 (2006), pp. 167–192, https://www.jstor.org/stable/23025482.


Reher, “The Demographic Transition Revisited as a Global Process,” p. 34.


Paul Collier, “Doing Well Out of War: An Economic Perspective,” in Mats Berdal and David M. Malone, eds., Greed and Grievance: Economic Agendas in Civil Wars (Boulder, Colo.: Lynne Rienner, 2000), p. 94; and Urdal, “A Clash of Generations?” p. 609.


Countries with expanding prime military-age cohorts in the second half of the twentieth century, in contrast, frequently experienced sharp declines in the unit costs of military personnel. Per soldier costs for the North Atlantic Treaty Organization powers were calculated from NATO, “Financial and Economic Data Relating to NATO Defence: Defence Expenditures of NATO Countries (1975–1999)” (Brussels: NATO, December 2, 1999), http://www.nato.int/cps/en/natolive/news_26178.htm?mode=pressrelease. Per soldier costs for all other countries were calculated from a combination of Erik Gartzke's dataset, which contains labor and total military costs for ninety-nine countries ranging from 1950 to 1997. Gartzke, “Disaggregated Military Expenditure Data” (San Diego, Calif.: website of Erik Gartzke, last modified February 18, 2010), http://dss.ucsd.edu/~egartzke/htmlpages/data.html; and various years of World Military Expenditures and Arms Transfers reports, which contain military personnel costs for all countries from 1965 to 1997, U.S. Department of State, “World Military Expenditures and Arms Transfers, 1964–1997” (Washington, D.C.: U.S. Department of State, n.d.), http://www.state.gov/t/avc/rls/rpt/wmeat/c50834.htm.


NATO, “Financial and Economic Data Relating to NATO Defence.” In contrast to the dominant trend for NATO countries, military personnel costs as a percentage of states' budgets plummeted in a majority of states in sub-Saharan Africa, South Asia, and the Middle East in the last decades of the twentieth century. See Gartzke, “Disaggregated Military Expenditure Data.”


David E. Bloom, David Canning, and Jaypee Sevilla, The Demographic Dividend: A New Perspective on the Economic Consequences of Population Change (Santa Monica, Calif.: RAND Corporation, 2003).


Martin C. Libicki, Howard J. Shatz, and Julie E. Taylor, Global Demographic Change and Its Implications for Military Power (Santa Monica, Calif.: RAND Corporation, 2011).


David E. Bloom and Jeffrey G. Williamson, “Demographic Transitions and Economic Miracles in Emerging Asia,” World Bank Economic Review, Vol. 12, No. 3 (September 1998), pp. 419–455, doi.org/10.1093/wber/12.3.419.


Andrew Mason, “Demographic Transition and Demographic Dividends in Developed and Developing Countries,” paper presented at the United Nations Expert Group Meeting on Social and Economic Implications of Changing Population Age Structures, Mexico City, Mexico, August 31-September 2, 2005, http://www.un.org/esa/population/meetings/Proceedings_EGM_Mex_2005/mason.pdf. David S. Reher estimates that for many developing countries the window for the demographic dividend will be even shorter, lasting between ten and thirty years. Reher, “Economic and Social Implications of the Demographic Transition,” Population and Development Review, Vol. 37, No. S1 (January 2011), p. 27, doi.org/10.1111/j.1728-4457.2011.00376.x.


Dave Turner et al., “The Macroeconomic Implications of Aging in a Global Context,” OECD Economics Department Working Paper No. 193 (Paris: Organisation for Economic Co-operation and Development [OECD], 1998), doi.org/10.1787/18151973. A 2016 study of U.S. states from 1980 to 2010 found that a 10 percent increase in the percentage of 60-year-olds and older reduced the growth rate of gross domestic product (GDP) per capita by 5.5 percent. See Nicole Maestas, Kathleen J. Mullen, and David Powell, “The Effect of Population Aging on Economic Growth, the Labor Force, and Productivity,” NBER Working Paper No. 22452 (Cambridge, Mass.: National Bureau of Economic Research [NBER], July 2016), http://www.nber.org/papers/w22452.pdf.


On these points, see Haas “A Geriatric Peace?” p. 119.


Tongfi Kim and Jennifer Dabbs Sciubba, “The Effect of Age Structure on the Abrogation of Military Alliances,“ International Interactions, Vol. 41, No. 2 (March 2015), p. 281, doi.org/10.1080/03050629.2014.948156


Sanders Korenman and David Neumark, “Cohort Crowding and Youth Labor Markets: A Cross-National Analysis,” NBER Working Paper No. 6031 (Cambridge, Mass.: NBER, 1997), doi.org/10.3386/w6031; and Onn Winckler, “The Demographic Dilemma of the Arab World: The Employment Aspect,” Journal of Contemporary History, Vol. 37, No. 4 (October 2002), pp. 617–636, doi.org/10.1177/00220094020370040701.


Saul Levine, “Youth in Terroristic Groups, Gangs, and Cults: The Allure, the Animus, and the Alienation,” Psychiatric Annals, Vol. 29, No. 6 (June 1999), p. 343, at p. 347, doi.org/10.3928/0048-5713-19990601-08. See also Hannes Weber, “Demography and Democracy: The Impact of Youth Cohort Size on Democratic Stability in the World,” Democratization, Vol. 20, No. 2 (2013), pp. 348–351, doi.org/10.1080/13510347.2011.650916.


See Jessica Stern, “Radicalization to Extremism and Mobilization to Violence: What Have We Learned and What Can We Do about It?” Annals of the American Academy of Political and Social Science, Vol. 668, No. 1 (November 2016), pp. 102–117, doi.org/10.1177/0002716216673807.


Jack A. Goldstone, Revolution and Rebellion in the Early Modern World: Population Change and State Breakdown in England, France, Turkey, and China, 1600–1850, 25th anniversary ed. (London: Routledge, 2016); and Herbert Moller, “Youth as a Force in the Modern World,” Comparative Studies in Society and History, Vol. 10, No. 3 (April 1968), pp. 237–260, doi.org/10.1017/S0010417500004898.


Kim and Sciubba, “The Effect of Age Structure on the Abrogation of Military Alliances,” pp. 280–281.


Stephen M. Walt, Revolution and War (Ithaca, N.Y.: Cornell University Press, 1996), pp. 18–45. On how ideological beliefs are often a key source of international conflict, see Mark L. Haas, The Ideological Origins of Great Power Politics, 1789–1989 (Ithaca, N.Y.: Cornell University Press, 2005).


Amy Oakes, Diversionary War: Domestic Unrest and International Conflict (Stanford, Calif.: Stanford University Press, 2012).


Cincotta, Engelman, and Anastasion, The Security Demographic, p. 44.


Directorate of Intelligence, “The Youth Bulge: A Link between Demography and Instability: A Research Paper“ (Washington, D.C.: Central Intelligence Agency [CIA], March 1986), https://www.cia.gov/library/readingroom/docs/CIA-RDP97R00694R000500680001-1.pdf.


Stephen E. Gent, “Going In When It Counts: Military Intervention and the Outcome of Civil Conflicts,” International Studies Quarterly, Vol. 52, No. 4 (December 2008), pp. 713–735, doi.org/10.1111/j.1468-2478.2008.00523.x; and Michael G. Findley and Tze Kwang Teo, “Rethinking Third-Party Interventions into Civil Wars: An Actor-Centric Approach,” Journal of Politics, Vol. 68, No. 4 (November 2006), pp. 828–837, doi.org/10.1111/j.1468-2508.2006.00473.x.


Therése Pettersson and Peter Wallensteen, “Armed Conflicts, 1946–2014,” Journal of Peace Research, Vol. 52, No. 4 (July 2015), pp. 536–537, 539, doi.org/10.1177/0022343315595927.


Calculated from Nils Petter Gleditsch et al., “UCDP/PRIO Armed Conflict Dataset,” ver. 17.1 (Uppsala, Sweden: Department of Peace and Conflict Research, Uppsala University, 2017), http://ucdp.uu.se/downloads/; described in Therése Pettersson and Kristine Eck, “Organized Violence, 1989–2017,” Journal of Peace Research, Vol. 55, No, 4 (July 2018), pp. 535–547, doi.org/10.1177/0022343318784101; and in Nils Petter Gleditsch et al., “Armed Conflict 1946–2001: A New Dataset,” Journal of Peace Research, Vol. 39 No. 5 (September 2002), pp. 615–637, www.jstor.org/stable/1555346.


A civil war is transformed into an interstate war if an intervener takes over the bulk of the fighting for the rebel groups. A civil war is transformed into an extra-state war if an intervener takes over the bulk of the fighting for the government. See Meredith Reid Sarkees, “The COW Typology of War: Defining and Categorizing Wars” (University Park: Correlates of War Project, Pennsylvania State University, n.d.), http://www.correlatesofwar.org/data-sets/COW-war/the-cow-typology-of-war-deining-and-categorizing-wars/view; and Meredith Reid Sarkees, “ExtraState Wars (Version 4.0): Definitions and Variables” (University Park: Correlates of War Project, Pennsylvania State University, n.d.), http://cow.dss.ucdavis.edu/data-sets/COW-war/extra-state-wars-codebook.


See the reports on social security programs in Europe, Asia and the Pacific, Africa, and the Americas at Social Security Administration, “Social Security Programs throughout the World” (Washington, D.C.: Social Security Administration and International Social Security Association, n.d.), http://www.ssa.gov/policy/docs/progdesc/ssptw/.


For details, see CIA, “Long-Term Global Demographic Trends: Reshaping the Geopolitical Landscape” (Washington, D.C.: CIA, 2001), p. 27, https://www.cia.gov/library/reports/general-reports-1/Demo_Trends_For_Web.pdf; and Richard Jackson, “The Global Retirement Crisis: The Threat to World Stability and What to Do About It” (Washington, D.C.: Center for Strategic and International Studies, 2002), p. 22.


The demand for greater governmental expenditures for the elderly as societies age may very well result in inter-cohort tensions, with older generations calling for increased health and welfare spending for seniors and younger generations resisting it. Who wins this intergenerational conflict will go a long way in determining the intensity of the crowding-out effect that we discuss in this section. Because the elderly are more reliable voters than the young, however, their political preferences are more likely to be satisfied.


We follow the United Nations' (UN) classification of the world's regions. See UN, “World Population Prospects: The 2017 Revision” (New York: UN Department of Economic and Social Affairs/Population Division, 2017), https://population.un.org/wpp/General/Files/Definition_of_Regions.pdf. All demographic data in this article were obtained from the 2010 to 2017 editions of UN, “World Population Prospects” (New York: UN), https://esa.un.org/unpd/wpp/; or from World Bank, “World Development Indicators” (Washington, D.C.: World Bank, 2012), https://datacatalog.worldbank.org/dataset/world-development-indicators.


Shiri Noy, “New Contexts, Different Patterns? A Comparative Analysis of Social Spending and Government Health Expenditure in Latin America and the OECD,” International Journal of Comparative Sociology, Vol. 52, No. 3 (June 2011), p. 227, doi.org/10.1177/0020715211408760.


Ito Peng and Joseph Wong, “Growing Out of the Developmental State: East Asian Welfare Reform in the 1990s,“ paper presented at the Annual RC19 Conference, Paris, France, September 2–4, 2004, p. 5, https://core.ac.uk/download/pdf/51177570.pdf.


Similar trends existed for leading OECD states from 1960 to 1980, which controls for effects of the end of the Cold War on defense spending. See Vito Tanzi and Ludger Schuknedcht, Public Spending in the 20th Century: A Global Perspective (Cambridge: Cambridge University Press, 2000), pp. 25, 28, 38, 41.


Military expenditures between 1984 and 1994 fell by 0.9 percent of GDP in Latin America and 0.7 percent in East Asia, with equal reductions before and after the end of the Cold War. See U.S. Department of State, “World Military Expenditures and Arms Transfers, 1964–1997” (various years, especially 1990, 1995, and 1996).


Douglas A. Sylva, “Europe's Strategic Future and the Need for Large-Family Pronatalism: A Normative Study of Demographic Decline,” in Yoshihara and Sylva, Population Decline and the Remaking of Great Power Politics, p. 99.


Erik Gartzke, “Democracy and the Preparation for War: Does Regime Type Affect States' Anticipation of Casualties?” International Studies Quarterly, Vol. 45, No. 3 (September 2001), pp. 467–484, doi.org/10.1111/0020-8833.00210.


Libicki, Shatz, and Taylor, Global Demographic Change and Its Implications for Military Power, pp. 101, 105, 109.


For details, see Fred Arnold and Sonia Albores, The Value of Children: A Cross-National Study (Honolulu: East-West Population Institute, 1975); and Bernhard Nauck and Daniela Klaus, “The Varying Value of Children: Empirical Results from Eleven Societies in Asia, Africa, and Europe,” Current Sociology, Vol. 55, No. 4 (July 2007), pp. 487–503, doi.org/10.1177/0011392107077634.


Rodolfo A. Bulatao, “Values and Disvalues of Children in Successive Childbearing Decisions,” Demography, Vol. 18, No. 1 (February 1981), p. 1, doi.org/10.2307/2061046.


Ibid., p. 11; and James T. Fawcett “Perceptions of the Value of Children: Satisfaction and Costs,” in Rodolfo A. Bulatao and Ronald D. Lee, eds., Determinants of Fertility in Developing Countries, Vol. 1 (New York: Academic Press, 1983), pp. 443–445.


Bernhard Nauck, “Value of Children and the Framing of Fertility: Results from a Cross-Cultural Comparative Survey in 10 Societies,” European Sociological Review, Vol. 23, No. 5 (December 2007), pp. 624–625, doi.org/10.1093/esr/jcm028.


Gang Zheng, Shaohua Shi, and Hong Tang, “Population Development and the Value of Children in the People's Republic of China,” in Gisela Trommsdorff and Bernhard Nauck, eds., The Value of Children in Cross-Cultural Perspective: Case Studies from Eight Societies (Berlin: Pabst Science, 2005), p. 257.


Bulatao, “Values and Disvalues of Children in Successive Childbearing Decisions,” p. 2.


For example, if carrying on the family name is a particularly important benefit associated with child-rearing—and survey data, especially from Asian countries, indicate that this frequently is the case—the loss of children in low-fertility households will be acute. See ibid., pp. 5, 7.


Thus, the argument linking decreasing fertility to increased casualty aversion is not that parents love children born later in the family-building process less than children born earlier in this process, but that the benefits of children born early in the birth order are particularly great. Critics of arguments that support fertility-based sources of casualty aversion, such as Ronald Krebs and Jack Levy, have attacked the idea that parents' love for their children is finite (which, if true, would result in lower levels of love for each child the bigger the family) without addressing the empirically based claim that the benefits of children vary substantially by birth order. See Krebs and Levy, “Demographic Change and the Sources of International Conflict,” p. 76.


Analyzing the level of parents' investments in terms of time and money (especially in terms of educational costs) in individual children leads to the same conclusion. Studies have demonstrated that parents' human capital expenditures per child are substantially higher when fertility is lower. As these investments in what have been labeled “high quality” children expand, we would expect casualty sensitivity in lower-fertility societies to also increase. See Ronald Lee and Andrew Mason, “Fertility, Human Capital, and Economic Growth over the Demographic Transition,” European Journal of Population, Vol. 26, No. 2 (2010), pp. 159–182, doi.org/10.1007/s10680-009-9186-x; and Reher, “Economic and Social Implications of the Demographic Transition,” p. 18.


See Gerhard Kummel and Nina Leonard, “Casualty Shyness and Democracy in Germany,” Security and Peace, Vol. 22, No. 3 (2004), pp. 119–126, www.jstor.org/stable/24231454. See also Natalie La Balme, “The French and the Use of Force: Public Perceptions and Their Impact on the Policy Making Process”; Jan van der Meulen and Marijke de Konink, “Risky Missions: Dutch Public Opinion on Peacekeeping in the Balkans”; Pierangelo Isernia, “Italian Public Opinion and the Use of Force in the Early 1990s”; and Philip Everts, “War Without Bloodshed? Public Opinion and the Conflict Over Kosovo,” all in Everts and Isernia, Public Opinion and the International Use of Force (London: Routledge, 2001). In addition, see Yagil Levy, “How Casualty Sensitivity Affects Civilian Control: The Israeli Experience,” International Studies Perspectives, Vol. 12, No. 1 (February 2011), p. 68, doi.org/10.1111/j.1528-3585.2010.00420.x; and Hugh Smith, “What Costs Will Democracies Bear? A Review of Popular Theories of Casualty Aversion,” Armed Forces & Society, Vol. 31, No. 4 (Summer 2005), p. 507, doi.org/10.1177/0095327X0503100403. For studies that claim an exception to this trend in the U.S. case, see Eric V. Larson, Casualties and Consensus: The Historical Role of Casualties in Domestic Support for U.S. Military Operations (Santa Monica, Calif.: RAND Corporation, 1996), p. 10; and Christopher Gelpi, Peter D. Feaver, and Jason Reifler, Paying the Human Cost of War: American Public Opinion and Casualties in Military Conflicts (Princeton, N.J.: Princeton University Press, 2009), pp. 23–66.


Yagil Levy, “An Unbearable Price: War Casualties and Warring Democracies,” International Journal of Politics, Culture, and Sociology, Vol. 22, No. 1 (March 2009), pp. 70–72, doi.org/10.1007/s10767-009-9048-x.


On the preceding points, see Levy, “How Casualty Sensitivity Affects Civilian Control,” pp. 72–82, at pp. 78–79.


John A. Gentry, “Casualty Management: Shaping Civil-Military Operational Environments,” Comparative Strategy, Vol. 30, No. 3 (July 2011), p. 244, doi.org/10.1080/01495933.2011.561737.


Benjamin A. Valentino, Paul K. Huth, and Sarah E. Croco, “Bear Any Burden? How Democracies Minimize the Costs of War,” Journal of Politics, Vol. 72, No. 2 (April 2010), pp. 528–544, doi.org/10.1017/S0022381609990831.


UN, “World Population Prospects: 2015 Revision” (Washington, D.C.: UN, July 2015), http://www.un.org/en/development/desa/publications/world-population-prospects-2015-revision.html


According to some reports, these calculations played an important role in pushing thousands of Japanese pensioners in 2015 to participate in mass protests against Prime Minister Shinzō Abe's efforts to weaken Japan's pacifist constitution. See Yoichi Funabashi “Japan's Gray-Haired Pacifism,” New York Times, August 12, 2015, http://www.nytimes.com/2015/08/13/opinion/japans-gray-haired-pacifism.html.


See the online appendix at doi.org/10.7910/DVN/EIASGF.


See the discussion in Andrew Gelman, “Graphs Showing Regression Uncertainty: The Code!” Statistical Modeling, Causal Inference, and Social Science blog, August 26, 2012, http://andrewgelman.com/2012/08/26/graphs-showing-regression-uncertainty-the-code/.


The figures present the results of a simulation exercise in which 1,000 independent estimates of the model coefficients were taken from the multivariate normal distribution described by the coefficients and variance-covariance matrix estimated from the model.


Nicole Speulda, “Youth and War: From Vietnam to Iraq, Generations Disagree about the Use of Military Force” (Washington, D.C.: Pew Research Center, February 21, 2006), http://www.pewresearch.org/2006/02/21/youth-and-war/.


Val Burris, “From Vietnam to Iraq: Continuity and Change in Between-Group Differences in Support for Military Action,” Social Problems, Vol. 55, No. 4 (November 2008), pp. 443–479, doi.org/10.1525/sp.2008.55.4.443. Burris finds that individuals from ages 30 to 64 tend to be slightly more bellicose than individuals from ages 18 to 29 in most cases; regardless, individuals ages 65 and older are less inclined to support war than both younger age groups across most of the interventions.


Gelpi, Feaver, and Reifler, Paying the Human Cost of War, pp. 35–36, 84–88, at p. 88. See also Peter D. Feaver and Christopher Gelpi, Choosing Your Battles: American Civil-Military Relations and the Use of Force (Princeton, N.J.: Princeton University Press, 2004), p. 174, also pp. 125–128, 164, 168, 170.


See Andrew Dugan, “On 10th Anniversary, 53% in U.S. See Iraq War as Mistake,” (Washington, D.C.: Gallup, March 18, 2013), http://www.gallup.com/poll/161399/10th-anniversary-iraq-war-mistake.aspx.


Ben Clements, “Examining Public Attitudes towards Recent Foreign Policy Issues: Britain's Involvement in the Iraq and Afghanistan Conflicts,” Politics, Vol. 31, No. 2 (June 2011), pp. 63–71, doi.org/10.1111/j.1467-9256.2011.01404.x; and Rachael Gribble et al., “British Public Opinion after a Decade of War: Attitudes to Iraq and Afghanistan,“ Politics, Vol. 35, No. 2 (June 2015), pp. 128–150, doi.org/10.1111/1467-9256.12073.


Harald Schoen, “Personality Traits and Foreign Policy Attitudes in German Public Opinion,” Journal of Conflict Resolution, Vol. 51, No. 3 (June 2007), pp. 408–430, doi.org/10.1177/0022002707300180.


Pascal Vennesson, “Civil-Military Relations in France: Is There a Gap?” Journal of Strategic Studies, Vol. 26, No. 2 (June 2003), pp. 29–42, doi.org/10.1080/01402390412331302965.


Deborah Jordan Brooks and Benjamin A. Valentino, “A War of One's Own: Understanding the Gender Gap in Support for War,” Public Opinion Quarterly, Vol. 75, No. 2 (January 2011), pp. 270–286, doi.org/10.1093/poq/nfr005.


Details about the experiment, including examples of their treatments, questionnaire wording, and YouGov sampling methodology, are discussed in ibid.


See the online appendix.


Control variables included gender, race, ethnicity, marriage, parenthood, parent of a military-age son, education level, partisanship, political interest, geographic region, church attendance level, and income.


On states' ability to disregard and even counter the pressures created by aging, see Jennifer Dabbs Sciubba, “A New Framework for Aging and Security: Lessons from Power Transition Theory,” in Jack A. Goldstone, Eric P. Kaufman, and Monica Duffy Toft, eds., Political Demography: How Population Changes Are Reshaping International Security and National Politics (New York: Oxford University Press, 2012), pp. 63–77.


UN, “World Population Prospects, 2017.”


See Howard W. French, “As China Ages, Shortage of Cheap Labor Looms,” New York Times, June 30, 2006, https://www.nytimes.com/2006/06/30/world/asia/30aging.html.


Steven V. Dunaway and Annalisa Fedelino, “Fiscal Policy in China,” in Jahangir Aziz, Steven V. Dunaway, and Eswar S. Prasad, eds., China and India Learning from Each Other: Reforms and Policies for Sustained Growth (Washington, D.C.: International Monetary Fund, 2006), p. 234.


Stockholm International Peace Research Institute (SIPRI), “Military Expenditure by Country as Share of GDP, 1988–2017” (Stockholm: SIPRI, 2018), https://www.sipri.org/databases/milex.


Richard Jackson, Neil Howe, and Tobias Peter, The Global Aging Preparedness Index, 2nd ed. (Washington, D.C.: Center for Strategic and International Studies, 2013), p. 16.


For much of the period, China claims that growth in personnel costs was the largest in its military's main spending categories, including in comparison to spending on equipment and operations. See Dennis J. Blasko et al., “Defense-Related Spending in China: A Preliminary Analysis and Comparison with American Equivalents” (Washington, D.C.: China Policy Foundation, 2007), pp. 18–19, http://www.uscpf.org/v2/pdf/defensereport.pdf; and Haas, “A Geriatric Peace?” p. 142.


See the data compiled in GlobalSecurity.org, “PLA Reductions” (Washington, D.C., GlobalSecurity.org, n.d.), http://www.globalsecurity.org/military/world/china/pla-reductions.htm; and Edward Wong, Jane Perlez, and Chris Buckley, “China Announces Cuts of 300,000 Troops at Military Parade Showing Its Might,” New York Times, September 2, 2015, https://www.nytimes.com/2015/09/03/world/asia/beijing-turns-into-ghost-town-as-it-gears-up-for-military-parade.html.


Blasko et al., “Defense-Related Spending in China,” p. 19; and “China's Military Spending: At the Double,” Economist, March 15, 2014, https://www.economist.com/china/2014/03/15/at-the-double.


On this point, see Peter E. Robinson, “Cost Disease in China's Military: The Headline Figures Showing China's Growing Military Spending Are Misleading,” Diplomat, May 13, 2015, https://thediplomat.com/2015/05/chinas-military-and-cost-disease/.


For details on this relationship, see ibid.


“China's Military Spending.”


For details, see Simon Denyer and Congcong Zhang, “China Drops One-Child Policy, but ‘Exhausted’ Tiger Moms Say One Is Plenty,” Washington Post, October 16, 2016, https://www.washingtonpost.com/world/asia_pacific/china-drops-one-child-policy-but-exhausted-tiger-moms-say-one-is-plenty/2016/10/14/336f1890-8ae7-11e6-8a68-b4ce96c78e04_story.html.


Didi Kirsten Tatlow, “With Fertility Rate in China Low, Some Press to Legalize Births outside Marriage,” New York Times, November 17, 2016, https://www.nytimes.com/2016/11/18/world/asia/china-fertility-birth.html.


Nicholas Eberstadt, “Asia-Pacific Demographics in 2010–2040: Implications for Strategic Balance,” in Ashley J. Tellis, Andrew Marble, and Travis Tanner, eds., Strategic Asia, 2010–11: Asia's Rising Power and America's Continued Purpose (Seattle, Wash.: National Bureau of Asian Research, 2010), p. 248. The middle-income trap refers to the tendency for countries' economic growth to slow after reaching middle-income levels, which makes the transition to high-income levels difficult.


See, for example, Graham Allison, Destined for War: Can America and China Escape Thucydides's Trap? (Boston: Houghton Mifflin Harcourt, 2017).


On how aging populations may incentivize aggression based on closing window-of-opportunity dynamics, see Mark L. Haas, “Population Aging and International Conflict,” in William R. Thompson, ed., The Oxford Encyclopedia of Empirical International Relations Theory, Vol. 3 (New York: Oxford University Press, 2018), pp. 93–95.


This is the case notwithstanding the significant decline in the U.S. fertility rate since the 2008 economic recession, from close to the replacement rate of 2.1 before the recession to 1.77 in 2017. If U.S. fertility remains at the latter level, some—but not all—of the United States' demographic advantage will be eroded in the coming decades. Even at such a lower level of fertility, the United States' population would still be expected to increase significantly over the remainder of this century, during which time the populations of China and Russia (as well as Germany and Japan) will markedly decline. This prediction assumes, however, that the United States will continue to receive high numbers of immigrants, which is an assumption that is in greater doubt after Donald Trump's election to the presidency. Based on the level of immigration that existed at the end of the Barack Obama administration, 88 percent of the forecasted growth in the U.S. population by 2065 will come from immigrants and their descendants. See Suzannah Gonzales, “Immigrants Will Drive U.S. Population Growth in Next Five Decades: Pew,” Reuters, September 28, 2015, https://www.reuters.com/article/us-usa-immigration-study/immigrants-will-drive-u-s-population-growth-in-next-five-decades-pew-idUSKCN0RS08V20150928.


See Haas, “A Geriatric Peace?”


See NATO, “Defence Expenditures of NATO Countries (2006–2016)” (Brussels: NATO, 2016), table 6a.


Haas, “A Geriatric Peace?” pp. 136, 142.


Keir Giles, “Where Have All the Soldiers Gone? Russia's Military Plans versus Demographic Reality” (Shrivenham, Oxfordshire: Conflict Studies Research Centre, Defence Academy of the United Kingdom, 2006), p. 1.


Vladimir Isachenkov, “Putin's Russia Seeks to Project Power with Modern Military,” Associated Press, December 6, 2016. See also section 2.4, “Personnel Costs,” in Susanne Oxenstierna, “Russia's Defense Spending and the Economic Decline,” Journal of Eurasian Studies, Vol. 7, No. 1 (January 2016), pp. 60–70, doi.org/10.1016/j.euras.2015.06.001.




Haas, “A Geriatric Peace?” p. 141.


In October 2016, for example, the Russian government announced a $12.7 billion increase in military spending to be paid for from a fund that it had created by delinking pensions from inflation rates, which resulted in a significant cut in welfare spending for the elderly. Michael Khodarkovsky, “Playing with Fear: Russia's War Card,” New York Times, October 26, 2016, https://www.nytimes.com/2016/10/27/opinion/playing-with-fear-russias-war-card.html.


On how this backlash may already be building, see Marc Bennetts, “Will Pensions End Putin? Russia's Retirement Age Debate,” Newsweek, August 20, 2018, https://www.newsweek.com/russia-pension-putin-retirement-age-1081420; and Ann M. Simmons, “Putin Dials Back Pension Plan after Backlash,” Wall Street Journal, August 31, 2018, https://www.wsj.com/articles/russian-proposal-to-raise-retirement-age-poses-problems-for-putin-1535533201.


Haas, “Population Aging and International Conflict,” pp. 93–95; and Jennifer Dabbs Sciubba, “Coffins versus Cradles: Russian Population, Foreign Policy, and Power Transition Theory,” International Area Studies Review, Vol. 17, No. 2 (June 2014), pp. 205–221, doi.org/10.1177/2233865914528823.


Nicholas Eberstadt, Russia's Peacetime Demographic Crisis: Dimensions, Causes, Implications (Seattle, Wash.: National Bureau of Asian Research, 2010), pp. 281–282.


“Putin Warns of ‘Senile Nation,‘” BBC News, July 8, 2000, http://news.bbc.co.uk/2/hi/europe/824748.stm.


Quoted in C.J. Chivers, “Putin Urges Plan to Reverse Slide in the Birth Rate,” New York Times, May 11, 2006, https://www.nytimes.com/2006/05/11/world/europe/11russia.html.


Quoted in “Vladimir Putin Vows to Reverse Russian Population Decline,” Telegraph, February 13, 2012, https://www.telegraph.co.uk/news/worldnews/vladimir-putin/9078672/Vladimir-Putin-vows-to-reverse-Russian-population-decline.html.


UN, “World Population Prospects, 2017.”


On casualty sensitivity in European nations, see many of the chapters in Everts and Isernia, Public Opinion and the International Use of Force; and Kummel and Leonard, “Casualty Shyness and Democracy in Germany.”


Calculated from NATO, “Defence Expenditure of NATO Countries (2010–2017)” (Brussels: NATO, 2017), table 2 (current price and exchange rate), https://www.nato.int/nato_static_fl2014/assets/pdf/pdf_2017_06/20170629_170629-pr2017-111-en.pdf.


Quoted in Thom Shanker, “Defense Secretary Warns NATO of ‘Dim’ Future,” New York Times, June 10, 2011, https://www.nytimes.com/2011/06/11/world/europe/11gates.html.


Quoted in Ashley Parker, “Donald Trump Says NATO Is ‘Obsolete,’ UN Is ‘Political Game,’” New York Times, April 2, 2016, https://www.nytimes.com/politics/first-draft/2016/04/02/donald-trump-tells-crowd-hed-be-fine-if-nato-broke-up/.


Quoted in Shanker, “Defense Secretary Warns NATO of ‘Dim’ Future.”


For an overview of deep engagement, see Stephen G. Brooks, G. John Ikenberry, and William C. Wohlforth, “Don't Come Home, America: The Case against Retrenchment,” International Security, Vol. 37, No. 3 (Winter 2012/13), pp. 7–51, doi.org/10.1162/ISEC_a_00107; and Stephen G. Brooks and William C. Wohlforth, America Abroad: The United States' Global Role in the 21st Century (New York: Oxford University Press, 2016).


See, for example, Barry Posen, Restraint: A New Foundation for U.S. Grand Strategy (Ithaca, N.Y.: Cornell University Press, 2014). President Trump has clearly indicated sympathy for these prescriptions. As he said in April 2016, “We have spent trillions of dollars over time on planes, missiles, ships, equipment, building up our military to provide a strong defense for Europe and Asia. The countries we are defending must pay for the cost of this defense, and if not, the U.S. must be prepared to let these countries defend themselves. We have no choice…In this time of mounting debt…not one single dollar can we waste.” “Transcript: Donald Trump's Foreign Policy Speech,” New York Times, April 27, 2016, https://www.nytimes.com/2016/04/28/us/politics/transcript-trump-foreign-policy.html.


As Jackson et al. explain, “Demographic aging is about as close as social science ever comes to a certain forecast. Every demographer agrees that it is happening and that, absent a global catastrophe—a colliding comet or a deadly super virus—it will continue to gather momentum,” Jackson et al., The Graying of the Great Powers, p.3.


Brooks, Ikenberry, and Wohlforth, “Don't Come Home, America”; and Brooks and Wohlforth, America Abroad.