The Population Implosion

Be careful what you wish for. After decades of struggling to contain the global population explosion that emerged from the healthcare revolution of the 20th century, the world confronts an unfamiliar crisis: rapidly decreasing birthrates and declining life spans that might set back the progress of human development.

It may not be the first way we think of ourselves, but almost all of us alive today happen to be children of the "world population explosion" -- the momentous demographic surge that overtook the planet during the course of the 20th century. Thanks to sweeping mortality declines, human numbers nearly quadrupled in just 100 years, leaping from about 1.6 or 1.7 billion in 1900 to about 6 billion in 2000.

This unprecedented demographic expansion came to be regarded as a "population problem," and in our modern era problems demand solutions. By century's end, a worldwide administrative apparatus -- comprised of Western foundations and aid agencies, multilateral institutions, and Third World "population" ministries -- had been erected for the express purpose of "stabilizing" world population and was vigorously pursuing an international antinatal policy, focusing on low-income areas where fertility levels remained relatively high.

To some of us, the wisdom of this crusade to depress birthrates around the world (and especially among the world's poorest) has always been elusive. But entirely apart from its arguable merit, the continuing preoccupation with high fertility and rapid population growth has left the international population policy community poorly prepared to comprehend (much less respond to) the demographic trends emerging around the world today -- trends that are likely to transform the global population profile significantly over the coming generation. Simply put, the era of the worldwide "population explosion," the only demographic era within living memory, is coming to a close.

Continued global population growth, to be sure, is in the offing as far as the demographer's eye can see. It would take a cataclysm of biblical proportions to prevent an increase in human numbers between now and the year 2025. Yet global population growth can no longer be accurately described as "unprecedented." Despite the imprecision of up-to-the-minute estimates, both the pace and absolute magnitude of increases in human numbers are markedly lower today than they were just a few years ago. Even more substantial decelerations of global population growth all but surely await us in the decades immediately ahead.

In place of the population explosion, a new set of demographic trends -- each historically unprecedented in its own right -- is poised to reshape, and recast, the world's population profile over the coming quarter century. Three of these emerging tendencies deserve special mention. The first is the spread of "subreplacement" fertility regimens, that is, patterns of childbearing that would eventually result, all else being equal, in indefinite population decline. The second is the aging of the world's population, a process that will be both rapid and extreme for many societies over the coming quarter century. The final tendency, perhaps the least appreciated of the three, is the eruption of intense and prolonged mortality crises, including brutal peacetime reversals in health conditions for countries that have already achieved relatively high levels of life expectancy.

For all the anxiety that the population explosion has engendered, it is hardly clear that humanity will be better served by the dominant demographic forces of the post-population-explosion era. Nobody in the world will be untouched by these trends, which will have a profound impact on employment rates, social safety nets, migration patterns, language, and education policies. In particular, the impact of acute and extended mortality setbacks is ominous. Universal and progressive peacetime improvements in health conditions were all but taken for granted in the demographic era that is now concluding; they no longer can be today, or in the era that lies ahead.


In arithmetic terms, the 20th-century population explosion was the result of improvements in health and the expansion of life expectancy. Human life expectancy at birth is estimated to have doubled or more between 1900 and 2000, shooting up from approximately 30 years to nearly 65 years. Population growth rates accelerated radically thanks to the concomitant plunge in death rates. Despite tremendous population growth, rough calculations suggest that the world's population would be over 50 percent larger today in the absence of any other demographic changes.

The world's population currently totals about 6 billion, rather than 9 billion or more, because fertility patterns also changed over the course of the 20th century. And of all those diverse changes, without question the most significant was secular fertility decline: sustained and progressive reductions in family size due to deliberate birth control practices by prospective parents.

Within the full sweep of the human experience, secular fertility decline is very, very new. It apparently had not occurred in any human society until about two centuries ago in France. Since that beginning, secular fertility decline has spread steadily, if unevenly, embracing an ever rising fraction of the global population. In the final decades of the 20th century, subreplacement fertility made especially commanding advances: According to estimates and projections by the U.S. Census Bureau and the United Nations Population Division, fertility levels for the world as a whole fell by more than 40 percent between the early 1950s and the end of the century -- a drop equivalent to over two births per woman per lifetime.

Indeed, subreplacement fertility has suddenly come amazingly close to describing the norm for childbearing the world over. In all, 83 countries and territories are thought to exhibit below-replacement fertility patterns today [see map on this page]. The total number of persons inhabiting those countries is estimated at nearly 2.7 billion, roughly 44 percent of the world's total population.

Secular fertility decline originated in Europe, and virtually every population in the world that can be described as of European origin today reports fertility rates below the replacement level. But these countries and territories today currently account for only about a billion of the over 2.5 billion people living in "subreplacement regions." Below-replacement fertility is thus no longer an exclusively -- nor even a predominantly -- European phenomenon. In the Western Hemisphere, Barbados, Cuba, and Guadeloupe are among the Caribbean locales with fertility rates thought to be lower than that of the United States. Tunisia, Lebanon, and Sri Lanka have likewise joined the ranks of subreplacement fertility societies.

The largest concentration of subreplacement populations, however, is in East Asia. The first non-European society to report subreplacement fertility during times of peace and order was Japan, whose fertility rate fell below replacement in the late 1950s and has remained there almost continuously for the last four decades. In addition to Japan, all four East Asian tigers -- Hong Kong, the Republic of Korea, Singapore, and Taiwan -- have reported subreplacement fertility levels since at least the early 1980s. By far the largest subreplacement population is in China, where the government's stringent antinatal population control campaign is entering its third decade.

The singularity of the Chinese experience, however, should not divert attention from the breadth and scale of fertility declines that have been taking place in other low-income settings. A large portion of humanity today lives in countries where fertility rates are still above the net replacement level, but where secular fertility decline is proceeding at a remarkably rapid pace.

A glance at the 15 most populous developing countries illustrates the magnitude of fertility change over the last quarter century [see graphs on pages 44-45]. These countries account for about three quarters of the current population of the "less developed regions," and three fifths of the total world population. In addition to China, Thailand is believed to be below the replacement level. Three other countries (Brazil, Iran, and Turkey) are thought to be just barely above the replacement level. Another four (Bangladesh, Indonesia, Mexico, and Vietnam) are slightly higher. Today, in other words, nine of the 15 largest developing countries are believed to register fertility levels lower than those that characterized the United States as recently as 1965. And over the last quarter century, fertility decline in this set of countries has been pronounced: In eight of those 15, fertility dropped by over half.

The regions where fertility levels remain highest, and where fertility declines to date have been most modest, are sub-Saharan Africa and the Islamic expanse to its north and east -- more specifically, the Middle East. Those areas encompassed a total population of about 900 million in 2000, less than a fifth of the estimated total for less developed regions, and a bit under a seventh of the world total. Even for this grouping, however, the image of uniformly high "traditional" fertility patterns is already badly outdated. A revolution in family formation patterns has begun to pass through these regions. In 2000, in fact, the overall fertility level for North Africa -- the territory stretching from Western Sahara to Egypt -- was lower than the U.S. level of the early 1960s. Perhaps even more surprisingly, secular fertility decline appears to be unambiguously in progress in a number of countries in sub-Saharan Africa. For instance, Kenya's total fertility rate is believed to have dropped by almost four births per woman over the past 20 years.

The remarkable particulars of today's global march toward smaller family size fly in the face of many prevailing assumptions about when rapid fertility decline can, and cannot, occur. Poverty and illiteracy (especially female illiteracy) are widely regarded as impediments to fertility decline. Yet, very low income levels and very high incidences of female illiteracy have not prevented Bangladesh from more than halving its total fertility rate during the last quarter century. By the same token, strict and traditional religious attitudes are commonly regarded as a barrier against the transition from high to low fertility. Yet over the past two decades, Iran, under the tight rule of a militantly Islamic clerisy, has slashed its fertility level by fully two-thirds and now apparently stands on the verge of subreplacement.

For many population policymakers, it has been practically an article of faith that a national population program is instrumental, if not utterly indispensable, to fertility decline in a low-income setting. Iran, for instance, achieved its radical reductions under the auspices of a national family planning program. (In 1989, after vigorous doctrinal gymnastics, the mullahs in Tehran determined that a state birth control policy would indeed be consistent with the Prophet's teachings.) But other countries have proven notable exceptions. Brazil has never adopted a national family planning program, yet its fertility levels have declined by well over 50 percent in just the last 25 years.

What accounts for the worldwide plunge in fertility now underway? The honest and entirely unsatisfying answer is that nobody really knows -- at least, with any degree of confidence and precision. The roster of contemporary countries caught up in rapid fertility decline is striking for the absence of broad, obvious, and identifiable socioeconomic thresholds or common preconditions. (Reviewing the evidence from the last half century, the strongest single predictor for any given low-income country's fertility level is the calendar year: The later the year, the lower that level is likely to be.) If you can find the shared, underlying determinants of fertility decline in such disparate countries as the United States, Brazil, Sri Lanka, Thailand, and Tunisia, then your Nobel Prize is in the mail.

Two points, however, can be made with certainty. First, the worldwide drop in childbearing reflects, and is driven by, dramatic changes in desired family size. (Although even this observation only raises the question of why personal attitudes about these major life decisions should be changing so commonly in so many disparate and diverse locales around the world today.) Second, it is time to discard the common assumption, long championed by demographers, that no country has been modernized without first making the transition to low levels of mortality and fertility. The definition of "modernization" must now be sufficiently elastic to stretch around cases like Bangladesh and Iran, where very low levels of income, high incidences of extreme poverty, mass illiteracy, and other ostensibly "nonmodern" social or cultural features are the local norm, and where massive voluntary reductions in fertility have nevertheless taken place.


Barring catastrophe, the world's total population can be expected to grow substantially over the coming quarter century: U.S. Census Bureau projections for 2025 would place global population at over 7.8 billion, almost 30 percent larger than today. Yet, due to declining fertility, population growth is poised to decelerate markedly over the coming generation. The projected annual rate of world population growth in 2025 is just under 0.8 percent, considerably slower than the current projected rate of 1.3 percent, and far below the estimated 2.0 percent annual growth rate of the late 1960s. The great global birth wave will have crested and begun ebbing by 2025. In fact, by those projections, slightly fewer babies will be born worldwide in the year 2025 than in any year over the previous four decades.

The prospective pace of population growth for the different regions of the world is highly uneven over the coming generation [see table on page 46]. The most dramatic increases will occur in sub-Saharan Africa, followed by countries in North Africa and the Middle East. By 2025 more people may be living in Africa than in all of today's "more developed countries" taken together.

The natural growth of population in the more developed countries has essentially ceased. The overall increase in population for 2000 in these nations is estimated at 3.3 million people, or less than 0.3 percent. Two thirds of that increase, however, is due to immigration; the total "natural increase" amounts to just over 1 million. Over the coming quarter century, in the U.S. Census Bureau's projections, natural increase adds only about 7 million people to the total population of the more developed countries. And after the year 2017, deaths exceed births more or less indefinitely. Once that happens, only immigration on a scale larger than any in the recent past can forestall population decline. (The specter of population decline in more developed countries looms even larger if the United States, with its relatively high fertility level and relatively robust inflows of immigrants, is taken out of the picture. Excluding the United States, total deaths already exceed total births by almost half a million a year.)

For Europe as a whole (including Russia), the calculated long-term volume of immigration required to avert overall population decline is nearly double the recent annual level -- an average of 1.8 million net newcomers a year, versus the roughly one million net entrants a year in the late 1990s. To prevent an eventual decline in the size of the 15 to 64 grouping (often termed the "working-age" population), Europe's net migration will have to nearly quadruple to a long-term average of about 3.6 million a year. Migration of this magnitude would change the face of Europe: By 2050, under these two scenarios, the descendants of present-day non-Europeans will account for approximately 20 to 25 percent of Europe's inhabitants.

Even more dramatic are the prospects for Japan, where current net migration levels are close to zero. To maintain total population size, Japan would have to accept a long-term average of almost 350,000 newcomers a year for the next 50 years, and it would need nearly twice that number to keep its working-age population from shrinking. Under the first contingency, over a sixth of Japan's 2050 population would be descendants of present-day gaijin (foreigners); under the second contingency, that group would account for nearly a third of Japan's total population.

Europe and Japan will not lack immigration candidates in the years ahead. If Europe's needed immigration flows continue to come largely from North Africa, the Middle East, sub-Saharan Africa, and South Asia, those migrants will account for only about 3 to 7 percent of the population growth in their home countries. By the same token if Japan, for reasons of history and affinity, relies upon China and Southeast Asia for all its new national recruits, it will require just 2 to 4 percent of those countries' total envisioned population increase over the next 25 years. And as long as a huge income gap separates these more developed and less developed locales, there will be a compelling motive for such migration.

The issue clearly will not be supply, but rather demand. Will Western countries facing population decline opt to let in enough outsiders to stabilize their domestic population levels? Major and sustained immigration flows will entail correspondingly consequential long-term changes in a country's ethnic composition, with accompanying social alterations and adjustments. Such inflows will also require a capability to assimilate newcomers, so that erstwhile foreigners (and their descendants) can become true members of their new and chosen society.

The current outlook for "replacement migration" varies dramatically within the more developed regions. Throughout Europe, vocal (but still marginal) antiforeign political movements have taken the stage in recent years, while more tolerant sectors of the public have worried about the impact of immigration on their welfare states. Yet the continent, populated as it has been by successive historical flows of peoples, possesses traditions and capacities of assimilation that are not always fully appreciated.

The situation looks very different for Japan, where no major influxes of newcomers have been recorded over the past thousand years, and where the delicate distinctness of the Japanese minzoku (race) is a matter of intense, if not always enunciated, public consciousness. Despite reforms in Japanese immigration laws, a community of ethnic Koreans in Japan -- many of them fourth-generation residents of the country -- still does not enjoy Japanese citizenship. Indeed, Japan naturalizes fewer foreigners each year than tiny Switzerland.

It is extraordinarily difficult to imagine any circumstances under which the Japanese public might acquiesce in "replacement migration." Socially and politically, long-term demographic decline seems likely to be a much more acceptable alternative. But these are the only two choices, and over the coming decades all the more developed countries must decide between them. For all societies with long-term fertility rates significantly below the replacement level, the only alternative to an eventual decline of the total population -- or of key age groups within that total population -- is steady and massively enhanced immigration.


The world's population is set to age markedly over the coming generation: The longevity revolution of the 20th century has foreordained as much. The tempo of social aging, however, has been accelerated in many countries by extremely low levels of fertility. In 2025, there will likely remain a few pockets of the world in which populations will remain as youthful as those from earlier historical epochs. For example, the median age in sub-Saharan Africa in 2025 will be just 20 years, that is, as many people would be under 20 as over 20. (Such a profile probably characterized humanity from the Neolithic era up until the Industrial Revolution.) Throughout the rest of the world, however, the phenomenon of aging will transform the structure of national populations, often acutely.

Population aging will be most pronounced in today's more developed countries. By the U.S. Census Bureau's estimates, the median age for this group of countries today is about 37 years. In 2025, the projected median age will be 43.

Due to its relatively high levels of fertility and immigration (immigrants tend to be young), the population of the United States is slated to age more slowly than the rest of the developed world. By 2025, median age in the United States will remain under 39 years. For the rest of the developed world, minus the United States, median age will be approximately 45 years. And for a number of countries, the aging process will be even further advanced.

In Germany, for example, the projected median age in the year 2025 is 46. Greece and Bulgaria are both ascribed median ages in excess of 47. Japan would have a median age of over 49. In this future Japan, more than a fifth of the citizenry would be over 70 years of age, and nearly one person in six would be 75 or older. In fact, persons 75 and older would outnumber children under 15 years of age.

Population aging, of course, will also occur in today's less developed regions. Current developed countries grew rich before they grew old; many of today's low-income countries, by contrast, look likely to become old first. One of the most arresting cases of population aging in the developing world is set to unfold in China, where relatively high levels of life expectancy, together with fertility levels suppressed by the government's resolute and radical population control policies, are transforming the country's population structure. Between 2000 and 2025, China's median age is projected to jump by almost 9 years. This future China would have one-sixth fewer children than contemporary China, and the 65-plus population would surge by over 120 percent, to almost 200 million. These senior citizens would account for nearly a seventh of China's total population.

Caring for the elderly will inexorably become a more pressing issue for China under such circumstances, but nothing remotely resembling a national pension system is yet in place in that country. Even with rapid growth over the next quarter century, China will still be a poor country in 2025. Coping with its impending aging problem promises to be an immense social and economic issue for this rising power.


Given the extraordinary impact of the 20th century's global health revolution, well-informed citizens around the world have come to expect steady and progressive improvement in life expectancies and health conditions during times of peace.

Unfortunately, troubling new trends challenge these happy presumptions. A growing fraction of the world's population is coming under the grip of peacetime retrogressions in health conditions and mortality levels. Long-term stagnation or even decline in life expectancy is now a real possibility for urbanized, educated countries not at war. Severe and prolonged collapses of local health conditions during peacetime, furthermore, is no longer a purely theoretical eventuality. As we look toward 2025, we must consider the unpleasant likelihood that a large and growing fraction of humanity may be separated from the planetary march toward better health and subjected instead to brutal mortality crises of indeterminate duration.

In the early post-World War II era, the upsurge in life expectancy was a worldwide phenomenon. By the reckoning of the U.N. Population Division, in fact, not a single spot on the globe had a lower life expectancy in the early 1970s than in the early 1950s. And in the late 1970s only two places on earth -- Khmer Rouge-ravaged Cambodia and brutally occupied East Timor -- had lower levels of life expectancy than 20 years earlier. In subsequent years, however, a number of countries unaffected by domestic disturbance and upheaval began to report lower levels of life expectancy than they had known two decades earlier. Today that list is long and growing. U.S. Census Bureau projections list 39 countries in which life expectancy at birth is anticipated to be at least slightly lower in 2010 than it was in 1990. With populations today totaling three quarters of a billion people and accounting for one eighth of the world's population, these countries are strikingly diverse in terms of location, history, and material attainment.

This grouping includes the South American countries of Brazil and Guyana; the Caribbean islands of Grenada and the Bahamas; the Micronesian state of Nauru; 10 of the 15 republics of the former Soviet Union; and 23 sub-Saharan African nations. As might be surmised from the heterogeneity of these societies, health decline and mortality shocks in the contemporary world are not explained by a single set of factors, but instead by several syndromes working simultaneously in different parts of the world to subvert health progress.

Russia has experienced a prolonged stagnation and even decline in life expectancy, and its condition illuminates the problems facing some of the other former Soviet republics [see graph on this page]. After recording rapid postwar reductions in mortality in the 1950s, Russian mortality levels stopped falling in the 1960s and began rising for broad groups of the population. By 1990, overall life expectancy at birth in Russia was barely as high as it had been 25 years earlier. With the end of communist rule in 1991, Russia suffered sudden and severe declines in mortality, from which it has not yet fully recovered. By 1999, overall life expectancy at birth in Russia had regressed to the point where it had been four decades earlier.

Although many aspects of Russia's continuing health crisis remain puzzling, it appears that lifestyle and behavioral risks -- including heavy smoking and extremely heavy drinking -- figure centrally in the shortening of Russian lives. A weak and rudderless public health system, combined with apparent indifference in Moscow to the nation's ongoing mortality crisis, also compromises health progress. Although Russia is an industrialized society with an educated population and a large indigenous scientific-technical cadre, such characteristics do not automatically protect a country from the sorts of health woes that have befallen the Russian Federation.

In sub-Saharan Africa, a different dynamic drives mortality crises: the explosive spread of the hiv/aids epidemic. In its most recent report, the Joint United Nations Programme on hiv/aids (unaids) estimated that 2.8 million died of aids in 1999, 2.2 million in sub-Saharan Africa alone. unaids also reported that almost 9 percent of the region's adult population is already infected with the disease. By all indications, the epidemic is still spreading in sub-Saharan Africa. As of 2000, unaids projected that in several sub-Saharan countries, a 15-year-old boy today faces a greater than 50 percent chance of ultimately dying from aids -- even if the risk of becoming infected were reduced to half of current levels.

Given sub-Saharan Africa's disappointing developmental performance and conspicuously poor record of governance over the post-independence period, the pervasive failure in this low-income area to contain a deadly but preventable contagion may seem tragic but unsurprising. Yet it is worth noting that the aids epidemic appears to have been especially devastating in one of Africa's most highly developed and best-governed countries: Botswana.

Unlike most of the region, Botswana is predominantly urbanized; its rate of adult illiteracy is among the subcontinent's very lowest; and over a generation in which sub-Saharan economic growth rates were typically negative, Botswana's was consistently positive. Yet despite such promising statistics, Botswana's population has been decimated by hiv/aids over the last decade. Between 1990 and 2000, life expectancy in Botswana plummeted from about 64 years to about 39 years, that is to say, by almost a quarter century. Recent projections for 2025 envision a life expectancy of a mere 33 years. If this projection proves accurate, Botswana will have a much lower life expectancy 25 years from now than it had nearly half a century ago.

One of the disturbing facets of the Botswanan case is the speed and severity with which life expectancy projections have been revised downward. Assuming most recent figures are accurate, as recently as 1994 expert demographers were overestimating Botswana's life expectancy for 2000 by about 30 years. Such abrupt and radical revisions raise the question of whether similar brutal adjustments await other sub-Saharan countries -- or, for that matter, countries in other regions of the world. This question cannot be answered with any degree of certainty today, but we would be unwise to dismiss it from consideration. hiv/aids may not be the only plague capable of wrenching down national levels of life expectancy over the coming quarter century. Twenty-five years ago, hiv/aids had not even been identified and diagnosed.

Surprisingly, sub-Saharan Africa's aids catastrophe is not projected to alter the region's population totals dramatically. That speaks to the extraordinary power of high fertility levels. Given the region's current and prospective patterns of childbearing, the subcontinent's population totals in 2025 may prove to be unexpectedly insensitive to the scope or scale of the disasters looming ahead. Yet it is the mortality patterns that will do much to define the quality of life for those human numbers -- and to circumscribe their economic and social potential.


Looking toward 2025, we must remember that many 20th-century population forecasts and demographic assessments proved famously wrong. Depression-era demographers, for example, incorrectly predicted depopulation for Europe by the 1960s and completely missed the "baby boom." The 1960s and 1970s saw dire warnings that the "population explosion" would result in worldwide famine and immiseration, whereas today we live in the most prosperous era humanity has ever known. In any assessment of future world population trends and consequences, a measure of humility is clearly in order.

Given today's historically low death rates and birthrates, however, the arithmetic fact is that the great majority of people who will inhabit the world in 2025 are already alive. Only an apocalyptic disaster can change that. Consequently, this reality provides considerable insight into the shape of things to come. By these indications, indeed, we must now adapt our collective mind-set to face new demographic challenges.

A host of contradictory demographic trends and pressures will likely reshape the world during the next quarter century. Lower fertility levels, for example, will simultaneously alter the logic of international migration flows and accelerate the aging of the global population. Social aging sets in motion an array of profound changes and challenges and demands far-reaching adjustments if those challenges are to be met successfully. But social aging is primarily a consequence of the longer lives that modern populations enjoy. And the longevity revolution, with its attendant enhancements of health conditions and individual capabilities, constitutes an unambiguous improvement in the human condition. Pronounced and prolonged mortality setbacks portend just the opposite: a diminution of human well-being, capabilities, and choices.

It is unlikely that our understanding of the determinants of fertility, or of the long-range prospects for fertility, will advance palpably in the decades immediately ahead. But if we wish to inhabit a world 25 years from now that is distinctly more humane than the one we know today, we would be well advised to marshal our attention to understanding, arresting, and overcoming the forces that are all too successfully pressing for higher levels of human mortality today.


Dehydrating Conflict

Remember the last time two nations went to war over water? Probably not, since it was 4,500 years ago. But today, as demands for water hit the limits of a finite supply, conflicts are spreading within nations. And more than 50 countries on five continents might soon be spiraling toward water disputes unless they move quickly to strike agreements on how to share the rivers that flow across international boundaries.

Talk of water wars reverberates around the globe these days. United Nations Secretary-General Kofi Annan said last March that "fierce competition for fresh water may well become a source of conflict and wars in the future," and a recent report of the U.S. National Intelligence Council concludes that the likelihood of interstate conflict will increase during the next 15 years "as countries press against the limits of available water."

Some dismiss these warnings as alarmist, and history seems to be on their side. The only recorded incident of an outright war over water was 4,500 years ago between two Mesopotamian city-states, Lagash and Umma, in the region we now call southern Iraq. Conversely, between the years 805 and 1984, countries signed more than 3,600 water-related treaties, many showing great creativity in dealing with this critical resource. An analysis of 1,831 international water-related events over the last 50 years reveals that two thirds of these encounters were of a cooperative nature. Nations agreed, for example, to implement joint scientific or technological work and signed 157 water treaties.

Others argue, however, that when it comes to water the past will not be a reliable guide to the future. A renewable but not infinite resource, fresh water is becoming increasingly scarce: The amount available to the world today is almost the same as it was when the Mesopotamians traded blows, even as global demand has steadily increased. Just since 1950, the renewable supply per person has fallen 58 percent as world population has swelled from 2.5 billion to 6 billion. Moreover, unlike oil and most other strategic resources, fresh water has no substitute in most of its uses. It is essential for growing food, manufacturing goods, and safeguarding human health. And while history suggests that cooperation over water has been the norm, it has not been the rule. One fourth of water-related interactions during the last half century were hostile. Although the vast majority of these hostilities involved no more than verbal antagonism, rival countries went beyond name-calling on 37 recorded occasions and fired shots, blew up a dam, or undertook some other form of military action.

Lost amidst this perennial debate over whether there will be water wars has been a serious effort to understand precisely how and why tensions develop, beyond the simplistic cause-and-effect equation that water shortages lead to wars. First, whether or not water scarcity causes outright warfare between nations in the years ahead, it already causes enough violence and conflict within nations to threaten social and political stability. And as recent events in the Balkans and sub-Saharan Africa demonstrated, today's civil conflicts have a nasty habit of spilling over borders and becoming tomorrow's international wars. Second, water disputes between countries, though typically not leading to war directly, have fueled decades of regional tensions, thwarted economic development, and risked provoking larger conflicts before eventually giving way to cooperation. The obsession with water wars begs more important questions: What are the early signs and likely locations of water-related disputes, and what can governments and international agents do to prevent the eruption of violence and political instability?


On July 6, 2000, thousands of farmers in the Yellow River basin of eastern China clashed with police over a government plan to recapture runoff from a local reservoir for cities, industries, and other users. The farmers had long relied on that runoff to irrigate their crops, and a bad drought had made the supply more critical than ever. The incident took place in Shandong, the last province through which the Yellow River runs before reaching the sea. The location is noteworthy because the geography of water-related tensions is beginning to show a pattern: Disputes are erupting within countries in the downstream regions of overtapped river basins. China's Yellow River has been running dry in its lower reaches on and off since 1972, and the dry spells have lengthened markedly in recent years, including a record 226 days in 1997. Likewise, water disputes seem to be brewing between Thailand's northern and southern regions as Bangkok's main river supply, the Chao Phraya, dwindles. And there is intensifying friction in the lower portions of the Indus River, where Pakistan's Punjab and Sind provinces have been feuding over water for several years. This past April, protests in Karachi turned violent as demonstrators shouting "Give us water" clashed with police.

These incidents should not be dismissed as isolated and unrelated. Water stress is spreading as populations increase. By 2015, nearly 3 billion people -- 40 percent of the projected world population -- are expected to live in countries that find it difficult or impossible to mobilize enough water to satisfy the food, industrial, and domestic needs of their citizens. This scarcity will translate into heightened competition for water between cities and farms, between neighboring states and provinces, and at times between nations.

The largest and most combustible imbalance between population and available water supplies will be in Asia, where crop production depends heavily on irrigation. Asia today has roughly 60 percent of the world's people but only 36 percent of the world's renewable fresh water. China, India, Iran, and Pakistan are among the countries where a significant share of the irrigated land is now jeopardized by groundwater depletion, scarce river water, a fertility-sapping buildup of salts in the soil, or some combination of these factors. Groundwater depletion alone places 10 to 20 percent of grain production in both China and India at risk. Water tables are falling steadily in the North China Plain, which yields more than half of China's wheat and nearly one third of its corn, as well as in northwest India's Punjab, another major breadbasket.

As farmers lose access to irrigation water and see their livelihoods deteriorate, they may not only resort to violent protest but migrate across borders and to restive and already overcrowded cities. Such has been the case in Pakistan, where falling agricultural output has prompted a massive rural migration to large urban centers, leading to renewed outbreaks of ethnic violence.

Internal water stresses will also shift international political alliances and add to the burden of humanitarian crises. Countries commonly adapt to water stress by importing more of their food, provided they have the foreign exchange to do so. It takes about 1,000 cubic meters of water to grow one ton of grain. By importing wheat and other staples, water-stressed countries can allocate more of their scarce fresh water to cities and industries, which generate far more economic value per liter than agriculture does. Israel, for example, has done very nicely with this approach. Currently water-stressed countries in Asia, Africa, and the Middle East account for 26 percent of global grain imports. As an additional billion people are added to water-stressed countries over the next 15 years and as more countries join the ranks of food importers, demand for international grain will increase. A good portion of that increase may come from China, India, and Pakistan -- all currently grain self-sufficient, but unlikely to remain so for reasons of water and land scarcity. Their governments will inevitably form stronger alliances with the nations from which they choose to import food. For those nations without sufficient foreign exchange to turn to imports, notably those in sub-Saharan Africa, higher world grain prices will likely mean greater hunger and more calls for humanitarian aid.

Finally, a new cause of water-related tensions has surfaced in just the last few years -- the transfer of water system ownership and/or management from public authorities to private multinational corporations. Driving privatization is a confluence of forces: the mounting costs and political liabilities of providing urban water services, increased pressure on governments from the International Monetary Fund (IMF) and the World Bank to reduce water subsidies and public-sector debt, and the growing power of private corporations seeking to profit from the sale of water and related services. Especially where privatization takes place in the presence of poverty and inequality, which is to say in most of the developing world, it can lead to civil protest and violence.

Recent events in Cochabamba, Bolivia's third largest city, underscore these risks. Following the privatization of Cochabamba's water system, which had been plagued by corruption and mismanagement, water rates skyrocketed -- resulting in water bills for some residents equal to a quarter or more of their income. Months of civil unrest culminated in April 2000, when the La Paz government sent soldiers into Cochabamba, where some 30,000 protesters had amassed in the central plaza. Several days of violence ensued, leaving one person dead and more than a hundred injured. The conflict abated only when the water system returned to public control.

Cochabamba is an extreme but not isolated case. Activists in Colombia and South Africa likewise have opposed the privatization of water and other municipal services. Meanwhile, IMF loan agreements with at least half a dozen countries last year called for some degree of water system privatization. The number of urban dwellers worldwide is projected to nearly double -- to 5 billion -- by 2025. Unless governments and lenders strengthen municipal water agencies and steer private-sector involvement toward equity as well as efficiency and toward social justice as well as shareholder profit, more violence like that in Cochabamba may be forthcoming.


Some 261 of the world's rivers are shared by two or more countries. These international watersheds account for about 60 percent of the world's freshwater supply and are home to approximately 40 percent of the world's people. Despite the absence to date of full-scale water wars, unresolved tensions over water have persistently irritated relations, fueled other hostilities, and occasionally led to military action that risked provoking a larger conflict.

Yet, the overarching lesson to draw from the basins of the Jordan, the Nile, and the Tigris and Euphrates rivers and other regions of water dispute is not that worsening scarcity will lead inevitably to water wars. It is rather that unilateral actions to construct a dam or river diversion in the absence of a treaty or institutional mechanism that safeguards the interests of other countries in the basin is highly destabilizing to a region, often spurring decades of hostility before cooperation is pursued. In other words, the red flag for water-related tension between countries is not water stress per se (as is the case within countries), but rather a unilateral attempt to develop an international river, usually by a regional power. In the Jordan River basin, for example, violence broke out in the mid-1960s over an "all-Arab" plan to divert the river's headwaters (itself a preemptive move to thwart Israel's intention to siphon water from the Sea of Galilee). Israel and Syria sporadically exchanged fire between March 1965 and July 1966. Water-related tensions in the basin persisted for decades and only recently have begun to dissipate.

A similar sequence of events transpired in the Nile basin, which is shared by 10 countries -- of which Egypt is last in line. In the late 1950s, hostilities broke out between Egypt and Sudan over Egypt's planned construction of the Aswan High Dam. The signing of a treaty between the two countries in 1959 defused tensions before the dam was built. But no water-sharing agreement exists between Egypt and Ethiopia, where some 85 percent of the Nile's flow originates, and a war of words has raged between these two nations for decades. Along with civil war and poverty, such verbal threats have likely inhibited Ethiopia's water development, leaving the Horn of Africa more vulnerable to drought and famine. Meanwhile Egypt, the regional power, has continued to pursue large-scale river basin schemes unilaterally. As in the case of the Jordan, only in recent years have the Nile nations begun to work cooperatively toward a solution.

Similar scenarios have unfolded in a number of other river basins. India unilaterally constructed a barrage during the 1960s and early 1970s on the Ganges River at Farakka, near the border with Bangladesh, in order to channel more river water to the port of Calcutta. This diversion left Bangladesh with significantly less water for irrigation during the dry season. A 20-year period of intermittent hostility and instability ensued, including increased migration of desperate Bangladeshis across the border to India.

These conflicts share a common trajectory: unilateral construction of a big dam or other development project, leading to a protracted period of regional insecurity and hostility, typically followed by a long and arduous process of dispute resolution. A two-year study of conflict and cooperation within international river basins by researchers at Oregon State University found that the likelihood of conflict increases significantly whenever two factors come into play. The first is that some large or rapid change occurs in the basin's physical setting (typically the construction of a dam, river diversion, or irrigation scheme) or in its political setting, especially the breakup of a nation that results in new international rivers. Secondly, existing institutions are unable to absorb and effectively manage that change -- i.e., when there is no treaty spelling out each nation's rights and responsibilities with regard to the shared river nor any implicit agreements or cooperative arrangements. Even the existence of technical working groups can provide some capability to manage contentious issues, as they have in the Middle East.

Looking ahead, then, which river basins are ripe for the onset of tensions or conflict over the next 10 years? Where are dams or diversions planned or under construction that may hurt other countries and where there is no mechanism for resolving resulting disputes? The accompanying map shows the location of 17 such basins, along with the four in which serious unresolved water disputes already exist or are being negotiated. These basins at risk encompass 51 nations on five continents in just about every climatic zone. Eight of the basins are in Africa, primarily in the south, while six are in Asia, mostly in the southeast. Few are on the radar screens of water-and-security analysts.

Consider, for example, the Salween River, which rises in southern China, then flows into Myanmar (Burma) and Thailand. Each of these nations plans to construct dams and development projects along the Salween, and no two sets of plans are compatible. China, moreover, has not lately been warm to notions of water sharing. It was one of just three countries that voted against a 1997 U.N. convention that established basic guidelines and principles for the use of international rivers. Add in other destabilizing factors in the Salween basin -- including the status of Tibet, indigenous resistance movements, opium production, and a burgeoning urban population in Bangkok -- and the familiar conflict trajectory emerges. Without a treaty in place, or even regular dialogue between the nations about their respective plans, there is little institutional capacity to buffer the inevitable shock as construction begins.

Consider, too, the Okavango, the fourth largest river in southern Africa. Its watershed spans portions of Angola, Botswana, Namibia, and Zimbabwe, and its vast delta in northern Botswana offers a world-renowned wildlife habitat -- the "jewel of the Kalahari." In 1996, drought-prone Namibia revived colonial plans to divert Okavango water to its capital city of Windhoek. Angola and especially Botswana object to the scheme because of its potential harm to the people and ecosystems that depend on the Okavango's flow for their existence. The main institution that can help manage the dispute is the fledgling Okavango Commission, formed in 1994 to coordinate plans in the basin. The commission has recently received renewed support from the Southern Africa Development Community, the U.S. Bureau of Reclamation, and other agencies, but the water dispute continues to simmer.

Several river basins are at risk of future disputes more because of rapid changes in their political settings than any specific dam or development scheme. The breakup of the Soviet Union resulted in several new international river basins almost overnight, and, not surprisingly, institutional capacity for managing water disputes in them is weak. The watershed of Central Asia's Aral Sea, for instance, spanned five Soviet republics that are now independent countries. Tensions among the young nations quickly arose both over how to share the Amu Darya and Syr Darya, the two rivers that feed the Aral Sea, as well as how to ameliorate the human and environmental tragedy caused by the sea's dramatic shrinking -- a result of 40 years of river diversions masterminded by Moscow to grow cotton in the Central Asian deserts. With assistance from international agencies, these young governments have taken tentative steps toward trying to resolve their water dilemmas.

Other recently internationalized basins are only beginning to establish channels of cooperation. The Kura-Araks river system, for example, runs through the politically volatile Caucasus, including the newly independent countries of Armenia, Georgia, and Azerbaijan. The river system is the source of drinking water for large portions of these nations, but millions of tons of untreated sewage and industrial waste regularly push the level of water pollution to 10 to 100 times international standards. On top of the pollution problems, some forecasts project severe water shortages within 10 years. These water strains exacerbate, and are exacerbated by, relations over other contentious issues in the region, notably those of Nagorno-Karabakh and the proposed pipeline to transport Caspian crude oil across the region to Turkey.


History supports the hopeful notion that fresh water may foster cooperation more often than conflict in the years ahead. Water sharing has regularly brought even hostile neighboring states together. But the unprecedented degree of current water stress is creating more zero-sum situations -- in which one party's gain is perceived as another's loss -- both within and between countries. The challenge to governments and international bodies is to recognize the new geography and causes of water-related conflict and to embrace three guiding principles as they act to promote water security.

First, efforts to increase the productivity of water use -- output per unit of water -- are key to defusing tensions as water stress worsens. Measures such as drip irrigation (a highly efficient technique that delivers water directly to the roots of crops), shifts in cropping patterns, recycling and reusing wastewater, and water-thrifty household appliances enable cities and farming regions to do more with less water. Since agriculture accounts for two thirds of water use worldwide, and 80 to 90 percent in many developing countries, increasing the productivity of irrigation water is particularly critical. Several water-short urban areas, including greater Los Angeles and Beijing, are investing in conservation improvements on nearby farms in exchange for the water those investments save. The farmers stay in production, the city obtains additional water supplies at a reasonable cost, and cooperation replaces competition. Moreover, where water conservation and productivity improvements eliminate the need for a new dam or river diversion, or allow a big project to be scaled down in size, they also address a major source of tension and conflict. As the costs of desalination decrease, the desalting of contaminated aquifers and of seawater may generate new drinking water supplies and thereby ease tensions in water-scarce regions as well.

Second, stronger policies are needed in most countries to regulate groundwater use, to price irrigation and urban water in ways that encourage thriftiness instead of waste, and to protect rivers and lakes from degradation. Greater assistance to governments from international agencies in carrying out these policy and management reforms could help lessen the likelihood of future water conflicts. Letting globalization loose in the form of poorly regulated privatization of water services or unconstrained private funding of dam construction will likely cause more problems than it solves. In this regard, the 2000 report of the independent World Commission on Dams, which establishes recommendations for more socially responsible planning and assessment of dams, is an important step forward. Among other things, the report calls for an open decision-making process that includes all those affected by a proposed dam; a thorough examination of the full range of alternatives to determine if a dam is really the best choice; negotiations with and adequate compensation for those adversely affected by dam construction; and, where international rivers are concerned, regional cooperation and collaboration. While some governments have publicly endorsed the commission's recommendations, others -- India, for instance -- have disavowed them.

Third, governments and international organizations must act early and constructively. Some of the most tense water disputes of the 20th century simmered for decades before the rival parties resolved their differences. After three decades of tension in the Jordan basin, Israel and Jordan included a water-sharing provision in the peace treaty they signed in 1994. Tensions among the Nile basin countries are finally easing, thanks in part to unofficial dialogues among scientists and technical specialists that have been held since the early 1990s and more recently to a ministerial-level "Nile Basin Initiative" facilitated by the United Nations and the World Bank. India and Bangladesh ended a 20-year dispute in 1996 with the signing of a treaty that sets out specific terms for sharing the dry-season flow of the Ganges.

The prevailing ad hoc pattern -- implementing agreements, sometimes decades after a crisis emerges -- is not only risky and inefficient, but in many cases preventable. The key is establishing a process of cooperation early in the trajectory before serious hostilities erupt that make it difficult for nations to sit around a negotiating table together. The Indus basin offers a good example. After their independence in 1947, India and Pakistan nearly went to war over the waters of the Indus, which were awkwardly divided by the new political borders. World Bank President Eugene Black used his good offices to mediate the dispute, a long but ultimately successful effort that culminated in the 1960 Indus Waters Treaty.

Strong institutions make a difference. Treaties that provide for effective monitoring and enforcement are often remarkably resilient, holding even when the signatories are engaged in hostilities over non-water issues. The Indus Waters Treaty survived two wars between the signatories and allowed each to pursue its agricultural and economic plans without risking the ire of the other. Long-term programs of joint fact-finding, technical cooperation, and other initiatives that establish a climate of cooperation among countries can pave the way for resolving disputes when they do arise. The U.S. State Department, other donor countries, and a number of U.N. agencies have established a Global Alliance for Water Security aimed at coordinating assistance in priority regions, which may help countries get ahead of the crisis curve.

Most of humanity's long history with water management has focused on developing ways to capture and deliver water in ever greater quantities to people, industries, and farms. We have more or less mastered the technologies that enable us to bend nature to our will. This success, however, has not created a water-secure world. Together, more effective technologies, policies, and international institutions can help prevent and resolve water disputes. But the stresses on rivers and water supplies are now so great and so widespread that we cannot wait for these measures to gradually evolve. We must implement them before long periods of verbal threats, hostilities, environmental degradation, and human suffering engulf more regions of the globe.