CIAO DATE: 5/99

States, Scarcity, and Civil Strife in the Developing World

Colin H. Kahl

First Draft, November 1997, Revised April 1999

Institute of War and Peace Studies, Columbia University

Since the early 1990s, the National Security Strategy (NSS) of the United States has identified civil strife in developing countries as an important threat to American interests and pointed to demographic and environmental pressures as potential sources of these conflicts. As the Clinton administration’s 1996 NSS notes, “America’s security imperatives... have fundamentally changed. The central security challenge of the past half century—the threat of communist expansion—is gone. The dangers we face today are more diverse... [L]arge-scale environmental degradation, exacerbated by rapid population growth, threatens to undermine political stability in many countries and regions.” ₁ President Clinton echoed these sentiments in a speech before the United Nations on June 26, 1997, declaring that efforts to preserve the planet’s natural resources were “crucial not only for the quality of our individual environments and health, but also to maintain stability and peace within nations and among them.” ₂ Similar concerns have been voiced outside Washington. In 1991, for example, then NATO secretary general Manfred Worner argued that “[t]he immense conflict potential building up in the Third World, characterized by growing wealth differentials, an exploding demography, climate shifts and the prospect for environmental disaster, combined with the resource conflicts of the future, cannot be left out of our security calculations...” ₃ And, in an influential and particularly apocalyptic article entitled “The Coming Anarchy,” Robert Kaplan went so far as to suggest that the environment was “the national-security issue of the early twenty-first century. The political and strategic impact of surging population, spreading disease, deforestation and soil erosion, water depletion, air pollution, and, possibly, rising sea levels in critical, overcrowded regions... will be the core foreign-policy challenge from which most others will ultimately emanate...” ₄

A growing number of international relations and security scholars have also begun to study the demographic and environmental sources of political instability and violent internal conflicts in developing countries. ₅ This burgeoning research program stems chiefly from the recognition of two important trends. One trend is the changing nature of warfare in the post-World War II era. In contrast to the pattern prevalent throughout most of modern history, most wars since 1945 have been located in the developing world and have been predominantly intrastate rather than interstate in character. ₆

A second trend is the accelerating pace and enormity of demographic and environmental change. In 1950, the global population stood at 2.5 billion; by 1990, that number had more than doubled to almost 5.3 billion. Over eighty-five percent of this increase occurred in the developing world, where the population ballooned by an astounding 2.4 billion in forty years. This ominous trend continues. While the rate of population growth has declined since its peak in the 1960s, the demographic momentum built into the system currently produces an absolute increase of 78 million additional people every year. The increase in population in the 1990s alone has exceeded the total global population in 1600. The United Nations estimates that global population will increase from its 1998 level of 5.9 billion to 8.9 billion by 2050, with the bulk of this growth occurring in the world’s poorest countries. ₇

The demographic explosion has combined with an incredible expansion in the world economy to place enormous pressure on the planet’s renewable resources and global life-support systems. Indeed, by most calculations, more natural resources have been consumed since the Second World War than in all of human history to that time. ₈ A growing chorus of scientists and environmentalists now fear that environmental degradation in many parts of the world has become, or is fast becoming, so severe that it threatens to undermine the very natural resource base upon which the economies and stable political order of developing countries ultimately depend. ₉

Although several major studies have successfully demonstrated that some link exists between population growth, environmental degradation, resource scarcity, and violence within countries, a more systematic theoretical framework is required. ₁₀ Without such a theory, existing causal claims are difficult to test and offer inadequate guidance for those interested in preventing or limiting the humanitarian, social, and strategic costs of intrastate conflict in the decades ahead. In this chapter, I seek to move the theoretical debate forward by providing a framework that places the state at the center of the analysis, and focuses as much on the social and political processes involved in producing conflict as the demographic and environmental ones. The theory I outline draws extensively on arguments advanced by Jack Goldstone and Thomas Homer-Dixon regarding the demographic and environmental sources of violence. To this foundation, I add other insights from long-standing theories of revolution and rebellion, as well as more recent theories of civil and ethnic conflict from the field of international relations. In doing so, I strive to build a theoretical bridge between the study of demographic and environmental conflicts and the broader study of internal wars.

My argument begins with the claim that demographic and environmental stress can produce three important first order effects (renewable resource scarcity, economic marginalization, and demographic shifts) and two key second order ones (increased social grievances and state weakness). By way of these ecological, economic, and social effects, population and environmental pressures reverberate into politics and produce two potential pathways to civil strife. I call these pathways state disintegration and state exploitation. State disintegration conflicts occur when demographic and environmental stress substantially weakens state authority, thereby reducing the government’s ability to maintain order and increasing the opportunities and incentives for antistate and intergroup violence. State exploitation conflicts, in contrast, occur when state elites seize on resource scarcities and related social grievances to instigate conflicts that advance their parochial interests. The final part of my argument contends that two key intervening variables, institutional inclusivity and groupness, play decisive roles in determining which countries are most prone to state disintegration and state exploitation conflicts. I argue that violence is particularly likely in countries with low degrees of institutional inclusivity (i.e., highly discriminatory and repressive political systems) and high degrees of groupness (i.e., numerous non-overlapping identity groups).

In developing this argument, I first provide a brief survey of existing hypotheses linking demographic and environmental stress to interstate and intrastate conflict, and point to their limitations. The state-centric theory of demographically and environmentally induced civil strife is then outlined. I conclude with a number of methodological suggestions for advancing the “environment and security” research program.

 

Existing Hypotheses

The literature on demographic and environmental stress as a cause of violent conflict is extensive and growing. Traditionally, scholars have focused the bulk of their attention on the various ways population growth and environmental pressures spark international conflict. Some claim that the depletion of natural resources contributes directly to “simple scarcity” conflicts between countries. ₁₁ Others contend that demographic and environmental stress can create large numbers of environmentally displaced persons that spawn conflicts as they cross borders in their quest for survival. ₁₂

While these claims are popular ones, there is scant empirical evidence supporting a strong link between population growth, environmental degradation, resource scarcity, and international violence. There is only limited empirical support for the simple scarcity hypothesis. Resource wars between countries are very rare. Nations have only occasionally clashed over oil and other nonrenewable mineral resources, and international violence driven by conflicts over renewable resources, such as cropland, forests, fresh water, and fisheries, is even less common. ₁₃ A simple cost-benefit analysis explains the infrequency of international resource wars. The destructive capacity of modern weaponry makes international warfare an incredibly costly and risky endeavor, creating a general presumption in favor of resolving conflicts peacefully. The relatively low stakes involved in most interstate resource disputes reinforces this general presumption toward non-violence. While countries may occasionally view certain natural resources, especially oil and water, as critical to short-term national survival, directly transferable into national power, and thus potentially worth fighting over, this is not the case with most natural resources most of the time. ₁₄ Moreover, even when resources are seen as vital, it is almost always cheaper to obtain them through trade or bilateral and multilateral agreements than it is to fight for them. ₁₅ Indeed, in the contemporary period, the countries most dependent on natural resources for their national security are either rich enough to purchase them (e.g., Japan), militarily powerful enough to guarantee access through arrangements backed up by the implied threat of force (e.g., the United States and Russia), or so poor and militarily impotent that external aggression is not a viable option (e.g., most countries in the developing world). ₁₆ Thus, given the high costs of war relative to the stakes involved, and the existence of non-violent alternatives, we would expect the majority of international resource conflicts to be resolved peacefully, if not always fairly. In most cases, nations competing for natural resources will avoid war through trade, mutual accommodation, or less symmetric arrangements dictated by the most powerful actors involved. ₁₇

The contention that environmentally induced population displacements produce international strife also has a weak empirical foundation. There are a few cases of conflict arising from demographically and environmentally induced population displacements, most notably the violence that followed the flow of Bangladeshi migrants into the Indian states of Assam and Tripura during the 1980s. ₁₈ But there are many other cases where similar population displacements have not incited violence. ₁₉

There is considerably more prima facie evidence for a causal connection between population growth, environmental degradation, and violence within countries. One prominent hypothesis, the “deprivation hypothesis,” claims that demographic and environmental stress can substantially reduce individual living standards, thereby producing absolute or relative deprivation. ₂₀ Deprivation stemming from hunger, poverty, unemployment, and resource competition leads, in turn, to frustration-aggression driven crime, riots, religious fundamentalism, terrorism, rebellion, and other forms of social upheaval.

Norman Myers, for example, has argued that people impoverished by population growth and environmental degradation “become desperate people, all too ready to challenge governments through... guerrilla groups...” And Jessica Tuchman Mathews has posited that the demographic and environmental impact on a country’s security is generally “felt in the downward pull on economic performance and, therefore, political stability... [E]conomic decline leads to frustration, resentment, domestic unrest or even civil war.” ₂₁

By itself, the deprivation hypothesis is highly problematic. First, it significantly overpredicts incidents of civil strife. If poverty and deprivation were sufficient to lead people to rebel against their governments or fight each other, the world’s poor would constantly be engaged in organized violence. This is, of course, not the case. The poor usually lack the requisite resources and opportunities to rebel. As Rod Aya notes, “[r]omantic myth aside, oppressed groups cannot rebel in a situation of complete impotence; powerless people ‘are easy victims.’” ₂₂ Moreover, individuals contemplating engaging in organized violence face significant collective action problems. At the individual level, the risks to one’s life and property inherent in antistate or intergroup violence generate high potential costs, and the choice to forgo wages and peaceful exchange with others creates large opportunity costs. On the benefit side of the equation, each individual’s contribution, in and of itself, has very little impact on the prospects for success, and the benefits to be accrued from joining a violent social movement are frequently “public”, or collective, in nature (i.e. they are non-rival and non-excludable). This can create powerful incentives for individuals to “free-ride” on the efforts of others, which, in the aggregate, work against the formation of organized conflict groups. ₂₃

Second, while the deprivation hypothesis seeks to explain a political outcome, it is curiously apolitical. Most substantially, the deprivation hypothesis ignores the activities of state elites and the characteristics of state institutions, despite the fact that numerous studies demonstrate that these factors are essential to any viable theory of civil strife. ₂₄ In Suharto-era Indonesia, for example, social grievances emanating from deforestation have long been managed and suppressed by the strength of the New Order regime. It has only been very recently that a significant weakening of the Indonesian state, stemming from East Asia’s economic crisis, has opened the door for the manifestation of these grievances in the form of organized violence. ₂₅

This insight lies at the heart of a second hypothesis, the “state weakness hypothesis,” put forth by Jack Goldstone, Thomas Homer-Dixon, and others. The state weakness hypothesis contends that population growth, environmental degradation, and resource scarcity only lead to internal wars when they generate incentives and opportunities for individuals and groups to engage in violence. On the incentive side of the equation, demographic and environmental stress contributes to deprivation and intergroup competition over scarce resources. This ratchets-up the level of social grievances and resource competition, increasing incentives for antistate and intergroup conflict. In terms of opportunities, it is argued that population growth and environmental pressures can undermine economic productivity, reduce state revenue, and increase societal demands. This can weaken the capacity, cohesion, and legitimacy of states and thereby increase the “political space” for collective violence. ₂₆

This line of argument points to a number of fundamental dynamics, but still faces two major limitations. First, despite an attempt to bring the state back in, the causal role of the state remains under-theorized. To the extent that state weakness triggers civil strife, it does so in a number of ways not recognized by the existing literature. State weakness not only expands the opportunities for challengers to attack the regime, it can also provide powerful incentives for antistate and intergroup violence by generating acute conflicts of interest and insecurity between social groups. Furthermore, the state need not totally collapse for violence to erupt. As I argue in greater depth below, violence can also result from the purposive actions of state elites and their allies who seek to exploit rising resource scarcities and related social grievances to advance political aims.

Second, while proponents of the state weakness hypothesis recognize that a number of intervening variables mediate the relationship between population growth, environmental stress, and civil strife, they have failed to systematically incorporate these intervening variables into their theoretical framework. Goldstone, for example, notes that “neither environmental degradation nor population growth by themselves act as motors of regional political crises.” ₂₇ In a similar vein, Homer-Dixon argues that “environmental scarcity produces its effects within extremely complex ecological-political systems... [W]hen it does contribute to violence... it always interacts with other political, economic, and social factors. Environmental scarcity’s causal role can never be separated from these contextual factors, which are often unique to the society in question.” ₂₈ And Günther Baechler contends that “passing the threshold of violence definitely depends on sociopolitical factors and not on the degree of environmental degradation as such.” ₂₉ The laundry list of important intervening variables identified by these authors includes: cultural conceptions of the environment; the degree and type of social cleavages; the nature of civil society and the quality of trust, norms, and networks between social groups; the nature of political institutions; the autonomy of the state; and the leadership skills, ideology, and organizational resources of challenger groups and governing elites; among others. ₃₀

Unfortunately, analysts have thus far failed to identify which of these myriad intervening variables are most important. They have also failed to sufficiently specify how these variables interact with demographic and environmental pressures or with the broader causal processes linking population growth, environmental degradation, and resource scarcity to conflict. Absent a more systematic causal theory, existing causal accounts are difficult to evaluate empirically. ₃₁ If every contextual variable is a potentially important intervening variable, then every case in which demographic and environmental pressures are positively correlated with civil strife automatically “confirms” the theory and suggests a causal connection when, in reality, there may not be one. Furthermore, without a clear idea of which intervening variables are most important, the practical utility of the theory for policymakers seeking to locate and head-off potential flashpoints is limited.

In the following section, I outline a systematic, state-centric theory of demographically and environmentally induced civil strife. This theory builds on the best causal insights from existing hypotheses, while also adding a number of new ones drawn from the broader study of internal wars.

 

A State-Centric Theory of Demographically and Environmentally Induced Civil Strife

Demographic and Environmental Stress, Civil Strife, and the State

Before sketching out the theoretical argument, three key concepts employed throughout must be clearly defined. First, I define demographic and environmental stress (DES) as a composite variable encompassing: (1) population growth; and (2) the degradation, depletion, and maldistribution of renewable resources. A number of important factors, including domestic and international economic forces, state policies, cultural practices, and natural conditions, usually combine to cause DES. The justification for analyzing demographic and environmental stress together stems from the high degree of interaction between the two. Guatemala provides an excellent example. In Guatemala, a heavily skewed distribution of land (3 percent of farmers control 65 percent of the country’s best farmland) directly deprives small farmers of an essential resource. This problem is greatly compounded by Guatemala’s rapid rate of population growth (3 percent a year and as much as 9 percent in some rural areas). Together, poor land distribution and population growth lead to environmental degradation by encouraging small farmers to migrate to steep hillsides and forests, converting these poorly suited areas into crop- and pastureland. The Guatemalan state further encourages this migration and degradation by choosing to open up marginal lands for cultivation instead of redistributing land. ₃₂

A second important concept employed throughout is civil strife, which I define as large-scale, sustained, and organized violent conflict within a political entity. This includes revolution, rebellion, insurgency, civil war, and sustained campaigns of terrorism. It does not include smaller scale or less organized forms of violence, such as riots and crime.

Finally, I seek to construct a state-centric theory of demographically and environmentally induced conflict. I define the state along Weberian lines as a set of governing institutions and organizations led and coordinated by individuals occupying offices that authorize them to make and implement binding rules for all people within a territorially demarcated area. ₃₃ This definition points to the dual role the state plays in my causal story: the state as an actor, in the sense of representing a set of individuals and organizations that act according to their preferences; ₃₄ and the state as a set of institutions that enable and constrain the behavior of both state elites and social groups seeking to influence state policy. ₃₅ This definition conceptualizes the state in terms of the broadest and most unique function it is authorized to carry out: binding rule-making. It resists the temptation to include in the definition the specific functions states have traditionally carried out, such as guaranteeing internal order by monopolizing violence, or the empirical existence of the state’s capacity to actually make and enforce binding rules upon society. The specific functions and empirical capacities are incredibly important for determining a state’s strength or weakness, but not its existence.

First Order Effects

The causal connection between population growth, environmental pressure, and civil strife is rarely direct and obvious. Rather, DES can generate several first and second order effects that, when refracted through certain types of political institutions and forms of social organization, make deadly conflict more likely. The challenge for any systematic theory is to clearly identify the connections and logics involved in this complex causal story. In this section, I describe three key first order social and ecological effects that can result from DES: renewable resource scarcity; economic marginalization; and demographic shifts. It should be noted that while it is analytically fruitful to distinguish these three first order effects, they are rarely independent from one another empirically. For example, renewable resource scarcity can lead to economic marginalization and vice-versa, and both interact with demographic shifts. Still, it is productive to take a look at each effect separately in order to best flesh out the various ways in which each contributes to violence.

Renewable resource scarcity. Population growth and the degradation, depletion, and maldistribution of renewable resources can lead to a scarcity of these resources in countries lacking the technological, social, and political ingenuity to adapt. ₃₆ Renewable resources are natural resources, such as fresh water, clean air, arable land, forests, and fisheries, that theoretically regenerate themselves indefinitely through normal ecological processes. ₃₇ Renewable resource scarcity emerges when the stock or flow of a renewable resource is quantitatively depleted or qualitatively degraded at a rate faster than the rate of regeneration. ₃₈

Population growth can lead to scarcity by increasing the demand for resources, while environmental depletion and degradation can generate scarcity by decreasing resource supply. Environmental degradation and depletion can also generate scarcity indirectly if the rising price of a resource makes it more difficult for some individuals to obtain, or if the purchasing power of individuals dependent on a degraded or depleted resource for employment declines, thereby reducing their ability to obtain other vital resources. For example, the Bangladeshi famine of 1974 was dramatically worsened when the flood ravaged rice crop led to increased unemployment, which, in turn, made it more difficult for individuals to purchase the primary food crop aman. ₃₉ Finally, a skewed distribution of resources can produce a condition of “structural scarcity” for large segments of a country’s population by concentrating a resource “in the hands of a few and subject[ing] the rest to greater scarcity.” ₄₀

Population growth and environmental pressures are already putting incredible strains on the Earth’s renewable resource base and promise to place even greater strains on the planet’s ability to provide for human needs in the future. Indeed, after a review of every major estimate of human carrying capacity made in recent decades, Joel Cohen concluded that

The human population of the Earth now travels in the zone where a substantial fraction of scholars have estimated upper limits on human population size. These estimates are no better than present understanding of humankind’s cultural, economic and environmental choices and constraints. Nevertheless, the possibility must by considered seriously that the number of people on the Earth has reached, or will reach within half a century, the maximum number the Earth can support in modes of life that we and our children and their children will choose to want. ₄₁

Economic marginalization. In developing countries with stagnant or slowly growing economies, demographic and environmental stress contributes to economic marginalization. Labor intensive sectors of the economy remain much larger than capital intensive ones in many developing countries. Unfortunately, individuals engaged in labor intensive activities, especially agriculture and other endeavors in which people are directly dependent on natural resources for subsistence and employment, are particularly vulnerable to DES induced marginalization. ₄₂ In rural areas of the developing world, population growth and environmental degradation can lead to chronic poverty and increasing landlessness by reducing the amount of land per person and lowering sustainable yields. An asymmetric distribution of property rights can also contribute to rural inequality and landlessness. These forces often work in tandem. The Philippines provides a classic case. Population growth in the Philippines averaged 2.8 percent annually from 1950 to 1985 (a doubling time of only 25 years). Growth was, and continues to be, particularly high among the rural poor. Rural women average 7 births during their lifetime, compared to 3.8 birth on average for urban women. It is no coincidence that poverty is particularly pervasive in rural areas. In 1985, approximately 28 percent of the Philippine population had incomes below the subsistence levels, and two thirds of these people lived in rural areas. The number of landless workers also increased dramatically during the same period. Population pressures combined with the concentration of land to increase the percentage of landless farm workers in the agricultural labor force from 40 to 56 percent between 1975 and 1980 alone. ₄₃

Population growth can also swell the ranks of the urban un- and under-employed. In sub-Saharan Africa, which has the highest average annual population growth rate of any region in the world (2.9 percent), bulging labor forces and stagnant economies resulted in urban unemployment rates 50-100 percent higher in the early 1990s than in the 1970s. ₄₄ Given the sizable increase in total population over the same period, the absolute number of un- and underemployed individuals is staggering. While the problem is most acute in sub-Saharan Africa, it is not unique to the region. Out of a total global labor force of 2.8 billion, 120 million people are unemployed and another 700 million are underemployed. ₄₅ Moreover, the International Labor Organization has estimated that the size of the global labor force will increase by nearly 1 billion over the next two decades. Most of this growth will occur in the developing world, where approximately 35 million additional people currently seek employment each year. ₄₆

As Goldstone has shown, the problem of economic marginalization is magnified in a stagnant economy by its non-linear nature. For example, if country A with a population of 150 can provide 100 jobs, 50 people are marginalized. If the population doubles to 300 people while the job market fails to expand, there will be 200 marginalized individuals. In this hypothetical example, the population has only doubled, but the number of marginalized individuals has increased by a factor of four. ₄₇ In addition, economic marginalization and DES often feed off each other, creating a positive feedback loop in which population and environmental pressures contribute to marginalization, which, in turn, encourages the poor to overexploit the environment to survive, and so on.

Finally, economic marginalization stemming from DES can be further aggravated by state and societal elites who capitalize on scarcities of cropland, timber, or other resources to enrich themselves at the expense of others. When population and environmental pressures lead to renewable resource scarcity, the price of the resource in question increases. As Homer-Dixon has notes, this “can open up opportunities for fast profits from speculation on resources. It also becomes easier to corner the market on key resources—that is, to capture such a significant fraction of the resource pool that monopolistic profits can be extracted.” ₄₈ Consequently, at the very time DES impoverishes many it can fill the coffers of the few. And, as elites seize control over critical resources and engage in rent-seeking behavior, those marginalized by the direct effects of scarcity are hurt further still. ₄₉

Demographic shifts. Population growth and environmental pressures frequently lead to important shifts in the demographic composition of countries, including changes in age structure, the balance of ethno-cultural populations, and levels of urbanization. Countries experiencing constant high rates of population growth experience “youth bulges,” as each generation produces, in absolute numbers, more children than the previous generation. This phenomenon is apparent today in Mexico. Despite a drop in the fertility rate from 6.4 to 3.4 over the past two decades, Mexico’s population continues to grow at a rapid rate; two million new people are added to the population annually. As a result, approximately 36 percent of Mexico’s population in 1996 was under the age of 15. With an annual increase in the labor force of almost 3 percent a year, the Mexican economy will have to grow at an average rate of 6 percent annually simply to provide menial jobs for the country’s increasing number of young people. In the wake of the 1995 devaluation of the peso, few believe the economy is up to the task. ₅₀ Mexico is not alone in this dilemma. In the developing world as a whole (excluding China) 38 percent of the population was under the age of 15 in 1996, and in sub-Saharan Africa it was a staggering 46 percent. ₅₁

Population growth can also change a country’s ethno-cultural composition. For example, the growth rate of the Palestinian population in Israel’s occupied territories, estimated at between 5.2 and 6 percent a year, is among the highest in the world. ₅₂ Consequently, while the Jewish population in Israel, including Jewish settlements in occupied territories and the Golan Heights, is projected to grow from 4.6 million in 1990 to 6.7 million in 2020 (a total increase of about 46 percent), the Palestinian population in the West Bank and Gaza strip is expected to grow from 1.8 million to 4.7 million over the same period (an increase of 161 percent!). ₅₃

Finally, demographic and environmental stress can contribute to urbanization. Population growth in urban areas can contribute directly to the “natural” growth of cities. ₅₄ Urbanization is also fostered by rural-to-urban migration, which appears to account for 40 to 60 percent of annual city growth in developing countries. “Pull” incentives for migration include greater perceived opportunities for education, employment, health care, and other social services in cities compared to rural areas. Other forces “push” rural residents into cities. Here, DES can lead to migration, as increasing poverty, maldistributions and degradation of land and water resources, and reduced agricultural employment all encourage people to leave rural areas in search of a better life. ₅₅

Natural growth and rural-to-urban migration are currently leading to an explosion of cities in developing countries. Between 1990 and 2025, the number of people living in urban areas will double to more than five billion; 90 percent of this growth will occur in the Third World, where 150,000 are currently added to urban populations every day. Globally, the least developed countries are growing at the spectacular rate of 5 percent a year, and some countries have annual rates of 7 percent or more. The size of cities is also reaching unprecedented levels. In 1950, there were only two “megacities” (cities with 8 million plus people), New York and London. In 1990, there were 21 such cities, 16 of which were in developing countries; by 2015, there will be 33 megacities, 27 of which will be in developing countries. In addition, by 2015, there will be 516 cities with more than one million people, nearly doubling the 1990 total of 270. ₅₆

Second Order Effects

These first order effects produce two key second order ones: increased social grievances and state weakness.

Increased social grievances. Demographic and environmental stress can increase the level of grievances within societies. The primary sources of such grievances are absolute and relative deprivation brought on by population and environmental pressures. Absolute deprivation occurs when there is a discrepancy between what people get and what they need. In other words, absolute deprivation implies hardship. Relative deprivation occurs when there is a discrepancy between what people get and what they feel they are entitled to. Relative deprivation frequently goes hand-in-hand with increasing hardship, but can be present even when individuals’ living standards are improving if others’ are improving even more. ₅₇

Demographically and environmentally induced resource scarcity and economic marginalization can lead to both types of deprivation. As absolute levels of hardship increase and the supply of economic and natural resources shift away from poorer segments of society, resentment will grow among those receiving the short end of the stick. In the Xinjiang province of western China, for example, Muslim poverty and feelings of absolute and relative deprivation result from a combination of high rates of Muslim population growth and state investment funds for irrigation and agricultural improvements that are heavily biased in favor of areas settled by Han Chinese immigrants. ₅₈

Another example of DES contributing to relative deprivation can be found in Israel’s occupied territories. A combination of population growth, intensive agricultural practices, and fragile water ecosystems result in a scarcity of fresh water in the West Bank and Gaza Strip. However, Palestinians experience this scarcity to a much greater degree than Israeli settlers. This stems from a conscious decision by the Israeli state to use its control over scarce water resources to curry favor with Jewish settlers while increasing its social control over Palestinians. The end result is profound sense of relative deprivation among Palestinians. ₅₉

In some cases, urbanization can also contribute to deprivation and ratchet-up the level of social grievances. While it is true that the standard of living and quality of life of urban dwellers is, on average, better than it is in rural areas, the benefits of urban life frequently do not extend to the poorest groups within Third World cities. Many city residents face appalling conditions. Presently, more than 220 million urban residents in developing countries lack access to clean drinking water. At least 420 million lack access to latrines, sewage is often released into waterways untreated, and one- to two-thirds of urban solid wastes are not collected. In excess of 1.1 billion people live in cities where air pollution is greater than healthful levels. Furthermore, poor urban migrants tend to cluster in squatter settlements, where they are frequently exposed to extreme environmental hazards, such as industrial emissions, and can easily become trapped in chronic poverty. At the same time, rising expectations generated by the promise of government services, coupled with greater access to diverse media information on the “good life” enjoyed by wealthier city residents at home and abroad, can significantly increase the sense of relative deprivation. While the majority of urbanites will believe their lives have improved, this is much less likely among the poor and dispossessed. Indeed, income inequality, and thus the potential for relative deprivation, is much wider in urban areas than rural ones. ₆₀

These problems resulting from urbanization can be seen throughout the developing world. Even before their recent economic troubles, infrastructure lagged far behind growing needs in the rapidly industrializing cities of South East Asia and Latin America. And in poorer regions, absolute deprivation can be extensive. Across Africa, for example, urban dwellers in Abidjan (Côte d’Ivore), Dakar (Senegal), Kinshasa (Congo [Zaire]), Lusaka (Zambia), Nairobi (Kenya), and other cities “are confronting rampant urban population growth, a breakdown in urban services such as water and sanitation, a deterioration in urban environmental quality, the AIDS epidemic, and growing social tensions—problems rendered all the more intractable by the extensive poverty of the region.” ₆₁

Other demographic shifts can also lead to a sense of deprivation. Youth bulges can result in large numbers of uneducated, unemployed, and otherwise frustrated young people who feel both absolutely and relatively deprived. Throughout the Central Asian republics of the former Soviet Union, for example, the labor market has collapsed at a time when the number of young people seeking employment has increased. This has contributed to a growing sense of deprivation and widespread bitterness in Kazakhstan, Uzbekistan, and elsewhere. ₆₂ Such youths are prone to radicalization, as the case of Sri Lanka demonstrates. Both the Sinhalese national insurgency in 1970 and the Tamil rebellion in the 1980s reached their peak levels of support when the numbers of 15-24 year olds exceeded 20 percent of the total population of their respective groups. ₆₃ Similarly, changes in the balance of ethno-cultural populations can lead to perceptions of relative deprivation or a fear of such deprivation down the line. The growth of a particular group can place political, social, and economic pressures on slower growing groups and create a great deal of insecurity about the future. ₆₄

A sense of deprivation will be translated into grievances if individuals or social groups come to blame or resent others for their predicament. Grievances may be directed against the state, other groups, or both. Antistate grievances will emerge when the state is seen as passive in the face of, or an active participant in, significant deprivation, whether it be absolute or relative. These grievances will often manifest themselves, at least initially, in rising demands for better state provided goods and services, in addition to growing calls for political and economic reforms, such as land reform, reductions in corruption, and greater political participation. More violent manifestations are also possible. Rapid population growth, soil erosion, deforestation, and unequal land distribution in the Philippines and El Salvador, for example, significantly increased antistate grievances and helped fuel communist insurgencies in both countries during the 1980s. ₆₅ In other situations, grievances will emerge between social groups themselves as a result of population and environmental pressures. This is likely when a sense of relative deprivation makes losers resentful and covetous of the gains of others. Intense resource competition can also increase intergroup grievances by creating zero-sum conflicts of interest between groups. Intergroup grievances will manifest themselves in such activities as increased legal disputes, protests, sporadic violence, and calls for the state to intervene. Finally, in some cases antistate and intergroup grievances will increase simultaneously. This is likely when disadvantaged groups come to blame state collusion with societal rivals for their plight.

State weakness. Another potential second order effect of DES is state weakness. A state’s “strength” represents its ability to actually realize, in the empirical sense, its binding rule-making authority. Two key factors determine the degree of state strength or weakness: a state’s functional capacity and its cohesion. ₆₆ A state’s functional capacity for binding rule-making depends on a combination of coercive power, administrative capacity, and legitimacy of authority. ₆₇ In the context of explaining demographically and environmentally induced conflict, these aspects of functional capacity should be evaluated in terms of their impact on the state’s ability to provide individuals and groups with the public goods of physical and economic security, and deter or repress collective violence.

A second key determinant of strength or weakness is a state’s cohesion. Cohesion represents how unified or divided state elites are, both in terms of their interests and their strategies for advancing these interests. While functional capacity determines the theoretical capacity of the state to make and implement binding rules, the degree of state cohesion determines the ability and willingness of disparate state elites to act as a unified actor when such collective action is necessary to carry out state functions. In terms of maintaining domestic order, the ability and desire of state elites to unify behind policies designed to address societal demands or otherwise head-off organized antistate or intergroup violence is especially important.

State strength can be mapped along a continuum, with “very strong” and “very weak” states representing opposite ends. As the functional capacity and/or cohesion of the state decreases, the state becomes weaker (see Figure 1). States “low” in both functional capacity and cohesion are prime candidates for disintegration. ₆₈ Such situations are likely to be characterized by what Charles Tilly calls “multiple sovereignty”: a situation in which members of a previously subordinate segment of the polity assert sovereignty, non-ruling contenders mobilize into blocs successfully exerting control over some portion of the state, and/or the ruling coalition fragments into two or more competing factions contending for power. ₆₉

Figure 1: State Strength

Demographic and environmental stress can weaken a state’s functional capacity in several interrelated ways. First, the pressures that drive increased societal grievances can also weaken the state. As we saw above, the demands placed on the state from suffering segments of the economy and marginalized individuals are likely to grow as population and environmental pressures mount. Demands may include calls for costly development projects, such as hydroelectric dams, canals, and irrigation systems, subsidies for fertilizer and other agricultural inputs, and urban demands for employment, housing, schools, sanitation, energy, and lower food prices. ₇₀ Demands by key societal elites, regime supporters, and urbanites are likely to be given top priority by the state. The demands of elites and regime supporters for compensations for economic losses, often in the form of subsidies, obviously must be taken seriously if the state wishes to preserve the foundation of its support. ₇₁ Rising urban needs also require governmental attention, because even the strongest states have to be wary of urban unrest. Overall, rising societal demands stemming from DES increase fiscal strains and thus erode a state’s administrative capacity by requiring budgetary trade-offs. A state’s legitimacy may also be cast in doubt if individuals and groups come to blame the government for their plight. ₇₂ The state’s legitimacy may decline, for example, if the government fails to implement land reform in the face of significant rural population growth or if state corruption and patronage appear widespread in a context of swelling urban unemployment rates.

To make matters worse, state interventions to meet rising demands can have unanticipated consequences that further reduces the administrative capacity and legitimacy of the state. Meeting urban demands, for example, often generates a positive feedback loop whereby improved access to goods and services increases “pull” incentives for rural-to-urban migration, which, in turn, increases urban demands and the threats perceived by central authorities. ₇₃ Furthermore, as Homer-Dixon notes, many of the interventions and development schemes state elites see as necessary to meet rising urban and rural demands in the short-term end up seriously distorting prices, encouraging corruption, and accelerating the kinds of economic inefficiencies that undermine administrative capacity in the medium- and long-term. ₇₄

Second, demographic and environmental stress can undermine overall economic productivity, thereby reducing the revenue available to local and central governments at the very time that rising demands require greater expenditures. ₇₅ In countries with stagnant economies, dysfunctional markets incapable of efficiently allocating critical natural resources, and government policies biased against labor (the most abundant factor of production in most developing countries), significant population growth can significantly undermine economic productivity. ₇₆ Rapid population growth increases dependency ratios and makes it more difficult for households to educate and pass on capital to children. By creating large numbers of young people who cannot be educated or productively employed, population growth can undermine the productivity of workers and reduce a country’s ability to compete in the global economy. ₇₇ Higher dependency ratios, in addition to other forms of economic marginalization stemming from demographic and environmental pressures, also force households to shift a higher proportion of financial resources toward basic consumption, limiting their ability to save. In the aggregate, lower domestic savings rates can undermine investments necessary for long-term economic growth or force public and private actors to borrow from abroad, thereby increasing foreign debt. ₇₈

Environmental stresses, including the loss of valuable agricultural land and reductions in crop yields due to soil erosion or pollution, the loss of timber and fuelwood due to unsustainable forestry practices, and the loss of hydroelectric power and transportation due to the siltation of rivers and reservoirs, can also lead to economic decline and reduce the flow of revenue to the state. ₇₉ In Central America, for example, renewable resources account for more than half of all economic production, half of all employment, and most exports. During the 1980s rapid population growth, soil erosion, deforestation, sedimentation of dams, waterways, and coastal fishing grounds, and grossly unequal land tenure systems all contributed to a precipitous decline in per capita income, and aggravated serious fiscal crises throughout the region. ₈₀

The price increases that accompany scarcities of renewable resources can lead to other, less direct, adverse effects that ripple throughout the economy as well. Severe land degradation or drought, for example, can force those dependent on agriculture for employment to search for alternative jobs. Simultaneously, scarcity-induced increases in the price of staple food crops forces people to sacrifice other goods in order to purchase enough food to survive. This can significantly reduce the demand for other products and cause business owners to scale down their operations. Therefore, one possible indirect effect of scarcity may be a reduction in the demand for labor in these businesses, and thus an increase in un- and underemployment, at the very time when the economy must generate more jobs to provide employment for newly displaced agricultural workers. ₈₁

Economic decline and a reduction in administrative capacity can also set in motion feedback loops and other negative consequences that confront states with additional challenges. As the economy declines, for example, states may face greater incentives to rapidly and unsustainably exploit their country’s natural resources in order to generate revenue. Even when this strategy is successful in providing a short-term economic boost, it may undermine the very natural resource base upon which medium- and long-term economic development depends. ₈₂ A reduction in a state’s administrative capacity can also encourage environmental depletion and degradation. In China, for example, a decline in the capacity of the central government to exert its authority over localities has hindered the state’s ability to prevent degradation of water resources. The emerging scarcity of water resources has, in turn, further eroded state capacity. ₈₃

Third, DES can undermine a state’s coercive power. Obviously, revenue short-falls can limit the funds available for police and armed forces. Rising rural-to-urban migration can also reduce a state’s relative coercive power by creating large concentrations of urbanites that are increasingly costly to control. ₈₄ Moreover, renewable resource scarcities sometimes contribute to growing concentrations of wealth and power among non-state elites capable of capturing the profits from the increased value of a scarce resource. This, in turn, can reduce a state’s coercive power. As local and regional elites increase their wealth and power, they may create competing loci of authority, thereby reducing the state’s relative coercive capacity and legitimacy over sub-national units. Population pressures and control over land and water resources in the KwaZulu-Natal region of South Africa, for example, helped solidify the political control of quasi-autonomous warlords during the Apartheid era. ₈₅

In addition to weakening the functional capacity of states, demographic and environmental stress can also undermine state cohesion by generating acute competition and disagreements between elites inside, or allied with, the government. Population pressures, as Goldstone observes,

can create splits between the state and business community, within the military, and within ruling parties... When a population grows, the number of people who aspire to elite positions—positions of leadership, wealth, and influence—is also going to grow. In a healthy society that is rich in resources and is building its institutions and its economy, there is usually room to accommodate a growing elite. But what happens if a country does not have a growing economy, but has invested in universities that are trying to develop an elite for the future, and that future does not come? That is precisely what we see in many developing countries. ₈₆

Furthermore, cohesion may erode as a result of environmental degradation and resource scarcity. In China, for example, provinces have engaged in fierce competition over scarce water resources necessary for agriculture, industry, residential use, and hydropower. As Elizabeth Economy argues, “[i]nter-provincial conflict over the financing of large-scale clean-up operations for major pollution disasters and grand-scale river diversion projects hinders the ability of the state to resolve water crises effectively.” ₈₇ In general, DES can create a situation in which some state and societal elites suffer from renewable resource scarcity while others gain from this scarcity or related remedial efforts. This will pose a challenge to state cohesion as losers turn to the state for help and winners seek to increase their influence. ₈₈ In a political context where demographic and environmental pressures are simultaneously weakening the state’s fiscal health, elites who traditionally benefit from their association with the state will be forced to compete for shrinking government budgets. The enrichment of some segments of the elite population and the deprivation of others, coupled with declining state revenue, can substantially narrow the social base of regime support, increase elite grievances, and potentially fragment the state.

Finally, cohesion may be undermined if state elites disagree amongst themselves about the desirability of responding to rising demands and grievances from marginalized segments of the populace. In extreme cases, disputes can paralyze governments by dividing state elites into rival factions of “reformers,” “moderates,” and “reactionaries” incapable of working together to meet growing societal needs.

State weakness resulting from DES will not necessarily occur in a linear fashion. Instead, the process whereby states disintegrate will likely exhibit non-linear “threshold effects:” the state will first become stressed, then brittle, and then vulnerable to a rapid collapse. In addition, the collapse of state authority may not occur to the same degree in all parts of a country at the same time. It is more likely that state authority will disintegrate unevenly across the regional landscape of a country. A state in decline will be forced to cede authority bit by bit. Eventually, the state may control little more than the capital city and its immediate surroundings. ₈₉ Consequently, the scenarios discussed below for the outbreak of violent conflict as a result of state weakness could conceivably occur at the local, regional, or national levels as state authority disintegrates at these different levels of governance.

Finally, the “straw that breaks the camel’s back” need not be demographic or environmental in character. Population and environmental pressures are frequently medium- and long-term sources of state weakness, whereas the proximate cause of state disintegration in any given case may be a set of short-term policy decisions or exogenous shocks. Such proximate causes may include economic mismanagement, externally imposed structural adjustment, significant changes in a country’s terms of trade, a global economic crisis, a short-term climatic shock, foreign military involvement, or other similar events. Nevertheless, the vulnerability of a state, i.e. the government’s ability to accommodate such poor policies or shocks, can be significantly conditioned by the rate and magnitude of demographic and environmental stress. In short, states made brittle by DES are more prone to a collapse brought on by any number of non-DES factors.

Two Pathways to Civil Strife

Recent scholarly work on internal wars suggests that state weakness makes countries particularly prone to civil strife. A substantial weakening of state capacity and cohesion can lead to violent “bottom-up” conflicts, whether they be antistate or intergroup in character, if the state loses its ability to maintain order. Alternatively, state weakness can lead to “top-down” conflicts initiated by state elites and their allies if they come to see violence as a viable means of perpetuating their rule. ₉₀ Using these arguments as building blocks, I contend that rising social grievances and weakened state authority emanating from DES create two potential pathways to civil strife: state disintegration conflicts and state exploitation conflicts.

State disintegration conflicts. State disintegration conflicts occur when the state is substantially weakened, thereby increasing the opportunities and incentives for antistate and intergroup violence. The few scholars who employ a state-centric approach to demographic and environmental conflict tend to focus on this pathway. Specifically, these authors contend that the erosion of state capacity and authority expands the opportunities for antistate challengers. ₉₁ This follows the line of argument put forth by many contemporary students of revolution and rebellion in sociology and comparative politics, and sees state weakness primarily as a permissive, or necessary, cause of strife. ₉₂

Strong states prevent organized antistate and intergroup strife by altering the cost-benefit calculations of individuals and groups contemplating violence as a possible means of advancing their interests. States possessing powerful coercive apparatuses (police, armed forces, and intelligence organizations) can dramatically increase the costs of intergroup and antistate violence, and thus deter or repress such strife before it threatens the status quo political order. This suggests that a state’s coercive power, commonly measured in terms of its monopoly on the means of physical force, is an incredibly important determinant of a state’s ability to prevent organized violence. Nevertheless, the administrative capacity and legitimacy of a state are also very important. States with a strong and capable administrative apparatus can limit incentives for violence by helping individuals and groups satisfy their basic needs for education, housing, social services, food, water, sanitation, energy, etc. Furthermore, high levels of legitimacy may substantially compensate for low levels of coercive power. A state’s legitimacy, or “right to rule,” is determined by the degree of confidence and trust among key social groups in the state’s intention to protect their fundamental interests and values. High levels of legitimacy ratchet-up the normative inhibitions and perceived costs associated with collective violence. On the other hand, a state whose authority is deemed completely illegitimate will find it much more difficult to maintain domestic order indefinitely, regardless of its coercive power. The absence of legitimacy means that individuals have little confidence in ability of the status quo to preserve their basic interests and values. This lowers the threshold required for individuals and groups to switch their allegiance to antistate challengers promising a better government and better life. ₉₃ Finally, a state’s cohesion matters as well. Cohesive states that maintain the loyalty of key factions and regime supporters are much less vulnerable to challenges from within.

Organized antistate and intergroup violence becomes a real possibility when individuals and groups calculate that the benefits to be accrued from violence outweigh the costs. State weakness makes it more likely that individuals and groups will choose violence as a means of advancing their interests by lowering the costs of such action. Exactly how weak the state must be to permit organized violence undoubtedly varies from context-to-context. If the costs associated with the status quo order are sufficiently intolerable, for example, desperate individuals and groups may take a risky gamble and engage in violence even in a political context where we might normally expect the state to be strong enough to deter such violence. Nevertheless, while the exact threshold of state weakness may vary, it is surely the case that as the state weakens, the opportunities, and therefore the prospects, for violence increase.

This conceptualization provides an essential corrective to analyses that focus solely on societal sources of violence and ignore the state. But current accounts of state disintegration conflicts miss a number of important dynamics linking demographically and environmentally induced state weakness to conflict. Not all instances of state weakness result in conflict; there are many more weak states in the world than there are antistate and intergroup conflicts. In other words, existing state-centric explanations overpredict conflict in the context of a weakened state. State strength is an extremely important factor shaping the cost-benefit calculations of potential combatants, but it is not the only one. Organized antistate and intergroup violence are very costly enterprises. As I discussed above, at the individual level, the cost-benefit calculation confronting potential participants in collective violence provide powerful reasons for them to sit on the sidelines and let others do the fighting. Furthermore, at the group level, there are high potential costs to communal violence, as well as benefits to be gained by continuing to cooperate peacefully and engage in exchange with other social groups. ₉₄ Therefore, we would not expect every instance of state weakness to lead to civil strife. The key questions then become: (1) Under what conditions is the erosion of state authority most likely to result in civil strife?; and (2) What role do demographic and environmental forces potentially play?

Like the boys deserted on the island in William Golding’s Lord of the Flies, actors interacting in a social environment lacking a strong central authority are prone to violent conflict when they come to fear that others will victimize them if they fail to fight. In other words, state disintegration conflicts are most likely when DES and state weakness produce a “security dilemma.” International relations scholars, especially those in the Realist tradition, have long argued that the anarchic nature of the international system not only permits conflict, but actually encourages violent clashes between countries via preventive and preemptive acts. ₉₅ These scholars argue that violent conflict arises not only from real and incompatible interests, evil, and greed, but also from tragedy. At the center of this argument is the notion of the security dilemma. The logic of the security dilemma can help us understand why violent conflicts sometimes occur within countries as well. ₉₆

First, as the state weakens, it generates a great deal of insecurity. An emerging anarchy within countries creates a situation in which no neutral central authority is capable of providing the collective good of physical and economic security to the populace. Social groups (ethno-cultural, religious, class based, organizational, or otherwise) will be left to fend for themselves. Because groups want to remain independent and physically and economically secure, they must be wary of any actions taken by other groups, including remnants of the state, that might threaten their security. Unfortunately, as each group attempts to provide for its security, their actions and preparations, even if ostensibly taken for defensive purposes, can incite fear in others. As a result, an anarchic political context can create a situation in which even defensive actions taken by groups with no interest other than maintaining their security can set off an action-reaction spiral that leaves all parties worse off. This is the security dilemma. It is a situation in which the actions of one group to enhance it’s own security automatically appears threatening to others, and causes reactions that, in the end, leave all less secure. ₉₇ It should be noted that the total absence of state authority (i.e., anarchy) is not required for the security dilemma to operate. The logic of the security dilemma only requires (1) the absence of a neutral central authority capable and willing to resolve intergroup conflicts before they escalate; and (2) a significant threat to a vulnerable group’s economic or physical security. Thus, the security dilemma may operate at levels of state weakness short of complete collapse.

Second, the general dynamics linking state weakness to the emergence of a security dilemma will be compounded in states plagued by high levels of DES, because groups will face other direct threats to their security emanating from resource scarcity, economic marginalization, demographic shifts, and rising social grievances and intergroup competition. These threats produce additional incentives for individuals and groups to restructure the political order to meet their needs, by force if necessary. Thus, even when state weakness alone may be insufficient to create a security dilemma, the combination of state weakness and other first and second order effects caused by DES produce a particularly volatile social context.

In the highly insecure landscape likely to emerge as state authority wanes and demographic and environmental threats to vulnerable groups escalate, it takes very little to start a spiral of violence. The initial spark could result from a random act of intergroup violence, a preemptive act taken under the assumption, reasonable or mistaken, that a rival group soon intends to attack, or any other similar event. Whatever its source, the nature of the security dilemma makes violence especially difficult to contain once it begins. ₉₈ The quest to guarantee security and independence under conditions of state weakness and scarcity conditions all groups to adopt “worst case” thinking; they must assume that violence and exploitation is possible from any other group at any moment. The occurrence of group violence makes this fear a reality, and forces all groups to quickly act to secure their position. Reprisal and counter-reprisal transform a situation in which all feel vulnerable into one in which all are vulnerable; apparent, and frequently illusory, conflicts of interest turn into real ones. ₉₉

The security dilemma stems from two fundamental problems facing social groups. One problem is uncertainty regarding the intentions and motivations of others. As John Herz observed:

[W]hether man is “by nature” peaceful and cooperative, or aggressive and domineering, is not the question... It is his uncertainty and anxiety as to his neighbors’ intentions that places man in this basic dilemma, and makes the “homo homini lupus” a primary fact of the social life of man. ₁₀₀

Strong and neutral states are capable of substantially mitigating the problem of uncertainty. By serving as third party facilitators of communication and arbitration between social groups, strong states reduce uncertainty by increasing the flow, reliability, and impartiality of information. ₁₀₁ Additionally, by providing physical security for groups against the potential attacks of others, strong states reduce the importance of any remaining uncertainty. Not surprisingly, a significant erosion of state authority creates a complex and confusing situation in which groups are likely to be much more uncertain regarding the intentions and motivations of others. Uncertainty is enhanced by the fact that some groups, seeking to exploit the emerging anarchy to advance their interests, have strategic incentives to misrepresent their own intentions to obtain bargaining advantages over rivals, or misrepresent the intentions of others in order to mobilize support. ₁₀₂ Finally, even if groups are fairly confident that others currently harbor no aggressive designs, there is no guarantee that these intentions will not change. ₁₀₃ In short, as Russell Hardin notes, “[o]ne need not hate members of another group, but one might fear their potential hatred or even merely their threat.” ₁₀₄

A second problem is the inability of actors to credibly commit not to exploit the cooperative and peaceful positions of others. Even if groups are fairly certain of each other’s current intentions and interests, they may still find themselves in a tragic spiral of insecurity and conflict if (1) they value their survival highly (a reasonable assumption); (2) there is some probability that a rival group will threaten this survival; and (3) it is difficult for their opponents to credibly commit to forsake violence. ₁₀₅ The situation is especially precarious when an imbalance of group power creates opportunities for a stronger group to exploit the weak, or creates fear on behalf of the weak that this will be the case. Under such circumstances, a credible commitment on behalf of the stronger group not to exploit its position is very difficult, and weaker groups, fearing oppression or destruction, face incentives to engage in preventive or preemptive violence. Unfortunately, this is exactly the scenario likely to emerge as the state weakens in countries with high levels of DES.

The renewal of civil war in the Sudan in 1983 provides an example of this dynamic. By the end of the 1970s, environmental stress in Northern Sudan, resulting in large part from mechanized farming, increased the value of water, land, and oil resources in the South. Northern elites, acting in support of allied Northern mechanized farm owners, pushed South to capture these resources. This posed an enormous threat to the physical and economic security of Southerners. The overall weakness of the Sudanese state, coupled with its Northern bias, left groups in the South to fend for themselves, encouraging them to restart the war against the North. ₁₀₆

Even if weaker groups are pessimistic about their chances of victory, an acute security dilemma raises the perceived costs on inaction, and thus the likelihood of violence. If the costs of doing nothing are high (because status quo threats to security are severe) and the future chances for success appear even more bleak, weaker groups will be encouraged to engage in a “now or never” gamble for survival. ₁₀₇ Under normal circumstances, the inherent collective action problems involved in forming conflict groups suggest that even a moderately weak state should be capable of preventing widespread antistate or intergroup violence. Nevertheless, if the combination of DES and declining state authority poses a significant threat to the survival of a disadvantaged group, the cost-benefit calculus facing individuals will be altered. Most obviously, the benefits to engaging in collective action increase substantially. The costs, relative to doing nothing, also decline. If doing nothing risks annihilation, joining a conflict group may seem rational even if the probability for group success is only marginally increased by any one individual’s participation. In other words, an individual with little to lose might as well take a gamble. Finally, conflicts over, or inspired by, resource issues, such as access to land or water, often provide spoils only to those who participate. This generates significant private gains to be had from joining conflict groups and reduces an individual’s incentive to free-ride.

The guerrilla war which began in 1988 on the island of Bougainville, Papua New Guinea, provides an illustration. Copper was discovered on the island in the 1960s, and in the 1970s, the Panguna copper mine became one of the largest mines in the world. Extensive cooperation occurred between the National Government and the copper company Bougainville Copper Limited, and copper mining soon took on central importance in Papua New Guinea’s economy. Between 1978 and 1987, 60 percent of the revenue from the mine went to the National Government, and since 1972, the mine has produced 16 percent of Papua New Guinea’s internally generated income and 44 percent of its exports. However, the devastating environmental effects of the mine threatened to destroy the livelihood of local landowners. When the National Government and copper company ignored the growing concerns of Bougainville’s indigenous population, local landowners started a sabotage campaign against the mine. The conflict soon escalated to a guerrilla war against troops of the National Government. ₁₀₈

Similarly, in 1994, individuals suffering severe demographic and environmental stress in Chiapas, Mexico, turned to the Zapatista National Liberation Army (EZLN) in a desperate gamble for survival. Large landowners in Chiapas have historically conspired with the ruling PRI to manipulate land tenure arrangements, agricultural prices, and access to credit to the disadvantage of small commercial and subsistence farmers. Moreover, these policies increased as environmental pressures in Chiapas grew. The result was the further marginalization of indìgenas and campesinos (peasants) already experiencing deprivation from rapid population growth (3.6 percent a year from 1970-1990) and severe land degradation. Together, demographic and environmental pressures, the PRI’s failure to serve its role as neutral arbiter of social disputes, and the ability of the Zapatista leadership to channel peasant grievances and class- and ethnicity-based affiliations resulted in organized violence against the Mexican state. ₁₀₉

Finally, the notion of a security dilemma caused by both state weakness and DES suggests a number of possibilities not sufficiently recognized by the existing literature. It suggests that population and environmental pressures can lead to conflict in countries that have had weak states for quite some time, but have managed to avoid conflict prior to the onset of severe DES. It also opens the possibility that population growth and environmental pressures can play an important causal roles in state disintegration conflicts even in situations where the state was weakened by non-demographic and non-environmental forces. For example, in Sierra Leone in the early 1990s, economic mismanagement crippled the functional capacity of the state, drastically reducing state-centered education and employment opportunities. In the context of 2 percent annual population growth, a fertility rate of 6.5, and a large youth cohort (more than 44 percent of the population was under the age of 15 in 1995) traditionally dependent on such opportunities, the recruitment efforts of the Revolutionary Unified Front (RUF) were significantly aided. ₁₁₀ In part, the mobilization of youth was facilitated by the prospect for private gains from armed banditry, which, in the absence of other employment opportunities, provided a viable means of economic survival. ₁₁₁ Thus, both state weakness and population pressures explain the mobilization of antistate challengers; either in isolation is inadequate.

The war in Somalia provides another example. The collapse of the Somali state stemmed largely from non-demographic and non-environmental factors. The economic troubles following the Somali defeat in the 1977-1978 Ogaden war, the proliferation of small arms (which reduced the state’s monopoly on violence), and the end of the Cold War were primarily responsible for the collapse of the Siyaad Barre regime. Nevertheless, population growth, overgrazing, soil erosion, and excessive fuelwood cutting all aggravated inter-clan competition over water and grazing rights during the 1980s. When crop failures and famines devastated Southern Somalia in the late 1980s, the movement of people and livestock across traditional clan-territorial boundaries helped spark deadly inter-clan conflict over access to wells and crops. ₁₁₂ “The militarized conflicts in the bush gave added power to warlords within the clans, and diminished the power of those who had peace interests. With the collapse of the regime, these forces engendered a dynamic that no indigenous actors had the power or legitimacy to stop.” ₁₁₃ The importance of demographic and environmental forces becomes even clearer when the experiences of Northern and Southern Somalia are compared. The relative calm maintained in Northern Somalia, where drought and its dislocations were far less severe, and the chaos in the South suggests that any explanation focusing on the collapse of the Barre regime as the sole cause of conflict is suspect. ₁₁₄ Both state weakness and DES were important.

State exploitation conflicts. State exploitation conflicts represent a second pathway whereby DES induced social grievances and state weakness can lead to violence. In contrast to state disintegration conflicts, however, state exploitation conflicts can occur at levels of state weakness far short of total collapse. State exploitation is also primarily a “top-down” process, rather than one driven chiefly by “bottom-up” dynamics.

Social schisms emanating from demographic and environmental pressures present state elites in the developing world with both incentives and opportunities to instigate violence. On the incentives side of the equation, rising grievances and social conflict, if left unchecked or unchanneled, can potentially threaten the actual or perceived viability of a regime. To stay in power, all regimes, even the most authoritarian ones, require some base of social support. In a country experiencing severe demographic and environmental pressures, state elites are likely to fear an erosion of this support if they are unable to meet rising societal demands. Prior to the 1978 Iranian revolution, for example, demographic pressure on the agricultural sector of the economy contributed substantially to mass opposition against the Shah. The Iranian population nearly doubled from almost 17 million to 32 million between 1950 and 1976. ₁₁₅ Despite the Shah’s efforts at land reform in the early 1960s, only 22 percent of Iran’s peasants in the 1970s acquired sufficient land for subsistence. Significant rural-to-urban migration contributed to unemployment, a profound sense of relative deprivation (as new urbanites quickly became aware of disparities in national income and lifestyles), and added fuel to the revolutionary fire. ₁₁₆

Threats to a regime encourage state elites to search for strategies that will stabilize their base, mobilize new supporters, and co-opt or crush political opponents. ₁₁₇ Tragically, state elites often conclude that the instigation of intergroup violence is an effective means of achieving these goals. Most obviously, inciting intergroup violence can be a brutally effective way of crushing political opponents. Less obviously, but no less importantly, violence can divert attention away from the regime’s failings while also making key social groups dependent upon the state for their physical well-being, thereby insuring their allegiance. The insecurity generated by state-sponsored intergroup violence provides existing and potential allies of the state with powerful incentives to support the regime; after all, if these groups back the regime, both their chances for survival and their share of the spoils from victory increase dramatically. ₁₁₈

Incentives for state-sponsored violence are only half the story, however, because state elites cannot create intergroup violence by fiat. Because violence is costly for participants, individuals and groups must have good reasons to attack their neighbors. It is here that DES provides threatened elites with golden opportunities to perpetuate their rule. The resource and economic competition engendered by mounting population and environmental pressures can easily create zero-sum distributional conflicts between social groups, and absolute and relative deprivation stemming from DES can expand the breadth and depth of intergroup animosities. This can create a volatile social climate in which state elites find ample opportunities to spark violent conflicts by playing on the fears, hatreds, and desires of contending groups. Elites may indirectly incite violence by capitalizing on propaganda advantages, such as control over newspapers, television, radio, and public appearances, to exaggerate the threats posed by certain groups or the potential gains to be had if one group attacks another. ₁₁₉ More directly, elites may actually engineer clashes by supporting or encouraging attackers and then standing by while opposing groups kill one another. This kind of direct encouragement is most likely to succeed when the state can insure participants concrete benefits, such as access to coveted natural resources, if they engage in violence.

The 1994 genocide in Rwanda provides a tragic example of state exploitation. By the early 1990s, the collapse of coffee and tea prices, severe drought and famine, structural adjustment policies, and civil war with the Tutsi-led Rwandan Patriotic Front (RPF), all combined to place enormous pressure on President Habyarimana’s Hutu regime. The prospect for continued Hutu dominance of the Rwandan state was further jeopardized by increasing international pressure to democratize and a power-sharing agreement negotiated under the 1992 Arusha Accords (which ended the war with the RPF). Fearing a loss of political power and economic privilege, Hutu elites in both the civilian and mititary sectors sought to maintain their authority by organizing militias and by formenting anti-Tutsi and anti-RPF sentiment among the general populace.

After the death of Habyarimana in a suspcious plane explosion in April 1994, the army and militias began to orchestrate and instigate widespread violence against Tutsis and moderade Hutus, during which between 500,000 and a million people were killed. ₁₂₀

Opportunities for state exploitation created by demographic and environmental stress were critical to the ability of Hutu extremists to execute their vicious campaign. From 1985 to 1990, the annual population growth rate in Rwanda was 3.3 percent. Moreover, with a population density of around 290 people per square kilometer (km), and an adjusted density of 422 per square km (if lakes, national parks, and forest reserves are excluded), Rwanda ranked as Africa’s most overcrowded country. Given the predominantly rural nature of Rwanda’s economy, these population dynamics inevitably encouraged competition over increasingly scarce land, competition that was further compounded by falling soil fertility stemming from overcultivation and erosion. By the late 1980s, most available land was under cultivation and agricultural production was no longer growing faster than the population. ₁₂₁ Thus, demographic and environmental pressures were already increasing intergroup tensions before the economic and military shocks of the late 1980s and early 1990s made conditions even more desperate for the populace and regime. ₁₂₂ Under conditions of acute land scarcity and diminishing opportunities, increasing numbers of individuals, especially uneducated and underemployed youths, began to look for any means of improving their condition and self-esteem. By creating a large pool of desperate recruits from which Hutu extremists could draw, DES provided opportunities for state elites to rally their brethren to violence against those supposedly to blame for their misfortune. ₁₂₃ In addition, scarcity created a situation in which individuals could reap private gains by possessing the land and livestock of those they killed, thereby encouraging violent collective action. Thus, as Gérard Prunier notes:

The decision to kill was of course made by politicians, for political reasons. But at least part of the reason why it was carried out so thoroughly by the ordinary rank-and-file peasants in their ingo [small administrative units comprised of ten households] was a feeling that there were too many people on too little land, and that with a reduction in their numbers, there would be more for the survivors... Villagers also probably had a vague hope that if things settled down after the massacres they could obtain pieces of land belonging to the victims, a strong lure in such a land-starved country as Rwanda. ₁₂₄

Due to the inherent threats to the long-term stability of any regime posed by society-wide warfare, we would expect state elites to prefer to keep state-sponsored violence from escalating to civil war. Nevertheless, even if this is what most state elites ultimately prefer, they may, like the sorcerer’s apprentice, lose control of the forces they unleash. State exploitation makes it incredibly difficult for ruling elites to credibly commit to prevent future aggression by the regime or its allies. Under such circumstances, victimized groups concerned with their security and fearful of future hostilities may feel compelled to initiate preventive or preemptive violence before the situation worsens, even if their prospects for success are low. ₁₂₅ As Hardin notes, intergroup violence can lead to the collapse of general expectations of reasonable behavior, “making preemption seem to be a compelling interest, thereby ensuring future violence.” ₁₂₆ In the spiral of attack and counter-attack that ensues, an entire society can easily be dragged into the Hobbessian vortex of all-out civil war.

If intergroup violence is inherently difficult to control, why would state elites ever step on the slippery slope to begin with? Two possible reasons suggest themselves. First, immediate needs to shore up state legitimacy or reduce DES induced threats to the stability of the regime may encourage state elites to adopt short time horizons. Under such circumstances, long-term costs tend to be heavily discounted, and actions that appear irrational in the long-term may be viewed as rational in the short-term. Second, since most of the potential benefit from such gambles will be accrued by those in power, while the bulk of the potential costs will be born by the citizenry and opposition, state elites may be willing to run the risk of broader conflict. ₁₂₇

Intervening Variables

Not all instances of DES induced state disintegration or state exploitation lead to large scale violence. In this section, I suggest that two intervening variables, institutional inclusiveness and groupness, play crucial roles in determining when state disintegration or state exploitation dynamics are most likely to spawn civil strife.

Institutional inclusivity. One important intervening variable is institutional inclusivity. The state, as defined above, is both a set of actors (state elites) and organizations, and a set of institutions that enable and constrain state elites and social actors. As conceptualized here, the inclusivity of state institutions refers to the degree to which key social groups are institutionally empowered to participate in, and influence, decision-making by state elites. Inclusive institutions guarantee a broad spectrum of social groups membership in the decision-making tiers of the executive (e.g., cabinets, the civil service, the military, etc.), or representation in a national legislature institutionally empowered to influence and constrain executive policies. Somewhat inclusive states allow structurally disadvantaged groups input into national decision-making. Highly inclusive institutions also provide groups with mutual vetoes or require supra-majorities for policy decisions impacting significantly on group rights. ₁₂₈ In exclusive states, on the other hand, decisions are made by a narrow clique of state elites and their allies. ₁₂₉

John F. Kennedy once observed that “[t]hose who make peaceful revolution impossible will make violent revolution inevitable.” ₁₃₀ Although institutional inclusivity is no panacea, countries with inclusive institutions will be less prone to demographically and environmentally induced civil strife. Institutional inclusivity provides several safeguards against state disintegration conflicts. First, and most obviously, inclusive institutions provide a mechanism for peaceful compromise. Because the costs and risks, even for stronger parties, of communal or antistate violence are high relative to the costs of many cooperative arrangements, groups have inherent incentives to reach non-violent resolutions. Inclusive institutions facilitate such compromise by providing a forum that makes it easier and cheaper for parties with a basic common interest in peaceful coexistence to resolve residual conflicts of interest without resorting to violence. In the language of the new institutional economics, the existence of inclusive institutions lowers the “transaction costs” involved in reaching cooperative arrangements relative to the costs of reaching an agreement in the absence of such institutions. ₁₃₁ Inclusive institutions also provide a forum for repeated conflict resolution, thereby lengthening the “shadow of the future.” Compromise is encouraged because groups feel they can trade concessions now on certain issues for advantageous deals on other issues down the line. Absent such an inclusive forum, prospects for future cooperation are lower and groups have fewer incentives to make sacrifices in the short-term. ₁₃₂

Furthermore, in countries where inclusive institutions have been in place for an extended period of time prior to the weakening of state authority, a long history of interaction, cooperation, and compromise can help establish reputations for trustworthiness and facilitate mutual expectations among groups regarding the prospects for future cooperation. In some cases, a history of repeated cooperation can foster a robust norm of intergroup respect and mutual trust by creating normative inhibitions against collective violence as a legitimate means of advancing group interests. ₁₃₃ This reduces the fears of disadvantaged groups when the state weakens. The more confident disadvantaged groups are that stronger groups will continue to respect their rights, the less they will fear becoming victims, and the easier it will be for stronger groups to make credible assurances. ₁₃₄ On the other side of the equation, robust norms of respect also discourage stronger groups from exploiting their advantageous position and initiating violence. Exclusive institutions, in contrast, assert that only certain groups are deserving of rights and protection. This will not foster much intergroup respect.

When the state is weak and institutions are exclusive, the rule of law is likely to be replaced by the rule of the jungle; weaker groups will fall prey to their more powerful rivals. If the state is too weak to impose a just solution, and if political institutions do not empower victims of severe deprivation and insecurity to peacefully influence government or business practices, disadvantaged groups are faced with the following menu of unpalatable and extra-institutional choices: (1) silent suffering at the hands of more powerful actors; (2) peaceful mobilization with little prospect for success; or (3) armed resistance. In Indonesia, for example, the demise of the Suharto regime has opened the political space for communities with long-standing grievances related to forest lands to express their anger. Given the lack of credible and inclusive legal and political institutions, these groups are left with few options for settling these disputes peacefully. Thus, groups increasingly resort to “spontaneous direct actions which often flare into violence.” ₁₃₅

Lebanon provides another example of the dangers of exclusive institutions in the context of a disintegrating state. Rapid Muslim population growth in Lebanon, including a large influx of Palestinian refugees, tipped the demographic balance in favor of Muslims in the 1960s and 1970s. At the same time, Maronite Christians refused to renegotiate the balance of ethno-cultural political power established by the Lebanese constitution and an unwritten 1943 agreement. The political primacy of Maronite Christians was established at a time when Christians made up the majority of the Lebanese population. When Lebanon’s inflexible institutions failed to accommodate changing demographic realities, Sunni, Shiite, and other Muslim sects joined together in violent opposition. In a political landscape lacking a strong and neutral state, the sense of deprivation on the part of Muslims engendered by population changes coupled with exclusive governing institutions contributed greatly to the tensions that exploded into civil war in 1975. ₁₃₆

It should be emphasized that institutional inclusivity will not prevent all state disintegration conflicts. When conflicts of interest between contending groups are intense and zero-sum, when stronger groups have clear intentions to exploit the opportunities provided by their superior capabilities, and where the stakes for weaker groups of being exploited include physical or cultural annihilation, inclusive institutions may be of little consequence. But, when all major political actors ultimately prefer continued peace to bloodshed, inclusive institutions help avoid conflicts arising from residual conflicts of interest, the security dilemma, and spiral dynamics. In both the Philippines and El Salvador, for example, the development of democratic institutions has helped restore a modicum of faith in government among the dispossessed, thereby decreasing their need to turn to communist rebels for security. ₁₃₇

Finally, inclusive institutions also reduce the likelihood that DES will result in state exploitation conflicts. Inclusive institutions allow structurally disadvantaged groups to limit the ability of more privileged groups to exploit them. In exclusionary states, state elites and their allies have both the power and the incentive to exploit resource scarcities to capture quasi-monopolistic rents or otherwise manipulate social schisms to advance their narrow self-interests, because the social costs of such policies are spread out across society while the benefits are accrued by the narrow clique at the top. Furthermore, control over vital resources, as well as the coercive apparatus of the state, provides state elites and their allies an enormous potential to harm disadvantaged groups. Under these circumstances, disempowered groups are likely to fear for their physical and economic survival, and the state and its allies will find it extremely difficult to credibly commit not to exploit their position of power and privilege. More inclusive institutions, however, put constraints on state exploitation. Because the very actors potentially hurt by state abuses are also empowered to influence government decisions, state exploitation, at least in its most extreme varieties, is less likely to occur or lead to violent conflict. ₁₃₈ Moreover, because privileged groups know that it will be difficult and costly to exploit weaker groups, they are less likely to try. In short, inclusive institutions dramatically reduce the uncertainty and credible commitments problems related to potential state exploitation by making trust and limits on state behavior easier to enforce. ₁₃₉ In Mali, for example, democracy has limited corruption and facilitated consensual decisions related to the use of scarce resources, thus helping one of the poorest countries in Africa escape the civil strife plaguing so many of its neighbors. ₁₄₀

Groupness. Demographically and environmentally induced state disintegration and state exploitation conflicts are also more likely in societies characterized by high degrees of groupness. Here, groupness is conceptualized in terms of the degree of societal segmentation. Individuals always have multiple identities and group affiliations. The degree of segmentation in any given society depends on the extent to which clusters of individuals depend on distinct identity-groups (whether they be ethno-cultural, kin-, tribe-, religious-, or class-based) for physical, economic, and psychic security, as opposed to a number of overlapping and cross-cutting identity-groups. ₁₄₁ Both homogenous societies dominated by a single identity-group, and heterogeneous societies with significant cross-cutting group affiliations and interests, are characterized by low segmentation and thus low groupness. Societies characterized by high groupness, on the other hand, tend to be heterogeneous with few overlapping and cross-cutting group affiliations and interests. Figure 2 provides an illustration to help clarify the distinction between “low” and “high” groupness.

Figure 2: Groupness*

* Note: A, B, C, X, and Y represent different identity-groups.

High degrees of groupness increase the prospects for both state disintegration and state exploitation conflicts by facilitating the formation of conflict groups. Strong group identification can contribute to conflict group formation by causing individuals to conceptualize the costs and benefits associated with their personal choices in collective rather than individualistic terms. Even when individuals strongly prefer change in the status quo and believe violence to be the best route for achieving such change, an individualistic cost-benefit calculation tends to encourage free-riding; if an individual’s contribution to the preferred outcome is slight and violence is costly, they will be inclined to sit on the sidelines and let others take all the risks. However, if costs and benefits are defined in collective terms, individuals realize that their individually rational behavior is collectively irrational, because, in the aggregate, widespread free-riding prevents the advancement of group interests. Thus, a collective conceptualization of interests makes collective action more likely. ₁₄₂

Furthermore, a high degree of groupness increases the prospects for conflict group formation even if individuals fail to define their preferences in collectivist terms. Michael Taylor has argued that the prospects for joining an organized conflict group are higher when individuals are part of a pre-existing community of individuals who share at least some beliefs and values in common, engage in direct and many-sided interactions, and engage in some form of reciprocity. ₁₄₃ A strong community allows individual actions to be monitored, and communities frequently have at their disposal a powerful array of positive and negative social sanctions that can be used as selective incentives to encourage participation in conflict groups. The existence of a community also facilitates conditional cooperation (i.e., cooperation based on direct or defuse reciprocity). Participation in organized violence “is conditioned on others continuing to participate, but the experience of conditional cooperation, the knowledge conditions which are necessary for successful conditional cooperation (e.g. knowing that others are cooperating), and the effectiveness of social sanctions used during a rebellion all derive from the fact that the participants in the rebellion are members of a pre-existing community and will continue to be members of the same community after the rebellion.” ₁₄₄ Thus, in game theoretic terms, cases of organized violence are embedded in a larger “iterated” game of social interaction within the community. Consequently, failure to participate in organized violence on behalf of the group might bring about social sanctions and threaten an individual’s ability to reap benefits from other aspects of participation in the community. This encourages individual participation in organized violence and facilitates the formation of conflict groups even if individuals conceive of costs and benefits solely in individualistic terms.

Finally, a high degree of groupness also lowers the costs for political entrepreneurs seeking to mobilize conflict groups in order to advance their own interests. Since discrete groups already exist, political entrepreneurs do not have to spend as many precious resources overcoming collective action problems to create them. Instead, state and local elites can devote their time, energy, and resources to assuring the support of pre-existing groups for their cause. A high degree of societal segmentation and group cohesion further facilitates elite driven mobilization by making it much easier for elites to target discreet groups for propaganda purposes. ₁₄₅

While high degrees of groupness make countries more conflict prone, low groupness has the opposite effect. Most obviously, significant cross-cutting group ties and interests increase the perceived costs for individuals contemplating intergroup violence. The more individuals see violence as self-defeating, the less likely it is to occur. ₁₄₆ Moreover, if individuals have several groups to which they can turn in order to meet their basic needs for economic and physical security, as well as psychic benefits from group membership, sanctions arising from a failure to participate in violence on behalf of any particular groups will be less likely to overcome collective action problems. Conflict group formation is also less likely because low groupness dramatically increases the organization costs for political entrepreneurs.

Kenya provides a cogent example of low degrees of groupness limiting strife. Between 1991 and 1993, rural ethnic violence erupted in Kenya’s fertile Rift Valley, Nyanza, and Western Provinces, where significant population growth, an acute scarcity of arable land, and high degrees of ethnic identification provided the Moi regime with a golden opportunity to stoke intergroup tensions and discredit the ongoing process of democratization. The violence failed to engulf the entire country, however, because critical urban areas stayed relatively calm. To a significant degree, this resulted from the fact that a growing urban middle class, with a vested economic interest in avoiding the kind of tribal violence that swept over neighboring Rwanda, cut across ethnic lines. Consequently, urbanites refused to join in the fray. ₁₄₇

Interactions between causal variables. To a significant extent, the degree of institutional inclusivity and groupness in any given polity is determined by factors that are prior to, and independent from, demographic and environmental forces. Nevertheless, population growth and environmental pressures, increased social grievances, and the processes of state disintegration and state exploitation can have some affect on the level of institutional inclusivity and/or groupness. Demographic and environmental stress, as well as the state disintegration and state exploitation DES engenders, can, under a limited set of circumstances, affect the degree of institutional inclusivity. In countries that already have fairly exclusive institutions, government discrimination may intensify if the threats produced by population growth, environmental degradation, and a significant weakening of the state encourage governmental repression. As Homer-Dixon notes, demographic and environmental pressures “might overwhelm the management capacity of institutions in developing countries” and encourage state elites to adopt policies that are more “extremist, authoritarian, and abusive of human rights” in order to address emerging threats to the regime’s survival. ₁₄₈ Similarly, already exclusive states may become more so if resource scarcities and state exploitation increase the concentration of political power in the hands of the few at the top.

Interaction effects are even more likely with regard to groupness. Groupness will likely increase as resource scarcity, economic marginalization, and demographic shifts increase hardships, as the state weakens, or as the state tilts strongly in favor of some groups while leaving others to fend for themselves. Under such circumstances, disadvantaged individuals will increasingly turn to identity-groups and local organizations to provide them with physical, economic, and psychic security. ₁₄₉ As Michael Ignatieff observes, individuals tend to retreat to their groups “when the only answer to the question ‘Who will protect me now?’ becomes ‘my own people.” ₁₅₀ In Egypt, for example, economic marginalization engendered by population pressures has led increasing numbers of people in Cairo and elsewhere to turn to radical Islamic groups for basic social services and hope of a better life. As the Egyptian journalist Mohammed Auda notes, desperate Egyptians are prone to “join people who believe this is a society that must be destroyed because it can’t be changed. And population is one of the main causes.” ₁₅₁

Moreover, threats to a disadvantaged group’s survival create powerful incentives for elites within the group to act as political entrepreneurs and provide selective incentives to encourage mobilization. The elites themselves may feel threatened, and if they have sufficient political, organizational, or economic resources to provide selective incentives, their individual action can make a powerful difference in the probable outcome. Following Olsonian logic, we would expect such elites to utilize their resources as selective incentives to facilitate the formation of conflict groups. ₁₅₂ Alternatively, opportunistic elites may capitalize on threats, real or imagined, to fellow members of their disadvantaged group as a means of gaining political support and power. Under normal circumstances, these elites may not have sufficient resources to convince members of their group to mobilize, but threats to group survival lower the threshold of required selective incentives by increasing individual incentives to cooperate.

Finally, there is likely to be some interaction, or at least an elective affinity, between institutional exclusivity and groupness. On the one hand, if institutions become more exclusive, groupness may increase as disadvantaged groups face added incentives to mobilize in opposition. On the other hand, highly segmented societies may be more inclined toward exclusive institutions in the first place. ₁₅₃ Therefore, institutional inclusivity and groupness should be conceptualized as partially endogenous intervening variables. In other words, they not only influence, but are to some degree influenced by, the effects of DES and each other.

Summary. The general causal argument of the state-centric theory is summarized schematically in Figure 3.

Figure 3: Causal Diagram*

 

Studying Demographically and Environmentally Induced Civil Strife

Marc Levy has recently argued that

[b]etter research and better advice can grow out of an understanding that environmental factors interact with a variety of other factors to spawn violent conflict... There appears to be no interesting mechanisms that are purely and discretely environmental. Therefore any research strategy aimed at deepening our understanding of security problems by studying only environmental connections can never succeed... Instead, for those who worry about global conflict, the attention ought to be on how the constellation of factors that promote or impede violence operate. ₁₅₄

Levy has a point. Demographic and environmental variables are clearly neither universally necessary nor wholly sufficient causes of civil strife in developing countries. Even if DES is a necessary condition for violence in a particular case, it is certainly not a necessary variable across the universe of possible cases. As conflicts in the former Yugoslavia and elsewhere demonstrate, civil strife can occur in the absence of significant population and environmental pressures. Moreover, DES is not a wholly sufficient cause of conflict because all countries experiencing significant population growth and environmental pressures do not descend into chaos. Even when DES is an important source of violence, it is only the combination of, and interaction between, DES and other political and social variables that produce civil strife. ₁₅₅ Acknowledging this serves as a useful corrective to the kind of environmental determinism that frequently creeps into popular discussions of the relationship between population growth, the environment, and conflict. ₁₅₆

Nevertheless, Levy’s suggestion that we abandon focused research on environmental (and presumably demographic) sources of violence in favor of a more general research program on civil and regional conflict forces a choice when one is not needed. Levy is correct that DES is a conjunctural variable; even when DES plays a role in conflict, it is a part of a causal chain that results in organized violence only when other intervening variables are also in play. That being said, some conjunctural variables are more important and inherently interesting than others. If a variable is not merely an additive component of a causal chain, but also interacts with and aggravates other key links in the chain, it is not simply conjunctural, it is synergistic. Demographic and environmental stress is such a synergistic variable and hence deserves careful study as a potentially powerful source of violence. Furthermore, adopting a generic research program focusing on every potential cause of civil and regional strife is not the only means of making a valuable contribution to the study and practice of international security. A research strategy focusing on DES is important because we need to know whether the intersecting trends of future population growth, accelerating economic activities, and mounting environmental degradation in the developing world are cause for concern. ₁₅₇

If empirical investigation is worthwhile, how should we go about testing the theoretical claims made in this chapter? The subjective aspects of a number of the key variables and hypothesized effects (e.g. scarcity, deprivation, elite cohesion, state legitimacy, group identity), in addition to the complex, often non-linear, causal connections, interactions, and feedback loops within and between ecological and social systems, suggests that in-depth cases studies provide the best method for testing the state-centric theory presented here. ₁₅₈

But if only a small number of cases are to be examined, analysts should chose cases that offer the greatest leverage for testing the theory. Traditionally, theories are tested by comparing a representative sample of cases with various values of the independent variable, checking for covariation with the dependent variable, and drawing causal inferences. If the processes and outcomes in the cases under study conform to the theory’s expectations, the theory is provided with provisional support; if they do not, the theory is called into question. ₁₅₉ However, making comparisons across countries with varying degrees of DES tells us little because it is admitted at the outset that DES is neither a universally necessary nor wholly sufficient cause of civil strife. Given the conjunctural and synergistic nature of DES as a causal variable, a more nuanced methodological approach is required.

Cases should be selected to fit three criteria. ₁₆₀ First, to facilitate useful case comparison, analysts should chose cases in which the value of the independent variable (DES) is high at some point in each case, but the values of the intervening variables (groupness and institutional inclusivity) vary across the cases. This allows for a comparison of the specific conditions under which DES is predicted by the state-centric theory to generate violence to those where violence is expected to be less likely (see Figure 4). Second, while DES should be high at some point in all the cases selected, the level of DES should nevertheless vary across time within each case. This guarantees some variation in the independent variable, making it possible to draw credible inferences regarding the causal significance of DES relative to case-specific alternative explanations that focus on non-demographic and non-environmental factors. Finally, as a methodological check against the possibility that the intervening variables are doing all the causal “work,” any temporal variation in the intervening variables within each case should not occur prior to changes in the value of DES.

Figure 4: Predictions of the State-Centric Theory
Degree of DES	Degree of Institutional Inclusivity	Degree of Groupness	Probability of Civil Strife
High	Low	High	Likely
High	High	High or Low	Not Likely (as a result of DES)
High	High or Low	Low	Not Likely (as a result of DES)
Low	High or Low	High or Low	Theory Indeterminate (likelihood of strife determined by variables unspecified by the theory)

 

Conclusion

Existing hypotheses linking demographic and environmental stress to violent internal conflict underemphasize and under-theorize the role of the state. The state-centric theory outlined above rejects a crude demographic or environmental determinism in favor of a thoroughly sociopolitical theory. Specifically, I have suggested that there are two pathways whereby demographic and environmental pressures interact with the state to cause civil strife. State disintegration conflicts occur when DES weakens the state, increasing the opportunities for social groups to engage in violence, and increasing their incentives to do so by creating a security dilemma. State exploitation represents a second pathway to bloodshed. These conflicts occur when population and environmental pressures provide state elites and their allies with incentives and opportunities to instigate violence that serves their own narrow self-interests.

State disintegration and state exploitation conflicts are most likely in countries characterized by low degrees of institutional inclusivity and high degrees of groupness. Inclusive institutions check violence by greasing the wheels of intergroup cooperation in the face of a weakened state and by making state exploitation more difficult, while exclusive institutions short-circuit intergroup cooperation and facilitate exploitation of the weak by the strong. A high degree of groupness encourages violence by facilitating the mobilization of conflict groups, whereas a low degree of groupness frustrates mobilization and reduces the prospects for organized violence.

This chapter has not provided a definitive test of this theory. Instead, I have sought to set the stage for an improved research program. The next challenge for scholars is to conduct additional empirical “tests.” Unlike much of the research conducted up to this point, however, analysts should pay much more attention to the political and social variables that mediate the relationship between DES, the state, and conflict. I have suggested a number of these intervening variables; there may, of course, be others. If so, they should be clearly specified and systematically incorporated into a rigorous theoretical framework. This will allow scholars to identify when and where demographic and environmental forces are most likely to lead to violence and how this complex process will likely unfold. Only then will we be able to fully understand the phenomenon and assist policymakers in addressing one of the most complex security challenges of our day.

 

___________________________________

I would like to thank Chris Ball, Richard Betts, Rachel Bronson, Nazli Choucri, Tim Crawford, Michael Desch, David Downie, Frank Gavin, Peter Gleick, Jack Goldstone, Tad Homer-Dixon, Robert Jervis, Robert Keohane, Marc Levy, Biza Repko, Aaron Seeskin, Jack Snyder, Leslie Vinjamuri-Wright, Jon Western, and all of the participants in the John M. Olin Institute National Security Seminar for helpful comments and criticisms on previous versions of this chapter. I would also like to thank the United States Institute of Peace and the Compton Foundation for financial support.

 

References

Ahlburg, Dennis A. 1994. “Population Growth and Poverty.” In Population and Development: Old Debates, New Conclusions. Robert Cassen, ed., 127-147. New Brunswick: Transaction Publishers.

André, Catherine and Jean-Philippe Platteau. 1998. “Land Relations Under Unbearable Stress: Rwanda Caught in the Malthusian Trap.” Journal of Economic Behavior and Organization 34, 1: 1-47.

Arnold. Guy. 1995. Wars in the Third World Since 1945. second edition. London: Cassell.

Aya, Rod. 1979. “Theories of Revolution Reconsidered.” Theory and Society 8, 1: 39-99.

Ayoob, Mohammed. 1996. “State Making, State Breaking, and State Failure.” In Managing Global Chaos: Sources of and Responses to International Conflict. Chester A. Crocker and Fen Osler Hampson, eds., pp. 37-53. Washington, D.C.: United States Institute of Peace Pres

Baechler, Günther. 1997. “Why Environmental Transformation Causes Violence: A Snythesis,” Environmental Change and Security Project Report. Woodrow Wilson Center. Issue 4 (Spring): 24-44.

Baechler, Günther, et al. 1996. Environmental Degradation as Causes of War: Environmental Conflicts in the Third World and Ways for their Peaceful Resolution, Volume I. Berne: Swiss Peace Foundation/Swiss Federal Institute of Technology.

Baechler, Günther, and Kurt R. Spillmann (editors). 1996 Environmental Degradation as Causes of War: Regional and Country Studies, Volume II and III. Berne: Swiss Peace Foundation/Swiss Federal Institute of Technology

Barber, Charles Victor. 1998. “Natural Resource Scarcity, Civil Violence, and the State: Indonesia’s Forests in a Period of Economic and Political Crisis.” Paper prepared for the Columbia University Environment and Security Project.

Bates, Robert H. 1981. Markets and States in Tropical Africa: The Political Basis of Agricultural Policy. Berkeley: University of California Press.

--------. 1989. Beyond the Miracle of the Market: The Political Economy of Agrarian Development in Kenya. New York: Cambridge University Press.

Berejikian, Jeffrey. 1992. “Revolutionary Collective Action and the Agent-Structure Problem.” American Political Science Review 86, 3 (September): 647-657.

Besteman, Catherine and Lee V. Cassenelli. 1996. The Struggle for Land in Southern Somalia: the War Behind the War. Boulder, CO: Westview Press.

Block, Fred. 1987. Revising State Theory. Philadelphia, PA: Temple University Press.

Böge, Volker. 1992. “Bougainville: A ‘Classical’ Environmental Conflict?” Occasional Paper, No. 3, Environment and Conflicts Project, Swiss Peace Foundation Berne, Center for Security Studies and Conflict Research (October).

Broad, Robin (with John Cavanagh). 1993. Plundering Paradise: The Struggle for the Environment in the Philippines. Berkeley, Calif: University of California Press.

Brown, Lester R. 1996. “The Acceleration of History.” In State of the World 1996. Lester R. Brown, et. al., pp. 3-20. New York: W. W. Norton & Company.

Brown, Michael E. 1996. “The Causes and Regional Dimensions of Internal Conflict.” In The International Dimensions of Internal Conflict. Michael E. Brown, ed., pp. 571-602. Cambridge, MA: The MIT Press.

Bull, Hedley. 1977. The Anarchical Society: A Study of Order in World Politics. New York: Columbia University Press.

Butts, Kent. 1996. “National Security, the Environment and DOD.” Environmental Change and Security Project Report. Woodrow Wilson Center. Issue 2 (Spring): 22-27.

Buzan, Barry, et. al. 1993. The Logic of Anarchy: From Neorealism to Structural Realism (New York: Columbia University Press.

Chase, Robert, et al. (editors) 1999. The Pivotal States: A New Framework for U.S. Policy in the Developing World. New York: W.W. Norton & Company.

Choucri, Nazli. 1992. “Environment, Development and International Assistance: Crucial Linkages.” In Resolving Third World Conflict: Challenges for a New Era. Sheryl J. Brown and Kimber M. Schraub, eds., pp. 89-118. Washington D.C.: United States Institute of Peace Press.

Chourcri, Nazli and Robert North. 1975. Nations in Conflict. San Francisco: Freeman.

Cohen, Joel E. 1995. How Many People Can the Earth Support. New York: W. W. Norton & Company.

Corr, Edwin G. 1996. “Application of Theory and Principle: The Case of El Salvador.” In Managing Contemporary Conflict: Pillars of Success. Max G. Manwaring and Wm. J. Olson, eds, 71-87. Boulder, Colo: Westview Press.

Dalby, Simon. 1999. “Threats from the South? Geopolitics, Equity, and Environmental Security.” In Contested Grounds. Daniel H. Deudney and Richard A. Matthew, eds., 155-185. Albany, NY: State University of New York Press.

David, Steven R. 1997. “Internal Wars: Causes and Cures.” World Politics 49, 4 (July): 552-576.

Davies, James C. 1962. “Toward a Theory of Revolution.” American Sociological Review 27 (February): 5-19.

de Blij, Harm J. 1995. “Geographic Factors in Ethnic Conflict in Africa.” In The Challenge of Ethnic Conflict to National and International Order in the 1990s: Geogrpahic Perspectives. Conference Report. Central Intelligence Agency, pp. 74-82. Washington D.C.: Central Intelligence Agency, RTT 95-11039 (October).

deFiueredo, Rui and Barry R. Weingast. 1997. “The Rationality of Fear: Political Opportunism and Ethnic Conflict.” unpublished ms. Stanford University.

Desch, Michael C. 1996. “War and Strong States, Peace and Weak States?” International Organization 50, 2: 237-268.

Deudney, Daniel H. 1990. "The Case Against Linking Environmental Degradation and National Security." Millennium 19, 3: 461-476.

--------. 1999. “Environmental Security: A Critique.” In Contested Grounds. Daniel H. Deudney and Richard A. Matthew, eds., 187-222. Albany, NY: State University of New York Press.

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper & Row.

Dréze, Jean and Amartya Sen. 1989. Hunger and Public Action. New York: Oxford University Press.

Eckstein, Harry. 1971. “On the Etiology of Internal Wars.” In Why Revolution? Clifford T. Paynton and Robert Blackey, eds., pp. 124-150. Rochester, VT: Schenkman Books, Inc.

Economy, Elizabeth. 1998. “The Implications of Water Scarcity for the Chinese State.”.” Paper prepared for the Columbia University Environment and Security Project.

Eggertsson, Thráinn. 1990. Economic Behavior and Institutions New York: Cambridge University Press.

Ehrlich, Paul R. and Anne H. Ehrlich. 1990. The Population Explosion. New York: Simon & Schuster.

Fearon, James D. 1993. “Ethnic War as a Commitment Problem.” unpublished ms. University of Chicago.

--------. 1995. “Rationalist Explanations for War.” International Organization 49, 3: 379-414.

Friends of the Earth. 1993. Africa’s Greenwars: The Ecological Roots of the Crisis in Somalia. Washington, D.C.: Friends of the Earth.

Fuller, Gary. 1995. “The Demographic Backdrop to Ethnic Conflict: A Geographic Overview.” In The Challenge of Ethnic Conflict to National and International Order in the 1990s: Geographic Perspectives. Conference Report. Central Intelligence Agency, pp. 151-154. Washington, D.C.: Central Intelligence Agency, RTT 95-10039 (October).

Gagnon, V.P. Jr. 1994/95. “Ethnic Nationalism and International Conflict: The Case of Serbia.” International Security 19, 3 (Winter): 130-166.

George, Alexander L. and Timothy J. McKeown. 1985. “Case Studies and Theories of Organizational Decision Making.” Advances in Information Processing in Organizations 2: 21-58.

Gizewski, Peter and Thomas Homer-Dixon. 1995. “Urban Growth and Violence: Will the Future Resemble the Past?” Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

--------. 1996. “Environmental Scarcity and Violent Conflict: the Case of Pakistan.” Occassional Paper. Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

Glaser, Charles L. 1997. “The Security Dilemma Revisited.” World Politics 50, 1 (October): 171-201.

Gleditsch, Nils Petter. 1998. “Armed Conflict and the Environment: A Critique of the Literature.” Journal of Peace Research 35, 1 (May): 381-400.

Gleick, Peter H. 1990. "Environment, Resources, and International Security and Politics." In Science and International Security: Responding to a Changing World. Eric Arnett, ed., pp. 501-523. Washington, DC: American Association for the Advancement of Science.

--------. 1993. "Water and Conflict: Fresh Water Resources and International Security." International Security 18, 1: 79-112.

--------. 1998. “Freshwater in Southern Africa as a Tool of Domestic and International Politics.” Paper prepared for the Columbia University Environment and Security Project.

Goldstone, Jack A. 1991. Revolution and Rebellion in the Early Modern World. Berkeley: University of California Press.

--------. 1996. “Debate.” Environmental Change and Security Project Report. Woodrow Wilson Center. Issue 2 (Spring):66-71.

--------. 1997. “Population Growth and Revolutionary Crises.” In Theorizing Revolutions. John Foran, ed. Routledge.

--------. forthcoming. “Demography, Domestic Conflict, and International Order.” In International Order and the Future of World Politics. T.V. Paul and John A. Hall, eds. Under review at Columbia University Press.

Goodwin, Jeff and Theda Skocpol. 1989. “Explaining Revolutions in the Contemporary Third World.” Politics and Society 17, 4: 489-510.

Gurr, Ted Robert. 1970. Why Men Rebel. Princeton: Princeton University Press.

Hall, Peter A. 1986. Governing the Economy: The Politics of State Intervention in Britain and France. New York: Oxford University Press.

Hardin, Russell. 1995. One For All. Princeton: Princeton University Press.

Hauge, Wenche and Tanja Ellingsen. 1998. “Beyond Environmental Scarcity: Causal Pathways to Conflict.” Journal of Peace Research 35, 3 (May): 299-317.

Hazarika, Sanjoy. 1993. “Bangladesh and Assam: Land Pressures, Migration and Ethnic Conflict.” Occassional Paper No. 3. Project on Environmental Change and Acute Conflict. University of Toronto, the Academy of Arts and Sciences.

Hechter, Michael. 1987. Principles of Group Solidarity. Berkeley, CA: University of California Press.

Herz, John H. 1951. Political Realism and Political Idealism: A Study in Theories and Realities. Chicago: University of Chicago Press.

Hobbes, Thomas. 1985 [1651]. Leviathan. New York: Penguin Books.

Holsti, Kalevi J. 1996. The State, War, and the State of War. New York: Cambridge University Press.

Homer-Dixon, Thomas F. 1991. “On the Threshold: Environmental Changes as Causes of Acute Conflict." International Security 16, 2: 76-116.

--------. 1994. "Environmental Scarcities and Violent Conflict: Evidence From Cases." International Security 19, 1: 4-40.

--------. 1995a. “The Ingenuity Gap: Can Poor Countries Adapt to Resource Scarcity.” Population and Development Review 21, 3 (September): 587-612.

--------. 1995b. “Strategies for Studying Causation in Complex Ecological Political Systems.” Occassional Paper. Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

--------. 1995/1996. “Correspondence: Environment and Security.” International Security 20, 3 (Winter): 189-194.

--------. 1999. Environment, Scarcity, and Violence. Princeton, NJ: Princeton University Press.

Homer-Dixon, Thomas F. and Valerie Percival. 1996. “Environmental Scarcity and Violent Conflict: Briefing Book.” Occassional Paper. Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

Hooglund, Eric J. 1982. Land and Revolution in Iran, 1960-1980. Austin, TX: University of Texas Press.

Horowitz, Donald. 1994. “Democracy in Divided Societies.” In Nationalism, Ethnic Conflict, and Democracy. Larry Diamond and Marc F. Plattner, eds., pp. 35-55. Baltimore: The Johns Hopkins University Press.

Howard, Philip and Thomas Homer-Dixon. 1996. “Environmental Scarcity and Violent Conflict: the Case of Chiapas, Mexico.” Occassional Paper. Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

Huntington, Samuel P. 1996. The Clash of Civilizations and the Remaking of World Order. New York: Simon & Schuster.

Ignatieff, Michael. 1997. The Warrior’s Honor: Ethnic War and the Modern Conscience. New York: Metropolitan Books.

International Labour Organization. 1995. World Labour Report 1995. Geneva: ILO.

Jepperson, Ronald L. 1991. “Institutions, Institutional Effects, and Institutionalism.” In The New Institutionalism in Organizational Analysis. Paul J. DiMaggio and Walter W. Powell, eds., pp. 143-163. Chicago: The University of Chicago Press.

Jacobson, Jodi L. 1989. “Environmental Refugees: A Yardstick of Habitability.” World Watch Paper 86. Washington, D.C.: World Watch Insitute.

Jervis, Robert. 1976. Perception and Misperception in International Politics. Princeton, NJ: Princeton University Press.

--------. 1978. “Cooperation Under the Security Dilemma.” World Politics 30, 2: 167-214.

Jones, Gregg R. 1989. Red Revolution: Inside the Philippine Guerrilla Movement. Boulder, Colo: Westview Press.

Kahl, Colin H. 1998a. “Population Growth, Environmental Degradation, and State-Sponsored Violence: The Case of Kenya, 1991-93.” International Security 23, 2 (Fall): 80-119.

--------. 1998b. “Chaos and Calm in East Africa: Population Growth, Land, and State-Sponsored Violence in Kenya, 1991-1993.” Paper prepared for the Columbia University Environment and Security Project.

Kaplan, Robert D. 1994. "The Coming Anarchy." The Atlantic Monthly, February, pp. 44-76.

--------. 1996. The Ends of the Earth: A Journey at the Dawn of the 21st Century. New York: Random House.

--------. 1997. “History Moving North.” The Atlantic Monthly, February, pp. 21-24, 31.

Katzenstein, Peter J. 1978. “Conclusion: Domestic Structures and Strategies of Foreign Economic Policy.” In Between Power and Plenty: Foreign Economic Policies of Advanced Industrialized States. Peter J. Katzenstein, ed., pp. 295-336. Madison, WI: The University of Wisconsin Press.

Kaufman, Chaim. 1996. “Possible and Impossible Solutions to Ethnic Civil Wars.” International Security 20, 4 (Spring): 136-175.

Kaufman, Stuart J. 1996. “Spiraling to Ethnic War: Elites, Masses, and Moscow in Moldova’s Civil War.” International Security 21, 2 (Fall): 108-138.

Kelly, Allen C. 1988. “Economic Consequences of Population Change in the Third World.” Journal of Economic Literature 26: 1685-1728.

Kelley, Allen C. and William Paul McGreevey. 1994. “Population and Development in Historical Perspective.” In Population and Development: Old Debates, New Conclusions. Robert Cassen, ed., 107-126. New Brunswick: Transaction Publishers.

Kelly, Kimberly and Thomas Homer-Dixon. 1996. “Environmental Scarcity and Violent Conflict: the Case of Gaza.” Occassional Paper. Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

Kennedy, John F. 1962. Public Papers of the Presidents of the United States: John F. Kennedy, 1962. Speech, March 13.

Kennedy, Paul. 1993. Preparing for the Twenty-First Century. New York: Random House.

Keyfitz, Nathan. 1991. “Population Growth Can Prevent the Development That Would Slow Population Growth.” In Preserving the Global Environment. Jessica Tuchman Matthews, ed., pp. 39-77. New York: W. W. Norton & Company.

King, Gary, et. al. 1994. Designing Social Inquiry. Princeton, NJ: Princeton University Press.

Krasner, Stephen D. 1978. Defending the National Interest. Princeton, NJ: Princeton University Press.

Kupchan, Charles A. 1994. The Vulnerability of Empire. Ithica, NY: Cornell University Press.

Lake, David A. and Donald Rothchild. 1996. “Containing Fear: The Origins and Management of Ethnic Conflict.” International Security 21, 2: 41-75.

Laitin, David D. 1997. “Somalia — Civil War and International Intervention.” unpublished ms. University of Chicago.

Lee, Shin-wha. 1994. “A Deadly Triangle: The Environment, Refugees, and Conflict.” Paper presented at the Annual Meeting of the American Political Science Association, New York, 1-4 September.

Leonard, H. Jeffrey. 1989. “Managing Central America’s Renewable Resources: The Path to Sustainable Economic Development.” International Environmental Affairs 1, 1 (Winter): 38-56.

Levi, Margaret. 1988. Of Revenue and Rule. Berkeley, CA: University of California Press.

Levy, Jack S. 1987. “Declining Power and the Preventive Motivation for War.” World Politics 40, 1 (October): 82-107.

--------. 1993. “The Diversionary Theory of War: A Critique.” In Handbook of War Studies. Manus I. Midlarsky, ed., pp. 259-288. Ann Arbor, MI: University of Michigan Press.

Levy, Marc A. 1995. “Is the Environment a National Security Issue.” International Security 20, 2: 35-62.

Libiszewski, Stephan. 1992. “What is an Environmental Conflict?.” Occasional Paper, No. 1, Environment and Conflicts Project, Swiss Peace Foundation Berne, Center for Security Studies and Conflict Research (July).

Lichbach, Mark I. 1989. “An Evaluation of ‘Does Economic Inequality Breed Political Conflict?’ Studies.” World Politics 41, 4: 431-471.

--------. 1994. “What Makes Rational Peasants Revolutionary? Dilemma, Paradox, and Irony in Peasant Collective Action.” World Politics 46, 3: 383-418.

Lijphardt, Arend. 1977. Democracy in Plural Societies. New Haven, CT: Yale University Press.

Lipschutz, Ronnie D. 1989. When Nations Clash: Raw Materials, Ideology, and Foreign Policy. New York: Ballinger Publishing Co.

--------. 1998. “Damming Troubled Waters: Conflict Over the Danube, 1950-2000.” Paper prepared for the Columbia University Environment and Security Project.

Lipschutz, Ronnie D. and John Holdren, 1990. "Crossing Borders: Resource Flows, the Global Environment, and International Security." Bulletin of Peace Proposals 21, 2: 121-133.

Lowi, Miriam R. 1993a. "Bridging the Divide: Transboundary Resource Disputes and the Case of West Bank Water." International Security 18, 1: 113-138.

--------. 1993b. Water and Power: The Politics of a Scarce Resource in the Jordon River Basin. Cambridge: Cambridge University Press.

Mann, Michael. 1988. States, War and Capitalism. Cambridge, MA: Blackwell.

Marshall, Ray. 1995. “The Global Jobs Crisis.” Foreign Policy 100 (Fall): 50-68.

Matthews, Jessica Tuchman. 1989. "Redefining Security." Foreign Affairs 68, 2: 162-177.

McKibben, Bill. 1998. “A Special Moment in History.” The Atlantic Monthly, May, pp. 55-78.

Merrick, Thomas W. 1994. “Population Dynamics in Developing Countries.” In Population and Development: Old Debates and New Conclusions. Robert Cassen and Contributers, pp. 79-106. Washington, D.C.: Overseas Development Council.

Migdal, Joel S. 1988. Strong States and Weak Societies: State-Society Relations and State Capabilities in the Third World. Princeton, NJ: Princeton University Press.

Moffett, George D.1994a. Critical Masses: The Global Population Challenge. New York: Penguin Books.

--------. 1994b. “The Population Question Revisited.” The Wilson Quarterly 18, 3 (Summer 1994): 54-79.

Muller, Edward N. and Karl-Dieter Opp. 1986. “Rational Choice and Rebellious Collective Action.” American Political Science Review 80, 2 (June): 471-487.

Myers, Norman. 1993. Ultimate Security: The Environmental Basis of Political Stability. New York: W. W. Norton & Company.

A National Strategy of Engagement and Enlargement. 1996. Washington, D.C.: The White House.

Nettl, J.P. 1968. “The State as a Conceptual Variable.” World Politics 20, 4: 559-592.

Nordlinger, Eric. 1981. On the Autonomy of the Democratic State. Cambridge, MA: Harvard University Press.

Ohlsson, Leif. 1999. Environment, Scarcity, and Conflict: A Study of Malthusian Concerns. Ph.D. Thesis. Department of Peace and Development Research. Göteborg University, Sweden.

Olcott, Martha Brill. 1998. “The Caspian’s False Promise.” Foreign Policy111 (Summer): 95-113.

Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, MA: Harvard University Press.

--------. 1993. “Dictatorship, Democracy, and Development.” American Political Science Review 87, 3: 567-576.

Orme, John. 1997/98. “The Utility of Force in a World of Scarcity.” International Security 22, 3 (Winter): 138-167.

Oye, Kenneth A. 1986. “Explaining Cooperation Under Anarchy.” In Cooperation Under Anarchy. Kenneth A. Oye, ed., pp. 1-24. Princeton, NJ: Princeton University Press.

Percival, Valerie and Thomas Homer-Dixon. 1995a. “Environmental Scarcity and Violent Conflict: the Case of Rwanda.” Occassional Paper. Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

--------. 1995b. “Environmental Scarcity and Violent Conflict: the Case of South Africa.” Occassional Paper. Project on Environment, Population, and Security. Washington, D.C: University of Toronto, the Academy of Arts and Sciences.

Pogi, Gianfranco. 1978. The Development of the Modern State: A Sociological Introduction. Stanford, CA: Stanford University Press.

Popkin, Samuel L. 1979. The Rational Peasant: The Political Economy of Rural Society in Vietnam. Berkeley, CA: University of California Press.

Population and Development Review 1998. “The 1998 Revisions of the United Nations Population Projections,” Population and Development Review 24, 4 (December): 891-895.

Population Reference Bureau. 1996. World Population Data Sheet. Washington, D.C.: Population Reference Bureau.

Posen, Barry R. 1993a. “The Security Dilemma and Ethnic Conflict.” In Ethnic Conflict and International Security. Michael E. Brown, ed., pp. 103-124. Princeton: Princeton University Press.

--------. 1993b. “Nationalism, the Mass Army, and Military Power.” International Security 18, 2: 80-124.

Postel, Sandra. 1993. "The Politics of Water." Worldwatch 6, 4 (July/August): 10-18.

Prunier, Gérard. 1995. The Rwanda Crisis: History of a Genocide. New York: Columbia.University Press.

Renner, Michael. 1996. Fighting For Survivial: Environmental Decline, Social Conflict, and the New Age of Insecurity. New York: W. W. Norton & Company.

--------. 1997. “Transforming Security.” In State of the World 1997. Lester Brown et. al., pp. 115-131. New York: W. W. Norton & Company.

Richards, Paul. 1995. “Rebellion in Liberia and Sierra Leone: A Crisis of Youth?” In Conflict in Africa. Oliver Furley, ed., pp. 134-170. London: I.B. Tauris Publishers.

Rosenberg, Morris. 1981. “The Self-Concept: Social Product and Social Force.” In Social Psychology: Sociological Perspectives. Morris Rosenberg and Ralph H. Turner, eds., pp. 593-624. New York: Basic Books.

Ross, Marc Howard. 1993. The Culture of Conflict: Interpretations and Interests in Comparative Perspective. New Haven, CT: Yale University Press.

Sandler, Todd. 1992. Collective Action: Theory and Applications. Ann Arbor, MI: University of Michigan Press.

Sen, Amartya. 1983. “Development: Which Way Now?” Economic Journal 93: 745-762.

--------. 1984. Resources, Values, and Development. Cambridge, MA: Harvard University Press.

--------. 1994. “Freedom and Needs,” New Republic, January 10 & 17, pp. 31-38.

--------. 1996. “Wars and Famines: On Divisions and Incentives.” Disarmament 19, 3: 18-37.

Sheriff, Muzufar. 1966. Group Conflict and Cooperation: Their Social Psychology. London: Routledge and Kegan Paul.

Simon, Julian L. 1981. The Ultimate Resource. Princeton: Princeton University Press.

Skocpol, Theda. 1979. States and Social Revolutions. Cambridge: Cambridge University Press.

Smil, Vaclav. 1993. China’s Environmental Crisis: An Inquiry into the Limits of National Development. New York: M.E. Sharpe.

--------. 1997. “China’s Environment and Security: Simple Myths and Complex Realities.” SAIS Review (Winter/Spring): 107126.

Snyder, Jack. 1991. Myths of Empire: Domestic Politics and International Ambition. Ithica, NY: Cornell University Press.

--------. 1993. “Nationalism and the Crisis of the Post-Soviet State.” In Ethnic Conflict and International Security. Michael E. Brown, ed., pp. 79-102. Princeton: Princeton University Press.

Snyder, Jack and Karen Ballentine. 1996. “Nationalism and the Marketplace of Ideas.” International Security 21, 2 (Fall): 5-40.

Snyder, Jack and Robert Jervis. 1997. “Civil War and the Security Dilemma.” unpublished ms. Columbia University.

Starr, Joyce. 1991. "Water Wars." Foreign Policy 82: 17-30.

Stinchcombe, Arthur L. 1968. Constructing Social Theories. Chicago, IL: University of Chicago Press.

Suhrke, Astri. 1996. “Environmental Change, Migration, and Conflict.” In Managing Global Chaos: Sources of and Responses to International Conflict. Chester A. Crocker and Fen Osler Hampson, eds., pp. 113-140. Washington, D.C.: United States Institute of Peace Press.

Suliman, Mohamed. 1993. “Civil War in the Sudan: From Ethnic to Ecological Conflict.” The Ecologist 23, 3 (May/June): 104-109.

Tarrow, Sidney. 1994. Power in Movement: Social Movements, Collective Action and Politics. New York: Cambridge University Press.

Tajfel, Henri and John C. Turner. 1986. “The Social Identity Theory of Intergroup Behavior.” In Psychology of Intergroup Relations. William G. Austin and Stephen Worchel, eds., pp. 7-24. Chicago: Nelson-Hall.

Taylor, Michael. 1982. Community, Anarchy, and Liberty. Cambridge: Cambridge University Press.

--------. 1988. “Rationality and Revolutionary Collective Action.” In Rationality and Revolution. Michael Taylor, ed., pp. 63-97. Cambridge: Cambridge University Press.

Tilly, Charles. 1975. “Food Supply and Public Order in Modern Europe.” In The Formation of National States in Western Europe. Charles Tilly, ed., pp. 380-455. Princeton, NJ: Princeton University Press.

--------. 1978. From Mobilization to Revolution. Reading, MA: Addison-Wesley.

--------. 1993. European Revolutions, 1492-1992. Cambridge, MA: Blackwell.

Tinker, Jon and Lloyd Timberlake. 1985. “Environment and Conflict.” World Press Review, April, pp. 29-31

Turner, John C. 1987. Rediscovering the Social Group. New York: Basil Blackwell, Inc.

Walter, Barbara F. 1997. “Exiting From War: Cooperating Under Even the Most Difficult Conditions.” unpublished ms. University of California-San Diego (February).

Waltz, Kenneth N. 1979. Theory of International Politics. New York: McGraw-Hill, Inc.

Weber, Max. 1964. The Theory of Social and Economic Organization, Talcott Parsons, ed. New York: The Free Press.

Weingast, Barry R. 1995. “Constructing Trust: The Political and Economic Roots of Ethnic and Regional Conflict.” unpublished ms. Stanford University.

--------. 1996. “The Fundamental Political and Economic Puzzles of Ethnic and Regional Violence.” unpublished ms. Stanford University (May).

Westing, Arthur. ed., 1986. Global Resources and International Conflict: Environmental factors in Strategic Policy and Action. Oxford - New York: Oxford University Press.

Williamson, Oliver. 1975. Markets and Hierarchies. New York: The Fress Press.

Woodrow Wilson Center. 1997. “Official Statements and Documents.” Environmental Change and Security Project Report. Woodrow Wilson Center. Issue 4(Spring): 73-79.

World Resources Institute. 1994. World Resources 1994-1995. New York: Oxford University Press.

World Resources Institute. 1996. World Resources 1996-1997. New York: Oxford University Press.

World Resources Institute. 1998. World Resources 1998-1999. New York: Oxford University Press.

Zartman, I. William. 1995. “Introduction: Posing the Problem of State Collapse.” In Collapsed States: The Disintegration and Restoration of Legitimate Authority. I. William Zartman, ed., pp. 1-14. Boulder, CO: Lynne Rienner Publishers.

Zucker, Lynn G. 1991. “The Role of Institutionalization in Cultural Persistence. ” In The New Institutionalism in Organizational Analysis. Paul J. DiMaggio and Walter W. Powell, eds., pp. 83-107. Chicago: The University of Chicago Press.

 

---

Endnotes

Note 1: A National Security Strategy of Engagement and Enlargement 1996, preface.  Back.

Note 2: Quoted in Woodrow Wilson Center 1997, 73.  Back.

Note 3: Quoted in Dalby 1999, 162. NATO’s Strategic Concept now includes environmental components, and members of the European Union seem to believe that demographic and environmental forces in East-Central Europe, the Middle East, and Africa pose threats to European security. See Butts 1996.  Back.

Note 4: Kaplan 1994, 58 (emphasis in original).  Back.

Note 5: A detailed discussion of this literature is provided below.  Back.

Note 6: Ayoob 1996, 37. Kalevi Holsti has recently calculated that only 18 percent of the the 164 wars since 1945 have been between states. In contrast, 77 percent of the wars occurred within states. Moreover, of the remaining 5 percent of wars stemming from external armed interventions, the vast majority were actually the result of internal violence and rebellion. See Holsti 1996, 21-25.  Back.

Note 7: Population and Development Review 1998; McKibben 1998, 56; Merrick 1994; World Resources Institute 1994, chap. 2; World Resources Institute 1998, 141.  Back.

Note 8: Mckibben 1998, 63.  Back.

Note 9: See L. Brown 1996; and Moffett 1994b.  Back.

Note 10: For a discussion of existing evidence, see Baechler et al. 1996; Baechler and Spillman 1996; Chase et al. 1999; Homer-Dixon 1994; and Homer-Dixon 1999. For critical reviews of this research see Deudney 1999; Gleditsch 1998; and Levy 1995.  Back.

Note 11: The term “simple scarcity” originates from the discussion in Homer-Dixon 1991, 106-108. Proponents of this hypothesis include Choucri and North 1975; Gleick 1990; Gleick 1993; Lowi 1993b; Orme 1997/98; Postel 1993; Starr 1991; and Westing 1986.  Back.

Note 12: See Chourcri 1992; Homer-Dixon 1991, 108-109; Homer-Dixon 1994, 20-23; Jacobson 1989; and Lee 1994.  Back.

Note 13: Homer-Dixon 1994; Lipschutz and Holdren 1990; and Lipschutz 1989.  Back.

Note 14: Homer-Dixon 1999, 138-139.  Back.

Note 15: Deudney 1990, 469-474; and Deudney 1999, 202-213.  Back.

Note 16: A similar point is made by Homer-Dixon 1999, 139.  Back.

Note 17: See, for example, Baechler 1997, 30-31 and 38; Gleick 1998; and Lipschutz 1998.  Back.

Note 18: Hazarika 1993  Back.

Note 19: Suhrke 1996.  Back.

Note 20: This hypothesis is grounded in the theoretical work of Davies 1962 and Gurr 1970. For a discussion, see Homer-Dixon 1991, 104-105 and 109-111. The difference between absolute and relative deprivation is discussed in greater detail below.  Back.

Note 21: Myers 1993, 22; and Mathews 1989, 166, 168. See also Ehrlich and Ehrlich 1990, 178-179.  Back.

Note 22: Aya 1979, 41. See also, Eckstein 1971, 128-129; and Skocpol 1979, 114-115.  Back.

Note 23: See Berejikian 1992, 649-651 and 655; Lichbach 1994; and Muller and Opp 1986, 471-473. This line of reasoning builds on the classic work of Olson 1965.  Back.

Note 24: See the discussion of state disintegration conflicts below.  Back.

Note 25: See Barber 1998.  Back.

Note 26: See Goldstone 1991; Goldstone 1997; Homer-Dixon 1991; Homer-Dixon 1994; Homer-Dixon 1999. A similar, but much less rigorous, argument is made by Kaplan 1994; Kaplan 1996; and Kennedy 1993.  Back.

Note 27: Goldstone 1996, 70 (emphasis added).  Back.

Note 28: Homer-Dixon 1999, 178.  Back.

Note 29: Baechler 1997, 32 (emphasis in original).  Back.

Note 30: Homer-Dixon 1991, 111; Homer-Dixon 1994, 26-27; Homer-Dixon 1999, 178; and Perival and Homer-Dixon 1995b, 7. See also, Baechler 1997, 32-35; Goldstone 1997, ms. 111-115.  Back.

Note 31: For a similar criticism, see Gleditsch 1998, 389.  Back.

Note 32: See Moffett 1994a, chapter 2. Homer-Dixon calls this dynamic “ecological marginalization.” For a discussion see Homer-Dixon 1994, 15-16; and Homer-Dixon 1999, 77-79.  Back.

Note 33: This definition combines elements found in Weber 1964, 156; Mann 1988, 4; Migdal 1988, 19; Nordlinger 1981, 11; and Pogi 1978, 1.  Back.

Note 34: This definition does not assume that all state preferences are autonomously generated, in the sense of arising independently from societal influence. I recognize that the degree of state autonomy vis-a-vis the preferences of societal actors is an empirical question. Likewise, how “unitary” an actor the state is will vary depending on the degree of state cohesion. As I explain below, the degree of state cohesion is an important component determining state “strength.”  Back.

Note 35: See Block 1987, 84; and Hall 1986, chapter 1.  Back.

Note 36: It should be made clear that DES is not synonymous with renewable resource scarcity. Rather, the former is a potential cause of the latter. If markets, governments, and other social institutions adjust to population and environmental pressures, societies will not necessarily experience scarcity as a result of DES. For an excellent discussion of the importance of ingenuity in adapting to DES, see Homer-Dixon, 1995a.  Back.

Note 37: Obviously, like all resources, renewable resources are “goods,” in the sense of being tangible commodities produced, consumed, and otherwise utilized by individuals. Renewable resources also provide “services.” By acting as sinks and buffers for wastes and other by-products of human activities, renewable resources provide a valuable service by limiting the harmful effects of human activities on society. For example, the ability of land and ocean plants to serve as carbon sinks helps limit the destabilizing effect industrial emissions of carbon dioxide have on the climate, which in turn limits the potentially harmful economic and societal effects of climate change. See Libiszewski 1992, 3-6.  Back.

Note 38: When operationalizing renewable resource scarcity in any given context, two caveats must be kept in mind. First, scarcity is relative and subjective, not absolute and objective. Scarcity is experienced by human beings. As such, the degree of scarcity is determined by what individuals or groups need, or feel they need, to survive or maintain their standard of living. Consequently, there is no absolute and objective level below which a reduction of the stock or flow of renewable resources will cause all individuals to experience “scarcity.” Second, measurements of scarcity need to be disaggregated. Because aggregate national measurements of the quantity and quality of renewable resources often mask the degree of deprivation experienced by individuals and groups at lower levels of analysis, scarcity must be analyzed at the regional, local, and group level. See Dréze and Sen 1989, 30; and Homer-Dixon 1994, 9.  Back.

Note 39: Dréze and Sen 1989, 29.  Back.

Note 40: Homer-Dixon 1994, 9. This tripartite discussion of the sources of renewable resource scarcity exactly paralells Homer-Dixon’s definition and discussion of “environmental scarcity.” I chose to use the term “renewable resource scarcity” rather than environmental scarcity primarily to avoid analytical confusion regarding my independent variable. In Homer-Dixon’s analyses, environmental scarcity is the key independent variable, whereas in my analysis it is a first order effect. Therefore, using the term environmental scarcity would likely produce unnecessary confusion on behalf of some readers regarding the nature of my causal variable, which is broader and not entirely “environmental.”  Back.

Note 41: Cohen 1995, 367  Back.

Note 42: Goldstone forthcoming, 5-6.  Back.

Note 43: World Resources Institute 1994, 35-36.  Back.

Note 44: Population Reference Bureau 1996; and Marshall 1995, 51.  Back.

Note 45: Marshall 1995, 50.  Back.

Note 46: International Labour Organization 1995; and World Resources Institute 1996, 11-12.  Back.

Note 47: See the discussion in Goldstone 1991, 33.  Back.

Note 48: Homer-Dixon 1999, 74. See also, Goldstone forthcoming, 10-11; and Homer-Dixon 1994, 11-14. For the general logic of this argument, see Levi 1988, 17-23.  Back.

Note 49: Homer-Dixon calls this particular form of rent-seeking behavior “resource capture.” See the discussion in Homer-Dixon 1994, 11-14; and Homer-Dixon 1999, 73-77.  Back.

Note 50: Kaplan 1997, 21; and Population Reference Bureau 1996.  Back.

Note 51: Population Reference Bureau 1996.  Back.

Note 52: Kelly and Homer-Dixon 1996, 8.  Back.

Note 53: Total and percentages were calculated from the data in Table 3 in Lowi 1993, 120  Back.

Note 54: World Resources Institute 1996, 11.  Back.

Note 55: Ibid, 10-12; and Gizewski and Homer-Dixon 1995.  Back.

Note 56: World Resources Institute 1996, 8-9.  Back.

Note 57: See Gurr 1970, especially chapters 1 and 2. For a recent discussion, see Lichbach 1989.  Back.

Note 58: Patrick E. Tyler, “In China’s Far West, Tensions with Ethnic Muslims Boil Over in Riots and Bombings.” New York Times, 28 February 1997, A8.  Back.

Note 59: See Kelly and Homer-Dixon 1996; and Lowi 1993. These authors tend to treat state exploitation as part of the context of broader structural inequality. A more powerful argument can and should be made. Supply- and demand-induced scarcity in this case actually aggravates this structural inequality. Furthermore, the evidence provided by the authors themselves suggests that state exploitation of this scarcity worsens the Israeli-Palestinian conflict.  Back.

Note 60: World Resources Institute 1996, chapter 1.  Back.

Note 61: Ibid., 4.  Back.

Note 62: Olcott 1998, 107-108.  Back.

Note 63: Fuller 1995, 152-154.  Back.

Note 64: See Huntington 1996, 259-265.  Back.

Note 65: For the Philippines, see Broad 1993; Homer-Dixon 1999, 152-155; Jones 1989; and Myers 1993, chap. 6. For El Salvador, see Myers 1993, chap. 8.  Back.

Note 66: This conceptualization of state strength and weakness is similar to the one put forth by Desch 1996, 240-241. It differs from those who, like J.P. Nettl, Stephen D. Krasner, and Peter J. Katzenstein, define “strong” and “weak” states in terms of how autonomous the state is from societal actors. See Nettl 1968; Krasner 1978; and Katzenstein 1978.  Back.

Note 67: Adapted from Buzan et. al. 1993, 35-36.  Back.

Note 68: See Zartman 1995.  Back.

Note 69: See Tilly 1993, chapter 1; and Tilly 1978.  Back.

Note 70: See Homer-Dixon 1994, 25; Goldstone 1991, chap. 1; and Goldstone 1997, ms. 108.  Back.

Note 71: Homer-Dixon 1994, 25.  Back.

Note 72: Homer-Dixon 1994, 26. The degree to which individuals blame the state may depend on how involved the state traditionally is in providing goods and services to the population. See Tilly 1975; and Holsti 1996, 109.  Back.

Note 73: Keyfitz 1991, 66. See also Gizewski and Homer-Dixon 1995.  Back.

Note 74: Homer-Dixon 1994, 25.  Back.

Note 75: Homer-Dixon 1994, 25.  Back.

Note 76: Kelly 1988, 1715.  Back.

Note 77: Cohen 1995, 352; and Kelly and McGreevey 1994.  Back.

Note 78: Ahlburg 1994, 136-137.  Back.

Note 79: Homer-Dixon 1991, 94-97.  Back.

Note 80: Leonard 1989. For other excellent examples, see Gizewski and Homer-Dixon 1996; and Smil 1993.  Back.

Note 81: See Bates 1989, 143-145.  Back.

Note 82: For example, this dynamic appears to be driving increased rates of deforestation in Indonesia. Barber 1998, 16-20.  Back.

Note 83: See Economy 1998.  Back.

Note 84: Goldstone, 1997, 112.  Back.

Note 85: See Percival and Homer-Dixon 1995b.  Back.

Note 86: Goldstone 1997, ms. 110.  Back.

Note 87: Economy 1998, 30.  Back.

Note 88: Homer-Dixon 1994, 25.  Back.

Note 89: For examples, see Kaplan 1994; and Kaplan 1996.  Back.

Note 90: For the bottom-up argument, see Fearon 1993; Hardin 1995; Lake and Rothchild 1996, Posen 1993a; Skocpol 1979; Tilly 1978; and Weingast 1996. For the top-down argument, see deFiueredo and Weingast 1997; Gagnon 1994/94; S. Kaufman 1996. For a general survey of recent theories of internal war, see David 1997.  Back.

Note 91: Goldstone 1991; Goldstone 1997; and Homer-Dixon 1994.  Back.

Note 92: See Eckstein 1971, 141-145; Skocpol 1979; and Tilly 1978. For a general discussion of the importance of opportunity structures and collective action, see Tarrow 1994.  Back.

Note 93: See Holsti 1996, chapter 5; and Migdal 1988, 32-33.  Back.

Note 94: See Fearon 1995; and Fearon and Laitin 1996.  Back.

Note 95: See, for example, Jervis 1976, chapter 3; and Waltz 1979. For a recent discussion of the security dilemma, see Glaser 1997.  Back.

Note 96: Posen 1993a; and Snyder and Jervis 1997. Ironically, the “logic” of anarchy long at the heart of Realist theories of conflict between states, provides even greater promise for understanding conflict within states. This is due to the simple fact that individuals and social groups are more vulnerable than nation-states to annihilation. See Hobbes 1968 [1651]; and Bull 1977, 46-51.  Back.

Note 97: See Jervis 1978.  Back.

Note 98: See Walter 1996; and C. Kaufman 1996, 147-148.  Back.

Note 99: See Jervis 1976, 75-76.  Back.

Note 100: Herz 1951, 3.  Back.

Note 101: See Lake and Rothchild 1996, 47-48.  Back.

Note 102: Groups may misrepresent their intentions to 1) bluff; 2) hide their aggressive intentions; or 3) maintain advantages on the battlefield. See Fearon 1995, 391-401; and Lake and Rothchild 1996, 46-47. On attempts to misrepresent the intentions of others, see Gagnon 1994/95; and S. Kaufman 1996, 109 and 116-118  Back.

Note 103: Jervis 1978, 168.  Back.

Note 104: Hardin 1995, 144.  Back.

Note 105: See Weingast 1995; and Weingast 1996. This argument is also implicit throughout Jervis 1978.  Back.

Note 106: See Suliman 1993.  Back.

Note 107: See Fearon 1993.  Back.

Note 108: See Böge 1992.  Back.

Note 109: See Howard and Homer-Dixon 1996.  Back.

Note 110: World Resources Institute 1996, 190 and 192.  Back.

Note 111: Richards 1995.  Back.

Note 112: Bestemen and Cassenelii 1996; and Friends of the Earth 1993.  Back.

Note 113: Laitin 1997, 5. It should be noted that Laitin concludes that this explanation is not wholly satisfactory primarily because it is not “generalizable” across cases. However, if the dynamics outlined in this chapter have merit, the explanation is more generalizable than Laitin assumes. Furthermore, the notion that DES exacerbated conflicts of interest between Somali clans is perfectly compatible with Laitin’s general model of inter-clan competition in the wake of state collapse.  Back.

Note 114: de Blij 1995, 76.  Back.

Note 115: Ibid., 193; and Hooglund 1982, 118-119 and 168-169 note 44..  Back.

Note 116: Hooglund 1982; and Tinker and Timberlake 1985, 29-30. See also Kaplan 1996, chapter 11. [117] Gagnon 1994/95, 135-136.  Back.

Note 117: For a general discussion of diversionary tactics as a cause of ethnic and civil war, see Brown 1996, 585-586; Gagnon 1994/94, 134-140; Levy 1993; and S. Kaufman 1996, 117.  Back.

Note 118: For a general discussion of diversionary tactics as a cause of ethnic and civil war, see Brown 1996, 585-586; Gagnon 1994/94, 134-140; Levy 1993; and S. Kaufman 1996, 117.  Back.

Note 119: See Snyder and Balentine 1996.  Back.

Note 120: Percival and Homer-Dixon 1995, 2-3. See also Ohlsson 1999, 87-96; and Renner 1996, 114-122.  Back.

Note 121: Percival and Homer-Dixon 1995, 6-7.  Back.

Note 122: Ohlsson 1999, 107-108.  Back.

Note 123: Ohlsson 1999, 99-108. Ohlsson primarily bases this conclusion on research conducted by André and Platteau 1998; and Prunier 1996. See also Renner 1996, 117-118; and Renner 1997, 124.  Back.

Note 124: Prunier 1995, 4 and 248, quoted in Ohlsson 1999, 99. In contrast, Percival and Homer-Dixon conclude that DES was not a major source of the civil strife in Rwanda. They base this claim on two arguments: (1) There is no evidence that a large number of Rwandans participated in the genocide; and (2) Hutus in southern Rwanda, where land pressures were most accute, initially abstained from killing and only joined in when they were coerced by northern militias and local authorities (Percival and Homer-Dixon 1995, 9-10). But, as Ohlsson demonstrates, there is evidence of mass participation, even if it was often coerced. The reason the south initially stayed calm was the unique restraining influence of Tutsi opposition leaders in positions of political power there. Once these leaders were purged, the killing began (See Ohlsson 1999, 112-118). Moreover, Percival and Homer-Dixon fail to find a causal connection in part because they chose to focus on bottom-up dynamics (i.e., rising social grievances as a cause of spontaneous violence) rather than top-down state exploitation of these grievances. To the extent that state exploitation is considered by Percival and Homer-Dixon, it is only with regard to the role DES played in increasing the salience of ethnic identity, which they conclude was negligible (Percival and Homer-Dixon 1995, 11-12). What this misses is the important role population growth and land scarcity played in providing incentives, in the form of private gains from seizing victims’ land and livestock, that elites could play on to stir up ethnic violence.  Back.

Note 125: See Fearon 1993; Lake and Rothchild 1996; and Weingast 1996.  Back.

Note 126: Hardin 1995, 155.  Back.

Note 127: See deFigueirdo and Weingast 1997, 3; Gagnon 1994/94, 138; and Synder 1991, 41-43.  Back.

Note 128: Lake and Rothchild 1996, 58-61; and Lijphardt 1977, chapter 2.  Back.

Note 129: While most inclusive states are democratic, and most exclusionary states are authoritarian, democracy is neither necessary nor sufficient for inclusivity. Democracies vary in the extent to which they include a broad spectrum of societal groups in the decision-making process, and some authoritarian states allow for minority participation and influence in executive decision-making processes. Nevertheless, it is undoubtedly the case that democracies, on average, tend to be more inclusive than authoritarian states. See Lake and Rothchild 1996, 59; Olson 1993, 579; and Horowitz 1994.  Back.

Note 130: Kennedy 1962.  Back.

Note 131: For discussions of transaction costs, see Eggertsson 1990, chapter 1; and Williamson 1975.  Back.

Note 132: See Oye 1986, 13-18.  Back.

Note 133: On the institutionalization of such norms, see Jepperson 1991; Sen 1994, 36; and Zucker 1991.  Back.

Note 134: See Lake and Rothchild 1996, 57-58.  Back.

Note 135: Barber 1998, 43.  Back.

Note 136: Arnold 1995, 485-487; and Holsti 1996, 93-94.  Back.

Note 137: For the Philippines, see Homer-Dixon 1999, 155; and Jones 1989, chap. 24. For El Salvador, see Corr 1996.  Back.

Note 138: This builds on arguments made by Downs 1957, 96-108; Olson 1993, 567-576; Sen 1983: Sen 1984; Sen 1994; and Sen 1996.  Back.

Note 139: Fearon 1993, 16-20; and Weingast 1996, 14.  Back.

Note 140: See Howard H. French, “In One Poor African Nation, Democracy Thrives.” New York Times, 16 October 1996, A3.  Back.

Note 141: For excellent discussions of individual identity and reference group selection, see Rosenberg 1981; and Turner 1987.  Back.

Note 142: See Muller and Opp 1986.  Back.

Note 143: See Taylor 1982, chapter 1.  Back.

Note 144: Taylor 1988, 68-69. See also, Hechter 1987, 50.  Back.

Note 145: See Snyder and Ballentine 1996.  Back.

Note 146: Convincing evidence for cross-cutting ties limiting intergroup violence is presented by Ross 1993, chapters 3 and 9.  Back.

Note 147: See Kahl 1998a; and Kahl 1998b.  Back.

Note 148: Homer-Dixon 1991, 113.  Back.

Note 149: This builds on an argument made by Snyder 1993, 82 and 86. See also, Homer-Dixon and Percival, 1996, 8.  Back.

Note 150: Ignatieff 1997, 45. See also, ibid., 7.  Back.

Note 151: Quoted in Moffett 1994a, 48. See also ibid., chapter 1 passim.  Back.

Note 152: See the discussion of “privileged” groups in Olson 1965, 48-50; and Sandler 1992, 9-10, 24, and 35. For a empirical assessment of the use of selective incentives in rebellious collective action, see Popkin 1979.  Back.

Note 153: Horowitz 1994.  Back.

Note 154: Levy 1995, 58 (emphasis mine).  Back.

Note 155: In a recent statistical study of the 1980-92 period, for example, Hauge and Ellingsen found that demographic and environmental pressures were positively correlated with civil strife, but that poverty, income inequality, and the degree of democracy were more important contributors to large-scale violence. Their study also found that a combination of all these variables significantly increased the probability of conflict. These findings are broadly compatible with the theoretical argument advanced here. See Hauge and Ellingsen 1998.  Back.

Note 156: For additional discussion, see Homer-Dixon 1999, 82-84; and Smil 1997, 109-110.  Back.

Note 157: This last point is also made by Homer-Dixon 1995/1996, 191.  Back.

Note 158: These methodological issues are discussed in depth in George and Mckeown 1985; and Homer-Dixon 1995b.  Back.

Note 159: See King et al. 1994; and Stinchcombe 1968.  Back.

Note 160: In my usage, “cases” refers to analyses of discrete historical events. Thus, a single country could potentially produce several “cases,” depending on the length of time and geographic variation within the country involved in the study.  Back.