No food wars.

Despite this work, there is a disjunction between scholarly research on environmental conflict and often-heard assertions in journalistic accounts and policy papers. The main purpose of this article is to close this gap by engaging in discussion with the policy community. While many in academia have been cautious about their claims, policy activists – seeking to underscore the urgent need to take action on climate change – continue to make dire predictions about looming ‘resource wars’, often basing their arguments on speculation or shoddy research. In this article, I also encourage researchers to incorporate theoretically informed political variables into their models of environmental conflict and call on the academic community to better communicate their findings to general audiences.

Let me be clear: climate change certainly presents a major challenge to economies, social relations, and livelihoods, and action must be taken quickly to attenuate its negative impacts and to adapt to this reality. That said, claims of environmental determinism leading seamlessly from climate change to open warfare are suspect. The overly structural logic linking climate change to armed conflict ignores human agency, ingenuity, the potential for technological innovation, and the vital role of political institutions in managing conflict (or failing to do so). Additionally, ignoring the role that governments play in managing and redistributing resources, as well as mediating conflict, leads to incorrect predictions and policy prescriptions, and allows decisionmakers to shift blame for civil wars and grave human rights violations. If scholars are serious about informing policy debates (see Mack, 2002), then communicating more effectively with the policy community is critical.

On a fundamental level, if we acknowledge that actors faced with environmental stress make decisions strategically, then we can see that violence is generally a poor response to resource scarcity, given the alternatives. Barring the defeat, subjugation, or extermination of the other party, armed conflict by itself does nothing to resolve the underlying incompatibility over the distribution of resources. Violence is typically used as a strategy used to influence outcomes during negotiations, whether in a domestic or international setting (Filson & Werner, 2002; Wagner, 2000); eventually, actors must come to the bargaining table. Moreover, there is good reason to think that civil wars are extremely disruptive to the natural environment, leaving fewer resources than there were to begin with. Warfare is, therefore, an inefficient and costly way to resolve conflicts over resources (Fearon, 1995). Failure to find a suitable bargain and forgo fighting stems from failures in the political process, not from the absolute level of resources. Thus, while environmental degradation is certainly not a necessary condition for armed conflict, neither is it a sufficient one, since states play a key role in containing or aggravating violence.

In the remainder of this article, I will assess the link between climate change and armed conflict. In doing so, I will point out the pitfalls of what I term ‘deterministic’ approaches, which place the emphasis on structural features of the environment rather than social processes and the decisionmaking capacity of actors. This environmental determinism – though becoming a ‘conventional wisdom’ in some policy and activist circles – is misleading and possibly counterproductive. Finally, I will suggest ways in which scholars can improve empirical research as well as communicate with the policy community.

The Environment, Conflict, and the State

The deterministic view argues that climate change and concomitant resource scarcities will lead directly to violent conflict or, at least, to a significantly higher risk of violence. This reasoning is most often heard in journalistic accounts, statements by activists, and policy papers by government agencies and IGOs.3 This view downplays the role of governments, political institutions, and social actors in mitigating resource pressures and conflict. Most people in the academic community, however, reject simple determinism and have made more nuanced arguments (e.g. Barnett, 2001; Diamond, 2005). Nonetheless, scholars making strong claims about the primacy of environmental conditions continue to argue that human ingenuity and the redistributive functions of governments are likely to be overwhelmed by environmental stress (Homer-Dixon, 1999, 2000). Below, I outline the importance of governance to regulate resources, manage the environment, and contain conflict. These factors are often missing or understated in popular accounts of the environment–conflict connection.

A few caveats are in order here. It is important to note, again, that the most severe effects of climate change are likely to be felt in the future, and the future is inherently uncertain.4 While fundamental shifts in the environment are not inconceivable, our best bet for predicting what is to come is to look at what has transpired in the past. Since it is frequently argued that climate change will lead to resource scarcities and exacerbate inequality, it is possible to draw upon past evidence regarding these factors to develop a sense of how conflicts might unfold given changes in the Earth’s atmosphere. Additionally, I do not take issue with the claim that climate change will present considerable challenges for human societies and ecosystems more generally. Humanitarian crises stemming, in part, from climate change have the potential to be severe, and steps must be taken quickly to attenuate such contingencies. Rather, my purpose here is to underscore the point that environmental processes, by themselves, cannot explain why, where, and when fighting will occur; rather, the interaction between environmental and political systems is critical for understanding organized armed violence.

First, the deterministic view has poor predictive power as to where and when conflicts will break out. For every potential example of an environmental catastrophe or resource shortfall that leads to violence, there are many more counter-examples in which conflict never occurs. But popular accounts typically do not look at the dogs that do not bark. Darfur is frequently cited as a case where desertification led to food scarcity, water scarcity, and famine, in turn leading to civil war and ethnic cleansing.5 Yet, food scarcity and hunger are problems endemic to many countries – particularly in sub-Saharan Africa – but similar problems elsewhere have not led to large-scale violence. According to the Food and Agriculture Organization of the United Nations, food shortages and malnutrition affect more than a third of the population in Malawi, Zambia, the Comoros, North Korea, and Tanzania,6 although none of these countries have experienced full-blown civil war and state failure. Hurricanes, coastal flooding, and droughts – which are all likely to intensify as the climate warms – are frequent occurrences which rarely lead to violence. The Asian Tsunami of 2004, although caused by an oceanic earthquake, led to severe loss of life and property, flooding, population displacement, and resource scarcity, but it did not trigger new wars in Southeast Asia. Large-scale migration has the potential to provoke conflict in receiving areas (see Reuveny, 2007; Salehyan & Gleditsch, 2006), yet most migration flows do not lead to conflict, and, in this regard, social integration and citizenship policies are particularly important (Gleditsch, Nordås & Salehyan, 2007). In short, resource scarcity, natural disasters, and long-term climatic shifts are ubiquitous, while armed conflict is rare; therefore, environmental conditions, by themselves, cannot predict violent outbreaks.

Second, even if local skirmishes over access to resources arise, these do not always escalate to open warfare and state collapse. While interpersonal violence is more or less common and may intensify under resource pressures, sustained armed conflict on a massive scale is difficult to conduct. Meier, Bond & Bond (2007) show that, under certain circumstances, environmental conditions have led to cattle raiding among pastoralists in East Africa, but these conflicts rarely escalate to sustained violence. Martin (2005) presents evidence from Ethiopia that, while a large refugee influx and population pressures led to localized conflict over natural resources, effective resource management regimes were able to ameliorate these tensions. Both of these studies emphasize the role of local dispute-resolution regimes and institutions – not just the response of central governments – in preventing resource conflicts from spinning out of control. Martin’s analysis also points to the importance of international organizations, notably the UN High Commissioner for Refugees, in implementing effective policies governing refugee camps. Therefore, local hostilities need not escalate to serious armed conflict and can be managed if there is the political will to do so.