Statistical evidence proves disputes escalate.

The typical version of the water war hypothesis posits that countries will be willing to wage war over threats to existing water resources that may arise, for instance, from upstream diversions. Some of the more developed of theoretical models incorporate additional explanatory variables such as riparian position, relative military might, measures of relative scarcity, and level of economic development (e.g., Frey and Naff, 1984; Hauge and Ellingsen, 1998). A stronger version of the water war hypothesis claims that countries experiencing acute water scarcity will be compelled by a "hydrological imperative" to obtain additional water supplies from neighboring countries, and will resort to violent means if needed (e.g., Cooley, 1984; Stauffer, 1982; Stork, 1983).

Still other theoretical models stress that water scarcity alone may not in and of itself often serve as a casus belli between nations, but rather, may aggravate existing international or domestic conflicts (e.g., Glcick, 1991). Water scarcity may, for instance, result in population displacement ("environmental refugees"), leading to competition and political destabilization (Homer-Dixon, 1994). Similarly, droughts may reduce returns from agricultural labor, thereby improving the relative returns from initiation of violent competition between groups or revolt, and as such may be a contributing factor to wars (Collier and Hoeffler, 2002). Water scarcity may be a result of a number of factors, including internal population increases, increased use by other riparian nations, increased per capita demand due to changes in standards of living, decreases in quality of available resources, or unequal access to available resources (Homer-Dixon, 1994). In addition, the impacts of climate change on water availability may also contribute to local water scarcity, raising the potential for conflict (Van der Molen and Hildering, 2005).** These theoretical constructs are in line with the larger literature on environmental security."

By far, the majority of the empirical evidence presented in support of the water war hypothesis consists of case studies and anecdotal references. Among the most popular are Israel's bombing of Syrian bulldozers attempting to divert the headwaters of the Jordan River and the Palestinian Liberation Organization's (PLO) attempt to blow up Israel's National Water Carrier (e.g.. Naff and Matson, 1984; Starr, 1991; Gleick, 1993; Lowi, 1993). While the Middle East is the most cited example, other regions have also been cited as having experienced violent conflict over water. Most recently, skirmishes in Sri Lanka have been described as water-related (Bajpaec, 2006; The Economist, 2006; Reddy, 2006), and several commentators, including U.N. Secretary General Ban Ki Moon, have characterized the violence in the Darfur region of Sudan as stemming, at least m part, from struggles over scarce water resources (Ban, 2007).

Until relatively recently, the water war hypothesis lacked any systematic empirical support. Over the past few years, however, several researchers have published large sample size statistical studies on the subject (although such work still pales in comparison to that relying on a few case studies). In one of the first such studies, Hauge and Ellingsen (1998) found water scarcity to be statistically correlated with the outbreak of civil wars between 1980 and 1992 and with armed conflict within nations between 1989 and 1992. Gleditsch et al (2006) found that dry countries have more conflict than humid ones. Toset et al (2000) and Furlong et al (2006) found that, holding other variables constant, countries that share a river are more likely to engage in violent conflict than neighboring countries that do not, and the higher the number of rivers, the higher the likelihood of conflict. Levy et al (2005) found that regions with high levels of variability in rainfall were statistically more prone to high (but not to low) intensity domestic conflict-Miguel et al (2004) also found a correlation between rainfall and violent conflict within Africa. Both Furlong et al (2006) and Hensel et al (2006), while noting the small number of militarized disputes over rivers, nonetheless found that militarized conflict over rivers is more likely when an initiating state is experiencing greater water scarcity.