Water conflict is one of those topics that inspires terrific divisiveness: it’s either entirely over-hyped or the secret source of all mass migrations and international power struggles for the next century. There is, of course, a large and well-argued middle ground, but policy circles are often dominated by the groups with the more thrilling, oversimplified narratives.
This tendency towards extremes extends to one particular type of conflict: water-related terrorism. One of the prevailing authorities in this arena is Peter Gleick of the Pacific Institute. They’ve compiled a Water and Conflict Chronology database that hosts a list of water-related terrorism events dating back to the 1800s. The WCC has been cited in many national intelligence reports focused on prevention of water-related terror attacks in the United State (quoth the Director of National Intelligence in a ’12 State Department report: “Some non-state actors (terrorists or extremists) will almost certainly target vulnerable water infrastructure to achieve their objectives”).
A cursory look at this database gave me a hunch that it’s biased towards inclusion of low-intensity events and threats in developed countries, while neglecting events in developing countries. If this is true, it’s worth pointing out for a few reasons: these reports drive spending, first of all, and a biased database could shift funds away from regions of actual threat and toward regions of perceived threat. A more internationally comprehensive database could help is identify trends in water terrorism (ex. the database only lists one school poisoning in Afghanistan, but there have been a few in recent years, extending to Pakistan and Nigeria). I’m sure more rigorous catalogs exist in secret places outside my grasp, but for the purpose of general scholarship I think a comprehensive and unbiased public listing is important. On a more philosophical note, I think reporting every vague threat in the United States while only reporting a portion actual death and destruction elsewhere underscores the notion that these things should be accepted as a norm elsewhere.
In the U.S. Federal Code of regulations, “terrorism” is defined as “the unlawful use of force and violence against persons or property to intimidate or coerce a government, the civilian population, or any segment thereof, in furtherance of political or social objectives.” To assess potential overinclusion of attacks and threats in high-income countries, I’ve broken this down into five conditions, all binary:
- Did the event involve the use of force, as assessed by property damage and/or use of physical, biological, or chemical weapons?
- Was there any violence that resulted in human injuries or fatalities?
- Was force or violence intended?
- Was there a civilian target?
- Were the events conducted on the basis of a political motive?
- “A drill simulating a terrorist attack on the Nacimiento Dam in Monterey County, California got out of hand when two radio stations reported it as real attack.” would receive a score of 0.
- “Up to 150 schoolgirls are reported sickened by poison in a school water supply in an intentional attack thought to be carried out by religious conservatives opposed to the education of women,” would receive a score of 5.
For a quantitative proxy for development, I used the human development index of each country during the year of the threat or attack (and estimated this value in years without public HDI rankings). I’ve tallied the number of US (henceforth “attack intensity”, for better or worse) and plotted that number against country HDI, color-coded by region:
If you see something that looks like a trend, you’re right: there’s a drop in mean HDI of 0.1 (range 0-1) for every condition met – R^2 of .3 for the nerds. US events also meet an average of 2.53 conditions for terrorism, as opposed to an international average of 3.91 (p<.01).
It’s also interesting to examine threat inclusion by region. A full 77% of listed US events did not actually occur, as opposed to 28% of international events:
A few other interesting tidbits:
- It’s unclear from this data threat/actual attack discrepancies are because events in developed regions are frequently caught by security and intelligence forces before they can manifest in actual violence, whether threats in less-developed regions such as Africa and Southeast Asia go chronically underreported, or a combination of the two. Qualitative data from news sources does not provide much clarity here: media searches for threats akin to water terrorism threats occurring in the U.S. produced few specific threats abroad, again meaning that they may either go unreported or may simply be acted upon with greater success.
- Events occurring in countries experience some sort of conflict are, as you might expect, more likely to include human violence (23% vs 50%, p>.05). I tried pouring through news reports to see whether local media describes these events as “terrorism” or as fragments of a broader conflict narrative, but came up short. Of the 57 listed events, I tracked down 15 in both international and local media sources: 11 are referred to as terrorism in international sources, 9 in local sources. Not a significant difference, but maybe a train of thought worth pursuing further.
- There’s essentially nothing from north or west Africa in the database. In areas of genuine water scarcity, extremist groups may be more interested in controlling water sources than in destroying those controlled by others or in weaponizing the water itself. It would be interesting to expand the definition of water-related terror activity to include some measure of territorial targeting.