‘Consumers are embedded in society and not in science,’ said the bold red letters on the screen. Tony Allan, Professor of Geography and head/member of the Water Issues Group at the School of Oriental & African Studies, was giving a seminar on his theory of ‘virtual water theory’ (henceforth: VWT) at the Stockholm World Water Week. He had just won the conference’s award for developing the theory he was explaining to the audience. In a nutshell, as a concept, virtual water (VW) is the water used in the production of agricultural commodities. Although VW remains largely absent from political agendas its simplicity makes it extremely effective. Most people are shocked to find that whilst they consume three to four litres of water a day, and perhaps use around 300 litres for domestic purposes, an astounding 3,000 to 6,000 litres go into the production of their daily diet (Allan, 2003, p, 5).

On their own, single disciplines such as international relations (IR) are insufficient to facilitate the understanding of water-scarcity related issues. Virtual water attempts to combine culture, society and political economy in an analytical framework for analysing international agricultural trade. Amongst the many disciplines that contribute to VW, international relations and economics stand out as front-runners in the discourse. Both distinguish between national and international levels of analysis and have produced theories to predict responses to water scarcity.

This paper aims to explore the added value of VW through its constituent disciplines. At the outset, this is a matter of political will and economic recognition because, as Allan states, VW is ‘economically invisible and politically silent.’ (Allan, 2005, p.185) Thus, this paper hopes to go beyond exploring the importance of this novel policy-making tool, towards an analysis answering the fundamental question: ‘why should policy makers take virtual water into consideration?’

First, the definition and significance of virtual water will be elaborated. Then, an exploration of the (limited) role of IR will reveal the rather counter-intuitive finding that national water scarcity and international conflicts are not intimately related. This leads to the chapter on VW’s coupling to the economic theory of comparative advantage.

Allan’s VW fits into gaps IR theories are unable to fill. VW ends up receiving critique from both disciplines. Despite these, it will be argued that VW retains policy relevance Allan suggests. Indeed, VW’s thoroughly interdisciplinary nature gives it the potential to address new water scarcity situations where single disciplines cannot.

Virtual water theory

Virtual water is, quite simply, the water needed to produce agricultural commodities. Where one normally thinks of a cup of cup of coffee as containing a few millilitres of water, the VW used in growing the beans for this cup is 140 litres (see figure 1 and 2 below. Chapagain and Hoekstra, 2003, p. 42). Producing a tonne of grain requires around 1000 tonnes of water and roughly 80% of the world’s water goes to agricultural production. Considering the massive scale on which grains are consumed the amount of water being traded in its virtual form is significant, to say the least.

Figure 1. (Chapagain and Hoekstra, 2003, p. 40)

Figure 2. Global average virtual water content of daily products (Chapagain and Hoekstra, 2003, p. 42).

Previously, the World Bank had attempted to introduce the term ’embedded water’ without much success. Allan developed VW in 1992 as a result of pondering the relationship between water scarcity and conflict in the Middle East, a region that ran out of water in the 1970s1. Allan calculated that the Middle East currently imports around 50 million tons of grain every year, requiring 50 billion cubic meters of water (Allan, 2003, p. 5). Accordingly, VW is of importance to trade, water and agricultural policies. Yet, linking a country’s water shortage to its agricultural production remains taboo (Allan, 2005, p. 185). Donors running water programs in such countries are commonly confronted with the arguments that better water management would follow if only they had ‘a little more water’ (Goleman, 1997).

Allan launched VW into international water discussions, not as a scientific theory but as a concept designed to bring about policy changes. Other authors who have adopted VW in their work hypothesize about the consequences that the recognition and incorporation of the virtual water values in trade policy will bring about (Hoekstra and Hung, 2003; Hoekstra and Chapagain, 2007; Renault, 2002). It is a quasi theory, still being entertained by the academic community.

Currently, some water-scarce countries are net exporters of virtual water, causing considerable stress on their water resources. VWT would predict that taking the VW values of crops into account in trade policy would lead to an adjustment in the agricultural commodities that are imported and exported. Water-scarce countries would import water-intensive commodities and thereby minimise the costs of water scarcity. Allan proposes VW as a solution because of its four virtues. Firstly, it is an effective trade solution for coping with water deficits. Secondly, it is politically silent as it is not a genuinely contentious political issue in within domestic politics, particularly in countries from which the water-intense crops would be imported. Thirdly, it is economically invisible as it is effectively contained within crops. Fourthly, the international trade of VW is very flexible compared to real water. Water embedded in staple foods is also more suited to coping with varying food demands. The storage of virtual water in grains guarantees far more security than engineering solutions to water storage problems (Allan, 2005, p.188).

The sternest critique of the VWT’s scientific robustness comes from agronomist Stephen Merrett (Merrett, 2003). Merrett raises linguistic objections to the use of the term ‘virtual’ when in fact the water is quite real. The meaning of ‘virtual’, Merrett says, is non-negotiable as its use is well-established, and ‘It is not good science to build theory on terms that are inherently misleading.’ (ibid., p. 3) Merrett calls upon Occam’s razor: ‘it is vain to do with more what can be done with fewer.'(Merrett, 2003, p. 4) Merrett feels that ‘virtual water’ could be replaced with ‘the crop water requirements of food exports’ (ibid.).

Discarding VWT on these grounds is not quite so simple, given the fact that it is rooted firmly in both political views on water as a resource and the age-old economic theory of ‘comparative advantage’, as will be explained later. True to its reputation as the study of scarcity, economics lends itself to analysing the international trade of commodities, even though Allan’s interpretation of water as the traded commodity is somewhat unusual.

Hoekstra and Hung of the UNESCO-IHE Institute for Water Education were the first to translate Allan’s idea into quantifiable models (Hoekstra and Hung, 2003). They multiplied the international crop trade flows with their virtual water content, whilst keeping in mind differences in climate, water scarcity and soil structure. The results showed that international virtual net water flows total 695 Gm3yr-1, which represents 13 percent of total water used by crops worldwide (see figure 3 for complete results). However, for water scarce countries in particular, this number does not reflect the proportion of the water they are importing.

Figure 3. (Chapagain and Hoekstra, 2003, p. 44)

On water security

Inevitably, any international agreements on coping with water scarcity will depend on politics. This in turn is determined by common perceptions of water security, as studied in the sphere of international relations. At the World Economic Forum in 2008, UN Secretary General Ban Ki-Moon stated that, ‘Too often, where we need water, we find guns instead.’ (Ban, 2008) The expression ‘water wars’ grossly exaggerates the relationship between conflict and water scarcity. Water shortage by itself is insufficient as a predictor armed conflict. Wolf and colleagues have shown that there have been considerably less water conflicts since the intensification of international water agreements largely due to efforts by the UN. Furthermore, most of the conflicts since 1946 were between Israel and its neighbours (Wolf et al., 1997).

This example demonstrates the ‘hydro-centricity’ of IR scholars (Allan, 2005, p. 183). IR theorists jump to conclusions based solely on the absolute water resources used without taking into account the ability of the affected community to adapt by making creative policy options. Dr. Thomas Homer-Dixon proposes a loop of social instability resulting from environmental degradation, population growth and uneven distribution of resources (Homer-Dixon, 1994). A typical feature of such models is the use of, often alarming, per-capita freshwater data to describe regional water availability (Gleick, 1993). Again, Allan rejects this model as simplistic and deterministic. Furthermore, such data is misleading as it makes the implicit assumption that where there is water there is also the political will to distribute it (Mustafa, 2007).

From the perspective of realist IR scholars the state is thoroughly self-interested as there is no supranational authority to maintain international order. Consequently, the maintenance of sufficient water levels through international cooperation is a riddle.

According to Dr. Kathryn Furlong IR theories underestimate the ability of states to engage in sustained multilateral cooperation. The threshold at which countries revert to violence is raised by the countries’ political investment in efforts to reach a negotiation (Furlong, 2006, p. 444). Being a bulky liquid and thus difficult to transport, water’s location is usually taken as a given. Consequently, on a sub-national level cooperation between societies sharing the same water source is regarded as unavoidable because of the fundamental need for water (ibid.). In fact, this interdependency, through shared water for example, is often seen as a guarantee that asymmetric gains will be harmonized.

Another prominent criticism of IR is that it is not capable of understanding water governance as a social construction on a sub-national level (ibid.). Many societies across the globe have developed under conditions of water scarcity or abundance. Decisions about water management often influence, and are themselves subject to, power dynamics. Furlong says that IR cannot reflect, ‘the new geographies of water that simultaneously reinforce and reflect distributions of privilege and disadvantage within and between households, communities and states.’ (ibid., p. 445) In many developing countries it is still the case that the elite have access to excess water whilst thirst prevails among the poor. Thus, IR’s realist school of thought’s assumption that the state should be the smallest unit of analysis renders it incapable of accounting for the immense influence of sub-national actors.

Clearly, theoretical linkages connecting water resources and political decisions tend to oversimplify water security. IR is not capable of including culture, society and political economy in its analytical framework (Allan, 2005, p. 183). Allan’s VW fits snugly into this knowledge gap but it does so only by coupling water scarcity to trade.

Where water meets trade

Wichelns eloquently summarises comparative advantage as the theory that, ‘nations can gain from trade if they concentrate or specialize in the production of goods and services for which they have a comparative advantage, while importing goods and services for which they have a comparative disadvantage.’ (Wichelns, 2004, p. 52) Although it does address resource endowment, VW calculations do not take production technologies or opportunity costs – the cost of picking one choice out of a series of mutually exclusive choices – into consideration. In fact, any economist will know predictions based on resource endowments are not always reflected in comparative advantage calculations (Appleyard et al., 2006). For example, a water-scarce country producing two goods at a lower price than a water-rich country will still have an absolute advantage in the production of both goods (ibid.). A large opportunity cost in VW translates to high water scarcity values as every extra unit of water consumed is worth more.

In response, Allan stated that, ‘virtual water is something of a descendant of the concept of comparative advantage.’ (Allan, 2003) More fundamental still, opportunity cost calculations are not predictive in the long run. Here, Wichlens points to government responsibilities, ‘When scarcity values are not reflected in market prices, or when governments administer production and marketing activities, policy intervention might be required to increase the values generated with limited resources.’ (Wichelns, 2004, p.60) Thus, including opportunity cost calculations within VWT will improve its policy relevance (ibid., p. 50). As of yet, water is not a factor in production and trade decisions. It is not a variable in comparative advantage because it is not a limiting factor (Wichelns, 2004). Allan himself pointed out that energy resources are a far more pressing agenda item (Allan, 2008).

Agricultural commodities have a special status within the World Trade Organisation, resulting in many loopholes for protectionist policies and subsidies (Healy et al., 1998). Highly controversial is the fact that recognising VW might well require farmers to pay for water. Thus, given that a large proportion of the labour force depends on agriculture, VW will require a period of acceptance and transition. (Allan, 1999) Furthermore, recognising water scarcity is a precondition for introducing the policy mechanisms that would secure efficient trade patterns.

Virtual water as an interdisciplinary concept

Merrett argues that the concept of virtual water inadequately addresses the economic, political and social processes it covers. He suggests that water resources management, traditionally dominated by engineering and hydrology, requires thorough interdisciplinarity (Merrett, 2003, p. 2). In order to do this, the social sciences are now forging an appropriate language, aided by philosophy. In this context, science and resulting policies, says Merrett, can only develop as fast as its constituent disciplines proceed in creating their own language.

It seems that, essentially, VWT is an umbrella term pushing for conceptual innovation from within many disciplines. VW, as a concept, lacks scientific exclusivity to any one discipline and suffers the consequences of being regarded, perhaps justifiably, as unscientific. However, one cannot dismiss the argument that VW’s unique interdisciplinary nature is a response to the practical necessity of the times.

Conclusion

The popularity of the VW concept has demonstrated the need for effective policies to cater to the newly arisen needs resulting from water scarcity. IR cannot explain enduring peace in water-scarce societies. Furthermore, the way it links water to food security and the livelihoods of millions of farmers worldwide leaves no doubt about VW being a political issue. However, as it shares several characteristics of comparative advantage, VW is inextricably economic in nature as well. Both disciplines, however, imply the need for the recognition of water scarcity. Thus, the policy-maker endeavouring to reform trade policy requires an interdisciplinary team to create a middle way. Dependency on other states can be minimized by creating a basket of trade relations. However, failing to undertake any action means taking a serious risk of which the opportunity cost could prove to be enormous. Not only in terms of crumbling agricultural sectors in semi-arid regions, but also of environmental, socio-economic damage that might eventually lead to political escalation without the safety net of VW.

Essentially, VW is a term that was created to embody this capacity to easily speak to one’s imagination because, ‘capturing the invisible is easier than communicating the invisible to move the minds of consumers.’ (Allan, 2008b) Indeed, virtual water is conceptual innovation in its own right and remains to be tested by society and the academic community alike.

Footnotes

1. On a per capita basis, these countries could be defined as water scarce (Allan, 2005, p. 187).