ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Water: A Critique of Three Concepts

Discussion of issues concerning water is largely based on three concepts - virtual water, water stress and water storage per capita. Virtual water and water stress have limited usefulness from the point of view of analysis or policy formulation, while the concept of water storage per capita is fallacious and needs to be abandoned.


be importing or exporting the water that

Water: A Critique of Three Concepts

had gone into its production.

(iii) “Water storage per capita” is arrived

at by dividing the total water storage, i e, Ramaswamy R Iyer the total capacity of reservoirs behind

Discussion of issues concerning water is largely based on three concepts – virtual water, water stress and water storage per capita. Virtual water and water stress have limited usefulness from the point of view of analysis or policy formulation, while the concept of water storage per capita is fallacious and needs to be abandoned.

Ramaswamy R Iyer ( is a well-known commentator and writer on issues relating to water.

Economic & Political Weekly january 5, 2008

his article examines three concepts relating to water that are current in the international water circuit and academia, namely, water stress, virtual water and water storage per capita. I have written about the first two earlier, but the present article takes the examination further.

Let us first take note of the definitions of these terms.

  • (i) “Water stress”, a concept that we owe to Malin Falkenmark of Sweden, is a measure of annual water resource (AWR) availability per capita, arrived at by dividing the freshwater availability in a country, or, roughly speaking, the water resource endowment of that country by its population.
  • (ii) “Virtual water” is the amount of water that is needed for the production of, and therefore, embedded in food or other products; it follows that if we import or export those products, we may be said to
  • dams (and other smaller structures) in a country divided by its population.

    Virtual Water Trade

    Taking “virtual water” (and “virtual water trade”) first, every import or export can be redescribed with reference to what has gone into the production. Exports of rice and wheat can be treated as exports of water (among other things); exports of aluminium can be regarded as exports of electric power; exports of iron and steel originate in iron ore and can be regarded as exports of the soil of the country; and so on. This can serve the purpose of drawing our attention to the implications of certain kinds of trade. For instance, a watershort country or area within a country can be cautioned against producing a waterintensive crop for markets outside its borders. However, in international conferences and forums the concept of virtual water trade tends often to be used as one


    more means of persuasion directed at developing countries, extolling the virtues of markets and imports as against domestic production.

    It is argued that developing countries that face growing pressure on their water resources need not use their water to grow water-demanding produce, but could import their requirements from water-rich countries, thus virtually importing water. This is part of the neoliberal economic prescription. Some of us may feel that a distinction needs to be made between imports of staple food and imports of other things such as fruits and vegetables and that a vulnerable dependence on food imports is not desirable, while others may argue in favour of exporting whatever we can and importing whatever we need, including food. Leaving that aside, it seems to me that we need to be wary of the doctrinaire presentations of the ideas of virtual water and virtual water trade in international forums. However, while resisting such advocacy, we can accept the term as a useful reminder of the water implications of an economic activity.

    Water Stress

    Turning to water stress, we must note that the term is well-established and has become part of conventional wisdom. It is widely and loosely used. For instance, we often come across statements to the effect that the per capita availability of water resources in India is declining, and that India is fast becoming a water-stressed country. There is a bit of confusion here that needs to be cleared up.

    The AWR per capita in a country needs to be distinguished from the basic water requirement (BWR) per capita of a human being. The former, as we have seen, measures the quantum of water resources available to a country divided by the population. According to accepted doctrine, an AWR of 1,700 m3 per capita means that only occasional and local stress may be experienced; an AWR of less than 1,000 m3 indicates a condition of stress; and one of 500 m3 or less means a serious constraint and a threat to life. On the other hand, the BWR is the quantity of water that a human being needs to meet his or her daily needs for drinking, cooking, bathing and sanitation, measured in litres per capita per day (lpcd). Peter Gleick has estimated this as 50 litres. If we consider that a low estimate and double or triple it (Falkenmark takes the BWR as 100 lpcd), it still comes to no more than 36.5m3 or 54.75 m3 per person per annum. Why then should a per capita availability of 1,000 m3 mean that a country is water-stressed? The answer is that in addition to our direct and personal water needs, we also need food, energy, consumer goods and various industrial products, and water goes into their production. It follows that the totality of our water needs goes well beyond the BWR in Gleick’s sense. That is what Falkenmark’s AWR norm measures.

    If the water available to a person falls below say 150 lpcd or 54.75 m3 in a year, that person, as a human being, may be in a state of physical stress. If the country as a whole has an annual water resource availability of less than 1,700 m3, it is argued that the country will be approaching a state of stress. The “stress” of the country in this sense is not the same thing as an individual’s stress in the absence of water. In the latter case, we are referring to a pathological state, whereas in the former stress is a metaphor for a difficult situation. This distinction is not always kept in mind in the discussions.

    In some cases, a country that is not wellendowed with natural resources may manage to meet a wide range of its requirements quite satisfactorily through imports. As we noted earlier, we may hold different views on how dependent we should be on imports for meeting essential requirements, but that is an economic or political point. What the water stress theory implicitly assumes is that a country needs a water resource (WR) endowment adequate to support a certain level of domestic production of goods and services, and that a country better endowed with water resources than another would be better placed both politically and economically than the latter. One can readily accept those propositions, but the use of the same term stress in the country context, where it has an economic connotation, as well as in the individual context, where it has a pathological meaning, has created much confusion. To repeat: a low AWR may mean economic difficulties for a country, but it does not necessarily imply physical stress or threats to health for its people.

    It is, of course, true that a poor country with a very low AWR level may find it difficult to meet the people’s minimum needs for a reasonable quality of life through imports, and that at that level the country’s economic stress may actually translate into physical stress for the people; but it is debatable whether the cause of that stress is the low AWR of the country or its



    poverty. It is not self-evident that a country with a low AWR must necessarily be poor. For instance, the oil-rich but water-poor Gulf countries.

    It may be argued that when such a country imports its requirements, and thus its water needs, it is in fact, adding to its AWR, and therefore, the water stress doctrine continues to be valid, in the sense that the AWR stands enhanced above the stress level; but this is a piece of sophistry. AWR is essentially a measure of resourceendowment. It postulates a certain level of that endowment as necessary for the country’s well-being in an economic sense.

    Certain Questions

    What then is the relationship between the concepts of virtual water and water stress? Superficially, it may seem – it had earlier seemed to me – to be one of contradiction, as the former implies that all needs can be met through trade and the latter considers a level of natural resource endowment necessary. However, if we recognise that neither concept is absolute, they can be reconciled. Yes, a country’s needs can be met partly through trade, but some degree of self-reliance is desirable; and yes, a certain level of resource endowment (AWR) is desirable. But that endowment can be supplemented through trade. If we accept that proposition, the apparent contradiction disappears, and both concepts serve some limited purposes.

    However, one more question needs to be asked about AWR. Why do we need to classify countries by the AWR measure at all? Variations exist not merely in water endowment, but also in other natural factors such as elevation above sea-level, nature of terrain, temperature, humidity, climate as a whole, proneness to cyclones, thunderstorms, tornadoes or earthquakes, etc, and people live their lives accordingly. Would that not be true of AWR differences as well (subject of course to the proviso that areas with an AWR level below a certain minimum will remain uninhabited)?

    Water Endowment

    I venture to suggest that we do not need the concept of AWR to ascertain that a country’s water endowment is poor; that fact would be clear enough to all. What follows from that recognition? The resource endowment

    Economic & Political Weekly january 5, 2008

    cannot be changed; and while the rate of growth of population can be controlled to some extent, existing numbers cannot be reduced. Reductions may, of course, occur because of epidemics or war or natural calamity, or, when the rate of population growth falls below the replacement level because of certain economic or social factors. What then can people living in a country or area with a low water endowment do?

    The following are the possibilities:

  • (a) They can abandon the country/area and move elsewhere, but that is not an easy option and it is not always available.
  • (b) They can develop ways of living that can be viably pursued with the given availability of water, and learn to use water with the utmost economy: that is what desert people do in India or elsewhere, and they have been doing so for centuries without calculating an AWR. The Rajasthan desert was never an uninhabited area, and it was in the past an area of high civilisation, not of primitive living; nor did the people regard themselves as “waterstressed” in the Falkenmark sense. They simply developed ways of using the available water to good purpose. Israel is an outstanding example of a country that manages a limited water resource endowment very well, though that statement needs to be qualified by the point that Israel has also captured a part of the water resources of the Palestinians and the Jordanians. (c) They, i e, the people of a country with poor water endowment, can supplement the AWR level of the country through imports of virtual water to the extent feasible. (d) They can live in waterprofligate ways, generating ever-increasing and unsustainable demands for water (and for other natural resources), undertaking supply-side projects with all their attendant impacts, environmental, social and human, and with long-term consequences for the ecological system and for the country. That is what humanity as a whole has been doing in relation to its habitat, planet earth. What precisely is the contribution that the concept of AWR and the related (misleading) use of the term water stress can make in this context, except to make two obvious truths even more obvious, viz, that some countries are better endowed with water than others,
  • and that they too will experience difficulties with the growth of population and the pace of urbanisation.

    As mentioned earlier, the postulation of a desirable level of AWR per capita may be useful as a counter to the virtual water theory, and the latter in turn may be useful as a counter to the water stress theory. Either concept can be used to moderate the other.

    Water Storage Per Capita

    The water stress theory gets connected to the concept of “water storage per capita” because the former becomes an argument for the latter. One of the things that countries do to augment the availability of water for use is to build “storages”, i e, dams and reservoirs. This does not change the natural endowment, and therefore, the AWR remains the same; but more of that AWR comes into the category of “usable WR”. The thinking behind this is the following: in the household, given intermittent water supply, we store water; so too, given the uncertainty, variability, and in some countries, seasonality of rainfall, we must store water in reservoirs for transferring water from wet season to dry, from good years to bad, and from water-abundant areas to water-short areas. Proceeding from this, the point is often made by many writers and speakers, including the World Bank in its report India’s Water Economy: Bracing for Turbulence, that water storage per capita is much lower in India than in the US, and that India must add substantially to its storage capacity – “water infrastructure” in the World Bank’s language.

    However, what is the justification for adopting “storage per capita” as a norm and a desideratum? What such a norm implies is that the extent of storage that a country needs to build must be determined by the size of its population. In other words, the country with a larger population must have more storage than a country with a smaller population. But is that proposition maintainable? Large dams/reservoirs are location-specific: they can be built only on certain kinds of terrain (ignoring for the moment the nature and characteristics of the rivers in question). Let us assume that there are two countries with populations of similar magnitudes. By the logic of the water


    storage per capita concept, they should have similar storage capacities. However, would that be possible if one is a hilly country and the other is flat and featureless? One country may tend to build more projects than the norm because the topography favours this, whereas another country may not be able to build any storage at all because of the nature of its terrain but will have to manage its water needs in some other way. Can we apply a standard norm of water storage per capita to, say, India, Nepal and Bangladesh? On the other hand, let us imagine two countries with equal physical possibilities of storage-creation in terms of terrain, topography and river characteristics, but one densely and the other sparsely populated; or one with rich biodiversity and the home of rare species of flora and fauna, and the other not sharing these characteristics: can they have similar magnitudes of water storage per capita? Paradoxically enough, the very magnitude of the population that seems to necessitate the building of more storage will render such construction difficult because there will be people everywhere.

    So far we have not considered the desirability of large storages. Two points must be noted: first, that all such projects have environmental, social and human impacts which must be reckoned and evaluated before they are undertaken. Secondly, that storage does not necessarily mean large reservoirs behind big dams, but could include small, local water-harvesting structures and underground storages in aquifers. Those – the World Bank and others – who talk about water storage per capita usually do not take these things into account, but use the concept as the basis for arguing in favour of building more dams or water infrastructure. Behind that recommendation lies the old and discredited thinking that water-use means water-abstraction; that “available” WR becomes usable WR by being stored behind a dam; that water flowing in a stream is not “used”; and that water that flows to the sea is “wasted”. However, even without going into all these points, the concept seems questionable enough.

    The conclusion that this examination leads to is that the concepts of virtual water and water stress based on AWR per capita have a limited usefulness as mutual correctives; and that the concept of water storage per capita is fallacious and needs to be abandoned.

    Dear Reader,

    To continue reading, become a subscriber.

    Explore our attractive subscription offers.

    Click here

    Back to Top