Occasional PapersPublished on Feb 01, 2021 PDF Download
ballistic missiles,Defense,Doctrine,North Korea,Nuclear,PLA,SLBM,Submarines

Finding Solutions to Water Scarcity: The Potential of Virtual Water Trade in Agricultural Products

  • Roshan Saha
  • Preeti Kapuria

    This paper analyses the patterns of virtual water trade (VWT) in agricultural products across the globe—VWT is the flow of water embedded in goods and services when they are traded—and the implications for alleviating water scarcity. Virtual water trade has been crucial in ameliorating water scarcity in virtual water-importing nations. At the same time, it has led to per capita water availability declining at a more rapid rate among the net virtual water-exporting countries like India, which are water-scarce, to begin with. This paper argues that despite a few shortcomings, virtual water trade remains a useful tool in water resources management. To be effective, it must be supported by economic instruments that capture the “scarcity value” of water and drive optimal decision-making in agricultural water resource management. 

Attribution:

Roshan Saha and Preeti Kapuria, “Finding Solutions to Water Scarcity: The Potential of Virtual Water Trade in Agricultural Products,” ORF Occasional Paper No. 298, February 2021, Observer Research Foundation.

1. Introduction

Demand for water has risen massively across the world in the last few decades, due to increasing world population, changes in living standards and consumption patterns, and expansion of irrigated agriculture.[1] The World Economic Forum, in its Global Risks Report 2020 has noted that “water crises” are the largest global risks of the near future.[2] While “water wars”, as some fear, are still not likely, water scarcity is a growing concern that requires immediate attention.[3] Certain regions like the Middle East and North Africa (MENA) are extremely water-scarce and therefore dependent on other countries to meet parts of their basic water needs, especially in agriculture. Most of their domestic food demand is met through imports. These imports make up for the water scarcity in the region by utilising the water resources of other countries through the transfer of “virtual water”.

While “water wars”, as some fear, are still not likely, water scarcity is a growing concern that requires immediate attention

Virtual water refers to the quantum of water used to produce a good or service; it is the volume of water embedded in that good or service.[4] When goods and services are traded, all the factors used in their production phases, in a virtual sense, also move across and within countries and regions. The movement of such embedded water is known as “virtual water trade”.[a],[5] Virtual water trade has the potential to reduce global water scarcity by distributing water resources ‘virtually’ from the relatively water-surplus regions, to the water-scarce. By ensuring that virtual water transfers also take place from regions with higher water productivity to those with lower, it can ensure efficiency in global water use.

Historical evidence from the Roman Empire underlines the importance of virtual water trade in determining the resilience of a region/country against the twin threats of growing demand and climatic variability.[6] In that pre-industrial era, Romans used the virtual water trade embedded in food grains through the Mediterranean and Black Sea region to link the food surplus and deficit areas within the empire. This helped ensure a steady supply of grain to the main cities of the empire even when local water resources were insufficient to meet cultivation and domestic needs.[7] Although evidence is available only for trade, it can be assumed that virtual water trade was an important factor that spatially integrated the empire and also enhanced its longevity.[8]

Since the mid-20th century, water resources management has undergone a paradigm shift—from supply-side interventions to demand management. There has also emerged certain interdisciplinary approaches to policymaking in the form of integrated water resources management (IWRM).[b] A central tenet of IWRM is “clear and strict prioritisation of various types of needs and demands for water.”[9] Optimal allocation of a scarce resource, in this case water, among multiple uses, is the canonical definition of economics, as proposed by Sir Lionel Robbins.[10] An optimal strategy would involve putting water to its best possible use. Economic instruments such as price (equivalent to the marginal productivity of water) can play an important role in this optimisation exercise. In the presence of a market that reflects this information, regions or countries can decide on the most productive use of their water resources. Markets for water do exist in Chile, California, and the Murray-Darling basin in Australia. On 17 September 2020, the US’ CME Group and NASDAQ[c],[11] together announced the launch of the NASDAQ Velles California water futures market. But IWRM is yet to be incorporated into decision-making at the required scale in other places.

The World Bank has found that the agriculture sector accounts for 70 percent of water use across the world. Implementing IWRM in agricultural water resource planning can therefore alleviate water scarcity.[12] Agro-climatic conditions determine the water intensity of agricultural production across regions. If certain regions are not suitable for the production of water-intensive crops,[d] it is prudent to utilise the scarce water resources available for other purposes such as domestic or industrial consumption. The water demand for these other uses, especially domestic, entails in-situ consumption. But water requirement for food products can be met through imports of those products wherever feasible. If regions that are unsuitable for the cultivation of water-intensive products like paddy, import them from regions agro-climatically more suitable, the former would be virtually importing water and thus increasing its availability, albeit in virtual form.[13] This puts virtual water trade strategy at the centre of policies envisioned under an IWRM paradigm.

The subsequent sections discuss the evolution of the concept of virtual water trade within a standard theory of international trade, and the patterns of such trade across the world. The paper then analyses whether water resource endowments have been the primary factor in virtual water trade, by mapping virtual water exports and imports against the per capita water availability in the top virtual water-exporting and -importing countries. The results suggest a need to incorporate demand-side management through valuation of water resources as an integral element of making virtual water trade an efficient tool of IWRM and alleviating water scarcity.

2. Virtual Water Trade: Theoretical Underpinnings

2.1 Heckscher-Ohlin and the Theory of Comparative Advantage

Virtual water trade has its theoretical foundations in the concept of comparative advantage propounded by David Ricardo in his 1817 treatise, “On the Principals of Political Economy and Taxation”.[14] However, only after Alfred Marshall, William Stanley Jevons, Carl Menger and Leon Walras had developed the neo-classical school of economics in the late 19th century, did it become possible to generalise David Ricardo’s theory, by going beyond the labour theory of value and examining why countries have comparative advantages in different goods and services. In 1936, Bertil Ohlin and Eli Heckscher formulated a compelling argument that explained the differences in comparative advantage across commodities and countries by incorporating a neo-classical general equilibrium framework for a single economy and subsequently for an economy with international trade.[15],[16], [17]  The Heckscher-Ohlin theorem states that a country/region exports a product which is intensive in the factor of production that is relatively abundant in the country/region.

The theorem linked export-import patterns to factor endowments.[18] A crucial assumption of the H-O model was that of similar production technology across countries. Ohlin’s definition of relative factor abundance rests on the pre-trade price ratio of factor prices in the two countries. The final price of a product is dependent upon the factor prices. According to Heckscher and Ohlin, if a factor of production is relatively abundant in a country/region, it is supposed to reflect in a relatively cheaper price for that factor.[e]

Introducing a relatively capital abundant foreign economy into the model, with the subscripts H and F for home and foreign countries, respectively—

where and represent the prices of textiles (T) and computers (C), respectively.

This implies that the relative price of computers is higher in the home country. The home country can be said to be relatively less abundant in capital, and therefore capital would be relatively more expensive in H under autarky. Labour is the relatively cheaper factor of production in the home country. Based on this relationship between factor endowments and relative price of commodities, the home country has a relative comparative advantage in textiles and will therefore export textiles to the foreign country, which in turn has a comparative advantage in computers.

Applying the basic observations of the Heckscher-Ohlin model to the study of trade in water-intensive products, it was argued that water-abundant regions should be exporting water-intensive products to those regions where water is relatively scarce, because the former would have a comparative advantage in water-intensive products. This was initially observed in the MENA countries, which were among the first regions to suffer water scarcity.[19] By importing water-intensive agriculture products, MENA countries were able to avoid over-exploitation of their already scarce water resources.

Virtual water trade as a strategy to ameliorate water scarcity can be explained using the concept of ‘scarcity’ value of water. It refers to the additional benefit gained by relaxing the limit on the availability of water by a single unit – it is the value that would have been generated if an extra unit of water was available. Due to water availability constraints this value cannot be realised and is thus aptly called the ‘scarcity’ value.[20] According to Ghosh and Bandyopadhyay (2009), scarcity value of a unit of water, in the production of a commodity (say, rice), can be formulated as:

where,

refers to the scarcity value of an additional unit of water,

is the value of marginal product of an additional unit of water, and

is the marginal cost of an additional unit of water usage.

Following the underlying assumptions of production and cost functions, Equation 2 has been graphically depicted in Figure 1.

Figure 1

Scarcity value of water

Source: Ghosh and Bandyopadhyay (2009)[21]

Scarcity value is also a reflection of the unmet demand for water. Figure 1 can be used to explain a hypothetical economy characterised by water scarcity. Under conditions of perfect competition, the equilibrium amount of water that a farmer (or central planner) would use is reflected by WE, corresponding to the point of intersection of the MP(W) and MC(W) curves. But since this is a water scarce economy, equilibrium at WE is not possible. Rather,  is the maximum water resource available in this country, and the equilibrium corresponds to this value of W. At the equilibrium, the difference between the MP and MC curves at  is the scarcity value of water. For water endowments below W1, an increase in the availability of water will increase scarcity value, but such situations are presumed to be rare. Therefore, if this economy resorts to virtual water imports, it can be conceived of as a rightward shift of , and a reduction in the scarcity value of water with respect to the commodity under consideration. Thus, virtual water embedded in the import of this commodity reduces the intensity of water scarcity in this economy.

This argument was examined in the context of paddy cultivation in Tamil Nadu and Karnataka.[22] By resorting to import of paddy instead of cultivating it, both states can increase their water savings.   There are challenges of doing so, especially when institutional support in the form of subsidies for irrigation and price support mechanisms play an instrumental role in deciding cropping patterns. Nonetheless, the concept of scarcity value of water provides the necessary insights for effective water management in agriculture.

In other words, if a region has limited water resources, an additional unit of water would be more valuable in it than in a region relatively well-endowed. Virtual water imports from a region where the scarcity value of water is relatively low, implying relatively water surplus conditions, can help ameliorate the problem of water scarcity. The H-O model, in the context of water-intensive agricultural goods trade, should be extended to include variables that capture the price of water used as a factor of production.

In markets with perfect information, the relative availability (supply vis-a-vis demand) of a factor is reflected in its price. Axiomatically, the price or value of water is equal to the marginal productivity of water. This assumes that markets for water resources exist. However, water is one of the most underpriced factors of production, and therefore it is not adequately reflected in the calculation of comparative advantage of a product. This may result in perverse trade flows, where water-intensive crops are exported on a large scale from areas where water is highly scarce and overexploited. Under these conditions, as long as markets for valuing water resources remain elusive, it is unlikely that trade will contribute to optimal production and trade outcomes from a water-resources perspective.[23] Scarcity value estimations can indicate the price of water in a region based on its relative availability and the marginal productivity and cost associated with usage of an additional unit of water. This can then be a tool to guide policy makers in water resource management. It is, therefore, of paramount significance to the relevance of the H-O model in explaining virtual water trade patterns that the price of water resources used in the production process is considered.

2.2 Alternative Models to Explain Virtual Water Trade

Empirical criticism of the H-O model is not new in international trade literature. Earlier studies such as those of Leontief (1953, 1956),[24], [25] Tatemoto and Ichimura (1959),[26] Baldwin (1971),[27] Heller (1976),[28] and Trefler (1995),[29] have shown that the theory is inconclusive empirically. Most of these studies examined the role of labour and capital endowments in shaping trade patterns. None of them focused on the role of natural resources, especially water.

Most of the paradoxes were attributed to the rigid assumptions in the model. International trade in reality is guided by a host of variables other than just factor endowments. While supply-side bias determined by cost of production was identified as the basis of comparative advantage in the H-O model, in the new trade theories, expounded by Krugman (1979)[30] and Helpman (1981),[31] demand, returns to scale, and imperfectly competitive markets begin to play a more ubiquitous role in determining the comparative advantage of countries. This is because international trade has changed remarkably from the time of the mercantilists and classical economists. With technological advancements and proliferation of information technology, production networks have become more integrated. Countries are no longer required to specialise in the production of an entire product, and can engage their resources only in production along a specific segment of the entire value chain. This has led to an increase in the share of intra-industry trade across the world with companies often engaging in strategic competition to capture one another’s market shares.[32]

Subsequently, the role of factor endowments was further diminished as interventions in research and development (R&D) and innovation of differentiated products emerged as new sources of comparative advantage. Increasing incomes also led to an increase in demand for products, and a change in the composition of demand.[33] Linder (1961)[34] found that countries with similar demand structures, represented by per capita income, were more likely to engage in international trade with one another.

Additionally, in the last two decades, especially since the formation of the World Trade Organisation (WTO) in 1995, there has been a proliferation of preferential or free trade agreements which may have led to different kinds of trade creation or trade diversion.[35] Trade agreements are in essence discriminatory to those countries which are not signatories to the treaty, and favourable to its members. If, as a result of a trade agreement, trade shifts from a low-cost producer to a high-cost producer because the latter is a member while the former is not, it leads to diversion of trade from its most efficient source. Such effects play an important role in determining trade patterns across the globe.

Therefore, even in the context of virtual water trade it is paramount that these alternative theories and factors are examined. Given the importance of the agriculture sector in the consumption of global water resources, it is necessary to analyse how agriculture trade across countries is influenced by these parameters. According to recent developments in international trade theory, the unit of analysis is no longer the country but the industry and individual firms, because countries do not trade, firms do.

This is important for analysing virtual water trade in agricultural products as well because the agriculture sector comprises individual units that are heterogeneous and have varying degrees of productivity. Growth in agricultural factor productivity has a positive influence on the agricultural exports of a country.[36] Investments in R&D to enhance agricultural water productivity can result in domestic agricultural output becoming more efficient in the use of its water resources. Additionally, agriculture production and exports are also supported by domestic policies and international trade agreements. An examination of the interaction of these factors with the relative water endowments of a region will provide a more holistic understanding of virtual water trade patterns.

3. Empirical Analysis

3.1 Global Trends in Virtual Water Trade

Virtual water trade has the potential to help ameliorate water scarcity across the world.  This section examines previous studies that have estimated virtual water trade patterns at the global and cross-country levels, and superimposes these patterns on the relative water endowments of the regions under consideration.

Most of these studies have focused on the trade in agricultural products, but a few have also included livestock and industries. Between 1986 and 2011 international trade in agricultural products and associated virtual water trade almost tripled.[37] Changes in the recent history of agricultural trade include the ever increasing presence of China as a food importer, particularly from South America; the increase of soybean exports from Brazil and Argentina to Southeast Asia; and the escalating exports of palm oil from Indonesia and Malaysia to China, India, Pakistan, and Europe. Network analysis of temporal virtual water trade has shown that countries like the US, Brazil, Argentina, India and Australia have been net exporters while Germany, Italy, Russia, Japan, and the Middle East countries are net importers.[38] This has led to the emergence of a few countries as key suppliers of agricultural products across the globe.

Cereals, meat, fats and oilseeds constitute more than 50 percent of the virtual water trade across the globe.[39] Figure 2 depicts the pattern of global virtual water trade from 1996 to 2005.[40] Recent estimates of virtual water trade suggest that the network of exporters and importers remains similar (refer to Appendix A1).[41] Therefore, using data from Hoekstra and Hung (2002) for the period 1996-2005,[f] this paper provides indicative evidence of the implications of the virtual water trade on water scarcity across the globe.  Figure 2 represents some of the major net virtual water importing countries. Virtual water imports and exports are classified as green, blue and grey water flows. This categorisation follows from the ‘water footprint’ concept developed by Arjen Hoekstra in 2002.[g] The green water footprint refers to the volume of rainwater or the moisture in the soil used in the production, especially of crops. The blue water footprint is the volume of surface and ground water used, and the grey water footprint is the volume of freshwater required to neutralise the load of pollutants, based on existing ambient water quality standards.

Figure 2

Net virtual water imports (billion cubic metres)

Source: Mekonnen and Hoekstra (2011)[42]

The two graphs in Figure 3 (Panels A and B) present the relationship between net virtual water exports, net virtual water imports, and the per capita renewable water resources in the same countries as in Figure 2, based on data from Mekonnen and Hoekstra (2011). Renewable fresh water resources per capita is representative of the water endowment of these countries.[43],[h],[i]

Figure 3

Net virtual water exports and imports[44] and renewable internal fresh water resources per capita[45]

Source: Authors’ calculations

Panel A of Figure 3 represents the top 30 net virtual water importers over the period 1996-2005. Most of the countries represented in the bottom left hand corner of the scatter plot are from the MENA region. Some of the net importers of cereal among them were Israel, the United Arab Emirates, Jordan, Libya, Saudi Arabia, Tunisia, Algeria and Yemen. These countries have been experiencing historical water scarcity. Other major net virtual water importers such as Japan, Italy, Germany, and Mexico too have very low per capita renewable fresh water resources. Thus, the decision to import virtual water has a strong negative relationship with the availability of water in the country. This result has been ratified in a study on water intensive cereals’ trade across the world.[46] It was observed that countries with per capita water availability below a threshold of 1,500 m3 per capita per year were net importers of cereals which account for a significant amount of crop water requirement.

In Panel B of Figure 3, the top 30 net virtual water exporters have been mapped against their respective per capita renewable water availability. Although there does exist a cluster of countries which have low per capita water resources and yet export virtual water, such exports are relatively small. Further, the relative per capita availability of water in most of the top net exporting countries is also higher than in the net importing countries. However, India, despite being one of the highest net virtual water exporting nations, has very low per capita water resource availability. The per capita availability of water in India is lower than that of several major virtual water importing nations like Mexico, Japan and Italy. This has further implications for water scarcity in India.

3.2 Implications for Water Scarcity

With the exception of India, most net virtual water exporters have a relatively higher per capita availability of water resources than the net virtual water importing countries. This is a pointer to whether virtual water trade in agricultural products does ameliorate water scarcity. It can also be gauged from the changes in per capita water availability in these countries in the period 1992-2005. The Falkenmark indicator measuring the volume of water available per capita per year is a widely used index of water scarcity.[j],[47] According to Falkenmark et. al., (1989), 1,700 m3 per year is the threshold for water scarcity.[48] If water availability falls below this, it leads to a high level of competition among water users and causes water stress. If the per capita availability is below 1,000 m3 per year, the area experiences high water scarcity, and if it falls below 500 m3 per year, it has absolute scarcity. Most of the major virtual water exporting nations, except India, are above the Falkenmark threshold of 1,700 m3 per year. However, a declining trend in per capita water availability is visible, with countries like Argentina, the US and Thailand tending towards this threshold (figure 4).

Figure 4

Per capita renewable water resources availability among virtual water exporters (in thousand cubic metres)

Source: World Bank Development Indicators Database[49]

Further, as depicted in Table 1, with the exception of the MENA countries, the annual decline[k] in per capita water availability, or increase in per capita water scarcity, is relatively higher among the exporting nations than among the importers. As water-scarce countries import virtual water to manage their demand for water-intensive products, they add to the burden on the water resources of exporting nations. For example, in importing countries like Japan and Germany, the per capita water resource availability has remained much the same throughout the period 1992-2014. Since these two are among the top virtual water importers, it highlights how importing water-intensive commodities helps manage domestic water scarcity. But Israel, Jordan and Egypt have continued to experience rapid decline in per capita water availability despite their virtual water imports. No doubt it is possible that their per capita water availability would have been even lower in the absence of such imports.

As water-scarce countries import virtual water to manage their demand for water-intensive products, they add to the burden on the water resources of exporting nations

Table 1

Change in per capita water availability in major virtual water trading nations

Country Name Status in virtual water trade Change between 1992 and 2014 (%) CAGR (%)
Belgium Net Importer -10 -2
China Net Importer -15 -3
Egypt Net Importer -35 -7
Germany Net Importer 0 0
Israel Net Importer -38 -8
Italy Net Importer -7 -1
Japan Net Importer -2 0
Jordan Net Importer -55 -13
Mexico Net Importer -28 -5
Spain Net Importer -16 -3
United Kingdom Net Importer -11 -2
Argentina Net Exporter -21 -4
Australia Net Exporter -25 -5
Brazil Net Exporter -24 -4
Canada Net Exporter -20 -4
India Net Exporter -30 -6
Indonesia Net Exporter -26 -5
Thailand Net Exporter -16 -3
United States of America Net Exporter -19 -4

Source: Authors’ calculations using World Bank data

Thus, there is evidence of decline in water scarcity in some countries that are net importers, excluding those from MENA. A statistically significant t-test statistic suggests that the decline in water resource availability is higher among net exporters as compared to net importers.[l]

Literature on the virtual water trade points to the growing importance of a few countries as the source of water-intensive products for a larger group of importing countries.[50] It is pertinent to examine the nature of the goods being exported from these few countries that have emerged as the source of water-intensive agricultural products. There is yet no empirical evidence to establish the hypothesis that water-intensive products are exported from relatively water-abundant regions or countries. For example, studies analysing the domestic trade of water-intensive agricultural products in India and China have shown that virtual water flows occur from relative less water endowed areas to relatively better water endowed areas.[51], [52] Indian exports of virtual water embedded in agricultural products despite very low per capita water availability highlights that factors other than water resource endowments also play a role in virtual water trade. However, there seems to be a negative correlation between net virtual water imports and water resources availability. This is true mostly for the top importers, and might represent the fact that these countries face acute water shortage or are short of arable land, and thus have to import agricultural commodities for food.

3.3 Commodity Analysis of Net Virtual-Water Exporting Countries

As highlighted in the earlier section, a few countries like India, Argentina, the US, Australia and Brazil are significant exporters of virtual water. Some of the products traded are cereals, cotton, oil seeds and meat. This section analyses the patterns of export of these commodities from these countries in 2010 and in 2018. So far it has been seen that virtual water imports have helped the major importing nations ameliorate water scarcity to some extent, while the exporting countries have seen a gradual increase in their relative water scarcity levels (see section 3.2). With the volume of virtual water trade rising, the link between production and consumption across countries has intensified. As Figure 5 points out, cereals, meat and oilseeds are some of the major commodities through which virtual water trade takes place. 

Figure 5

Major commodities in global virtual water trade

Source: P D’ Odorico et. al. (2019)[53]

The share of some of these commodities in the total value of merchandise exports of Brazil, Argentina, Australia, India, the US, Canada, Thailand and Indonesia is given below (Figures 6 and 7).

Figure 6

Share in value of merchandise exports (%)

Source: Authors’ own calculations (based on WITS database)[54]

Figure 7

Share in value of merchandise exports (%)

Source: Authors’ own calculations (based on WITS database)[55]

As depicted in Figures 6 and 7, in both 2010 and 2018, the share of cereals and oilseeds in the value of total merchandise exports was relatively small compared to that of machinery, transport, minerals and chemicals. Only Argentina and Brazil were exceptions, whose cereals and oilseeds, respectively, showed a sharp increase in their share of the total value of their merchandise exports between 2010 and 2018. Cereals and oilseeds are important because they account for a significant share of the total virtual water trade across the world (Figure 5). Additionally, these are also some of the most water intensive commodities.[56] Yet, their contribution to the total value of merchandise exports is not concomitant and indicates low economic returns on the water used in the production process. This could perhaps be due to the relative undervaluation of water resources used in the production process compared to other factors of production. For example, in the US, cereal exports are the highest (Figure 8), but its corresponding share in value of merchandise exports is negligible.

Figure 8

Total export volume (in million tonnes)

Source: Authors’ own (Based on FAOSTAT)[57]

No doubt, in Brazil, the share of oilseeds in the total value of merchandise exports increased substantially between 2010 and 2018, but this was alongside an increase in the total volume of exports. Indeed, it is clear that the rise of the share of oilseeds in the total value of exports was due to the increase in the volume of exports, and not due to any rise in the price of oilseeds, since during the same period, the vegetable oil price index experienced a slump (Figure 9), as seen from the Primary Commodity Price estimates of the International Monetary Fund.

Further, it is seen that there is a decline in the global price levels of most commodities, except meat. Brazil, the US, Australia and Canada are the largest exporters of meat products. Although the total volume of meat exports is very small compared to the exports of cereals and oilseeds from these countries, the share of meat in the total value of exports is relatively high. Brazil and Australia, in particular, have a high share of meat in the total value of their merchandise exports. The export of meat products was around 4 million tonnes and 2 million tonnes from Brazil and Australia, respectively, in 2018, while that of cereals was approximately 20 million tonnes. Yet the share of cereals (in the total value of merchandise exports) from Brazil was 1.9 percent, while that of meat was 5.5 percent. The share of cereals in Australia’s exports was 1.8 percent and that of meat is 3.9 percent. Thus, despite a much lower volume of exports, the corresponding share of meat products in total value of exports was much higher. This is significant because both meat and cereals are water intensive commodities and have very high water footprints.

Figure 9

Global price index

Source: International Monetary Fund[58]

Competitiveness of export commodities is determined by the relative price of the commodity. The cheaper the product, the more competitive it is in the export market. This could be another explanation for the relatively lower share of cereals and oilseeds in the total value of merchandise exports. Domestic policies to support agro-based industries, in the form of subsidies, help to keep the price of these commodities artificially low. This has been a major bone of contention at the WTO, especially between countries like the US, China and India.[59]

Such policies also promote excess production which can have detrimental consequences on the natural environment, especially water resources. For example, in Brazil, more than two-thirds of the total Producer Support Estimates (PSE) to farmers is in the form of trade distorting price support and credit subsidies.[60] Similarly, in the US, a large share of farm support accrues to the oilseeds sector, especially soybean.[61] This is reflected in the increase in the volume of oilseed exports, and a decline in the share of oilseeds in the total value of merchandise exports. It is also estimated that farm subsidies are important for agricultural exports in the US. A 1 percent decrease in subsidies would reduce farm exports from the US by 0.40 percent per annum.[62] In India, virtually all of the budgetary transfers to the agricultural sector are in the form of subsidies for variable input use – especially fertilisers, electricity and irrigation water.[63]

With per capita water availability declining in the virtual water exporting nations, and already being extremely low in India, policies supporting agricultural trade have major implications for the sustainability of water resources in these countries. A re-examination of the Heckscher-Ohlin model to explain virtual water trade is needed. The next section identifies factors other than water endowments which drive virtual water trade. Policy makers and academics can use these factors, in conjunction with the availability of water resources, to model the impact of the trade in alleviating water scarcity, and as a tool in water resources management.

4. Factors beyond Water Endowments

Water is an important input in agricultural production. Agricultural or agro-based exports such as cereals, oilseeds, and meat, account for most of the virtual water trade. Allan (1993) pointed out that water endowments play an important role in determining virtual water trade.[64] But there are also other factors.[65] As the analysis in the earlier sections  suggests, water endowments alone cannot adequately explain virtual water trade patterns.[66], [67] Summarising the findings of several studies, the drivers of virtual water trade can be identified as: the Gross Domestic Product (GDP) of the trading countries, their population, the precipitation per capita in these countries, geographical distance between trading partners,[68] agricultural production of the exporting countries,[69] the extent of cropped land[70], [71] and the water footprint.[72] Among these factors, demand for food products captured by the GDP and population were identified as the most significant ones. This was followed by rainfall, irrigated or cropped land, and water footprint of production.[73]

Availability of cropped or irrigated land is an important factor in determining virtual water trade patterns. The more arable land a country has, the better it can use its water resources..[74] Higher amounts of arable land imply greater volumes of soil moisture. Soil moisture, also called green water, is a critical factor in agriculture production. Water resource availability per hectare of land is better at explaining existing patterns of virtual water trade, rather than absolute water resource availability.[75] It is observed that countries which are relatively water rich but land scarce, such as Portugal and Japan, are importers of virtual water. Despite being well endowed with water resources, limited land is a major constraint in agricultural production for them.[76] Gross cropped area has significant impact on water use efficiency and explains the strong positive relationship between gross cropped area per capita and virtual water export.[77] Similarly, in India, availability of arable land, and yield, were found to be the primary criteria for agriculture production decisions.[78]

As noted earlier, trade agreements ease market barriers for members but discriminate against non-members. With an unprecedented growth in free and preferential trade agreements, there has been an increase in the volume of trade.  Recent estimates by the Organisation for Economic Cooperation and Development (OECD) suggests that the share of trade in agricultural products due to the signing of regional trade agreements (RTAs) has been rising during 1998-2009.[79] Due to these RTAs, the preferential margins – the difference between the Most Favoured Nation’s (MFN) rate of duty and the preferential duty rates for members – has increased for most agricultural products. Therefore, despite the presence of non-tariff barriers in agricultural goods trade, members of RTAs found it easier to get access to one another’s markets compared to other suppliers. On average, a 1 percent reduction in preferential tariff generated a 2 percent increase in agricultural trade relative to other countries. Therefore, tariff preferences had a favourable impact in increasing agricultural trade flows for both pre-existing and new trade partners.[80]

With lower prices, the demand for imported agricultural goods will also rise, and thus lead to an increase in the volume of virtual water trade between members of trade agreements. This is particularly true in the case of the US and Mexico trade agreement for agriculture (which was part of the NAFTA[m]) introduced in 1994. It allowed the US to satisfy the rising demand for meat and cattle feed that emerged as Mexico’s per capita income rose in the 1990s. Similarly, China’s imports of soybeans from Brazil, the US and Argentina, increased significantly due to increase in demand and removal of trade restrictions on soy imports in 2000-01.[81] This pattern has shifted largely in favour of Brazil in recent times due to the trade war between the US and China, and is reflected in the increase in exports of oilseeds from Brazil.[82]

Further, the trend of certain water scarce countries like India exporting water intensive products to water productive nations has been interpreted as an illustration of countries’ disregard for water resources management relative to other inputs, such as labour and land, mainly due to the historical deficiency of incentives to regulate and manage water efficiently.[83] Despite being more water productive in certain agriculture products than their exporting counterparts, Austria, Britain, and the Netherlands resorted to water imports. Austria and the Netherlands’ decision to import is attributed to the lack of adequate land resources, while Britain’s historical emphasis on industries over agriculture is cited as the primary reason behind the imports.[84]

The trend of certain water scarce countries like India exporting water intensive products to water productive nations has been interpreted as an illustration of countries’ disregard for water resources management relative to other inputs, such as labour and land, mainly due to the historical deficiency of incentives to regulate and manage water efficiently

In certain instances, farmers in developing countries are incentivised by international agribusiness firms to produce crops specifically designed for exports. These firms take advantage of the cheap labour and land in these regions to meet the demands of the global North. This leads to a switch from local subsistence-based farming to commercial farming. Although remunerative for farmers, this also shows that developing regions of the world lack optimal water management policies, due to which, despite being water scarce, they end up exporting virtual water to the more water productive developed nations.[85]

Existing patterns of virtual water trade are thus the result of various factors – lack of efficient pricing and market mechanisms, differences in productivity of land and water across different sectors in different regions, government policies, availability of land vis-a-vis water resources, trade relations between countries, demand for particular products, labour costs, and capital and knowledge.  The virtual water trade, at times, cannot be sufficiently explained only on the basis of relative water endowments within the framework of the H-O model. In their critique of the concept of virtual water, Erik Gawel and Kristina Bernsen have commented that by neglecting the role of other factors of production, virtual water flows fail to capture all the information necessary for sound policy advice. However, questions can be raised about their conclusion: that “both international environmental and trade policies should refrain from drawing upon virtual water calculations and their implied and often misleading policy implications”.[86] The present analysis has sought to show that virtual water trade can be a useful tool for management of water resources.

Conclusion

This paper provides evidence of the high volume of virtual water exports from a highly water-stressed country like India, and the increasing decline in per capita availability of water in key virtual water-exporting nations. At the same time, virtual water-importing nations like Italy, Japan and Germany have seen little change in their per capita water resource availability between 1992 and 2014. Therefore, by resorting to virtual water imports, these countries have been able to solve part of their domestic water management problem and avoided a decline in the per capita availability of water. However, for the big virtual water-exporting countries like India, Indonesia, Brazil, Australia and Canada, the decline in per capita water availability has been more rapid over the same period. It is alarming for India, whose per capita water availability level is already down to the Falkenmark threshold of 1,700 m3 per capita per year. The other water-exporting countries are yet to reach this point, but their trend of rapid decline in per capita water availability indicates an immediate need to implement efficient water management policies.

Virtual water trade can still be a sound water management tool. However, it is essential that there be markets for water resources. As highlighted earlier, the Heckscher-Ohlin model is based on the assumption of a perfect market for the factors of production. For global, regional and intra-regional virtual water trade patterns to reflect the comparative advantage of countries in terms of water resources, the scarcity value of water must be reflected in the price of water. Only then can water endowments be reflected directly in the factor price of water and comparative advantage be determined. Water use in agriculture, industry, and livestock production, must consider the associated economic and ecological costs. Institutional support for water use must not be at the cost of ecology, as has been observed in many parts of the world. This in turn has resulted in a false perception regarding the actual water availability scenario and has artificially reduced its scarcity value. Subsequent trade in virtual water embedded in agricultural products has led to harmful consequences for the exporting countries, most immediate of which is a gradual decline in per capita water availability. Therefore, a key learning from the present analysis is the need for markets for water resources, especially in commercial sectors like agriculture and industry. In this context, the September 2020 announcement of the launch of the NASDAQ Velles California water futures market is a positive development.[87] Virtual water trade in agriculture, in accordance with the scarcity value of water, can be then employed as a tool in IWRM. Further, institutional support must be reoriented towards the production of crops suitable to a region, and support adoption of production techniques that do not put unnecessary pressure on the natural resources of the region.

The key to ensuring that virtual water trade relative endowments of water resources is deeply ingrained in the economic concept of markets. Markets for natural resources like water are not as prevalent as those for other factors of production. This results in a skewed representation of variables like land, labour, and electricity, in the costs of production relative to water. Water is nonetheless a ubiquitous factor in agricultural production and an important component of agricultural trade globally. Moreover, trade between countries is also influenced by trade agreements, geopolitics, and demand. Along with these factors, virtual water trade, in the presence of markets for water resources, can play a better role in optimising the use of scarce water resources across the world.


Appendices

Appendix A1

Global Pattern of Virtual Water Trade 1986-2011

Countries shown in purple are net virtual water exporters, and those in pink are net virtual water importers of agriculture products. As evident from the figure, the volume of virtual water trade has increased manifold during the period, although the major exporting and importing nations remain the same.

Source: P D’ Odorico et. al. (2019)

Appendix A2

The t-test: Paired two samples for means

Source: Authors’ own

Variable 1 refers to the sample of rate of decline in per capita water resources of net importers (excluding Israel, Egypt and Jordan), and Variable 2 refers to the same for the sample of net exporters.


Endnotes

[a] The term was coined by John Anthony Allan in 1993.

[b] In IWRM, water is viewed as an integral component of the global hydrological cycle and not as a stock of material resource to be utilised solely for the satisfaction of human needs.

[c] The company that owns the NASDAQ stock exchange.

[d] Due to factors like arid climate, for example.

[e] In a hypothetical two-good, two-factor economy, comprising, say, computers and textiles, and labour and capital, if the factor price of capital is higher than that of labour, the economy is considered to be relatively abundant in labour and scarce in capital. This pattern of factor abundance will then reflect in the relative price of the products. Production of computers can be assumed to be relatively more capital intensive, while that of textiles is relatively labour intensive.

[f] This was the most recent data on virtual water flows available.

[g] Water footprint is a measure of human appropriation of water resources.

[h] Per capita renewable internal freshwater resources in figure 3 refers to the average of such resource measured in the years 1992, 1997, 2002 and 2007 in the respective countries. Average figures have been taken because the virtual water export/import data is an aggregate value for the period 1996-2005. Using per capita renewable internal fresh water resources data for a single year would render the analysis incompatible.

[i] Compiled from the World Development Indicators database of the World Bank

[j] It is the ratio of total volume of water available within a domain to the number of people living there.

[k] The Compound Annual Growth (or decline) Rate is being considered.

[l]Refer to Appendix A2.

[m] North American Free Trade Agreement; it was replaced by the US-Mexico-Canada agreement in July 2020

[1]Mesfin M. Mekonnen and Arjen Y. Hoekstra, “Four billion people facing severe water scarcity,” Science Advances Vol.2 (2) (February, 2016).

[2]World Economic Forum, The Global Risks Report 2020, 15th edition, World Economic Forum, Geneva, Switzerland (January, 2020).

[3]Wendy Barnaby, “Do nations go to war over water?” Nature 458 (2009).

[4] John Anthony Allan, “The Water, Food and Trade Nexus. Useful Concept or Misleading Metaphor?” Water International 28, 2003.

[5] John Anthony Allan, “Fortunately there are Substitutes to Water Otherwise our Hydro-political Futures would be Impossible,” in Priorities for Water Resources Allocation and Management (London, United Kingdom: ODA, 1993)

[6]B.J. Dermody et al., “A virtual water network of the Roman world,” Hyrology and Earth System Sciences 18 (December, 2014).

[7]Dermody et al., “A virtual water network of the Roman world”

[8] B.J. Dermody et al., “A virtual water network of the Roman world,” Hyrology and Earth System Sciences 18 (December, 2014).

[9] Soumya Bhowmick, Nilanjan Ghosh and Roshan Saha, “Tracking India’s Progress in Clean Water and Sanitation: A Sub-National Analysis,” ORF Occasional Paper 250 (June, 2020).

[10] Lionel Robbins, An Essay on the Significance of Economic Science, (London: Macmillan and Company Ltd., 1932).

[11] Ryan Dezember, “CME and NASDAQ to Launch Water Futures Contract, The Wall Street Journal, September 23, 2020.

[12] Dik Roth and Jeroen Warner, “Virtual Water: Virtuous Impact? The unsteady state of virtual water,” Agriculture and Human Values 25, 2008.

[13] Nilanjan Ghosh, “Three propositions on transboundary water conflicts in light of the new emerging paradigms,” ORF Expert Speak (July 14, 2020).

[14] David Ricardo, On the Principles of Political Economy and Taxation, (London: John Murray, 1817).

[15] Eli Heckscher, “The Effect of Foreign Trade on the Distribution of Income.” EconomiskTidskrift XX1 (1919)

[16]Heckscher, “The Effect of Foreign Trade on the Distribution of Income.” 

[17]Bertil Ohlin, Interregional and International Trade. (Cambridge: Harvard University Press, 1933)

[18] Ronald W. Jones, “Factor Proportion and the Heckscher-Ohlin Theorem,” Review of Economic Studies 24, No.1 (1956-57)

[19] P. Rogers, “Water in a future Middle East,” in Water in the Arab World, eds. P. Rogers and P. Lydon (Cambridge, Massachusetts: Harvard University Press, 1994).

[20] Nilanjan Ghosh and Jayanta Bandyopadhyay, ” A scarcity value based explanation of trans-boundary water disputes: the case of the Cauvery River Basin in India,” Water Policy 11 (2009).

[21]Ghosh and Bandyopadhyay, ” A scarcity value based explanation of trans-boundary water disputes: the case of the Cauvery River Basin in India.”

[22]Ghosh and Bandyopadhyay, “A scarcity value based explanation of trans-boundary water disputes: the case of the Cauvery River Basin in India.”

[23] Arjen Y. Hoekstra, “The relation between international trade and freshwater scarcity,” WTO Staff Working Paper (2010).

[24] Wassily Leontief, “Domestic Production and Foreign Trade: the American Capital Position Re-examined,” Proceedings of the American Philosophical Society 97, September 28, 1953.

[25] Wassily Leontief, “Factor Proportions and the Structure of American Trade: Further Theoretical and Empirical Analysis,” The Review of Economics and Statistics 38, November, 1956.

[26] M. Tatemoto and S. Ichimura, “Factor proportions and foreign trade: the case of Japan,” Review of Economics and Statistics 41, 1959.

[27] Robert Baldwin, “Determinants of the Commodity Structure of U.S. Trade,” American Economic Review 61, March1, 1971.

[28] Peter S. Heller, “Factor endowment change and comparative advantage: the case of Japan,” The Review of Economics and Statistics 58, August, 1976.

[29] Daniel Trefler, “The Case of the Missing Trade and Other Mysteries,” American Economic Review 85, 1995.

[30] Paul Krugman, “Increasing returns, monopolistic competition, and international trade,” Journal of International Economics 9, November, 1979.

[31] E. Helpman, “International trade in the presence of product differentiation, economies of scale and monopolistic competion: a Chamberlain-Heckscher-Ohlin approach,” Journal of International Economics 11, 1981.

[32] H.G. Grubel and P.J. Lloyd, Intra-Industry Trade: The Theory and Measurement of International Trade in Differentiated Products. (London: Macmillan Press, 1975).

[33]Kym Anderson, ” Globalization’s effects on world agricultural trade: 1960-2050,” Philosophical Transactions of the Royal Society of London B Biological Science (September, 2010).

[34] Staffan Burenstam Linder, An Essay on Trade and Transformation, (Uppsala: Almqvist and Wiksells, 1961).

[35]Jacob Viner, The Customs Union Issue, Carnegie Endowment for Intellectual Peace (New York, 1950).

[36]Munisamy Gopinath and P. Lynn Kennedy, “Agricultural Trade and Productivity Growth: A State-Level Analysis,” American Journal of Agricultural Economics 82 (2000).

[37] Paolo D’Odorico et al., “Global virtual water trade and the hydrological cycle: patterns, drivers and socio-environmental impacts,” Environmental Research Letters 14 (2019).

[38] D’Odorico et al.,.”Global virtual water trade and the hydrological cycle: patterns, drivers and socio-environmental impacts.”

[39]D’Odorico et al.,.”Global virtual water trade and the hydrological cycle: patterns, drivers and socio-environmental impacts.”

[40]M. M. Mekonnen and Arjen Y. Hoekstra, National Water Footprint Accounts: The Green, Blue and Grey Water Footprint of Production and Consumption, Volume 1: Main Report, Value of Water Research Report Series No. 50 (IHE Delft: The Netherlands; May 2011)

[41] D’Odorico et al.,.”Global virtual water trade and the hydrological cycle: patterns, drivers and socio-environmental impacts.”

[42]M. M. Mekonnen and Arjen Y. Hoekstra, National Water Footprint Accounts: The Green, Blue and Grey Water Footprint of Production and Consumption, Volume 2: Appendices, Value of Water Research Report Series No. 50 (IHE Delft: The Netherlands; May 2011)

[43]World Development Indicators, Renewable internal freshwater resources per capita (cubic metres), Washington D.C.: The World Bank.

[44] Author’s calculation based on data from Mekonnen and Hoekstra (2011).

[45] World Development Indicators, Renewable internal freshwater resources per capita (cubic metres), Washington D.C.: The World Bank.

[46]Hong Yang, Peter Reichert, Karim C. Abbaspour and Alexander J. B. Zehnder, “A Water Resources Threshold and Its Implications for Food Security,” Environmental Science and Technology 37 (2003).

[47]J. Liu et. al., “Water scarcity assessments in the past, present and future,” Earth’s Future 5, 2017.

[48]M. Falkenmarket. al., “Macro-scale water scarcity requires micro-scale approaches, Natural Resources Forum 13, 1989.

[49]World Bank, “World Development Indicators,”Washington, D.C. The World Bank.

[50]Joel A. Carr, Paolo D’Odorico, F. Laio and L. Rudolfi, “Recent History and Geography of Virtual Water Trade,” PLOS One 8, 2013.

[51]  Shilp Verma, D.A. Kampman, P, van der Zaag and Arjen Y. Hoekstra, Going against the flow: A Critical Analysis of virtual water trade in the context of India’s National River Linking Programme. Value of Water Research Report Series No. 31 (IHE Delft: The Netherlands, March 2008).

[52] Jing Ma, Arjen Y. Hoekstra, Hao Wang, Ashok K. Chapagain and Dangxian Wang, “Virtual versus real water transfers within China,” Philosophical Transactions of the Royal Society 361, October 2005.

[53]Paolo D’Odorico et al., “Global virtual water trade and the hydrological cycle: patterns, drivers and socio-environmental impacts,” Environmental Research Letters 14, 2019.

[54]World Bank, World Integrated Trade Solution (WITS).

[55]World Bank, World Integrated Trade Solution (WITS).

[56]M. M. Mekonnen and Arjen Y. Hoekstra, National Water Footprint Accounts: The Green, Blue and Grey Water Footprint of Production and Consumption, Volume 2: Appendices, Value of Water Research Report Series No. 50 (IHE Delft: The Netherlands; May 2011)

[57]Food and Agriculture Organisation, FAOSTAT Statistical Database, Rome: FAO.

[58]International Monetary Fund, “Primary Commodity Price System,” IMF Data Access to Macroeconomic and Financial Data.

[59]Dipti Jain, “Why do India and US disagree on farm subsidies? Livemint, August 16, 2018.

[60]OECD, “Brazil- Agricultural Policy Monitoring and Evaluation“.

[61]Mike Dorning, “Trump’s $28 Billion Trade War Bailout is Overpaying Farmers, Bloomberg, December 4, 2019.

[62]Lanh Anh Tong, “The Effects of Farm Subsidies on Farm Exports in the United States, American Journal of Agricultural Economics 101, no. 4, 2019.

[63]OECD and ICRIER, Executive Summary to Agricultural Policies in India, OECD Food and Agricultural Reviews, OECD Publishing, Paris, 2018.

[64]J. A. Allan, “Fortunately there are substitutes for water: otherwise our hydro-political futures would be impossible,” In: ODA, Priorities for water resources allocation and management, ODA, London, 1993.

[65]Dennis Wichelns, “Virtual Water: A helpful perspective, but not a Sufficient Policy Criterion,” Water Resources Management 24, 2010.

[66]A. Fracasso, ” A gravity model of virtual water trade,” Ecological Economics 108, 2014.

[67]H. Chouchaneet. al., “Virtual water trade patterns in relation to environmental and socio-economic factors- A case study of Tunisia,” Science of the Total Environment 613, 2018.

[68]S. Tameaet. al., “Drivers of virtual water trade,” Water Resources Research 50, 2014.

[69]R. Wang et. al., “(Virtual) Water flows uphill toward money,” Environmental Science and Technology 50, 2016.

[70]Dinesh M. Kumar and O.P. Singh, “Virtual Water in Global Food and Water Policy Making: Is There a Need for Rethinking?” Water Resources Management 19, 2005.

[71]Charlotte de Fraiture, Ximing Cai, Upali Amarasinghe, Mark Rosegrant and David Molden, Does International Cereal Trade Save Water? The Impact of Virtual Water Trade on Global Water Use. Comprehensive Assessment Research Report 4.(Comprehensive Assessment Secretariat: Colombo, Sri Lanka; 2004).

[72]I. K. El-Gafy, “System dynamic model for crop production, water footprint and virtual water nexus,” Water Resources Management 28, 2014.

[73]Paolo D’Odorico et al., “Global virtual water trade and the hydrological cycle: patterns, drivers and socio-environmental impacts,” Environmental Research Letters 14, 2019.

[74]Dinesh M. Kumar and O.P. Singh, “Virtual Water in Global Food and Water Policy Making: Is There a Need for Rethinking?” Water Resources Management 19, 2005.

[75]Esther Delbourg and Shlomi Dinar, “The globalisation of virtual water flows: Explaining trade patterns of a scarce resource,” World Development 131, 2020.

[76] Dinesh M. Kumar and O.P. Singh, “Virtual Water in Global Food and Water Policy Making: Is There a Need for Rethinking?” Water Resources Management 19 (2005).

[77]Kumar and Singh, “Virtual Water in Global Food and Water Policy Making: Is There a Need for Rethinking?”

[78]Government of India, “Natural resource aspects of sustainable development in India”, paper presented at 5th Session of the United Nations Commission on Sustainable Development, New Delhi, India, 1997.

[79] OECD, “Regional trade agreements and agriculture,” OECD Food, Agriculture and Fisheries Papers No. 79, OECD Publishing, Paris, 2015.

[80]“Regional trade agreements and agriculture”

[81] Carole Dalin et al., “Evolution of the global virtual water trade network,” PNAS 16 (April, 2012).

[82]Mike Dorning, “Trump’s $28 Billion Trade War Bailout is Overpaying Farmers, Bloomberg, December 4, 2019.

[83]Esther Delbourg and Shlomi Dinar, “The globalisation of virtual water flows: Explaining trade patterns of a scarce resource,” World Development 131 (2020).

[84] Delbourg and Dinar, “The globalisation of virtual water flows: Explaining trade patterns of a scarce resource.”

[85]Nigel Key and David Runsten, “Contract Farming, Smallholders, Rural Development in Latin America: The Organisation of Agroprocessing Firms and the Scale of Outgrower Production,” World Development 27 (1999).

[86]  Erik Gawel and Kristina Bernsen, “What is wrong with Virtual Water Trading? On the Limitations of the Virtual Water Concept,” Environment and Planning C: Government and Policy 37 (2013).

[87] Ryan Dezember, “CME and NASDAQ to Launch Water Futures Contract, The Wall Street Journal, September 23, 2020.

The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.

Authors

Roshan Saha

Roshan Saha

Roshan Saha was a Junior Fellow at Observer Research Foundation Kolkata under the Economy and Growth programme. His primary interest is in international and development ...

Read More +
Preeti Kapuria

Preeti Kapuria

Preeti Kapuria was a Fellow at ORF Kolkata with research interests in the area of environment development and agriculture. The approach is to understand the ...

Read More +

Contributors

Roshan Saha

Roshan Saha

Preeti Kapuria

Preeti Kapuria