This study analyzed direct and indirect relationships in water consumption by Macedonian economic sectors using virtual water in an input–output framework. Macedonia was found to be characterized by intensive water consumption, with some sectors having a significant virtual water content in their products. Virtual water multipliers were used to analyze the trade balance and determine whether national commercial trade strategies are in line with Heckscher–Ohlin (HO) theory. It was found that Macedonia trade strategies in terms of virtual water were generally in line with HO theory. However, as a consequence of significant exports and high virtual water content in vegetables, fruit, grapes and sheep and goat products within the agriculture sector and in food and basic metal products in the manufacturing sector, the region was a net exporter of virtual water, losing about 124 million m3 of water at 2005 level or 18% of total water consumption. Reducing exports of the most water-intensive products with significant net positive exports would result in substantial water savings of 42% of total water consumption. The results presented here can help policy makers in promoting production specializations that are more environmentally sustainable or in redesigning existing water pricing policies at national level to encourage rational use of water.

INTRODUCTION

Fresh water is an irreplaceable, limited natural resource and an essential factor for existence and for social and economic development of society. Water scarcity, arising mainly because of mismanagement of water resources, is becoming a severe problem. Wang et al. (2009, p. 894) noted that ‘promoting sustainable development of water resources must involve a consideration of the interactions between water use and the economic sectors’. To mitigate water scarcity and ensure proper allocation and management of the resources available, information on the economy-wide implications is needed.

The term ‘virtual water’ was first coined by Allan (1996). It refers to water that is ‘embodied’ during the production process. This embodied water comprises not only the physical content of water in the product, but also the water consumed during its production (Dietzenbacher & Velázquez 2007). Since this principle only concerns agricultural commodities, Allan (1998) extended the definition to other commodities and services. To investigate the importance of water in terms of exports and imports, Hoekstra (2003) proposed two broad categories of virtual water, which are suitable for this purpose, namely real and theoretical virtual water. Real water is the water actually used in domestic production, while theoretical virtual water refers to potential water that would have been used in the production in the country of destination. This concept has been applied in recent years to study international flows of virtual water (Lenzen & Foran 2001; Hoekstra & Hung 2003; Dietzenbacher & Velázquez 2007; Wang et al. 2009; Yu et al. 2010). It provides a rationale for water demand management, with different alternatives for reducing water use. It also provides an approach to investigate whether national trade strategies are in line with Heckscher–Ohlin (HO) theory, whereby only geographical regions with water abundance should export goods produced in a water-intensive way (Reimer 2012). The essence of HO theory is that ‘traded commodities are a bundle of factors, which through commercial trade become mobile and provide services to regions where they are scarce’ (Leamer 1995; Reimer 2014).

Macedonia is currently rich in freshwater resources, with a total estimated endowment of surface water and groundwater of around 6.4 and 0.9 billion m3, respectively (UNECE 2011). This gives a comparative advantage to the country in terms of endowment, but the country has a high Water Exploitation Index (WEI) of 34% (EEA 2009), meaning that it is similar to Spain and Italy when it comes to intensity. WEI has a threshold value of 20% indicating a transition from non-stressed to water scarce region. A severe situation occurs when the index is greater than 40%. Agriculture has been identified as a key water-consuming sector where high quantities of water embodied in agricultural products are incorporated into other sectors when used as intermediate products, or traded abroad.

The aim of this study was to assess the direct and indirect relationships in water consumption by various economic sectors in Macedonia by means of virtual water in an input–output framework and to investigate the sectors characterized by rational use of water. In addition, Macedonian trade strategies were assessed in terms of commercial trade in virtual water, in order to empirically test the validity of HO theory and provide deeper insights into the traded virtual water content and the associated factor intensity of water consumption. The agricultural sector was first disaggregated into seven crop and four livestock sub-sectors, and then the direct and indirect transfers between different economic sectors and commercial trade were analyzed in an environmentally extended input–output model framework. The input–output model provides a simplified way to track the ‘virtual water contents of all intermediate inputs to the virtual water content of the final product without the need to revert to the detailed stages of the production process’ (Ip et al. 2007, p. 1981). Since trading virtual water is an instrument to control and promote sustainable natural resource management and to reduce the pressure on the existing water resources (Velázquez 2007), this analysis was also intended to help identify water saving policies to ensure more rational use of water.

METHODS AND DATA

According to the basic Leontief (1936) model, the input–output model of an economy is represented as a set of linear equations which in matrix formulation is 
formula
1
where X is a vector of output quantities, (IA)−1 the Leontief inverse matrix (also denoted by L), with A representing the matrix of technical coefficients of intersectoral dependencies and F is the final demand vector, which includes categories such as household consumption, investment and exports. The elements in the Leontief inverse matrix (lij) indicate the extra direct and indirect increase in the output of sector i as a result of a unit extra final demand by sector j.
Given the definition of virtual water as the water necessary for, and directly consumed during, production, in order to adjust the input–output model for virtual water analysis, it is essential to define direct water input coefficients for each respective sector as follows: 
formula
2
Defining the row vector W as the annual direct sectoral water consumption (million m3), the new W* element may be interpreted as a technical coefficient in a traditional input–output model, i.e. it gives the share of water consumption per 1 million MKD Macedonian Denar of output produced (Yu et al. 2010).
By pre-multiplying the lij elements from Equation (1) by Equation (2), the virtual water multipliers (V) are obtained as follows: 
formula
3
where Ŵ* denotes a diagonal matrix with the elements of W* on the leading diagonal.

The elements of V (denoted by vij) indicate the quantity of total water (m3) directly and indirectly consumed by sector i for each additional unit increase in final demand by the respective sector j.

Considering the trade components in the input–output framework, it is possible to calculate and quantify the virtual water content associated with exports (Fe) and imports (M). 
formula
4a
 
formula
4b
The difference between Equation (4a) and (4b) defines the net export virtual water content, which allows the respective sector to be identified as a net exporter or importer of virtual water. The production technology outside the country is assumed to be similar to the one used domestically (Ip et al. 2007; Feng et al. 2012).
Owing to the trade in virtual water content, in the promotion of sustainable water management at national level, Dietzenbacher & Velázquez (2007) propose an alternative way of saving water by nullification of foreign trade. In the present study, in order to assess the trade strategies, the sectors that consumed the highest quantities of water and displayed significant net positive exports in terms of virtual water were nullified. The effect of export nullification was quantified by summing up the virtual water multipliers of the respective sectors obtained by Equation (4a). It is also important to investigate the effects on imports due to export nullification. The trade balance (exports minus imports) for sector j deteriorates by the respective Fe,j value. A chain reaction occurs because the export reduction affects production, with the decline in production then denoting fewer inputs from imports, which amounts to the following: 
formula
5
where mi are the direct import coefficients obtained in the same way as the technical coefficients a and water coefficients w, i.e. by dividing the import value by the total input value for each respective sector. The total deterioration in the Macedonian trade balance due to the reduction in exports of sector j was obtained as follows: 
formula
6
The methodology described above was applied to the Macedonian economy in 2005 by using the officially published national input–output table (SSO 2008). To improve the analysis and obtain a more accurate representation of the country's agricultural sector, it was broken down into 11 sub-sectors (see Table 1). The disaggregation was conducted by following the Lindberg & Hansson (2009) procedure and using several relevant sources of data, namely the Common Agricultural Policy Regionalised Impact (CAPRI) dataset for Macedonia (CAPRI 2013), EUROSTAT data on economic accounts in agriculture, the Farm Monitoring System (FMS) on farm performance and reports published by the State Statistical Office in Macedonia. FMS is an annual survey carried out by the National Extension Agency, in line with the EU Farm Accountancy Date Network principles. The agricultural water account was disaggregated using data on the irrigated area from the 2008 Annual Agricultural Report (MAFWE 2009), as well as combined figures on crop water requirements calculated specifically for Macedonia by Hoekstra & Hung (2003), and crop level estimates by Iljoski (1990). For the livestock sectors, the water requirements reported in Galev & Arsovski (1990) were used.
Table 1

Sectoral water use in Macedonia and associated virtual water content

iSector [1]Output (mil. MKD) [2]Water use (mil. m3) [3]Water use coefficient (m3/1 mil. MKD) [4]Virtual water multiplier (m3/1 mil. MKD) [5]
Cereals 6,383.31 26.11 4,090.35 4,588.99 
Rice 487.61 14.07 28,854.80 31,986.19 
Raw tobacco production 4,159.07 11.27 2,709.74 2,971.23 
Vegetables 17,016.58 41.98 2,467.01 3,525.80 
Fruits 3,728.33 23.65 6,343.31 7,988.57 
Grapes and wine production 5,505.97 49.82 9,048.36 12,595.49 
Other crops 3,318.59 43.73 13,177.28 13,956.73 
Cattle 7,554.43 25.59 3,387.56 8,917.93 
Pigs 2,696.69 5.02 1,863.27 3,878.17 
10 Sheep/goat 2,004.89 11.05 5,509.17 9,776.16 
11 Other livestock 2,811.32 2.81 1,000.29 2,079.90 
12 Forestry 1,492.08 0.00 0.00 199.79 
13 Fishing 72.34 0.00 0.00 1,125.04 
14 Mining and quarrying 5,807.57 161.33 27,778.39 28,461.38 
15 Other mining 3,639.12 26.43 7,262.74 8,896.20 
16 Food and beverages 30,950.86 103.37 3,339.88 6,292.08 
17 Tobacco products 8,975.01 0.40 44.46 1,413.84 
18 Textiles 2,986.92 0.45 151.66 495.99 
19 Wearing apparel 27,024.94 0.97 36.00 134.43 
20 Leather products 2,877.10 0.04 15.29 202.19 
21 Wood products 1,673.83 2.20 1,311.96 1,649.25 
22 Pulp and paper 5,155.36 0.08 16.10 230.40 
23 Printed matter 1,057.81 0.29 273.20 342.83 
24 Coke, refined petroleum 22,243.00 0.28 12.77 98.94 
25 Chemical products 5,785.97 4.71 814.73 1,345.31 
26 Rubber and plastic 4,307.92 0.08 17.41 274.16 
27 Other non-metallic product 6,574.01 1.46 222.09 1,321.57 
28 Basic metals 30,138.25 68.13 2,260.65 5,260.18 
29 Fabricated metal products 4,230.76 0.09 21.75 286.56 
30 Machinery and equipment 2,117.49 4.52 2,133.18 2,491.89 
31 Electrical machinery 5,157.20 42.81 8,301.59 8,727.21 
32 Medical/optical instruments 399.75 0.01 25.02 145.99 
33 Motor vehicles 1,815.25 0.17 95.30 486.18 
34 Other transport equipment 659.08 0.13 201.80 514.06 
35 Furniture 1,781.82 0.31 173.98 405.58 
36 Secondary raw materials 684.13 0.00 2.92 288.78 
37 Electrical energy 14,537.49 1.00 68.79 4,830.42 
38 Collected/purified water 2,649.43 0.00 0.00 614.62 
39 Construction work 42,405.96 1.00 23.58 474.27 
40 Services 253,812.42 8.40 33.10 342.56 
 Total  683.78   
 Average   3,327.24 4,490.42 
iSector [1]Output (mil. MKD) [2]Water use (mil. m3) [3]Water use coefficient (m3/1 mil. MKD) [4]Virtual water multiplier (m3/1 mil. MKD) [5]
Cereals 6,383.31 26.11 4,090.35 4,588.99 
Rice 487.61 14.07 28,854.80 31,986.19 
Raw tobacco production 4,159.07 11.27 2,709.74 2,971.23 
Vegetables 17,016.58 41.98 2,467.01 3,525.80 
Fruits 3,728.33 23.65 6,343.31 7,988.57 
Grapes and wine production 5,505.97 49.82 9,048.36 12,595.49 
Other crops 3,318.59 43.73 13,177.28 13,956.73 
Cattle 7,554.43 25.59 3,387.56 8,917.93 
Pigs 2,696.69 5.02 1,863.27 3,878.17 
10 Sheep/goat 2,004.89 11.05 5,509.17 9,776.16 
11 Other livestock 2,811.32 2.81 1,000.29 2,079.90 
12 Forestry 1,492.08 0.00 0.00 199.79 
13 Fishing 72.34 0.00 0.00 1,125.04 
14 Mining and quarrying 5,807.57 161.33 27,778.39 28,461.38 
15 Other mining 3,639.12 26.43 7,262.74 8,896.20 
16 Food and beverages 30,950.86 103.37 3,339.88 6,292.08 
17 Tobacco products 8,975.01 0.40 44.46 1,413.84 
18 Textiles 2,986.92 0.45 151.66 495.99 
19 Wearing apparel 27,024.94 0.97 36.00 134.43 
20 Leather products 2,877.10 0.04 15.29 202.19 
21 Wood products 1,673.83 2.20 1,311.96 1,649.25 
22 Pulp and paper 5,155.36 0.08 16.10 230.40 
23 Printed matter 1,057.81 0.29 273.20 342.83 
24 Coke, refined petroleum 22,243.00 0.28 12.77 98.94 
25 Chemical products 5,785.97 4.71 814.73 1,345.31 
26 Rubber and plastic 4,307.92 0.08 17.41 274.16 
27 Other non-metallic product 6,574.01 1.46 222.09 1,321.57 
28 Basic metals 30,138.25 68.13 2,260.65 5,260.18 
29 Fabricated metal products 4,230.76 0.09 21.75 286.56 
30 Machinery and equipment 2,117.49 4.52 2,133.18 2,491.89 
31 Electrical machinery 5,157.20 42.81 8,301.59 8,727.21 
32 Medical/optical instruments 399.75 0.01 25.02 145.99 
33 Motor vehicles 1,815.25 0.17 95.30 486.18 
34 Other transport equipment 659.08 0.13 201.80 514.06 
35 Furniture 1,781.82 0.31 173.98 405.58 
36 Secondary raw materials 684.13 0.00 2.92 288.78 
37 Electrical energy 14,537.49 1.00 68.79 4,830.42 
38 Collected/purified water 2,649.43 0.00 0.00 614.62 
39 Construction work 42,405.96 1.00 23.58 474.27 
40 Services 253,812.42 8.40 33.10 342.56 
 Total  683.78   
 Average   3,327.24 4,490.42 

mil. = million; 1 € = 61.5 MKD.

RESULTS

The main agricultural exports from Macedonia, such as wine, fruit, vegetables, tobacco and lamb, either require a lot of water for irrigation purposes or contain a high amount of embodied water. However, there is general unawareness of exactly how much water is embodied in these products and how much enters or exits the country due to commercial trade.

An overview of Macedonia's economic sectors, i.e. their respective output and water used to produce that output, the water technological coefficients and the total embodied virtual water, is provided in Table 1. The results reveal a huge variation in water use intensity. The average direct water coefficient was around 3,327 m3 per unit of output. Almost all agriculture sub-sectors (sectors 1–11 in Table 1) consumed high quantities of water per unit of output, generally well above the average. In particular, rice (2) and other crops (7) (this agricultural sub-sector mainly includes forage production and more especially alfalfa) required large amounts of direct water, around 28,855 and 13,177 m3 per unit of output, respectively. Water consumption for irrigation purposes in grape and wine production and fruit production and water use in sheep/goat production were of lower intensity (9,048, 6,343 and 5,509 m3 per unit of output, respectively), but should not be neglected, because these are major exports from Macedonia. Regarding the manufacturing and service sectors, most of these had small coefficients and well below average water consumption. However, a few had similar consumption to the highest water intensity agricultural sub-sectors, namely mining and quarrying (14), other mining (15) and electrical machinery (31), which had significantly higher coefficients than the average per unit of output (27,778, 7,262 and 8,301 m3, respectively, in column [4] of Table 1).

According to the virtual water multipliers, average consumption increased to 4,490 m3 per unit of output. There was again a distinction between the agriculture sub-sectors and the remaining sectors, since the multipliers capture the indirect water consumption. For instance, almost all livestock sub-sectors consumed almost twice the respective direct water coefficient. This means that large amounts of water are consumed indirectly elsewhere to produce the products used as input in livestock production. The high direct water coefficient (13,177 m3) and virtual water multiplier (13,956 m3) of the sector other crops (7), which is dominated by forage production, is thereby explained. The manufacturing sectors cited above again had virtual water multipliers that were above the average. In addition, sectors such as food and beverages (16), basic metals (28) and electrical energy (37) should not be neglected, because they showed a similar pattern of driving indirect consumption as the livestock sectors.

This study sought to assess the Macedonian economy in terms of HO theory and the virtual water multipliers obtained provided a tool to quantify the virtual water exports and imports, i.e. the exact real and theoretical virtual water exiting or entering Macedonia as a consequence of commercial trade (Table 2, columns [4]–[6]). It was found that in general, the Macedonian economy was characterized by large positive virtual water exports of 344 million m3, on one hand, meaning that around 50% of the total water consumed left the country by means of trade with other countries. On the other hand, a large amount of water entered the country through imports (219 million m3). However, these imports of virtual water were not sufficiently large to compensate for the exports, leading to a net deficit of virtual water of around 125 million m3 at 2005 level.

Table 2

Estimated virtual water trade in Macedonia, 2005

iSector [1]Export distributiona (%) [2]Import distribution (%) [3]Virtual water exports (mil. m3) [4]Virtual water imports (mil. m3) [5]Virtual water net exports (mil. m3) [6]
Cereals 0.05 0.52 0.30 2.97 –2.68 
Rice 0.01 0.00 0.26 0.16 0.09 
Raw tobacco production 0.01 0.01 0.03 0.04 –0.01 
Vegetables 1.77 0.63 8.24 2.75 5.49 
Fruits 0.40 0.10 4.22 1.01 3.21 
Grapes and wine production 2.30 0.01 38.18 0.20 37.98 
Other crops 0.01 0.01 0.22 0.14 0.08 
Cattle 0.15 1.44 1.79 15.91 –14.12 
Pigs 0.07 0.76 0.34 3.64 –3.30 
10 Sheep/goat 0.85 0.00 10.89 0.06 10.83 
11 Other livestock 0.09 0.88 0.25 2.27 –2.02 
12 Forestry 0.07 0.22 0.02 0.05 –0.04 
13 Fishing 0.01 0.01 0.01 0.01 0.00 
14 Mining and quarrying 0.46 0.74 17.18 26.11 –8.93 
15 Other mining 0.66 0.23 7.75 2.52 5.23 
16 Food and beverages 6.44 3.96 53.42 30.80 22.62 
17 Tobacco products 0.58 0.37 1.08 0.65 0.43 
18 Textiles 2.23 1.07 1.46 0.66 0.80 
19 Wearing apparel 18.36 12.46 3.25 2.07 1.18 
20 Leather products 2.43 1.44 0.65 0.36 0.29 
21 Wood products 0.20 0.17 0.44 0.34 0.10 
22 Pulp and paper 0.31 1.01 0.09 0.29 –0.19 
23 Printed matter 0.07 0.10 0.03 0.04 –0.01 
24 Coke, refined petroleum 5.14 15.14 0.67 1.85 –1.18 
25 Chemical products 2.76 1.50 4.89 2.49 2.40 
26 Rubber and plastic 0.95 1.39 0.34 0.47 –0.13 
27 Other non-metallic product 1.93 0.86 3.37 1.40 1.96 
28 Basic metals 21.27 10.16 147.37 66.03 81.34 
29 Fabricated metal products 0.94 2.01 0.36 0.71 –0.35 
30 Machinery and equipment 1.05 0.78 3.45 2.39 1.06 
31 Electrical machinery 1.80 1.74 20.71 18.81 1.90 
32 Medical/optical instruments 0.04 0.09 0.01 0.02 –0.01 
33 Motor vehicles 0.92 0.70 0.59 0.42 0.17 
34 Other transport equipment 0.50 0.20 0.34 0.13 0.21 
35 Furniture 0.40 0.52 0.21 0.26 –0.05 
36 Secondary raw materials 0.00 0.15 0.00 0.05 –0.05 
37 Electrical energy 0.00 2.43 0.00 14.50 –14.50 
38 Collected/purified water 0.00 0.26 0.00 0.20 –0.20 
39 Construction work 2.22 7.54 1.39 4.42 –3.03 
40 Services 22.54 28.05 10.17 11.88 –1.71 
 Total   343.95 219.11 124.84 
iSector [1]Export distributiona (%) [2]Import distribution (%) [3]Virtual water exports (mil. m3) [4]Virtual water imports (mil. m3) [5]Virtual water net exports (mil. m3) [6]
Cereals 0.05 0.52 0.30 2.97 –2.68 
Rice 0.01 0.00 0.26 0.16 0.09 
Raw tobacco production 0.01 0.01 0.03 0.04 –0.01 
Vegetables 1.77 0.63 8.24 2.75 5.49 
Fruits 0.40 0.10 4.22 1.01 3.21 
Grapes and wine production 2.30 0.01 38.18 0.20 37.98 
Other crops 0.01 0.01 0.22 0.14 0.08 
Cattle 0.15 1.44 1.79 15.91 –14.12 
Pigs 0.07 0.76 0.34 3.64 –3.30 
10 Sheep/goat 0.85 0.00 10.89 0.06 10.83 
11 Other livestock 0.09 0.88 0.25 2.27 –2.02 
12 Forestry 0.07 0.22 0.02 0.05 –0.04 
13 Fishing 0.01 0.01 0.01 0.01 0.00 
14 Mining and quarrying 0.46 0.74 17.18 26.11 –8.93 
15 Other mining 0.66 0.23 7.75 2.52 5.23 
16 Food and beverages 6.44 3.96 53.42 30.80 22.62 
17 Tobacco products 0.58 0.37 1.08 0.65 0.43 
18 Textiles 2.23 1.07 1.46 0.66 0.80 
19 Wearing apparel 18.36 12.46 3.25 2.07 1.18 
20 Leather products 2.43 1.44 0.65 0.36 0.29 
21 Wood products 0.20 0.17 0.44 0.34 0.10 
22 Pulp and paper 0.31 1.01 0.09 0.29 –0.19 
23 Printed matter 0.07 0.10 0.03 0.04 –0.01 
24 Coke, refined petroleum 5.14 15.14 0.67 1.85 –1.18 
25 Chemical products 2.76 1.50 4.89 2.49 2.40 
26 Rubber and plastic 0.95 1.39 0.34 0.47 –0.13 
27 Other non-metallic product 1.93 0.86 3.37 1.40 1.96 
28 Basic metals 21.27 10.16 147.37 66.03 81.34 
29 Fabricated metal products 0.94 2.01 0.36 0.71 –0.35 
30 Machinery and equipment 1.05 0.78 3.45 2.39 1.06 
31 Electrical machinery 1.80 1.74 20.71 18.81 1.90 
32 Medical/optical instruments 0.04 0.09 0.01 0.02 –0.01 
33 Motor vehicles 0.92 0.70 0.59 0.42 0.17 
34 Other transport equipment 0.50 0.20 0.34 0.13 0.21 
35 Furniture 0.40 0.52 0.21 0.26 –0.05 
36 Secondary raw materials 0.00 0.15 0.00 0.05 –0.05 
37 Electrical energy 0.00 2.43 0.00 14.50 –14.50 
38 Collected/purified water 0.00 0.26 0.00 0.20 –0.20 
39 Construction work 2.22 7.54 1.39 4.42 –3.03 
40 Services 22.54 28.05 10.17 11.88 –1.71 
 Total   343.95 219.11 124.84 

aThe export and import percentages are relative to total intermediate exports and imports, i.e. 132,127 and 123,588 million MKD, respectively (SSO 2008).

mil. = million.

At first sight, it appeared that the trade balance in Macedonia in terms of virtual water was satisfactory and in line with stated HO theory, i.e. for the water-intensive sectors, the exports exceeded the imports in terms of virtual water. A closer look at sector level revealed some deviations, however (Table 2, column [6]). For instance, cereals (1), rice (2), raw tobacco production (3), other crops (7), cattle (8), pigs (9) and other livestock (11) showed negative or very marginally positive net exports. Hence, most of the agriculture sub-sectors did not follow the virtual water trade strategy outlined in HO theory. The manufacturing and service sectors followed similar water trade strategies, with significant imports of virtual water by the mining and quarrying (14) and electrical energy (37) sectors (about 26 and 14 million m3 in column [5] of Table 2).

The general trend in the water trade pattern was influenced by several sectors in the economy. The basic metal sector (28) consumed around 10.0% of total water consumption, but according to the calculated virtual water multiplier, around 42.8% of the total exported virtual water, or 21.5% of total water consumption in 2005, left the country in such products. Although a substantial amount of theoretical virtual water was imported (30%) with the total metal imports, it was not sufficient to offset the high exported amount. The food and beverage sector (16) followed a comparable pattern but at a lower level, with net positive exports of 3.3% of total water consumption. Through the agriculture sub-sectors that were intensive direct water users, exports of grapes and wine (6) and sheep/goat (10) products caused Macedonia to lose around 5.5% and 1.6% of water, respectively, relative to total water consumption.

POLICY ASPECTS

The virtual water multipliers and trade patterns identified here can help in the formulation of different policy scenarios through a cost–push model, hence assessing the price changes in the products (Dietzenbacher & Velázquez 2007). By setting the current price of products to one unit, it is possible to measure by how much the cost of production for a specific product increases when the price for 1 m3 of water is raised by 1 MKD. To obtain this result, Dietzenbacher & Velázquez (2007, p. 190) relate the price change of each product j to water costs through the cost–push relationships associated with the Leontief model (expression (1)). Then, assuming that all current output prices are equal to unity (1 MKD) and rearranging the cost–push relationships, it can be shown that the price change of output j is given by the following expression: 
formula
7
where denotes the new water costs (MKD) in sector i and is equal to 0.001 × wi and wi denotes the water use (column [3]) in Table 1.

Based on the above expression and the estimated virtual water multipliers presented in column [5] of Table 1, it is possible to determine the increase in the price of each output j resulting from an increase by 1 MKD/m3 of the price of water. In this process, the virtual multipliers are divided by 1,000. To illustrate, let us take sector 4 (vegetables), which has an output of 17,106 million MKD requiring the use of 41.98 million m3 of water. Assuming an increase in water cost of 1 MKD/m3, total water cost in this sector increases by 41.98 million MKD. Then, the current price in this sector increases by 3,525.80/1000 = 3.5%. Although this assumption was arbitrary given the linearity property in the input–output model, it indicated that the cost of production will increase in proportion to the revised water costs. An inspection of the estimated virtual water multipliers in Table 1 reveals that an increase in the water price of 1 MKD/m3 would result in significant increases in the prices of agricultural products. Hence, the most price-sensitive products in the agricultural sub-sectors would be rice (2), fruit (5), grape and wine production (6), other crops (7), cattle (8) and sheep/goat (10), with a price increase of around 32%, 8%, 12.6%, 14%, 8.9% and 9.8%, respectively. The most price-sensitive manufacturing sectors would be the highest water users, with a price increase of 28.5%, 8.9%, and 8.7% for mining and quarrying (14), other mining (15) and electrical machinery (31), respectively. Such sensitivity in the output prices indicates that water pricing policy can be used in water saving policies to promote the rational use of water.

In the alternative way of saving water by nullification of foreign trade for the sectors consuming the largest quantities of water and displaying significant net positive exports in terms of virtual water, the total amount saved would be 290.78 million m3 (Table 3). Assuming that the exports of the other economic sectors remain the same, eliminating exports of vegetables (4), fruit (5), grapes and wine (6), sheep and goats (10), other mining (15), food and beverages (16), basic metals (28) and electrical machinery (31) would result in around 42.5% less water being consumed at national level. Examining the results presented in column [4] of Table 3, the basic metal sector (28) showed a significant reduction in water use (21.5%) relative to total water consumption at national level, followed by food and beverages (7.8%) and grape and wine production (5.6%). This exercise is rather hypothetical, because Macedonia could not stop its most important exports, which drive the economy and comprise around 35.5% of total exports. However, from an environmental point of view, the exercise shows the consequences for natural resources and what can be done in future at policy level to promote substantial savings in water consumption.

Table 3

Effects of nullification of foreign exports of some sectors

iSector [1]Exports decline relative to total exports (%) [2]Reduction in water use (mil. m3) [3]Relative to total water demand (%) [4]Decrease in imports relative to total imports (%) [5]Reduction in water use (mil. m3) [6]Relative to total water demand (%) [7]Total decline in trade balance (mil. MKD) [8]
Vegetables 1.77 8.24 1.21 0.14 0.60 0.09 2,168.07 
Fruits 0.40 4.22 0.62 0.02 0.23 0.03 499.34 
Grapes and wine production 2.30 38.18 5.58 0.01 0.14 0.02 3,020.26 
10 Sheep/goat 0.85 10.88 1.59 0.00 0.05 0.01 1,108.55 
15 Other mining 0.66 7.75 1.13 0.08 0.93 0.14 767.30 
16 Food and beverages 6.44 53.41 7.81 2.01 15.62 2.28 6,007.23 
28 Basic metals 21.27 147.37 21.55 14.14 91.90 13.44 10,546.09 
31 Electrical machinery 1.80 20.71 3.03 1.03 11.13 1.63 1,098.05 
 Total 35.50 290.78 42.53 17.43 120.59 17.64 25,214.91 
iSector [1]Exports decline relative to total exports (%) [2]Reduction in water use (mil. m3) [3]Relative to total water demand (%) [4]Decrease in imports relative to total imports (%) [5]Reduction in water use (mil. m3) [6]Relative to total water demand (%) [7]Total decline in trade balance (mil. MKD) [8]
Vegetables 1.77 8.24 1.21 0.14 0.60 0.09 2,168.07 
Fruits 0.40 4.22 0.62 0.02 0.23 0.03 499.34 
Grapes and wine production 2.30 38.18 5.58 0.01 0.14 0.02 3,020.26 
10 Sheep/goat 0.85 10.88 1.59 0.00 0.05 0.01 1,108.55 
15 Other mining 0.66 7.75 1.13 0.08 0.93 0.14 767.30 
16 Food and beverages 6.44 53.41 7.81 2.01 15.62 2.28 6,007.23 
28 Basic metals 21.27 147.37 21.55 14.14 91.90 13.44 10,546.09 
31 Electrical machinery 1.80 20.71 3.03 1.03 11.13 1.63 1,098.05 
 Total 35.50 290.78 42.53 17.43 120.59 17.64 25,214.91 

mil. = million; 1 € = 61.5 MKD.

When the export reduction was obtained using Equation (6), decrease in imports and resulting reduction in water use are estimated and presented in columns [5]–[7] of Table 3. Hence, imports were reduced by 17.4%, with significant reductions in imports by the basic metals sector (28) of around 14.1%. This implies that it would consume 91.9 million m3 less water or around 13.4% relative to total water consumption. The other sectors showed marginal reductions in imports in monetary terms due to export nullification. However, from the environmental point of view, food and beverages (16) and electrical machinery (31), the most intensively water-demanding sectors, would decrease water requirements at national level by 2.3% and 1.6%, respectively, relative to total water consumption (column [7] of Table 3). If the exports were nullified, the trade balance and the Macedonian economy would decline by only 25,214 million MKD, which is 3.3% of total output in 2005. Total output of the economy in Macedonia in 2005 was 773,759 million MKD (SSO 2008). From an economic perspective, such a practice is reasonable, because the economic returns are not greatly reduced compared with the environmental benefits.

DISCUSSION AND CONCLUSIONS

Although trade in virtual water is still a new concept in the literature, it is of crucial importance for defining policies promoting more rational use of water. By conducting a deeper analysis of the relationships between the product processes and the commercial trade in terms of virtual water at disaggregated level, this study revealed an important issue in sustainable water use. In general, Macedonian trade in terms of virtual water is in line with HO theory, i.e. net exports of water-intensive products exceed the net imports. The Macedonian agriculture and manufacturing sectors are the main drivers of this. The significant exports of vegetables, fruit, grapes and wine, sheep and goats from the agriculture sector and of food and basic metal products from the manufacturing sector make Macedonia a net exporter of virtual water. Owing to this trade pattern, based on the 2005 input–output framework, a net quantity of 124 million m3 of water exited the country. These findings are important because these sectors are also the largest water-consuming economic sectors, with highly water-intensive production. Macedonia currently has relatively abundant water resources, but if intensive use continues over the years, combined with climate change effect and the uneven spatial and seasonal water availability will have a significant impact upon the environment and the economy. The water supply is also hampered due to problems with the irrigation schemes, which are old and poorly maintained (World Bank 2003). Hence, there is a presence of water scarcity, especially during the summer months when irrigation is most important. Some agriculture and manufacturing sectors did not follow HO trade strategies, such as cereals, rice, cattle and pig production, mining, etc. and offset their intensive and high consumption of water by importing products with a virtual water content exceeding that of the corresponding exports.

If policy makers aim to promote sustainable consumption of the natural water resource in Macedonia, specialization in some products that require less water during the production process ought to be carefully considered. Such production specialization would mitigate the water scarcity issue that is expected to occur in the long run due to climate change, mismanagement of water resources, intensive consumption and unfavorable trade practices. Dietzenbacher & Velázquez (2007) propose a way to overcome the issue of water scarcity in the long run by investing in technological change, for instance in irrigation systems to significantly reduce the intensity of the water coefficients. Agriculture at national level is the largest water-consuming sector, and thus, technological changes to more efficient irrigation practices such as drip irrigation, combined with capital investment, would be the most desirable policy option.

This study also explored options that provide insights for future changes in a more environmentally sustainable economy. If policy makers were to decide to nullify the exports that display net positive and significant export of virtual water, the trade balance would decrease by 3.2%. However, from an environmental perspective, total water consumption would be reduced by 42.5%. Thus, at national level, the environmental benefits would exceed the economic losses in the long run. Many countries in the Middle East have adjusted their trade policies in order to resolve the problems of water scarcity (Velázquez 2007). For instance, Jordan and Israel have implemented policies to promote sustainable water demand by significantly reducing or completely abandoning the export of intensive water use products and starting to import products that allow them to optimally allocate their scarce water resources (Van Hofwegen 2003). In addition, the policy option to increase water pricing would result in a chain reaction of price increases in products. This study showed how sensitive different products would be given a cost–push in water prices. This increase might also stimulate the producers of water-intensive goods to use water more efficiently and thus decrease the cost of their products. The question is open for further research and is worth exploring at a more detailed level with a different model, such as computable general equilibrium. However, this would require more comprehensive data on prices and demand elasticities of water in Macedonia, which are not available at the moment. It is unlikely that the measures mentioned above would be acceptable in the long run. Thus, the most realistic option with a significant impact upon water consumption would be capital investment in improving and developing the existing irrigation infrastructure in Macedonia, combined with technological changes in irrigation practices.

ACKNOWLEDGEMENT

The authors gratefully acknowledge the financial support received from the Swedish International Development Cooperation Agency (SIDA) under the Unicoop project. The views expressed in this paper are solely those of the authors and do not represent the views of SIDA.

REFERENCES

REFERENCES
Allan
J. A.
1996
The political economy of water: reasons for optimism but long-term caution
. In:
Water, Peace and the Middle East: Negotiating Resources in the Jordan Basin
(
Allan
J. A.
, ed.).
Tauris Academic Studies
,
London
.
CAPRI
2013
CAPRI model as a tool for impact assessment of agricultural and trade policies within EU
. .
Dietzenbacher
E.
Velázquez
E.
2007
Analysing Andalusian virtual water trade in an input–output framework
.
Regional Studies
41
(
2
),
185
196
.
EEA
2009
Water Resources across Europe – Confronting Water Scarcity and Drought
.
EEA Report No.2/2009
.
European Environment Agency
,
Copenhagen
.
Feng
K.
Sui
Y. L.
Guan
D.
Hubacek
K.
2012
Assessing regional virtual water flows and water footprints in the Yellow River Basin, China. A consumption based approach
.
Applied Geography
32
,
691
701
.
Galev
T.
Arsovski
B.
1990
Organization of Agricultural Enterprises and Organization of Livestock Production. Faculty of Agricultural Sciences and Food
.
University Ss. Cyril and Methodius
,
Skopje, FYR Macedonia
.
Hoekstra
A. Y.
2003
Virtual water: an introduction
. In:
Proceedings of the International Expert Meeting on Virtual Water Trade. Values of Water Research Report Series
,
vol. 12
.
IHE
,
Delft, The Netherlands
.
Hoekstra
A. Y.
Hung
P. Q.
2003
Virtual water trade: a quantification of virtual water flows between nations in relation to international crop trade
. In:
Proceedings of the International Expert Meeting on Virtual Water Trade. Values of Water Research Report Series
,
vol. 12
.
IHE
,
Delft, The Netherlands
.
Iljovski
I.
1990
Irrigation: Compendium for the Students. Faculty of Agriculture Sciences and Food
.
University Ss. Cyril and Methodius
,
Skopje, FYR Macedonia
.
Ip
W. C.
Wong
H.
Jun
X.
Zhu
Y.
Shao
Q.
2007
Input–output analysis of virtual water trade volume of Zhangye
. In
Proceedings of the MODSIM 2007, International Congress on Modelling and Simulation
.
Modeling and Simulation Society of Australia and New Zealand
, pp.
1980
1984
.
Leamer
E. E.
1995
The Heckscher–Ohlin Model in Theory and Practice. Princeton Studies in International Finance
.
No. 77, Department of Economics
,
Princeton University Princeton
,
NJ, USA
.
Lenzen
M.
Foran
B.
2001
An input–output analysis of Australian water usage
.
Water Policy
3
(
4
),
321
340
.
Leontief
W.
1936
Quantitative input–output relations in the economic system of the United States
.
The Review of Economics and Statistics
18
(
1
),
105
125
.
Lindberg
G.
Hansson
H.
2009
Economic impacts of agriculture in Sweden: a disaggregated input–output approach
.
Food Economics – Acta Agriculturae Scandinavica C
6
(
2
),
119
133
.
MAFWE
2009
2008 Agricultural Annual Report
.
Ministry of Agriculture, Forestry and Water Economy
,
Skopje, FYR Macedonia
.
Reimer
J. J.
2012
On the economics of virtual water trade
.
Ecological Economics
75
,
135
139
.
Reimer
J. J.
2014
Virtual water trade means ‘trade in water services
. In:
Global Water: Issues and Insights
(
Grafton
R. Q.
Wyrwoll
P.
White
C.
Allendes
D.
, eds).
Australian National University Press
,
Canberra
.
SSO
2008
Symmetric Input–Output Tables for the Republic of Macedonia, 2005
.
State Statistical Office of the Republic of Macedonia
,
Statistical review
,
Skopje, FYR Macedonia
.
UNECE
2011
2nd Environmental Performance Review: The Former Yugoslav Republic of Macedonia
.
United Nations Economic Commission for Europe
,
New York and Geneva
.
Van Hofwegen
P.
2003
Virtual water trade-conscious choices
. In
Synthesis e-Conference on Virtual Water Trade and Geopolitics World Water Council
,
Marseille, France
.
World Bank
2003
Water Resources Management in South Eastern Europe. Country Water Notes and Water Fact Sheets
, Volume II
.
Environmentally and Socially Sustainable Development Department, Europe and Central Asia Region
,
Washington, DC
.
Yu
Y.
Hubacek
K.
Guan
D.
Feng
K.
2010
Assessing regional and global footprints for the UK
.
Ecological Economics
69
(
5
),
1140
1147
.