Key points are not available for this paper at this time.
From a value-added trade perspective, this study utilizes the world input–output tables and the water footprint data of each industry in each country in the Eora database to explore the virtual water resources of 19 countries (the G20 countries excluding the European Union) in 2016. We calculated nine value chain decompositions and the pattern of virtual water flows and then explored the implied virtual water use due to the trade of intermediate goods and final goods, and we also analyzed the unequal trade of virtual water and added value among countries. The results indicate the following. Firstly, in most countries, the largest portion of virtual water is attributed to exports of intermediate inputs that are produced in the source country and fully utilized by the direct import countries, followed by the foreign value-added component of intermediate goods, while the smallest share of virtual water is returned to the country. Secondly, in value-added trade, excluding the rest of the world (ROW), China, France, Italy, Japan, Mexico, South Korea, South Africa, Saudi Arabia, and Germany are net importers in the virtual water trade between G20 countries, and the USA is the largest net exporter of virtual water. Thirdly, intermediate product trade is the dominant form of implied virtual water trade among countries, which leads to a net flow ratio of implied virtual water of about 80% to 90%. Lastly, the Virtual Water Inequality Index shows that thirteen country combinations, including Brazil and Argentina, exhibit significant inequality, and most countries are in a relatively equal state. In addition, the virtual water and added value of the relatively economically developed regions benefit more from the virtual water trade. Therefore, it is crucial for countries to reduce their consumption of virtual water when trading intermediate products to develop high-value-added and low-water-consumption industries and to choose appropriate virtual water trade targets.
Deng et al. (Wed,) studied this question.