مدل سازی تعادل عمومی اثرات اقتصادی مستقیم و غیر مستقیم از بهبود کیفیت آب در هلند در مقیاس ملی و حوضه رودخانه

The main objective of the study presented in this paper is to estimate the direct and indirect economic impacts of water quality policy scenarios in the Netherlands focusing on the reduction of emission levels of nutrients and a number of eco-toxicological substances. For this purpose, an Applied General Equilibrium (AGE) model consisting of 27 production sectors is extended to water through the inclusion of substitution elasticities between labour, capital and emissions to water in the sectors' production functions. The macro-economic costs of a 10, 20 and 50% reduction of the emission levels in the year 2000 of ten priority substances in the EU Water Framework Directive vary between 0.2 and 9.4% of Net National Income (NNI). A large share of the total economic costs are borne by important sources of pollution like commercial shipping, the chemical and metal industry. However, important spin-off effects due to adaptation take place in the tertiary service sector. Besides the estimation of the economy-wide impacts of water quality improvements, the novelty of the study presented here is found in the downscaling of national and sector results to river basin level and the estimation of shadow prices for water-polluting substances through the introduction of an emission permits market.

Integrated river basin models aim to evaluate and predict the impacts of policy interventions on both economic and water systems, based on the idea that aquatic ecosystems perform valuable functions in economic consumption and production processes. Water is used as an input factor (source) in economic production processes (e.g. food processing, electricity production etc.) and at the same time also as a sink for the unwanted by-products of economic production processes (e.g. emission of pollutants to water). In the Netherlands, there exist no comprehensive hydro-economic models to estimate and predict the economic consequences of water policy on both the water and economic system (see Reinhard and Linderhof, 2006 and Brouwer et al., 2007 for recent overviews). Most water related models are hydrological models supplemented by a simple economic module. For instance, the impact of varying groundwater levels on agricultural production is estimated with the help of physical dose–effect relationships and subsequently multiplied by an average market price for agricultural output. Another example is the impact of surface water flow on commercial shipload. Changes in shiploads as a result of different flow levels are valued in economic terms with the help of an average market price per ton shipload. A more systematic and comprehensive economic assessment, accounting for the direct and indirect economic consequences of water policy, is often lacking. Integrated water quality models that link biogeochemical water and substance flow models to economic models are rare. Most of the time ad hoc analyses are carried out to assess, for instance, the cost and effectiveness of possible water quality measures to achieve certain water quality objectives (e.g. Lohani and Thanh, 1978, Beavis and Walker, 1979, Schleich et al., 1996, Gren et al., 1997, Van der Veeren, 1999, Yang and Weersink, 2004, Kramer et al., 2006, Bonham et al., 2006, Wang, 2006 and Clasen et al., 2007). Costs are based on the direct financial engineering costs of a technical pollution abatement measure like wastewater treatment, while the environmental impact assessment is based on estimated dose–effect relationships or expert judgment of the emission reduction capacity of the specific technical measure. More comprehensive integrated modelling of the wider economic impacts of water pollution abatement is missing in the international literature, the recently published input–output model for the emission of nutrients in wastewater in the Chinese City of Chongqing in this journal (Okadera et al., 2006) being an exception. A few new examples are given in this special issue, but none of these examples are based on AGE models.1 Until now the existing models seemed to suffice to underpin water policy. However, the demand for integrated economic assessments increases, including the estimation of the indirect economic impacts of water policy. In the Netherlands this was explicitly acknowledged for the first time in the Fourth National Water Policy Document (Ministry of Transport, Public Works and Water Management, 1998). At European level, the Water Framework Directive (WFD) (2000/60/EC) adopted in 2000 is the first European directive to explicitly recognize the importance of the interdependency between aquatic ecosystems and their socio-economic values and advocates a more integrated river basin approach to water policy. Investments and water resource allocations in river basin management plans are guided by cost-effectiveness and cost recovery. An important challenge here is to collate and present data and information about the water system and the economy at the level of river basins. For the assessment of what the WFD refers to as ‘disproportionate costs’ in relation to the Directive's environmental objectives, the estimation of both the direct and indirect costs of policy measures is expected to be relevant, especially when the corresponding changes in water use and water prices are substantial. In order to meet this growing demand for integrated water policy assessment tools and methods at the scale of river basins, an integrated water economics information system called the National Accounting Matrix including Water Accounts for River Basins (NAMWARiB) was developed in the Netherlands (Brouwer et al., 2005). NAMWARiB provides information about the interlinkages between the physical water system and the economy at national and river basin scale and is an extension of the National Accounting Matrix (NAM) with physical water and substance flow satellite accounts. Based on this information system, interactions between economic activities and the water system can be modelled in a systematic way to predict future changes in water and economic systems. As a first step, we use an existing static applied general equilibrium (AGE) model of the Dutch economy in the year 2000, and extend this to include the emission of a number of polluting substances to surface water, such as nutrients (N and P), heavy metals such as Arsenic (As), Chromium (Cr), Cadmium (Cd), Copper (Cu), Mercury (Hg), Nickel (Ni), Lead (Pb), Zinc (Zn) and the chemical compound Polycyclic Aromatic Hydrocarbon (PAH). Information from NAMWARiB is used to disaggregate the macro-economic model results to river basin level. The direct and indirect economic consequences of a number of emission reduction scenarios are estimated with this model. The paper's main objective is to present the integrated hydro-economic model, the economic results of the emission reduction policy scenarios in the context of the WFD and the disaggregation procedure of the macro-economic results to river basin level.2 The paper is structured as follows. First, the AGE model is presented in Section 2. Section 3 provides information about the calibration procedure and the way emissions to surface water are integrated into the model, including investments in pollution abatement. Section 4 introduces the policy scenarios and discusses the model results at macro-economic level, sector level and river basin level. Finally, Section 5 concludes.

This paper presents the use of an AGE model of the Dutch economy extended with water-related emissions, to assess the economy-wide impacts of the implementation of three different water quality policy scenarios related to the WFD. The total economic costs of a 10, 20 and 50% reduction of the national emission levels in 2000 are estimated at three different levels: national, sector and river basin level. The total economic costs at national level range between 700 million and 32 billion euro, equivalent to 0.2 and 9.4% of NNI in the year 2000. At sector and river basin level, most costs accrue to commercial shipping, the chemical industry, oil refineries and the metal industry in the largest river basin in the Netherlands Rhine West, followed by the rubber and plastics industry and the energy sector in the Meuse basin. The agricultural sector is hit relatively hard in Rhine West, the Meuse basin and Rhine East under nutrient emission reduction policies of 20 and 50%. The novelty of the study presented here is found in the fact that it is the first study to explicitly address the total direct and indirect economic impacts of WFD implementation in Europe with the help of an AGE model. Integrated water quality and economic models are rare. Also the consistent and coherent up and down scaling of the predicted economy-wide impacts across sectors and regions is of particular interest here. The model results show that the expected indirect impacts of large-scale interventions as foreseen due to the implementation of the WFD in the Netherlands can be substantial. Considering total abatement costs as direct economic costs, the indirect economic costs of the emission standards are by definition given by the difference between loss in value added and total abatement costs. In this case, the share of indirect costs can be as high as 70–90% of the total economic costs. However, this result has to be interpreted with the necessary care as we will discuss below. The expected high share of indirect costs is also reflected in the fact that besides important sources of pollution such as commercial shipping and the chemical and metal industry, trade and commercial support services in the tertiary sector bear a large share of the total economic costs. The introduction of an emission permit market is also novel in hydro-economic modelling. The emission permit market gives rise to further substantial indirect economic effects through the substitution of relatively expensive abatement technology and changing production and consumption patterns as a result of changes in relative prices and changes in the tax structure. Based on abatement technology only the total abatement costs in the new equilibrium would have been substantially higher as total abatement costs for a maximum reduction of 35% of toxic substances are as high as 18 billion euros (and 2.8 billion euros for a maximum reduction of 60% of nutrients). Although the costs associated with the purchase of emission units can be considered direct costs from the perspective of an individual producer or sector, they are not an economic cost at macro-economic level as the revenues are fully recycled into the economy in the model used here. Hence, the model structure makes it hard if not impossible to provide a straightforward estimate of the different components underlying the total economic costs. An important reason for this is the complex dynamics underlying AGE models, especially the reshuffling of scarce production factors, including water as ‘natural capital’, across economic activities due to relative price level changes caused by the restriction on pollution emission levels. Given the model's static nature, it is also unknown what the flow of total annual economic adjustment costs are as a result of the implementation of the emission reduction scenarios. The model merely shows what the total cost is if the economy adjusts instantaneously based on a comparison of the new economic equilibrium situation with the baseline situation. Research is ongoing to convert and update the static AGE model used in this study into a dynamic version. This will allow us to better assess the economic adaptation costs of the imposed emission constraints in different economic sectors on an annual basis, and provide more insight in the direct and indirect economic effects of the proposed water policy interventions. Dutch policy makers generally welcome the integrated international river basin approach advocated by the WFD. The Netherlands are often referred to as the ‘sink of Europe’ in view of the fact that large international rivers such as the Rhine, Meuse and Scheldt and the pollution they carry drain into the North Sea in the Netherlands. Objectives and measures taken in surrounding countries therefore have an important impact on water quality in the Netherlands. In this study, we distinguish two different approaches to account for what happens in surrounding countries. The first variant assumes that changes in the Dutch economy have no impact on world market prices, hence affecting the international trade balance, while the second variant assumes that world market prices change in the same way as domestic price levels, leaving international trade conditions under different policy scenarios as they were in the baseline situation. The total economic costs are considerably lower under the first variant when the emission levels are reduced by half, largely because of the possibility to import cheaper emission-intensive products from abroad. However, given the fact that Dutch water systems are part of large international river basins, pollution exported abroad may under this less expensive variant literally flow back into the country, leaving the water environment as bad off as before, requiring additional investments in clean-up. This potentially important feedback mechanism is missing in the model presented here. Future work will focus on the development of international river basin models, which are able to account for important spatial upstream–downstream relationships in an international basin context. Many European countries deal with the same problem that they are part of an international river basin, and decisions upstream affect water quality and economic costs and benefits downstream. Also the potential cost savings in some industries as a result of improved water quality, another important feedback relationship, are not included in the total cost estimations presented here, possibly resulting in an overestimation of the total economic costs. The interaction between the economic activities and the water system is captured here uni-directional in the activities’ production function through constant elasticities of substitution between the production factors labour, capital and emission units. These emission units refer to a limited set of WFD priority substances (N, P, As, Cr, Cd, Cu, Hg, Ni, Pb, Zn and PAHs). Water flows, which have an important influence on substance flows and water quality, are not part of the modelled interaction. An important next step is to link the AGE model and the emission of polluting substances by economic activities to a water and substance flow model, accounting for pollution pathways and hydro-biological decay processes. Modelling of the interactions between pollution sources and water and substance flows will allow for a more detailed assessment of the spatial water quality impacts of proposed WFD policy scenarios. The spatial disaggregation procedure of the macro-economic impacts of the emission reduction scenarios presented here is based on a simple weighted allocation procedure accounting for the distribution of economic activities and their emission levels of nutrients and toxic substances across river basins. Linking emissions to water quality through spatially differentiated water and substance flow models will address an important source of uncertainty underlying the current model results, i.e. how will water quality improve across all river basins and water bodies due to economic adaptation to emission restrictions. This work should focus simultaneously on the above mentioned feedback relationship of water quality on economic production and consumption. In the case of consumption one could consider extending the model even further by including consumptive water uses which are currently kept outside the analysis such as water-based recreation, and which are expected to have important welfare implications too.