Content

1. INTRODUCTION

This paper presents some of the key issues in the economic analysis underlying the European Water Framework Directive (WFD) and provides an overview of the available economic methods and models to address these issues. The WFD is one of the first European directives in the domain of water, where economics is an integral part of the decision-making process surrounding its implementation in Member States (MS). The implementation process has been an important learning process so far for both experts and policy makers. Many decision issues have been clarified in the past years, but equally as many policy and research questions related to the development of the first integrated river basin management plans in 2009 remain open.

The identified key issues for the future implementation of the WFD across European river basins include:

• integrated water-economic river basin information systems and indicators;
• selection of cost-effective programs of measures to reach the WFD objectives;
• definition and assessment of disproportionate costs to reach the WFD objectives;
• cost recovery of water services including environmental and resource costs;
• design and evaluation of incentive water pricing instruments.

These issues surrounding the development of integrated river basin management plans are inherently multi-disciplinary in nature, requiring the input from hydrology, civil engineering, ecology and economics. For this interdisciplinary approach to be successful, the involvement of the users of the information (policy makers) is paramount (Brouwer et al., 2003). The issues have partly been identified based on the economic requirements in the WFD and partly on specific national and European economic research programs in support of specific unresolved issues in the implementation of the WFD.

The paper’s main objective is to highlight the key issues in the economic analysis in the WFD towards the first integrated river basin management plans in 2009, and to present an overview of available economic methods and tools to address these issues. The paper is written for economic practitioners, water policy advisors and consultants working on WFD implementation, who are familiar with economics.

The key issues and tools and methods are illustrated using examples largely taken from own work in this field and are therefore inevitably biased towards North European case studies. However, in view of the authors’ participation in pan-European networks (including Common Implementation Strategy working groups) surrounding the WFD implementation, the key issues raised in this paper are believed to be wider applicable. Another important disclaimer is that the examples do not necessarily reflect state-of-the-art practices in the actual WFD implementation across MS. Political-economic conditions are highly influential in the way many if not most of the issues are addressed and resolved in practice in individual MS. The overview of European integrated hydro-economic models presented in the annex to this paper is based on the international workshop by the same name organized by the corresponding author in November 2005 in the Institute for Environmental Studies, Vrije Universiteit Amsterdam (see www.ivm.falw.vu.nl/watereconomics).

The paper is organized as follows. Section 2 first outlines the economic elements in the WFD. This is followed in Section 3 by a presentation and discussion of the key issues surrounding the economic analysis in the WFD. Section 4 presents a description of the available economic methods and tools to address these issues, including cost-effectiveness analysis, cost-benefit analysis, non-market valuation of the economic benefits of water policy, multi criteria analysis, integrated river basin models, and economic instruments. A summary of the key issues and the available economic tools, methods and instruments and conclusions are provided in Section 5.

2. THE ECONOMIC ANALYSIS IN THE WFD

The WFD is one of the first European Directives in the domain of water, which explicitly recognizes the role of economics in reaching environmental water quality objectives. The Directive calls for the application of economic principles (e.g. polluter pays principle), approaches and tools (e.g. cost-effectiveness analysis) and for the consideration of economic instruments (e.g. water pricing methods) for achieving good water status for all European water bodies in the most effective manner. The Guidance Document on the Economic Analysis prepared in 2002 by the European Water and Economics Working Group (WATECO) advises that the various elements of the economic analysis should be integrated in the policy and management cycle in order to adequately aid decision-making when preparing the river basin management plans. The integration of economic reason and arguments throughout the WFD policy and decision-making cycle is presented in Figure 2.1.

Source: WATECO Guidance Document
Figure 2.1: The role of economics throughout the WFD implementation process

The main elements of the economic analysis are found in Articles 5 and 9 and Annex III in the WFD. Economic arguments also play an important role in the political decision-making process surrounding the preparation of River Basin Management Plans in Article 4 where derogation can be supported by the strength of economic arguments when setting environmental objectives. The economic analysis can be summarised as follows:

The main steps in the economic analysis identified by WATECO (2002) and the associated time path are illustrated in Figure 2.2. The first step, the economic characterisation of river basins, has been completed. In the next years the preliminary findings so far in the different European river basins regarding the main water users (sources of pollution) will be further elaborated (including a more detailed definition of environmental objectives) and a start has been made with the identification of additional measures needed to reach good water status in a second step.

By the end of 2007 each EU Member State has to produce an overview of its basic and additional measures according to Article 11, from which the most cost-effective programme of measures will be selected in step 3 by the end of 2008. Based on a cost-effectiveness analysis of programmes of measures, the question whether the total costs of additional measures to reach good water status are disproportionate will be addressed at the same time. Finally, the financial implications of the basic and additional measures for different groups in society has to be evaluated by 2009, including the level of cost recovery, the use of economic instruments (e.g. levies, taxes, water prices) and their role in achieving a more efficient and sustainable water use.

Source: modified from the WATECO Guidance Document
Figure 2.2: Steps in the economic analysis in the WFD and corresponding time path

3. KEY ISSUES IN THE ECONOMIC ANALYSIS IN THE WFD

This section discusses the key issues in the economic analysis underpinning the implementation of the WFD. These key issues are directly related to the different steps in the economic analysis presented in Figure 2.2., and include:

1)Integrated water-economic river basin information systems and indicators.
2)Selection of cost-effective programs of measures to reach the WFD objectives.
3)Definition and assessment of disproportionate costs to reach the WFD objectives.
4)Cost recovery of water services including environmental and resource costs.
5)Economic instruments to reach the WFD objectives.

The first issue refers to the compilation of integrated policy (progress) indicators at river basin level and is relevant for the first two steps in Figure 2.2. Integrated progress indicators enable river basin characterization as required in step 1 from an environmental and economic system perspective and show the possible trade-offs involved (i.e. the core of economics), and the identification of possible water management problems as required in step 2. The following issues refer to step 3. The assessment of disproportionate costs follows the assessment of the least cost way to reach the WFD environmental objectives of good chemical and ecological status (cost-effective program of measures). This is followed subsequently by an assessment of the distribution of the financial burden across different groups in society (households, agriculture and industry) based on the principles of cost recovery and the polluter pays through existing or future market-based pricing policies as formulated in Article 9 in the WFD. The key issues will be elaborated in more detail in the sub-sections below.

3.1. Integrated water-economic river basin information systems and indicators

The implementation of the WFD has increased policy and decision-maker demand for integrated hydro-economic information at the level of river basins. The WFD requires that river basins across Europe are described in both physical and economic terms. According to Article 5 in the WFD, the economic characterisation of river basins should include an assessment of the economic significance of current water use and future water use up to 2015 (Wateco, 2002). Integrated hydro-economic accounting systems are considered to be useful tools, displaying information about the interactions between the physical water system and the economy at national and river basin scale. This involves not only linking water related environmental data to economic data, but also presenting these data at the relevant spatial and

Box 1: Example of integrated river basin accounts and policy progress indicators The economic significance of water use in economic production processes in different river basins in the Netherlands is measured in the National Accounting Matrix including Water Accounts for River Basins (NAMWARiB) in two different ways. Economic significance is measured through production values and value added generated in river basins per sector (expressed in euros). Water use is measured through water extraction (expressed in cubic metres) by economic activities and the emission of polluting substances (expressed in kilograms) per sector. Water use can furthermore be measured through wastewater discharge per sector in each river basin (expressed in inhabitant equivalents). Based on time series analysis possible trends can be identified. Trends in economic driving forces are linked to pressures such as water consumption, wastewater production and the emission of polluting substances (nutrients, metals etc.). An example is given in the figure below. At the national level, real economic growth (in terms of GDP in constant prices) over the period 1996-2001 was 18 percent (on average 3 percent per year). Total wastewater production remained more or less the same over that same period, whereas the emission of nutrients decreased by approximately 15 percent and the emission of metals by about 10 percent. Hence economic activities seem to use the water environment in a more efficient way over this time period.

important insight in the environmental efficiency of economic activities. They also provide a basis for trend analysis and the calibration and validation of integrated hydro-economic optimization and simulation models. Based upon the observed development of economic activities within sectors and corresponding water use over say the past 10 years, one can extrapolate this development into the future.

(right) river basin over the period 1996-2000 (1996=100) Source: Brouwer et al. (2005).

temporal scales through appropriate (dis)aggregation procedures. It is especially this issue of matching available data across various spatial scales, which has proven to be one of the major challenges in the compilation of integrated river basin information systems (Brouwer et al., 2005).

According to the European Task Force on Water Satellite Accounts (2002, p.4), the implications of the WFD include more focus on the geographical boundaries of available data, i.e. water bodies and river basin districts. Also the United Nation Handbook on Integrated Environmental and Economic Water Accounting acknowledges this (UN, 2006, p.17): ‘… it is important that the spatial reference is the same for hydrological and economic data.’ However, ‘river basins for which hydrological data is usually available, do not generally coincide with administrative regions for which economic information is collected’. No consistent pan-European water economic information system exists, which allows for a consistent (dis)aggregation and comparison between the national and river basin level. Most efforts in the area of integrated water accounting are based on national statistics. An example of an integrated river basin information system, its use and usefulness, is presented in Box 1.

3.2. Selection of cost-effective programs of measures to reach the WFD objectives

Environmental efficiency has always been an important guiding principle in European environmental legislation, including the WFD. According to the WATECO guidance, the definition of the program of measures and the ranking of basic and supplementary measures based on cost-effectiveness criteria is the key economic input into the preparation of the RBMP. The main steps identified in the guidance include the estimation of the costs of each measure, the estimation of the effectiveness (environmental impact) of each measure and the ranking of cost-effectiveness of measures. Hence, the information needed basically relates to the costs of potential measures and their effectiveness (see the example presented in Box 2). According to WATECO, specific care needs to be given to the choice of the effectiveness indicator as different effectiveness indicators may lead to different rankings of measures. Furthermore, specific attention may be required as the effectiveness of measures can often be assessed quantitatively only for a few environmental indicators, and not for the range of environmental issues encompassed in the definition of good water status.

WATECO states that uncertainty about costs, effectiveness and time-lagged effects of measures needs to be dealt with throughout the economic analysis process, and more generally throughout the process of identifying measures and developing the river basin management plan. Sources for uncertainty are highly diverse according to situations and river basins, but will exist with regard to the assessment of pressures, impacts, baseline scenario, costs and effectiveness. The main sources of uncertainty underlying the selection of a cost-effective programme of measures include (Brouwer, 2005; Refsgaard & Nilsson 2003):

Based on experiences in various regional and international pilot river basins, (e.g. Boeters and Brouwer et al., 2006; Trémolet Consulting, 2006) the following key issues arise when trying to identify a cost-effective program of measures:

Identification of the environmental objective(s) involved
The concrete environmental WFD objectives for different types of water bodies (artificial, heavily modified, natural) are still unknown. Especially when aiming to develop an

An example of a cost-effectiveness analysis is presented in the table below (adapted from Baan, 1996). The table presents possible measures to improve water quality in the West Scheldt estuary, located in the south-east of the Netherlands (red area in the map), their effects and costs. A measure’s cost-effectiveness is calculated by simply dividing the annual cost (column 3) by the annual effect (column 2). The measure with the lowest cost per unit pressure reduction is the most cost-effective measure. In this particular study, dredging is the most cost-effective solution, followed by the treatment of drinking water and the dislocation of the effluent from the wastewater treatment plant (WWTP) outside the estuary. Replacement of zinc plates on ships is the most costly measure per unit pressure reduction. The pressure reduction in the table below is expressed in ‘Pressure Equivalents’ (PE). This PE is based on the measures’ emission reduction of different pollutants (N, P, Cd, Cu, Zn) in kg per year (estimated through expert judgment) compared to the threshold concentration levels of each pollutant in kg/m3.

Second, measures can be taken at different points along the environmental source-effect chain. Preventing instead of treating water pollution may be another important decision criterion for the selection of measures. Examples here are dredging versus the dislocation of the WWTP effluent discharge point. Third, given the spatial and temporal complexity of many water systems, pressure reduction in terms of the emission of kilograms of a pollutant per year is often more easily estimated and predicted than the actual impact of this pressure reduction on water quality in terms of milligrams of a pollutant per liter. Although it is water quality that is of prime interest in the selection and assessment procedure, pressure indicators are often used as proxies for the impacts on water quality given the uncertainties surrounding dose-response relationships and environmental damage functions.

international river basin management, including transboundary programs of measures, some degree of common understanding and consensus is needed about the environmental objectives involved, and the measurement of progress in reaching these objectives through compatible monitoring programs.

Identification of the sources of pollution, pressures and impacts now and in the future over the appropriate time horizon

The exact relative contribution of different sources of pollution to the environmental water quality problem is often unknown, making it hard if not impossible to target different polluting sectors with specific measures to bridge the gap between the expected and desired state of the water body and river basin by 2015.

Quantification of the gap to be bridged

Related to the previous point, the gap to be bridged (problem definition) appears to be hard and in some cases even impossible to quantify given the lack of scientific knowledge and the uncertainties surrounding dose-effect relationships. In many cases a qualitative assessment of the problem at hand based on expert judgement is the best available estimation.

Identification of measures to bridge the gap between the reference (baseline) situation and target situation

The WFD distinguishes between basic and additional measures. Basic measures include measures, which help to achieve existing European water related legislation (e.g. Nitrates Directive or the Urban Wastewater Directive). Although the objectives of existing European and national policy are usually clear, it is less clear which types of measures have to be and will be implemented in the future in order to reach these objectives, not least because of the given uncertainties about their impact on water quality. Another important issue is the distinction between technical measures and economic instruments (see section 3.6). In practice, the distinction is often not clear, and measures and instruments are used interchangeably in the cost-effectiveness analysis. Evaluating technical measures and instruments simultaneously has the risk of double counting. Often, instruments and measures are implemented sequentially, not simultaneously, and there is therefore a real risk of double counting.

Evaluation of the effectiveness of measures

The evaluation of the effectiveness of the proposed measures is probably the most troublesome in the entire decision process given the lack of knowledge and information and hence substantial uncertainties involved. Different indicators are used (e.g. chemical, biological), focusing on different points along the environmental dose-effect chain (e.g. pressure indicators and impact indicators), expressed in different units (e.g. kg/ha, mg/litre) making it hard to compare results from different types of measures in a compatible, meaningful way. Multi-criteria analysis (MCA) techniques (see the method description in section 4) may be one possible way of dealing with this plurality of measurement units in a systematic and coherent way. Significant differences also exist between the temporal and spatial scale at which measures are implemented in a basin and the associated differences in temporal and spatial impacts.

Evaluation of the costs of measures

The distinction between different cost types (financial versus economic; fixed versus variable; direct versus indirect), cost items (e.g. one-time-off investment costs, annual management and maintenance costs, exploitation costs, overhead and depreciation costs) and their representation in time (e.g. annuity, (net) present value) is of utmost importance to arrive at a meaningful comparative analysis across measures and basins. It has to be made explicit which cost type and (proxy) estimation is being used, as there may exist significant differences between direct financial (engineering) costs and the usually broader defined direct and indirect economic costs. Different bottom-up and top-down calculation and estimation methods are available to assess the direct and indirect financial and economic costs of programs of measures (see the method description in section 4). The way costs are estimated has important implications also for the assessment of disproportionate costs (see the next section).

Obviously, the ranking of measures in terms of increasing costs and based on this the selection of the least cost way to reach the environmental WFD objectives can only take place if the extent of the water quality problem (gap) is known and hence the necessary means (programs of measures) to solve the problem or bridge the gap, and the costs and effects of programs of measures are measured in comparable units. This requires inevitably the introduction of some degree of quantification in the analysis, including the qualification and quantification (in terms of orders of magnitude) of the main uncertainties involved.

3.3. Definition and assessment of disproportionate costs to reach the WFD objectives

The procedure to establish the environmental WFD objectives are described in Article 4. This article also allows for so-called derogation, i.e. to lower environmental objectives or delay them in time based on (i) technical feasibility of achieving the objectives and/or (ii) disproportionate costs (paragraph 3-7). The concept of disproportionate costs is not a standard economic concept. The assessment of disproportionate costs is subjective (WATECO, 2002), depends on the political economy of a country or river basin region, and proves to be surrounded by a lot of uncertainty as to their exact definition at what scale in the practical implementation of the WFD. It refers to a skewed distribution (discrepancy) between objectives and measures for reaching these objectives, but concrete national or international benchmarks do not exist, also not in comparable European Directives such as the IPPC Directive where the concept of Best Available Techniques Not entailing Excessive Costs (BATNEC) is introduced or the Habitats Directive, which refers to ‘imperative reasons of overriding public interests, including those of a social or economic nature’ to justify exemptions.

In theory, the assessment of disproportionate costs follows the assessment of the least cost way to reach the WFD environmental objectives and the corresponding distribution of the financial burden across different groups in society (households, agriculture and industry) through existing or future pricing policies as formulated in Article 9 in the WFD. However, in practice excessive or disproportionate costs often also play an important role already during the first stages of selecting a cost-effective program of measures (see previous section) as a political screening criterion1. Whether the costs of reaching the defined environmental WFD objectives are disproportionate depends on willingness and ability to pay by different socio-economic groups and sectors in society to whom these costs are transferred and who pay for them. Willingness to pay (WTP) depends on political, sectoral and public preferences for the environmental objectives involved and the environmental and resource costs of current water use (see section 3.5) and ability to pay. Ability to pay in turn depends on the financial strength and capacity of the public and private actors and sectors involved, which is in part determined by the current state of the economy.

Given the definition of disproportionate: unequal relationship between two units (in this case the objective and the means to achieve the objective), the key question to assess disproportionate costs is to what extent the costs of WFD measures outweigh the benefits of reaching the WFD objectives. Cost-benefit analysis (CBA) would be the appropriate economic tool to assess this relationship (see the method description in section 4), using the net present value or benefit-cost ratio as the decision criterion, comparing all positive and negative welfare effects of the WFD measures to achieve good chemical and ecological status. Assuming that all positive and negative welfare effects are accounted for in the CBA, a negative NPV or a benefit-cost ratio less than one, indicating a national welfare loss, would result in principle in a rejection of the proposed program of measures to reach the proposed WFD objectives. However, given the public good nature of the WFD objectives and in the interest of future generations, the government may be willing to accept a benefit-cost ratio of less than one. The question then becomes how many times the costs can exceed the benefits before the program of measures involved is considered disproportionate. Uncertainty is expected to play an important role here too, related to the assessment of the costs and the environmental and social benefits of WFD implementation.

Another important issue is the scale at which the assessment of disproportionate costs takes place. Costs may be considered disproportionate at the level of an individual region or sector given its financial buoyancy, but economically beneficial at the higher aggregated level of a nation as a whole. On the other hand, the skewed distribution of costs and benefits across time and space and across different sectors may be an important reason for policy makers to consider a particular solution undesirable given for example the disproportionate effects on certain vulnerable groups in society. It is important to realize that the answer to the question whether the estimated costs of programs of measures are disproportionate remains subjective in all these cases (see also Box 3).

3.4. Cost recovery of water services

According to Paragraph 1 in Article 9 in the WFD, ‘member states shall take account of the principle of recovery of the costs of water services, including environmental and resource costs, having regard to the economic analysis conducted according to Annex III, and in accordance in particular with the polluter pays principle’. The assessment of environmental and resource costs will be addressed in the next section. This section focuses specifically on the economic cost assessment of water services. Water services are defined in paragraph 38, article 2 in the WFD: ‘Water services means all services which provide, for households, public institutions or any economic activity: (a) abstraction, impoundment, storage, treatment and distribution of surface water or groundwater; (b) waste water collection and treatment facilities which subsequently discharge into surface water.’

In practice, the distinction between water use and water services is not very clear-cut and surrounded by many political obstacles related to the discrepancy between current and potentially future pricing of water use and services. The WATECO guidance regards water services as the link between the environment and water use, consisting of activities, which change important characteristics of water or wastewater. It is this change in water or wastewater characteristics, which the WATECO guidance calls a water service. Water use is defined as an activity, which has a significant impact on a water body’s status. It is unclear, however, what is considered a ‘significant’ impact.

The issue of water use and water services can also be addressed from a different perspective, namely that of the water system instead of the human activities related to the water system. This approach is in line with the ecological principles underlying the WFD, and is based upon the international environmental and ecological economics literature, in which it is nowadays generally acknowledged that ecosystems, including aquatic ecosystems, provide goods and services (functions), which are beneficial to human society (e.g. Adamus and Stockwell, 1983; Nichols, 1983; Larson et al., 1989; Turner, 1993; Barbier, 1994; Gren et al., 1994; Maltby et al., 1995; Costanza et al., 1997). Following this approach, a distinction can be made between the supply of and demand for water services. This introduces basic economics back into the economic analysis of the WFD, where demand for water services is reflected in water use.

Water services are supplied first of all by the water system itself. Examples are basic services such as surface or groundwater used for drinking water purposes or growing crops, but also the nutrient and wastewater absorption capacity of aquatic ecosystems are important services provided by the water system. In the environmental economics literature the former are called direct functions and the latter indirect ecosystem functions. Human activities may also add value to these services or try to (partly) replace these services, but the economic significance of these activities usually depends largely on the original services provided by the water system. Examples are pumping up surface or groundwater, transportation of water, water purification, water distribution etc. In efficiently organised modern societies, these activities are nowadays considered essential services in themselves, satisfying consumer demand for water availability and water quality. Hence, depending on demand, water services may be modified to different degrees by human activities in order to make them suitable for final consumption. Examples of human activities, which (partly) replace original water services, are wastewater collection and treatment. Often, the original waste absorption capacities of aquatic ecosystems have been impaired by excess pollution and hydro-morphological modification. The original water services or the water services modified by human activities are supplied to various groups of consumers, such as households, agriculture and industry. In some cases these consumers are also the suppliers of the water service. This is labelled ‘self service’ in the WATECO guidance. Examples are farmers or industries, who pump up surface or ground water for their own use such as crop irrigation, cooling water, paper production or food processing.

The principle of cost recovery implies that the economic production costs of a good or service – measured through their opportunity costs - are fully recovered. An example is given in Box 4. The cost of a productive activity consists, in theory, of the opportunity costs of the necessary inputs. The opportunity cost of employing an input is the highest net benefit generated had it been employed elsewhere. If input and output markets exist and function well, the opportunity costs are reflected in the market prices paid for the inputs. For example, the price of the necessary labour, equipment or electricity to produce a good or service.

However, in some cases, inputs are used or outputs produced for which no markets exist and hence also no market prices are available to reflect their opportunity costs or which are traded in distorted markets, which do not reflect the real opportunity costs of the inputs used or outputs produced. These inputs or outputs usually involve so-called public goods or services, i.e. goods or services (consumed or produced), which can technically (and often also institutionally or politically) not be broken down in separate and individually purchasable units. Examples include resources such as air and water and their natural purification processes (or their function as a disposal sink for emitted pollutants). Here, environmental damages caused by water use (e.g. water abstraction or the emission of pollutants) often do not result in a private cost to individual agents, but do result in a social cost to society (Brouwer et al., 2004). This is what the WFD refers to as environmental and resource costs (see the next section).

Besides environmental damage not accounted for through market prices (i.e. market prices only reflecting the private opportunity costs and not the social opportunity costs), Renzetti and Kushner (2004) distinguish a number of other reasons why a government agency or firm’s private accounting of the costs of its activities may differ from the full economic cost of those activities. First, there may be accounting guidelines set out in government regulations, which dictate the way in which costs are recorded. For example, utilities may be prevented from assigning a competitive rate of return as part of the opportunity cost of their purchased capital goods.

Secondly, a water utility may receive subsidies from other agencies. These could include direct subsidies such as capital grants from higher levels of government or indirect subsidies as might occur if a municipal water utility were to receive services from the city’s legal department without charge.

Thirdly, analysts and utility regulators may disagree in how completely they wish to see an agency move to full cost accounting. For example, some may argue that it is sufficient to see that operation and maintenance costs are fully accounted for. Others may argue that operating and maintenance plus capital costs must be fully accounted. Still others would argue that also external costs such as environmental damages and the opportunity cost of raw water supplies must be included.

A fourth factor that inhibits implementation of full cost recovery is a lack of standardized guidelines for these types of accounting. Although there exists a large body of economic research devoted to measuring whether a household, government agency or firm’s accounting of the costs of its activities accurately reflect the costs borne by all of society, such as road transportation or the generation of electricity2, there is a limited literature that is concerned with applying the principles of full cost accounting to water and sewage utilities.

Finally, because of the novelty of this approach to water and sewage utility accounting, there may be difficulties in collecting the data needed to estimate some cost components.
Other important issues that play a role when trying to assess the level of cost recovery of water services and which add to the level of complexity and uncertainty when trying to assess cost recovery of water services include:

• In some cases there does not exist a clear direct relationship between the service in question, the provider of the service and the user of the service, making it hard if not impossible to assess to what extent different water users contribute adequately to cost recovery as required in article 9. For example, the management of rivers and watercourses (e.g. dredging) serves both commercial and recreational shipping and flood control. Flood control management is another example of the provision of a public good, which cannot technically be broken down in individual units and sold to individual beneficiaries.
• In other cases there exist interdependencies between water services and their costs. For example, an increase in the level of sewerage provision levels (e.g. increase in connection rate) will also result in an increase of water treatment costs.
• Not all costs are always clearly visible. For example, environmental and resource costs (see the next section). Or costs are financed through different financial channels and by different institutions (e.g. local tax revenues and general income taxation). This too adds to the problem of assessing the adequate contribution of different water users.

3.5. Environmental and resource costs

Environmental and resource costs are mentioned in article 9 related to cost recovery (see section 3.4), but are expected to also play an important role in the assessment of disproportionate costs in terms of the public benefits foregone of not reaching a good chemical and ecological status. In the Wateco guidance’s glossary of terms, environmental costs are defined as representing the costs of damage that water uses impose on the environment and ecosystems and those who use the environment (e.g. a reduction in the ecological quality of aquatic ecosystems or the salinisation and degradation of productive soils). Resource costs are defined as the costs of foregone opportunities which other uses suffer due to the depletion of the resource beyond its natural rate of recharge or recovery (e.g. linked to the over-abstraction of groundwater).
However, the distinction in the Wateco guidance between environmental and resource costs is not clear-cut. Also their estimation and use are still surrounded by a lot of discussion. It is not so much the general idea that water use imposes a cost on the water environment and those who use this environment, which often remains unaccounted for in policy and decision-making, that is disputed. Cause for debate is primarily (1) the explicit distinction between environmental costs and resource costs, (2) their practical measurement, either through cost or benefit based approaches, and (3) the extent to which these costs have already been accounted for in existing pricing and financing mechanisms.

Although no explicit distinction is made in the existing environmental economics literature between environmental and resource costs (for an overview see Schaafsma and Brouwer, 2006), the terms have been interpreted and defined as two separate concepts in the Wateco guidance and subsequently also in the ECO2 information sheet3. The distinction between environmental and resource costs in the Wateco guidance seems to be based primarily on the question whether the water resource is depletable (non-renewable) or not, while the ECO2 information sheet introduces the notion of use and nonuse values into the equation, where environmental damage costs could be argued to refer to nonuse values primarily attached to a healthy functioning aquatic ecosystem and the opportunity costs of those who use the water environment to the corresponding use values. Both the Wateco guidance and the ECO2 information sheet use the term ‘opportunity costs’ to describe and delineate resource costs and this seems to be considered a distinctive feature between environmental and resource costs. However, the concept of opportunity costs is a generally applicable concept in environmental economics, and may equally apply to environmental and resource costs.

An important reason for their distinction is also to account for the fact that across Europe important differences exist in terms of water quantity and water quality management issues. The distinction is largely driven by concerns in South European MS facing (extreme) water scarcity and allocation problems, nowadays in the context of the WFD more generally referred to as resource costs. Resource costs here refer to competing and/or conflicting water use for different socio-economic activities (domestic household, agriculture, industry), primarily as a result of its limited quantitative availability in time and space (see also section 4.7), but not exclusively as there exists a close relationship between water quantity (scarcity) and water quality. Water quality issues (pollution, eutrophication, loss of biodiversity) generally are the most important water management concern in North European MS, usually referred to as environmental costs.

More important than the discussion about the exact definition and distinction between the two terms is the question (1) what exactly constitutes damage, to the water environment and those who use the water environment due to excess pollution and/or abstraction, (2) what are the benefits of WFD implementation aimed at more sustainable water use (reducing or eliminating excess pollution and/or abstraction), and (3) how are these damage costs and WFD benefits best measured (in a valid and reliable way).

In theory, damage arises when there is a discrepancy between some reference and target point or situation. The latter can be measured, for instance, through existing environmental norms or standards or the right people attach to a clean environment and the provision of sufficient and clean water. In practice, sometimes also a point in the past, when pollution levels and corresponding damage costs were lower, is taken to represent the target situation. An example is the discharge of waste water into a water course at a rate (e.g. tons of N per year), which exceeds some permitted rate (in tons per year) and hence results in a eutrophic water system with negative consequences for both the biological diversity of the water system and the recreational amenities provided by the water course, including the possible negative effects on human health if the specific water body is also used for recreational swimming. In the context of the WFD, it seems logic to use the expected water status in 2015 as the reference situation, but other reference situations may also be appropriate. The same applies to the target situation. It seems logic to relate the target situation to the environmental objectives of the WFD, i.e. good ecological water status in 2015. However, other target situations may also apply. Hence, there is a strong dependence of environmental and resource costs on the physical status of the water system and knowledge and information about this physical status. This includes the damage caused to the water system as a result of pressures exerted on the water system, such as the extent to which the system’s natural rate of recharge or recovery has been impaired by a specific water use. Other general damage categories include eutrophication, salinisation, dessication, loss of biological diversity and morphological changes to a water system. This type of information is crucial to the subsequent economic estimation of environmental costs. The physical status of a water body or water system provides the basis for the estimation of the environmental and resource costs in economic terms. If this information is not available, environmental and resource costs cannot be assessed.

Coming to the economic valuation of the costs and benefits of water use, a first important observation is that such economic valuation studies associated with the specific scope and objectives of the WFD are rare or non-existent in Europe and elsewhere. An example is given in Box 5. The main problem when considering economic choices related to water is that a competitive, freely functioning market does not exist for many water related uses. Many services provided by water resources are cost free, and users are not charged for their use (see section 3.4). The reasons for this stem from water characteristics such as the fact that (a) water is an essential commodity the value of which for a basic survival amount is infinite, (b) water has natural monopoly characteristics, (c) property rights for water resources are often absent and difficult to define, (d) water is a ‘bulky’ commodity, thereby restricting the development of markets beyond the local area (see the next section about market based instruments).

Given the absence of a functioning market mechanism for many water uses and services, and in line with the increasing water conflicts and need for more efficient allocation, it is necessary to have knowledge and information of the marginal value or benefits of the resource in its alternative uses. The main aim of economic valuation is to provide the value information that would normally be provided in the market for efficient water resource allocation, including pollution (see the method description in section 4).

Finally, it is important to point out the difference between price and value. Price and value are two distinct concepts. Water per se or water of a certain quality has value, but may have no price. Alternatively, water (of a certain quality) may have a price, but this price does not reflect the true economic value of water (or the value of a specific quality level). Hence the need to calculate and impute shadow prices for water use and the goods and services provided by aquatic ecosystems under circumstances where existing prices are believed to undervalue or overvalue the resource and economic efficiency is an important guiding principle in water (allocation) management. For the estimation of shadow prices different methods are available (see the method description in section 4). These estimated shadow prices can subsequently be used to calculate the total economic value of a specific water use (e.g. the amount of water consumed times its shadow price), including environmental and resource costs of water use. Furthermore, existing pricing systems can be modified based on these shadow prices to allow economic market mechanisms of supply and demand re-allocate scarce water resources across different water uses and users and reduce possible inefficiency in current resource use (see the next section about economic instruments).

3.6. Design, application and evaluation of economic instruments

The WFD requires in article 9 that water pricing policies by the year 2010 provide adequate incentives for users to use water resources efficiently and thereby contribute to the WFD environmental objectives. The WFD furthermore refers to the use of economic instruments as part of the program of measures in its consideration 38 and in article 11. However, the role of economic instruments in the programs of measures is unclear, and their costs of implementation highly uncertain (Brouwer, 2005) (see section 4.8). In many cases they are included as separate ‘measures’ in the programs of measures. In practice, economic instruments and technical measures are often complementary and implemented consecutively.

Currently, many water users do not pay for many if not most types of water use in many EU MS. The most important pricing mechanisms currently in use in EU MS are market prices for drinking water and taxes and charges for wastewater collection and wastewater treatment paid by domestic households and industry. Often, the charges and taxes are set at levels, which just cover direct financial costs, but lack incentives for more efficient and sustainable water use. Domestic household and industrial water demand are generally characterized by low price elasticities (e.g. Dalhuisen et al., 2003), the same applies for irrigation water demand in agriculture (OECD, 1998; 1999). Additionally, in a number of EU MS agriculture and industry also pay an emission or pollution tax, charge or levy for direct discharges of pollutants to surface water and/or a groundwater or surface water abstraction charge, tax or levy. For all other kinds of water use economic instruments for efficient water use are usually absent.

Economic instruments, often also referred to as market-based instruments, are used to correct for the failure of markets to protect the (water) environment and face increasing popularity (e.g. EEA, 2005; 2006). The most important economic instruments are emission charges and environmental taxes where economic agents pay for their use of scarce environmental assets, and emission trading where emission quota are allocated and traded so that a unit price results, which reflects the environmental scarcity. Although the latter type of instrument has distinct economic advantages over ‘command and control’ (e.g. pollution source based standards) and taxes and charges, its application is very limited across Europe. Most experiences with tradable emission permits and water quality are found in the US (see Box 6).

Article 10 in the WFD requires MS to ensure the implementation of emission control based on best available techniques (BAT), emission limit values or, in the case of diffuse pollution, best environmental practices, cross-referencing to other EU legislation such as the IPPC Directive (96/61), the UWWT Directive (91/271) and the Nitrates Directive (91/676). The combined approach of pollution source based and water quality based requirements seems to leave little opportunity for economic instruments. Each and every source has to comply with stringent BAT or comparable standards. Allowing less efficient pollution sources to buy pollution credits for example from more efficient pollution sources and hence to exceed the imposed standards, is only possible if the WFD and other relevant EU laws were to be changed. This is what happened in the case of the EU greenhouse gas emissions trading system (Directive 2003/87). The IPPC Directive was amended so as to exclude emission limit values for greenhouse gases in permits for installations participating in emissions trading. Similar changes for water quality trading in an emission permits market will only be feasible if such trading is introduced on an EU-wide scale given the transboundary nature of many large rivers. Such an emission trading system is expected to be limited to fill the ‘gap’ between the water quality level that can be achieved with the currently available BAT to reduce pollution levels at source and the WFD required water quality level. This seems to reduce the scope for the introduction of a market based trading scheme, but filling the remaining ‘gap’ with market based instruments such as an emission permits system may be highly relevant in view of the fact that the ‘gap’ is found on the right-hand side of the cost-effectiveness curve (see method description in section 4), where the marginal cost per unit pollution (emission) reduction is relatively high.

4. ECONOMIC METHODS, MODELS AND INSTRUMENTS

4.1. Introduction

In this section, we will present an overview of the available economic methods, models and instruments to address and deal with the key issues identified and described in the previous section. Figure 4.1. once again shows the different steps in the economic analysis in the WFD (see section 2) and relates the steps and the activities under these steps to the available methods, models and instruments.

Figure 4.1: The role of economic methods, models and instruments in the economic analysis in the WFD

The main methods include cost-effectiveness analysis (CEA), cost-benefit analysis (CBA) and multi-criteria analysis (MCA). It goes without saying that CEA underpins the selection of a cost-effective program of measures in step 3. However, if these measures are screened and selected on the basis of other additional criteria too, as we saw in the previous section they may be, MCA may be a more appropriate method. Depending on the scale of WFD implementation and the required level of detail regarding the economic impacts of the proposed programs of measures, hydro-economic models can also be used to estimate the full economic costs and effects of the measures.

In the previous section the use and usefulness of CBA was already briefly discussed in the context of disproportionate costs. Non-market valuation methods provide important additional information for a fully monetized CBA given the public good nature of the expected benefits derived from the WFD implementation. One specific valuation method will be addressed in more detail here and that is benefits transfer. Benefits transfer is often used in policy analysis and appraisal. It is an attractive low cost alternative for carrying out original valuation research to value the economic benefits of WFD implementation.

As for the assessment of the full costs and effects of proposed programs of measures in the CEA, also in the case of the evaluation of possible disproportionate costs extended hydro-economic models based on a full coverage of the economic system (e.g. input-output models or applied general equilibrium models) may play an important role. These models can also be used for shadow pricing purposes in the context of article 9, and the modeling of the effect of the introduction of new or modified economic instruments such as an emission permit system to stimulate more efficient and sustainable water use on the economy as a whole and the various economic sectors.

4.2. Cost-effectiveness analysis

The purpose of a cost-effectiveness analysis (CEA) is to find out how predetermined targets, e.g. threshold values for nutrients or other pollutant loads in a catchment, can be achieved at least costs. Theoretically speaking, the least cost allocation of pollution abatement strategies is found if the marginal costs of the proposed measures are equal. The marginal costs of these abatement measures can for example be defined as the increase in total abatement costs when pollution loads are decreased by 1 ton or 1 kilogram per year. As long as marginal costs are not equal, it is theoretically possible to obtain the same level of pollution reduction at lower costs by shifting emission reduction from high cost measures to lower cost measures. In the WFD a CEA has to be carried out at catchment level. Hence, also the spatial distribution of costs plays an important role and the question where measures should be taken. This is visualised in Box 7.

The various steps distinguished in a CEA are described below:

These steps are taken in sequence, but important feed-backs may exist between steps. As information becomes available about the problem, the source-effect pathway and possible solutions, the same step may be revisited several times. The outline of the various steps shows that carrying out a CEA is a multi-disciplinary exercise, requiring the input of and collaboration between different scientific disciplines, such as natural scientists, economists and technical engineers, but also the input of policy and decision-makers as they determine the scope and objective of the analysis.

A number of approaches are used in practice at varying levels of complexity, scale, comprehensiveness and completeness for carrying out a CEA. These are discussed, for example, in Zhang and Folmer (1995). A distinction is made between bottom-up and top-down approaches. The bottom-up approach focuses on technological details of measures and their impact on individual enterprises (micro level), whereas top-down approaches usually consider the wider economic impacts of pollution abatement measures and strategies, often without detailed technical specification of the proposed measures (macro level). Bottom-up approaches can also be characterised as technical engineering approaches, often including detailed information about the technical characteristics of production processes and only limited information about the financial engineering costs of emission abatement technologies. Top-down approaches on the other hand focus more on the economic relationships and consequences involved and less on the technical specification of measures. Examples of bottom-up approaches include ad-hoc approaches comparing a limited number of abatement technologies usually on a very local scale based on their engineering costs and emission reduction capacity (see Box 8) and the use of dynamic optimisation models where various abatement measures and technologies at enterprise or sector level are automatically prioritised with the help of linear programming (LP) techniques. Examples of top-down approaches are input-output and computable general equilibrium models. The inclusion of indirect effects depends on their relevance and the role they are expected to play in the decision-making procedure.

4.3. Cost-benefit analysis

Cost-benefit analysis (CBA) is carried out in order to evaluate and compare the economic efficiency of alternative actions. The benefits from an action are contrasted with the associated costs within a common analytical framework. To allow comparison of these costs and benefits related to a wide range of scarce productive resources, including water resources, measured in widely differing units, money is used as the common denominator. This is where most problems usually start for environmental policy and project appraisal since many environmental resources such as water are often not priced in monetary terms. For many goods and services provided by water resources, there is no market on which they are traded, and therefore no market price is available which reflects their economic value. There are, however, several economic valuation methods, which allow placing a value on non-marketed goods and services (see section 4.4). The economic valuation of water uses and services compares the willingness to pay and opportunity costs of the goods and services supplied by water resources and the water environment. This means that a wide range of environmental goods and services can be explicitly recognised in the CBA process

In general, the following steps are followed in a CBA:

Like CEA, carrying out a CBA is a multi-disciplinary process, involving expertise from different fields and the input from policy and decision-makers. While economists are involved in all steps, environmental expertise of many kinds is also needed, especially in steps 2 and 5. In order to ensure that the policy options are technically feasible, input from technical engineers is required especially in step 3, and often also in step 4 to specify the exact nature of the policy measure and estimate the associated investment costs. Policy and decision-maker input is essential when defining the objective, which the policy measures are supposed to achieve, and when defining the baseline and policy scenarios, including current policy. A key role of the economist in the whole process is to frame the issue and develop the CBA framework so that all relevant socio-economic stakes and the stakeholders are included and the multitude of environmental studies that need to be undertaken are working towards answering the following two questions:
1)Is the action economically speaking worthwhile, that is, do the benefits outweigh the costs?
2)And if so, which policy option yields the highest net benefit?

A CBA compares the costs and benefits of different policy options in monetary terms. The results of this analysis can be interpreted as a B-C ratio, that is total benefits divided by total costs, where a ratio larger than one indicates that the policy measure is economically beneficial, or as a NPV, that is the present value of the net benefits, where a positive NPV indicates a welfare improvement. Strictly speaking, only those costs and benefits are included in a CBA that can be quantified in monetary terms. However, it will hardly ever be possible to monetise all impacts all the time: those impacts that cannot be monetised are often left out of the analysis (see Box 9). Non-monetised impacts, if considered relevant, can nonetheless be included in a qualitative discussion accompanying the discussion of the CBA results.

While a textbook CBA requires that all impacts be monetised, in practice different approaches exist on how non-monetised impacts are included in the CBA. In some approaches they are listed as ‘Pro Memoriam’ items on the balance sheet, expressed in qualitative or quantitative form (Brouwer and Van Ek 2004). Pearce (1998) argues that in early CBA conducted in the UK, such impacts would have been either ignored entirely, left for a subsequent environmental impact analysis, or monetised only partly. This approach of monetising impacts where possible, and including them in another form where monetisation is not possible marks a deviation from the textbook ideal, but does not discredit the method as such.

An important distinction in CBA is that between a financial and economic CBA:

A financial CBA, also referred to as a cash-flow or a financial analysis, evaluates advantages and disadvantages of a project or measure in terms of the expenditures and earnings directly associated with its implementation for the investor. Originally devised for investment decisions, the tool can also be used to assess budgetary impacts of policies.

An economic CBA evaluates the costs and benefits of a project or measure in a broader sense, taking into account all positive and negative welfare effects, on the (national) economy as a whole and all relevant stakeholders. The costs and benefits addressed in an economic CBA may include indirect (second-order) effects and non-priced external effects on society and the environment. If such externalities are included in the analysis in monetary terms, it is usually also referred to as an extended CBA.

In practice, government policies are often evaluated primarily on the basis of their financial (budgetary) costs, as these can be assessed relatively easily. The calculation of economic costs and benefits, especially non-priced external environmental effects, is a more difficult task. This will be addressed in the next section. An economic or extended CBA is the more appropriate method for evaluating public water policies, since government interventions are often related to the provision of public goods, having an impact on society as a whole. Such impacts should consequently be valued and evaluated from a societal perspective, not the perspective of the investor only (e.g. government).

4.4. Non-market valuation methods

In economics, value is expressed as the degree to which people want to give up scarce resources, such as money or time, to acquire or retain something. Value exists in this sense only through the interaction between a subject (individual) and an object, and is therefore not considered an intrinsic quality of something (Pearce and Turner, 1990). As in other social sciences, the value people attach to something is based upon a hypothesised positive relationship between their observed behaviour or verbal responses and that value. In economics, an individual’s value is usually revealed through market behaviour and measured in money terms by an individual’s willingness to pay (WTP), for instance for a water quality improvement, or willingness to accept (WTA) compensation for a water quality decline. WTP and WTA cannot be used interchangeably. Although the WTP approach has become the most frequently applied and has been given peer review endorsement through a variety of studies (e.g. Arrow et al., 1993), the question which concept is most appropriate in the context of water resources depends on the specific circumstances and the property rights regime associated with the specific water use.

Aggregated across those who benefit from natural resources and their services and who will hence be affected by any change in their provision level, including quality level, the aggregated WTP or WTA amount provides an indicator of their total economic value (TEV) (see Box 10). Environmental economists have introduced a taxonomy of this TEV, distinguishing between use values and non-use values, in order to account for the various reasons and motives people may have to value environmental change. Use values are associated with the actual or potential future use of a natural resource (e.g. drinking water, fish consumption, irrigation water). Non-use values are not related to any actual or potential future use, but refer to values attached to the environment and natural resource conservation based on considerations that, for example, the environment should be preserved for future generations or because plants and animals also have rights.

A range of valuation techniques exists for assessing the economic value of the functions performed by aquatic ecosystems (Box 11). Many water resources provide goods and services, which are not traded in markets and therefore remain un-priced. The valuation techniques developed over the past decades attempt to assess the relative economic worth of these goods and services using non-market valuation techniques. Depending on the nature of

the specific environmental change in the water system, the presence of a market where the goods and services involved are exchanged and data availability, economic values can be estimated using direct or indirect market and non-market based valuation techniques. More detailed information about the underlying theory and practical implementation of these techniques can be found in general texts like Randall (1988), Braden and Kolstad (1991), Freeman (1993), Hanley and Spash (1993), Pearce et al. (1994) and Young (2005).

An important distinction is between valuation techniques, which estimate benefits directly and those, which estimate costs as a proxy for benefits. For instance, estimating damage costs avoided, defensive expenditures, replacement costs, substitute costs or restoration costs as part of an economic valuation exercise suggests that these costs are a reasonable approximation of the benefits that society attributes to the resources in question. The underlying assumption is that the benefits are at least as great as the costs involved in repairing, avoiding or compensating the damage involved. These techniques are widely applied due to their relative ease of estimation and availability of data, but it is important to be aware of the limitations in terms of the information they convey with respect to economic benefits. Investment by public bodies in conserving ecosystems may also represent a surrogate for aggregated individual WTP and hence social value. These ‘public prices’ paid for resources are also used to approximate the value society places upon them, for instance the costs of designating an ecosystem as a nature reserve.

Market valuation means that existing market behaviour and market transactions are used as the basis of the valuation exercise. Economic values are derived from existing market prices for inputs (production values) or outputs (consumption values), through more or less complex econometric modeling of dose-response or damage functions. Examples include the economic value of fish, which is sold on a fish market (market analysis), the costs of replacing impaired environmental riparian functions such as nutrient retention and export through the installation of a wastewater treatment plant (replacement costs) or the costs of a water filter on tap water (avertive behaviour or defensive expenditures).

Where market prices exist for water resources, these may have to be adjusted for market distortions such as taxes or subsidies in order to obtain the real or shadow prices, but otherwise they are likely to provide a relatively simple means of assessing economic value. However, theoretically these estimations based on market prices are still not the same as the total economic value since they do not include the consumer surplus.

In the absence of market prices for water resources, the economic value of the goods and services provided by water systems can be estimated with the help of direct and indirect non-market valuation methods (e.g. Johansson, 1987; Mitchell and Carson, 1989; Freeman, 1993; 2003). Non-market valuation means deriving economic values in cases where such markets are non-existent or distorted. Direct methods (also called stated preference methods) refer to contingent valuation (CV), discrete choice experiments (CE), and contingent ranking (CR) techniques, where individuals are asked directly, in a social survey format, for their WTP for a pre-specified environmental change. WTP can also be measured indirectly by assuming that this value is reflected in the costs incurred to travel to specific sites (travel cost studies) or prices paid to live in specific neighborhoods (hedonic pricing studies) (also called revealed preference methods). The latter two approaches measure environmental use values through revealed preferences, while CV and CE are believed to be able to also measure non-use or passive use values through stated preferences. Of these methods, CV is probably the most widely applied method in contemporary valuation research (Carson et al., 1995; Bateman and Willis, 1999). However, the validity and reliability of contingent valuation methods and their results are contested (e.g. Foster, 1997), especially in the context of non-use values. Moreover, conducting a contingent valuation survey can be a lengthy and resource-intensive exercise.

Irrespective of these criticisms, it is clear that valuation studies have a role to play in contemporary environmental policies, as they provide additional knowledge to support better decision-making (see www.aquamoney.org). It is important to apply and interpret economic valuation results in their appropriate context and to be aware of the pitfalls involved. However, this applies to most methods and techniques, in economics and any other field. Following best-practice recommendations is essential when applying the method in the field of water resources (for best practice recommendations, see for instance Arrow et al., 1993 and Bateman et al., 2002).

4.5. Benefits transfer

Instead of carrying out a new original valuation study, existing economic value estimates from previous studies can be used too. This is generally referred to as benefits transfer (Brouwer, 2000). Benefits transfer is a technique where the results of previous environmental valuation studies are applied to new policy or decision-making contexts. In the literature, benefits transfer is commonly defined as the transposition of monetary environmental values estimated at one site (study site) to another site (policy site). The study site refers to the site where the original study took place, while the policy site is a new site where information is needed about the monetary value of similar benefits.

In the field of environmental valuation, benefits transfer has been applied extensively in various contexts, ranging from water quality management (e.g. Luken et al., 1992) and associated health risks (e.g. Kask and Shogren, 1994) to waste (e.g. Brisson and Pearce, 1995) and forest management (e.g. Bateman et al., 1995). Costanza et al. (1997) extrapolated the monetary values of existing valuation studies to the flow of global ecosystem services and natural capital, and thereby raised a number of questions as well as heavy criticism about the reliability of benefits transfer (Box 12).

The most important reason for using previous research results in new policy contexts is that it saves a lot of time and money. Applying previous research findings to similar decision situations is a very attractive alternative to expensive and time consuming original research to inform decision-making. In practice, several approaches to benefits transfer can be distinguished, which differ in their degree of complexity, data requirements and the reliability of the results.

The first two approaches are most frequently applied, as they require relatively little data or expertise, and are usually not very time consuming. First, the average WTP value from another study can be used to predict the economic value of the benefits involved at the policy site. The old study ideally focuses on the same environmental goods and services, but may have been carried out at a different location or at a different point in time. A second approach is to estimate an average WTP value from the mean value estimates of several similar studies.

A third approach is to use an average WTP value, which is adjusted for one or more factors that are expected to influence the new value estimate at the policy site. For instance, average WTP may be adjusted for differences in income levels between the study and policy site. This means that in the old study information has to be available about the way income influences WTP.

Fourth, the entire WTP function from an original study can be used to predict mean WTP at the policy site. The estimated coefficients in the WTP function are multiplied by the average values of the explanatory factors in the new policy context to predict an adjusted average WTP value. This approach is expected to be more robust than the transfer of unadjusted average WTP values, since effectively more information is transferred (Pearce et al., 1994). However, it is also more data intensive as secondary information about all relevant factors has to be collected.

A fifth approach is to use a WTP function, which has been estimated based on the results of various similar valuation studies. The WTP function is in this case estimated on the basis of the summary statistics of different studies. This approach is usually referred to as meta-analysis.

A sixth approach is the use of a WTP function - either based on a single or multiple studies – where the coefficient estimates are adjusted when transferring the estimated value function based on prior knowledge about the impact of the explanatory factors on WTP in the new policy situation.

A number of criteria have been identified in the literature for benefits transfer to result in reliable estimates (e.g. Desvousges et al., 1992; Loomis et al., 1995; Brouwer, 2000):

1.good quality data
2.similar environmental goods and services
3.similar sites where these goods and services are found
4.similar populations of beneficiaries
5.similar market constructs
6.similar market size (number of beneficiaries)
7.similar number and quality of substitute sites where the environmental goods and services are found.

Study quality is an important criterion, which can be assessed in a number of ways. Above all, one can look at the internal validity of the study results, that is, the extent to which findings correspond to what is theoretically expected. This internal validity has been extensively researched over the past three decades in valuation studies. Studies should contain sufficient information to assess the validity and reliability of their results. This refers, among others, to the adequate reporting of the estimated WTP function, including the applied statistical techniques and the definition of variables.

Thus, while benefit transfer provides a quick and cheap alternative to original valuation research, some conditions must be met if it should provide reliable results. Above all, the local circumstances and conditions in the new decision-making context need to be close enough to the ones prevailing in the original research. The risk of obtaining misleading results may be controlled and reduced by integrating more explaining variables into the transfer, however this also increases the data requirements and the complexity of the analysis. Also, the possibilities of conducting a sound and reliable benefits transfer hinge on the number, quality and diversity of valuation studies available – the larger, the better and the more diverse the existing set of studies is, the more likely there will be a primary study that is close enough to the policy site for results to be transferable.

4.6. Multi-criteria analysis

MCA is applicable to cases where a single objective or single criterion approach like CBA, which compares options based on their economic efficiency only, is insufficient. Instead, a MCA accommodates a range of social, environmental, technical, economic, and financial criteria. MCA is especially suitable in situations where multiple environmental and social objectives are pursued, which cannot easily be expressed in monetary terms. MCA techniques are often integrated with participatory approaches, involving different policy and stakeholder groups (see for example the UNEP ‘Value Base Assessment’ procedure and use of MCA in relation to transboundary water management).
One setback of MCA in these participatory approaches is that the method is often difficult to use for lay people. Most of them require an expert to explain how the method works, help users to define options, criteria and weights, as well as to choose the appropriate aggregation procedure. An important question is which approach the experts and the end-users of the results feel most comfortable with when making effects comparable and commensurable, that is through economic valuation of effects or standardisation and weighting procedures of impacts based on the statistical standardisation of impacts and criteria weights as in MCA.

The general steps in MCA are described below:

As for a CBA, MCA starts off by identifying and framing the problem to be assessed, and formulating the objectives that are to be reached through the policy intervention. Such objectives may include environmental targets as well as social cohesion, employment generation or income distribution. This is followed in the second step by the identification of different options to achieve the objectives, a step that may involve the participation of different stakeholders and the wider public. If too many potential measures are identified, an initial screening may be helpful to pre-select the available options. This screening can take place based on a qualitative MCA.

A third step is to identify the criteria and possible sub-criteria against which the different options should be evaluated, along with their suggested indicators and measurement units. The identification of the criteria will be closely related to the objectives identified in the first step. The criteria can be measured in monetary and non-monetary terms, but they may also be qualitative in nature. An important difference between MCA and CBA is that the former is able to relate information in different forms to multiple objectives measured through multiple criteria. MCA incorporates environmental, economic and social effects as in a CBA, but trades them off against each other in a different way, namely based upon multiple criteria measured in different units. While the weighing procedure and trade-off of different criteria is more complex for a MCA than for a CBA, it means that a MCA does not require the use of monetary valuation studies in the same way that a CBA does.

In the following step, the policy options identified in step 2 are assessed in terms of the criteria defined in step 3. To do this, the policy options are evaluated compared to a common pre-defined baseline scenario, mostly a ‘business as usual’ scenario. The performance of the different options against the baseline scenario are then presented in a scorecard, also referred to as an effects table, performance or impact scoring matrix (Box 13). A score is provided for each option against each criterion, measured in different units. These scores are then standardized between 0 and 1. For this, several techniques are available depending on the starting point and end point of the effect scores and the expected distribution of intermediate values.

In step 7 the different criteria are prioritized by attaching weights to them. These weights can be assigned by the analyst, the decision maker, or they can be elicited from stakeholders through some form of participatory consultation process. Another option is to involve a panel of experts (Delphi method). A direct analysis of the performance matrix through policy maker screening of alternative options is of course also possible. The determination of weights is crucial, at the same time it also introduces an element of subjectivity into the decision making process. The weighting procedure basically corresponds with the valuation step in a CBA. A sensitivity analysis is essential to assess the effect of different weights on the outcome of the analysis.

For the calculation of a global effect score, the weighted scores have to be aggregated. For this aggregation a number of approaches have been developed. The following two main approaches can be distinguished (DETR, 1998): the compensation method and the outranking method. The compensation method is the best-known method where a global score is calculated for each policy measure in the form of a weighted arithmetic average of the scores attributed to that measure for the different criteria. The method called ‘compensatory’ because the calculation of the weighted average makes it possible that low scores on one criterion are compensated by high scores on another. The outranking method on the other hand is used if not all criteria are considered commensurable, and therefore no weighted aggregate score can be calculated. The analysis is based on paired comparisons of measures with respect to individual criteria. All possible comparisons are analyzed and synthesized in terms of ‘policy measure A is at least as good as policy measure B in relation to a majority of criteria, without being altogether too bad in relation to the other criteria’ (DETR, 1998).

The final step in a MCA is to prioritise, compare and order the alternative policy measures based on the evaluation results. Different methods can be used to this end. Depending on the method chosen, a MCA can be used to identify a single most preferred option, to rank options, to short-list a limited number of options for subsequent detailed appraisal, or to distinguish acceptable from unacceptable options. The prioritisation of different options should also involve a sensitivity analysis, investigating how the ranking order of the options changes if key parameter values and criteria weights were to be changed.

4.7. Integrated river basin models

The complexity of interactions between water and economy is increasingly translated into formal (mathematical) reduced-form models by linking relevant hydrological and biogeochemical structures and processes to economic ‘laws’ of supply and demand of water services. Historically, these models have been developed by water and civil engineering scientists, focusing on single and multiple objective decision-making and trade-offs (e.g. Dudley, 1972; Braat and Lierop, 1987; McKinney and Cai, 1996; Andreu et al., 1996; Cai and McKinney, 1997; Rosegrant et al., 2000). Water use in agriculture (irrigation) has been the prime focus of many of these models given the fact that agriculture is the largest water consumer in the world (FAO, 2004). These models often include a detailed coupled hydrological and geological module - and in some cases also hydraulic and biogeochemical modules - to control for the hydro-geological heterogeneity in a basin area and are based on node networks.

Node networks are graphical delineations of water flows and stocks in a watershed or basin into different water quantity and/or quality balance (monitoring) stations linked to specific water demand and supply. For each geo-referenced node, a water demand and supply function is estimated based on the geographical unit’s hydro-geological and biogeochemical characteristics. In the case of agriculture, the demand and supply functions are based on an agronomic model, such as a crop yield function, which depends on factors like soil, crop acreage, rainfall, crop evapotranspiration and irrigation system characteristics. Economic behavior is usually included through a profit maximization objective function, where fixed and variable production costs are subtracted from the yield benefits subject to the natural resource constraints such as land and water availability. The latter is obviously dependent on the hydro-geological conditions involved, including water supply and water quality constraints.

The key to integrated water modelling is that water systems perform economic functions and can be used as a source and a sink. After some degree of transformation, water is used as a source for economic consumption like drinking or recreation, and in economic production as an input factor in crop and food production, energy, paper, textile or metal production. At the same time, water is also used as a sink for the by-products of economic production and consumption processes such as the emission of polluting substances into surface and groundwater bodies. The interaction between the water and economic realm works both ways: water is transformed for economic use and the impact of economic use on water availability and quality consequently has implications in both the short and long term for the transformation process to adapt water for economic use. It is this latter dynamic feedback mechanism, which is often ignored and omitted in integrated water models.

Following Braat and Lierop (1987), a distinction is made in the literature between two different approaches to integrated hydro-economic model development, i.e. (1) models which allow for an effective transfer of information from one component to the other: the compartment or modular approach and (2) the holistic approach.

In the modular approach a connection is built between the hydrological and economic model, and output data from one module usually provides the necessary input for the other. In principle, the modules operate independently of each other and systems of equations are solved in an exogenous way (input variables from one model into the other are exogenous). In holistic models, variables that are exogenous in a modular approach are solved endogenously in a system of equations (Cai and Wang, 2003). However, under the modular approach, a loose connection exists between the different hydrologic and economic components. The various sub-models can be very complex and the main problem is to find the right transformation of data and information between sub-models. In the holistic approach there is one single unit with both the hydrologic and economic component tightly interwoven in a consistent endogenous model (see the example in Box 14). In order to be able to solve the complexity of simultaneous equations the different components have to be represented in a very simple way (McKinney et al., 1999).

So, whereas information transfer between the various compartments or sub-models is one of the most important technical obstacles in the modular approach, the most important issue in the holistic approach is to find one single technique and denominator for the variable quantities and represent both the simplified hydrological and economic component in a meaningful way.

Important limitations to integrated modelling include (McKinney et al., 1999):

• Hydrological models are often based on simulation techniques, whereas economic models usually use optimization techniques.
• Different spatial scales: water bodies, water sheds or basins usually are the geographical unit in hydrological models, while economic models often refer to administrative boundaries of a region (county, province, state) or a country as a whole.
• Different time intervals and horizons: time scales in hydrological models often refer to days, months or seasons (summer and winter), while in economic models the time scales (intervals and horizon) are usually larger than that (years).

Besides the distinction between holistic and modular approaches, integrated hydro-economic models can furthermore be classified as (1) water use models, (2) water system models and (3) effects models. An overview of these different model types is given in Figure 4.2.

Figure 4.2: Different types of integrated hydro-economic models

Most integrated hydro-economic models are water system models, i.e. models based on detailed node networks of water and substance balances throughout the river basin, linked to an economic activity through a demand function. This demand function often depends on fixed (exogenous) technical input-output parameters of the economic production process involved (e.g. irrigation demand from agriculture), and reflects at best a partial economic equilibrium system of demand and supply equations.

Water use models are integrated models based on economic demand (consumption) and supply (production) functions, which are related to different forms of water use where water is an essential input in consumption and production processes. As most existing water system based hydro-economic models, water use models often reflect partial equilibrium states. Examples include input-output models of direct and indirect water use (e.g. Velázquez, 2005). More recently computable general equilibrium models are also used to link economic activities to water use, including water extraction and water pollution through the emission of polluting substances to water bodies (Brouwer and Hofkes, forthcoming).

Effects models incorporate changes in the water system and their effect on the economic system. Most existing models are partial equilibrium models, based on production function approaches, where water is one of the input factors and changes in the availability or quality of water needed for the production process at hand is assessed through estimated dose-effect relationships, which are subsequently related to market prices in order to derive a marginal value for water use.

Sometimes also ‘meta-models’ are distinguished as a separate class of modelling tools. Generally, meta-models are used to construct and develop general frames around specific problems analyzed with the help of a variety of data, expeert judgements and models. An example is given in Box 15. Meta-models integrate simulation results from sub-models (e.g. an economic optimisation model and a water quality simulation model) in a cause-effect framework. They are called meta-models because they only include the results of underlying models, where underlying model response surfaces are summarised for example in conditional probability distributions.

Finally, in the context of the WFD, a distinction can be made across European Member States between modelling efforts focusing on water scarcity, water allocation problems and the estimation of shadow prices and opportunity costs of alternative water use (especially in southern Europe), and modelling efforts focusing on water pollution, applying primarily bottom-up linear programming approaches to assess the cost-effectiveness of water quality measures (especially in northern Europe). An overview of existing integrated hydro-economic models is provided in the annex to this paper, based on the international workshop about the potential role of integrated river basin modelling in the WFD held at in Amsterdam in November 2005 (www.ivm.falw.vu.nl/watereconomics).

One of the workshop outcomes was that integrated models have a role to play in the economic analysis in the WFD in article 5 (river basin characterization), article 11 (selection of a cost-effective programme of measures), article 9 (economic instruments), and article 4 (disproportionate costs).

In order to be able to predict the impact of water policy changes to reach the environmental WFD objectives on both the water and economic system, the causal relationship between water and economic production and consumption processes has to be established first, for example through the article 5 reports and the use of integrated river basin information systems, as described in section 3.1. This includes for the selection of a cost-effective programme of measures the physical dose-effect relationships in order to be able to assess the impact of end-of-pipe or process-integrated measures on the water system. Furthermore, if large-scale interventions are foreseen, the economic impacts may not be limited to the sector involved, but cause a chain reaction across interlinked sectors. For example, if measures in agriculture result in substantial changes in inputs and outputs, the whole agribusiness may be af