Water Resources and Economics 7 (2014) 66–81

Contents lists available at ScienceDirect

Water Resources and Economics

journal homepage: www.elsevier.com/locate/wre

Experimental valuation of Dutch water resources

according to SNA and SEEA
Bram Edens n, Cor Graveland
National Accounts Department, Statistics Netherlands, The Hague, Netherlands

a r t i c l e in f o a b s t r a c t

Article history: The valuation of water resources is an important element for

Received 15 May 2014 sustainable water resource management. Valuation exercises
Received in revised form depend on their purpose and context of use. The objective of this
13 October 2014
paper is to investigate methods to value water resources that are
Accepted 14 October 2014
consistent with national accounts principles. This implies that we
look for exchange values that exclude consumer surplus. While
Keywords: water resources provide a range of benefits, in this paper we follow
System of Environmental Economic the measurement boundary of the System of Environmental
Accounting (SEEA)
Economic Accounting (SEEA) Central Framework and restrict
Valuation of water
ourselves to the extractive use or provisioning service of various
Wealth accounting
Ecosystem services types of water resources, surface- ground-, and soil water as
Resource rent recorded in the physical supply and use tables. The main users/
Replacement cost uses that we value are agriculture (predominantly soil water), the
water supply sector (mainly surface and groundwater) and
industries (surface and sea water primarily for cooling). We obtain
a value of water resources for the Netherlands in 2010 of
approximately 26 billion euros, which is about 10 percent of the
value of natural capital currently included in the Dutch balance
sheet. We find that the resource rent method has limited feasibility
due the occurrence of negative rents or difficulties to identify the
return to water in the rent. The replacement cost method has
better potential but more research is clearly needed to improve
estimates.
& 2014 Elsevier B.V. All rights reserved.

n
Correspondence to: Postbus 24500, 2492 JP Den Haag, The Netherlands. Tel.: þ31 70 337 5428;
fax: þ31 70 337 5994.
E-mail address: b.edens@cbs.nl (B. Edens).

http://dx.doi.org/10.1016/j.wre.2014.10.003
2212-4284/& 2014 Elsevier B.V. All rights reserved.
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 67

1. Introduction1

The need to treat water as an economic good has been recognized as an essential component of
integrated water resources management [1]. For instance, the valuation of water is requested by the
Water Framework Directive [2] in articles 4 and 9. In the statistical community, concurrent with the
surge of interest in natural capital and wealth accounting (e.g. the World Bank's Wealth Accounting
and the Valuation of Ecosystem Services partnership), there is a growing demand to value water
resources as they constitute an important part of the natural capital base. According to the 2008
System of National Accounts (SNA) [3], water resources need to be valued as part of the National
balance sheet in situations when their scarcity leads to use restrictions. The System of Environmental
Economic Accounting (SEEA) Central Framework [4] – a satellite system to the SNA – discusses
valuation of water and the SEEA Water [5] dedicates a whole chapter to the valuation of water
resources. To date there are only a few countries that have experimented with valuing water resources
in an accounting context, most notably Mexico (focussing on the depletion of water resources) [6],
Australia [7] and Namibia [8]. It appears that so far no country has been able to come up with credible
estimates.
In the environmental economics literature there is a growing experience with the valuation of
various types of water resources according to a range of techniques [9,10]. For the choice of valuation
technique, the context of use is very important [11]: many of the traditional valuation techniques such
as contingent valuation are welfare based assessments which include consumer surplus whereas
integration in the national accounts requires consistency with national accounts principles. The
National accounts aim to measure so-called exchange values, which are the prices willing buyers and
sellers transact goods and services for, hence excluding consumer surplus.
The objective of this article is to obtain an experimental valuation of Dutch water resources that is
consistent with national accounts principles. Hereto, we will assess several valuation techniques, in
particular the resource rent and replacement cost approach. The implication of using an accounting
approach is that we try to be comprehensive, covering water uses by all economic activities, rather
than focusing in depth on the valuation of water use in one specific sector.
Water resources provide a range of benefits, which can be described and classified in various ways.
The SEEA Central Framework has a broad recognition of natural resources in physical terms (in
national accounts parlance ‘extends the asset boundary of the SNA') and supplements the monetary
description of economic activities contained within the national accounts with a physical picture of
the dependency of the economy on flows from and to the environment. Its valuation scope is fully
aligned with the SNA (in national accounts parlance it ‘keeps the SNA production boundary intact’).
It is only in the context of the SEEA Experimental Ecosystem Accounting (SEEA EEA) [12] that the
production boundary is extended in order to recognize the contribution ecosystems make in the form
of ecosystem services. In that regard, water resources (e.g. considered as aquatic ecosystems) may be
analyzed in terms of various services they provide, such as provisioning services (e.g. drinking water);
regulating services (e.g. breakdown of pollutants in surface and ground water); and, cultural services
(e.g. recreation). Some of these services will contribute to outputs (or benefits) that are already
captured in the national accounts, although the contribution of water resources is not being explicitly
recognized; others will lie outside the SNA production boundary.
The scope of this article is restricted to valuing the physical flows that are recognized in the SEEA
Central Framework [4] as crossing the boundary from the environment to the economy. This can be
considered as the extractive use of water resources or from an ecosystem accounting perspective as
the provisioning service of water. The values that we obtain although conceptually consistent with
accounting principles, will therefore likely go beyond the scope of the SNA.
Statistics Netherlands has a long tradition in water accounting. The NAMWA (National accounting
Matrix including Water Accounts) [13] consists of three types of accounts: water use, emissions to
water, and regional water accounts. The NAMWA also separately identifies transactions including

1
The views expressed are those of the authors and do not necessarily correspond with the official point of view of Statistics
Netherlands.
68 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81

taxes related to water from the national accounts. Eventually, valuation of water resources could be an
addition to the Dutch water accounts and the derived values be included on the Dutch balance sheet.
The outline of the paper is as follows. Section 2 provides an overview about the use of water resources by
Dutch economic activities and the main flows from the environment that we intend to value. Section 3
discusses valuation methods and data sources. Section 4 provides the analysis and results for the main
economic activities (agriculture, water supply sector, other industries) that abstract water as well as a
discussion of the valuation methods used. Section 5 contains a discussion of the outcomes and concludes.

2. Importance of water abstraction for Dutch economy

Water accounts provide a comprehensive picture about the dependency of economic activities on water
abstraction and use and the impact that economic activities have on water resources via emissions,
consistent with the national accounts. There are various types of water accounts. The physical supply and
use tables show the abstraction (or withdrawal) of water from various sources (groundwater, surface water
and soil water) by economic activities [4, p. 72–75]. An important aspect in this regard is the boundary that
is chosen in order to delineate water flows that cross the boundary between the economy and the
environment from water flows that remain within the environment. This boundary is based on the
production boundary of the SNA i.e. only water inputs that are used for productive activities in the SNA
sense2 are in scope [4, para 3.19] excluding exchanges of water that occur within the environment.3
Abstraction includes water used for cooling although the recommendation is to separately identify these
flows due to their magnitude, and the fact that these flows often are ‘in-stream’ and thus returned to the
inland water system. Abstraction of soil water is also included and defined as ‘the uptake of water by plants
equal to the amount of water transpired by plants plus the amount of water that is embodied in the
harvested product.’ [4, para 3.198]. This measure appears to be equivalent to the concept of ‘green water’ as
used in the hydrological literature [45].
Brouwer et al. [10, Table 3.1] provide a classification of water use outcomes. The use of water
resources that the SEEA Central Framework measures is close to ‘extractive use’ or ‘consumptive use’.
From an ecosystem services perspective the physical supply and use tables describe the provisioning
service of water by ecosystems [12].
Table 1 provides an overview of the abstraction from various types of water resources by economic
activities following the International Standard Industrial Classification of All Economic Activities (ISIC Rev. 4).
The largest user of water is the electricity and gas industry at 44 percent followed by agriculture at 32
percent and manufacturing at 16 percent. It is apparent that these economic activities often draw upon
different types of water resources, such as the electricity sector from surface water and agriculture from soil
water. The electricity and gas industry uses water predominantly for cooling purposes. In fact, about 55
percent of total water abstraction is for cooling purposes. The abstraction from non-fresh water sources
(marine and brackish water) abstracted along or near the Dutch coastline is completely used for cooling
purposes. Total use of water by Dutch economic activities in 2010 was 22,159 million m3. The water supply
and waste management industry (Section E) is responsible for only 8 percent of water use (of which the
water supply industry – ISIC 36 – for 5.5 percent).
In the period 2001–2010 agriculture is responsible for on average about 7 percent of total groundwater
abstraction and less than 1 percent of the total abstracted fresh surface water. The evapotranspiration that
occurs on agricultural land – which is classified in the SEEA Central Framework as use of soil water –
however gives a much larger picture of water use compared to the abstractions of just ground and surface
water (see Table 1). With an agricultural area of over 2 million hectares in the country, crops take water

2
Production is defined in the SNA as "processes or activities carried out under the control and responsibility of institutional
units that use inputs of labour, capital, goods and services to produce outputs of goods and services" [3, para 5.5], hence
excluding natural processes. Own account abstraction for household consumption is included.
3
Where the boundary has to be drawn precisely, has been the subject of debate during the SEEA revision process. In fact,
one of the revision issues was whether water in artificial reservoirs should be considered as a produced asset. It was eventually
decided that water enters the economy at the moment it leaves the reservoir (for hydropower generation), not at the moment it
enters the reservoir, see: 〈http://unstats.un.org/unsd/envaccounting/seearev/issue.asp?iID=25&vID=1〉.
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 69

Table 1
Abstraction of water by Dutch economy, 2010.
Source: [14–16] with minor adjustments.

Water Unit Economic activity

resource
Agriculture, Mining Manufacturing Electricity Watersupply and Other Total
forestry and and and gas waste water
fishing quarrying supply management use

Ground mln m3 96 0 142 5 763 . 1006

Cooling mln m3 . 0 65 0 . . 65
Other use mln m3 96 0 77 5 763 . 941
Surface mln m3 26 1 3350 9693 1006 . 14,076
Fresh mln m3 26 1 804 – 1006 . 1837
Fresh – mln m3 . . 2273 5699 . . 7972
cooling
Salt and mln m3 . . 273 3994 . . 4267
brackish
Soil water mln m3 7076 – – – – . 7076
3
Total mln m 7198 1 3492 9699 1769 . 22,159

. Data not available

– Nil or less than half of unit concerned

from the available soil moisture in both the unsaturated and for a limited part from the saturated soil zones
with their roots. With an uptake (crop transpiration) of on average over 300 mm per growing season and
over 350 mm for the evapotranspiration as a whole, the agricultural sector takes over 7000 million m3 from
soil water resources, compared to the national actual evapotranspiration figure of over 17,000 million m3
annually. This implies that roughly 45 percent of the evapotranspiration occurring at the national territory
concerns flows ‘from the environment to economy’. Only a small part (o1 percent) of the total uptake by
plants is actually included in agricultural products. Part of the evapotranspiration volume is used up in
urban areas, which contain for instance sports fields. Although these constitute economic activities with
relevant soil water uptake, this type of soil water use is ignored in this study. Dutch land use statistics show
that the land area of the Netherlands consists of around 1 percent of sports grounds so the effect of this
omission on our results is small.

3. Methods and data sources to value water resources consistent with accounting principles

Numerous methods exist to value water resources [10,9,5].4 These methods each have their
advantages and drawbacks, but importantly often also differ in terms of the concept of value they try
to measure. Valuation consistent with SNA principles requires that the exchange value of transactions
between statistical units are measured. This implies that valuation excludes consumer surplus [11]
which is often called marginal valuation [12].
The 2008 SNA [3, A3.85] recommends that ‘water bodies should in principle be valued in a manner
parallel to the valuation of mineral resources but with an indication that more pragmatic alternatives
may have to be used such as estimates based on access fees.’ Moreover [3, para 13.51] states that ‘Non-
cultivated biological resources, water and other natural resources are included in the balance sheet to
the extent that they have been recognized as having economic value that is not included in the value of
the associated land. As observed prices are not likely to be available, they are usually valued by the
present value of the future returns expected from them.’ The method that is alluded to here is using the
resource rent of companies involved in the abstraction of water such as drinking water companies or

4
Brouwer et al. [10] distinguish between: market based transactions; derived demand functions; production and cost
functions; residual imputation; mathematical programming and optimization, choice modeling, hedonic price methods; travel
cost methods; contingent valuation; defensive expenditures; and, replacement/restoration cost/cost savings methods.
70 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81

farmers. This is an example of what is described as a residual imputation method in Brouwer et al. [10].
The relationship between the value of water resources and the valuation of land will be discussed in
Section 4.2.
On the other hand, the SEEA Central Framework observes that due to its public goods character,
water is often supplied at prices below costs of production and/or prices that are unrelated to the
actual volume supplied [4, para 5.489]. ‘Given this situation, the standard approaches to valuation of
environmental assets, and in particular the net present value approach ..do not work because the resource
rent that is derived following standard definitions is negative.’ Given these conflicting recommendations,
by using the resource rent method we hope to shed light on the conditions to its applicability.
In addition to the resource rent method, we will use a replacement cost approach, as this is a
method that is considered to result in a valuation consistent with accounting principles [12, p. 126].
There are other valuation methods that could be applied, for instance as suggested by the 2008 SNA
based on ‘access fees’. In the Netherlands, currently there are however no tradable water rights (as in
Australia, see Comisari and Vardon [17]) that could be used for such a valuation approach.

3.1. Resource rent method

As advocated in the SEEA Central Framework Section 5.4 [4] in the absence of market prices
environmental assets can be valued through the net present value of future resource rents. This method is
being applied for the valuation of mineral and energy resources by Statistics Netherlands [18] and recently
also for the valuation of wind energy [19]. It was applied by Comisari et al. [7] for the valuation of water
resources in Australia. Due to its consistency with SEEA principles we will also use it here as our starting
point. According to the SEEA Central Framework [4, p. 153], the resource rent can be estimated through
the operating surplus of companies involved in extraction of the resources in question, as follows:
Output (at basic prices i.e. excluding taxes on products, including subsidies on products)
 Intermediate consumption (cost of inputs).
 Compensation of employees (wages and pensions).
 Other taxes on production (plus subsidies on production).
¼ Gross operating surplus.
þ Specific taxes on extraction (minus specific subsidies on extraction).
 User cost of fixed capital.
¼ Resource rent.
The resource rent is obtained as a residual that measures the contribution of the environmental
asset to production.

3.2. Replacement cost method

The replacement cost method is recognized in the SEEA Experimental Ecosystem Accounting
[12, p. 126] as a feasible method for valuing ecosystem services as it excludes consumer surplus.
Precursors for using such an approach can already be found in the SNA itself, although it is clearly not
applied for valuing ecosystem services. For instance, national accounts often use a capital stock model
to estimate the value of fixed assets, based upon a depreciated (or ‘written down’) replacement cost
method, which values these assets as ‘the current acquisition price of an equivalent new asset less the
accumulated depreciation’ [3, para 13.23]. A second element concerns the idea of using “prices of
same or similar items” [3, para 3.123] in the absence of market prices, as is applied in the SNA for
instance in the valuation of subsistence agriculture.
The replacement cost method consists in estimating the value of certain flows based on the costs of
the next best alternative for replacing the service in question. Necessary conditions for its application
are that valuation is based on the least-cost alternative and that replacement of the service is to be
expected in case it would indeed be lost. The reason this approach may be feasible for valuing water
resources, is that there are few substitution possibilities for the use of water resources and demand of
water may be considered fairly inelastic.
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 71

3.3. Data sources

This article draws upon multiple data sources, focusing on the years 2005–2010. First of all in order
to calculate the resource rent two types of national accounts data are combined. Information from the
supply and use tables regarding the water supply industry and agriculture [21] in combination with
information on capital stocks for these activities [22]. The latter type of information is available from
the Dutch growth accounts – satellite accounts to the national accounts. The data used in this article
predate the revision of the Dutch national accounts.5
Secondly, information from the Dutch water accounts is used [14]. The water accounts themselves are
based upon multiple data sources as described by Graveland and Baas [16] such as the agricultural sample
survey (Farm Accountancy Data Network, FADN), information from the Dutch association of water
companies (Vewin), environmental reports by manufacturing companies, and additional types of water
statistics. Use of soil water needs to be estimated. Hereto, we have used data from a report commissioned
by Eurostat on the water balance [16] that quantified precipitation and evapotranspiration for different land
use categories in 2009. The assessment was based upon spatially and temporally explicit remote sensing
data and precipitation radar data that was obtained from WaterWatch (E-Leaf).
Third, for the replacement costs calculations numerous research reports, articles and data sources
are used [23–32]. A crucial source in this regard is the benchmark analysis of water supply agencies,
which includes information on the operational costs per m3 produced drinking water of various
suppliers that is routinely performed by Vewin [20]. In addition, several auxiliary data source are used
such as energy statistics and energy accounts.

4. Results and analysis

In this section we discuss the experimental values that we obtain when we attempt to value the physical
flows described in Table 1 by applying the above mentioned methods. This section is structured according
to the main water uses/users: the water supply sector (mainly surface and groundwater), agriculture
(predominantly soil water), and other industries (primarily use of surface and seawater for cooling).

4.1. Water supply sector

Table 2 provides a time series of economic data related to the water supply sector6 that is obtained
from the Dutch national accounts. We observe that the value of production has declined slightly over
time, while intermediate consumption has increased. As a result value added declined during the
period of investigation.
Production in the national accounts is recorded in basic prices, excluding any product based taxes
and/or subsidies and excluding value added tax. In the Netherlands, a range of water related taxes
applies (Table 3); see also Schuerhoff et al. [34]. According to Vewin [20, p. 26] ‘the first three are
considered as cost-price increasing taxes, that are covered via the drinking water rate, the last two
mentioned taxes . . . are consumer taxes which the water company pays to the tax authority on behalf of
consumers’.7 Tap water tax and value added tax are therefore not included in the Vewin figures on
turnover. In the Dutch national accounts, all water related taxes are considered as product based taxes
and therefore excluded from the value of production and therefore also from the resource rent.

5
EU Member states are required to comply with the 2010 European System of Accounts which is based upon the 2008 SNA.
The results of the revision were not yet available at the time this research was undertaken.
6
The ‘water supply sector’ refers to ISIC 36 (International Standard Industrial Classification of All Economic Activities
Rev. 4). In the Netherlands at the time of the research it is formed by 10 (public) water companies that supply drinking water
(and limited amounts of ‘industrial water’) to the industries and households. The sector has undergone rapid changes, as in
2000 there still existed more than 20 companies.
7
In Dutch, this would be called a ‘kassiersfunctie’.
72 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81

Table 2
Water supply sector: resource rent calculation.
Source: Author's calculations based on [21,22].

Item Unit 2005 2006 2007 2008 2009 2010

Production mln euro 1684 1629 1604 1606 1617 1647

Water sales mln euro 1579 1520 1491 1491 1503 1522
Intermediate consumption mln euro 707 716 726 764 770 768
Value added mln euro 977 913 878 842 847 879
Wages mln euro 336 338 343 354 368 386
Net taxes on production mln euro 22 24 23 24 32 34
Gross operating surplus mln euro 619 551 512 464 447 459
Water related operating surplus mln euro 580 514 476 431 415 424
Capital stock mln euro 8253 7885 8037 8138 8356 8305
Return to capital mln euro 370 405 488 513 325 283
Consumption of fixed capital mln euro 413 399 408 418 416 427
User cost of fixed capital mln euro 783 804 896 931 741 710
Resource rent mln euro  202  290  420  500  326  286

Table 3
Characterization of Dutch water related taxes.

Type Product based Part of turnover

Groundwater taxes Yes Yes

Provincial groundwater levies Yes Yes
Distribution and concession reimbursements Yes Yes
Tap water taxes Yes No
Value added tax Yes No

On average 93 percent of the production value of the water supply industry consists of water;
additional production is realized as a result of processing and other services. We therefore assume
that 93 percent of the gross operating surplus is due to the supply of water.
Information about the capital stock and consumption of fixed capital is available from the Dutch
growth accounts [22]. The rate of return to fixed capital consists of the internal reference rate to which
a small risk premium of 1.5 percent is added [18].8 The rate of return together with the consumption
of fixed capital provide an estimate of the user cost of fixed capital. After subtracting the user costs of
fixed capital from the water related gross operating surplus we obtain the resource rent. We find that
the resource rent is negative for all years, which implies that the value of water resources is zero.
To test the plausibility of these results, a comparison can be made with business accounts data i.e.
data that can be obtained directly from annual company reports. Small differences are to be expected
between the national accounts as data source and the business accounts (profit and loss accounts) of
the water supply agencies, due to:

 Differences in definitions between for example the sales concept of business accounting and the
production concept of the national accounts. These differences have to do with the treatment of
changes in inventories, treatment of taxes and subsidies, treatment of interest and financial
services.
 Differences that may arise due to the integration process of the national accounts in which for
instance supply of water is confronted with demand for water (by households and businesses).
 Strategy to focus in the Dutch national accounts on measuring correct growth rates of
macroeconomic indicators rather than focus on measuring correct levels at all times.

8
The internal reference rate indicates the interest rate charged between banks for borrowing money. Alternatively,
interbank lending rates such as Euribor could be used.
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 73

Table 4
Water supply sector: resource rent using business accounts (mln €).
Source: Author's calculations based on financial reports of Dutch water companies 2010 and [20].

Type Nr. 1 Nr. 2 Nr. 3 Nr. 4 Nr. 5 Nr. 6 Nr. 7 Nr. 8 Nr. 9 Nr. 10 Total
Ground Mixed Mixed Surface Ground Surface Ground Ground Surface Surface

Production 454 76 114 295 218 141 55 45 192 102 1691

Intermediate consumption 136 42 48 127 112 53 23 16 87 . .
Value added 317 35 66 168 107 88 32 29 105 . .
Wages 83 14 25 37 48 33 13 11 40 . .
Gross operating surplus 235 21 40 130 58 55 19 19 66 . .
Capital stock 1682 187 525 904 605 488 125 145 724 . .
Consumption fixed capital 86 13 25 65 20 35 7 7 40 . .
User costs of fixed capital 143 19 43 95 41 51 11 12 65 . .
Resource rent 91 2 3 35 17 4 7 7 1 . .

Data not available

 Different valuation methods for capital stocks such as (i) current replacement costs;
(ii) depreciated replacement costs, or (iii) net present value (see Comisari et al. [7]).

Table 4 provides information gathered from the financial reports of 9 out of ten Dutch water supply
companies that existed in 2010 (excluding Waternet9). According to Vewin [20] total water sales in
2008 for Waternet amounted to 102 million euro. Assuming that this figure has not changed much, we
derive that the total revenues in 2010 equaled 1691 million euro, which is only slightly higher than
the production value reported in the national accounts. We conclude therefore that differences
according to definitions and the integration process (items 1–3) are small.
A more significant difference is due to the valuation of the capital stock. According to the Dutch
national accounts shown in Table 2 the value is about 8.3 billion euros in 2010. This value is the
outcome of a capital stock model which is based upon a depreciated (or ‘written down’) replacement
cost method. According to Vewin the value of tangible fixed assets at the end of 2010 was 5.5 billion
euro [35, p. 36] which is 34 percent less than the national accounts estimate. The reason these
valuations are so different is due to differences between business accounts and national accounts
conventions. The Vewin figure is based on individual company statements some of which value assets
using historic costs.10 A lower capital stock value commands a lower depreciation value. Depreciation
according to the annual statements would be around 330 million euros in 2010. As a result the user
cost of capital would be only 460 million euros in 2010 and positive rents would be obtained for all
but one of the companies in 2010. However, in the absence of market based valuation (e.g. when water
infrastructure would be sold from one company to another), a depreciated replacement cost method
is clearly the preferred alternative in a national accounts context. Moreover, these positive rents are to
a large extent the result of very low interbank lending rates in 2010 compared to their long time
average, which drives down the user cost of capital. Therefore, we conclude that use of the resource
rent approach is problematic for valuing the extractive use of water by water companies.
As explained in Section 3, an alternative valuation approach consists in using a replacement cost
method. For instance, the value of the provisioning service of water could be based upon the second
best alternative for supplying water of a similar quality [40]. In order to apply the replacement cost
approach we need to estimate least cost alternatives for provisioning (fresh) water from various types
of water resources used by the water supply sector.

9
Waternet is a foundation and involved in both water supply and waste water activities.
10
For instance, the annual report of WMLsays (author's translation from Dutch): ‘the buildings and terrains, plants and
machinery, main and connecting pipes, water meters, other fixed assets and tangible fixed assets under construction are valued
at acquisition or production cost less accumulated depreciation’.
74 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81

Fig. 1. Operational costs of drinking water production for various water sources, 2010.
Source: author's calculations based on Vewin data [36, Fig. 55] and [27,29].

Valuation of the provisioning service of groundwater may be obtained by estimating what it would
cost if ground water resources would not be available at all. This would require that surface water
resources would have to be used, with additional operational costs due to necessity to clean the raw
water source, as the water quality of surface water, generally speaking, is lower than that of
groundwater. In doing so we assume that surface water is indeed available under comparable
conditions for abstraction and transport and not subject to depletion. Indeed, Table 4 demonstrates
that ground water companies tend to have higher resource rents than companies that predominantly
use surface water or ‘mixed water companies’. Based upon Vewin data a crude estimate of these costs
can be made, by comparing the operational costs of the production of drinking water per m3 of the
three companies with lowest costs with those of the three companies with highest costs [36, Fig. 55],
where we assume that the former three represent groundwater companies while the latter three
represent surface water companies.
The least cost alternative for using surface water for making drinking water would be to use
desalination. Another alternative would be to import water from neighboring countries such as
Germany or Belgium, but it can readily be seen that importing water is not a least cost alternative. For
instance, we obtain a rough indication by calculating the lowest price that can be estimated from our
international trade in goods statistics, which concerned a bulk import from a neighboring country,
resulting in a price of 2.7 euro/m3.11
The costs of desalination have been estimated based upon Bernat et al. [29]. The costs are to a large
extent driven by energy costs which are dependent on the type of desalination technique. Using
Spanish data about an existing desalination plant in the Mediterranean sea as proxy, we assume that the
energy requirement per m3 of tap water has improved from 5 kWh/m3 in 1998 towards 2.9 kWh in 2010
[29, Fig. 7]. The Netherlands has an energy cost structure that distinguishes households from industrial users
and has differentiated rates depending on use. For the Dutch energy costs per euro, we use a representative
electricity tariff for industrial users using more than 150,000 MWh [27]. Furthermore, we assume that
electricity costs are 43 percent of total costs for desalination when using reverse osmosis technologies [29, p.
341] to arrive at a cost of 0.52 euro cents per m3 of drinking water (see Fig. 1).
In order to estimate the value of the provisioning services, we take the additional costs (or the
saved costs) involved in replacing a water resource for its least cost alternative. The additional costs
becomes 0.13 euro cent for fresh surface water as alternative for groundwater and 0.26 euro cents for
seawater as alternative for fresh surface water. The efficiency of water companies in producing
drinking water from abstracted water is almost 90 percent in 2010. After multiplying with this time
specific efficiency factor we obtain the additional costs per m3 abstracted water. Combined with
information on ground- and surface water abstraction for ISIC 36 we then obtain an estimate of
provisioning service of 197 million euros of ground and surface water together in 2010 used by the

11
A more radical alternative would be to estimate what it would cost to import water resources, based upon for instance
large floating plastic bags: 〈http://www.volkskrant.nl/vk/nl/2844/Archief/archief/article/detail/442068/1996/07/03/Waterbe
drijf-wil-Noors-drinkwater-in-zak-invoeren.dhtml〉.
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 75

Table 5
Agriculture sector: resource rent calculation.
Source: Author's calculations based on [21,22].

Item Unit 2005 2006 2007 2008 2009 2010

Production mln euro 19297 20,958 21,799 22,428 20,833 22,770

Intermediate consumption mln euro 11,997 12,828 13,849 15,411 14,787 15,249
Value added mln euro 7300 8130 7950 7017 6046 7521
Wages mln euro 1753 1776 1873 1983 2024 2053
Net taxes on production mln euro 147  132  295  379  269  218
Gross operating surplus mln euro 5400 6486 6372 5413 4291 5686
Capital stock mln euro 34,852 35,875 37,143 38,691 40,923 42,564
Return to capital mln euro 1561 1844 2255 2438 1592 1451
Consumption of fixed capital mln euro 2635 2719 2797 2933 3015 3129
User costs of fixed capital mln euro 4196 4563 5052 5371 4607 4580
Resource rent mln euro 1204 1923 1320 42  316 1106

water supply industry. It should be mentioned that in addition to additional cleaning costs,
desalination would also require additional costs (both capital and operational) for distributing the
water from the desalination plants to where it is used. Such an estimate is however very difficult to
make and not attempted here. The calculations of additional costs provided here should therefore be
considered as conservative

4.2. Agriculture

The areal being irrigated at least once during the growing season ranges between 5 and 9 percent
of cultivated land [37]. In warmer and particularly dryer years like 2006, 2010 and 2011 a larger area is
irrigated, increasing the amounts of ground- and surface being abstracted. Irrigation is done for two
main reasons. First, in arable farming, irrigation water is given to a limited number of crops, just
during the start of the growing season in spring close to the moment of sowing or planting, to ensure
that the crops get a good start, enabling the seeds to germinate and the small plants to survive the
first stages of cultivation. Secondly, there is irrigation in larger volumes that is given to facilitate the
crop growth itself and reduce the water shortage or to let the crop survive under particular hot and
dry circumstances (i.e. relevant for grassland). Irrigation during the 2001–2011 period varied between
36 (2008) and 160 million m3 (2003) and shows an average of 66 million m3 in this period. This
compares to only 1.1 percent of the total evapotranspiration volume by agricultural crops. Even in the
driest year like 2003, at most 2.3 percent of the water needed by the crops was provided via irrigation.
This demonstrates that agricultural crops in the Netherlands are grown predominantly under rain
fed conditions, as the water provided from the agricultural land (and managed soil) originally and
predominantly stem from precipitation. The uptake of water represents one of the ecosystem services
provided by the agricultural land. Other services consist for instance in the provisioning of nutrients.
Table 5 similar to Table 2 derives the resource rent for the agriculture sector (defined here as ISIC
01). We find that the rent is positive in all years except 2009, but volatile. As with Table 2 the capital
stock items in the table solely represent the so-called fixed assets such as buildings, animal houses,
sheds, milking parlors, storage facilities, machinery etc. The capital embodied in agricultural land is
not included in this capital stock representation. Land is considered a non-produced asset in the SNA
(apart from land improvement which is considered produced capital) and rent payments are
considered as income transfers which are not part of production.
A difficulty with the resource rent calculation is that the residual, the determined resource rent,
measures not only the return to natural capital used in production, but may also include mixed
income i.e. compensation of unpaid labor by the farmer or other members of the household. The
national accounts [38, Table D 14.1] provide information about mixed income in the broader
agriculture, forestry and fishing sector (ISIC Section A) which in 2010 equals about 2.7 billion euro and
as a result would imply negative resource rents for the agriculture sector.
76 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81

Table 6
Estimation of soil water provisioning service through rent payments for agricultural land.
Source: Author's calculations based on [24,32].

Item Unit 2005 2006 2007 2008 2009 2010

Agricultural area km2 22,758 22,696 22,643 22,590 22,534 22,478

Rent/ha euro 440 451 454 528 516 506
Rent payments mln euro 1001 1024 1028 1193 1163 1137
Provisioning soil water mln euro 330 338 339 394 384 375

In order to investigate this issue further, we employ in Table 6 an alternative valuation strategy,
which consists in analyzing the rent payments by farmers for the use of agricultural land.
In the Netherlands in 2013 around 57 percent of agricultural land is owned by the farmer, the
remainder is leased (or similarly ‘given in use’ although with different institutional/organizational
arrangements). When we combine information about total agricultural land use with the prices for
the land that is rented/leased we obtain an estimate of total rent/revenue of lease from the
agricultural area under the assumption that all the land would have been leased.12
The rent payments for agricultural land use will reflect a range of characteristics such as soil
quality, but also proximity to population centers, nature reserves, infrastructure, etc. There is indeed
strong variation in land rent per ha in the Netherlands over time but also between regions, plots, and
individual parcels [32,39].
The type of rent also has an impact. Nature rent i.e. rent of land (mostly grassland) by farmers in
areas destined for nature conservation purposes show the lowest rents (e.g. in 2011: 32 euro/ha),
while one year rent, crop rent (normally two years at the maximum), and long lease (with additional
rights to the lessee; in Dutch ‘erfpacht’) show higher values (in 2011 respectively: 690, 1668 and
577 euro/ha). The average price that we use is 506 euro/ha in 2010 which lies in-between the
common short and long term rates of regular rent..
In order to separate the provisioning service of soil water in the rent payments by farmers, we use
the fact that according to Zwart and Bastiaansen [33] about 33 percent of the variation in crop yield of
wheat can be explained by evapotranspiration. In the absence of more accurate estimates, we apply
this factor to the total rent payment to estimate the value of the provisioning service of soil water. This
results in a value of 375 million euros in 2010.

4.3. Water use by other industries

In this Section we discuss the valuation of water abstraction by other industries. As we already saw
when discussing Table 1, by far the largest usage of water in the Dutch economy is the abstraction of
large amounts of both fresh and non-fresh surface water and to a lesser extent groundwater for
cooling purposes in manufacturing and electricity production. These large volumes are used in once-
through type of cooling systems sometimes referred to as in-stream cooling. We will value this water
use based on the replacement cost method.
The total heat discharged to the environment is around 9 percent of Dutch total net domestic
energy use in recent years, of which 7 percent is discharged via cooling water [25,31]. In fact, the
energy discharged as heat via cooling water to the environment (in PJ) is more or less similar to the
electricity produced (in PJ).
The summer season of 2003 was particularly hot and dry in the Netherlands. Although power
plants were not running at their maximum, normally peak demand for electricity occurs in winter

12
Subtracting this value from the resource rent could be interpreted as providing an indirect estimate of mixed income,
which would be rather volatile, and negative in 2008 and 2009. This illustrates what is commonly understood that farmers or
entrepreneurs and often their family members in primary agriculture sector in income terms are not compensated on an equal
basis compared to other sectors.
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 77

season, the power demand from the Dutch economy in the summer of 2003 could hardly be satisfied
due to cooling difficulties. The water from the rivers at several locations reached 27 1C already and
therefore could only absorb little additional heat from the cooling cycle of power plants before
reaching the maximum discharge temperature of 30 1C. The response chosen at that time was to allow
the discharge of slightly warmer cooling water beyond the legal limit of 30 1C, namely 311. As a
consequence of this episode, the Minister of Economic Affairs argued for the construction of cooling
towers in the near future. This is the alternative technology that we therefore will value.
There are various cooling tower techniques possible, such as natural draft or mechanical draft
towers. To obtain an estimate of the costs we have used information about a recently fan assisted
natural draught cooling tower that was built in Karlsruhe (2009–2011) at a cost of 19 million euro
(Alpine 2012). The reason we chose this benchmark is because it is close to the Dutch situation.
The plant for which the cooling tower was constructed has a capacity of more than 900 MW, and
uses in stream cooling (‘once through’) from the Rhine, but when temperature becomes too high,
switches to using the cooling tower.13 In order to obtain an estimate we make the following
assumptions. First of all by dividing the total installed Dutch capacity of electricity production by coal
and gas fired plants in 2010 [26] by the capacity of this plant we first of all estimate the number of
plants required in the Netherlands, which we correct for the percentage of plants that is located at the
seashore and already cooled using sea water (using information from Table 1).
Second, we estimate the annual capital costs (user costs of capital) using the same rate of return
rates as in Tables 2 and 5 applied to the annual investment required, which we assume constant
during the period of investigation by lack of more data.14 Furthermore, we assume a service life of the
cooling tower of 20 years and we assume that the operating costs of cooling towers are 25 percent of
their construction costs [30, p. 40 and 70].
Hence we obtain an estimate of the total replacement cost per year. When we divide this number
by the actual annual cooling water use from Table 1, we obtain a provisional estimate of 1.4 euro cent
per m3 of the provisioning service of cooling water in 2010. Herewith we obtain a value of 109 million
euros in 2010 for the provisioning service of surface water used as cooling water. Obviously, this
estimate is subject to large uncertainty.
Non-cooling water used by other industries is valued with the replacement costs for groundwater
and surface water that we obtained in Section 4.1.

4.4. Asset values

Now that we have valued all water abstraction by economic activities, we can turn these annual
values into experimental stock values by calculating the net present value. Assuming sustainable use
by the government and using a discount rate of 4 percent – this is the same rate as used in the
valuation mineral and energy resources by Statistics Netherlands – we would obtain a value of water
resources ranging between 25 and 28 billion euro over the 2005–2010 period (Fig. 2). Surface water is
the largest contributor (on average about 57 percent).
The value of water resources corresponds in 2005 with 19 percent of the value of natural capital (defined
here as consisting of the sum value of agricultural land and mineral and energy reserves) that is currently
included in the Dutch balance sheet [23]. This share decreases towards 10 percent in 2010 primarily due to
strong increases in the value of the mineral and energy reserves and agricultural land during this period.
How should this experimental asset value be recorded in the national accounts? The 2008 SNA [3]
recognizes that water resources (a form of non-produced assets) may be drawn into the asset boundary
when ‘used for extraction to the extent that their scarcity leads to the enforcement of ownership or use rights,
market valuation and some measure of economic control.’ [3, para 10.184]. [3, para 12.22] states that:
‘[s]imilarly, drawing water from a natural spring does not bring an aquifer into the asset boundary of the SNA,
but a significant diversion of groundwater does. A move to charge for regular extraction from a body of surface

13
For more information regarding the plant see: 〈http://www.alstom.com/Global/Power/Resources/Documents/Brochures/
RDK8-germany-steam-power-plant-editorial.pdf〉.
14
By doing so we effectively assume that the effect of technological progress and inflation balance.
78 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81

Fig. 2. Experimental stock values of Dutch water resources in current prices (2005–2010).

water may also bring a water resource into the balance sheet.’ In the Netherlands, groundwater use is
restricted and can be considered a scarce resource especially in summer when desiccation has become an
issue due to lowering of the groundwater table [46]. Also, companies involved in abstraction of ground
water have to pay groundwater levies. So it is clear that according to the SNA groundwater resources in
the Netherlands are considered assets and should be valued.
Whether surface water resources fall within the SNA asset boundary is debatable: no levies are
paid for extraction although use of surface water is highly regulated especially in summer when there
are competing uses (e.g. minimum river flow for navigation; irrigation etc.). The SEEA Central
Framework has an extended asset boundary compared to the SNA, but follows the same valuation
principles as the SNA. It is likely that the stock value of soil water that we estimated here is already
captured in existing balance sheet items such as the value of agricultural land, and therefore would
need to be separately identified (as an of which item). This is arguably not the case for the value of
groundwater that we obtained, which may be added to the balance sheet.

5. Discussion and conclusions

We have discussed a number of approaches towards valuing the provisioning services of various
water resources used by the Dutch economy. Our main approach has been to apply sector-specific
valuation methods using a variety of approaches.
The resource rent method when applied to the water supply sector resulted in a negative resource
rent for the time period under investigation. It is not uncommon to find negative or zero resource rents
for example in the valuation of fish stocks (e.g. Harkness and Aki [41]). Our findings resemble the
conclusion reached by Comisari et al. [7], who found for Australia that ‘for the majority of water suppliers,
no positive resource rent valuations were generated. . . . [W]ater prices . . . barely allow a return on assets and
never deliver an implied value on their respective water stocks’. Furthermore, the conclusion reached in
Comisari et al. [7] was that: ‘businesses in the Australian water supply industry usually generate minimal or
no profit . . . even though the businesses would view themselves as ‘for-profit’ entities. The economic benefits
of water infrastructure assets do not appear to reside in their ability to generate a commercial return. Instead
much of the benefit appears to reside with the multitude of Australian businesses and households who
receive a clean and reliable water supply at a price kept artificially low by a dedicated regulatory regime’.
The Australian institutional setting is comparable to the Netherlands. Most of the Dutch water supply
companies are so-called ‘Naamloze Vennootschappen’ with municipalities and provinces as share-
holders. Consumers are not free to choose a particular water supply company, as the drinking water
distribution systems are closed. In general, the shareholders decide upon the tariffs where cost-recovery
is the guiding principle (as stipulated by the EU Water Framework Directive).15 In short, the situation in
Australia is similar to the Netherlands in the sense that there also is a strongly regulated environment.
These observations seem to call in question the applicability of the resource rent valuation method

15
According to the Dutch Waterleidingwet (Dutch Water Act), water companies should ask for prices that allows for at
minimum cost-recovery [42, p. 19].
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 79

under circumstances of strong governmental regulation in the area of public goods, a conclusion that
was also reached in SEEA Central Framework.
We have therefore used an alternative approach for valuing the provisioning service of water for
the drinking water sector. The replacement cost method results in estimates that are consistent with
national accounts valuation principles. However, the rough estimates that we have obtained especially
concerning desalination costs and cooling tower costs, may need to be refined based upon future
research. The replacement cost method does yield values that have a straightforward interpretation.
For instance, when the quality of groundwater resources would deteriorate, the wedge in difference of
cleaning costs compared to surface water would diminish its stock value. Likewise, when due to
technological improvements, the costs of constructing and/or operating cooling towers would
decrease, the value of the provisioning service of surface water would diminish. That said, a crucial
assumption underlying the replacement cost approach is that a reduction or loss of provisioning
service of water would indeed be replaced. The Australian water accounts [43] however suggest that
urban and agricultural water use is in fact elastic, showing responses to changes in market, regulatory
and climatic circumstances. This may be a bit different for the Dutch situation, where we observe that
in dryer years the irrigated area and frequency of irrigation is substantial higher.
Finally, there may be other alternatives for replacing the use of cooling water from surface water by
cooling towers such as generating electricity from alternative sources (solar, waste, etc.) or through
increased imports of electricity, that we did not assess in this article.
In agriculture, applying the resource rent method also proved difficult as it measures not only a
return to land used, but also contains a component which captures the compensation for mixed
income. Therefore, we used actual rent payments for land leases to obtain an estimate for the return
to land in combination with an estimate from the literature to separate the soil water provisioning
component.
A superior method may consist in using a variant of the production function approach to
investigate the relation between agricultural output and soil water use. Hereto, micro data at farm
level with various characteristics including yield per crop type per year could be combined with
estimates of evapotranspiration at grid level that were obtained through remote sensing and
modeling as explained in Section 2. Such an approach is possible in theory, but would require linking
various spatially explicit data sets, which is something we would like to investigate in future research.
Alternatively, it would also be interesting to apply a hedonic price method to research the relation
between agricultural land prices (as transacted in the market) and a number of agronomical features
of land using micro data. For instance, Kuhlman et al. [44], show the strong influence of the type of
economic activity on the market for land. On average agricultural land shows substantial lower prices
than other types of land use.
Given the numerous assumptions that were necessary in the absence of more detailed information,
and considering the objections formulated above regarding the appropriateness of some of the
applied methods, it is important to stress that the values and results obtained here should be
considered experimental (hence the title) and in some instances due to the absence of a quantification
of changes in distribution costs as conservative. Especially the experimental stock values that we have
obtained for the various types of water resources distinguished within SEEA should be considered
with caution, as they are greatly dependent upon the chosen discount rate and the assumption of
sustainable use. What we have tried to demonstrate in the article is that a valuation of water resources
using accounting principles is possible. More research in this area is clearly necessary especially
concerning improving the estimate obtained using the replacement cost approach. The large number
of required data sources including the existence of national accounts and water accounts in sufficient
detail, may limit the feasibility of the approach developed here to countries less abundant in data.
In this article we have restricted ourselves to valuing the extractive use of water, and hence obtain
only a partial value of Dutch ground and surface water resources. There may be several non-
consumptive uses such as use of rivers for navigation, or recreation that when valued would increase
the value of water resources (or aquatic ecosystems). The recognition of these values necessitates an
extension of the production boundary of the SNA as proposed within the SEEA Experimental
Ecosystem Accounting framework. The valuation of such water related ecosystem services could be
undertaken as part of the development of comprehensive ecosystem accounts.
80 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81

Acknowledgments

This research was undertaken in the context of the EC funded FP7 project CREEA (Compiling and
Refining Environmental Economic Accounts), specifically WP3 – Water Accounts. This paper does not
represent any official position or endorsement of the funding organization. The authors would like to
acknowledge helpful comments and suggestions provided by Maarten van Rossum, Sjoerd Schenau,
Carl Obst and Peter Comisari, colleagues within the CREEA project and three anonymous referees.

References

[1] United Nations, 1992. Agenda 21. 〈http://www.un.org/esa/dsd/agenda21/res_agenda21_08.shtml〉.

[2] European Commission, Directive 2000/60/EC of the European Parliament and of the Council establishing a framework for
the Community action in the field of water policy (EU Water Framework Directive), Official Journal (OJ L 327), 2000.
[3] UN (United Nations), EC (European Commission), IMF (International Monetary Fund), OECD (Organisation for Economic
Co-operation and Development) and World Bank 2009, System of National Accounts 2008, New York.
[4] UN (United Nations), EC (European Commission), FAO (Food and Agriculture Organisation), IMF (International Monetary
Fund), OECD (Organisation for economic Co-operation and Development) and World Bank, 2014. System of Environmental-
Economic Accounting 2012: Central Framework, New York.
[5] UN (United Nations), System of Environmental-Economic Accounting for Water (SEEA-W), ST/ESA/STAT/SER.F/100,
Department of Economic and Social Affairs, Statistics Division, UN, New York, 2012.
[6] INEGI 2014, 〈http://www.inegi.org.mx/est/contenidos/proyectos/cn/ee/〉 (accessed January 2014).
[7] P. Comisari, L. Feng, B. Freeman, Valuation of water resources and water infrastructure assets. Paper presented to the 17th
London Group Meeting, Stockholm Sweden. Available at: 〈http://unstats.un.org/unsd/envaccounting/londongroup/meet
ing17/LG17_12.pdf〉, 2011.
[8] G.-M. Lange, R. Hassan, The Economics of Water Management in Southern Africa, Edward Elgar, UK, 2006.
[9] L.M. Brander, R.J.G. Florax, J.E. Vermaat, The empirics of wetland valuation: a comprehensive summary and a meta-analysis
of the literature, Environ. Resour. Econ. 33 (2006) 223–250.
[10] R. Brouwer, D. Barton, I. Bateman, L. Brander, S. Georgiou, J. Martín-Ortega, S. Navrud, M. Pulido-Velazquez, M. Schaafsma,
A. Wagtendonk, Economic Valuation of Environmental and Resource Costs and Benefits of Water Uses and Services in the
Water Framework Directive: Technical Guidelines for Practitioners, 2009.
[11] C. Obst, B. Edens, L. Hein, Ecosystem services: accounting standards, Science, vol 342, 25 October 2013.
[12] EC (European Commission), OECD (Organisation for Economic Co-operation and Development), UN (United Nations),
World Bank, System of Environmental-Economic Accounting 2012 Experimental Ecosystem Accounting. White cover
publication, pre-edited text subject to official editing, 2013.
[13] M. Rossum, I. van Geloof, S. Schenau, NAMWA 2010 Water in de nationale rekeningen (2007), CBS Den Haag, 2010.
[14] Statistics Netherlands, . Environmental accounts; water use and abstraction, The Hague/Heerlen, the Netherlands, 〈http://
statline.cbs.nl〉, 2014 (accessed January 2014).
[15] OECD/Eurostat, OECD/Eurostat – Joint Questionnaire Inland Waters. Data available at: 〈http://epp.eurostat.ec.europa.eu/
portal/page/portal/statistics/search_database〉, 2012 (accessed July 2013).
[16] C. Graveland, K. Baas, Improvement of the national water balance; water stocks; feasibility of water balances per river
basin. Final Report on Eurostat Water Statistics Grant: Grant Agreement NO. 50303.2010.001–2010.564, Heerlen/The
Hague, 2012.
[17] P. Comisari, M. Vardon, Valuation and treatment of water resource stocks. Paper presented to the 19th Meeting of the
London Group on Environmental Accounting, London. Available at: 〈http://unstats.un.org/unsd/envaccounting/lon
dongroup/meeting19/LG19_8_3.pdf〉, 12–14 November 2013.
[18] E. Veldhuizen, C. Graveland, D. van der Bergen, S.J. Schenau, Valuation of oil and gas reserves in the Netherlands 1990–
2005, Statistics Netherlands, 2009.
[19] Statistics Netherlands, Environmental Accounts of the Netherlands 2010, Heerlen/The Hague, 2011.
[20] Vewin, Dutch Drinking Water Statistics 2008, 2010.
[21] Statistics Netherlands, Dutch national accounts (input-output tables and supply use tables), The Hague/Heerlen, the
Netherlands. 〈http://www.cbs.nl/nl-NL/menu/themas/macro-economie/cijfers/incidenteel/maatwerk/2012-i-o-cm.htm〉,
2014b (accessed January 2014).
[22] Statistics Netherlands, Table ‘Groeirekeningen; kapitaalgoederenvoorraad’, The Hague/Heerlen, the Netherlands. 〈http://
statline.cbs.nl〉, 2014c (accessed January 2014).
[23] Statistics Netherlands, Table ‘Inkomens- en vermogensrekeningen; niet-financiële balansen 1996-2012’, The Hague/
Heerlen, the Netherlands. 〈http://statline.cbs.nl〉, 2014 (accessed January 2014).
[24] Statistics Netherlands, Table ‘Bodemgebruik’, The Hague/Heerlen, the Netherlands. 〈http://statline.cbs.nl〉, 2014 (accessed
January 2014).
[25] Statistics Netherlands, Table ‘Environmental accounts; energy use by companies and households’, The Hague/Heerlen, the
Netherlands. 〈http://statline.cbs.nl〉, 2014 (accessed January 2014).
[26] Statistics Netherlands, Table ‘Elektriciteit; productie en productiemiddelen', The Hague/Heerlen, the Netherlands. 〈http://
statline.cbs.nl〉, 2014 (accessed January 2014).
[27] Statistics Netherlands, Table ‘Aardgas en elektriciteit, gemiddelde prijzen van eindverbruikers’, The Hague/Heerlen, the
Netherlands. 〈http://statline.cbs.nl〉, 2014 (accessed January 2014).
[28] Alpine, Cooling tower construction. Available at: 〈http://www.alpine.at/wp-content/uploads/ALP_Folder_210x297_Kuehl
turmbau_EN_4c_20120130_72dpi.pdf〉, 2012 (accessed Jan 2014).
[29] X. Bernat, G. Oriol, R. Guiu, J. Tobella, C. Campos, The Economics of Desalination for Various Uses, 2009.
 B. Edens, C. Graveland / Water Resources and Economics 7 (2014) 66–81 81

[30] M. Choi, Leon R. Glicksman, Computer Optimization of the MIT advanced Wet/Dry Cooling Tower Concept for Power
Plants. Energy Laboratory Report No. MIT-EL 79-035 Heat Transfer Laboratory Report No. 87162-102, 1979.
[31] Statistics Netherlands, PBL, Wageningen UR, Koelwaterverbruik en warmtelozing door elektriciteitscentrales, 1981-2011
(indicator 0123, versie 09, 28 januari 2014). CBS, Den Haag; Planbureau voor de Leefomgeving, Den Haag/Bilthoven en
Wageningen UR, Wageningen. 〈http://www.compendiumvoordeleefomgeving.nl〉, 2014.
[32] LEI – Binternet, Rent area and prices, average of farms with rent, in: Farm results and income on agriculture and
horticulture holdings. 〈http://www3.lei.wur.nl/BIN_ASP/show.exe〉, 2014.
[33] S.J. Zwart, W.G.M. Bastiaansen, Review of measured crop water productivity values for irrigated wheat, rice, cotton and
maize, Agricult. Water Manag. 69 (2) (2004) 115–133.
[34] M. Schuerhoff, H.P. Weikard, D. Zetland, The life and death of the Dutch groundwater tax, Water Policy 15 (6) (2012)
1064–1077.
[35] Vewin, 2012. Dutch Drinking Water Statistics 2012.
[36] Vewin, Reflections on Performance 2012, Benchmarking in the Dutch drinking water industry, 2013.
[37] Statistics Netherlands, Environmental accounts of the Netherlands 2012, Heerlen/The Hague, 2013.
[38] Statistics Netherlands, National accounts of the Netherlands 2012, Heerlen/The Hague, 2013.
[39] J. Luijt, R. W. van de Meer, M. J. Voskuilen, et al., Hoogst toelaatbare pachtprijzen per pachtprijsgebied in 2013. Gegevens
over de periode 2007-2011. LEI-nota 13-044. Projectcode 2272000314. LEI Wageningen UR, Den Haag, 2013.
[40] L. Hein, Economic benefits generated by protected areas: the case of the Hoge Veluwe forest, the Netherlands, Ecol. Soc.
16 (2) (2011) 13. ([online] URL: 〈http://www.ecologyandsociety.org/vol16/iss2/art13/〉).
[41] J. Harkness, L. Aki, A numerical comparison of fish quota values and standard resource rent calculations using New
Zealand's commercial fish resource, Paper presented to the London Group on Environmental Accounting Brussels,
Statistics New Zealand, 2008.
[42] Ministerie van Verkeer en Waterstaat, Kostenterugwinning van waterdiensten in NL, 2005.
[43] Australian Bureau of Statistics, Water Account, Australia, 2011–12, cat. No. 4610.0, 2013.
[44] J.W. Kuhlman, J. Luijt, J.P.M. van Dijk, A.D. Schouten, M.J. Voskuilen, Grondprijskaarten 1998–2008, Wettelijke
Onderzoekstaken Natuur & Milieu (WOT Natuur & Milieu) Werkdocument 185, Wageningen UR Digital Library, 2010.
[45] M. Falkenmark, J. Rockstrom, The new blue and green water paradigm: Breaking new ground for water resources planning
and management, J. Water Resourc. Plann. Manag. 132 (3) (2006) 129–132.
[46] Statistics Netherlands, PBL, Wageningen UR, 2008. Oorzaken en effecten van verdroging (indicator 0278, versie 03, 4 april
2008). 〈http://www.compendiumvoordeleefomgeving.nl〉. CBS, Den Haag; Planbureau voor de Leefomgeving, Den Haag/
Bilthoven and Wageningen UR, Wageningen.