A pilot index for environmental policy in The Netherlands

A pilot index for environmental policy in The Netherlands

ELSEVIER Energy Policy,Vol. 26, No. 2, pp. 95-101, 1998 © 1998 Published by Elsevier Science Ltd. All rights reserved Printed in Great Britain 0301-4...

665KB Sizes 7 Downloads 63 Views

ELSEVIER

Energy Policy,Vol. 26, No. 2, pp. 95-101, 1998 © 1998 Published by Elsevier Science Ltd. All rights reserved Printed in Great Britain 0301-4215/98 $19.00 + 0.00

Pli:S0301-4215(97)00010-4

A pilot index for environmental policy in The Netherlands F. A. G. den Butter* and J. A. C. van der Eyden Free University Amsterdam, Faculty of Economics, De Boelelaan 1105, 1081 HV Amsterdam, The Netherlands

A pilot environmental index for The Netherlands shows a downward trend of environmental pressure in the last decade. The index measures environmental pressure from the policy perspective and the observed downward trend does not imply that environmental quality has increased. The aggregation weights which are used to combine various theme indicators of environmental policy to one overall index are derived from public opinion polls on the concern for environmental problems. Following Hope c.s., who constructed a similar pilot environmental index for the UK, we used the analogy of compilation of quality of life indices. © 1998 Published by Elsevier Science Ltd. Keywords. Environmental index; Environmental pressure; Quality of life index; Environmental concern

Introduction There is growing awareness that the quality of the environment is paramount to the quality of our life. Modern economics considers questions regarding quality of life part of welfare economics. That is why environmental economics and welfare economics have become more and more integrated. A very practical offspring of welfare economics is the construction of one, or a small number of welfare indicators. Time series of these indicators show the development of welfare in the course of time. These welfare indicators are also used for comparison of levels of welfare in various countries. This paper follows the tradition of calculating welfare indicators by the construction of a pilot index for environmental policy in The Netherlands. We imitate the methodology used by Hope et al (1992) for the United Kingdom on data for environmental theme-indicators collected by Adriaanse (1993) for The Netherlands. The major problem when we are to combine a set of environmental indicators to an index is to determine the aggregation method and to select the set of weights used in the aggregation. In this respect we follow the procedure by Hope c.s. and use results from the public opinion poll of the European Omnibus Survey to assign weights to the components of the index. This assignment problem consists

*Author to whom correspondence should be addressed. The author is also affiliated to the Tinbergen Institute.

of two parts. Firstly, the environmental themes from the opinion poll are to be translated to the themes of the indicators for which Adriaanse has collected data for The Netherlands. Secondly, a selection has to be made from the different sets of results from the opinion poll and from the various waves of the European Omnibus Survey. We perform a sensitivity analysis in order to investigate how both choices affect the time pattern of the overall index. The rest of this paper is organized as follows. "Welfare theory and measures of environmental quality" summarizes how the environment is embedded in economic welfare analysis. It also contains a discussion on the role of indicators of environmental quality in this analysis. Moreover, this section pays attention to the various ways of measurement and taxonomy of environmental quality indicators. "Components of an environmental index for The Netherlands" provides further information about the constituent parts of our construction method of the environmental index for The Netherlands. It gives a short introduction to the methodology of Hope c.s., to environmental theme-indicators of Adriaanse and to the European Omnibus Survey. "Constructing the pilot environmental index for The Netherlands" describes the construction methods of our overall index and shortly discusses its results. In this section we also provide arguments for our specific selection of the aggregation weights in constructing the basic version of the index. These assumptions are subjected to a sensitivity analysis in "Sensitivity analysis". Finally, some conclusions are drawn.

96 A pilot index for environmental policy in The Netherlands: F A G den Butter and J A C van der Eyden

Welfare theory and measures of environmental quality Economic welfare theory assumes that rational economic agents maximize their personal or collective welfare over some finite (or even infinite) horizon. In this context, welfare or well-being is very broadly defined. In the context of this paper a good quality of the environment most certainly is part of this broad definition of welfare. It means that in the economic decision making process consequences of these decisions for environmental quality are weighted against the consequences for other aspects of material or non-material welfare (or wealth). In this way, the welfare function describes a trade-off between environmental quality and, eg consumption expenditures and/or leisure. These trade-offs in the welfare function relate to relative preferences for the various aspects of welfare. Obviously, in order to make welfare analysis operational from this perspective, some kind of index number should be constructed for the concept of environmental quality. That is why the construction, in one way or another, of a measure of environmental quality is important for applied welfare analysis. For that reason, recently much attention has been paid to the relationship between environmental quality and economic growth. The fact that a welfare function describes a trade-off between both of these targets of economic policy, does not necessarily imply that environmental quality can only be enhanced by a relative reduction of economic growth. Recently, Grossman and Krueger (1995) investigated the relationship between economic growth and environmental quality empirically for 42 countries. They examined the reduced-form relationships between a per capita income and four types of environmental indicators, ie urban air pollution, the state of the oxygen regime in river basins, faecal contamination of river basins, and contamination of river basins by heavy metals. The authors found no evidence that environmental quality deteriorates steadily with economic growth. Rather, from most indicators, economic growth brings an initial phase of deterioration followed by a subsequent phase of improvement. The turning points for the different types of pollution indicators vary, but in most cases they come before a country reaches a per capita income of US $ 8000. This study by Grossman and Krueger shows that the relationship between environmental quality and income (economic growth) is very complex and may have the shape of a reversed U-curve (see also World Bank, 1992). From the point of view of welfare analysis it is, therefore, desirable to avail of a separate index of environmental quality: environmental quality, or the abatement made in order to enhance environmental quality to sustainability levels, should not be linked directly to the measurement of national income, as protagonists of an environmental correction of GNP advocate. For that reason we consider the construction of a 'green' or eco-GNP not very useful in welfare analysis, and concentrate on the construction of a non-monetarized index of environmental quality (see also den Butter and Verbruggen, 1994). In doing this we acknowledge that a number of

such indices has already been designed and have been given flowery names, such as Mirror of Cleanliness or Ecological Dow Jones Index. Such an environmental index should provide a general assessment of all aspects of environmental quality which are important to individual or collective well-being. However, the environment is so diverse in its various dimensions, both in time and space, that some regard it as utterly impossible to construct such an overall index. This is of course true in case a 'perfect' index is wanted. But one can also start to dig the tunnel from the other end, and realize that imperfect information on environmental quality is of more help to policy analysis than no quantitative information at all. From that point of view we should try to collect as much information as possible on environmental quality and combine this information into an overall index or a small number of indicators, which give at least some impression of the evolution of environmental quality in the course of time. This is the very modest goal of our study. Moreover, such a pilot index can teach us what further data collection is essential for the construction of a more comprehensive and sophisticated index for environmental quality. The many dimensions of environmental quality have resulted in different types of indicators for the state of the environment. We distinguish between effect-indicators, pressure indicators and sustainability indicators. The effect indicators come closest to an actual description of environmental quality. The pressure indicators measure the direct consequences of the economic process for the environment. Environmental pressure relates to the quantities of emissions, extractive environmental services, waste disposal, etc. Rather long lags may be involved in the spill-overs from environmental pressure to environmental quality, so that an improvement in terms of environmental pressure does not necessarily imply that environmental quality has also improved. Sustainability indicators relate the values of effect indicators and pressure indicators to fixed reference values, which are regarded as their sustainability levels. For that reason, sustainability indicators have a normative connotation. The environmental theme-indicators of Adriaanse, which are used in this study as component parts of the overall environmental index, have the character of pressure indicators. In fact, these indicators are connected with the goals of environmental policy, so that these indicators measure, in a way, the impact of environmental policy. Our overall index should, therefore, be regarded as a pilot index for environmental policy, and does, by no means, yield a direct measure for environmental quality. Economists often try to cast an abstract notion in one number or index. Examples are gross national product, the state of the cycle and the level of welfare. It seems, therefore, worthwhile to borrow from the experience of constructing cyclical indicators or welfare indicators for the design of an environmental index. Exploiting the analogy with the cyclical indicator in den Butter (1992) did not prove very successful, as some component parts of the environmental index in that study showed opposite movements over prolonged periods

A pilot index for environmentalpolicy in The Netherlands: F A G den Butter and J A C van der Eyden

of time. The construction method of cyclical indicators does not allow for such opposite movements. Therefore, in this study we pursue the analogy with welfare indicators or quality of life indicators in constructing our environmental index. Ott (1978) already notes that environmental indices can be constructed in a similar way as quality of life indices. The first indicators for the quality of life were developed in the United States in the 1960s. It is a much broader concept than the notion of, eg air quality or water quality. According to Ott, it is, in general, supposed to reflect all aspects of a person's sense of well-being; ie it includes all factors which contribute to human satisfaction or dissatisfaction, factors which determine a person's 'happiness' or 'unhappiness'. In that sense, the concept is much broader than per capita national income as a measure of welfare, of which the history of measurement goes back to the 17th century. Various examples of quality of life indicators have been proposed and calculated (see eg Ram, 1982; Slottje, 1991; Diener, 1995). Similar indices are the basic needs index (see eg Hicks and Streeten, 1979), and life product index (Lind, 1993) or the human development index, which is published in the Human Development Report by the United Nations Development Program (1990) in a vast comparative study of the quality of life in all UN member states.

Components of an environmental index for The Netherlands In our methodology we follow Hope et al (1992) who constructed a pilot environmental index for the United Kingdom, covering the period 1980 to 1988. This UK index consists of nine environmental indicators which are selected on the basis of two criteria. These criteria are data availability on a monthly basis and environmental problem priority. Hope c.s. take the latter problem as a selection criterium because they have used a public opinion poll to see what general categories are important to people. The nine constituent indicators of the pilot index by Hope c.s., and their units of measurement, are listed in Table 1. In our environmental index for The Netherlands we use the data collected by Adriaanse which are available on an annual basis for the period 1980-1991. In the selection of his indicators Adriaanse has very carefully tried to cover the main themes of Dutch environmental policy. The seven theme

Table 1

Components of the pilot index of Hope

Unit

NO~,-emissions SO2 urban concentrations Low level ozone concentrations CO2-emissions Oil spills requiring clean-up % of length of river of poor or bad quality Resident population Fertilizer deliveries to agricultural use New dwellings started

1000 tonnes NO2-equivalents micrograms per ms average monthly 99th percentile million tonnes of carbon number percentage

thousands

indicators, reported in Table 2, are all compound indicators. For instance, the indicator for the theme 'change of climate' is calculated as the weighted summation of the Dutch annual discharges of CO> CH 4 and N20, and the Dutch use of CFCs and Halons, expressed in CO2-equivalents. The production of CFCs and Halons are also component parts of the theme indicator on 'the depletion of the ozone layer'. The aggregation reckons with the fact that Halon 1301 damages the ozone layer more than 10 times as much as the reference substance CFC-I 1. The emission of Halon 1301 is therefore given a 10 times greater weight in the indicator concerned. Another example is the theme indicator for 'the disturbance of local environment', which consists of various aspects of noise disturbances and other disturbances at the local level (eg odour). Figure 1 depicts the time profile of the seven theme indicators of Adriaanse, and shows that this disturbance indicator is the only one with a clear upward trend. Most other indicators show a downward trend. We should note that Adriaanse also combines these indicators to two alternative overall 'theme indices', He derives weights from target values of environmental policy in the year 2000, and alternatively from some normatively determined 'sustainability levels'. Hence, his methodology differs in two ways from the construction method of the index by Hope c.s., which we follow. Firstly, he needs for each theme indicator some reference value which is chosen as policy target or which is selected as 'sustainability level'. Necessarily this introduces an element of policy or expert's judgement in the construction method. Secondly, given these reference values, each theme indicator would obtain the same weight when all indicators were at an equal distance from their reference values. On the other hand, the methodology of Hope c.s. assigns different weights to the theme indicators in conformity with measured differences in preferences. However, in all fairness it must be said that the design of an overall environmental index is not a major aim of the data construction by Adriaanse. The main concern of his study is to calculate the contribution of a number of target groups such as agriculture, air transport, industry, consumers, etc. to the environmental pressure as measured by the seven theme indicators. As mentioned before, we follow Hope c.s. and use the results from the European Omnibus Survey to determine the set of weights for the aggregation of the theme indicators to one overall index. The survey is carried out in 1982, 1986 and regularly up to the most recent survey of 1996. It asked

c.s.

Indicator

millions 1000 tonnes

97

Table 2 The indicators of Adriaanse (1993) as components of the environmental index

Indicator

Unit

Change of climate (contribution to greenhouse effect) Depletion of the ozone layer Acidification Eutrophication Dispersion of toxic substances Disposal of solid waste Disturbance of local environments

CO2-equivalents Ozone depletion-equivalents Acidification-equivalents Eutrophication-equivalents Dispersion-eqiuvalents Disposal-equivalents Disturbance-equivalents

98 A pilot index for environmental policy in The Netherlands." F A G den Butter and J A C van der Eyden 110

-

-

100

~ 9o II

o

80

or,

~

70

x

,-

60 50

I

40

I

I

I

I

r

I

I

I

I

I

1

1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Year 120 110 Q --

100

II O oo

o~

90

.~ 80 70

_

I

60

Eutrophication

0

I

I

I

I

~

I

[

I

I

D

I

I

I

1980 1981 1982 1983 19841985 1986 1987 1988 1989 1990 1991 Year 140 D 130 ~

Dispersion

0 Disposal

/~

~ ' ~

120

Constructing the pilot environmental index for The Netherlands

II 0

~

110

o

I00

t-

90 80

ie the results for The Netherlands in 1982 and 1996, the results for all 12 members states in the EC in 1996, and the results for Italy in 1996. In the European Omnibus Survey, the measure of concern, taken for each problem, is to multiply the proportion who said they worry a great deal by 3, a fair amount by 2, and not very much by 1, and add the multiplied proportions together. Hence, when everybody in the survey worried a great deal about some problem, this problem would have a measure of concern of 3.0--the highest possible note for the concern. The table shows that the concern on local environmental problems is much smaller than the concern on national problems and problems on a worldwide scale. In The Netherlands the concern on environmental problems has, on average, not changed very much in 1996 as compared to 1982. The Dutch are very concerned about industrial waste disposal, whereas the Italians regard marine pollution and air pollution as the most important national problems. The concern for the destruction of the ozone layer is relatively high as compared to the other global environmental problems. The data for The Netherlands indicate that the concern for the greenhouse effect has increased quite substantially between 1982 and 1996. This is probably due to increased attention for this problem in the media. It is remarkable that the concern of the Italians about environmental problems is above the European average, and also much higher than the concern of the Dutch. This may, however, be attributed to a general national attitude of being concerned about things. In the calculation of the aggregation weights for the environmental index it is, due to normalization, the relative concern and not the absolute concern that matters.

II

I

I

I

I

I

I

I

I

I

I

I

1980 1981 1982 1983 19841985 1986 1987 1988 1989 1990 1991 Year

Figure 1 Time pattern of envrionmental indicators for The Netherlands

over 10,000 adults across the EC whether they were worried a great deal, a fair amount, not very much, or not at all about 14 environmental problem~ These 14 environmental problems are listed in Table 3. It should be noted that in the survey of 1996 the question on the concern for the loss of farmland, a local environmental problem which does not play a further role in our calculations, was replaced by a question on the concern for the destruction of the ozone layer. This change in the questionnaire enabled us to give a weight to Adriaanse's theme indicator 'depletion of the ozone layer' without further need of a translation key. Table 3 shows four different results from the opinion poll,

The first set of assumptions for the aggregation weights relate to the transformation of the 14 environmental policy problems of the European Omnibus Questionnaire to the seven themes of environmental policy in The Netherlands, which become part of the overall index. Table 4 gives the transformation key which is used in the basic version of the index. Our assumptions on this transformation are as much as possible in conformity with the transformation used by Hope c.s. However, we have to make some assumptions of our own, because the nine component parts of the pilot index for the United Kingdom differ from the seven themes of environmental policy which are the components for the index for The Netherlands. The major additional assumptions of our own needed to be taken for the themes acidification and eutrophication. Our arguments for selecting the specific transformation key are explained in the table. Next we have to decide which results from the European Omnibus Survey we use for composing the weights of the Dutch index. In our basic version we have taken, quite obviously, the results for The Netherlands of the last European Omnibus Survey, which is available, ie that of 1996. Application of the transformation key to these results and normalization gives the weights for our basic version of the index.

A pilot index for environmentalpolicy in The Netherlands." F A G den Butter and J A C van der Eyden Table 3

99

Weights from 'European Omnibus Survey' with respect to concern about environmental problems

Environmental problem

Net82

Net96

EC96

lta96

Local 1 Drinking water 2 Noise 3 Air pollution (local) 4 Waste disposal 5 Access to countryside 6 Damage to landscape

0.27 0.63 0.67 0.76 0.32 0.89

0.35 0.69 0.81 0.74 0.54 0.88

0.90 0.97 1.17 1.05 0.87 1.12

1.42 1.37 1.69 1.71 1.37 1.56

National 7 Pollution of rivers and lakes 8 Marine pollution 9 Air pollution (national) 10 Industrial waste disposal

2.25 2.38 2.11 2.45

1.85 2.03 1.93 2.11

2.32 2.39 2.29 2.36

2.44 2.56 2.63 2.55

Global 11 Extinction of species 12 Loss of natural resources 13 Greenhouse effect 14 Destruction of the ozone layer~

2.12 1.95 1.68 -

2.32 2.16 2.13 2.30

2.40 2.32 2.38 2.52

2.45 2.52 2.58 2.67

Source: European Omnibus Survey Net82=Weights from European Omnibus Survey for The Netherlands in 1982.Net96=idem in 1996; EC96=idem for the whole EC in 1996; lta96=idem for Italy in 1996. a For Net82, survey results on destruction of the ozone layer are not available; see Table 4 for the translation key which has been used.

Table 4

Key to transformation from components of the European Omnibus Survey to the compound indicators of Adriaanse (basic version)

Indicators in environmental index

Transformation key to questions in European Omnibus

Climate change (greenhouse effect only) Depletion of ozone layer~ Acidification h

Greenhouse effect Destruction of the ozone layer (NOx+SO2+NH3)=0.06*greenhouse effect+2* [air pollution (local+national)] + drinking water Drinking water + pollution of rivers and lakes Pollution of rivers and lakes Local waste disposal + industrial waste disposal Noise

Eutrophication" Dispersion of toxic substances Disposal of solid waste Disturbance

Weight in basic version Net96 0.12 0.13 0.33 0.12 0.10 0.16 0.04

~'The destruction of the ozone layer was not part of the questionnaire for 1982; for that year we have calculated the weight in conformity with Hope et al (1992) as 0.12*Greenhouse effect+[Air pollution (local + national)]. h Adriaanse ascribes acidification to NO~, SO 2 and NH3; according to Hope c.s., NOx is weighted as {0.06*greenhouse effect + [air pollution (local+national)] SO 2 obtains the weight [Air pollution (local+national)]; NH 3 relates mainly to use of fertilizers and manure; Hope c.s. assign the weight of the quality of 'drinking water' to this component; this gives the transformation formula from the table. Eutrophication is mainly caused by the use of fertilizers and manure which is given the weight of the quality of drinking water by Hope c.s.; moreover, with respect to this component water pollution due to waste disposal of households and industry is of importance as well; therefore this transformation key has 'pollution of rivers and lakes' as an additional aspect in its weight.

They are reported in the third column of Table 4. It appears that in our index, acidification obtains the highest weight. The Netherlands is one of the European countries which is most affected by acidification and apparently this theme of environmental policy is both of quantitative importance and major concern in this country. The next themes in order of importance are depletion of the ozone layer and disposal of solid waste. Surprisingly, disturbance of the local environment by, amongst others, noise, obtains the smallest weight. T h i s is p a r t l y a c o n s e q u e n c e o f t h e f a c t t h a t a c c o r d i n g t o t h e

European Omnibus Survey local environmental problems c a u s e less c o n c e r n t h a n n a t i o n a l o r w o r l d w i d e p r o b l e m s .

Figure 2 pictures the basic version of our pilot environmental index. The time profile of the index shows an almost steady decline. It implies a reduction of environmental p r e s s u r e , w h i c h is e s p e c i a l l y a p p a r e n t

a f t e r 1985. O n t h e

o t h e r h a n d , t h e i n d e x h a s r e m a i n e d v i r t u a l l y c o n s t a n t in t h e period 1980-1985.

106 104

m

m

102 100 98 96 94

-~

92 90

-= 88 86 84 82

80

78 76 74

I

I

1980

I

I

1982 1981

I

I

1984 1983

I

I

1986 1985

I

I

1988 1987

I

I

1990 1989

1991

Year Figure 2 T h e p i l o t i n d e x f o r T h e N e t h e r l a n d s basic version

a c c o r d i n g to t h e

100

A pilot index for environmental policy in The Netherlands: F A G den Butter and J A C van der Eyden

rather small. Anyhow, as shown in Figure 3, the changes in the time profile of the index are much smaller under these alternative assumptions than when we had just given an equal weight to each theme-indicator. The change in the time profile of the index is even more pronounced when we give a large weight to disturbance (indicated by AO4 in Table 6 and Figure 3).

Table 5 Alternative keys to transformation; changes with respect to basic version AOI:

AO2

AO3

it is assumed that disturbance due to odour is also part of disturbance of the local environment due to noise; according to Adriaanse 'odour' is equal to 0,4* noise so that we have 'disturbance' as 1,4" 'noise'. Because of the massive publicity on climate change the concern with respect to this component is supposed to be doubled as compared to the basic version: 'climate change' now obtains two times the weight of 'greenhouse effect'. This alternative assumes much concern on dispersion: 'dispersion' now gets two times the weight 'pollution of rivers and lakes'.

Conclusion This paper applies the methodology of Hope c.s. for the construction of a composite overall index for environmental policy in The Netherlands by means of aggregation of the annual time series data collected by Adriaanse on seven theme indicators of environmental policy. The results from the European Omnibus Survey of opinion polls on concern with environmental problems are used to determine the aggregation weights. The basic version of our index shows a downward trend after 1985. This may indicate that environmental pressure has decreased during that period, but it does, by no means, imply that environmental quality has increased. In most cases it takes a long time before a reduction in environmental pressure results in an enhanced quality of the environment. The sensitivity analysis on the assumptions needed to determine the weights shows that our

Sensitivity analysis This section investigates how the time profile of the index is affected by changes in the assumptions on the transformation key and the results from the European Omnibus Survey. Table 5 lists three alternative assumptions with respect to the transformation key. Table 6 presents a variety of aggregation weights, which are obtained when these alternative assumptions are combined with the four different sets of results of the European Omnibus Survey reported in Table 3. The table illustrates that the weights do not change very much under these alternative assumptions. It implies that the sensitivity of the overall index for these alternative assumptions is

Table 6

Alternative weights in the sensitivity analysis

Indicator

Weights

Climate change Depl. of ozone layer Acidification Eutrophication Dispersion Disposal Disturbance

AO l/Net96 0.12 0.12 0.33 0.12 0.10 0.16 0.05

AO2/Net96 0.21 0.12 0.30 0.11 0.09 0.14 0.03

AO3/Net96 0.11 0.12 0.30 0.11 0.19 0.14 0.03

Bas/Net82 0.09 0.15 0.31 0.13 0.12 0.17 0.03

Climate change Depl. of ozone layer Acidification Eutrophication Dispersion Disposal Disturbance

AOI/Net82 0.09 0.15 0.30 0. t 3 0.12 0.16 0.05

Bas/EC96 0.11 0.11 0.35 0.14 0.10 0.15 0.04

Bas/Ita96 0.09 0.10 0.37 0.14 0.09 0.16 0.05

equal weights I/7 1/7 1/7 1/7 1/7 1/7 I/7

Climate change Depl. of ozone layer Acidification Eutrophication Dispersion Disposal Disturbance

AO2/ita96 0.17 0.09 0.34 0.13 0.08 0.14 0.05

AO2/Net82 0.16 0.14 0.29 0.12 0.11 0.15 0.03

AO3/Net82 0.08 0.14 0.27 0.12 0.21 0.15 0.03

AO 1/ita96 0.09 0.09 0.37 0.14 0.09 0.15 0.07

Climate change Depl. of ozone layer Acidification Eutrophication Dispersion Disposal Disturbance

AO3/lta96 0.09 0.09 0.34 0.13 0.16 0.14 0.05

AO4 1/12 I/12 1/12 1/12 1/12 1/ 12 1/2

A pilot index for environmentalpolicy in The Netherlands." F A G den Butter andJ A C van der Eyden

106 104 102 100 98 96 94 ~ 92 •~ 90 --~ 88 86 84 82 80 78 76 74

Figure 3

-

-_ -

rn Bas/Net 96 + AOl/Net 96 0 AO2/lta 96 A Equal x AO3/lta96 v AO4

1

-I

I

I

I

I

"~,k'N~ ~,~>~ r~ l ~ ta~ x t3~,~ x~i I

I

I

I

I

I

101

Another possibility is to let the weights be determined by experts, possibly in some kind of Delphi-procedure where they react to each others opinion, rather than by public opinion polls. Moreover, the choice of the component parts of the environmental index is very much restricted by the availability of data. More data collection a n d more methodological research is needed in order to come to a comprehensive environmental index, which can be used actively in policy analysis and which can be incorporated into macroeconomic policy models.

I

1980 1982 1984 1986 1988 1990 1981 1983 1985 1987 1989 1991 Year The envrionmental index according to different weights

aggregation method is not very sensitive to alternative assumptions. Yet our study takes a very specific look at the aggregation problem in the construction of an overall environmental index. For instance, following Hope c.s., .we assume a linear aggregation and assume the outcomes of the opinion polls exogenous with respect to the relative importance of the environmental problems. However, the results could very well be endogenous as the concern on a specific problem would become greater in case the environmental pressure with respect to that problem increases. For instance, noise disturbance obtains a very low weight in our index, whereas its indicator is the only one which shows a steady rise during the reference period. Hence, its weight may become larger in the future. In that case we need non-linear aggregation, eg by using a CES-function or some kind of spline functions where the weight is proportional to the distance between the actual indicator and some target or sustainability level, t In such an aggregation scheme, noise disturbance would have obtained a much larger weight, and the time profile of the index for The Netherlands would have changed much more than in most variants of our sensitivity analysis. iIt should be noted that the constructionmethod of Adriaanse ['orhis two alternative overall indicesimpliesthat larger weights are given to indicators which are further away from their target or sustainabilitylevels.

Acknowledgements Useful remarks of a referee on a previous version of this article are gratefully acknowledged.

References Adriaanse, A. (1993). Environmental Policy PerformanceIndicators, Sd u Uitgeverij, The Hague. den Butter, E A. G. (1992). The mirror of cleanliness:on the construction and use of an environmental index, in: J. J. Krabbe and W. J. M. Heijman (Eds.), National lncome and Nature. Externalities. Growth and Steady State, Kluwer Academic, Dordrecht, 49-75. den Butter, F. A. G., & Verbruggen,H. (1994). Measuring the trade-off between economic growth and a clean environment. Environmental and Resource Economics, 4, 187-208. Diener, E. (1995). A value based index tbr measuring national quality of life. Social Indicators Research, 36, 107-127. Grossman, G. M., & Krueger, A. B. (1995). Economic growth and the environment. Quarterly Journal of Economics, 112, 353-377. Hicks, N., & Streeten, P. (1979). Indicators of development: the search lbr a basic needs yardstick. WormDevelopment, 7, 567-580. Hope, C., Parker, J., & Peake, S. (1992). A pilot environmental index tbr the UK in the 1980's.Energy Policy, 20, 335-343. Lind, N. C. (1993). A compound index of national development. Social Indicators Research, 28, 267-284. W. R. Ott, Environmental Indices; Theory and Practice, Ann Arbor Science, Michigan, 1978. Ram, R. (1982). Composite indices of physical quality of life, basic needs fulfilment, and income; a 'principal component' representation. Journal of Development Economics, 11,227-247. Slottje, D. J. (1991). Measuring the quality of life across countries. The Review of Economics and Statistics, 73, 684-693. United Nations Development Program, Human Development Report 1990, Oxtbrd University Press, New York, 1990. World Bank, The development of the environment, WorldDevelopment Report 1992, Ox~brd University Press, New York, 1992.