The Economic Valuation of Green Electricity

SPRINGER BRIEFS IN ENVIRONMENTAL SCIENCE
Simona Bigerna · Paolo Polinori
The Economic
Valuation of
Green Electricity
123
SpringerBriefs in Environmental Science
SpringerBriefs in Environmental Science present concise summaries of cutting-edge
research and practical applications across a wide spectrum of environmental fields,
with fast turnaround time to publication. Featuring compact volumes of 50 to
125 pages, the series covers a range of content from professional to academic.
Monographs of new material are considered for the SpringerBriefs in Environmental
Science series.
Typical topics might include: a timely report of state-of-the-art analytical tech-
niques, a bridge between new research results, as published in journal articles and a
contextual literature review, a snapshot of a hot or emerging topic, an in-depth case
study or technical example, a presentation of core concepts that students must
understand in order to make independent contributions, best practices or protocols
to be followed, a series of short case studies/debates highlighting a specific angle.
SpringerBriefs in Environmental Science allow authors to present their ideas and
readers to absorb them with minimal time investment. Both solicited and unsolicited
manuscripts are considered for publication.
More information about this series at http://www.springer.com/series/8868

Simona Bigerna • Paolo Polinori
The Economic Valuation

of Green Electricity
Simona Bigerna Paolo Polinori
Università di Perugia Università di Perugia
Perugia, Italy Perugia, Italy
ISSN 2191-5547 ISSN 2191-5555 (electronic)

SpringerBriefs in Environmental Science
ISBN 978-94-024-1572-8 ISBN 978-94-024-1574-2 (eBook)
https://doi.org/10.1007/978-94-024-1574-2
Library of Congress Control Number: 2018952352
© The Author(s), under exclusive licence to Springer Nature B.V. 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature B.V.
The registered company address is: Van Godewijckstraat 30, 3311 GX Dordrecht, The Netherlands
Introduction
The aim of the book is to analyze the relationship between renewable (or green)
electricity and citizens, focusing on both the demand side and the supply side. Today
the consequences of the use of fossil energy are seen from a different perspective
because issues related to climate change are evident worldwide. Thus, climate
change and resource depletion are real problems to be addressed for the welfare of
society. Renewable energy sources are essential to reduce polluting emissions, but
they can produce a range of environmental effects, which might be detrimental to
human activities, as attested to by several types of “Not in My Back Yard” (NIMBY)
reactions. This is because renewable energy infrastructure siting usually implies
several pros and cons to the local stakeholders involved in the projects. For example,
in Italy, according to the last report available in 2016 (Nimbyforum 2017), there are
359 contested facility projects and, among these, 45% involve renewable energy.
Nevertheless, empirical evidence (from the Eurobarometer survey, among other
sources) shows that in Italy, as in several European countries, citizens are willing to
pay a significant amount to produce renewable electricity. Renewable electricity
production involves citizens from two opposite points of view. Indeed, they are
involved both as end users and as stakeholders in the construction of the facilities
and in the local production process. In this book we analyze this dual role played by
citizens in order to evaluate the actual and global public acceptance of renewable
electricity generation in Italy.
We address a specific and important area of the economic analysis—the stated
preferences method—focusing on two welfare measures: willingness to pay and
willingness to accept.
Consequently, the research evaluates the attitudes of citizens toward the end use
of renewable electricity and the likelihood of acceptance of new infrastructure
related to renewable electricity generation. Our aim is not to consider all technolo-
gies; we focus only on site-specific cases that involve siting controversies. To do this
we focus on empirical results in Europe, including Italian case studies, comparing
them with our contingent valuation field experiments. Furthermore, in our empirical
v
vi Introduction
analysis we explicitly take into account the uncertainty associated with the respon-
dents in order to obtain more robust results.
The book, therefore, is not about how to reconcile consumers’ and citizens’
behavior regarding renewable electricity consumption and production; rather, it
explores the main determinants of people’s behavior, on the two sides of the market,
for better understanding of this phenomenon, to obtain useful information for public
and private decision makers. The structure of the book is as follows.
In Chap. 1 we use a meta-analysis to collect and analyze related literature about
renewable electricity adoption, taking into account the double role played by the
citizens. The main determinants of citizens’ behavior are analyzed for better under-
standing of the renewable electricity adoption process.
In Chap. 2 we investigate whether existing wind farms affect respondents’
attitudes and perceptions toward potential enlargement of wind farms, using a
contingent valuation method. We explicitly take into account the existence of
respondents’ heterogeneity in perceiving the new project externalities referring to
potential land use conflicts and local opposition. To do this, we use both willingness
to pay and willingness to accept measures, and we also appraise the impact of
uncertainty, taking into account several degrees of uncertainty, using the numerical
scale method.
In Chap. 3 we estimate Italian households’ willingness to pay for renewable
electricity, comparing our results with those of other similar studies conducted
worldwide. Furthermore, we use two approaches to treat uncertainty, appraising
consumers’ willingness to pay for renewable electricity, to provide more robust
results.
Contents
1 Citizens’ Versus Consumers’ Attitudes Toward Renewable

Electricity: What the Literature Tells Us in a Contingent
Valuation Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.1 Meta-analysis Approach . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.2 Review of Related Literature . . . . . . . . . . . . . . . . . . . . . . 4
1.2.3 Topics of Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2.4 What the Literature Tells Us . . . . . . . . . . . . . . . . . . . . . . 7
1.2.5 Quantitative Analysis: Metaregressions . . . . . . . . . . . . . . . 10
1.2.6 Qualitative Analysis: Local Survey . . . . . . . . . . . . . . . . . . 12
1.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3.1 Willingness to Pay for Renewable Electricity
on the Demand Side . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.3.2 Willingness to Accept and Willingness to Pay
for Renewable Electricity on the Supply Side . . . . . . . . . . 15
1.3.3 Local Survey Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2 Evaluating an Onshore Wind Farm Enlargement Project:
A Contingent Valuation Study in Central Italy . . . . . . . . . . . . . . . . . 27
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2 Related Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3 Method for Valuing Wind Farm Enlargement . . . . . . . . . . . . . . . . 31
2.3.1 The Scenario: Wind Power Generation in Italy . . . . . . . . . 31
2.3.2 Case Study: The Monte Cucco Regional Park . . . . . . . . . . 31
2.3.3 Survey Method and Questionnaire . . . . . . . . . . . . . . . . . . 33
vii
viii Contents
2.4 Theoretical and Econometric Framework . . . . . . . . . . . . . . . . . . . 34

2.4.1 Theoretical Model and Elicitation Format . . . . . . . . . . . . . 35
2.4.2 Econometric Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.5 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.5.1 Descriptive Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.5.2 Econometric Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.5.3 Welfare Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3 Consumers’ Willingness to Pay for Renewable Electricity
in Italy: A Comparative Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.2 Related Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.3 Renewable Electricity in Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3.1 Cost of Renewable Electricity in Italy . . . . . . . . . . . . . . . . 63
3.4 Methods and Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.4.1 Theoretical Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.4.2 Survey Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.4.3 Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.4.4 Elicitation Format and Econometric Model . . . . . . . . . . . . 68
3.5 Empirical Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.5.1 Estimation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.6 Comparative Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Abbreviations
ANEV National Wind Energy Association

CEx Choice experiment
CV Contingent valuation
DBDC Double bound dichotomous choice
DETR Department of Environment, Transport and the Regions
DK Do not know
DN Definitely no
DY Definitely yes
EU European Union
HB Higher bound
LB Lower bound
LHS Left hand side
MBDC Multiple bound dichotomous choice
NAP National Action Plan
NCS Numerical certainty scale
NIMBY “Not in My Back Yard”
NSM Numerical scale method
OE Open ended
OLS Ordinary least squares
PC Payment card
PN Probably no
PY Probably yes
REn Renewable energy
REnS Renewable energy sources
RE Renewable electricity
RHS Right hand side
SBDC Single bound dichotomous choice
SPC Stochastic payment card
SUR Seemingly Unrelated Regression
WTP Willingness to pay
WTA Willingness to accept
ix
List of Figures
Fig. 1.1 WTP and WTA kernel density estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Fig. 2.1 Umbria and Monte Cucco Regional Park . . . . . . . . .. . . . . . . .. . . . . . . .. . . . . 32
Fig. 2.2 Elicited willingness to pay (WTP) and willingness to accept
(WTA) according to degrees of uncertainty (2014 EUR) . . . . . . . . . . . . 40
Fig. 2.3 Unweighted and weighted willingness to pay (WTP)
and willingness to accept (WTA) distributions . . . . . .. . . . . . . . . . .. . . . . . . 40
Fig. 2.4 Unweighted willingness to pay (WTP) and willingness to accept
(WTA) distributions according to six degrees of certainty . . . . . . . . . . . 41
Fig. 2.5 Estimated willingness to pay (WTP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Fig. 2.6 Estimated willingness to accept (WTA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Fig. 3.1 Uncertainty preferences model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Fig. 3.2 Uncertainty preferences model according to different
approaches. a The Welsh and Poe approach. b The Broberg
and Brännlund approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Fig. 3.3 Diagonal response patterns .. . . . .. . . . .. . . .. . . . .. . . . .. . . .. . . . .. . . . .. . . .. . . 72
Fig. 3.4 Survivor function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
xi
List of Tables
Table 1.1 Metaregression: willingness to pay (WTP) for renewable

electricity (RE) .. . .. . . .. . .. . . .. . .. . . .. . .. . . .. . .. . . .. . .. . . .. . .. . . .. . . .. . 14
Table 1.2 Metaregression: willingness to pay/willingness
to accept (WTP/WTA) for an onshore wind farm . . . . . . . . . . . . . . . . . 16
Table 1.3 Subsample descriptive statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
Table 1.4 Partial correlations . .. . . . .. . . . .. . . . .. . . . . .. . . . .. . . . .. . . . .. . . . . .. . . . .. . . 20
Table 2.1 Sample answers: descriptive statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Table 2.2 Estimated willingness to pay (WTP) . .. . . .. . . . .. . . . .. . . .. . . . .. . . .. . . 42
Table 2.3 Estimated willingness to accept (WTA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Table 3.1 Willingness to pay (WTP) in the renewable electricity
(RE)–related literature . . .. . . .. . . . .. . . .. . . .. . . .. . . . .. . . .. . . .. . . .. . . . .. . 56
Table 3.2 A3 component analyzed according to different types
of users (2007) . . . . . . . . . .. . . . . . . . . . . . .. . . . . . . . . . . .. . . . . . . . . . . . .. . . . . . . . 65
Table 3.3 Elicitation format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Table 3.4 Variables and descriptive statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Table 3.5 Details of payment responses .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . .. . . 75
Table 3.6 Interval data regression for willingness to pay (WTP)
to support renewable electricity (RE) in Italy per model:
the Broberg and Brännlund approach . .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . 77
Table 3.7 Interval data regression for willingness to pay (WTP)
to support renewable electricity (RE) in Italy per model:
the Welsh and Poe approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Table 3.8 Market sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Table 3.9 Comparable willingness to pay (WTP) for renewable
electricity (RE): studies that consider WTP
and the target cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Table 3.10 Willingness to pay (WTP) for renewable electricity (RE):
studies that provide individual and/or aggregate WTP . . . . . . . . . . . . 83
Table A.1 Survey respondents’ and country residents’
characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
xiii
Chapter 1
Citizens’ Versus Consumers’ Attitudes
Toward Renewable Electricity: What
the Literature Tells Us in a Contingent
Valuation Framework
Abstract In dealing with renewable electricity (RE), individuals are involved both
as end consumers on the demand side and as stakeholders (citizens) in the local
production process on the supply side. Empirical evidence shows that in many
countries, consumers are willing to pay a significant amount to facilitate adoption
of RE. In contrast, environmental externalities are often the cause of strong opposi-
tion to RE adoption if local communities are involved as stakeholders in wind, solar,
or biomass investment projects. Looking at the literature on willingness to pay and
on willingness to accept, we have investigated RE acceptance mechanisms. In this
chapter, we use a meta-analysis to assess the major determinants of RE acceptance
on both the demand and supply sides. This meta-analysis has provided some insights
that are useful for managing field research on an onshore wind farm enlargement
project located in the Umbria region of Italy. The meta-analysis and survey results
confirm that the local community plays a central role in local RE acceptance.
Furthermore, people who have previous experience of windmills require less com-
pensation, or are willing to pay more, for RE development. The results suggest that
these attributes should be included in future research to improve understanding of
determinants of RE acceptance.
Keywords Renewable electricity · Meta-analysis · Willingness to pay · Willingness

to accept · Local community
1.1 Introduction
Although climate change has been recognized as a scientific problem of global

interest since 1980, in the past, policy makers and the media have frequently asserted
that climate science was uncertain and that more scientific evidence was necessary.
In the political context and as part of public opinion, this evoked uncertainty has
been used as an argument against adopting incisive measures to reduce greenhouse
gas emissions.
In recent years the scientific nature of climate change has been accepted world-
wide (Mu and Mu 2013), given that the consequences of the use of fossil energy are
© The Author(s), under exclusive licence to Springer Nature B.V. 2019 1

S. Bigerna, P. Polinori, The Economic Valuation of Green Electricity, SpringerBriefs
in Environmental Science, https://doi.org/10.1007/978-94-024-1574-2_1
2 1 Citizens’ Versus Consumers’ Attitudes Toward Renewable. . .
clearly evident. Climate change and resource depletion have became real problems
to be addressed for the welfare of society; consequently, renewable electricity
(RE) has been confirmed as an important tool to contribute to reducing polluting
emissions. Consumers seem to share this perception of the important function that
RE plays in the environmental and energy scenario, given that empirical evidence
shows that in many countries, individuals are willing to pay to support RE; that is,
consumers are willing to pay an extra premium for electricity produced using
renewable sources.
On the other hand, RE generation can produce a range of environmental local
externalities that might be detrimental to human activities, as confirmed by the “Not
in My Back Yard” (NIMBY) syndrome (Wolsink 2000). Local communities are
involved in the infrastructure-siting process with different pros and cons, and often
negative environmental externalities cause strong opposition to RE diffusion.
These results seem to suggest that individuals have different behavior according
to the roles they take on. As stakeholders, on the supply side, they are more affected
by local negative externalities, while as consumers, on the demand side, they are
more affected by global positive externalities due to RE.
In this chapter, we survey the existing literature related to RE adoption, taking
into account this double role played by individuals. We address an important area of
the economic analysis by investigating the literature on willingness to pay (WTP)
and willingness to accept (WTA) related to RE. Consequently, we analyze the
relationship between RE, consumers, and citizens, focusing on the literature related
to demand and supply. We survey research evaluating both the attitudes of con-
sumers toward RE use and the likelihood of citizens’ acceptance of new RE
investment projects.
On the demand side we focus on WTP for RE use. We want to point out that we
consider neither “green power products” nor “labeled green electricity” such as in
Wüstenhagen et al. (2003) and Ek and Söderholm (2008), respectively.1
In this chapter we analyze a generic stated intention to purchase RE. In the papers
collected and reviewed, respondents are generally asked if they want to voluntarily
contribute to RE development, purchasing more RE. In this context we focus on the
determinants of respondents’ choices.
With regard to the supply side we have collected papers using the contingent
valuation (CV) method and a choice experiment (CEx) in evaluating acceptance of
green energy investment programs. We focus on the local community reaction in
terms of opposition to and acceptance of investment projects, pointing out the main
determinants of projects outcomes.
We want to underline that this chapter is not about how to reconcile consumers
and citizens behaviors; rather, it explores the main determinants of people’s behavior
for better understanding of the preferences involved in the double role considered.
The method used is mixed. First, we analyze the related literature by meta-
analysis regression; then, using the results obtained, we conduct a local survey
1
For a recent review see Herbse and Frienge (2017).
1.2 Methods 3
assessing perceived net benefits associated with an onshore wind farm enlargement
project in the Umbria region of Italy.
The general aim of this chapter is to explore the mechanisms related to RE
acceptance. In detail, we expect to improve knowledge from three points of view.
First, we investigate consumers and citizens’ attitudes toward RE to help understand
whether common determinants in the acceptance process exist. Second, we analyze
local community characteristics, focusing on the collective dimension such as “sense
of place” or “place attachment,” given that they are important psychological and
emotional determinants. Finally, we assess how the daily windmill experience that
has affected the local community since 1990 currently affects the local community
attitude toward the new wind project. In the literature these points have seldom been
investigated, so we are confident that our results can improve knowledge of these
issues. Finally, we also generalize our results, extending them from the Umbria case
to RE sector development.
The chapter is organized as follows. In Sect. 1.2, after presenting the related
literature, we briefly discuss the method used. In Sect. 1.3 we illustrate the models
applied, discussing the main results. In Sect. 1.4 we draw conclusions and make
generalizations for the RE sector.
1.2 Methods
Operationally, we base our framework of analysis on four steps. According to the

meta-analysis we initially examine and classify the collected papers (primary
sources) according to the criteria specified below. Then, we clarify the research
questions improving the knowledge according to the aim of this chapter. In the next
step we apply metaregression analysis, primarily using primary information. Finally,
the insights from the metaregression results are used as key to interpretation in the
local survey on the enlargement project of the existing windmill in a central area in
Umbria.
In this way we have used a hybrid design approach that allows us to capture both
qualitative and quantitative aspects entailed in our research question (Edmonson and
McManus 2007).
1.2.1 Meta-analysis Approach
Meta-analysis is a useful tool for comparison and synthesis of various studies and
research. By the end of the twentieth century, the meta-analysis method had been
widely used in the fields of experimental psychology and clinical medicine, allowing
aggregation and synthesis of information from multiple experiments and/or clinical
research, in order to produce more robust and generalized results. Using a wide range
of individual publications to integrate and summarize literature results, meta-
analysis is defined as “the study of studies” (Holmgren 2007); consequently, the

scope of meta-analysis is quite complex and varied because it involves both social
sciences and the environmental dimension.
In social science and behavioral studies, meta-analysis allows comparison of
similar results, given the semicontrolled nature of the experiments that are conducted
in these fields (Stanley et al. 2013). Furthermore, this method gives more informa-
tion through synthesis and interpretation keys, starting from primary sources, to
improve the available knowledge.
1.2.2 Review of Related Literature
The literature is presented, highlighting the path that allows us to formulate our
research questions. We use more than 30 papers, which is a usual sample in meta-
analysis. The papers are almost equally divided between the demand and supply
sides, published from 1999 to 2014, and mainly related to European, North Amer-
ican, and Asian countries. This sample allows us to manage about 300 primary
information sources, which is a dimension comparable with the international liter-
ature. According to the double perspective described in the introduction, only
Groothius et al. (2008) and Ek and Matti (2014) point out the incongruence between
the RE demand and supply sides. Indeed, each of the reviewed papers deals with
only one of the two perspectives. Following the approach taken by Nelson and
Kennedy (2009), we analyze the papers taking into account both the methodological
and factual heterogeneity.
Methodological heterogeneity is related to differences in empirical analysis
employed by researchers; consequently, it is due to the study design adopted, survey
method used, information collected, type of data, and econometric techniques.
Among several welfare measure available, WTP is the measure most frequently
used; to the best of our knowledge, only Groothius et al. (2008, 2010), Strazzera
et al. (2012), and Mirasgedis et al. (2014) have used WTA.
The elicitation methods used in the literature are mainly CEx (Álvarez-Farizo and
Hanley 2002; Bergmann et al. 2006; Borchers et al. 2007; Byrnes et al. 1999;
Dimitropoulos and Kontoleon 2009; Ek 2005; Ek and Matti 2014; Goett et al.
2000; Grösche and Schröder 2011; Meyerhoff et al. 2010; Meyerhoff 2013;
Strazzera et al. 2012) and CV. According to Lancaster’s (1966) theory of consump-
tion, CEx allows us to obtain a more detailed representation of good under evalu-
ation linking the utility of respondents to the primary characteristics of the good
evaluated. On the other hand, CV outcomes are more accurate especially when the
scenario proposed involves a policy target influencing the good evaluated (Hanley
et al. 1998). In the CV approach the open-ended (OE) elicitation method is used by
Mirasgedis et al. (2014) whereas the single bound (SB) elicitation method is used by
Batley et al. (2000, 2001), Ivanova (2005, 2012), and Zoric and Hrovatin (2012).
1.2 Methods 5
The double bound (DB) elicitation method is applied by Nomura and Akay (2004),
Groothius et al. (2008), Koundouri et al. (2009), Yoo and Kwak (2009), Groothuis
(2010), and Kim et al. (2012); finally, a payment card (PC) is utilized by Hansla et al.
(2008), Bollino (2009), and Bigerna and Polinori (2014).
Among several types of survey, face-to-face interviews represent the most fre-
quent approach for the papers related to the supply side (Álvarez-Farizo and Hanley
2002; Dimitropoulos and Kontoleon 2009; Strazzera et al. 2012); some researchers
also use this type of interview on the demand side (Borchers et al. 2007; Yoo and
Kwak 2009; Kim et al. 2012; Zoric and Hrovatin 2012).
Factual heterogeneity concerns the period and context of the investigation and
characteristics such as the type of respondents, type of infrastructure, or localization
and site characteristics. In the case of RE studies the first kind of heterogeneity is
usually represented by variables related to three main categories of determinants:
(1) the type of respondents; (2) institutional and contextual conditions; and (3) pro-
ject and service characteristics. The type of respondent can be an individual (such as
in Byrnes et al. (1999), Goett et al. (2000), Batley et al. (2001), Álvarez-Farizo and
Hanley (2002), Bergmann et al. (2006), Borchers et al. (2007), Groothius et al.
(2008), Hansla et al. (2008), Koundouri et al. (2009), Meyerhoff et al. (2010),
Hanemann et al. (2011), Strazzera et al. (2012), Zhang and Wu (2012), Botetzagias
et al. (2013), Meyerhoff (2013), Ek and Matti (2014), and Mirasgedis et al. (2014),
the head of the family (Yoo and Kwak 2009), or the homeowner (Ek 2005), but
researchers often refer to the “family” as a “unified” research unit (such as in
Nomura and Akay (2004), Ivanova (2005, 2012), Bollino (2009), Dimitropoulos
and Kontoleon (2009), Grösche and Schröder (2011), Kim et al. (2012), Zoric and
Hrovatin (2012), and Bigerna and Polinori (2013, 2014). The use of the family in
economic research is an open research question. Casimir and Tobi (2011) have
highlighted that in several branches of social sciences—such as economics, sociol-
ogy, and transportation—a theoretical or operational definition of the concept of a
household is seldom provided. Consequently, we have limited our research to papers
in which the respondent is an individual or the household is treated as a single unit.
Nevertheless, since publication of seminal papers by Portney (1994), Blamey et al.
(1995), and Sagoff (1998) it has been well known that another important source of
heterogeneity is related to the respondents according to the consumers versus
citizens’ dilemma, as underlined by Huijts et al. (2012). These authors investigate
this dilemma in term of energy technology acceptability, distinguishing between
citizen and consumer acceptance (Huijts et al. 2012, p. 526): in the first case, the
preferences involve being faced with a “siting decision”; in the second case, respon-
dents are mainly asked about the purchase of pre-existing technology services. We
use this distinction, which mainly reflects differences between the demand and
supply sides even where a distinction exists within the demand-side group (Batley
et al. 2001).
Apart from characteristics such as age, gender, education, environmental or
cultural association membership, and political attitude, we have also found papers
that analyze attitudes toward several types of environmental risks or peculiar aspects
referring to the history of the respondents and their families. Given that empirical
evidence confirms that newcomers generally differ from long-term residents regard-
ing both RE acceptance and WTP/WTA (Groothuis 2010), we take into account how
long interviewees and their ancestors have lived in the area, because this character-
istic matters.
The scenario proposed in the empirical analysis is also a factual characteristic.
Indeed, in several papers, respondents are informed about policy scenarios incorpo-
rating clear environmental targets (such as in Batley et al. (2000, 2001), Ivanova
(2005, 2012), Bergmann et al. (2006), Bollino (2009), Koundouri et al. (2009), Yoo
and Kwak (2009), Grösche and Schröder (2011), Kim et al. (2012) and Bigerna and
Polinori (2013, 2014).
CO2 emission reduction is the benefit associated with the change proposed to
respondents in almost the entirety of the papers collected. Only in a couple of cases
(Bergmann et al. 2006; Ek and Matti 2014) are additional job opportunities consid-
ered. A subgroup of papers (Álvarez-Farizo and Hanley 2002; Bergmann et al. 2006;
Groothuis 2010; Meyerhoff et al. 2010; Meyerhoff 2013) focus on the visual impact
of windmills in value areas, trying to explain the main determinants of the accep-
tance (or of the opposition), while other papers simply consider the economic
constraint of the RE development, especially on the demand side. Other important
characteristics refer to the impact on fauna and the landscape impact; however, the
NIMBY syndrome is the main issue considered in the literature.
According to Wolsink (2000, p. 57), the location, such as project characteristics,
is important in order to distinguish among different types of NIMBY syndrome
coherently. Several authors deal with this crucial aspect (see, among others, Navrud
and Bråten (2007)), also trying to quantify the impact of the institutional factor
and/or the role of the local community attitude toward windmills. Other authors
deeply investigate local communities, focusing on social relations and on residents
attachment to the territory.
Finally, important characteristics can be connected to the type of evaluated good.
The main characteristic considered on the demand side is related to the energy mix
used in RE generation (see, among others, Goett et al. (2000), Borchers et al. (2007),
and Kim et al. (2012)) if a generic RE is not proposed. The literature on the supply
side is characterized by greater heterogeneity. The characteristics involved are
related to the localization of the windmills (Bergmann et al. 2006; Strazzera et al.
2012); height of the towers and their numbers are also important characteristics
(Bergmann et al. 2006; Meyerhoff 2013), both related to the wind farm size. In some
cases the type of camouflage (Groothuis 2010) and institutions involved in the
realization and management of the wind turbines (Strazzera et al. 2012) are also
considered.
In conclusion, a large number of papers are taken into account in this chapter,
focusing on aspects useful for our local survey. Consequently, attention is paid to
papers in which a policy scenario, or a clear project, has been proposed; furthermore,
the local community perspective is highly represented in our review.
1.2 Methods 7
1.2.3 Topics of Interest
On October 23–24, 2014, the European Council approved new energy climate
objectives for 2030,2 increasing the CO2 reduction percentage up to 40%. In the
Umbria region, RE is mainly obtained by hydropower and photovoltaic power, and
both are close to their maximum capacity; consequently, to achieve these new targets
it is crucial to develop more wind power.3 Given that wind power could be strategic
in the Umbria region, it has been investigated on the supply side.
Drivers of public acceptance/opposition toward onshore wind farms are widely
investigated in the literature (Rowlands et al. 2003; Devine-Wright 2007; Jobert
et al. 2007; Cass and Walker 2009; Aitken 2010; Read et al. 2013; Stigka et al. 2014;
van Rensburg et al. 2015), underlining that four main interesting and relevant
questions for Umbria exist. Attitudes toward windmills are affected by the roles of
individuals and household characteristics (I). Also, amenity risk perception (II) and
local community attachment to the territory (III) influence these attitudes. Finally,
how pre-existing experience of windmills affects attitudes toward new windmill
projects is considered (IV).4 We highlight that in the literature this last topic has not
been investigated adequately and our findings can improve the literature.5
1.2.4 What the Literature Tells Us
In the literature the relationships between individual characteristics and RE accep-

tance are investigated on both market sides.
Gender is frequently investigated in empirical papers. The expectation is that
males and females value costs and benefits related to RE differently, given that
women are willing to pay less due to their lower income. Ivanova (2005), Bollino
(2009), and Bigerna and Polinori (2014) confirm this finding; Zhang and Wu (2012)
2
The new European Union (EU) targets involve (1) a reduction of 40% in greenhouse gas
emissions, with binding targets for Member States for non–Emissions Trading System (non-ETS)
sectors; (2) increasing the share of renewable energy sources (REnS) by 27% of final consumption
of energy without binding targets at the Member State level; and (3) a 27% increase in energy
efficiency, which could be revised with a rise to 30%.
3
This new scenario points out the relevance of knowledge of the determinants of RE acceptance/
opposition, on both the demand and supply sides, in designing new energy policy agenda. The
relevance is strengthened by the fact that consumers perceive RE as a clean and environmentally
friendly good even if establishment of REnS infrastructure meets strong local opposition regarding
siting processes, such as in wind energy, biomass, and large photovoltaic plant projects
(Wüstenhagen et al. 2007; Kaldellis et al. 2013).
4
This last topic is related to the impact of existing wind farms on the attitudes and preferences of
respondents.
5
In the meta-analysis, controlling for heterogeneity, we have also tried to take into account these
aspects whenever possible.
report that in the Jiangsu province of China, females are willing to contribute less,
especially as the amount of payment required increases. For several wind farm
attributes, Ek and Persson (2014) confirm gender differences whereas Koundouri
et al. (2009) do not. Finally, several researchers find gender differences on the
demand side, but nonsignificant results are reported by Borchers et al. (2007), Ek
and Söderholm (2008), and Zoric and Hrovatin (2012).
The age of the respondent is often analyzed, providing heterogeneous results. In
general, younger people are more likely to support RE on both sides (Ek 2005;
Ivanova 2005; Bergmann et al. 2006; Ek and Söderholm 2008; Ivanova 2012; Zoric
and Hrovatin 2012; Bigerna and Polinori 2014) even if the age of respondents may
be nonsignificant in some cases (Bollino 2009; Koundouri et al. 2009; Yoo and
Kwak 2009; Groothuis 2010; Zhang and Wu 2012).
Results can require more careful interpretation in some cases. Kim et al. (2012),
testing for several REnS, find that age negatively affects WTP only for hydropower.
Groothius et al. (2008) find that age does not affect WTA even where increases in
age reduce participation in a green program. Finally, Borchers et al. (2007) point out
that RE is preferred by both respondents aged over 50 and those aged under 30.
With regard to education, general agreement exists on its positive impact on both
sides (Koundouri et al. 2009; Ek 2005; Groothuis 2010; Ek and Persson 2014).
Nevertheless, some exceptions exist, such as in Ek and Söderholm (2008), where the
level of education is not significant, or in Kim et al. (2012) and Zhang and Wu
(2012), where higher level of education reduces elicited WTP. Finally, Groothius
et al. (2008) find that the required WTA is reduced by education.
Other individual characteristics have been investigated, such as income, house-
hold size, homeownership, environmental awareness, and knowledge of RE.
In Groothuis (2010) the sign for income is negative, suggesting that a change in
mountain views due to windmills is a normal good. Ek (2005) also finds a negative
relationship between income and support for RE due to occupational effects on
lower-income respondents. Nevertheless, income is mainly positively related to
WTP (or negatively related to WTA) for RE or increased participation in RE projects
(Borchers et al. 2007; Groothius et al. 2008; Bollino 2009; Yoo and Kwak 2009;
Grösche and Schröder 2011; Kim et al. 2012; Zoric and Hrovatin 2012; Bigerna and
Polinori 2014). Grösche and Schröder (2011), Ivanova (2005, 2012), Koundouri
et al. (2009), and Hanemann et al. (2011) find no significant relationship between
income and attitudes to RE, while Zhang and Wu (2012) point out that different
relationships exist between WTP and income according to bid levels proposed to the
respondents.
The family size characteristic negatively affects support for RE, as highlighted by
Koundouri et al. (2009), Zoric and Hrovatin (2012), and Bigerna and Polinori (2014).
Homeownership is sometimes investigated under the hypothesis that
homeowners are more involved in the proposed scenario because property is an
important asset. For example, homeowners may be more affected by electricity price
variation due to RE diffusion; on the other hand, the installation of new windmills
close to homeowners’ properties could reduce the value of their houses. Even if the
expectation is that homeownership reduces support, RE variables related to
1.2 Methods 9
homeownership are not significant (Bollino 2009; Zoric and Hrovatin 2012; Bigerna
and Polinori 2014).
Support for RE programs (Bollino 2009; Kim et al. 2012; Zoric and Hrovatin
2012; Bigerna and Polinori 2013, 2014) and participation in RE programs (Borchers
et al. 2007; Groothius et al. 2008) seem to be increased by environmental awareness
and knowledge of RE. Environmental awareness is taken into account as “amenity
risk perception” referring mainly to two environmental traits: CO2 concentrations
and visual intrusion. CO2 reduction has been commonly used to describe the benefits
due to RE diffusion, while the latter trait is the most important negative impact
associated with onshore wind farm projects. In Italy, Strazzera et al. (2012) confirm
that visual impact is crucial especially if residents involved in the projects have
experience of windmills, while Álvarez-Farizo and Hanley (2002) find that land-
scape impact is valuated less in monetary terms than impacts on flora and fauna.
Groothius et al. (2008) estimate a significant reduction in the acceptance of
electricity windmills due to mountain view intrusion, and Mirasgedis et al. (2014)
obtain a similar result in terms of WTP reduction to support siting of an onshore
wind farm. A deep analysis is conducted by Meyerhoff et al. (2010), pointing out
that distances from residential areas and small windmill size both contribute to
increased acceptance of wind projects, reducing landscape intrusion. These results
are confirmed by Meyerhoff (2013) using a spatial approach. Different results are
obtained by van Rensburg et al. (2015) in Ireland, analyzing 345 wind farm
applications and planning authority decisions. These authors point out that proximity
to residential areas does not affect planning outcomes; more important aspects
appear to be visual intrusion and the identity of the appellants. These results confirm
that amenity risk perception seems to be a central aspect in the siting process for
windmills especially if local communities are directly involved in term of environ-
mental externalities.
Several authors confirm the central role of local communities in the wind farm–
siting decision and in the development process. For example, Jobert et al. (2007)
point out that local actors affect acceptance of wind energy, and van Rensburg et al.
(2015) obtain the same result, focusing on local authority decisions. Read et al.
(2013) identify several local community attributes related to wind farm project
opposition, partially confirming the results reported by Jobert et al. (2007). Two
key variables that are able to jointly predict the opposition process are the presence
of community members able to influence local residents and the existence of past
opposition behaviors. Less relevant variables appear to be proximity, visual percep-
tion, and sense of place. Place attachment is also analyzed by Groothius et al. (2008),
taking into account how long respondents have lived in the place involved in the
wind project. Their results point out that respondents whose ancestors have lived in
the area require higher compensation to accept windmills. Strazzera et al. (2012)
obtain similar results.
A few studies analyze the impact of respondents’ experience of a wind farm on
local acceptance, such as those by Koundouri et al. (2009), Meyerhoff et al. (2010),
Strazzera et al. (2012), and Meyerhoff (2013), which explicitly take into account this
aspect, referring to an onshore wind farm, and confirm that respondents’ experience
of windmills influences welfare measures. Indeed, Strazzera et al. (2012) emphasize

that those residents in a municipality who have experience of a wind farm devote
more attention to the technical characteristics of the new wind energy project, and
Meyerhoff et al. (2010) and Meyerhoff (2013) find that respondents who live in
proximity to windmills are more willing to accept new wind projects.
1.2.5 Quantitative Analysis: Metaregressions
WTA and WTP are neoclassical welfare measures used for policy analysis. In detail,
WTA is the amount of money that would make an individual indifferent between a
given amount of good and a lesser amount of the same good plus monetary
compensation. Conversely, WTP is the amount of money that would make an
individual indifferent between a lesser amount of good and a greater amount of
good without money compensation. Formally, we use an indirect utility function v
defined over a vector of goods X purchased at prices p, a rationed good RG (in this
case, RE), and income M (Bigerna and Polinori 2014); consequently, we can obtain
WTP and WTA by solving the following equations:
vðM; X; RG0 Þ ¼ vðM WTP; X; RG1 Þ ð1:1aÞ

vðM; X; RG1 Þ ¼ vðM þ WTA; X; RG0 Þ ð1:1bÞ
The WTP/WTA for a marginal increase in production of RG is defined as the ratio

of the marginal utility of the production of RG and the purchase price of property
attributes:

∂U=∂RG
WTP=WTA ¼ ð1:2Þ
∂RG=∂P
We assume that U is a linear stochastic utility function for the individual

including the production of RG, the purchase price P, and a vector of other variables.
We must therefore define welfare measures as the ratio of the coefficient of the
production of RG, βRG, and the purchase price βP:

βRG
WTP=WTA ¼ ð1:3Þ
βP
We estimate these parameters using information drawn from the existing litera-
ture. The estimates of WTP/WTA depend on the reference level used for the
calculation,6 but a measure that circumvents this problem is compensating variation,
6
In the literature, these reference levels are usually calculated as the average level of production of
RE and the purchase price used in the survey. Reference levels can vary considerably between
primary studies, so comparison of WTP/WTA values obtained using different utility function
specifications can be difficult and puzzling.
1.2 Methods 11
which reflects the change in the economic welfare of an individual caused by a

change in the level of consumption of a good. In other words, compensating
variation is the monetary compensation necessary after a change in the level of the
good consumed to restore the same level of utility that the individual had before the
occurrence of the changes. By assuming a linear utility function in the production of
RG, the compensating variation is simply equal to the product of the marginal
WTP/WTA associated with the change proposed.7
Following Borenstein et al. (2007), Nelson and Kennedy (2009), and
Dimitropoulos et al. (2013), we have formulated Eqs. (1.4a) and (1.4b), which
illustrate the metaregression models used to explain the variation of the logarithm
of the estimates of WTP and WTA.
WTP equation for RE
XA
ln ðWTPn Þ ¼ α þ βRG ln ðRGn Þ þ δ PU a, n
a¼1 PU a
XB
þ δ FAC b, n ð1:4aÞ
b¼1 FAC b
XC
þ δ
c¼1 MET c
MET c, n þ En
In these equations, ln (RGn) is the natural logarithm of the average level of

consumption of RGn reported in, or inferred by, the n-primary study. PUn is a vector
of A dummy variables indicating whether the primary study is published in a peer-
reviewed journal and if the period of publication is after 2007 (in particular, we focus
on pre- or postcrisis). FACn denotes a vector of B dummy variables related to factual
heterogeneity and indicating the location of the study, the different composition of
REnS used in electricity generation, the type of respondent, whether environmental
awareness and knowledge of REnS are verified, and whether a clear policy target is
proposed to interviewees. METn is a vector of C dummy variables related to
methodological heterogeneity such as the elicitation format, survey type, and treat-
ment of uncertainty. Finally, εn is an error term with mean zero and variance σ n.
WTP/WTA equation for an onshore wind farm
XA
ln ðWTAn =WTPn Þ ¼ α þ βRG ln ðRGn Þ þ δ PU a, n
a¼1 PU a
XB
þ δ FAC b, n ð1:4bÞ
b¼1 FAC b
XC
þ δ
c¼1 MET c
MET c, n þ En
Appropriately transforming Eq. (1.4a) it is possible to analyze WTP/WTA for an

onshore wind farm–siting process. The previous model has been extended, including
7
For each study, we take into account, as far as possible, the initial and the final value of
rationed good, RG0 and RG1 respectively. The change in rationed good (ΔRG) objective of the
environmental policy.
in the FACn vector some technological characteristics such as the distance of the
windmills, the installed capacity, and if visual intrusion is considered. Furthermore,
we have embedded the type of opposition and other characteristics related to the
local community. Finally, the welfare metrics used are also considered in the
methodological category.
1.2.6 Qualitative Analysis: Local Survey
We have interviewed individuals who live in a municipality close to an onshore wind

farm to investigate their attitude both toward RE in general and to the new wind
project. The aim is twofold. First we want to evaluate the acceptability of improve-
ment of an existing wind farm; then we want to appraise the support that inter-
viewees are willing to provide to RE production. Operationally the survey8 was
conducted by the end of June 2012 and the respondents addressed a development
project of an existing onshore wind farm.9 Face-to-face interviews were conducted
by the authors using a subsample population that was representative in terms of
demographic features according to the 2011 population census.
The questionnaire includes three sections. Initially, we test respondents’ specific
knowledge of REnS and sustainability. In the second section, we investigate respon-
dents’ attitudes toward potential benefits and environmental damage associated with
the project.
The CV scenario includes the possibility for respondents to state if they want
(or do not want) to support the project, accounting for their choice. In the first case,
WTP is asked for, while in the second case, WTA is required.10 In accordance with
Lunander (1998) and Frew et al. (2004), among several elicitation formats, we adopt
the OE contingent valuation as providing the most prudent estimates of “true”
economic value. The last section of the questionnaire is aimed at measuring respon-
dents’ sociodemographic and economic characteristics, ensuring a means for us to
investigate how socioeconomic factors influence WTP and WTA.
8
This survey is included in more wide CV studies in which the monetary evaluation is obtained
through the development of a hypothetical market. We use the results of the preliminary survey to
test the questionnaire.
9
The project involves the installation of 16 towers for wind generation of electricity. Precisely, the
plan should provide for the installation of four towers reaching 40–60 m in Pian di Spilli (in the
municipality of Costacciaro) and in Val di Ronco (in the municipality of Sigillo), and eight similar
towers in the municipalities of Scheggia Pascelupo and Fossato di Vico. Each tower will have a
maximum power of 1 MW.
10
Initially, respondents are asked if they perceive the project as positive and consequently if they
want to support the project (i.e., WTP). Otherwise respondents can declare their opposition to the
project, due to their negative perception, and consequently they are asked if they are willing to
accept monetary compensation for the project (WTA).
1.3 Results and Discussion 13
1.3 Results and Discussion
A metaregression analysis is estimated11 for WTP/WTA related to RE, considering

both the demand and supply sides. We pay attention to the impact of different
empirical methods used, rather than only providing a summary of the main results
available in the literature, and we classify all of the variables, considering if they are
calculated (I) or are related to the publication process (II) or are ascribable to the
factual (III) or methodological (IV) category.
To enhance comparability of the statistical information the original data collected
are converted into kilowatt hours consumed or produced.12 To implement the full
comparative assessment of the studies, the estimated values of WTP and WTA are
converted to 2013 EUR with compensation for income differences.13 Finally, the
other three categories simply refer to different sources of heterogeneity described
earlier in this chapter.
1.3.1 Willingness to Pay for Renewable Electricity

on the Demand Side
Table 1.1 shows the empirical results obtained using studies related to WTP for RE
on the demand side.
A negative effect on WTP caused by the level of total RE14 consumption
emerges, while the first group of variables differently affect the left hand side
(LHS) because while the period of publication is not significant, papers published
in International Scientific Indexing (ISI)–indexed journals positively affect WTP.
11
Both equations are estimated using unweighted and weighted least squares estimators. In partic-
ular, the weighted ordinary least squares (wOLS) estimator is superior to the conventional random
effects estimator when the meta-analysis refers to a small sample (Stanley and Doucouliagos 2013),
such as in this chapter. We have reduced selection distortion, using, as far as possible, published
papers and working papers, by correcting for heteroscedasticity. Finally, we want to underline that
in both models we have used a log-linear specification because transformed data are less sensitive to
the problem of heteroscedasticity.
12
In other words, both WTP for RE households’ consumption and WTP/WTA for wind farm
production refer to kilowatt hours.
13
Consumption data are available from the World Energy Council website (http://www.wec-
indicators.enerdata.eu/thermal-electricity-use.html). For the UK, additional information is available
from the Department of Energy and Climate Change website (https://www.gov.uk/government/
uploads/system/uploads/attachment_data/file/65940/7341-quarterly-energy-prices-december-2012.
pdf). For Italy, additional information is available from the Terna (http://www.terna.it/) and Gestore
dei Servizi Energetici (GSE) (http://www.gse.it) websites. Information on capacity factors is
gathered by the websites https://www.eia.gov and https://community.ieawind.org. Finally, infor-
mation on exchange rates and deflators is gathered by the websites https://www.bloomberg.com and
https://www.oanda.com.
14
RE consumption is expressed in logarithm.
Table 1.1 Metaregression: willingness to pay (WTP) for renewable electricity (RE)
RHS uOLS wOLS
Calculated ln(RE) 0.804a 0.611a
0.058 0.046
Publication PU: After 2007 NS NS
PU: ISI journal 0.111b 0.051c
0.053 0.033
Factual FAC: Europe 0.515b 0.521b
0.122 0.037
FAC: USA 0.532c 0.299c
0.299 0.087
FAC: Wind 0.622a 0.659a
0.077 0.061
FAC: Green generic NS NS
FAC: Individual NS NS
FAC: Environmental awareness 0.045b 0.032c
0.021 0.019
FAC: Knowledge of RE 0.025b 0.015c
0.011 0.009
FAC: Specified target NS NS
Methodological MET: Payment card 0.491a 0.467a
0.034 0.034
MET: Choice experiment NS NS
MET: Double bound DC NS NS
MET: Face to face NS NS
MET: Uncertainty 0.335b 0.387b
0.148 0.122
MET: Small sample 0.276b 0.347a
0.098 0.045
Const 0.457a 0.502a
0.054 0.071
Standard error values are italicized
Obs 189; R-sq adj: uOLS ¼ 0.753, wOLS ¼ 0.835; RMSE: uOLS ¼ 0.324, wOLS ¼ 0.232; Prob F:
uOLS ¼ 1.5 E04, wOLS ¼ 2.4 E04
Const constant, DC dichotomous choice, ln(RE) logarithm of RE consumption, NS not significant,
Obs observations, Prob F F-value probability, RHS right hand side, RMSE root-mean-square error,
R-sq adj adjusted R-squared, uOLS unweighted ordinary least squares, wOLS weighted ordinary
least squares
a
Coefficient that is significant at the 1% level
b
c
1.3.1.1 Factual Variables
A negative effect is evident in studies done in the USA, while the effect is positive in
European studies. Studies that differentiate the renewable energy mix negatively
affect WTP when wind energy is taken into account. These are interesting results
because they confirm that wind energy is perhaps seen by the public as a potentially
environmentally dangerous REnS. Results related to individual characteristics are
also interesting, given that studies in which environmental awareness is taken into
account positively affect WTP for RE.
Furthermore, a smaller positive effect is also due to “knowledge of REnS,”
confirming that providing information on RE to consumers increases the acceptabil-
ity of RE. Surprisingly, the parameter associated with the policy target is not
significant. This means that the definition of a clear environmental target does not
affect the WTP of respondents.15
1.3.1.2 Methodological Variables
Controlling for methodological aspects we find that a negative effect is due to studies
that use a PC or small samples, even if the results are robust when using both normal
and weighed ordinary least squares (wOLS). Finally, studies that take into account
uncertainty contribute to increased WTP on the demand side.
1.3.2 Willingness to Accept and Willingness to Pay

for Renewable Electricity on the Supply Side
On the supply side we use studies in which various externalities related to an onshore
wind farm are considered using both WTP and WTA measures. The results are
illustrated in Table 1.2.
Also, in this case a negative effect on WTP caused by the level of total RE
production is significant. In this model both the period of publication and the type of
publication are not significant.
1.3.2.1 Factual Variables
Beyond the confirmation of the positive effects characterizing the European studies,
other interesting results related to other factual characteristics emerge. First, the
parameter estimates confirm that distance increases WTP for RE produced by wind.
Indeed, citizens are willing to pay to move windmills far away and this means that
the increase in distance positively affects the attitude toward wind RE generation
even if this is an open question in the “proximity versus reverse proximity” debate.
The parameter estimate related to installed capacity is significant and negative
(suggesting that people are willing to pay less if energy projects include a large
wind farm), strengthening the literature results.
15
A possible explanation is that current environmental policy uses too many strategies and
consequently environmental targets are not clear to the citizens who perceive a lack of policy
efficacy (European Environmental Bureau 2010).
Table 1.2 Metaregression: willingness to pay/willingness to accept (WTP/WTA) for an onshore

wind farm
RHS uOLS wOLS
Calculated ln(RE) 0.361a 0.305a
0.031 0.023
Publication PU: After 2007 NS NS
PU: ISI journal NS NS
Factual FAC: Europe 0.253b 0.259b
0.056 0.021
FAC: NIMBY 0.912b 0.969a
0.023 0.015
FAC: Distance 0.832a 0.818a
0.003 0.005
FAC: Installed capacity 0.053c 0.077c
0.041 0.034
FAC: Visual intrusion 0.031c 0.027c
0.021 0.020
FAC: Individual NS NS
FAC: Environmental awareness 0.108c 0.097c
0.075 0.052
FAC: Knowledge of RE NS NS
FAC: Experience of WF 0.081b 0.075b
0.033 0.030
FAC: Local community NS NS
FAC: Scenario NS NS
Methodological MET: Choice experiment 0.473a 0.406a
0.038 0.032
MET: Face to face 0.863a 0.925a
0.008 0.008
MET: WTA 0.830b 0.617b
0.012 0.018
MET: Small sample 0.130b 0.183a
0.046 0.029
Const 0.220a 0.247a
0.030 0.042
Obs 119; R-sq adj: uOLS ¼ 0.604, wOLS ¼ 0.671; RMSE: uOLS ¼ 0.392, wOLS ¼ 0.344; Prob F:
uOLS ¼ 1.3 E04, wOLS ¼ 1.5 E04
Const constant, ln(RE) logarithm of RE consumption, NIMBY “Not in My Back Yard”, NS not
significant, Obs observations, Prob F F-value probability, RE renewable electricity, RHS right hand
side, RMSE root-mean-square error, R-sq adj adjusted R-squared, uOLS unweighted ordinary least
squares, WF wind farm, wOLS weighted ordinary least squares
a
b
c
The environmental impact of wind projects is analyzed using several factual

variables. First, the results confirm that NIMBY syndrome16 consistently reduces
RE acceptance and that visual intrusion due to the windmills has the same effects
even if its parameter estimate is smaller.
An interesting result is that the parameter associated with knowledge of REnS is
not significant on the supply side, suggesting that information on REnS could be not
decisive in increasing the acceptability of RE in the context of investment projects.
The parameter associated with the “environmental awareness” dummy variable is
greater and becomes negative in this model with respect to the parameter estimated
on the demand side. This means that environmentally negative externalities are
perceived more intensely when the respondents and their community are directly
involved in electricity generation. This result is strengthened by the fact that “visual
intrusion” is estimated separately in this model.
Dummy variables related to the local community are not significant. These results
are not surprising, for at least two reasons: first because only a few studies have been
conducted at the local level, and second because the local community needs a more
detailed specification than a simple generic dummy variable—a specification that we
introduced in our local survey.
The variable related to the policy scenario is not significant; we deem that this is
for the same reason underlined in the demand-side model. Finally, our estimation
confirms that studies conducted in areas where respondents have experienced wind-
mills positively affect WTP for RE. In other words, it seems that citizens exposed to
windmills for medium or long periods are more supportive toward wind power
development.
1.3.2.2 Methodological Variables
First, CEx negatively affects WTP for RE production using wind energy. This can be
explained considering that the CEx approach produces more conservative WTP
estimates. Furthermore, according to the theoretical expectations, the use of WTA
produces an increase in the amounts elicited, consistent with the fact that this metric
is not in any way bound to the income of the respondents. Finally, we see that small-
sample studies tend to provide lower WTP even if a wOLS estimator has been used.
1.3.3 Local Survey Results
We compare WTP and WTA for kilowatt hours of RE using the results of the local
survey. From the first section of the questionnaire it emerges that all of the respon-
dents know of both fossil energy and REnS. Among nonrenewable energy sources
16
The type of resistance is dichotomized into a variable, which takes the value one if it is NIMBY
syndrome and zero otherwise.
Table 1.3 Subsample descriptive statistics

Variable Type Abbreviation Unit Supportersa Opponentsb
Bimonthly electricity bill Continuous bill EUR 73.062 72.232
9.256 9.210
Head of family age Continuous agehf NR 54.030 48.595
19.771 14.72
Gender Dummy sex # 0.455 0.486
(1 ¼ female) 0.505 0.506
Family years of residence Continuous resanni NR 41.545 33.297
(ancestors included) 21.21 18.65
Household members Continuous ncomp NR 2.939 2.757
1.058 0.954
Household income Continuous incom EUR 2.781 2.500
(10,000) 1.156 1.001
Leftist party affinity Dummy leftp # 0.697 0.189
(1 ¼ yes) 0.466 0.397
Amenity perception Scale (1–10) amenity NR 4.909 6.432
(10 ¼ max) 1.588 1.993
Years of education Continuous educy NR 9.545 11.541
4.024 4.226
Environmental association Dummy assambcc # 0.091 0.189
membership (1 ¼ yes) 0.291 0.397
Homeowner Dummy homeow # 0.788 0.730
(1 ¼ yes) 0.415 0.450
Positive attitude to wind Dummy wfexp # 0.212 0.135
farm project in 1999 (1 ¼ yes) 0.415 0.346
Mean WTP Continuous mWTP EUR 7.003
8.797
Mean WTA Continuous mWTA EUR 7.935
9.980
max maximum, NR not reported, WTA willingness to accept, WTP willingness to pay
a
103 observations
b
107 observations
the most widely known are methane (57%), while wind (64%) is the best known
among renewable sources. The majority of respondents (75%) have stated that they
are aware of both positive and negative forms of environmental impact associated
with electricity production from renewable and conventional sources.
Table 1.3 shows that respondents who support wind farm development projects
and perceive positive externalities make up almost the same percentage as those who
oppose the same project because of perceived local negative externalities; further-
more, the two types of respondents are quite similar in terms of sociodemographic
and economic characteristics. Down in the details it arises that both groups pay a bill
of around EUR 70 bimonthly and have declared an annual income that is
approximately EUR 26,000, with a difference smaller than 10%, and in both cases
more than 78% of respondents are homeowners. The households’ characteristics
confirm their relevance. First, the group that requires compensation for the project
includes families that have been resident long term and have a younger head of the
family than the respondents in the other group, while the number of components is
similar in the two groups.
With regard to individual respondents’ characteristics, these sometimes differ
consistently between the two groups. In particular, respondents are less educated in
the supportive group, which also includes fewer members of environmental associ-
ations but more leftist members. The characteristics are quite similar only in terms of
gender.
In comparison with the other group, we find that respondents who are willing to
pay for the project believe that amenity does not characterize their dwelling place;
this group also was more supportive than other respondents toward the first onshore
wind farm project in 1999. As expected, the mean WTA is significantly higher than
the mean WTP, while the median values are more similar. The difference between
WTA and WTP is still positive and statistically significant.
Figure 1.1 shows the kernel density estimation of the WTP and WTA distribu-
tions, showing the skewness and the kurtosis of the distributions; WTA also has a
higher maximum value, with a higher density for EUR 20 and EUR 35.
Partial correlations among WTP, WTA, and sociodemographic and economic
variables are shown in Table 1.4. We emphasize that this type of analysis does not
allow an explanation of causality; it only investigates whether a linear relationship
exists among these variables. Females exhibit a higher WTA for the project; this is
consistent with the findings of some recent studies (see, among others, Kontogianni
et al. (2014)). Amenity perception and education are positively related to WTA;
furthermore, how people perceive the place where they live and WTA are closely
related, confirming the existing related literature. This literature has mainly investi-
gated negative perceptions due to a wind farm (see, among others, Ladenburg
Fig. 1.1 WTP and WTA

kernel density estimation
Table 1.4 Partial correlations

Variable WTP WTA
Bimonthly electricity bill 0.1573 0.1791
Head of family age 0.2135 0.3040
Gender (1 ¼ female) 0.2561 0.3926a
Family years of residence (ancestors included) 0.4266a 0.4806a
Household members 0.1892a 0.3276a
Household income (10,000) 0.3293a 0.3936a
Leftist party affinity (1 ¼ yes) 0.0859 0.0399
Amenity perception 0.5690a 0.4957a
Years of education 0.4833a 0.3671a
Environmental association membership (1 ¼ yes) 0.1510 0.4827a
Homeowner (1 ¼ yes) 0.2406 0.2637
Positive attitude to wind farm project in 1999 (1 ¼ yes) 0.3995a 0.1018
a
Significant at 5%
WTA willingness to accept, WTP willingness to pay
(2008), Krueger et al. (2011), and Ladenburg and Lutzeyer (2012)) instead of direct
amenity perception.
Table 1.4 also illustrates that several family characteristics are positively corre-
lated with WTA. The first concerns how long the households of respondents have
resided in the area. The literature on “place attachment” and on “sense of place”
describes the importance of this attribute (Devine-Wright 2005; Devine-Wright
2009) for determining NIMBY syndrome, while Groothius et al. (2008) underline
that respondents whose ancestors have lived in the area involved in the wind farm
project exhibit higher WTA. We also confirm this last result.
Education is positively related to higher compensation. This result is in contrast to
that reported by Groothius et al. (2008) but consistent with that reported by Landry
et al. (2012). According to the existing literature, household size and income are
correlated positively and negatively (Groothius et al. 2008) with WTA, respectively.
Finally, WTA is positively associated with environmental association member-
ship. This result seems logical and consistent with expectations; it is not easily
comparable with the literature, given that, to the best of our knowledge, no other
studies have used this variable. Among respondents who have declared themselves
to be in agreement with the project, those who had a positive attitude toward the
wind farm project in 1999 exhibit higher WTP; this confirms that prior experiences
matter (Wosink 2007; Ladenburg 2009; Meyerhoff 2013). The same expected
relationship arises between WTP and household income, between WTP and how
long the family and ancestors have lived in the place (Groothius et al. 2008), and
between WTP and the educational level of respondents. A negative correlation is
obtained between WTP and household size and between WTP and amenity.
Integrating a meta-analysis with a local survey, we have investigated the main
determinants of RE acceptance and development, taking into account several expli-
cative variables categorized as factual variables. We have paid particular attention to
1.4 Conclusions 21
the environmental externalities and how individuals and local community perceive
them, given that perceptions impact on citizens’ and consumers’ preferences and
behaviors when both deal with RE acceptance. Some useful insights emerge.
Environmental awareness differently affects the two market sides considered.
Indeed, it increases WTP for RE on the demand side but opposite results arise on the
supply side, confirming that externalities are more strongly perceived if respondents
are directly involved in the project. Both meta-analysis and survey results support
this hypothesis, given that we have controlled for visual intrusion and for exposure to
windmills. The impact of knowledge of REnS is clearly asymmetrical. Indeed, it
impacts only on consumers’ behavior increasing WTP for RE, but this variable does
not matter in the local context if a wind investment project is proposed. This means
that the information is not sufficient to mitigate stronger perceptions of negative
externalities. Even if information on wind energy does not increase RE acceptance,
our results suggest that to increase acceptability of wind power, new wind invest-
ment should take place near a previously established wind farm, confirming the
dynamism of preference toward wind farms. Indeed, Wosink (2007) points out that
preferences change following a U-shaped scheme. Attitudes are very positive in two
particular periods: when citizens are not confronted with wind farms, and at some
reasonable time after the construction of wind farms. In our cases, this result is also
indirectly confirmed by the fact that people who had a positive attitude in 1999 are
still positive toward the new project.
1.4 Conclusions
The global reduction of CO2 emissions needs appropriate policies supported by

adequate incentive mechanisms. Furthermore, a considerable cost is incurred in
achieving an appreciable result. A cultural change is also needed because of the
different behaviors that characterize individuals facing renewable electricity
(RE) consumption and production. Indeed, if consumers state high support for RE,
citizens exhibit low acceptance of new infrastructure for RE production when this
project involves their community. Consequently, harmonization of these opposing
behaviors is required.
In this chapter we have first analyzed the existing RE-related literature by use of
meta-analysis and then we have conducted a local survey involving a community
that has experienced wind energy production and is faced with a new wind invest-
ment project.
Three main results emerge from the metaregression analysis. Consumers are
willing to support RE when they are end users but are less supportive if a new
project is proposed. Second, distance is an important attribute in the windmill-siting
process. Third, the local dimension is relevant, given that at the local level the
perception of technology is stronger in comparison with the supralocal scale.
These results support our choice of conducting a survey emphasizing the local
dimension to investigate citizens’ and consumers preferences’ in the local
community.
This integrated analysis confirms literature results in terms of individual charac-
teristics such as education (which positively affects RE acceptance on two market
sides) and age (which negatively influences both consumers’ and citizens’ attitudes),
but some interesting insights have also been pointed out.
First, the confirmation of Wosink (2007) suggests that past experience of
implemented wind investment projects could be a useful tool for an institutional
communication strategy. Furthermore, in the siting decision process, communities
already involved in wind projects seem to be an ideal counterpart in the bargaining
process due their pragmatism and also because they focus more on technical and
institutional aspects that on “ideological questions.” However, the local dimension
strengthens the perception of negative externalities, so careful planning is required in
the definition of wind investment projects, taking into account both technical
characteristics and environmental impact. In conclusion, our research results suggest
that sociological and psychological aspects should be included in future economic
research to facilitate new installation of onshore wind farms by reducing the related
infighting. On the other hand, policy agenda should include this type of research in
the toolbox for at least three reasons: first, to better understand wind investment
project outcomes; second, to guarantee active participation of the local community in
the decision process; and third, because the combined effects of the first two reasons
allow the probability of positive outcomes of this type of project to be increased.
References
Aitken M (2010) Wind power and community benefits: challenges and opportunities. Energy Policy
38:6066–6075
Álvarez-Farizo B, Hanley N (2002) Using conjoint analysis to quantify public preferences over the
environmental impacts of wind farms: an example from Spain. Energy Policy 30:107–116
Batley SL, Fleming PD, Urwin O (2000) Willingness to pay for renewable energy: implications for
UK green tariff offerings. Indoor Built Environ 9:157–170
Batley SL, Colbourne D, Fleming PD, Urwin O (2001) Citizen versus consumer: challenges in the
UK green power market. Energy Policy 29:479–487
Bergmann A, Hanley N, Wright R (2006) Valuing the attributes of renewable energy investments.
Energy Policy 34:1004–1014
Bigerna S, Polinori P (2013) A bidding game for Italian households’ WTP for RES. Atl Econ J
41:189–190
Bigerna S, Polinori P (2014) Italian households’ willingness to pay for green electricity. Renew Sust
Energ Rev 34:110–121
Blamey R, Common M, Quiggin J (1995) Respondents to contingent valuation surveys: consumers
or citizens. Aust J Agric Econ 39:263–288
Bollino CA (2009) The willingness to pay for renewable energy sources: the case of Italy with socio
demographic determinants. Energy J 30:81–96
Borchers AM, Dukea JM, Parsons GR (2007) Does willingness to pay for green energy differs by
source? Energy Policy 35:3327–3334
References 23
Borenstein M, Hedges L, Rothstein H (2007) Introduction to meta-analysis, 1st edn. www.Meta-

analysis.com. Accessed 24 Sep 2014
Botetzagias I, Malesios C, Kolokotroni A, Moysiadis Y (2013) The role of NIMBY in opposing the
siting of wind farms: evidence from Greece. J Environ Plan Manag 58:252–269
Byrnes B, Jones C, Goodman S (1999) Contingent valuation and real economic commitments:
evidence from electric utility green pricing programs. J Environ Plan Manag 42:149–166
Casimir GJ, Tobi H (2011) Defining and using the concept of household:a systematic review. Int J
Consum Stud 35:498–506
Cass N, Walker G (2009) Emotion and rationality: the characterization and evaluation of opposition
to renewable energy projects. Emot Space Soc 2:62–69
Devine-Wright P (2005) Beyond NIMBYism: towards an integrated framework for understanding
public perceptions of wind energy. Wind Energy 8:125–139
Devine-Wright P (2007) Reconsidering public attitudes and public acceptance of renewable energy
technologies: a critical review. Research Council Energy Programme—E.S.R.C. 2007, W.P. n.4
Devine-Wright P (2009) Rethinking NIMBYism. J Community Appl Soc Psychol 19:426–441
Dimitropoulos A, Kontoleon A (2009) Assessing the determinants of local acceptability of wind-
farm investment: a choice experiment in the Greek Aegean Islands. Energy Policy
37:1842–1854
Dimitropoulos A, Rietveld P, van Ommeren JN (2013) Consumer valuation of changes in driving
range: a meta-analysis. Trans Res Part A Policy Pract 55:27–45
Edmonson AC, McManus SE (2007) Methodological fit in management field research. Acad
Manag Rev 32:1115–1179
Ek K (2005) Public and private attitudes towards “green” electricity: the case of Swedish wind
power. Energy Policy 33:1677–1689
Ek K, Matti S (2014) Valuing the local impacts of a large scale wind power establishment in
northern Sweden: public and private preferences toward economic, environmental and socio-
cultural values. J Environ Plan Manag 58:1327–1345
Ek K, Persson L (2014) Wind farms—where and how to place them? A choice experiment approach
to measure consumer preferences for characteristics of wind farm establishments in Sweden.
Ecol Econ 105:193–203
Ek K, Söderholm P (2008) Norms and economic motivation in the Swedish green electricity
market. Ecol Econ 68:169–182
European Environmental Bureau (EEB) (2010) Future of EU environmental policy. Towards the
7th environmental action programme—sustainability. Annual conference report
Frew EJ, Wolstenholme JL, Whynes DK (2004) Comparing willingness-to-pay: bidding game
format versus open-ended and payment scale formats. Health Policy 68:289–298
Goett AA, Hudson K, Train KE (2000) Customers’ choice among retail energy suppliers: the
willingness to pay for service attributes. Energy J 4:1–28
Groothius PA, Groothuis JD, Whitehead JC (2008) Green vs. green: measuring the compensation
required to site electrical generation windmills in a viewshed. Energy Policy 36:1545–1550
Groothuis PA (2010) Land use issues: the last settler’s syndrome. J Agric Appl Econ 42:357–365
Grösche P, Schröder C (2011) Eliciting public support for greening the electricity mix using random
parameter techniques. Energy Econ 33:363–370
Hanemann M, Labandeira X, Loureiro ML (2011) Climate change, energy and social preferences
on policies: exploratory evidence for Spain. Clim Res 48:343–348
Hanley N, MacMillan D, Wright RE, Bullock C, Simpson I, Parsisson D, Crabtree B (1998)
Contingent valuation versus choice experiment: estimating benefits of environmentally sensitive
areas in Scotland. J Agric Econ 49:1–15
Hansla A, Gamble A, Juliusson A, Garling T (2008) Psychological determinants of attitude towards
and willingness to pay for green electricity. Energy Policy 36:768–774
Herbse C, Frienge C (eds) (2017) Marketing renewable energy—concepts, business models and
cases. Springer, Cham
Holmgren J (2007) Meta-analysis of public transport demand. Trans Res Part A Policy Pract
41:1021–1035
Huijts NMA, Molin EJE, Steg L (2012) Psychological factors influencing sustainable energy
technology acceptance: a review-based comprehensive framework. Renew Sust Energ Rev
16:525–531
Ivanova G (2005) Queensland consumers’ willingness to pay for electricity from renewable energy
sources. In: Proceedings of the Ecological Economics in Action conference, Palmerston North,
11–12 Dec 2005
Ivanova G (2012) Are consumers’ willingness to pay extra for the electricity from renewable energy
sources. An example of Queensland Australia. Int J Renew Res 42:758–766
Jobert A, Laborgne P, Mimler S (2007) Local acceptance of wind energy: factors of success
identified in French and German case studies. Energy Policy 35:2751–2760
Kaldellis JK, Kapsali M, Kaldelli E, Katsanou E (2013) Comparing recent views of public attitude
on wind energy, photovoltaic and small hydro applications. Renew Energy 52:197–208
Kim J, Park J, Kim H, Heo E (2012) Assessment of Korean customers’ willingness to pay with RPS.
Renew Sust Energ Rev 16:695–703
Kontogianni A, Tourkolias C, Skourtos M, Damigos D (2014) Planning globally, protesting locally:
patterns in community perceptions towards the installation of wind farms. Renew Energy
66:170–177
Koundouri P, Kountouris Y, Remoundou K (2009) Valuing a wind farm construction: a contingent
valuation study in Greece. Energy Policy 37:1939–1944
Krueger AD, Parsons GR, Firestone J (2011) Valuing the visual disamenity of offshore wind power
projects at varying distances from the shore: an application on the Delaware shoreline. Land
Econ 87:268–283
Ladenburg J (2008) Attitudes towards on-land and offshore wind power development in Denmark;
choice of development strategy. Renew Energy 33:111–118
Ladenburg J (2009) Visual impact assessment of offshore wind farms and prior experience. Appl
Energy 86:380–387
Ladenburg J, Lutzeyer S (2012) The economics of visual disamenity reductions of offshore wind
farms—review and suggestions from an emerging field. Renew Sust Energ Rev 16:6793–6802
Lancaster KJ (1966) A new approach to consumer theory. J Polit Econ 74:132–157
Landry CE, Allen T, Cherry T, Whitehead JC (2012) Wind turbine and coastal recreation demand.
Resour Energy Econ 34:93–111
Lunander A (1998) Inducing incentives to understate and to overstate willingness to pay within the
open-ended and the dichotomous-choice elicitation formats: an experimental study. J Environ
Econ Manag 35:88–102
Meyerhoff J (2013) Do turbines in the vicinity of respondents’ residences influence choices among
programs for future wind power generation? J Choice Model 7:58–71
Meyerhoff J, Ohl C, Hartje V (2010) Landscape externalities from onshore wind power. Energy
Policy 38:82–92
Mirasgedis S, Tourkolias C, Tzovla E, Diakoulaki D (2014) Valuing the visual impact of wind
farms: an application in south Evia, Greece. Renew Sustain Energy Rev 39:296–311
Mu Y, Mu X (2013) Energy conservation in the Earth’s crust and climate change. J Air Waste
Manag Assoc 63:150–160
Navrud S, Bråten KG (2007) Consumers’ preferences for green and brown electricity: a choice
modeling approach. Rev Econ Politiq 117:795–811
Nelson JP, Kennedy PE (2009) The use (and abuse) of meta-analysis in environmental and natural
resource economics: an assessment. Environ Resour Econ 42:345–377
Nomura N, Akay M (2004) WTP for green electricity in Japan as estimated through contingent
valuation method. Appl Energy 78:453–463
Portney PR (1994) The contingent valuation debate: why economists should care. J Econ Perspect
8:3–17
References 25
Read DL, Brown RF, Thorsteinsson EB, Morgan M, Price I (2013) The theory of planned behavior
as a model for predicting public opposition to wind farm development. J Environ Psychol
36:70–76
Rowlands IH, Scott D, Parker P (2003) Consumers and green electricity: profiling potential
purchasers. Bus Strategy Environ 12:36–48
Sagoff M (1998) Aggregation and deliberation in valuing environmental public goods: a look
beyond contingent pricing. Ecol Econ 24:213–230
Stanley TD, Doucouliagos H (2013) Better than random: weighted least squares meta-regression
analysis. School Working Paper—Economic Series—Deakin University, 2/2013. http://www.
deakin.edu.au/buslaw/aef/workingpapers/papers/2013_2.pdf. Accessed 14 Feb 2014
Stanley TD, Doucouliagos H, Giles M, Heckemeyer JH, Johnston RJ, Laroche P, Nelson JP,
Paldam M, Poot J, Pugh G et al (2013) Meta-analysis of economics research reporting
guidelines: reporting guidelines for meta-regression analysis in economics. J Econ Surv
27:390–394
Stigka EK, Paravantis JA, Mihalakakou GK (2014) Social acceptance of renewable energy sources:
a review of contingent valuation applications. Renew Sust Energ Rev 32:100–106
Strazzera E, Mura M, Contu D (2012) Combining choice experiments with psychometric scales to
assess the social acceptability of wind energy projects: a latent class approach. Energy Policy
48:334–347
van Rensburg TM, Kelley H, Jeserich N (2015) What influences the probability of wind farm
planning approval: evidence from Ireland. Ecol Econ 111:12–22
Wolsink M (2000) Wind power and the NIMBY-myth: institutional capacity and the limited
significance of public support. Renew Energ 21:49–64
Wosink M (2007) Wind power implementation: the nature of public attitudes: equity and fairness
instead of ‘backyard motives’. Energy Policy 11:1188–1207
Wüstenhagen RW, Markard J, Truffer B (2003) Diffusion of green power products in Switzerland.
Wüstenhagen R, Wolsink M, Bürer MJ (2007) Social acceptance of renewable energy innovation:
an introduction to the concept. Energy Policy 35:2683–2691
Yoo SH, Kwak SY (2009) Willingness to pay for green electricity in Korea: a contingent valuation
study. Energy Policy 37:5408–5416
Zhang L, Wu Y (2012) Market segmentation and willingness to pay for green electricity among
urban residents in China: the case of Jiangsu Province. Energy Policy 51:514–523
Zoric J, Hrovatin N (2012) Household willingness to pay for green electricity in Slovenia. Energy
Policy 47:180–187
Chapter 2
Evaluating an Onshore Wind Farm
Enlargement Project: A Contingent
Valuation Study in Central Italy
Abstract In many European countries the most suitable onshore sites for wind
installations are almost fully engaged; furthermore, the existing onshore wind farm
capacity will be replaced in the next 10 years, given that wind power plants are
progressively aging; without an adequate policy intervention, the Italian installed
wind power capacity would return to the size of 2011 by 2030. In this scenario, two
opportunities exist for further growth in wind energy generation: repowering or wind
farm enlargement. The choice between these two options mainly depends on local
characteristics. The aim of this chapter is twofold. First, we investigate whether
existing wind farms affect respondents’ attitudes and perceptions towards the poten-
tial enlargement of a wind farm, using a contingent valuation (CV) method. Second,
we investigate the perception of the risk associated with the enlargement of a wind
farm. In this case we explicitly take into account the existence of respondents’
heterogeneity in perceiving the new project externalities. To do this, we use both
willingness to pay (WTP) and willingness to accept (WTA) measures in order to
appraise welfare change due to the enlargement project. Each of these measures is
elicited jointly with the respective appraised externality impact perceived by the
respondents. The findings can offer useful insights for planning and design of
enlargement schemes in order to achieve further growth in wind energy generation.
Keywords Wind farm enlargement · Contingent valuation · Willingness to pay ·

Willingness to accept · Uncertainty
2.1 Introduction
Historically, the European Union (EU) has devoted great attention to determinants of
climate change. In particular, the need to reduce CO2 emissions has been formalized
by several legislative tools, as confirmed by the EU Commission decision on new
climate strategies taken in October 2014.
The increase in renewable energy sources (REnS) in the electricity generation
mix has represented an important response to climate change, given the importance
of renewables in diversifying the energy supply and preserving the environment.

28 2 Evaluating an Onshore Wind Farm Enlargement Project. . .
Furthermore, REnS produce other important positive externalities such as preserva-

tion of fossil fuels and increased spread of new technologies, small-scale plan
development, and job opportunities. Typically, these general advantages are bal-
anced by negative externalities due to the investment projects necessary to promote
REnS development.
Focusing on wind power, we think of its visual intrusion; local land use, which
conflicts with other economic activities; impacts on flora and fauna; the noise
impact; and the shadow flicker effects caused by the turbines (Leung and Yang
2012).
Balancing local negative externalities and global positive externalities, the former
are often not sufficient to convince local communities to accept the installation of
new onshore wind farms in their territories. This is particularly true in Italy, as
emphasized in Sect. 2.2. Nevertheless, what is less investigated in the literature is
local communities’ response to an enlargement project at an existing wind farm. This
is an important gap given that (1) the most suitable onshore sites for wind installa-
tions are almost fully engaged, especially in countries that have developed wind
energy facilities earlier; and (2) the existing onshore wind farm capacity should be
replaced in the next decade, given that wind power plants are progressively aging.
In Italy the average age of a typical wind farm will be 21 years in 2030. Given this
technical obsolescence, without an adequate policy intervention to support the wind
power industry, this means that the Italian installed wind power capacity would
return to the size of 2011 by 2030.
In this scenario, two opportunities exist for further growth in wind energy
generation to achieve new challenging environmental targets stated by the
European Council according to its climate and energy policies: repowering or
wind farm enlargement.1 The choice between these two options mainly depends
on both local acceptance and project characteristics. In this chapter we only focus on
wind farm enlargement.
The aims of this chapter are threefold. First, we investigate whether existing wind
farms affect respondents’ attitudes and perceptions toward the potential enlargement of
a wind farm, using a contingent valuation (CV) method. In this way we fill the gap in
the literature about the stated preferences regarding the enlargement of a wind farm.
Second, we investigate the perception of the risk associated with the enlargement
of a wind farm. In this case we explicitly take into account the existence of
respondents’ heterogeneity in perceiving the new project externalities referring to
potential land use conflicts and local opposition. To do this, we use both willingness
to pay (WTP) and willingness to accept (WTA) in order to measure welfare change
due to the enlargement project. In detail, both WTP and WTA are elicited jointly
with the subjective estimates of the respective main impact perceived by the
respondents.
1
Offshore development is the third opportunity. In this chapter we do not take it into account. In
Italy the National Action Plan (NAP) implemented Directive 2009/28/EC identifying a target of
680 MW by 2020. Until now in Italy we have not installed offshore wind farms.
2.2 Related Literature 29
Third, we appraise the impact of uncertainty, taking into account several degrees
of uncertainty that affect the WTP and WTA elicited from respondents, using the
Numerical Scale Method (NSM). The chapter is organized as follows. Section 2.2
provides a review of the literature. A method for valuing wind farm enlargement is
presented in Sect. 2.3. Section 2.4 describes the theoretical and econometric frame-
works. Section 2.5 presents an empirical analysis discussing the main results.
Section 2.6 provides the conclusions.
2.2 Related Literature
Recent contributions—such as those from Mattmann et al. (2016), Rand and Hoen
(2017), and Zerrahn (2017)—have examined the factors affecting local community
behavior toward onshore wind farm development, assessing the positive and nega-
tive externalities of this type of investment. The positive externalities are associated
with global and environmental effects, while the negative externalities refer to the
localized impact of investment projects necessary to promote wind energy develop-
ment. The negative externalities are associated with visual intrusion; local land use,
which conflicts with other economic activities; impacts on flora and fauna; the noise
impact; and the shadow flicker effects caused by the turbines (Leung and Yang
2012). Positive externalities refer to lower greenhouse gas emissions, air quality
improvement, fossil fuel independence, and job opportunities (Mattmann et al.
2016). Nevertheless, these positive features are often not sufficient to convince
local communities to accept the installation of new onshore wind farms in their
territories.
A review of the literature includes at least two types of contributions. The first
type deals with local acceptance/opposition, while other contributions deal with
economic quantification of local community attitudes, using the CV method.
Wüstenhagen et al. (2007) emphasize the innovation aspects, underlining that
wind energy is particularly affected by constrained social factors. Huijts et al. (2012)
investigate the psychological aspects, identifying three types of acceptance. The first
two are related to citizen and consumer behavior, respectively. The third one refers
to sociopolitical acceptance. While the first two types directly involve people,
affecting their social and economic interests, the third type of acceptance refers to
the supralocal scale, so it does not necessarily directly affect individuals.
The literature on acceptance of wind farm development based on CV can be
divided into three major groups. In the first group, researchers focus on environ-
mental impacts due to wind farm projects (Álvarez-Farizo and Hanley 2002;
Bergmann et al. 2006; Ek and Matti 2014; Mirasgedis et al. 2014). The papers in
the second group mainly refer to the technological characteristic of wind farms
(Dimitropoulos and Kontoleon 2009; Meyerhoff et al. 2010; Meyerhoff 2013),
while the last group of papers takes into account both of these aspects, also including
other institutional or economic topics (see, among others, Ek (2005) and Strazzera
et al. (2012)). Focusing on the issue discussed in this chapter—repowering and wind
farm enlargement—the literature is more limited, as highlighted by Frantàl (2015).
Kaldellis et al. (2013) report a high level of acceptability for enlargement and new
renewable energy projects. Indeed, no significant differences arise in this study
among wind energy, photovoltaic energy, and small hydropower applications in
term of acceptability. Furthermore, these authors point out that local communities
that have experienced operation of renewable energy applications maintain compar-
atively high acceptance of new renewable energy projects. Investigated communities
near wind farms in Texas, USA (Swofford and Slattery 2010; Slattery et al. 2012),
confirm that large public support exists in favor of building more wind farms in their
proximity.
The attitudes and perceptions of local communities toward the wind power
development process have been analyzed from several points of view. Arguments
in favor of or in opposition to wind farm projects have been addressed in both local
and global dimensions. In Europe, Rygg (2012), analyzing 13 communities in
Norway, underlines that favorable positions mainly refer to economic benefits, job
opportunities, and modernization. One important result is that different types of
externalities refer to a new project or to an enlargement project, although the
uncertainty regarding the proposed project always plays a central role. Jones and
Eiser (2010) underline that endorsement received for onshore sites is negatively
affected also by the anticipated impact due to “speculation, myth-propagation and
social media amplification.” Although visual intrusion and proximity2 are frequently
invoked by researchers to explain opposition processes, others factors are also
highlighted (Frantàl 2015). Read et al. (2013) have found that behavioral intentions
to oppose wind farm developments is not predicted by visual perception and sense of
place but is most strongly associated with past oppositional behavior. Jones et al.
(2011) stress the importance of cumulative effects assessment, while Friedl and
Reichl (2016) underline the importance of the institutional context and the partici-
pation process in order to minimize cost and delays in project completion.
To the best of our knowledge, this is the first contribution on assessment of wind
farm enlargement using the stated preference approach. Furthermore, the literature
on wind farm location assessment rarely includes the WTA approach; in this chapter
we also adopt this welfare measure. Finally, we analyze the impact of respondents’
uncertainty on welfare measures, adopting a different way to manage the uncertainty
measure. Taking into account all of these aspects, our contribution fills an important
gap in the literature.
2
The “proximity versus reverse proximity” debate is still an open question. The empirical literature
seems to support both the proximity and the reverse proximity hypothesis (Frantàl 2015), and in this
chapter we do not directly analyze these aspects, although our results may contribute to the current
academic debate.
2.3 Method for Valuing Wind Farm Enlargement 31
2.3 Method for Valuing Wind Farm Enlargement
This is the first experiment designed to apply the stated preference approach to
evaluate a wind farm enlargement project. However, several papers have used the
stated preference method to estimate welfare change associated with an onshore
wind farm project, using different points of view (see, among others, Álvarez-Farizo
and Hanley (2002), Bergmann et al. (2006), Moran and Sherrington (2007),
Groothius et al. (2008), Dimitropoulos and Kontoleon (2009), Meyerhoff et al.
(2010), Meyerhoff (2013), and Mirasgedis et al. (2014)). We introduce the experi-
ment—describing the wind power sector, the case study, and its relevance—and then
we present our survey method.
2.3.1 The Scenario: Wind Power Generation in Italy
In the EU, Italy ranks fifth among the nations for wind power capacity, with a
capacity of 9 GW. In the ranking, Italy follows Germany (44.9 GW), Spain (23 GW),
the UK (13.6 GW), and France (10 GW). In 2010 the National Action Plan (NAP) in
Italy implemented Directive 2009/28/EC identifying a target capacity of 12,000 MW
to achieve by means of onshore wind farms.
By the end of 2016, approximately 9250 MW were installed in Italy, with a
planned gap of about 2750 MW. Thus, to reach the threshold of 12,000 MW, about
700 MW per year should be installed. From the technical point of view this figure is
potentially achievable by the Italian industry, but it is an arduous task due to local
opposition phenomena. In Italy, local opposition has blocked several wind projects;
as at 2014, about 50 new onshore wind farm projects were beset by “Not in My Back
Yard” (NIMBY) syndrome, with 14 new projects also being disputed by local
communities in the subsequent 2 years (Nimbyforum 2016).
Further, given that Italian wind power plants are progressively aging without an
adequate policy intervention, the Italian wind power capacity never will achieve the
NAP target; nevertheless, the national potential capacity is estimated to be around
17 GW. In particular, a recent study conducted by the National Wind Energy
Association (ANEV) (2017) underlines that in the Italian scenario, power capacity
by 2030 will decrease from the current 9250 MW to around 7000 MW—the same as
the 2007 capacity.
2.3.2 Case Study: The Monte Cucco Regional Park
Mount Cucco (height: 1566 m) dominates the mountain system of the Apennines in
Umbria in central Italy. In 1999 a wind farm of two turbines of 1 MW (Monti Mutali)
Fig. 2.1 Umbria and Monte Cucco Regional Park
was located in this area close to the Monte Cucco Regional Park (Fig. 2.1). The
turbines are 50 m high and the diameter of the rotors is 44 m.
The current project involves installation of 16 towers for wind generation of
electricity. Precisely, the plan should provide for installation of four towers reaching
40–60 m in Pian di Spilli (in the municipality of Costacciaro) and in Val di Ronco
(in the municipality of Sigillo), and eight similar towers in the municipalities of
Scheggia Pascelupo and Fossato di Vico. Each tower will have a maximum power of
1 MW. To date, this is the only operational wind farm in Umbria, and it is an
important case study, given the characteristics of Umbria.
Umbria is known worldwide for its man-made landscape and for its environmen-
tal quality, earning the title of the “Green Heart of Italy.” In other words, the main
externalities due to wind energy generation involve the most valuable characteristics
2.3 Method for Valuing Wind Farm Enlargement 33
of this region3 and then greatly involve residents’ interests in the wind energy
development issue.
2.3.3 Survey Method and Questionnaire
In this chapter we analyze the relationship between local communities and power
plant facilities thorough a CV method jointly considering the perceptions of the
impact of the existing operational turbines and the degree of acceptance of wind farm
enlargement.
Data on the attitude toward the additional wind farm project have been collected
to investigate if the residents’ preferences have been affected by the experience of
living near an operational wind farm. Secondly, we have investigated wind farm
enlargement acceptance, focusing on the related local community risk perception. To
do this we have investigated the general attitude of the local community, taking into
account the respondents’ heterogeneity in perceiving negative and positive exter-
nalities associated with the new project. In other words, we clearly identify wind
power as the source of positive and negative externalities in order to jointly take into
account elicited value and externality risk perception. Externalities have mainly been
considered according to the existing literature (Mattmann et al. 2016) in order to
clearly define positives and negatives involved in the enlargement project. This
allows us to mitigate the hypothetical nature of the CV method, setting out the
trade-offs between the amenity in question and the monetary good. According to an
analysis by Shaikh et al. (2007), this is another way to reduce respondents’
uncertainty.
Another important item emerging from the literature review is that although
several findings refer to the monetary valuation problem of the wind power gener-
ation externalities, few studies exist in which estimates are available in terms of both
WTP and WTA. We allow respondents to assess the projects in both positive and
negative ways, and whether they are willing to pay to support the project or are
willing to accept monetary compensation. These monetary evaluations are drawn
using a simulated market in which the WTP to support the project or the WTA to
accept it are directly obtained by respondents jointly with the respectively perceived
externalities. The robustness of this assessment process requires that the respondents
know the good under valuation. In our research, the wind farm enlargement project
and the associated externalities are well known by the respondents, given that they
have experienced the wind farm operation for more than 15 years.
3
These characteristics have been promoted by Sensational Umbria, a photography exhibition by
Steve McCurry, held in Perugia, Italy, in 2014–2015. Sensational Umbria has been an innovative
use of a photographic exhibition for the purpose of promoting tourism worldwide. Through his
pictures, McCurry has highlighted relationships between man-made landscapes and the
environment.
Two preliminary analyses were conducted in May and July 2011 by focus groups
composed of energy managers, experts, members of local government, residents,
and academics in order to define the questionnaires’ questions and description of the
scenario, and how residents would be directly or indirectly affected by the change
due to the proposed project. These two points were crucial to guarantee the respon-
dents’ understanding of the proposed change. Finally, given that the payment vehicle
influences responses, it also needed to be properly defined during the preliminary
analysis.
In September 2011 we conducted the pilot survey4 to test the understandability of
the scenario proposed and the related questions.5 The preliminary results provided us
with an operative questionnaire that initially describes the pros and cons of the wind
farm enlargement scenario. The first section is aimed at testing respondents’ knowl-
edge on: (1) wind energy issues at the European and national levels; and (2) envi-
ronmental and wind energy issues at the regional level.
In the second section, questions are aimed at investigating respondents’ attitudes
toward potential benefits and environmental damage associated with the project,
taking into account the fact that the wind farm has been operational since 1999. In
particular, we jointly examine residents’ preferences and the potential impacts of the
externalities associated with wind farm enlargement as perceived by the respondents.
In the third section, the questionnaire allows respondents to state if they want to
support the project or if they want to be compensated for it. The reasons for their
choice are asked by quantifying the risk perceived.
The degree of WTP/WTA certainty is also asked according to the NCS approach.
In this way we consider an additional type of heterogeneity, taking into account each
respondent’s uncertainty in defining the amount of money that he/she is willing to
pay or to accept. We ask each respondent to match his (or her) WTP (or WTA) to a
level of certainty by choosing a number from 0 to 10, where 10 is equivalent to
“absolutely sure.” This information is used to directly ponder the answers (Evans
et al. 2003), using various thresholds of certainty, which take an answer as definite.
At the end of the questionnaire, residents are asked about their socioeconomic and
demographic attributes.
2.4 Theoretical and Econometric Framework
In this study a new scenario is proposed to residents, and their WTP to support the
project is then elicited if they agree with the enlargement; otherwise, their WTA is
recorded in order to accept the project. As with any CV application, a risk of
incurring bias always exits. Nevertheless, international empirical results demonstrate
4
Respondents are expected to believe in the plausibility of the proposed simulated transaction.
5
Guideline suggestions (Soderqvist and Soutukorva 2006) are applied in the questionnaire to test
for understanding and acceptance of the scenario proposed.
2.4 Theoretical and Econometric Framework 35
that a well-designed and carefully administered survey can provide consistent,

coherent, and credible information on welfare change estimates.
2.4.1 Theoretical Model and Elicitation Format
It is assumed that the total utility (U) of a family is a function of the indirect utility (v)
depending on the income (I ), on the landscape (L ), and on the air quality (A), with
p as the price vector.
U ¼ vðp; I; L; AÞ ð2:1Þ
Families who claim to perceive a negative impact of the project attach greater
importance to L; therefore, we state that the landscape worsens L1 < L0 with the
realization of the project, and a form of compensation will be requested:

v p0 ; L0 ; I ¼ v p0 ; L1 ; I þ WTA ð2:2Þ
They are also asked to express the degree of risk characterizing the landscape on a
scale from 1 to 10, with 10 representing the highest risk.
Families who express a positive attitude toward the project, by contrast, attach
greater importance to A. The realization of the project improves air quality A1 > A0,
and the family will be willing to contribute:

v p0 ; A0 ; I ¼ v p0 ; A1 ; I WTP ð2:3Þ
These families are also asked to express the risk degree that characterizes air
healthiness on a scale of 1–10, with 10 representing the maximum risk level.
Comparing (2.2) and (2.3) it is important to note that while WTA is not bound in
any way to income, WTP can at most assume a value equal to the latter. As a result,
using WTA may perhaps lead to overestimation of the wellness mutation. Solving
(2.2) and (2.3) for WTA and WTP, respectively, we can point out the following
implied relations dealing with the estimation of the well-being measures:
WTA ¼ f ðI; LÞ ð2:4Þ

WTP ¼ f ðI; AÞ ð2:5Þ
We have adopted the open-ended (OE) CV method, given that it provides the
most prudent estimates of the inspected economic value (Frew et al. 2004; Lunander
1998). Typically, the main drawbacks of this format (e.g., difficulty in asking the
payment question, resulting in many missing values) are strongly limited in our
context, given that the respondents are closely involved in the proposed scenario and
highly confident about the proposed project.
However, economic values elicited by stated preferences methods can differ from
actual values, given that the hypothetical nature of the questions can lead to biased
responses, weakening the findings (Mitchell and Carson 1989). Nevertheless, the
accumulated literature has improved the empirical methodology, reducing the gap
between actual and hypothetical values (Loomis 2011, 2014).
In particular, one of the main sources of bias is the respondents’ uncertainty. In
this chapter an NCS from 1 to 10 (maximum certainty) is used, allowing respondents
to indicate the certainty level for their declared WTP and WTA. At this preparatory
stage of the study, these data are used not to ponder the information within the
likelihood function but to ponder the answers directly detected by the literature and
using various thresholds of certainty, which take an answer as definite.
Using this approach we estimate both WTP and WTA measures, taking into
account respondents’ uncertainty; furthermore, we contribute to an increase in the
literature on wind farm enlargement, which represents one of the major challenges
for the European and Italian wind power sectors.
2.4.2 Econometric Model
Besides income, however, features such as the related riskiness perception and the
private characteristics of respondents may affect both WTP and WTA.
We have decided to use a Seemingly Unrelated Regression (SUR) (Zellner 1962)
estimator in order to take interactions between both WTP and WTA into account.
WTP and WTA are then estimated using two independent equation systems, each
weighting the well-being measure conjointly regressed to a risk measurement,
obtained through environmental risk scales, underlining the existent correlation
among errors. Formally:
System 1:
yia ðWTAÞ ¼ X ia βia þ εia ð2:6aÞ

yib ðRisk LÞ ¼ X ib βib þ εib ð2:6bÞ
System 2:
y ja ðWTPÞ ¼ X ja β ja þ ε ja ð2:7aÞ
y jb ðRisk AÞ ¼ X jb β jb þ ε jb ð2:7bÞ
The individual and family characteristics of the two groups of respondents—i and
j, respectively—are represented by the matrix X of the explanatory variables, while β
is the relevant parameter vector of each equation and ε is the vector of the error.
This approach allows us to integrate behavioral and perceptual features in a
purely economic framework, taking into account psychological aspects (Mozumder
et al. 2009). Indeed, many unobservable household and individual factors might
influence WTP and WTA—first of all, the perceived impact of externalities due to
wind power generation. Using the SUR approach, more precise parameters are
estimated in a more efficient way (Park and Loomis 1996; Riddel and Loomis 1998).
2.5 Results and Discussion
A local survey with 715 interviews was administered from the end of July to
September 2012, using a stratified sample representative of local residents living
in four municipalities involved in the project. Professional interviewers conducted
the survey, using a computer-aided telephone interviewing method.
The majority of the questionnaires (99%) were completed.
2.5.1 Descriptive Results
The respondents show a favorable attitude with regard to wind in electricity gener-
ation.6 The majority of the sample believes that climate change will worsen in the
next 10 years and considers that wind energy represents a strategic opportunity at the
national and local levels. Some preoccupations arise referring to visual intrusion,
although many respondents believe that the improvement in air quality and the
reduction of emissions are the most important externalities associated with wind
power generation.
The majority of the respondents (83%) state that they are aware of both positive
and negative externalities associated with wind power generation. Opponents of the
project indicate that the visual intrusion is the main externality (74%), while damage
to fauna and noise pollution are reported at 17% and 9%, respectively. It is important
to stress that 53% of the full sample express a certain disappointment with the lack of
involvement in the preview decision-making process in 1999, and 38% hope for
greater involvement in any future projects.
Faced with questions concerning the economic scenario, residents are invited to
declare a possible WTP or WTA associated with their attitude toward the project, as
described in the interview. In cases with a positive attitude, respondents are asked to
contribute to the project by assuming an increase in the A3 component7 of their
electricity bill, dedicated to support for renewables. Of the sample, 50.2% agree to
6
For the sake of brevity we comment on the survey results without presenting the related table and
figures, which are available from the authors upon request.
7
This is a component of electricity bills that also covers the cost of REnS used in electricity
generation. In addition to REnS the A3 component includes subsidies for power plant production
based on conventional fuels using alternative production techniques. The A3 component is used to
construct a reliable WTP/WTA scenario; indeed, the respondents are first asked to state the amounts
of their last bills, then they are informed of the A3 component.
contribute financially to the project. The main motivation (81%) is a reduction in the
CO2 emissions associated with electricity production from fossil fuels. The effects
on employment are less important for the sample; in fact, only 15% of respondents
state a preference for improvement of this feature.
A negative attitude implies that respondents are compensated through reduced
electricity bills. The main motivation (78%) is the visual intrusion associated with
the wind farm project. The effects on fauna are considered less important, with only
12% of respondents stating a preference for a reduction in this externality.
More deeply, Table 2.1 shows that the sample is divided into two groups: those
who support the wind farm enlargement to eliminate the environmental impacts due
to power generation from fossil fuels, and those who are opposed to the project
because of the visual impact due to the wind farm enlargement.
The majority of respondents support the project (54% of the sample) and believe
that the deployment of wind power according to NAP and EU environmental targets
should be promoted. Overall, 56% of respondents are willing to pay higher prices to
minimize negative externalities due to fossil fuel power generation.
Consistent with the negative opinion concerning the proposed project, 46% of
respondents demand financial compensation for the realization of the wind farm
enlargement.
Finally, a brief discussion on the results related to the degree of uncertainty
associated with the monetary metrics elicitation is necessary.
Although we have not formally analyzed it through the theoretical model, we
asked each respondent to evaluate his or her WTP (WTA), in terms of certainty by
choosing a score from 0 to 10, where 10 was equivalent to absolutely sure. In this
chapter the obtained scores are used not to weight the information within the
likelihood function but to directly weight the answers (Evans et al. 2003); further-
more, we use various thresholds of certainty to define a more robust certainty level.8
The elicited WTP and WTA according to different degrees of uncertainty are shown
in Fig. 2.2. Figure 2.2a, b refer to WTP measures. The median WTP lies between
EUR 12 and EUR 16 (Fig. 2.2a); the minimum value refers to the weighted measure.
The confidence intervals (Fig. 2.2c) are quite close to the mean values; the upper
bounds range from EUR 12.61 to EUR 17.00, while the lower bounds lie between
EUR 11.32 and EUR 14.92. Except for the weighted WTP, the mean is quite stable
at around EUR 15. According to economic theory, WTA measures are higher than
WTP ones. Fig. 2.2b shows that the median WTA lies between EUR 19.60 and
EUR 27.00, while the mean values (Fig. 2.2d) are more variable, ranging from
EUR 20.03 to EUR 28.79. Also, in this case the minimum value is obtained with the
weighted procedure, while the maximum value is associated with the unweighted
measures.
Welfare measure distribution functions are shown in Figs. 2.3 and 2.4.
Unweighted and weighted WTP and WTA distributions are compared in Fig. 2.3.
8
With regard to the NCS used (Table 2.1), in the case of WTP, we have an average of 7.67 and a
standard deviation of 2.13, whereas for WTA, the two values are 7.23 and 1.93, respectively.
Table 2.1 Sample answers: descriptive statisticsa

Variable Type Abbreviation Unit Supportersb Opponentsc
Bimonthly electricity bill Continuous bill EUR 70.11 70.28
8.58 8.62
Head of family age Continuous agehf n 54.46 49.95
16.17 14.44
Gender Dummy sex # 0.44 0.44
(1 ¼ female) 0.50 0.50
Family years of residence Continuous res_years n 36.12 34.05
(ancestors included) 20.75 17.97
Household members Continuous fam_s n 3.09 2.93
1.15 1.00
Household income Continuous incom EUR 2.93 2.73
(10,000) 1.41 1.17
Leftist party affinity Dummy leftp # 0.64 0.67
(1 ¼ yes) 0.48 0.47
Amenity perception Scale (1–10) qol n 3.44 6.01
(10 ¼ max) 3.10 3.43
Externality perception (air Scale (1–10) air_q n 6.06 –
quality) (10 ¼ max) 2.25 –
Externality perception Scale (1–10) landrisk n – 6.06
(visual intrusion) (10 ¼ max) – 2.25
Years of education Continuous educ n 9.50 10.52
3.75 3.79
Environmental association Dummy env_ass # 0.19 0.14
membership (1 ¼ yes) 0.39 0.34
Homeowner Dummy homeow # 0.82 0.72
(1 ¼ yes) 0.38 0.45
Positive attitude to wind Dummy att_t0 # 0.21 0.15
farm project in 1999 (1 ¼ yes) 0.45 0.38
Wind farm experience Dummy wfexp # 0.29 0.22
(1 ¼ yes) 0.45 0.61
Mean WTP Continuous wtp EUR 15.65 –
6.94 –
Mean WTA Continuous wta EUR – 28.79
– 10.35
Certainty degree (NCS) Scale (1–10) n 7.67 7.23
(10 ¼ max) 2.13 1.93
max maximum, NCS Numerical Certainty Scale, WTA willingness to accept, WTP willingness to
pay
a
b
364 observations
c
336 observations
Fig. 2.2 Elicited willingness to pay (WTP) -median (panel a) and mean (panel c)- and willingness to
accept (WTA) -median (panel b) and mean (panel d)- according to degrees of uncertainty (2014 EUR)
Fig. 2.3 Unweighted and weighted willingness to pay (WTP) and willingness to accept (WTA)
distributions
The distributions of the unweighted welfare measures, shown at the top of Fig. 2.3,
marginally overlap, confirming the distance between the two measures.
The distance between the two welfare measures is smaller using the weighted
distributions, as shown at the bottom of Fig. 2.3, confirming the importance of taking
into account the uncertainty associated with the elicitation procedure.
Fig. 2.4 Unweighted willingness to pay (WTP) and willingness to accept (WTA) distributions
according to six degrees of certainty
Finally, Fig. 2.4 compares WTP and WTA distributions according to several
degrees of uncertainty, from 5 to 10. Reducing the degree of certainty, the distance
between the two distributions increases and the overlapping areas are small, partic-
ularly for low levels of certainty.
2.5.2 Econometric Analysis
The obtained results are shown in Table 2.2. In each model the dependent variable
from the first equation (2.6a and 2.7a) is the monetary metric (WTP and WTA), and
in the second equation (2.6b and 2.7b) the left hand side (LHS) is the measures of the
riskiness of the externality as perceived by the respondents. Table 2.2 shows that
only three variables are significant in equation system 1, regardless of uncertainty
degrees: homeown in the first equation and age and income in the second equation.
This first result confirms the importance of uncertainty treatment in the CV studies. It
is also important to underline that even if the degree of uncertainty has an impact on
the magnitude of the estimated coefficients, uncertainty rarely changes the sign of
the relationship in our experiment in both the first and second equation systems.
Focusing on the first equation system, less significant models are associated with
Table 2.2 Estimated willingness to pay (WTP)

UnW_wtp W_wtp U_wtp_10 U_wtp_9 U_wtp_8 U_wtp_7 U_wtp_6 U_wtp_5
[Eq. 1: 0.1168a 0.0725b 0.0890 0.0167 0.0424 0.0670c 0.0886a 0.0962a
wtp] bill
0.0393 0.0346 0.0581 0.0455 0.0417 0.0385 0.0346 0.0375
agehf 0.0341 0.0106 0.0454 0.0452 0.0416 0.0441c 0.0559b 0.0339
0.0265 0.0233 0.0350 0.0299 0.0278 0.0263 0.0242 0.0251
sex 1.7921b 1.8704a 0.6629 1.8266b 1.8845b 2.5339a 2.5125a 2.2114a
0.7687 0.6766 1.0493 0.8772 0.8305 0.7756 0.7103 0.7505
res_years 0.0471b 0.0516a 0.0380 0.0230 0.0385c 0.0497a 0.0383b 0.0506a
0.0190 0.0167 0.0259 0.0212 0.0200 0.0189 0.0172 0.0181
fam_s 0.5154c 0.1651 0.7046c 0.8575a 1.1218a 1.1483a 0.9226a 0.6360b
0.2924 0.2574 0.3617 0.3150 0.3061 0.2965 0.2678 0.2888
incom 1.1656a 0.3527 1.1635a 2.1779a 2.1625a 1.9368a 1.9067a 1.3940a
0.2517 0.2216 0.4517 0.3391 0.3353 0.3061 0.2716 0.2751
leftp 1.1254 0.6990 0.4972 0.7061 0.9016 0.4956 0.4291 0.5466
0.6893 0.6067 0.9184 0.7445 0.7048 0.6787 0.6261 0.6646
qol 0.2491b 0.0088 0.3506b 0.2593b 0.4316a 0.3792a 0.3488a 0.3792a
0.1094 0.0963 0.1574 0.1219 0.1183 0.1115 0.1014 0.1084
educ 0.2459b 0.3331a 0.2223 0.0471 0.1840c 0.2361b 0.2223b 0.2089b
0.0967 0.0851 0.1518 0.1215 0.1097 0.1001 0.0914 0.0965
env_ass 1.8653b 3.7801a 1.7732 1.7579 0.2358 1.0656 0.9682 1.8959b
0.8217 0.7232 1.2795 1.1288 1.0115 0.9318 0.8712 0.8644
homeow 2.0878b 1.4594b 1.7723c 1.8455b 1.6412b 1.7333b 1.5944b
0.8401 0.7395 1.0275 0.8716 0.8246 0.7876 0.7399 0.8039
wfexp 1.9535a 1.8718a 0.3505 0.0812 0.5488 1.1128 1.3428b 1.4203
0.7218 0.6353 0.9320 0.8016 0.7369 0.6936 0.6346 0.6843
_cons 3.2143 0.8159 3.9786 4.3971 0.5388 1.6540 3.7272 2.7025
3.4030 2.9953 5.0621 4.1062 3.6568 3.3388 2.9940 3.2544
[Eq. 2: 0.0239b 0.0239b 0.0254 0.0078 0.0164 0.0164 0.0173c 0.0214
air_q] bill
0.0114 0.0114 0.0226 0.0154 0.0127 0.0112 0.0102 0.0105
agehf 0.0208a 0.0208a 0.0269b 0.0190c 0.0227a 0.0200a 0.0230a 0.0242
0.0077 0.0077 0.0136 0.0101 0.0085 0.0076 0.0071 0.0070
sex 0.1303 0.1303 0.2470 0.0207 0.0571 0.0582 0.0095 0.0897
0.2233 0.2233 0.4080 0.2976 0.2528 0.2248 0.2091 0.2098
res_years 0.0077 0.0077 0.0189c 0.0164b 0.0130b 0.0085 0.0099c 0.0095
0.0055 0.0055 0.0101 0.0072 0.0061 0.0055 0.0051 0.0051
fam_s 0.1459c 0.1459c 0.0584 0.1161 0.0413 0.0231 0.0014 0.0100
0.0849 0.0849 0.1406 0.1068 0.0932 0.0859 0.0788 0.0807
incom 0.2727a 0.2727a 0.3702b 0.2432b 0.2807a 0.2379a 0.1753b 0.1658
0.0731 0.0731 0.1756 0.1150 0.1021 0.0887 0.0799 0.0769
leftp 0.0728 0.0728 0.1090 0.1908 0.0723 0.1240 0.1077 0.0118
0.2003 0.2003 0.3571 0.2525 0.2146 0.1967 0.1843 0.1858
qol 0.0786b 0.0786b 0.0797 0.0642 0.0686c 0.0856a 0.0629b 0.0712b
0.0318 0.0318 0.0612 0.0413 0.0360 0.0323 0.0299 0.0303
educ 0.0540c 0.0540c 0.0116 0.0264 0.0459 0.0591b 0.0560b 0.0486c
(continued)
Table 2.2 (continued)

UnW_wtp W_wtp U_wtp_10 U_wtp_9 U_wtp_8 U_wtp_7 U_wtp_6 U_wtp_5
0.0281 0.0281 0.0590 0.0412 0.0334 0.0290 0.0269 0.0270
env_ass 0.2129 0.2129 0.5033 0.5018 0.5410c 0.4829c 0.5686b 0.3777
0.2387 0.2387 0.4975 0.3829 0.3079 0.2700 0.2564 0.2417
homeow 0.1909 0.1909 0.0405 0.6809b 0.5778b 0.5200b 0.4846b 0.4725b
0.2441 0.2441 0.3995 0.2957 0.2510 0.2282 0.2178 0.2247
wfexp 0.4655b 0.4655b 0.5144 0.4340 0.3596 0.4358b 0.3775b 0.2433
0.2097 0.2097 0.3624 0.2719 0.2243 0.2010 0.1868 0.1913
_cons 0.8041 0.8041 0.9578 2.0711 0.7879 0.6919 0.7041 0.4563
0.9887 0.9887 1.9683 1.3929 1.1133 0.9675 0.8812 0.9099
Obs
Eq. 1 345 345 74 143 205 262 299 314
Eq. 2 345 345 74 143 205 262 299 314
RMSE
Eq. 1 5.7995 5.1047 3.3713 3.9949 4.5541 4.8848 4.7616 5.2864
Eq. 2 1.6850 1.6850 1.3109 1.3552 1.3864 1.4155 1.4015 1.4780
R-sq
Eq. 1 0.2617 0.2298 0.4045 0.3494 0.4033 0.4078 0.4097 0.3382
Eq. 2 0.1531 0.1531 0.2503 0.2166 0.2154 0.1957 0.1729 0.1555
χ2(12)
Eq. 1 122.29 102.94 50.26 76.81 138.56 180.39 207.52 160.5
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Eq. 2 62.38 62.38 24.71 39.55 56.28 63.75 62.51 57.8
(0.00) (0.00) (0.02) (0.00) (0.00) (0.00) (0.00) (0.00)
Res corr 0.1188 0.1174 0.0522 0.0457 0.1530 0.1529 0.1772 0.1562
B-P test 4.872 4.722 0.202 0.299 4.796 6.126 9.386 7.663
χ2(1)
B-P test 0.027 0.029 0.653 0.585 0.029 0.013 0.002 0.006
-Prob-
WTP 18.24a 12.38a 9.29b 12.67a 9.36a 13.85a 14.32a 12.58a
mean
2.28 1.36 4.35 3.57 3.39 0.99 0.89 1.67
lo.bo 13.78 9.71 0.77 5.67 2.72 11.92 12.57 9.30
up.bo 22.71 15.05 17.82 19.67 15.99 15.78 16.067 15.85
B-P Breush–Pagan, Obs observations, Res corr residual correlation, RMSE root-mean-square error,
R-sq R-squared
a
b
c
higher certainty thresholds (e.g., an NSM index greater than 8). We also notice that
these models use fewer observations. Hereafter, we comment only on variables that
are significant at a level of at least 5% in at least half of the estimated models.
Economic variables highly affect WTP, which increases with income and is
higher for homeowners; these relationships are significant in seven out of eight
estimate models.
The electricity bill is positively related to WTP, suggesting that higher expendi-
ture for electricity does not reduce economic support for the project. This relation
may be explained by the existence of a possible correlation between income and the
electricity bill, but even checking for the interaction between these variables,9 the
parameter for the electricity bill is still significant and positive. A possible explana-
tion refers to respondents’ expectation of a decreasing electricity price due to
renewable deployment.
Demographic variables are statistically significant and consistent with expecta-
tions. WTP decreases as the number of family members increases. WTP is positively
related to the educational level of respondents, and women contribute more to
support the project, but the parameter associated with the age of the head of family
is not significant. Parameters associated with respondents’ years of residence suggest
that people who are more locally rooted support the project more strongly, although
the statistical significance is small.
Social aspects are not significant, since WTP is not influenced by membership of
environmental associations or by the political orientation of the respondents, but
psychological characteristics affect the stated WTP. Higher amenity perception is
associated with higher WTP, and familiarity with wind farm operation increases the
elicited WTP.
In the second equation the number of significant variables is notably reduced,
given that only five variables are significant at 5% in at least half of the estimated
models. Economic variables are significant and positively related to air quality risk
perceptions; this suggests that a higher standard of living is associated with higher
risk perception. This result is also confirmed by qol and educ; unfortunately, the
latter is often significant only at 10%.
Demographic variables also influence risk perception. In particular, agehf
increases the perception of the risk associated with the worsening of air quality.
This can be explained by the fact that older people are more sensitive to respiratory
diseases associated with air pollution. The qol variable is also positively associated
with risk perception. Conversely, people who have experienced a wind farm are less
sensitive to this type of risk. This may suggest that this class of respondents is more
familiar with this characteristic, tending to underevaluate the perception of this
externality, and that the perception of air quality improvement due to wind power
generation is influenced by psychological variables that contribute to modifying the
perception and the importance related to this positive externality.
The second equation system refers to respondents with negative attitudes toward
the project, who need to be compensated. Table 2.3 shows that also in this equation
system, the more significant models associated with lower certainty thresholds are
the models in which responses are weighted using an NCS. Economic variables that
positively affect WTA are the homownershio and the price of the bill,10 while
9
We estimate models in which spending on the electricity bill is interacted with higher-income
individuals identified by appropriate dummy variables. The interaction variable is never significant,
while the coefficient associated with the electricity bill remains positive and significant.
10
In this case we confirm comments we have made according to the other equation system.
Table 2.3 Estimated willingness to accept (WTA)

UnW_wta W_wta U_wta_10 U_wta_9 U_wta_8 U_wta_7 U_wta_6 U_wta_5
[Eq. 1: 0.0977a 0.1298b 0.1686 0.3198b 0.1399a 0.1759b 0.1822b 0.1337c
wta] bill
0.0582 0.0399 0.1228 0.0971 0.0735 0.0619 0.0615 0.0600
agehf 0.1324b 0.0726c 0.1747 0.1546c 0.2112b 0.1393b 0.0796a 0.1064c
0.0468 0.0320 0.1553 0.0726 0.0577 0.0480 0.0465 0.0465
sex 3.4735b 2.0643b 7.8378b 6.8727b 5.7162b 4.0501b 3.0054b 3.1313b
1.0554 0.7225 2.4157 1.6401 1.3670 1.1280 1.0962 1.0776
res_years 0.0711c 0.0816b 0.1655a 0.0986a 0.0261 0.0249 0.0940c 0.0723c
0.0363 0.0248 0.0936 0.0548 0.0444 0.0380 0.0373 0.0367
fam_s 0.6111 0.1654 2.0349 0.6893 0.1500 0.1966 1.2016c 0.9995a
0.5525 0.3782 1.3116 0.8539 0.7497 0.6330 0.5940 0.5716
incom 0.0160 1.2215b 1.6013 0.0889 0.0764 0.0134 1.7054b 0.5299
0.4956 0.3393 1.5516 0.9806 0.7660 0.6284 0.5680 0.5213
leftp 0.4769 2.4013b 7.3357c 4.6926c 1.3338 1.6685 1.2314 1.2436
1.1336 0.7760 3.0586 2.0329 1.5427 1.2617 1.2147 1.1750
qol 0.3974b 0.1070 0.6543c 0.2700 0.2008 0.2731a 0.1554 0.2988a
0.1545 0.1058 0.3266 0.2164 0.1768 0.1517 0.1519 0.1532
educ 0.2955c 0.1833a 1.0994b 0.0657 0.0058 0.0471 0.4588b 0.3606c
0.1466 0.1003 0.4084 0.2194 0.1857 0.1612 0.1506 0.1497
env_ass 4.1960b 0.6615 0.6951 1.6311 0.5221 1.7800 4.0879c 4.6189b
1.5042 1.0297 4.8187 3.0520 2.4846 1.8442 1.7599 1.6139
homeow 2.9023c 0.9196 0.5411 3.2940c 2.8324c 3.2476b 2.6330c 3.2749b
1.1814 0.8087 2.6176 1.6312 1.3640 1.1616 1.1591 1.1792
wfexp 0.2909 2.5677b 4.4150 4.1384c 3.5513c 4.0862b 3.9193b 3.6310b
0.8217 0.5625 2.8446 1.8772 1.5335 1.3655 1.3375 1.3240
_cons 0.9930 1.9683 5.5994 17.0310a 1.8907 2.0434 1.2504 0.5594
5.3342 3.6516 12.3956 8.8919 6.6092 5.7802 5.6611 5.4995
[Eq. 2: 0.0111 0.0111 0.0081 0.0287 0.0004 0.0182 0.0270c 0.0175
landrisk]
bill
0.0114 0.0114 0.0236 0.0215 0.0176 0.0146 0.0123 0.0117
agehf 0.0107 0.0107 0.0031 0.0051 0.0171 0.0114 0.0043 0.0078
0.0091 0.0091 0.0298 0.0160 0.0138 0.0114 0.0093 0.0091
sex 0.3639a 0.3639a 0.0978 0.7087a 0.7886c 0.5945c 0.4807c 0.4553c
0.2058 0.2058 0.4634 0.3626 0.3266 0.2667 0.2196 0.2105
res_years 0.0225b 0.0225b 0.0165 0.0251c 0.0130 0.0128 0.0320b 0.0289b
0.0071 0.0071 0.0180c 0.0121 0.0106 0.0090 0.0075 0.0072
fam_s 0.0681 0.0681 0.5145 0.2954 0.2188 0.1738 0.1425 0.0828
0.1078 0.1078 0.2516c 0.1888 0.1791 0.1497 0.1190 0.1117
incom 0.0432 0.0432 0.6059 0.2551 0.0587 0.0197 0.1019 0.0235
0.0967 0.0967 0.2977 0.2168 0.1830 0.1486 0.1138 0.1018
leftp 0.1425 0.1425 0.7213 0.7272 0.2966 0.4383 0.2739 0.0379
0.2211 0.2211 0.5868 0.4494 0.3685 0.2983 0.2434 0.2295
qol 0.1045b 0.1045b 0.0431 0.0617 0.0886c 0.0840c 0.0621c 0.0909b
0.0301 0.0301 0.0627 0.0478 0.0422 0.0359 0.0304 0.0299
(continued)

UnW_wta W_wta U_wta_10 U_wta_9 U_wta_8 U_wta_7 U_wta_6 U_wta_5
educ 0.1326b 0.1326b 0.1190 0.1592b 0.1034c 0.1087b 0.1293b 0.1266b
0.0286 0.0286 0.0784 0.0485 0.0444 0.0381 0.0302 0.0292
env_ass 1.0186b 1.0186b 1.2520 0.5418 1.0141a 1.0531c 0.8937c 0.8263b
0.2934 0.2934 0.9244c 0.6748 0.5935 0.4361 0.3526 0.3153
homeow 1.6813b 1.6813b 1.0009 1.6872b 1.6834b 1.8669b 1.7038b 1.7269b
0.2304 0.2304 0.5022 0.3606 0.3258 0.2747 0.2323 0.2303
wfexp 0.3646c 0.3646c 0.6298 0.3584 0.4601 0.6910c 0.7258b 0.6266c
0.1603 0.1603 0.5457c 0.4150 0.3663 0.3229 0.2680 0.2586
_cons 0.6799 0.6799 4.8189 0.1548 1.7296 0.4548 0.0621 0.3487
1.0404 1.0404 2.3780 1.9659 1.5789 1.3667 1.1343 1.0743
Obs
Eq. 1 346 346 44 116 157 204 294 320
Eq. 2 346 346 44 116 157 204 294 320
RMSE
Eq. 1 9.2620 6.3404 6.5894 7.6331 7.4186 7.2801 8.6193 9.0013
Eq. 2 1.8064 1.8064 1.2641 1.6876 1.7722 1.7214 1.7270 1.7584
R-sq
Eq. 1 0.2005 0.3001 0.6226 0.4126 0.3267 0.2783 0.2772 0.2389
Eq. 2 0.3501 0.3501 0.5371 0.4025 0.3618 0.3694 0.4123 0.3828
χ2(12)
Eq. 1 86.77 148.33 72.58 81.49 76.19 78.67 112.73 100.46
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Eq. 2 186.38 186.38 51.06 78.13 89.01 119.51 206.27 198.46
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Res corr 0.2930 0.1274 0.1810 0.1965 0.2270 0.2069 0.2307 0.2602
B-P test 29.701 5.612 1.441 3.334 8.090 8.736 15.647 21.658
χ2(1)
B-P test 0.000 0.018 0.2992 0.0068 0.0045 0.0031 0.0001 0.0000
-Prob-
WTA 26.54b 18.04b 1.74 26.69b 21.97b 23.93b 29.69b 32.98b
mean
2.13 1.39 12.62 4.24 3.93 3.29 2.01 2.21
lo.bo 22.36 15.31 23.00 18.38 14.27 17.49 25.74 28.64
up.bo 30.72 20.77 26.49 34.99 29.68 30.38 33.64 37.32
B-P Breush–Pagan, Obs observations, Res corr residual correlation, RMSE root-mean-square error,
R-sq R-squared
a
b
c
income, which is significant in only two models, is negatively related to WTA, as

highlighted by Groothuis et al. (1998). This is due to the diminishing marginal utility
of income. Respondents with higher incomes are less interested in the compensation
associated with the negative externalities related to the project. Demographic vari-
ables concur to explain respondents’ WTA. The age of the head of the household, the
level of education, and female gender contribute significantly in increasing the

compensation required, while the number of household members and years of
residence have limited impacts.
Social variables have no influence on WTA, and the psychological component is
significant only for familiarity with wind turbines, contributing to reducing the
amount of compensation. This result can be explained by taking into account the
fact that experience of living near a wind farm can familiarize respondents with
negative externalities (Parkhill et al. 2010) through adaptation to living in an altered
environment (Frantàl 2015).
The perception of risk for the landscape is influenced by psychological variables,
and while amenity increases the importance of the externality perceptions, the
experience of having lived near a wind farm greatly reduces the importance of this
attribute.
The variable years of residence is characterized by limited and positive impacts,
while women and more highly educated respondents perceive higher landscape risk.
Among economic variables, only homeown increases the landscape risk associated
with the project, while the other variables are not significant. This result suggests that
homeowners can be afraid that the project will decrease their property values.
2.5.3 Welfare Measures
Estimation of the parameters allows us to calculate the relative values of WTP and
WTA, combining the coefficients obtained from the SUR model and the mean values
of the right hand side (RHS) variables. The computation is done according to the
several models estimated. The estimated values are bimonthly, since the payment
vehicle proposed is the electricity bill. In accordance with the theory, WTA estimated
values are higher than those for WTP although the magnitude of this gap is smaller in
comparison with literature results. It is a well-known finding that scenarios exist in
which respondents do not overstate elicited WTA (Cooper and Osborn 1998).
Our results are in line with the literature, although our WTP is a little higher in
comparison with the cited papers. Higher WTP values are particularly significant for
two reasons.
First, these values have been obtained in the historical and social context of a
macroeconomic crisis, which was even worse in Umbria than the national average.
This means that despite the economic crisis, families are willing to support this type
of intervention to reduce the environmental impact caused by greenhouse gas
emissions in electricity generation.
Second, this result has been obtained in an area where there are two turbines and
the respondents have long experience of an operational wind farm.
The facts that more than 50% of respondents are in favor of wind farm enlarge-
ment and that over 70% of them state their willingness to contribute economically,
with a significant degree of certainty, symbolize an important political factor. The
presence of a wind farm is compatible with pro–wind power residents’ attitudes and
25
22.71
20 19.67
17.82 16.07 15.85
18.24 15.05 15.99 15.78
15 13.85 14.32
13.78 12.38 12.67 12.58
10 11.92 12.57
9.71 9.29 9.36 9.30
5 5.67
2.72
0.77
0
mean lo.bo up.bo
Fig. 2.5 Estimated willingness to pay (WTP)
50
40 33.64 37.32
30.72 34.99 29.68 30.38
30
26.54 20.77 26.49 26.69 29.69 32.98
20 22.36 21.97 23.93 25.74 28.64
18.04 18.38
10 15.31 14.27 17.49
0 1.74
-10
-20 -23.00
-30
mean lo.bo up.bo
Fig. 2.6 Estimated willingness to accept (WTA)
financial supporting behavior. Focusing on elicited WTA, the estimated values seem
to be in line with those in the few papers available, although the scarcity of such
papers reduces the significance of the comparison. According to all of the estimated
models, WTP ranges between EUR 9.29 and EUR 18.24 (Fig. 2.5).
More reliable values are those associated with weighted models and with the
models in which the thresholds lie between 5 and 7. Consequently, the valid interval
is narrower, ranging from EUR 12.38 to EUR 14.32. These results confirm the
impact of the certainty in the CV studies, given that it can double the amount of the
welfare measure, as arises if we compare U_wtp_8 and UnW_wtp models. Figure 2.6
shows that the compensation amounts required by respondents range from EUR 1.74
2.6 Conclusions 49
to EUR 32.98, highlighting the greater heterogeneity of the results due to the
different sizes of the samples. More robust results are associated with the weighted
model and with the models in which the thresholds lie between 6 and 8. This means
that the estimated WTA ranges from EUR 18.04 to EUR 29.69.
On average, the amount of compensation required is only slightly higher than the
related WTP. This is a reassuring result, given that according to some researchers
(Freeman 1993; Cooper and Osborn 1998), WTA is both one of the most appropriate
measures concerning facility-siting decisions in specific places and the most
incentive-compatible measure if respondents believe that their responses may influ-
ence policy decisions. Our scenario presents both characteristics, and our analysis
provides useful and robust results, suggesting that for a similar case study it is
possible to conduct WTA surveys to reduce a large number of conflicting scenarios
through the “siting decision.” This is very interesting in an European scenario in
which the most suitable onshore sites for wind installations are almost fully engaged
and the existing wind power plants are progressively aging.
2.6 Conclusions
In Europe, the current energy policy aims to increase the share of renewables in the
electricity generation mix in order to reduce emissions. In this scenario, wind energy
plays a central role, given that this source could continue to grow rapidly until 2020
due to the cost reductions that have taken place over the last decade (EEA 2017).
Unfortunately, barriers such as spatial planning issues, local opposition, and admin-
istrative delay often slow down onshore wind energy deployment. In this context,
wind farm enlargement represents a useful strategy to overcome these barriers in a
context of progressive aging of facilities and/or in regions characterized by a
particularly evaluable landscape.
The case study proposed reflects all of these characteristics assuming relevance in
terms of energy policy. We have analyzed a territory characterized by a valuable
man-made landscape in which turbines have been operational for more than
15 years. The majority of the sample agree with the proposed enlargement, and
they are willing to economically support the project due to the positive externality
associated with wind power generation. The rest of the sample require compensa-
tion, mainly due to the visual intrusion externality, declaring a willingness to accept
(WTA) that is only 20% higher than the willingness to pay (WTP). Further, in this
subsample, respondents who have long experience of an operational wind farm
require lower compensation.
These are very interesting results, given the appropriateness of the WTA mea-
sures and its incentive-compatible characteristics. Indeed, the WTA elicitation
format could greatly aid the achievement of European climate policy targets to
spur the deployment of a mature and low-cost technology such as wind power
generation.
Finally, in the “proximity versus reverse proximity” debate, our results support
the hypothesis that spatial proximity seems not to be a crucial factor in the wind farm
location process; more important factors are the environmental and economic
benefits.
References
ANEV (2017) Eolico in Italia: inerzia o full support? [in Italian]. http://www.anev.org/wp-content/
uploads/2017/05/Brochure-ANEVelements195x280_2017web.pdf. Accessed 20 Mar 2017
Bergmann A, Hanley N, Wright R (2006) Valuing the attributes of renewable energy investments.
Cooper JC, Osborn CT (1998) The effect of rental rates on the extension of conservation reserve
program contracts. Am J Agric Econ 80:184–194
Dimitropoulos A, Kontoleon A (2009) Assessing the determinants of local acceptability of wind-
farm investment: a choice experiment in the Greek Aegean Islands. Energy Policy
37:1842–1854
EEA (2017) Renewable energy in Europe 2016—recent growth and knock-on effects. http://www.
eea.europa.eu/publications/renewable-energy-in-europe-2016/download. Accessed 7 June 2017
Ek K (2005) Public and private attitudes towards “green” electricity: the case of Swedish wind
power. Energy Policy 33:1677–1689
Ek K, Matti S (2014) Valuing the local impacts of a large scale wind power establishment in
northern Sweden: public and private preferences toward economic, environmental and socio-
cultural values. J Environ Plan Manag 58:1327–1345
Evans MF, Flores NE, Boyle KJ (2003) Multiple-bounded uncertainty choice data as probabilistic
intentions. Land Econ 79:549–560
Frantàl B (2015) Have local government and public expectations of wind energy project benefits
been met? Implications for repowering schemes. J Environ Policy Plan 17:217–236
Freeman AM (1993) The measurement of environmental and resource values. Resources for the
Future, 1st Ed. Washington, DC
Frew E, Wolstenholme J, Whynes D (2004) Comparing willingness-to-pay: bidding game format
versus open-ended and payment scale formats. Health Policy 68:289–298
Friedl C, Reichl J (2016) Realizing energy infrastructure projects: a qualitative empirical analysis of
local practices to address social acceptance. Energy Policy 89:184–193
Groothius PA, Groothuis JD, Whitehead JC (2008) Green vs. green: measuring the compensation
required to site electrical generation windmills in a view shed. Energy Policy 36:1545–1550
Groothuis PA, van Houtven G, Whitehead JC (1998) Using contingent valuation to measure the
compensation required to gain community acceptance of a LULU: the case of a hazardous waste
disposal facility. Public Finance Rev 26:231–249
Huijts NMA, Molin EJE, Steg L (2012) Psychological factors influencing sustainable energy
technology acceptance: a review-based comprehensive framework. Renew Sust Energ Rev
16:525–531
Jones CR, Eiser JR (2010) Understanding ‘local’ opposition to wind development in the UK: how
big is a backyard? Energy Policy 38:3106–3117
Jones CR, Orr BJ, Eiser JR (2011) When is enough, enough? Identifying predictors of capacity
estimates for onshore wind-power development in a region of the UK. Energy Policy
39:4563–4577
Kaldellis JK, Kapsali M, Kaldelli E, Katsanou E (2013) Comparing recent views of public attitude
on wind energy, photovoltaic and small hydro applications. Renew Energy 52:197–208
References 51
Leung DYC, Yang Y (2012) Wind energy development and its environmental impact: a review.
Loomis J (2011) What’s to know about hypothetical bias in stated preference valuation studies? J
Econ Surv 2:363–370
Loomis JB (2014) Strategies for overcoming hypothetical bias in stated preference surveys. J Agric
Resour Econ 39:34–46
Lunander A (1998) Inducing incentives to understate and to overstate willingness to pay within the
open-ended and the dichotomous-choice elicitation formats: an experimental study. J Environ
Econ Manag 35:88–102
Mattmann M, Logar I, Brouwer R (2016) Wind power externalities: a meta-analysis. Ecol Econ
127:23–36
Meyerhoff J (2013) Do turbines in the vicinity of respondents’ residences influence choices among
programs for future wind power generation? J Choice Model 7:58–71
Meyerhoff J, Ohl C, Hartje V (2010) Landscape externalities from onshore wind power. Energy
Policy 38:82–92
Mirasgedis S, Tourkolias C, Tzovla E, Diakoulaki D (2014) Valuing the visual impact of wind
farms: an application in South Evia, Greece. Renew Sust Energ Rev 39:296–311
Mitchell RC, Carson RT (1989) Using surveys to value public goods: the contingent valuation
method. Resources for the Future, Washington, DC
Moran D, Sherrington C (2007) An economic assessment of windfarm power generation in
Scotland including externalities. Energy Policy 35:2811–2825
Mozumder P, Helton R, Berrens RP (2009) Provision of a wildfire risk map: informing residents in
the wild land urban interface. Risk Anal 29:1588–1600
Nimbyforum (2016) Osservatorio media permanente Nimby Forum [in Italian]. http://www.
nimbyforum.it/. Accessed 21 May 2017
Park T, Loomis JB (1996) Joint estimation of contingent valuation survey responses. Environ
Resour Econ 7:149–162
Parkhill KA, Pidgeon NF, Henwood KL, Simmons P, Venables D (2010) From the familiar to the
extraordinary: local residents’ perceptions of risk when living with nuclear power in the
UK. Trans Inst Br Geogr 35:39–58
Rand J, Hoen B (2017) Thirty years of North American wind energy acceptance research: what
have we learned? Energy Res Soc Sci 29:135–148
Read DL, Brown RF, Thorsteinsson EB, Morgan M, Price I (2013) The theory of planned behaviour
as a model for predicting public opposition to wind farm developments. J Environ Psychol
36:70–76
Riddel M, Loomis JB (1998) Joint estimation of multiple CVM scenarios under a double bounded
questioning format. Environ Resour Econ 12:77–98
Rygg BJ (2012) Wind power—an assault on local landscapes or an opportunity for modernization?
Shaikh SL, Sun L, van Kooten GC (2007) Treating respondent uncertainty in contingent valuation:
a comparison of empirical treatments. Ecol Econ 62:115–125
Slattery MC, Johnson BL, Swofford J, Pasqualetti MJ (2012) The predominance of economic
development in the support for large-scale wind farms in the US Great Plains. Renew Sust Energ
Rev 16:3690–3701
Soderqvist T, Soutukorva A (2006) An instrument for assessing the quality of environmental
valuation studies. Swedish Environmental Protection Agency, Stockholm https://www.
naturvardsverket.se/Documents/publikationer/620-1252-5.pdf. Accessed 11 Feb 2010
Strazzera E, Mura M, Contu D (2012) Combining choice experiments with psychometric scales to
assess the social acceptability of wind energy projects: a latent class approach. Energy Policy
48:334–347
Swofford J, Slattery M (2010) Public attitudes of wind energy in Texas: local communities in close
proximity to wind farms and their effect on decision-making. Energy Policy 38:2508–2519
Wüstenhagen R, Wolsink M, Bürer MJ (2007) Social acceptance of renewable energy innovation:

an introduction to the concept. Energy Policy 35:2683–2691
Zellner A (1962) An efficient method of estimating seemingly unrelated regressions and tests for
aggregation bias. J Am Stat Assoc 57:348–368
Zerrahn A (2017) Wind power and externalities. Ecol Econ 141(C):245–260. https://doi.org/10.
1016/j.ecolecon.2017.02.016
Chapter 3
Consumers’ Willingness to Pay
for Renewable Electricity in Italy:
A Comparative Analysis
Abstract European Union (EU) directive 2009/72/CE imposes environmental and

energy targets on European countries. The goal of Italy is to attain 26.4% renewable
electricity (RE) production from renewable energy sources (REnS) by 2020. This
goal imposes an extra cost on households; consequently, it is important to estimate
their willingness to pay (WTP) to attain this target. Our research is based on a
nationwide survey of households conducted in November 2007 in Italy, explicitly
considering uncertainty and the compulsory burden on the electricity bill. The results
obtained with different models indicate that there is noticeable WTP among Italian
households for the RE goal and that the estimated WTP differs according to
uncertainty. Indeed, the median WTP is between EUR 4.62 and EUR 8.05 every
two months per household. In this chapter, the relevance of these findings today is
discussed.
Keywords Willingness to pay · Renewable electricity · Contingent valuation ·

Uncertainty · Economic downturn · Policy evaluation
3.1 Introduction
Attention to paying for the environment seems to remain an important issue for
Italians, and the development of renewable energy (REn) is currently considered a
crucial target. Indeed, in 2014, 90% of respondents to the Eurobarometer survey
judged that it was important or very important to set governing targets increasing
REn consumption by 2030. Despite the financial crisis, increased REn use is still a
shared goal among Italians, given that according to the Eurobarometer survey,
Italians’ share of agreement with European Union (EU) Directive 2009/72/CE
goals has been on average higher than 64% since 2008.
This directive, known as the “climate and energy package,” has set three targets
for 2020 (known as “20–20–20”): a 20% reduction in polluting emissions, achieve-
ment of an energy portfolio with a 20% share of renewables, and 20% energy
consumption savings. According to this directive the Italian goal is to attain a 17%

54 3 Consumers’ Willingness to Pay for Renewable Electricity in. . .
share of renewable energy sources (REnS), which means 26.4% of electricity

produced by REnS.
In this context, new binding constraints have been set by the EU. First, by the
COM(2014)15 released on January 22, 2014, the EU has issued a nonlegislative
resolution (2013/2135 (INI)) outlining three targets: (1) an increase in the REnS
share of final consumption by 30%; (2) a 40% reduction in greenhouse gas emis-
sions; and (3) a 40% reduction in energy consumption. Later, on October 23–24,
2014, the European Council approved new energy climate objectives for 2030. The
targets are as follows: a reduction of 40% in greenhouse gas emissions, with binding
targets for Member States for non–Emissions Trading System (non-ETS) sectors; an
increase of 27% in the REnS share of final consumption of energy (this target is
binding only at the European level, not at the Member State level); and a 27%
increase in energy efficiency (currently, this target is not binding but it could be
revised with a rise to 30%).
As illustrated in Chap. 1, the positive perceptions of REnS in renewable electric-
ity (RE) production are due to reduced polluting emissions and to the preservation of
fossil fuel. These pros are balanced by uncertainty in electricity generation and by
higher generation costs, which prevent widespread uptake of RE; consequently,
public funding is needed to support RE development. Anyway, if consumers take
into account environmental issues, they could be likely to start paying for RE or to
increase the premiums that they are willing to pay for RE, reducing public financing
over time. In this context, it becomes important to estimate this price premium to
assess how the market is able to support the RE policy, considering some crucial
elements.
To better define the target of the research we deal with three aspects related to
RE. The first is the uncertainty. RE is a complex good; consequently, according to
Salmela and Varho (2006), uncertainty arises in an individual’s evaluation process.
Further, in the RE market, uncertainty characterizes the institutional context
(Jacobsson and Bergek 2004) and influences both the supply side, including gener-
ation (de Vries et al. 2007), and the demand side (Nemet 2009). In addition, the
stated preference method is itself characterized by uncertainty (Accent and Rand
Europe 2010; Broberg and Brännlund 2008; Evans et al. 2003; Haab and McConnell
1997; Loomis 2011; Murphy et al. 2005b; Pearce et al. 2008; Wang and He 2011;
Wang and Whittington 2005; Welsh and Bishop 1993). Nevertheless, few
researchers have explicitly included uncertainty in their contingent valuation
(CV) studies on RE (Ackura 2015; Vossler et al. 2003; Whitehead and Cherry 2007).
In this chapter, among several sources of uncertainty associated with the evalu-
ation process (Wang 1997a, b), we approach uncertainty related to the individual’s
preferences and characteristics, using ex ante and ex post uncertainty bias correction
methods (Loomis 2011).
The second aspect is related to end users. In Italy, RE subsidies are paid by all
consumers by a feed-in tariff mechanism; in other words, all consumers pay for an
increasing share of REnS. Consequently, the consumer’s sovereignty is reduced
such as in many network goods and services. Anyway, to the best of our knowledge,
with the exception of the study by Grösche and Schröder (2011), previous studies
have not considered that people are committed to pay for RE. In our questionnaire
we inform respondents about the feed-in tariff mechanism, together with the actual
and current cost of REnS paid by each consumer.
Finally, there are several types of consumers in the electricity market, who differ
in terms of usage, type of consumption, and electricity bill amounts. In this chapter
we only consider households, which are the mainly component of low-voltage
usage.
This chapter has three aims. First, we estimate households’ willingness to pay
(WTP) for RE, explicitly taking into account the issues discussed above. The
aggregate WTP obtained from CV studies is compared with the cost of the 20–20–
20 Italian RE target to assess its market sustainability. Furthermore, we increase the
knowledge on RE by examining related households’ perceptions, knowledge, and
attitudes in Italy.
Second, we compare our results with those of other similar studies conducted
worldwide. This is useful to contextualize our results in a worldwide perspective.
Thirdly, according to the first chapter framework, using two approaches to
appraise consumers’ WTP for RE, we provide more insights into the impact of
different methods on econometric results.
This chapter is structured as follows. Section 3.2 describes the related literature.
Section 3.3 presents the status of RE in Italy, including the incentive mechanisms
and the national cost of the 20–20–20 target. Section 3.4 provides the theoretical and
econometric framework and presents details of the survey design and collected data.
Section 3.5 discusses the empirical results and current policy implications. In
Sect. 3.6 we compare our results in a worldwide perspective. Section 3.7 concludes
the chapter. Additional details on the theoretical and econometric models used and
on the data collected are provided in the Appendix.
Several studies (Table 3.1) have focused on the financial viability and economic
sustainability of government programs for REnS development, generally finding two
main results: (1) a divergence between stated and actual consumer behavior in green
energy markets; and (2) moderate consumer WTP (Diaz-Rainey and Ashton 2008).
In the empirical works regarding RE, several elicitation formats for contingent
valuation (CV) are used. The main formats are single bound and double bound
dichotomous choice (SBDC and DBDC, respectively) and a payment card (PC). In
the literature there is a lack of consensus on the appropriate format to use; conse-
quently, we take into account all of them. In DBDC, respondents are asked two
closed-ended questions, which include a binary response (“yes” or “no”) to the initial
and follow-up questions. This format increases statistical efficiency but it is not bias
free; indeed, there exists strategic behavior for respondents who answer “yes” to the
first offer price because some answers to the second bid appear to be incoherent
Table 3.1 Willingness to pay (WTP) in the renewable electricity (RE)–related literature
56
Sample Survey Uncertainty

Study Country Data years size type Method Main findings correction
Ackura (2015) UK 2008–2009 2000 Internet CV-DBDC; zero-inflated Under mandatory payment, respon- Yes
dents are more supportive and less
uncertain than under a voluntary pay-
ment scheme (information on elicited
WTP is not provided)
Batley et al. UK 1996 (Mar) 746 Mail CV-OE; multiple regression Target: national goal of 10% electricity No
(2000) and factor analysis from REnS according to DETR climate
change draft UK program; households’
WTP: mean premium of 19.11%
(35.85% of consumer sample)
Batley et al. UK 1997 (Mar); 742; 692 Mail CV-OE; correlation Target: national goal of 10% electricity No
(2001) 1999 (Apr) analysis from REnS according to DETR climate
change draft UK program; households’
WTP: mean premium 16.6% (34% of
citizen sample), mean premium 18.5%
(34.86% of consumer sample) for
100% RE; consumers’ and citizens’
WTP are fare from national target
Bigerna and Italy 2007–2008 1600 Internet Bidding game; interval Target: 26% of RE by 2020 (+11%); Ex post;
Polinori (Nov) regression mean households’ WTP: numerical
(2013) EUR 40.90–50.40/year, around 35% of likelihoods
national target cost depending on
uncertainty
Bigerna and Italy 2007 (Nov) 1019 Internet CV-MBDC Target: 26% of RE by 2020 (+11%); Ex post;
Polinori mean households’ WTP: numerical
(2014) EUR 12.76–15.09/bill, around 50% of likelihoods
national target cost depending on
uncertainty
3 Consumers’ Willingness to Pay for Renewable Electricity in. . .
Bollino (2009) Italy 2006 (Nov) 1601 Internet SPC treated as dichotomous Target: 22% of RE by 2010 (+7%); Ex post;
choice model; probit mean households’ WTP: numerical
EUR 2.44–9.39/2 months, likelihoods
15.97–61.46% of national target cost
depending on uncertainty
Borchers et al. USA 2006 (May) 625 Face to CEx; nested logit model Mean households’ WTP: No
(2007) face (a) USD 1.08–21.54/month depending
on scenarios proposed and green

sources, (b) USD 8.44–17.00/month
for generic RE depending on payment
methods and scenarios
Byrnes et al. USA 1992 600–500 Telephone CEx; censored regression Mean households’ WTP: USD 1.63 No
(1999) (Sep)–1994 and mail and 1.72/month
(Nov)
Dagher and Lebanon 2013 600 Face to CV-OE Target: 12% of RE by 2020; four sce- No
Harajli (2015) face narios for better integrating RES into
the electricity system; median WTP:
USD 20–50/month, monthly mean
WTP: USD 24.79–59.50
Goett et al. USA 2000 1205 Telephone CEx; mixed-logit regression Mean households’ WTP: USD 0.0145/ No
(2000) and mail kWh, 100% wind; USD 0.02/kWh,
100% hydro
Grösche and Germany 2008 2948 Internet CEx; random parameter Target: 30% RES scenario (+15%) No
Schröder model replacing nuclear with RES by 2020;
(2011) median household WTP: EUR 0.0203/
kWh, EUR 0.0237/kWh if fossil fuel is
also replaced; financial scope to sup-
port renewable fuels is basically
exhausted
Guo et al. China 2010 (May) 700 Face to CV-SBDC Average WTP for RE: USD 2.70–3.30/ No
(2015) face month
57
(continued)
58
Sample Survey Uncertainty

Study Country Data years size type Method Main findings correction
Hanemann Spain 2009 (Nov– 233 Telephone CV-SBDC Median households WTP: With “do
et al. (2011) Dec) EUR 29.91/month not know”
option
Hansla et al. Sweden 855 Mail PC Households WTP: sample majority No
(2008) SEK 0.001–0.02/kWh
Harajli and Lebanon 2013 200 Face to CV-OE Commercial sector promises to be a No
Gordon (2015) face significant propagator of RE only if RE
sources can displacediesel gensets
(through use of battery storage), which
is possible for smaller offices; monthly
WTP lies between USD 210 and
USD 613
Ivanova Australia 2004 (Aug) 213 Mail CV-OE Mandatory renewable energy target: No
(2005) 12.5% (+2%); mean household WTP:
AUD 22/quarter if policy support,
AUD 28/quarter if voluntary payment
Ivanova Australia 2004 (Aug) 820 Mail CV-OE; latent class Mandatory renewable energy target: No
(2012) analysis 12.5% (+2%); mean WTP: “concerned
class” USD 29/quarter, “protest class”
USD 13/quarter, “WTP class” USD 36/
quarter
Kim et al. South 2010 (Aug) 720 Face to CV-DBDC; spike model Target: renewable portfolio standard by No
(2012) Korea face 2012; mean households WTP:
USD 1.35/month, 58.2% of govern-
ment budget for renewable program
Nomura and Japan 2000 (Oct) 379 Mail CV-DBDC; Weibull distri- Median household WTP: No
Akay (2004) bution function JPY 2000/month
Wiser (2007) USA 2001 (Feb– 1574–202 Face to CV-SBDC WTP depends on payment and provi- No
May) face sion context
Yoo and Kwak South 2006 800 Face to CV-DBDC: (a) parametric Target: 7% of RE by 2011 (+6.8%); No
(2009) Korea face (spike model), (b) nonpara- mean household WTP:
metric (interpolation model) (a) USD 1.8–2.0/month,
(b) USD 1.6–2.7/month; annual bene-
fits USD 150.5–194.2 million; no
information about target achievement

Zhang and Wu China 2010 652–536 E-mail and CV; PC; mixed logit Mean household WTP: No
( 2012) (May–Jun) mail USD 1.15–1.51/month
Zografakis Greece 2006 1440 Face to Face to face interview Mean household WTP: No
et al. (2010) (Crete) (Sep)–2007 face EUR 16.33/quarter
(Feb)
Zoric and Slovenia 2008 450 Internet CV-OE; Tobit, probit; trun- Mean household WTP: EUR 4/month, No
Hrovatin and field cated regression median EUR 4/month
(2012) interviews
CEx choice experiment, CV contingent valuation, DBDC double bound dichotomous choice, DETR Department of the Environment, Transport and the Regions,
MBDC multiple bound dichotomous choice, OE open ended, PC payment card, RES renewable electricity sources, SBDC single bound dichotomous choice,
SPC stochastic payment card
59
(De Shazo 2002). This means that the incentive properties of SBDC are not shared
by the DBDC format (Genius and Strazzera 2005).
Nomura and Akay (2004), using a DBDC format, investigate consumers’ WTP to
increase RE, by sending out a postal questionnaire in a sample of Japanese cities
before introduction of green energy funding. A WTP median estimate is about
JPY 2000 per month, one of the highest for Japan. This is because REnS are
becoming more familiar and their efficiency is becoming more widely known. A
similar result is obtained by Zografakis et al. (2010) in major Cretan cities in Greece,
confirming that households are well disposed toward RE implementation. The mean
WTP is calculated at EUR 70.76 per person per year. The results highlight that
climate change mitigation is firmly linked with more REnS (Hanemann et al. 2011;
Roe et al. 2001). Byrnes et al. (1999) assess the reliability of CV by the method’s
criterion validity. They employ two surveys that refer to two different programs—
the Colorado and Wisconsin Renewable Energy Programs—using telephone and
mail questionnaires.
Even though 73% of respondents support REnS investment, only 13% participate
in the program. Yoo and Kwak (2009) apply a DBDC method to evaluate RE in
South Korea. Even if a considerable number of respondents refuse to pay for a green
electricity policy, they find a monthly WTP of USD 1.80. This figure denotes an
increasing interest in REn combined with actual willingness to support RE. Ackura
(2015) investigates how payment options impact on WTP for RE and on respon-
dents’ certainty of paying their stated bids. Applying a DBDC method it arises that a
mandatory scheme is more effective than a voluntary one in terms of both the WTP
amount and the certainty of WTP. Using the same method, Kim et al. (2012) find that
South Koreans have an annual WTP of USD 277.4 million for RE, which is
approximately 58% of the government’s budgetary allocation.
An SBDC format is used by Ivanova (2005), Hanemann et al. (2011), and Wiser
(2007). The main advantages of this elicitation format are that it is incentive
compatible, cognitively manageable, and free of psychological biases, even if it is
not statistically efficient. Ivanova (2005) use a mail questionnaire to evaluate the
market sustainability of the Australian federal government’s renewable program.
The results show that 65% of Queensland respondents are willing to pay AUD 22 per
quarter to increase REnS use from 10% to 12%, but these figures are way off target.
Enlarging the sample and taking into account three respondent categories by latent
class analysis (“concerned,” “protest,” and “willing to pay”), Ivanova (2012) finds a
wider WTP interval ranging from EUR 42.37 to EUR 117.35/year. Hanemann et al.
(2011) find that Spanish households support the implementation of a green electric-
ity program to mitigate climate change. Using a telephone survey they estimate a
monthly household median WTP of EUR 29.90 against an average monthly bill of
EUR 40. The lead-up to the Copenhagen summit could explain this exceptional high
value. Finally, Wiser (2007) explores WTP under different payment methods (vol-
untary and mandatory) and under different ways in which REnS could be provided
(private and government) in the USA. The concept of free riding is used to explain
large differences between environmental attitudes reported by the general population
and weak actions of those groups engaged in voluntary environmental behavior.
The results confirm that the elicited WTP for REnS is higher under a mandatory
payment method then under a voluntary one, such as in Yoo and Kwak (2009), and
WTP under private supply is greater than WTP under government provision.
A PC method has been recommended by government guidelines and scholars
(Bateman et al. 2001; Champ et al. 2003; Holt and Holt 2004). This method avoids
the anchoring bias due to the CV approach; further, it maintains a direct question
approach, increasing the response rate for WTP questions. The PC method assumes
that respondents have a valuation distribution in mind instead of single-point
economic value estimation. The cons are value cues and range bias (Mitchell and
Carson 1989), but Rowe et al. (1996) suggest that when the range is sufficiently large
it does not constrain respondents and consequently it is possible to avoid range and
centering bias. According to these results, Zhang and Wu (2012) establish a starting
point and its interval through a pilot survey. They also distinguish respondents who
do not display a WTP, finding that the majority of respondents have a conservative
WTP mean of approximately USD 1.15–1.51 per month. Using other methods,
several studies focus on WTP for RE at national and regional levels. Among
these, Batley et al. (2001) find that 34–35% of respondents declare that they are
willing to pay an additional 16.6–18.5% of their actual expenditure to support RE,
but this effort is insufficient to achieve the national target. Bollino (2009) finds that
Italian households support paying 30% of the national target annual cost. Bigerna
and Polinori (2012, 2013, 2014) find similar results. In Germany, Grösche and
Schröder (2011) find that “the possible financial scope to support renewable fuels
is basically exhausted.”
In two recent publications, Harajli investigates WTP for RE in Lebanon in both
households (Dagher and Harajli 2015) and firms (Harajli and Gordon 2015), using
SBDC. In the first case the results show that RE is an important option for house-
holds to substitute for diesel generators in electricity production. The WTP amounts
elicited are comparable to the results of other similar research. Information plays a
crucial role; improving RE knowledge results in higher WTP. A commercial sector
survey highlights that firms can play an important role in RE diffusion. Nevertheless,
the essential prerequisite is full replacement of diesel generators by RE. The answers
provided by respondents confirm the need for well-designed policy tools, which
have to be transparent, clear, and accountable. Guo et al. (2015) use the same method
to value WTP for RE in Beijing. In this study, environmental concerns are the main
motivation in supporting RE, while low income determines negative WTP. The
results point out that also in Beijing well-known sociodemographics and institutional
characteristics matter. Knowledge of RE, a positive attitude toward RE, and the
payment vehicle increase the stated WTP; furthermore, a mandatory scheme con-
firms its superiority to a voluntary scheme. The authors underline the high WTP for
RE of Beijing residents, suggesting that opportunities exist for the local government
to develop RE.
Other studies, even if they estimate WTP, do not provide the cost of the renew-
able target proposed (see, among others, Mozumder et al. (2011), Álvarez-Farizo
and Hanley (2002), Roe et al. (2001), and Longo et al. (2012)). Finally, we also
analyze studies that take into account uncertainty in the RE WTP elicitation format.
A way to introduce uncertainty in CV studies is combining an ordinary PC and a

polychotomous choice question (Broberg and Brännlund 2008). This method is
employed by Welsh and Poe (1998), Evans et al. (2003), and Alberini et al.
(2003). All of these authors directly incorporate a certainty level into the discrete
choice decision framework. In other studies, introduction of a certainty level can be
regarded as an ex post adjustment to the dichotomous choice response (Champ et al.
1997).
Using an RE pricing program, Vossler et al. (2003) explore these methods; in
particular, these authors use a field validity test to compare the contributions of
Champ et al. (2003) and Welsh and Poe (1998), but the results do not provide a
definitive answer as to which of these methods is preferable.
All considered, uncertainty is still an open research question in CV studies. In this
chapter we use two different approaches handling data obtained from multiple bound
dichotomous choice (MBDC) data in order to check the robustness of our results.
3.3 Renewable Electricity in Italy
The International Energy Agency’s (IEA’s) World Energy Outlook ( 2014) predicted
that renewable energy would become the world’s second largest source of electricity
generation by 2015 and by 2035 achieving coal as the main source in fuels energy
mix. Actually, renewables had already achieved second place worldwide by 2014,
although in 2013 RE rose worldwide more slowly than predicted, mainly due to
lower hydropower availability (IEA 2014). The majority of REn is consumed in the
residential, commercial, and public service sectors, especially in Organization for
Economic Co-operation and Development (OECD) countries.
In this context, Europe plays a leading role; indeed, according to the New Policies
Scenario,1 RE is expected to increase from 24% in 2010 to 44% in 2035 and among
the leading countries, Italy increased its share from 14% to 27% between 2000 and
2011, becoming the world’s second-biggest user after Germany. Italy is also
included in the top five countries for total investment in renewables and it is in the
top seven countries for nonhydro RE capacity (REN21 2015). In 2014, Italy
confirmed its pre-eminent position in Europe. It is the fourth leading country
worldwide for nonhydro renewable power capacity after China, the USA, and
Germany (REN21 2015).
Focusing on RE, Italy is a leading Member State in the EU—together with
Germany, Spain, France, the UK, and Sweden, accounting for 70% of EU27
production (Jäger-Waldau et al. 2011). Furthermore, Italy is the third leading
1
A scenario in the World Energy Outlook (IEA 2014) takes account of broad policy commitments
and plans that have been announced by countries, including national pledges to reduce greenhouse
gas emissions and plans to phase out fossil energy subsidies, although the measures to implement
these commitments are yet to be identified or announced.
3.3 Renewable Electricity in Italy 63
biodiesel market, the second leading country for renewable consumption in the
transport sector, and the leading country for heat pump use in Europe (REN21
2015). The Italian leading position is mainly explained by public expenditure and
support policies.
In Italy, incentive mechanisms are based on both market regimes (such as the
quantity-oriented mechanism, or “green certificates”) and administrative regimes
(such as the price-oriented mechanism, or “feed-in tariffs,” capital incentives, and
tax credit incentives). In particular, these mechanisms include the following
(Bigerna and Polinori 2014): (1) incentive rates for REn and assimilated sources
(before 1999); (2) a system of green certificates for REnS (since 1999); (3) a system
of feed-in tariffs for REn installations to power less than 1 MW (200 kW for wind
power) since 2005; (4) a feed-in premium for solar power plants, particularly for
photovoltaic systems (since 2007); and (5) capital grants (local) for some renewable
sources (since 2003). Government intervention through taxes and subsidies trans-
lates into higher energy prices in the short run, affecting both households and
business. In particular, these policy measures are a burden on electricity bills;
consequently, end users are already paying part of the cost to increase the REnS
share in the fuel electricity mix.
3.3.1 Cost of Renewable Electricity in Italy
Since 1997 (Regulatory Authority for Electricity and Gas (RAEG) decree 70/97)
there has been a component of electricity bills (named the A3 component) in Italy
that covers the cost of REnS use in electricity generation. Beyond REnS, the A3
component includes several types of subsidies for different kinds of power plant
production; consequently, the A3 component overestimates the actual support pro-
vided to REnS.
Comparison of the data related to 3 years characterized by a different degree of
REnS development (2007, 2010, and 2013) allows us to investigate the progressive
and massive development of REnS in Italian electricity production.
In 2007, the A3 component amounted to 3.33 million EUR—that is, a mean
additional cost of between EUR 2.20 and EUR 2.60 per month per household—but if
we consider fees and charges strictly related to RE, these figures decrease to between
EUR 0.95 and EUR 1.30. For a better understanding of the Italian end user market it
is pointed out that in 2007, households represented 21% of total final electricity
consumption and used 65% of the total points of injection in the Italian electricity
market (RAEG 2008).
In 2010 there was a substantial increase of REnS in Italy. Considering all of the
items present in the electricity bill, the total amount was EUR 5.81 billion of EUR
and the A3 component was EUR 3.97 billion, but REnS subsidies were only 69% of
the component—that is, a figure of EUR 2.76 billion. This means that in 2010, the
average additional cost attributable to REnS ranged from EUR 1.40 to EUR 2.50 per
month per household.
In 2013 the A3 component amounted to EUR 12 billion. This figure refers to all
charges included in the A3 component; the REnS burden amounted to EUR 6.7
billion for photovoltaic sources and EUR 3 billion for other sources. This means that
in 2013 each household spent (via its electricity bill) EUR 9.85–17.60 bimonthly to
support REnS. This information has been used to make interviewees aware of the
compulsory burden on the electricity bill. In Table 3.2 we illustrate the structure of
the A3 component in 2007, the year of the survey.
Finally, for the aim of this chapter it is important to estimate the cost of attaining
26.4% of electricity being produced using REnS by 2020. In Italy, among several
estimated figures, we use the amount of EUR 3.5 billion per year, according to the
Centre for Research on Energy and Environmental Economics and Policy (IEFE 2009).
3.4 Methods and Data
In CV analysis, a policy scenario is proposed to interviewees and their WTP for

attaining national goals is then elicited. As in all cases of CV study, there is a risk of
bias (Loomis 2011). However, it has also been shown in the literature that a well-
designed and carefully administered survey provides consistent, coherent, and
credible information on WTP.
3.4.1 Theoretical Model
In this study, we consider the Italian household as the typical consumer unit. It is
assumed that households maximize their utility subject to budget constraints.
The demand for “RE use” can be viewed like the demand for any other good or
service, and we therefore model consumer choice within the utility (expenditure)
maximization (minimization) framework.
If we allow expenditure (E to be a function of “RE use” services2 (R), private
goods (XP), and composite of public goods (XG) subject to the utility (U ) constraint,
we obtain (see Appendix for further details):
2
The literature point out that the definition of RE is controversial because it is a new concept for
consumers who need more information to buy green electricity (see Salmela and Varho (2006) and
Diaz-Rainey and Ashton (2008)). According to these researchers, the green energy market is poorly
defined and this implies consumer confusion. Finally, Litvine and Wüstenhagen (2011) focus on the
perceived benefit as a key factor to increase the purchasing of green electricity. These characteristics
reduce the related consumers’ WTP. Consequently, we conduct focus group analyses using people
who are well informed about the characteristics of green electricity, to avoid consumer confusion
due to lack of information.
3.4 Methods and Data
Table 3.2 A3 component analyzed according to different types of users (2007)

Low voltage Power (KW)
Household use Other final use Medium voltage High voltage
Feature 3 >3 10 100 500 1000 3000 10,000
Use (h/year) 880 1167 1200 1500 2000 2500 2500 3500
Annual consumption (MWh/year) 2.64 3.5 12 150 1000 2500 7500 35,000
Average rate A3 component (EUR/kWh) 0.0073 0.0075 0.0121 0.0093 0.0075 0.0074 0.0068 0.0067
Monthly expenditure (EUR/month) 1.60 2.20 12 116 627 1545 4236 19,481
REnS only 0.95 1.30 7 69 370 913 2502 11,507
Tax levy (million EUR) 630 1228 1192 279
Total (million EUR) 3329
This table shows our elaboration of data from Gestore dei Servizi Energetici (GSE) (an Italian electricity services operator)
REnS renewable energy sources
65
Min E ðPP ; PR ; X G Þ ð3:1Þ
subject to: U ¼ U (XP, R, XG)

where PP and PR are the prices of private goods and of “RE use” services,
respectively. Given the utility constraint, the representative household faces expen-
ditures for “RE use” services, private goods, and composite public goods; thus, the
household will attempt to minimize the following expenditure function:
E ¼ E ðPP ; PR ; X G ; U Þ ð3:2Þ
However, given the characteristics of REnS, it makes sense to think of this as a

restricted demand problem where the consumer does not observe PR and chooses
R but, rather, is offered R and can then choose to either pay for these services (R1) or
not (R0). Therefore, PR is replaced with R, and the expenditure function can then be
rewritten as follows:
E ¼ E ∗ ðPP ; R; X G ; U Þ ð3:3Þ
In such a restricted case, WTP for “RE use” is simply the difference between two
expenditure functions (with R1 > R0); the compensating surplus (CS) welfare esti-
mate can be derived in terms of the difference:

CS ¼ E ∗ PP ; R0 ; X G ; U E ∗ PP ; R1 ; X G ; U ð3:4Þ
The above estimate of the CS is a measure of households’ WTP for “RE use”
services, i.e., the amount that each Italian household is willing to pay without
changing its utility level.
3.4.2 Survey Design
In order to derive estimates of households’ WTP, a nationwide survey with 1019

interviews was administrated at the end of November 2007, using a stratified sample
that was representative of Italian residents (see Appendix).
The survey was conducted by Istituto Piepoli, a marketing and consulting com-
pany, using a computer-aided web interviewing method. The full raw data set was
transferred to the authors for this research, so in principle, no hidden nonstochastic
distortions (such as recoding mistakes) would affect the results. A preliminary
analysis was conducted in May 2007 by a focus group to define the price vector,
taking into account potential bias associated with the PC method.
Considering the difficult to defined RE (Diaz-Rainey and Ashton 2008; Holt and
Holt 2004; Salmela and Varho 2006) we managed a focus group including energy
managers, experts, members of energy authorities, and academics. The fact that
these participants were confident with RE allowed us to obtain an unbiased
3.4 Methods and Data 67
economic valuation of RE without value judgment, in order to ascertain the percep-

tions of customers and to avoid consumer confusion about the RE market.
Information obtained from the focus group allow us to reduce price vector bias to
obtain a more reliable WTP. Furthermore, to take into account the consumer
confusion that affects the RE market (Diaz-Rainey and Ashton 2008; Salmela and
Varho 2006) we introduce validity test questions. Guidelines suggest various ways
to test for understanding and acceptance of the scenario proposed, such as follow-up
questions, test questions, etc. One common way is to include questions whose
answers should confirm each other (Soderqvist and Soutukorva 2006). In our
questionnaire, respondents are first asked if they know of RE and are then asked to
identify RE sources among a set of energy sources. In this way we investigate if
respondents perceive the importance of the description of the valuation scenario and
understanding of it, and how people are directly or indirectly affected by the
environmental change proposed.
Indeed, the literature has clarified that customers’ knowledge and awareness of
REnS are crucial factors in the success of CV studies (see, among others, Farhar
(1999)).
Finally, it is well known that in European RE markets a divergence between
stated and actual consumer behavior exists (Diaz-Rainey and Ashton 2008). To
reduce this divergence we directly introduce uncertainty in the elicitation format
according to the multiple bound uncertainty approach (Welsh and Poe 1998).
3.4.3 Questionnaire
Pros and cons of an RE development scenario are provided to respondents via the
questionnaire. The survey questionnaire provides a set of questions concerning
(1) general knowledge, awareness, and potential development of REnS in electricity
production; (2) knowledge about the Italian energy system; (3) WTP amounts (bids)
to support the Italian RE target; and (4) respondents’ attributes.
Respondents are first asked whether they are aware and if they believe that RE
could play an important role in Italy (“do not know/not sure” options are also
available), as follows:
Today, there is a heated debate on the opportunity to develop REnS. Are you for or
against REnS?
In your opinion, what is the Italian situation with regard to the need for energy
(electricity, heating, and transport) production activities?
In your opinion, can the development of REnS in Italy improve/worsen the current
energy situation in the country?
After the first section, respondents are questioned about their knowledge of RE,
using the validity test questions. Validity questions allow us to construct a dummy
variable that concerns the accuracy of the answers provided by respondents about the
scenario proposed—in particular, about the degree of knowledge of REnS. If the
interviewee answers “yes” to the first question and correctly identifies the different
types of REnS in the second question, the dummy variable is equal to one—
otherwise, zero.
Afterward, respondents are asked if they are in favor of and if they would or
would not contribute to REnS use in electricity production for environmental
reasons, according to the elicitation format. In this way we investigate respondents’
attitudes:
Italy has committed to increase the production of energy from renewable sources by
2020, bringing the ratio of REnS to 17%. The share is based on the gross final
energy consumption and if we only consider electricity generation the target
share is 26.4%. How much do you support this commitment?
Then we investigate the respondents’ WTP to support RE in Italy. To make
respondents confident with the elicitation proposed, the respondents are first asked to
state the amounts of their last three bills, and then they are informed of the A3
component, according to Sect. 3.3.1:
For the scenario described, what is the maximum amount that you are willing to pay
to support REnS as a surcharge on your bill? Please be careful about your
degree of certainty.
At the end of the questionnaire, interviewees are asked about their attributes
(demographic characteristics, age, education, etc.).
3.4.4 Elicitation Format and Econometric Model
In this CV study we assume that respondents may have in mind a whole valuation
distribution instead of a single-point economic value. Consequently, the uncertainty
typically affecting CV studies is taken into account by a stochastic elicitation format
and by an appropriate econometric model.
In detail, we use a method that refers to both a stochastic payment card (SPC) and
multiple bound dichotomous choice (MBDC).3 It can be seen as a variant of the PC
approach, which allows the possibility that consumers may have a range of eco-
nomic values in their minds. Furthermore, the PC method is consistent with impor-
tant guidelines (Atkinson et al. 2005) and many researchers suggest that this method
could be more intensively employed in CV studies (Mitchell and Carson 1989;
O’Garra and Mourato 2007).
3
This method is similar to the MBDC proposed by Welsh and Bishop (1993), which allows us to
consider that consumers have a range of economic values or a valuation distribution in mind,
instead of single-point economic value estimation. SPC was introduced by Wang and Whittington
(2005); more recently, Fonta et al. (2010) and Ichoku et al. (2009) have used the same approach.
Differences between these two approaches are also illustrated by Wang and He (2011).
By dealing with the uncertainty we use ex ante and ex post approaches to reduce
hypothetical bias.4 Specifically, among ex ante methods, we adopt a cheap talk script
(Bulte et al. 2005; Cummings and Taylor 1999; Nayga et al. 2007) so that partici-
pants are explicitly warned about hypothetical bias and are asked to respond to
valuation questions as if the payment were true.
However, cheap talk may have little or no effect on some people (Nayga et al.
2007; Samnaliev et al. 2003). Other characteristics determine the effective approach
of a cheap talk script, as underlined in Loomis (2011), Murphy et al. (2005b), and
MacKerron et al. (2009). The mixed success of cheap talk is also confirmed by Little
and Berrens (2004).
We use a script similar to that used by Cummings and Taylor (1999), appropri-
ately modified to be consistent with our scenario, elicitation format, and good
evaluated. We explain hypothetical bias and its effect on WTP, using information
reported by Murphy et al. (2005a, b).
Among ex post mitigation approaches (Broberg and Brännlund 2008; Evans et al.
2003; Vossler et al. 2003; Wang 1989; Welsh and Poe 1998) we adopt a certainty
correction method to reduce the overestimation risk by proposing five types of
acceptance intensity. These are “definitely yes” and “definitely no” (DY and DN),
“probably yes” and “probably no” (PY and PN), and “not sure or do not know”
(DK)—also using numerical likelihood information (100% and 0%, 75% and 25%,
and 50%, respectively).
Consequently, we adopt a variant of an SPC in which we propose a lesser
likelihood of acceptance but expressed also in a verbal form, or we can consider
our method as a MBDC variant in which we propose to respondents more bids
jointly with a acceptance likelihood also expressed numerically. Indeed, we propose
to the respondents 17 bids from EUR 0 to EUR 200 and five numerical likelihoods
for each bid. We use both numerical and verbal likelihood information because
“probably yes,” “unsure,” and “probably no” may be perceived differently among
individuals. Consequently, if only a verbal likelihood is directly asked for, an
ambiguous interpretation could result for some responses. Using both numerical
and verbal likelihood allows us to obtain more reliable results according to cognitive
reasons. Table 3.3 shows, in detail, the elicitation format structures used in the
survey.
Formally (Fig. 3.1), each respondent is asked to exchange income and an
environmental amenity level, which in our research is a lower CO2 concentration
in the atmosphere, due to RE production.
Starting from the status quo z0, each bid smaller than Ml M0 is paid by the
respondent to achieve z1. Conversely, each bid equal to or greater than Mu M0 is
rejected to achieve z1.
Between these two extremes cases we can locate the indifference curve of each
respondent we ask to state his/her WTP. Data collected by this format may be
analyzed in several ways (see, among others, Alberini et al. (2003), Broberg and
4
Ex ante and ex post approaches to reduce hypothetical bias are reviewed by Loomis (2011).
Table 3.3 Elicitation format

Definitely Probably Not sure/do Probably Definitely
Bid no (%) no (%) not know (%) no (%) yes (%)
EUR 0 0 25 50 75 100
EUR 0.05 0 25 50 75 100
EUR 0.10 0 25 50 75 100
EUR 0.15 0 25 50 75 100
EUR 0.30 0 25 50 75 100
EUR 0.50 0 25 50 75 100
EUR 0.75 0 25 50 75 100
EUR 1 0 25 50 75 100
EUR 1.50 0 25 50 75 100
EUR 2 0 25 50 75 100
EUR 5 0 25 50 75 100
EUR 10 0 25 50 75 100
EUR 15 0 25 50 75 100
EUR 20 0 25 50 75 100
EUR 30 0 25 50 75 100
EUR 50 0 25 50 75 100
EUR 100 0 25 50 75 100
EUR 200 0 25 50 75 100
+
Respondents are instructed to circle an answer for each of 17 prices
Money
Bid M0
0.05
M1
5
DY[100%]
10
PY[75%]
30
DK[50%]
Mu
PN[0%]
200 DN[0%]
z0 z1 Amenity
level
Fig. 3.1 Uncertainty preferences model

Money Money
Bid M0 Bid M0
0.05 0.05
M1 M1
5 5
DY [100%] 10 DY [100%]
10
PY [75%] PY [75%]
30 30
Mu Mu
200 200
z0 z1 z0 z1 Amenity
Amenity
level level
Panel (a) - Welsh and Poe approach Panel (b) - Broberg and Brännlund approach
Fig. 3.2 Uncertainty preferences model according to different approaches. a The Welsh and Poe
approach. b The Broberg and Brännlund approach
Brännlund (2008), Evans et al. (2003), Wang and He (2011), and Welsh and
Poe (1998).
In this chapter we apply both the Broberg and Brännlund (B&B) and Welsh and
Poe (W&P) approaches, recoding probabilistic answers according to four models:
1. “Higher bound” (HB) in which only DN ¼ no; others ¼ yes
2. “Unsure” (DK) in which DN and PN ¼ no; others ¼ yes
3. “Probably yes” (PY) in which DY and PY ¼ yes; others ¼ no
4. “Lower bound” (LB) in which only DY ¼ yes; others ¼ no
To explain our method we illustrate both approaches, comparing the B&B
approach with the well-known W&P approach.
Figure 3.2 illustrates these different ways of recoding data. Figure 3.2a shows the
PY model according to the W&P approach. In this case, allowing for uncertainty
(that is, a 25% reduction of certainty in the example proposed), the interval shifts
downward entirely, given that both the lower and upper bounds move together. The
B&B approach is illustrated in Fig. 3.2b. In this case the same change in probability
due to certainty reduction determines the expansion of the interval considered;
indeed, only the upper bound shifts downward, while the lower bound interval
remains at the initial level.
In other words, in the W&P approach the entire WTP interval shifts as the
probability statement changes, while using the B&B approach the interval does
not move entirely but only the upper bound shifts, introducing uncertainty; in
other words, the WTP interval is expanded instead of moved. Anyway, in both
approaches each respondent’s WTP lies in an interval that includes the highest WTP
in W&P while in B&B each respondent’s WTP is bound by the highest bid he or she
accepts and the lower bid he or she does not accept. Consequently, the B&B
approach uses the most reliable information about each respondent without
discarding “certain” responses (Broberg and Brännlund 2008). In this way it is
possible to reduce the overestimation risk, especially in the higher bound model.
The difference between these two approaches can also be explained by considering
Fig. 3.3, which shows an example of a response pattern. In detail, in the “probably
Bid DY PY DK PN DN
0.00
0.05 XXX
0.10 XXX
0.15 XXX
0.30 XXX
0.50 XXX
0.75 XXX
1.00 XXX
1.50 XXX
2.00 XXX
5.00 XXX
10.00 XXX
15.00 XXX
20.00 XXX
30.00 XXX
50.00 XXX
100.00 XXX
200.00 XXX
Fig. 3.3 Diagonal response patterns
yes” model the W&P interval is EUR 0.5–1.5 whereas in the B&B approach the
WTP interval is EUR 0.05–1.5.
Considering the higher bound model, the WTP intervals are EUR 30–50 and
EUR 0.30–50, respectively. Higher and lower bound models set a maximum confi-
dence interval for WTP estimates, given that they are “certainty models” (Broberg
and Brännlund 2008). These models are based on DY and DN responses providing
the extreme values of estimated WTP. According to the aim of our chapter, obtaining
a confidence interval for WTP makes the results more suitable for policy analysis.
From the econometrics point of view, in both approaches used it is possible to treat
the data collected as interval data; further, the limited number of zeros and point-
estimated WTP values, together with the small size of the intervals, confirm that the
use of an interval regression method is appropriate (Mitchell and Carson 1989;
Cameron and Huppert 1989; Whitehead et al. 1995). (See Appendix for further
details on econometric models.)
3.5 Empirical Findings
A favorable attitude with regard to REnS in electricity generation arises from

answers provide by the respondents. The majority of our sample believes that the
Italian energy scenario will worsen in the next 10 years and considers RE to
represent a strategic opportunity for Italy. More than 80% of interviews profess to
3.5 Empirical Findings 73
have “good” knowledge of REnS, while 10–12% report that they are not aware of
them. We notice that respondents affirm having accurate knowledge of REnS and are
able to correctly identify different types of REnS in more than 80% of cases. This
means that the respondents have good knowledge of the evaluated good and
understand the scenario evaluated. This information is an important check for the
survey. Indeed, if a respondent does not take the decision process seriously or does
not understand the questions posed, their responses will not reveal their true prefer-
ences (Solino et al. 2009).
Table 3.4 shows the location and scale parameters of several important variables.
The profile of the typical interviewee is a highly educated, married, 47-year-old man
who lives in a family that has one child, owns its own home, and has an income of
approximately EUR 35,000. The details of the WTP responses are presented in
Table 3.5. The first column refers to the amount (from lowest to highest) that
consumers are asked to pay for RE use, while the second column, showing the
frequency, provides information on consumer WTP to achieve the national target.
The third column, showing the cumulative frequency, reports the number of con-
sumers paying at least the indicated amount, and the column showing the survival
probability gives the related percentage. Figure 3.4 shows how uncertainty may
affect WTP distribution. As expected, the percentage of respondents willing to pay a
given amount decreases with the amount submitted but increases with weaker
certainty levels where accepted as a “yes.”
This is especially evident at the rightmost end of the tail for amounts greater than
EUR 10. Finally, in the present study, the sample does not report a large percentage
of zero WTP bids.
3.5.1 Estimation Results
Our model could be interpreted as a generalization of the MBDC and SPC methods;
consequently, our data might be estimated according to an interval regression model.
We estimate parameters under an assumption of log distribution and we compute the
mean and median WTP according to Eqs. (A.17) and (A.18).
It is still controversial whether the mean or median should be used to compute the
aggregate WTP (Pearce et al. 2008), and many researchers suggest using both
measures (Pearce et al. 2008; Hanemann 1989; Harrison and Krsitröm 1995).
Nevertheless, the median is sometimes preferred; indeed, it is more robust with
high bids and strategic responses, so it is a more conservative measure of WTP
(O’Garra and Mourato 2007; Arrow et al. 1993). Further, the median has been
preferred when it was “the tax-price that was offered to subject in the elicitation
format” (Harrison and Krsitröm 1995).
In our study, we use both indexes to estimate the aggregate WTP. The estimated
results confirm prior expectations, as shown in Tables 3.6 and 3.7. Variables related
to respondents’ characteristics explain differences in WTP for RE. Age and sex are
negatively related to WTP, while income, education, and professional status
Table 3.4 Variables and descriptive statistics

Standard
Variable Mean deviation Minimum Maximum
Income Continuous variable: household 38.42 12.01 0.15 77.46
(000) yearly income
Geo5 Categorical variable: 2.75 1.39 – –
1 ¼ northwest;
2 ¼ northeast;
3 ¼ central Italy;
4 ¼ south;
5 ¼ islands
D.Geo5 Dummy variable: 1 if Geo5 < 4 0.69 0.44 – –
D.City Dummy variable: 1 if municipality 0.25 0.09 – –
>100,000 inhabitants
D.Sex Dummy variable: 1 if male 0.42 0.5 – –
Age Continuous variable: age of a 46.3 17.91 20 78
respondent
Agec Categorical variable: from 1 to 5 2 1.26
Professional Categorical variable: from 1 to 10 6.03 3.24 – –
category
D.Profes- Dummy variable: 1 if professional 0.78 0.45 – –
sional status category ¼ enterprise or profes-
sional class
Education Continuous variable: number of 12.02 3.98 5 21
years for which a respondent has
attended school
D.Higher Dummy variable: 1 if education 0.56 0.5 – –
education (>13 years)
D.Scenario Dummy variable: 1 if response 0.21 0.19 – –
<3a
D.Know Dummy variable: 0 ¼ wrong; 0.88 0.18 – –
REnS 1 ¼ rightb
Household Discrete variable: number of 3.13 1.22 1 6
size household members
Child Discrete variable: number of 1.13 1.04 0 4
children
D.House Dummy variable: 1 if the family 0.67 0.34 – –
owns the house
D.Test Dummy variable: validity test 0.33 0.47 – –
questionc
REnS renewable electricity sources
a
The question is: “In your view, will the current Italian energy scenario worsen in the next
10 years?” (1 ¼ extremely likely; 2 ¼ very likely; 3 ¼ rather likely; 4 ¼ a little likely; 5 ¼ not at
all likely)
b
Knowledge of renewable energy sources: the interviewee is asked to identify renewable energy
sources
c
See Sect. 3.4.3
Table 3.5 Details of payment responses
Modela
Lower bound Probably yes Do not know Higher bound
Bidb Freq Cumul Surv Freq Cumul Surv Freq Cumul Surv Freq Cumul Surv
EUR 0 43 1019 1 37 1019 1 33 1019 1 28 1019 1
976 0.958 982 0.964 986 0.968 991 0.973
3.5 Empirical Findings
EUR 0.05 0 976 0.958 0 982 0.964 0 986 0.968 0 991 0.973
EUR 0.10 14 976 0.958 6 982 0.964 6 986 0.968 4 991 0.973
EUR 0.15 15 962 0.944 10 976 0.958 10 980 0.962 6 987 0.969
EUR 0.30 18 947 0.929 11 966 0.948 12 970 0.952 8 981 0.963
EUR 0.50 21 929 0.912 15 955 0.937 16 958 0.940 10 973 0.955
EUR 0.75 40 908 0.891 32 940 0.922 33 942 0.924 25 963 0.945
EUR 1 53 868 0.852 38 908 0.891 40 909 0.892 35 938 0.921
EUR 1.50 84 815 0.800 67 870 0.854 52 869 0.853 40 903 0.886
EUR 2 102 731 0.717 82 803 0.788 67 817 0.802 50 863 0.847
EUR 5 198 629 0.617 210 721 0.708 96 750 0.736 75 813 0.798
EUR 10 165 431 0.423 210 511 0.501 265 654 0.642 210 738 0.724
EUR 15 137 266 0.261 161 301 0.295 196 389 0.382 210 528 0.518
EUR 20 100 129 0.127 111 140 0.137 150 193 0.189 211 318 0.312
EUR 30 22 29 0.028 22 29 0.028 34 43 0.042 79 107 0.105
EUR 50 5 7 0.007 5 7 0.007 7 9 0.009 20 28 0.027
EUR 100 2 2 0.002 2 2 0.002 2 2 0.002 6 8 0.008
EUR 200+ 0 0 0.000 0 0 0.000 1 0 0.000 2 2 0.002
Cumul cumulative frequency, Freq frequency, Surv survival probability
a
Lower bound: definitely yes ¼ yes; others ¼ no. Probably yes: definitely yes, probably yes ¼ yes; others ¼ no. Do not know: definitely yes, probably yes ¼ yes;
others ¼ no. Higher bound: definitely no ¼ no; others ¼ yes
b
The largest amount the respondents would be willing to pay
75
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
LB PY DK HB
Fig. 3.4 Survivor function
positively affect estimates. These results show that younger citizens are more likely
to support the target, while women are less supportive than men. As expected,
citizens earning a higher income, such as those who are more highly educated,
support the RE target more. Differences between the two approaches used are
sometimes really small; indeed, the parameter estimates are significant, but they
are also really similar for age and sex. The differences for two crucial variables—
income and education—are slightly more substantial. Income is always significant
and the parameter estimates in the W&P approach are on average 10% higher than
those in the B&B approach. Education is significant in a few models, but the
parameter magnitude is again higher in the W&P approach.
Our results show that household size is negatively related to WTP, in accordance
with other studies conducted in Europe (Bollino 2009, Zoric and Hrovatin 2012),
and coherently with the expectations, the impact is greater using the W&P approach.
The empirical findings regarding the influence of geographical characteristics are
mixed. Indeed, if residents in northern and central Italy exhibit a higher WTP, the
municipality size is not always significant, especially in the B&B approach.
Considering the first two sections of the questionnaire, the results show that WTP
is significantly and positively influenced by REnS knowledge and by the conviction
that RE could play an important role in Italy even if this happens only for extreme
models.
Finally, the dummy variable, which is a proxy for scenario understanding,
reduces the actual WTP estimates, confirming that overestimation occurs if respon-
dents do not fully understand the scenario proposed. This result confirms the
importance of a carefully tested questionnaire because econometric analysis is not
likely to provide a remedy for an unclear valuation scenario. The importance of this
result is also stressed by the fact that it arises in both approaches, confirming that
methodological aspects are crucial in CV studies.
Table 3.6 Interval data regression for willingness to pay (WTP) to support renewable electricity
(RE) in Italy per model: the Broberg and Brännlund approach
Model
Variable Lower bound Probably yes Do not know Higher bound
Income (factual) 0.0479a 0.0600a 0.0602a 0.0663a
0.0047 0.0045 0.0051 0.0062
D.Geo5 (factual) 0.1271a 0.1060b 0.1701a 0.1866a
0.0330 0.0473 0.0562 0.0598
D.City (factual) 0.0806 0.1152c 0.1074 0.1043c
0.0462 0.0647 0.0912 0.0610
D.Sex (factual) 0.2905b 0.3589b 0.3418c 0.3025c
0.1400 0.1625 0.1701 0.1509
Age (factual) 0.3409b 0.4674b 0.5331b 0.6202b
0.0815 0.1110 0.1310 0.1513
D.Professional status (factual) 0.1416a 0.1601a 0.1888a 0.2074a
0.0367 0.0345 0.0450 0.0522
D.Higher education (factual) 0.1377c 0.1045c 0.2048 0.2409c
0.0627 0.0675 0.1389 0.1203
D.Scenario (methodological) 0.5025b 0.7832b 0.8045b 0.8210b
0.2513 0.3653 0.3990 0.4102
D.Know REnS (factual) 0.6463c 0.7070 0.7816 0.7934d
0.4110 0.5621 0.5419 0.5210
Household size (factual) 0.1094a 0.2318a 0.2712a 0.2710a
0.0467 0.0587 0.0643 0.0633
D.Test (methodological) 0.2087b 0.2961b 0.3095b 0.4007b
0.1480 0.1456 0.1989 0.1979
Constant 0.9581a 1.0312a 1.0406a 1.0468a
0.2019 0.2216 0.2899 0.2910
Lnsigma 0.5305a 0.5471a 0.5553a 0.5601a
0.0262 0.0279 0.0225 0.0230
Sigma 1.8091a 1.6089a 1.6043a 1.5708a
0.0422 0.0485 0.0508 0.5435
Obs 1019 1019 1019 1019
McKelvey and Zavoina’s R2 0.1263 0.1302 0.1467 0.1645
LR chi-squared (11) 139.98 152.37 183.90 191.67
Median WTP 4.67 7.32 7.59 9.02
95% CI 4.32–5.12 6.73 – 7.92 6.92–8.18 8.92–9.92
Mean WTP 12.44 13.16 13.57 15.08
95% CI 10.64–14.25 12.02–14.29 12.32–14.84 13.57–16.67
CI confidence interval, LR likelihood ratio, Obs observations, REnS renewable electricity sources
a
b
c
Table 3.7 Interval data regression for willingness to pay (WTP) to support renewable electricity
(RE) in Italy per model: the Welsh and Poe approach
Model
Variable Lower bound Probably yes Do not know Higher bound
Income (factual) 0.0479a 0.0632a 0.0667a 0.0730a
0.0047 0.0050 0.0048 0.0070
D.Geo5 (factual) 0.1271a 0.0911b 0.2701b 0.3932b
0.0330 0.0433 0.0989 0.0940
D.City (factual) 0.0806 0.2110a 0.1999c 0.2104b
0.0462 0.0588 0.0679 0.0875
D.Sex (factual) 0.2905b 0.2901c 0.3111c 0.3025
0.1400 0.1502 0.1123 0.1509
Age (factual) 0.3409b 0.3867c 0.5023b 0.5492a
0.0815 0.1881 0.1663 0.1051
D.Professional status (factual) 0.1416a 0.2001a 0.2319b 0.1807a
0.0367 0.0301 0.0699 0.0492
D.Higher education (factual) 0.1377c 0.0911b 0.2105b 0.2877b
0.0627 0.0455 0.0989 0.1003
D.Scenario (methodological) 0.5025b 0.8993a 0.6905b 0.7072b
0.2513 0.2501 0.4310 0.3910
D.Know REnS (factual) 0.6463c 0.4357 0.8222 0.9793c
0.4110 0.7121 0.5882 0.5882
Household size (factual) 0.1094a 0.1112b 0.2007a 0.2101a
0.0467 0.0657 0.0444 0.0558
D.Test (methodological) 0.2087b 0.2520b 0.1895b 0.3119b
0.1480 0.1011 0.1663 0.1120
Constant 0.9581a 1.1094a 1.1406c 1.4683a
0.2019 0.2145 0.2899 0.2910
Lnsigma 0.5305a 0.5000a 0.5001a 0.5216a
0.0262 0.0112 0.0441 0.0352
Sigma 1.8091a 1.6487a 1.6489a 1.6847a
0.0422 0.0273 0.0322 0.4444
Obs 1019 1019 1019 1019
McKelvey and Zavoina’s R2 0.1263 0.1413 0.1554 0.1638
LR chi-squared (11) 139.98 159.88 191.32 206.14
Median WTP 4.67 8.11 11.25 16.06
95% CI 4.32–5.12 7.47–8.75 10.24–12.21 14.35–17.75
Mean WTP 12.44 17.21 26.85 51.42
95% CI 10.64–14.25 15.73–18.75 23.99–29.67 44.33–58.65
CI confidence interval, LR likelihood ratio, Obs observations, REnS renewable electricity sources
a
b
c
Focusing on WTP, we notice that the expansion model provides more conserva-
tive estimates, a narrower confidence interval, and more stable parameter estimates
when we account for uncertainty in comparison with the W&P approach. That being
stated, the WTP estimates according to the W&P approach are 1.3–3.7 higher when
compared with the WTP estimates obtained by the B&B approach, depending on the
models and average type.
The median WTP lies between EUR 4.67 and EUR 9.02 bimonthly per house-
hold, using B&B, while the HB model provides a median WTP of EUR 16.10 if the
W&P approach is used.
A wider value range is obtained using the mean WTP, which is more affected by
high bids; indeed, the bimonthly amount goes from EUR 12.40 to EUR 15.10 in the
expansion model, but these figures increase noticeably using W&P, given that the
mean WTP ranges from EUR 12.40 to EUR 51.40 bimonthly per household.
Considering that in 2007 the electricity bill of a typical Italian household was
approximately EUR 70.50, this means that households’ support for RE lies between
6.6% and 70.3% of electricity bills, depending on the approach, models, and average
index considered. If we consider the B&B approach, similar percentages are
achieved by Brochers et al. (2007) and Zoric and Hrovatin (2012), while Hanemann
et al. (2011) find a WTP amount that is twice the bill, which is more comparable with
the W&P approach.
Finally, focusing on the annual WTP range obtained by the B&B approach, we
can see that they are comparable to the values discussed in Sect. 3.6. Indeed, the
annual median WTP obtained by the LB model is EUR 28, similar to Bollino’s
(2009) results, and the mean annual WTP (equal to EUR 90.48) obtained with the
higher bound model is close to the Grösche and Schröder (2011) figure obtained in
no-nuclear scenario. The W&P approach provides higher figures; the minimum
annual average WTP is equal to EUR 74.61, while the maximum value amounts to
EUR 308.51. A similar value is provided by Hanemann et al. (2011), who appraise
an annual mean equal to EUR 396.75. Considering that WTP estimates obtained
using data collected before the economic downturn maintain their validity in a long-
run perspective (see Sect. 3.6), the next step is to aggregate the household mean and
median WTP to compute the national WTP, comparing it with an estimate of the
total annual subsidy (IEFE 2009) needed in Italy to comply with EU climate change
package goals by 2020. Table 3.8 illustrates the market sustainability of the 20–20–
20 target according to the approaches used. The results confirm that high variability
exists also due to the models and averages used. The mean and median differ due to
the distribution’s skewness; consequently, the market sustainability ranges from
8.7% using the median WTP in the LB model to 56.4% using the median WTP in
the HB model. This wide range of results not only depends on the choice of statistical
index but also is affected by uncertainty.
These figures are obtained considering that households pay the entire target cost,
but if we assume that households pay only for their target cost share (around one fifth
of the full cost) the results change substantially. Using the B&B approach the median
WTP allows us to cover from 43.6% to 84.3%, while according to the W&P
approach the median WTP estimates are more than sufficient to achieve the national
Table 3.8 Market sustainability

Lower Probably Do not Higher
Model Variable Measure bound yes know bound
B&B approach WTP (per 2 months per household)
(expansion Median EUR 4.67 7.32 7.59 9.02
model) Mean EUR 12.44 13.16 13.57 15.08
Annual electricity n 6
bills
Households n 21,810,676
Total annual WTP
Median Million 305.30 479.24 496.43 590.17
EUR
Mean Million 1627.32 1722.05 1775.41 1973.52
EUR
Annual subsidy cost Million 3500
EUR
Market sustainability of RE
Median % 8.72 13.69 14.18 16.86
Mean % 46.49 49.20 50.73 56.39
(households’ share) EUR
Median % 43.61 68.46 70.92 84.31
Mean % 232.47 246.01 253.63 281.93
W&P approach WTP (per 2 months per household)
Median EUR 4.67 8.11 11.25 16.06
Mean EUR 12.44 17.21 26.85 51.42
Annual electricity n 6
bills
Households n 21,810,676
Total annual WTP
Median Million 305.30 530.88 736.04 1050.82
EUR
Mean Million 1627.32 2251.57 3513.68 6728.80
EUR
EUR
Median % 8.72 15.17 21.03 30.02
Mean % 46.49 64.33 100.39 192.25
(households’ share) EUR
Median % 43.61 75.84 105.15 150.12
Mean % 232.47 321.65 501.95 961.26
B&B Broberg and Brännlund, RE renewable electricity, W&P Welsh and Poe, WTP willingness to
pay
3.6 Comparative Analysis 81
target; indeed, market sustainability lies between 43.6% and 150.1%. In other words,
both the DK and HB models allow us to fully cover the target cost attributable to
Italian households. Using the mean instead of the median market sustainability of
20–20–20, the Italian target increases impressively, given that the more conservative
model (LB) allows us to cover twice the target cost.
Finally, we want to highlight that initially we have assumed that the full incre-
mental cost would be arbitrarily ascribed entirely to households. Computations made
under this assumption provide an underestimation of market sustainability of the
renewable national target. In fact, if we consider all client typologies and attribute to
households only their share of the cost, market sustainability could noticeably
increase.
3.6 Comparative Analysis
As shown in Sect. 3.2, several studies have focused on RE valuation from different
point of view (Menegaki 2008), such as economic and welfare, financial, ecological,
and engineering aspects. Our contribution belongs to the first type of studies
considered because we estimate an aggregate WTP to appraise if it is possible to
achieve the Italian Directive 2009/72/CE target by 2020 by market force. According
to this aim, we have investigated the related literature on RE (Tables 3.9 and 3.10) in
the main selected papers that estimate WTP to develop RE according to policy
programs in order to appraise their market financing capacity.
However, we have also reviewed papers that provide aggregate or individual
WTP for completeness.
Obviously, given the number of paper on the topic of RE, our choice leaves out
several studies that investigate this issue from different points of view such as
environmental impact, attributes (Roe et al. 2001; Kotchen and Moore 2007;
Longo et al. 2008), and determinants of the acceptance of REnS facilities installation
(Álvarez-Farizo and Hanley 2002; Oliver et al. 2011; Litvine and Wüstenhagen
2011; Mozumder et al. 2011). In Table 3.9 we consider studies that explicitly take
into account a quantitative policy target. Using several elicitation formats,
researchers find generally low financial economic support for RE, with the exception
of Italy, according to findings by Bollino (2009) and Bigerna and Polinori (2012,
2013, 2014). Comparing annual households’ WTP at a constant price, we note that a
very wide range of values exist in Europe and in the USA.
Restricting the analysis to Europe the annual amount ranges from EUR 10.20 in
Sweden to EUR 324.80 in Japan and EUR 396.80 in Spain. In the USA, depending
on the energy sources, policy program, and type of payment, the annual WTP per
household ranges from EUR 10.28 to EUR 309.10. The interval is narrower if we
consider countries with more comparable electricity markets (Sileo 2011), such as
Germany, Italy, and the UK. In these cases the annual amounts lie between
EUR 17.20 and EUR 92.70, depending on uncertainty and the energy mix.
Table 3.9 Comparable willingness to pay (WTP) for renewable electricity (RE): studies that consider WTP and the target cost
82
Sample Survey Mean annual WTP household

Study Country Data years size type Method (2014 EUR)
Batley et al. UK 1999 (Mar) 746 Mail CV-OE; multiple regression and factor analysis 87.92a
(2000)
Batley et al. UK 1997 (Mar); 742; Mail CV-OE; correlation analysis 81.25; 84.88a
(2001) 1999 (Apr) 692
Bigerna and Italy 2007–2008 1600 Internet Bidding game; interval regression From 46.87 to 57.76
Polinori (2013) (Nov)
Bigerna and Italy 2007 (Nov) 1019 Internet CV-MBDC From 76.56 to 90.54; median
Polinori (2014) from 27.72 to 48.30
Bollino (2009) Italy 2006 (Nov) 1601 Internet SPC treated as a dichotomous choice model; From 17.19 to 66.08
probit
Dagher and Lebanon 2013 600 Face to CV-OE From 271.18 to 650.88; median
Harajili (2015) face from 218.78 to 546.96
Grösche and Germany 2008 2948 Internet CEx; random parameter model 79.42; 92.65b
Schröder (2011)
Harajili and Gor- Lebanon 2013 200 Face to CV-OEc From 2297.23 to 6704.64
don (2015) face
Kim et al. (2012) South 2010 (Aug) 720 Face to CV-DBDC; spike model 13.89
Korea face
Ivanova (2005) Australia 2004 (Aug) 213 Mail CV-OE; regression analysis 57.98; 79.58
Ivanova (2012) Australia 2004 (Aug) 820 Mail CV-OE; latent class analysis From 42.37 to 117.35
Yoo and Kwak South 2006 800 Face to CV-DBDC: (a) parametric (spike model), (a) from 19.39 to 21.87, (b) from
(2009) Korea face (b) nonparametric (interpolation model) 18.15 to 29.29
CEx choice experiment, CV contingent valuation, DBDC double bound dichotomous choice, MBDC multiple bound dichotomous choice, OE open ended,
SPC stochastic payment card
a
According to the Department of Energy and Climate Change quarterly energy prices (Dec 2012), we assume annual bills of GBP 283 in 1997 and GBP 266 in
1999
b
According to the World Energy Council energy efficiency indicators, we assume consumption of 3500 KWh/year in 2008
c
This study refers to the commercial sector
Table 3.10 Willingness to pay (WTP) for renewable electricity (RE): studies that provide individual and/or aggregate WTP
Sample Mean annual WTP household (2014
Study Country Data years size Survey type Method EUR)
Ackura (2015) UK 2008–2009 2000 Internet CV-DBDC; zero-inflated Not applicable or not computable
Borchers et al. USA 2006 (May) 625 Face to face CEx; nested logit model From 11.78 to 235.04; from 92.14 to
(2007) 185.50a
Byrnes et al. USA 1992 (Sep)– 600–500 Telephone and CEx; censored regression 20.22; 22.28
(1999) 1994 (Nov) mail
3.6 Comparative Analysis
Goett et al. USA 2000 1205 Telephone and CEx; mixed-logit 224.24; 309.12b
(2000) mail regression
Guo et al. (2015) China 2010 (May) 700 Face to face CV-SBDC From 26.41 to 32.28; from 28.37 to
40.11; from 23.48 to 31.30
Hanemann et al. Spain 2009 (Nov–Dec) 233 Telephone CV-SBDC 396.75
(2011)
Hansla et al. Sweden 855 Mail PC; regression analysis From 10.20 to 20.40c
(2008)
Nomura and Japan 2000 (Oct) 379 Mail CV-DBDC; Weibull distri- 324.82
Akay (2004) bution function
Wiser (2007) USA 2001 (Feb–May) 1574–202 Face to face CV-SBDC Not applicable or not computable
Zhang and Wu China 2010 (May–Jun) 652–536 E-mail and mail CV; PC; mixed logit From 12.38 to 16.23
(2012)
Zografakis et al. Greece 2006 (Sep)– 1440 Face to face Face to face interview 70.76
(2010) (Crete) 2007 (Feb)
Zoric and Slovenia 2008 450 Internet and field CV-OE; Tobit, probit; 56.25; 53.50
Hrovatin (2012) interviews truncated regression
CEx choice experiment, CV contingent valuation, DBDC double bound dichotomous choice, MBDC multiple bound dichotomous choice, OE open ended,
PC payment card, SBDC single bound dichotomous choice
a
Voluntary and mandatory payments are considered with 10% or 25% of the REnS target
b
According to the World Energy Council energy efficiency indicators, we assume consumption of 11,055 KWh/year in 2000
c
According to the World Energy Council energy efficiency indicators, we assume consumption of 8900 KWh/year in 2008
83
Regardless of the energy mix used in RE production, uncertainty plays a crucial

role in WTP, as shown in a few studies that consider this issue. Vossler et al. (2003)
find that in a pricing program, certainty correction accounts for between 9.5% and
75% of uncorrected WTP in the DC model, while in the MBDC approach a less
conservative model provides a WTP value four times greater than the conservative
one. In Bollino (2009) uncertainty accounts for between 8.5% and 45.5% of the
Italian target cost. Similar results are obtained by Bigerna and Polinori (2012, 2014)
controlling also for different elicitation formats.
Another crucial aspect concerns the impact of the economic crisis on consumers’
WTP. This is especially true in this chapter because our aim is to assess if the EU
climate change policy (20–20–20 target) can be achieved by market forces in Italy.
This is a long-run policy target that needs to be confronted with stable and actual
households’ WTP. This is so because policy analysis should be based on reliable and
constant structural preference parameters, and should not be affected by exceptional
events.
In this respect, our survey fits this objective quite well because it was adminis-
trated at the end of November 2007. The financial crises that began at the end of
2007 led to a deep and prolonged global economic downturn, which afterward
significantly altered consumers’ long-run perceptions. This inevitably changed cit-
izens’ spending decisions. This means that our data refer to a period of macroeco-
nomic stability; thus, they have the necessary characteristics for a long-term policy
evaluation.
According to the literature it is doubtful whether an economic downturn reduces
WTP for environmental goods. Loureiro et al. (2009) highlight that in the period
2006–2009, WTP estimates for environmental damage (the Prestige oil spill in
Spain) were down by half even though this reduction did not affect regions more
directly involved in the environmental disaster. They explain this fact by the
economic crisis: in 2009 the Spanish yearly GDP dropped by 4% and negative
figures were expected for 2010. Conversely, in 2009, Hanemann et al. (2011)
estimated a very high WTP to support RE in Spain. They used the forthcoming
Copenhagen summit to justify this result.
Controversial results have also been obtained by Metcalfe and Baker (2012).
They compare two identical surveys conducted in 2008–2010 for an environmental
improvement in water sector. The results highlight that the economic downturn
consistently drops payment card WTP but does not reduce WTP elicited by
CV. The few studies on the impact of the economic downturn on WTP provide
conflicting results. Nevertheless, the results reported by Metcalfe and Baker (2012)
show some sensitivity of the PC method to economic crises. Considering that we
utilize a variant of the PC method, we think it is crucial to use precrisis data in a long-
run framework of analysis.
On the basis of these considerations we deem it appropriate to use our data to
elicit a stable WTP structure.
References 85
3.7 Conclusions
In Europe the current energy policy aims to increase the share of renewables in the
electricity generation mix. In Italy the current and binding constraint is to attain
26.4% of electricity production from renewable energy sources. This chapter has
investigated Italian households’ willingness to pay (WTP) to implement the
European Union climate change policy in a comparative perspective, also focusing
on the impact of methodological heterogeneity.
Our survey highlights an appreciable knowledge of REnS and broad consensus
regarding the development of renewable electricity (RE). The annual median WTP is
estimated at between EUR 305.3 million and EUR 1.0508 billion, which is a share
between 8.7% and 30% of the national target cost. Comparing annual households’
WTP, at constant prices, we notice that a very wide range of values exist in Europe.
Our results are close to those of similar countries. In particular, our WTP is
reasonably comparable to the German amounts. This is a comfortable result, given
the similarity of the Italian and German electricity markets.
In addition, we estimate that monetary value uncertainty amounts to 9% of the
target cost if the Broberg and Brännlund approach is used. This is a low figure
compared with previous results, and it depends on the more conservative method
(Broberg and Brännlund 2008) and the more conservative average index used in this
chapter. Indeed, using the Welsh and Poe approach, uncertainty increases to 21% of
the national target cost. This result also confirms that methodological heterogeneity
is important to explain the variability of the results that exist in the literature.
Finally, the analysis of the A3 component burden shows that in Italy the actual
additional cost to consumers due to support for RE is less than the WTP estimates
obtained in our models. This means that a further margin could exist for additional
policy action to implement appropriate education campaigns aimed at providing
information to reduce the uncertainty that affects the RE market.
References
Accent & Rand Europe (2010) Review of stated preference and willingness to pay methods—
prepared for Competition Commission. Available via Competition Commission. http://www.
competitioncommission.org.uk/our_role/analysis/summary _and_report_combined.pdf.
Accessed 12 Jan 2014
Ackura E (2015) Mandatory versus voluntary payment for green electricity. Ecol Econ 116:84–94
Alberini A, Boyle K, Welsh M (2003) Analysis of contingent valuation data with multiple bids and
response options allowing respondents to express uncertainty. J Environ Econ Manag 45:40–62
Arrow K, Solow R, Portney RP, Leamer EE, Radner R, Schuman H (1993) Report of the NOAA
panel on contingent valuation. http://www.cbe.csueastbay.edu/~alima/courses/4306/articles/
NOAA%20on%20contingent%20valuation%201993.pdf. Accessed 1 Nov 2011
Atkinson G, Healey A, Mourato S (2005) Valuing the cost of violent crime: a stated preferences
approach. Oxf Econ Pap 57:559–585
Bateman IJ, Langford IH, Jones AP, Kerr GN (2001) Bound and path effects in double and triple
bounded dichotomous choice contingent valuation. Resour Energy Econ 23:191–213
Batley SL, Fleming PD, Uwin P (2000) Willingness to pay for renewable energy: implications for
UK green tariff offerings. Indoor Buil Environ 9:157–170
Batley SL, Colbourne D, Fleming PD, Urwin P (2001) Citizen versus consumer: challenges in the
UK green power market. Energy Policy 29:479–487
Bigerna S, Polinori P (2012) Households’ willingness to pay for renewable energy sources in Italy:
a bidding game approach. In: Uvalic M (ed) Electricity markets and reforms in Europe. Franco
Angeli, Milan, pp 61–84
Bigerna S, Polinori P (2013) A bidding game for Italian households’ WTP for RES. Atl Econ J
4:189–190
Bigerna S, Polinori P (2014) Italian households' willingness to pay for green electricity. Renew Sust
Energ Rev 34:110–121
Bollino CA (2009) The willingness to pay for renewable energy sources: the case of Italy with socio
demographic determinants. Energy J 30:81–96
Borchers AM, Dukea JM, Parsons RM (2007) Does willingness to pay for green energy differ by
source? Energy Policy 35:3327–3334
Broberg T, Brännlund R (2008) An alternative interpretation of multiple bounded WTP data-
certainty dependent payment card intervals. Resour Energy Econ 30:555–567
Bulte E, Gerking S, List JA, de Zeeuw A (2005) The effect of varying the causes of environmental
problems on stated WTP values: evidence from a field study. J Environ Econ Manag
49:330–342
Byrnes B, Jones C, Goodmanf S (1999) Contingent valuation and real economic commitments:
evidence from electric utility green pricing programmes. J Environ Plan Manag 42:149–166
Cameron TA, Huppert DD (1989) OLS versus ML estimation of non-market resource values with
payment card interval data. J Environ Econ Manag 17:230–246
Champ PA, Bishop RC, Brown TC, McCollum DW (1997) Using donation mechanisms to value
non-use benefits from public goods. J Environ Econ Manag 33:151–162
Champ PA, Boyle KJ, Brown TC (2003) A primer on nonmarket valuation. Kluwer, Dordrecht
Cummings RG, Taylor LO (1999) Unbiased value estimates for environmental goods: a cheap talk
design for the contingent valuation method. Am Econ Rev 89:649–665
Dagher L, Harajli H (2015) Willingness to pay for green power in an unreliable electricity sector:
part 1. The case of the Lebanese residential sector. Renew Sust Energ Rev 50:1634–1642
De Shazo JR (2002) Designing transactions without framing effects in iterative question formats. J
Environ Econ Manag 43:360–385
de Vries BJM, van Vuuren DP, Hoogwijk MM (2007) Renewable energy sources: their global
potential for the first-half of the 21st century at a global level: an integrated approach. Energy
Policy 35:2590–2610
Diaz-Rainey I, Ashton JK (2008) Stuck between a ROC and a hard place? Barriers to the take up of
green energy in the UK. Energy Policy 36:3053–3061
Evans MF, Flores NE, Boyle KJ (2003) Multiple-bounded uncertainty choice data as probabilistic
intentions. Land Econ 79:549–560
Farhar BC (1999) Willingness to pay for electricity from renewable resources: a review of utility
market research. http://www.nrel.gov/docs/fy99osti/26148.pdf. Accessed 21 Jul 2013
Fonta M, Ichoku HE, Ogujiuba KK (2010) Estimating willingness to pay with the stochastic
payment card design: Further evidence from rural Cameroon. Environ Dev Sustain 12:179–193
Genius M, Strazzera E (2005) Modeling elicitation effects in contingent valuation studies. In:
Scarpa R, Alberini A (eds) Applications of simulation methods in environmental and resource
economics. Springer, Dordrecht, pp 223–246
Goett AA, Hudson K, Train KE (2000) Customers’ choice among retail energy suppliers: the
willingness to pay for service attributes. Energy J 4:1–28
References 87
Grösche P, Schröder C (2011) Eliciting public support for greening the electricity mix using random
parameter techniques. Energy Econ 33:363–370
Guo X, Liu H, Mao X, Jin J, Chen D, Cheng S (2015) Willingness to pay for renewable electricity: a
contingent valuation study in Beijing, China. Energy Policy 68:340–347
Haab TC, McConnell KE (1997) Referendum models and negative WTP: alternative solutions. J
Environ Econ Manag 32:251–270
Hanemann WM (1989) Welfare evaluations in contingent valuation experiments with discrete
responses data: replay. Am J Agric Econ 66:1057–1061
Hanemann WM, Labandeira X, Loureiro ML (2011) Climate change, energy and social preferences
on policies: exploratory evidence for Spain. Clim Res 48:343–348
Hansla A, Gamble A, Juliusson A, Garling T (2008) Psychological determinants of attitude towards
and willingness to pay for green electricity. Energy Policy 36:768–774
Harajli H, Gordon F (2015) Willingness to pay for green power in an unreliable electricity sector:
part 2. The case of the Lebanese commercial sector. Renew Sust Energ Rev 50:1643–1649
Harrison GW, Krsitröm B (1995) On the interpretation of responses of contingent valuation
surveys. In: Johansson PO, Krsitröm B, Mäler KG (eds) Current issue in environmental
economics. Manchester University Press, Manchester, pp 35–57
Holt EA, Holt MS (2004) Green pricing resource guide. http://www.env.state.ma.us/dpu /docs/
electric/08-54/62512ecatexr2.pdf. Accessed 10 Oct 2012
Ichoku HE, Fonta WM, Kedir A (2009) Measuring individuals’ valuation distributions using a
stochastic payment card approach: application to solid waste management in Nigeria. Environ
Dev Sustain 11:509–521
IEA (International Energy Agency) (2014) World energy outlook. OECD/IEA, Paris
IEFE (Centre for Research on Energy and Environmental Economics and Policy) (2009)
Prospettive di sviluppo delle energie rinnovabili per la produzione di energia elettrica.
Opportunità per il Sistema Industriale Nazionale—Research Report. http://portale.unibocconi.
it/wps/allegatiCTP/ Research%20Report%203_1.pdf. Accessed 8 Jun 2010
Ivanova G (2005) Queensland consumers’ willingness to pay for electricity from renewable energy
sources. In: Proceedings of the Ecological Economics in Action conference, Palmerston North,
11–12 Dec 2005
Ivanova G (2012) Are consumers’ willing to pay extra for the electricity from renewable energy
sources? An example of Queensland, Australia. Int J Renew Energy Res 2:758–755
Jacobsson S, Bergek A (2004) Transforming the energy sector: the evolution of technological
systems in renewable energy technology. In: Klaus J, Binder M, Wieczorek A (eds) Governance
for industrial transformation. Proceedings of the 2003 Berlin conference on the Human Dimen-
sions of Global Environmental Change. Environmental Policy Research Centre, Berlin, pp
208–236
Jäger-Waldau A, Szabó M, Scarlat N, Monforti-Ferrario F (2011) Renewable electricity in Europe.
Kim J, Park J, Kim H, Heo E (2012) Assessment of Korean customers’ willingness to pay with RPS.
Kotchen MJ, Moore MR (2007) Private provision of environmental public goods: household
participation in green-electricity programs. J Environ Econ Manag 53:1–16
Little J, Berrens R (2004) Explaining disparities between actual and hypothetical stated values:
further investigation using meta-analysis. Econ Bull 3:1–13
Litvine D, Wüstenhagen R (2011) Helping “light green” consumers walk the talk: results of a
behavioral intervention survey in the Swiss electricity market. Ecol Econ 70:462–474
Longo A, Markandya A, Petrucci M (2008) The internalization of externalities in the production of
electricity: willingness to pay for the attributes of a policy renewable energy. Ecol Econ
67:140–152
Longo A, Hoyos D, Markandya A (2012) Willingness to pay for ancillary benefits of climate
change mitigation. Environ Res Econ 51:119–140
Loomis J (2011) What’s to know about hypothetical bias in stated preference valuation studies? J
Econ Surv 25:363–370
Loureiro ML, Loomis JB, Vazquez MX (2009) Economic valuation of environmental damages due
to the prestige oil spill in Spain. Environ Res Econ 44:537–553
MacKerron GJ, Egerton C, Gaskell C, Parpia A, Mourato S (2009) Willingness to pay for carbon
offset certification and co-benefits among (high-)flying young adults in the UK. Energy Policy
37:1372–1381
Menegaki A (2008) Valuation for renewable energy: a comparative review. Renew Sust Energ Rev
12:2422–2437
Metcalfe PJ, Baker W (2012) The sensitivity of willingness to pay to an economic downturn. (Paper
presented at the Envecon 2012—Applied Environmental Economics—conference, London).
http://eprints.lse.ac.uk/43316/. Accessed 9 Mar 2012
Mitchell RC, Carson RT (1989) Using surveys to value public goods: the contingent valuation
method. Resources for the Future, Washington, DC
Mozumder P, Vásquez WF, Marathe A (2011) Consumers’ preference for renewable energy in the
southwest USA. Energy Econ 33:1119–1126
Murphy JJ, Allen PG, Stevens TH, Weatherhead D (2005a) A meta-analysis of hypothetical bias in
stated preference valuation. Environ Res Econ 30:313–325
Murphy JJ, Stevens TH, Weatherhead D (2005b) Is cheap talk effective at eliminating hypothetical
bias in a provision point mechanism? Environ Res Econ 30:327–343
Nayga RM, Wu X, Brummett RG (2007) On the use of cheap talk in new product valuation. Econ
Bull 2:1–9
Nemet GF (2009) Demand-pull, technology-push, and government-led incentives for
non-incremental technical change. Res Policy 38:700–709
Nomura N, Akay M (2004) WTP for green electricity in Japan as estimated through contingent
valuation method. Appl Energy 78:453–463
O’Garra T, Mourato S (2007) Public preferences for hydrogen buses: comparing interval data, OLS
and quantile regression approaches. Environ Res Econ 36:389–411
Oliver H, Volschenk J, Smit E (2011) Residential consumers in the Cape Peninsula’s willingness to
pay for premium priced green electricity. Energy Policy 39:544–550
Pearce D, Atkinson G, Mourato S (2008) Cost–benefit analysis and the environment—recent
developments. OECD Publishing, Paris
RAEG (2008) Annual report on the state of services and the regulatory activities. http://www.
autorita.energia.it/allegati/inglese/annual_report/relann2008english.pdf. Accessed 30 May
2012
REN21 (2015) Renewables 2015 global status report. REN21 Secretariat, Paris
Roe B, Teisl MF, Levy A, Russell M (2001) US consumers’ willingness to pay for green electricity.
Rowe RD, Schulze WD, Breffle WS (1996) A test for payment card biases. J Environ Econ Manag
31:178–185
Salmela S, Varho V (2006) Consumers in the green electricity market in Finland. Energy Policy
34:3669–3683
Samnaliev M, Stevens TH, More T (2003) A comparison of cheap talk and alternative uncertainty
calibration techniques in contingent valuation. http://nrs.fs.fed.us/pubs/jrnl/2003/ ne_2003_
samnaliev_001.pdf. Accessed 9 Sep 2011
Sileo V (2011) Dentro i prezzi dell’energia elettrica in Italia. Crescita libera senza limite? http://
www.adamsmith.it/download/download/UP120110927123222UPDentro%20i%20prezzi%
20Adam%20Smith%20paper.pdf. Accessed 15 Jan 2012
Soderqvist T, Soutukorva A (2006) An instrument for assessing the quality of environmental
valuation studies. Swedish Environmental Protection Agency, Stockholm
Solino M, Farizo BA, Campos P (2009) The influence of home site factors on residents’ willingness
to pay: an application for power generation from scrubland in Galicia, Spain. Energy Policy
37:4055–4065
References 89
Vossler CA, Ethier RG, Poe GL, Welsh MP (2003) Payment certainty in discrete choice contingent
valuation responses: result from a field validity test. South Econ J 69:886–902
Wang H (1989) Treatment of “don’t know” responses in contingent valuation surveys: a random
valuation model. J Environ Econ Manag 32:219–232
Wang H (1997a) Treatment of don’t-know responses in contingent valuation surveys: a random
valuation model. J Environ Econ Manag 32:219–232
Wang H (1997b) Contingent valuation of environmental resources: a stochastic perspective. PhD
dissertation, University of North Carolina at Chapel Hill
Wang H, He J (2011) Estimating individual valuation distributions with multiple bounded, discrete
choice data. Appl Econ 43:2641–2656
Wang H, Whittington D (2005) Measuring individuals’ valuation distribution using stochastic
payment card approach. Ecol Econ 55:143–154
Welsh MP, Bishop RC (1993) Multiple bounded discrete choice models. http://fes.forestry.
oregonstate.edu/sites/fes.forestry.oregonstate.edu/files/PDFs/W133%206th%20Interim%
20Report%201993.pdf. Accessed 22 Apr 2012
Welsh MP, Poe GL (1998) Elicitation effects in contingent valuation: comparisons to a multiple
bounded discrete choice approach. J Environ Econ Manag 36:170–185
Whitehead J, Cherry T (2007) Willingness to pay for a green energy program: a comparison of
ex-ante and ex-post hypothetical bias mitigation approaches. Resour Energy Econ 29:247–261
Whitehead JC, Hobant TJ, Cliffordt WB (1995) Measurement issues with iterated, continuous-
interval contingent valuation data. J Environ Manag 43:129–139
Wiser RH (2007) Using contingent valuation to explore willingness to pay for renewable energy: a
comparison of collective and voluntary payment vehicles. Ecol Econ 62:419–432
Yoo SH, Kwak SY (2009) Willingness to pay for green electricity in Korea: a contingent valuation
study. Energy Policy 37:5408–5416
Zhang L, Wu Y (2012) Market segmentation and willingness to pay for green electricity among
urban residents in China: the case of Jiangsu province. Energy Policy 51:514–523
Zografakis N, Sifaki E, Pagalou M, Nikitaki G, Psarakis V, Tsagarakis KP (2010) Assessment of
public acceptance and willingness to pay for renewable energy sources in Crete. Renew Sust
Energ Rev 14:1088–1095
Zoric J, Hrovatin N (2012) Household willingness to pay for green electricity in Slovenia. Energy
Policy 47:180–118
Appendix
Theoretical Model
Let us consider a household’s direct utility function:
U ¼ U ðX P ; R; X G Þ ðA:1Þ
This function is positively related to the private goods XP (XP1, ..., XPN), the
composite public good XG, and the public good R (renewable electricity
(RE) services). XG is a composite commodity of all public goods with unit prices
and value equal to the tax charged to the household. Households maximize U subject
to their budget constraints—that is:
M ¼ X P PP þ X G ðA:2Þ
where M is the nominal income and PP is a price vector of private goods. Each
household spends all of its disposable income by purchasing private goods:
Md ¼ M X G ðA:3Þ
A maximization framework provides a set of conditional demand functions:
d i ∗ ¼ dðPP ; PR ; X G ; Md Þ ðA:4Þ
By substituting di* into U, we obtain a conditional indirect utility function:
V ¼ V ðPP ; PR ; X G ; Md Þ ðA:5Þ

in Environmental Science, https://doi.org/10.1007/978-94-024-1574-2
92 Appendix
Inverting V for Md, we obtain the conditional expenditure function:
E∗ ¼ Md ¼ E ∗ ðPP ; PR ; X G ; U Þ ðA:6Þ
Minimizing expenditures on both private and public goods subject to the utility
level, we obtain the restricted expenditure function:
E ¼ EðPP ; PR ; X G ; U Þ ðA:7Þ
The conditional expenditure function and restricted expenditure function are

related as follows:
E ðPP ; PR ; X G ; U Þ ¼ E ∗ ðPP ; PR ; X G ; U Þ þ X G ðA:8Þ
We assume that the consumer does not observe PR and choose R; rather, the
consumer is offered R and can choose to pay for it or not. Therefore, PR is replaced
with R, and we can rewrite the relationship as follows:
EðPP ; R; X G ; U Þ ¼ E ∗ ðPP ; R; X G ; U Þ þ X G ðA:9Þ
By changing the scenario, we assume that the restricted expenditure function

varies according to R: R0 ¼ scenario without renewable energy sources (REnS) in
electricity production; R1 ¼ scenario with REnS in the energy portfolio. By holding
M constant, we find that the CS gives the willingness to pay (WTP) for the use of
REnS in electricity production:

CS ¼ E PP ; R0 ; X G 0 ; U 0 E PP ; R1 ; X G 0 ; U 0 ðA:10Þ

CS ¼ E ∗ PP ; R0 ; X G 0 ; U 0 þ X G 0 E∗ PP ; R1 ; X G 0 ; U 0 þ X G 0 ðA:11Þ

CS ¼ E ∗ PP ; R0 ; X G 0 ; U 0 E ∗ PP ; R1 ; X G 0 ; U 0 ðA:12Þ
where U0 is the utility level of the household without the RE program. This estimate
of compensating surplus is a measure of the WTP for the “RE use” service.
Econometric Model
Following Cameron and Huppert (1989), the WTP probability associated with the
choice of the respondent is:
Pðt i Þ ¼ Pðt li < WTPi t ui Þ ðA:13Þ

Appendix 93
Because WTP is non-negative and its distribution is skewed, we use a lognormal

conditional distribution:
log WTPi ¼ xi T β þ εi ðA:14Þ
where εi is distributed normally, with mean zero and standard deviation σ. The
probability of choosing ti can be written:

Pðt i Þ ¼ Φ log t ui xi T β =σ Φ log t li xi T β =σ ðA:15Þ
where Φ is the standard normal cumulative density function. The corresponding log
likelihood function can be written:
X
log L ¼ i
log Φ log t ui xi T β =σ Φ log t li xi T β =σ ðA:16Þ
We estimate the optimal values of β and σ and the mean and median WTP:

Median WTP ¼ exp xi T β ðA:17Þ

Mean WTP ¼ exp xi T β =exp σ 2 =2 ðA:18Þ
and we compute the confidence interval by bootstrap methods with 3000

replications.
Statistical Details
Table A.1 provides the sample characteristics and shows that the sample is highly
representative of the Italian population in terms of the male to female ratio, geo-
graphical and urban locations, demographic characteristics, education, and income
distribution.
Table A.1 Survey respondents’ and country residents’ characteristics

Variable Survey respondents Country residents
Gender
Male 47.78% 48.40%
Female 52.22% 51.60%
Macro region
Northwest 26.11% 26.21%
Northeast 19.69% 18.66%
Central 19.64% 19.14%
South (including Sicily and Sardinia) 34.55% 36.00%
(continued)
94 Appendix
Table A.1 (continued)

Municipality size
5000 people 17.47% 18.58%
5001–10,000 people 13.67% 14.11%
10,001–30,000 people 23.69% 22.81%
30,001–100,000 people 21.96% 21.29%
100,001–500,000 people 11.65% 10.98%
>500,000 people 11.55% 12.23%
Age
15–17 years 3.55% 3.54%
18–24 years 9.92% 9.53%
25–34 years 16.78% 17.98%
35–44 years 18.85% 17.77%
45–54 years 16.68% 15.52%
55–64 years 14.36% 13.89%
>64 years 19.84% 21.77%
Marital status
Single 27.99% 27.76%
Divorced 1.14% 1.23%
Separated 1.58% 1.92%
Married or cohabiting 61.75% 61.19%
Widowed 6.71% 7.90%
Status not stated 0.84% –
Highest education
None or primary school 33.50% 31.16%
Lower secondary school or professional training 35.60% 32.50%
Upper secondary school 23.90% 29.30%
University and/or higher degree 7.00% 7.04%
Income (EUR)
Mean 28,658.80 24,893.70
Quantile
10% 9822.22 8918.90
25% 14,801.18 13,175.46
50% 24,682.57 20,152.32
75% 34,088.30 30,998.86
90% 47,981.99 44,049.82
Professional status
Entrepreneur 6.32% 1.36%
Professional class 1.83%
Cooperative member 1.36%
Self-employed 5.70% 6.92%
Civil servant or earning employee 33.27% 31.45%
Unemployed worker 4.05% 5.62%
(continued)
Appendix 95
Table A.1 (continued)

Student 12.44% 11.34%
Housewife 13.38% 15.30%
Pensioner 23.89% 20.64%
Other 0.96% 4.17%
Household size
1 member 10.71% 24.89%
2 members 23.20% 27.08%
3 members 23.74% 21.58%
4 members 32.03% 18.96%
5 members 8.49% 5.80%
6 members 1.83% 1.69%
Acknowledgments
The authors are thankful to Carlo Andrea Bollino and to the participants in the 2007
Research Projects of National Relevance (PRIN) seminar (in Milan on June
27, 2011), especially Paolo Bruno Bosco, Lucia Visconti Parisio, and Matteo
Pelagatti, for their helpful suggestions. We also thank the participants in the 27th
US Association for Energy Economics/International Association for Energy Eco-
nomics (USAEE/IAEE) North American Conference (in Houston on September
16–19, 2007), the 30th USAEE/IAEE North American Conference (in
Washington, DC, on October 9–12, 2011), and the Italian Economic Association
(SIE) 52nd Annual Conference (in Rome on October 14–15, 2011). Finally, we want
to express our gratitude for editorial team assistance in polishing this manuscript.

Index
A Electricity market, 55
Age and sex, 79, 80 Elicitation format, 71
and econometric model, 70
Empirical findings, 76
B Estimation results, 79
Broberg and Brännlund (B&B), 72 European and Italian wind power
sectors, 36
European Council, 54
C European scenario, 49
Choice experiment (CEx), 2
Climate and energy package, 53
Climate and energy policies, 28 F
Comparative analysis, 84 Fuel electricity mix, 63
Computer-aided web interviewing method, 68
Contingent valuation (CV), 2, 55
analysis, 66 G
method, 28, 33 Government intervention, 63
Controversial results, 84 Green electricity policy, 60, 61
Green energy funding, 60
The Green Heart of Italy, 33
D Greenhouse gas emissions, 54
Demographic variables, 46, 47
Descriptive results, 38, 40
Diagonal response patterns, 75 H
Double bounded dichotomous choice Household’s direct utility function, 91
(DBDC), 55
I
E Istituto Piepoli, marketing & consulting
Econometric analysis, 44 company, 68
Econometric model, 36, 92–93 Italian electricity market, 64
Economic downturn, 81, 84 Italian electricity production, 64
Electricity bill, 44, 66 Italy incentive mechanisms, 63

100 Index
L homeownership, 8
Lancaster theory, 4 individual characteristics, 7, 8
Local community, 2, 3, 6, 9, 12, 17, 21, 22 installations to power, 63
institutional communication, 22
Italy, 63
M literature, 4–6
Market sustainability, 81, 84 local communities, 2, 9
Meta-analysis, 2–4, 20, 21 local survey, 17–21
Monte Cucco Regional Park, 32 meta-analysis approach, 3, 4
Multiple bounded dichotomous choice meta-analysis regression, 2
(MBDC), 71 in mountain view, 8
non-hydro RE capacity, 63
OECD countries, 62
N place attachment, 3
National Action Plan (NAP), 32 pricing program, 62
National Wind Energy Association, 32 production, 54
Negative externalities, 29 psychological aspects, 22
Not In My Back Yard (NIMBY) qualitative analysis, 12
syndrome, 2, 31 quantitative analysis, 10–12
Numerical Certainty Scale (NCS) approach, 35 REnS, 8
Numerical Scale Method (NSM), 29 SBDC, 60
in Umbria region, 7
USD, 60
P wind energy project, 10
Payment card (PC), 55 windmill, 9
method, 61 WTA and WTP, 15–17
Payment responses, 76 WTP, 2, 13, 15, 56, 61, 62
Policy evaluation, 84 Renewable energy sources (REnS), 27, 54
Positive externalities, 29 advantages, 28
characteristics, 67
cost, 55
Q EUR, 66
Questionnaire, 70 fossil fuels, 28
greenhouse gases emissions, 54
increasing share, 54
R Italy, 54
Regulatory Authority for Electricity and Gas MBDC, 62
(RAEG), 64 mitigation, 60
Renewable electricity (RE), 2 PC method, 61
AUD, 60 policy intervention, 28
in Beijing, 61 willing to pay, 54, 55, 61
climate change, 1 Reviewing literature, 29
CO2 emissions, 21
CO2 reduction, 9
consumers, 21 S
and citizens, 2 Seemingly Unrelated Regression (SUR), 36
DBDC, 60 Single bounded dichotomous choice
empirical methods, 13 (SBDC), 55
environmental and energy scenario, 2 Social variables, 49
framework, 3 Spanish households, 60
government provision, 61 Statistical index, 84
green energy, 2 Stochastic Payment Card (SPC), 71
heterogeneous, 8 Survey design, 67
Index 101
Survey method and questionnaire, 33 meta-regression, 16

Survey questionnaire, 68 methodological variables, 17
Survey respondent and Country resident partial correlations, 20
characteristics, 93 WTP, 17, 19
Survivor function, 79 Willingness to pay (WTP), 2, 4, 8–13, 15, 28
factual variables, 14–15
meta-regression, 14
T methodological variables, 15
Theoretical model, 66 WTA (see Willingness to accept (WTA))
Theoretical model and elicitation format, 36 Wind energy generation, 28
Typical family income, 73 Wind farm development, 29
Wind farm enlargement project
behavioral and perceptual features, 37
U economic values, 36
Umbria, 32, 33 SUR approach, 37
Uncertainty, 41 WTP and WTA, 36, 37
Uncertainty preferences model, 71, 73 CV method, 33
Italian, 31
preliminary analysis, 34
V questionnaire, 35
Variables and descriptive statistics, 73 theoretical model and elicitation format, 36
in Umbria, 32
uncertainty, 34
W wind power sector, 31
Welfare measurement, 38 WTP and WTA, 34
Welfare measures distribution functions, 41 Wind power, 28
Welsh and Poe (W&P) approaches, 72, 74 Wind power development process, 30
Willingness to accept (WTA), 4, 8, 10–13, 28 Wind power generation, 31
factual variables, 15–17 Wisconsin Renewable Energy Programs, 60

The Economic Valuation of Green Electricity

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

The Economic Valuation of Green Electricity

Загружено:

Авторское право:

Доступные форматы

SPRINGER BRIEFS IN ENVIRONMENTAL SCIENCE

Simona Bigerna · Paolo Polinori

More information about this series at http://www.springer.com/series/8868

The Economic Valuation

ISSN 2191-5547 ISSN 2191-5555 (electronic)

Library of Congress Control Number: 2018952352

© The Author(s), under exclusive licence to Springer Nature B.V. 2019

1 Citizens’ Versus Consumers’ Attitudes Toward Renewable

2.4 Theoretical and Econometric Framework . . . . . . . . . . . . . . . . . . . 34

ANEV National Wind Energy Association

Fig. 1.1 WTP and WTA kernel density estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Table 1.1 Metaregression: willingness to pay (WTP) for renewable

Keywords Renewable electricity · Meta-analysis · Willingness to pay · Willingness

Although climate change has been recognized as a scientiﬁc problem of global

© The Author(s), under exclusive licence to Springer Nature B.V. 2019 1

Operationally, we base our framework of analysis on four steps. According to the

1.2.1 Meta-analysis Approach

analysis is deﬁned as “the study of studies” (Holmgren 2007); consequently, the

1.2.2 Review of Related Literature

1.2.3 Topics of Interest

1.2.4 What the Literature Tells Us

In the literature the relationships between individual characteristics and RE accep-

of windmills inﬂuences welfare measures. Indeed, Strazzera et al. (2012) emphasize

1.2.5 Quantitative Analysis: Metaregressions

vðM; X; RG0 Þ ¼ vðM WTP; X; RG1 Þ ð1:1aÞ

The WTP/WTA for a marginal increase in production of RG is deﬁned as the ratio

We assume that U is a linear stochastic utility function for the individual

which reﬂects the change in the economic welfare of an individual caused by a

In these equations, ln (RGn) is the natural logarithm of the average level of

Appropriately transforming Eq. (1.4a) it is possible to analyze WTP/WTA for an

1.2.6 Qualitative Analysis: Local Survey

We have interviewed individuals who live in a municipality close to an onshore wind

1.3 Results and Discussion

A metaregression analysis is estimated11 for WTP/WTA related to RE, considering

1.3.1 Willingness to Pay for Renewable Electricity

1.3.1.1 Factual Variables

1.3.1.2 Methodological Variables

1.3.2 Willingness to Accept and Willingness to Pay

1.3.2.1 Factual Variables

Table 1.2 Metaregression: willingness to pay/willingness to accept (WTP/WTA) for an onshore

The environmental impact of wind projects is analyzed using several factual

1.3.2.2 Methodological Variables

1.3.3 Local Survey Results

Table 1.3 Subsample descriptive statistics

Fig. 1.1 WTP and WTA

Table 1.4 Partial correlations

The global reduction of CO2 emissions needs appropriate policies supported by

Borenstein M, Hedges L, Rothstein H (2007) Introduction to meta-analysis, 1st edn. www.Meta-

Keywords Wind farm enlargement · Contingent valuation · Willingness to pay ·

© The Author(s), under exclusive licence to Springer Nature B.V. 2019 27

Furthermore, REnS produce other important positive externalities such as preserva-

2.2 Related Literature

2.3 Method for Valuing Wind Farm Enlargement

2.3.1 The Scenario: Wind Power Generation in Italy

2.3.2 Case Study: The Monte Cucco Regional Park

Fig. 2.1 Umbria and Monte Cucco Regional Park

2.3.3 Survey Method and Questionnaire

2.4 Theoretical and Econometric Framework

that a well-designed and carefully administered survey can provide consistent,

2.4.1 Theoretical Model and Elicitation Format

WTA ¼ f ðI; LÞ ð2:4Þ

2.4.2 Econometric Model

yia ðWTAÞ ¼ X ia βia þ εia ð2:6aÞ

2.5 Results and Discussion