Is it really all about the return on investment? Exploring private wind energy investors' preferences

Energy Research & Social Science 14 (2016) 22–32

Contents lists available at ScienceDirect

Energy Research & Social Science journal homepage: www.elsevier.com/locate/erss

Original research article

Is it really all about the return on investment? Exploring private wind energy investors’ preferences Johannes Gamel, Klaus Menrad ∗ , Thomas Decker Chair of Marketing and Management of Biogenic Resources, Straubing Center of Science, Weihenstephan-Triesdorf University of Applied Sciences, Petersgasse 18, 94315 Straubing, Germany

a r t i c l e

i n f o

Article history: Received 10 June 2015 Received in revised form 8 January 2016 Accepted 13 January 2016 Keywords: Choice experiment Investors’ preference Private investment Renewable energy

a b s t r a c t Achieving EU climate targets requires an immense volume of investments in renewable energies, especially in the field of wind energy. Private individuals can play an essential role in raising significant parts of the necessary financial resources. This requires, however, a thorough understanding of investors’ preferences. Based on choice experiments by 725 German respondents who intend to invest in wind energy in the near future, this article shows that private individuals’ investment decisions are not only made with profit maximization in mind. Furthermore, this study reveals that an individual’s age, asset valuation and environmental attitude significantly affect the preference for different wind energy investment attributes. The findings of this study have important implications for financial institutions and for policy, as the findings indicate that private individuals are not well informed about many aspects of wind energy investments. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction The EU strategy for a competitive low-carbon economy by 2050 describes scenarios to keep global warming below 2 ◦ C. To reach that goal, an “energy technology revolution” is necessary in order to halve the global CO2 emissions by 2050 compared with 2005 levels [1,2]. Therefore, use of Renewable Energy (RE) has strongly increased within the last decade. However, RE has not yet reached its full potential and contributes only a small fraction (4.5%) to global electricity production [3]. This is partly due to the fact that the reduction of CO2 emissions requires an immense volume of investments in sustainable energy technologies [4–6]. The International Energy Agency estimates that $ 39 trillion worth of cumulative investments in energy supply will be required to keep global warming below 2 ◦ Celsius [2]. Of this amount, $ 8 trillion are necessary for the RE sector, of which wind energy requires most new investments (39%), followed by hydropower (27%), solar (23%) and bioenergy (11%). The required investments can be provided by the public sector through taxation and government expenditure (e.g., feed-in tariffs). In addition, the private sector can play an essential role in obtaining the necessary financial resources [7] and will be indispensable

∗ Corresponding author. Fax: +49 9421187211. E-mail addresses: [email protected] (J. Gamel), [email protected] (K. Menrad), [email protected] (T. Decker). http://dx.doi.org/10.1016/j.erss.2016.01.004 2214-6296/© 2016 Elsevier Ltd. All rights reserved.

in future, because the required sums cannot be provided by government investments alone [8]. Therefore, a combination of public and private financing will be necessary to achieve the targets for a reduction of greenhouse gas emissions [9]. In the field of private financing, individuals are of great importance, with private households contributing a significant share (9%) to global climate financing in 2012 with investments of $ 33 billion in REs [10]. Citizens, in addition to other investor groups such as utilities and other corporate or other financial actors, have provided an important source of finance for RE projects in some countries [11,12] or more specifically for wind energy (13). In Germany, private households owned 50% of onshore wind energy in 2010 [13], whereas energy providers (7%), project developers (21%), funds/banks (16%), and industry (2%) only invested a small amount in wind energy [13]. Thus, private individuals mainly drove the rapid expansion of wind energy in Germany, although this kind of investment can be seen as a special type of investment as they are considered to be riskier than more common financial investments such as fixed-term deposits or savings bonds. For example, they usually require larger minimum investments which are directed towards a concrete wind turbine. This demonstrates the relevance of this group of investors and reveals the importance of a better understanding of private individuals’ investment decisions in wind energy. Behavioral finance examines such investment decisions and argues that investors do not exclusively employ rational decisionmaking [14]. Furthermore, some authors point out that in

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

additional to rational factors, behavioral factors can play an important role in the decision-making process [15]. Nevertheless, private individuals’ investment preferences in RE have not been analyzed systematically. The available literature on the financial engagement of citizens and investment professionals show that investments in REs are motivated by several reasons. Demographic variables tend to affect “willingness to invest” in REs and “willingness to pay” for REs [16–18] and are used for customer segmentation in financial services [19–23]. Further, cognitive variables like knowledge (financial and technical) [24–26] and attitudes (e.g., towards the environment or the power generation system) [27–29] influence individual’s investment preferences. The literature also shows that economic variables such as access to financial resources or household income [30,31] affect the propensity to invest in RE. Against this background and by taking into account private individual’s preferences for wind energy investments, we address the following research questions: (1) which attributes of direct wind energy investments are of particular importance to private individuals and (2) to what extent is the willingness to invest in wind energy influenced by an individual’s age, asset valuation and environmental attitude? 2. Data and methods 2.1. Experimental design The objective of this study is to investigate the investment preferences of private individuals in wind energy. The objective of this study refers to direct investment in specific wind turbines, not to indirect investments such as the purchase of green electricity. One particularly popular method of analyzing individuals’ preferences is conjoint experiment (CE) [32], also referred to as conjoint analysis in marketing literature. This study is based on the prerequisite that investment decisions comply with the fundamental assumptions underlying conjoint analysis [33]. Specifically, it is assumed in a CE that the characteristics of an investment give rise to utility, not the investment per se. Furthermore, wind energy investments are complex products and thus have to be characterized by more than just one attribute. This implies that wind energy investments may have different characteristics to wind turbines and “typical” investment attributes like ROI, duration et cetera. In particular, the location of the wind park can be of great importance for private individuals’ investment decisions. The basic form of conjoint analysis has been adapted over the years in order to overcome certain weaknesses in the traditional method [34–37]. Among the advances are two particular variations of conjoint analysis: (1) Full profile methods, such as choice-based conjoint analysis (CBC), where respondents make simultaneous trade-offs between all attributes of the choice alternatives. (2) Partial profile methods, such as adaptive conjoint analysis (ACA), where respondents are first asked to rank the importance of attributes followed by choice tasks that gradually build up complexity [38]. The term “adaptive” refers to the fact that the computer-administered interview is individualized for each respondent. Adaptive choice-based conjoint analysis (ACBC) is a hybrid method between CBC and ACA that combines the specific characteristics of both methods [39]. For this reason, ACBC is the preferred choice of method, as we argue that a private individual’s choice among different opportunities to invest in wind energy is, in principle, similar to the decision by a customer to buy a product. The over-all complexity of real-life decision making cannot be reached

23

with a survey instrument. Therefore, it is essential to imitate the real decision-making process as closely as possible. ACBC is a well-established method in marketing research to measure customer preferences [40]. Research has utilized conjoint analysis in the discourse on clean energy, energy-efficiency and energy policy [23,41,42] as well as environmental economics [43–47]. Furthermore, conjoint analysis is well suited for investment decisions [48] and has been successfully applied to analysis of investor preferences or financial choices in other studies [32]. Most respondents pay attention to only a few attribute levels when making product choices, especially when it comes to complex product concepts as is the case in this study [49]. Therefore, ACBC screens a wide variety of product concepts but focuses on the subset of most interest to the respondent [38]. This is provided by a fixed sequence of various choice sections. Typically, the computer-administered interview includes three sections that build on each other:

(1) In the Build Your Own (BYO) section, respondents answer questions to identify attributes and levels, as well as to let the respondent determine the preferred level for each attribute. (2) In the Screening Section, the software generates a series of hypothetical investments based on the first section. The customized designs are near-orthogonal, generated by the software “on-the-fly” based on the information provided by the respondent in the BYO section and by following a controlled, randomized process. This allows for “controlled” randomized designs, and leads to a relatively high degree of level balance and statistical efficiency [39]. The different developed product concepts are presented to the respondent in groups of three per screen. Individuals “are not asked to make final choices, but rather just indicate whether they would consider each one a possibility or not a possibility” [39]. (3) Those concepts that passed the Screening Section are transferred to the Choice Task Section (cf. Appendix A) where the alternatives are presented in choice-groups of three. In each task, respondents have to indicate their most favored option. In the subsequent rounds of the tournament, the winning alternatives are measured against each other until the preferred concept is identified [50].

In spite of the adaptive approach and the resulting reduced number of choice tasks, ACBC surveys require more time than conventional approaches to CE, but they are perceived to be more interesting and engaging [39]. Moreover, it produces better predictions for a choice set that was custom-designed for each respondent from concepts preferred in previous choice sets [51]. It is recommended that ACBC is most appropriate for surveys with a large number of attributes (5–12) and with no more than seven levels per attribute [52]. Within this range, ACBC yields lower standard errors than conventional CE approaches, [53]. Therefore, the challenge for CEs is to find the right balance between important standard criteria that would make the choice experiment as realistic as possible and further attributes that would reflect the social influences on private individuals’ decision making—and all this while keeping the complexity for respondents at an appropriate level [54]. A further requirement of CEs is, on the one hand, to include all important attributes of a product and, on the other hand, not to overwhelm the participants with too much information. Therefore, ACBC provides a smart solution by avoiding the cognitive overload of respondents despite a high number of attributes. A particular feature allows respondents to eliminate alternatives with unacceptable attribute levels from their consideration set and then to choose among the remaining alternatives using a more refined

24

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

method [55]. Therefore, ACBC is only possible with the assistance of a computer. 2.2. Attributes and levels In CE, the choice of attributes and their dimensions is sophisticated and crucial. The attributes must be chosen which meet a number of requirements. They have to (1) be relevant to the problem being analyzed, (2) be credible/realistic, (3) be capable of being understood by the sample population and (4) provide a meaningful context [56]. To develop the attributes, an extensive literature review was carried out. The literature on investors’ preferences towards investment attributes is rather scarce, especially for investments in renewable energy sources (RES) or wind energy. Most studies dealing with the evaluation of investment attributes indicate that private individuals as well as investment professionals put their main focus on ROI [29,56–58] and investment amount [59–61]. Further, literature shows that a firm’s reputation and therefore its experience may be crucial to potential investors [60,62]. More recent studies point out the decisive role of participation opportunities, not only in financial terms but also in corporate decisions [61,63]. Additionally, the literature points out that the investment term [9,64] as well as the location of the investment [27,62,65] influence private individuals’ investment decisions. In addition to the literature review, expert interviews with financial service providers, consultants for REs, and wind energy project developers were conducted in 2014 in Germany. The market professionals were asked to list and explain the most important attributes for private investments in wind energy projects. The interviews revealed that in addition to those attributes that have already been discussed in the literature, three other attributes are also considered crucial for private individuals’ wind energy investment decisions and therefore were included in this study. Those are (1) the possibility to exit the investment during the term of the investment, (2) the date of the first payment return, and (3) the type of institution offering the investment. Based on the findings of previous studies and the expert interviews, the CE in this study includes nine attributes with no more than four levels per attribute, which is considered to be a reasonable number [52]. Table 1 shows each attribute, together with the levels presented in the survey. The levels of the attributes are based on relevant real world examples and on the information obtained in the expert interviews. Apart from the attributes Experience and Participation, all attributes consist of four levels. Thus, the applied conjoint design is asymmetric. With respect to the attributes Term and Repayment, it must be emphasized that in the case that an investment has a duration of three years (Term), both, level 2 and level 4 of the attribute Repayment describe the same date of the first payment return. 2.3. Estimation algorithm We estimated the ACBC data using a Hierarchical Bayes (HB) model. This model is called “hierarchical” because it consists of two levels: (1) at the higher level, it is assumed that individuals’ partworths are described by a multivariate normal distribution; (2) at the lower level, it is assumed that the probability that a respondent will choose a particular alternative is governed by a multinominal logit model [66]. In statistical terms, the higher level can be written as ˇi ∼Normal (˛, D) where ␤i is “a vector of part worths for the ith individual, ␣ is a vector of means of the distribution of individuals’ part worths, and

D is a matrix of variances and covariances of the distribution of part worths across individuals” [66]. At the lower level, choices are described by a multinominal logit model. The utility uk that the individual i refers to the kth alternative is defined asuk = xk’ ␤i . The probability of the ith individual choosing the kth alternative in a particular task is



pk =



exp xk ˇi











exp xj ˇj j

where P is “the probability of an individual choosing the kth concept in a particular choice task, and xj is a vector of values describing thejth alternative in that choice task“ [66]. In order to estimate the parameters ␤i , ␣ and D, two different Monte Carlo Markov Chain methods were used. “As the overall procedure to estimate the parameters” [54] a particular technique of Metropolis-Hasting algorithm called Gibbs sampling was used. The estimation of the vectors ␤i of part worths for each individual was operationalized “by a more complex iterative process of Metropolis-Hasting algorithm” [54] by using present estimates of ␣ and D [66]. For a more detailed discussion of the iterative estimation of the parameters see [66]. To ensure convergence, the parameter estimation used 20,000 draws as burn-in of a total of 40,000 draws per respondent. 2.4. Measuring environmental attitude, financial knowledge and asset valuation In order to test the influence of environmental attitudess on private individuals’ wind energy investment preferences, a set of statements (cf. Table 2) were used. We adapted Haws, Winterichs and Naylos “Green Consumer Value” [64] to gauge respondents’ environmental attitudes. Answers were requested on a 5-point Likert scale with values ranging from 1 (totally disagree) to 5 (totally agree). Respondents with higher values are perceived to be more environmentally aware. Additionally, two statements of a questionnaire on investment typology of the German investment company “DekaBank” [67] were adopted to measure financial knowledge of survey participants. Respondents were asked to assess their level of experiences with real and monetary values (cf. Table 2). Answers were given on a 4-point Likert scale with values ranging from 1 (no experience) to 4 (a lot of experience). In order to test the influence of asset valuation, participants rated their financial situation on a 5-point Likert scale with values ranging from 1 (very bad) to 5 (very good). The mean values for the variables environmental attitude, financial knowledge and asset valuation are shown in Appendix B. 2.5. Data collection and characteristics of the sample The sampling frame for this study includes Germans aged 18 years or above who have an intention to invest in wind energy and are experienced in financial investments. To ensure the investment intention and the experience in financial investments of participants, two selection criteria were defined at the beginning of the questionnaire: (1) Respondents’ investment portfolios had to include real1 or monetary2 values at the time of the survey (very conservative financial products in the portfolio also led to exclusion3 ), and (2) they

1 Property, shares, open equity funds, closed equity funds, real estate funds, raw materials, precious metals. 2 Fixed-income securities, bonds, balanced funds, warrants, certificates. 3 Fixed-term deposit, savings bond, savings book, savings plan, call money account.

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

25

Table 1 Attributes and attribute levels in the CE experiment. Attribute

Description

Levels

Investment

The minimum investment amount in D to enter the offer

Term

The duration of the investment offer in years

ROI

Return on investment per year

Location

Distance of the investment object (wind turbine) to the customer

Exit

Possibility to exit the investment during the duration

Participation

Investor participation in firm´ıs decision making process

Repayment

Date of the first payment return

Issuer

Type of institution offering the investment

Experience

Experience of the company in charge of the realization of the investment objective

500D 3,000D 10,000D 50,000D 3 years 7 years 10 years 20 years 2.5% 5.5% 8.5% 11.5% Neighborhood (radius of 5 km) Region (radius of 30 km) Germany (outside 30 km radius) Outside Germany Possible at any time Possible after the first year Possible from the mid-term Not possible One voice per stakeholder Prorated on business assets No voting rights (no participation) After the first year After the third year In the mid-term At the end of the term Citizens’ cooperative Regional company/bank Nationwide company/bank National/international fund New entrants to the market (no experience) Recently on the market (small experience) Established on the market (great experience)

had to intend to invest in wind energy projects within the next three years. The questionnaire, including the computer-based ACBC experiment, was designed and conducted with Sawtooth Software, which is the most common software solution for the design and analysis of CE in marketing research [68]. To recruit respondents, we subcontracted a market research company which has a panel of 70,000 active users (100,000 total users). The panel users were recruited though social media marketing, search engine marketing, on-site surveys, and mingle Blogs as well as from affiliate partner companies. As an incentive, participants received fixed compensation depending on the time they needed to complete the questionnaire (1.50 D < 15 min, 2.00 D > 15 min). In June 2014, 18,736 adult panel users were randomly selected by the market research company and invited to participate via e-mail in several rounds until the desired number of participants who met the selection criteria was reached. The invitation included no indication of the survey topic. 11,726 panel users accepted the invitation, which corresponded to a response rate of about 63%. Out of these, 2875 respondents met

the selection criterion (1), but only a minority of them stated that they had an intention to invest in wind energy within the next three years, leading to a reduced sample of 725 respondents. A comparison of key variables between the sampling frame and our sample is presented in Table 3. The sample fits the target population quite well in terms of socio demographic structure. Only the gender distribution displays greater differences with a higher proportion of men. Furthermore, people with a university education seem to be somewhat overrepresented in the sample, but similar differences with respect to education level have been found in other studies [69,70].

3. Results 3.1. Relative importance of attributes Our results are based on the data from725 private individuals with experience in real or monetary value investments as well as an intention to invest in wind energy. A total of 28,117

Table 2 Environmental attitude and financial knowledge statements. No. 1 2 3 4 5 6 1 2

Statement Environmental attitude It is important to me that the products I use do not harm the environment. I consider the potential environmental impact of my actions when making many of my decisions. My purchase habits are affected by my concern for our environment. I am concerned about wasting the resources of our planet. I would describe myself as environmentally responsible. I am willing to be inconvenienced in order to take actions that are more environmentally friendly. Financial knowledge What knowledge or experience do you have in the field of monetary assets (e.g. fixed-income securities, bonds, balanced funds) ? What knowledge or experience do you have in the field of real assets (e.g. shares, open equity funds, precious metals) ?

26

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

Table 3 Socio-demographic characteristics of the sample. Variable

Population (Sampling frame)

Sample

Gender Male Female

0.55 0.45a

0.63 0.37

Age 18 to under 30 years 30 to under 40 years 40 to under 50 years 50 to under 60 years 60 to under 70 years 70 years or older

0.06a 0.14a 0.26a 0.22a 0.14a 0.17a

0.07 0.17 0.24 0.28 0.19 0.05

Education Without school degree Secondary modern school degree High school degree Academic high school degree University or college degree

0.01a 0.33a 0.29a 0.12a 0.25a

0.00 0.11 0.27 0.21 0.41

a

Net household income per month under 900D 900 to under 1.300D 1.300 to under 1.500D 1.500to under 2.000D 2.000 to under 2.600D 2.600 to under 3.600D 3.600 to under 5.000D 5.000D and more a b

0.04b 0.06b 0.05b 0.12b 0.17b 0.24b 0.21b 0.11b

0.02 0.04 0.04 0.11 0.16 0.24 0.22 0.17

MDS online [71]. DESTASIS statistisches Bundesamt [72].

choice tasks were performed, resulting in an average of 38.8 tasks per respondent (including build-your-own, screening and choice tournament).4 Table 4 displays the average importance values of attributes as a result of the HB estimation procedure [66]. The attributes are ranked according to their importance. As expected, typical capital investment criteria (Investment, Term and ROI) were the most important to the respondents in our sample. This is in line with results of previous RE investment studies [58,73] as well as with existing literature on private investor behavior [60,62]. The high importance of ROI is in line with previous studies which show that ROI is indeed one of the most important investment criteria but usually not the most important one [9,62]. The large utility of the attribute Investment can be explained by the attribute’s limiting character if one intends to invest in a specific project. Furthermore, Table 4 shows that the attributes Location, Exit, Participation,

4 The exact number of choice tasks per respondent differs due to the adaptive nature of the ACBC interviewing procedure.

Table 4 Average relative importance values of attributes based on hierarchical Bayes model. Attribute

Average importance%a

Std. Dev.

Investment Term ROI Location Exit Participation Issuer Repayment Experience Total

21.11 17.13 15.30 8.79 8.40 7.90 7.73 7.59 6.05 100

8.358 7.163 7.997 3.765 3.569 4.038 4.174 3.471 3.565

a The relative importance values for each attribute are calculated by taking the difference between the highest and the lowest part-worth utility within each attribute and scaling this value to 100% across attributes [50].

Table 5 Part-worths utilities of the different attribute levels for the decision to invest in wind energy. Attributes and levels

Coefficienta

Std. Dev.

Investment 500D 3,000D 10,000D 50,000D

−1.072 −1.267 −0.296 −2.635

1.683 1.058 1.158 1.865

Term 3 years 7 years 10 years 20 years

−1.039 −0.767 −0.359 −2.165

1.289 0.836 0.718 1.511

ROI 2.5% 5.5% 8.5% 11.5%

−1.740 −0.060 −0.684 −0.995

1.700 0.587 0.708 1.483

Location Neighborhood Region Germany Outside Germany

−0.028 −0.392 −0.584 −0.947

0.445 0.588 0.432 0.723

Exit At any time After the first year From the mid-term Not possible

−0.687 −0.325 −0.184 −0.828

0.639 0.390 0.493 0.610

Participation One voice per stakeholder Prorated to business assets No voting rights

−0.428 −0.433 −0.860

0.537 0.522 0.528

Repayment After the first year After the third year From the mid-term At the end of the term

−0.512 −0.153 −0.014 −0.679

0.635 0.556 0.425 0.560

Issuer Citizens cooperative Regional company/bank Nationwide company/bank National/international fund

−0.129 −0.351 −0.004 −0.484

0.791 0.417 0.544 0.715

Experience New entrance to the market Recently on the market Established on the market Number of observations

−0.326 −0.311 −0.638 28,117

0.468 0.405 0.639

a Coefficient estimates are equal to the posterior population means across the saved draws, interval-scaled and zero-centered within attributes.

Repayment, and Issuer are on almost the same moderate importance level. Although it has been demonstrated in a previous study that great experience can be much more important for RE investment decisions than lowest price (Investment) or best technology [23], the investment attribute Experience is ranked lowest according to our analysis. However, it must be taken into account that experience refers to the project developer (i.e., company in charge of the realization of the investment objective), in our experiment but not to the institution offering the investment. 3.2. Utilities of attribute levels The following results provide additional information of the average effect of a particular attribute level on the respondent’s decision to invest in wind energy. Positive values indicate an increase; negative values show a decrease in the individual utility level. Table 5 displays effect-coded raw utilities [50]. Since part-worth utilities are interval data, which are scaled to an arbitrary additive con-

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

stant and summed to zero within each attribute, it is not possible to directly compare utility values across attributes [54]. The utility for the attribute investment is highest for 3000D . Smaller and especially larger investment amounts obtain lower utility levels (cf. Table 5). The preferences for the attribute levels related to Term, ROI, Exit and Repayment all follow a distinct order i.e., the levels indicating the shortest bond (3 years), highest financial gain (11.5%), the most flexible binding contract (exit at any time), and fastest return (after the first year) gain the highest part-worth estimates. The utility for the attribute ROI is relatively low for both the 2.50% level and the 5.50% level but increases for the 8.50% level as well as for the 11.50% level. This might indicate that respondents regard wind energy as a potentially risky investment and expect a return in the range of the two upper ROI levels for such investments. The part-worths for the first two levels of the attribute Participation are approximately equal. The utility is negative for the level “no voting rights”. The results for the attribute Experience show that private individuals prefer project developers which are established on the market. Project developers who are new on the market or who have just recently started operating have negative part-worth utilities. Furthermore, the results show that “citizens’ cooperative” or “regional company/bank” is clearly preferred over “nationwide company/bank” or “national/international fund”. Special attention should be paid to the preferences for the attribute Location. The “Not In My Backyard” (NIMBY) principle [74] seems to be confirmed, as the part-worth utilities increase with distance. But when the investment object is situated outside Germany, the part-worth drops strongly to negative. One possible explanation might be that respondents are familiar with the German wind energy regulations (in particular the EEG (Renewable Energy Act)) but have low knowledge about (regenerative) energy policies in other countries. Nevertheless, the slightly negative part-worth utility representing a wind turbine in the neighborhood (5 km) could indicate that respondents prefer the direct neighborhood as a location in comparison to wind turbines outside Germany.

3.3. Influence of attitudes and socio-demographics One objective of this study is to test whether socio-demographic and psychographic variables influence private individuals’ preferences for wind energy investments. Therefore, four variables (age, asset valuation, environmental attitude, and financial knowledge) were included as covariates into the model estimation. Generally, the integration of covariates could provide additional information and therefore improve parameter estimates and preference share predictions. To investigate the influence of environmental attitude and financial knowledge, the mean value of the asked statements (cf. Table 2) was calculated and used within the parameter estimation. Respondents with higher values are perceived to have a more positive environmental attitude and greater knowledge about realand monetary investments. To facilitate the interpretation of the results of the parameter estimation (cf. Table 6), the factor or item scores were zerocentered [74]. In this case, the arithmetic mean is subtracted from all measured values. Thus, the average of the centered data is zero, and positive centered scores indicate an above-average agreement to the statements formulated in the items. The Intercept displays the utility value of an attribute level in case the model parameters of all the four integrated variables are equal to zero. Depending on one of the covariates, the individual utility␤x of a person for any attribute level x can be calculated as ˇx = Interceptx + Paramerx × ExpressionCovariate

27

The None-Parameter can be interpreted as an investmentthreshold. This value represents the utility for not investing. It is assumed that a person invests in the respective investment concept if the sum of the utility values for the attribute levels exceeds the None-Parameter. The None-Parameter is not directly part of the ACBC but is estimated from the binary choices of the Screening Section (cf. section 2.1). Therefore, it must be stressed that the ACBC’s None-Parameter is generally calculated slightly too high [39]. The estimates presented in Table 6 show that age, environmental attitude and asset valuation affect private individuals’ preferences for wind energy investments, whereas financial knowledge seem to have almost no significant influence. With increasing age, the utilities for wind energy investments decrease with a large investment amount, long term, inflexible exit options, and late onset of repayments. At the same time, the investment-threshold (None-Parameter) increases significantly. This suggests that wind energy investments are of lower attractiveness for older people, which is in line with existing literature [21,24]. With growing environmental attitude, the investmentthreshold decreases, indicating that pro-environmental individuals are more willing to invest in wind energy. Furthermore, they seem to accept financial disadvantages as their utilities for lower ROI, no exit options and a late repayment of deposits and profits at the end of the term increase significantly. Previous studies of investment motivations in wind energy [31], REs [28] or other Social Responsible Investments [29] describe similar findings. Moreover, for people with greater environmental awareness, the aforementioned NIMBY principle cannot be confirmed as their utility for the attribute Location increases significantly for a wind energy investment in the neighborhood. This is in line with other research that explored the NIMBY principle in general [75,76] and with Van der Loo [77] who even identified the opposite phenomenon to NIMBY, called PIMBY (Please In My Backyard), which appears when wind turbines are seen as a socially acceptable financial investment. With a more positive asset valuation, the utilities for large investments, long-term investments and lower ROI increase significantly. The decreasing investment-threshold when assets are assessed more positively indicates that people with greater financial resources have a higher willingness to invest in wind energy. This finding is consistent with other studies analyzing differences between investors and non-investors in REs [9,17,30]. In contrast to the parameter estimation for the covariates age, environmental attitude and asset valuation, we found hardly any significant influences from financial knowledge on private individuals’ preferences for wind energy investments. This finding is in contrast to existing literature which suggests that individuals with financial knowledge [25,78] are the potential investors with the highest “willingness to invest” in RE.

4. Discussion 4.1. Methodological discussion The study data were collected through an online questionnaire. Generally, online surveys are considered to be more flexible, offer faster data collection and are more cost effective than traditional written surveys. However, internet access varies significantly between countries and population groups within countries [79]. A clear majority of the population has continuous access to the internet in Germany [80]. There is only one significant difference in access and use of the internet in Germany between people older than 60 years and the rest of the population [81]. Although this group is underrepresented in the sample, the total number of

28

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

Table 6 Results of the parameter estimation with use of covariates. Attributes and levels

Intercept

Age

Financial knowledge

Environmental attitude

Asset valuation

Investment 500D 3,000D 10,000D 50,000D

−1.18* −1.20* 0.18 −2.56*

−0.14* −0.20* −0.05* −0.40*

−0.07* −0.11* −0.08* −0.10*

−0.00* −0.03* −0.02* −0.05*

−0.79* −0.36* −0.38* −0.76*

Term 3 years 7 years 10 years 20 years

−1.27* −0.76* −0.26* −2.29*

−0.26* −0.11* −0.10* −0.27*

−0.00* −0.06* −0.06* −0.12*

−0.04* −0.16* −0.05* −0.15*

−0.22* −0.17* −0.13* −0.52*

ROI 2.5% 5.5% 8.5% 11.5%

−1.50* 0.00 −0.61* −0.89*

−0.16* −0.10* −0.01* −0.05*

−0.42* −0.09* −0.16* −0.17*

−0.45* −0.07* −0.20* −0.32*

−0.30* −0.01* −0.00* −0.29*

Location Neighborhood Region Germany Outside Germany

0.08 −0.39* −0.47* −0.94*

−0.05* −0.09* −0.00* −0.14*

−0.12* −0.11* −0.04* −0.19*

−0.14* −0.09* −0.02* −0.21*

−0.02* −0.10* −0.07* −0.06*

Exit At any time After the first year From the mid-term Not possible

−0.63* −0.39* −0.23* −0.79*

−0.08* −0.08* −0.01* −0.15*

−0.00* −0.01* −0.02* −0.00*

−0.20* −0.01* −0.03* −0.22*

−0.11* −0.06* −0.00* −0.05*

Participation One voice per stakeholder Prorated to business assets No voting rights

−0.53* −0.34* −0.87*

−0.03* −0.08* −0.11*

0.00* 0.00* −0.01*

−0.11* −0.16* −0.05*

−0.05* −0.06* −0.00*

Repayment After the first year After the third year From the mid-term At the end of the term

−0.51* 0.07 −0.13* −0.72*

−0.10* −0.01* −0.02* −0.12*

−0.09* −0.03* −0.03* −0.03*

−0.12* −0.04* −0.06* −0.22*

−0.01* −0.04* −0.03* −0.01*

Issuer Citizens cooperative Regional company/bank Nationwide company/bank National/international fund

0.13 −0.40* −0.03* −0.50*

−0.09* −0.03* −0.07* −0.05*

−0.05* −0.01* −0.04* −0.01*

−0.22* −0.02* −0.27* −0.07*

−0.10* −0.01* −0.04* −0.07*

Experience New entrance to the market Recently on the market Established on the market None-Parameter

−0.33* −0.43* −0.76* −3.51*

−0.07* −0.04* −0.11* −0.59*

−0.05* −0.03* −0.02* −0.20*

−0.16* −0.06* −0.10* −0.56*

−0.03* −0.03* −0.06* −0.67*

* Significant at the 0.05 level (Parameter estimates are significantly positive/negative if more than 95% of the estimated parameter values in each iteration of the algorithm are positive/negative [74]).

people aged 60 or over is sufficient for reliable estimations of agedrelated effects. The sampling procedure may give rise to statistical issues such as “panel conditioning”, which means that respondents’ answers are influenced by prior (other) interviews. This could affect the resulting estimates. However, web-based access panels can also be advantageous since respondents can be looked upon as experienced participants who give more precise and truthful answers [82]. To select the respondents, restrictions were incorporated into the questionnaire to ensure that (1) respondents were 18 years or older, (2) their investment portfolio included real or monetary values and (3) that they intended to invest in wind energy projects within the next three years. These restrictions were important because people with financial experience can assess wind energy investments more realistically .In addition, it can be expected that people with intentions to invest in wind energy are not only better informed about the financial aspects of wind turbines but also about technical and social aspects. For complex products such as wind energy investments, the implementation of a CE with unexpe-

rienced people in the field of private investments would therefore make little sense. Due to the strict restrictions, the age and gender distribution as well as education levels had some deviations from the sampling frame. Although this might weaken the representative character of this study, we would like to highlight the fact that this study is based on 28,117 experimental investment decisions by 725 private individuals. In contrast to other studies dealing with investment decisions the validity of this study is not limited due to its small sample size and/or the specific subgroups of respondents. When comparing sample sizes, it should be taken into account that some of the previous studies investigated professional investors, which represent a much smaller group than in the case of retail investors [32,83]. The environmental attitude of participants was determined by the use of Haws, Winterichs and Naylos “Green Consumer Value” [64]. Green consumers are defined as “those who have a tendency to consider the environmental impact of their purchase and consumption behaviors. As such, consumers with stronger GREEN values will tend to make decisions consistent with environmen-

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

tally sustainable consumption” [64]. This study did not investigate environmental awareness (no measurement of environmental attitudes) but used the “Green Consumer Value” as a predictor of environmental attitudes. Instead of using the absolute amount of assets to test the influence of private individuals’ assets on investment preferences, the variable asset valuation was used as a covariate within the HBestimation, where participants had to assess their own financial situation on a 5-point Likert scale ranging from 1 (very bad) to 5 (very good). As this study found almost no evidence of influence from financial knowledge on private individuals’ wind energy investment preferences, the method used to measure this has to be re-examined. Respondents were asked to rate their experience in real and monetary values in the form of two self-assessment questions. As mentioned, this method is indeed used by financial institutions to analyze the knowledge of their clients, but it might not have worked well in the context of our study using a computerbased experiment.One important requirement of CEs is, on the one hand, to include all important attributes of a product and, on the other hand, not to overwhelm the participants with too much information. Due to this issue, an adaptive form of CE, namely Adaptive Choice Based Conjoint (ACBC) was used in order to counteract the possibility of cognitively overwhelming the respondents by only using those attributes which the participants assessed as being important for them. Nevertheless, it may have been the case that some individuals were overwhelmed by the number of attributes included in the experiment despite this adaptive approach.

29

financial support of more pro-environmental people, wind energy projects can also be realized at non-optimal wind locations, and thus make a major contribution to a decentralized energy supply. The parameter estimation of this study revealed that people with greater financial resources have a higher willingness to invest in wind energy, which is in line with the findings of Bollinger and Gillingham as well as Drury et al. concerning the diffusion of photovoltaic energy in California (U.S.A). 5. Conclusion In order to increase the proportion of RE in Germany’s energy portfolio and to achieve EU climate targets, an immense volume of investments in RE is required. This can only be achieved with support from retail Investors. As shown by previous studies in the field of social science, those investment decisions often necessitate tradeoffs between differing interests or the defined attributes of a specific investment [84]. Consequently, some characteristics of an investment are perceived to be more important than others. A better understanding of how (potential) investors react to such tradeoffs when making decisions about investing in wind energy can contribute important insights to both energy research and social science [85]. As our paper investigates private individuals’ preferences for wind energy investments as well as the role of defined factors which influence these preferences, it makes a contribution to this field of research. 5.1. Implications

4.2. Discussion of study findings This study reveals that the minimum investment amount to enter a specific wind energy investment is the most important attribute for the survey participants who preferred low investment sums. This finding is in line with previous studies of investment decisions in RE: Ku and Yoo [59] (Korea) and Bergmann et al. [56] (Scotland) investigated the willingness to invest in RE by financing RE projects through an annual surcharge of the electricity bill and found that in two countries the willingness to invest in RE largely depends on a low annual increase in household electricity costs resulting from the expansion of RE projects. Therefore, the minimum investment amount to enter a specific wind energy investment should be relatively low to increase the interest of potential private investors. This seems to be the case for different countries, but could be substantiated in future research. The study further shows that the term of the investment and the ROI have a major impact on private individuals’ investment decisions. The preferences related to the investment term and ROI all follow a distinct order in the sense that the levels indicating the shortest term (3 years) and the highest financial return (11.5%) gain the highest utility. Aguilar and Cai [9] as well as Clark-Murphy and Soutar [62] found similar findings in studies investigating investment decisions made by individuals in the U.S.A and in Australia respectively, indicating that private investors in different countries seem to have at least similar interests concerning the economic outcomes of their RE investments. Moreover, our study found influencing effects from sociodemographic (age), psychographic (environmental attitude) variables and the self-assessment of the individual financial situation (asset valuation) on both the willingness to invest and investment preferences in wind energy. It was shown that people with higher environmental attitudes are more likely to invest in wind energy and even seem to accept financial disadvantages for such “environmentally-friendly” projects. This finding is consistent with previous studies investigating investment decisions in RE by Swiss citizens [31] and European individual investors [29]. With the

The decrease in RE investments in Germany in the last few years is mainly due to the sharply decreasing feed-in tariffs for RE power according to the EEG. In the near future, investments in wind energy will have to be economically viable without (high direct public) subsidies, as the feed-in tariffs decrease every year by1.5% in Germany. Thus, investments in wind turbines situated in Germany have to prove their economic self-sufficiency in the years to come. Under the expected future conditions, the decision of private individuals in Germany to invest in wind energy is comparable to investment decisions made by households in countries outside Germany without government subsidies for wind energy. Additionally, potential investors in Germany also have the possibility to invest in wind turbines located outside Germany, as simulated in our CE experiment. However, the respondents clearly rejected this option showing that “this international investment horizon” might not be in the focus of potential investors in Germany so far, or they did not assess it as being attractive. In order to change this situation, financial institutions could inform their customers about the situation of RE in foreign markets and offer subsequently appropriate investment products. The study shows that as expected, older people tend to avoid large investments, long investment terms and inflexible exit options. However, the older age group is of particular importance, since they own large available assets [86]. Therefore, financial institutions should put a focus on older people by aligning their investment offers specifically to this group. This applies especially to locally based banks, since a large proportion of seniors put their trust in these financial institutions. And such regionally operating or cooperative banks are also positively assessed as issuers of wind energy projects according to the results of our study. Not only financial institutions but also the policy must contribute to the further expansion of wind power in the coming years. Since investments in wind turbines are often not perceived as attractive, policy should ensure more favorable conditions, not necessarily in the form of direct subsidies but in form of indirect subsidies such as a preferred connection of RE plants to the

30

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

Fig. A1. Screenshot of ACBC choice task section.

electricity grid or guaranteed feed-in tariffs for a certain time period.

Appendix A. (See Fig. A1)

5.2. Limitations and further research It must be emphasized that the context of this paper is limited to the use of private individuals’ preferences for wind energy investments as a predictor for their likelihood to invest. The actual investment decision could not be studied due to methodological issues, practical hindrances and reasons of privacy policy. Further, it must be pointed out that the findings on the relative importance of different wind energy investment attributes only allow conclusions about the nine attributes included in this study. Even though the extensive literature review and qualitative expert interviews show the relevance of these attributes, unobserved attributes and factors could also affect the results of the analysis. The influence of factors and attributes not included in our study could be investigated in future research. Another limitation that is worth mentioning is the measurement of the variable asset valuation. Participants had to assess their own financial situation on a 5-point Likert scale ranging from 1 (very bad) to 5 (very good), which is quite a subjective measure as the reference level for what is perceived as “very good” and “very bad” varies per person. Nevertheless, despite this limitation we used this subjective rating as respondents refrained from indicating their financial situation directly. Although this article reveals important implications for other countries, this study focuses on German individuals and does not allow comparison between different nations. Therefore, future studies should take up the research question in order to allow a detailed analysis of similarities and differences in wind energy investment preferences between countries. Furthermore, a comparison between private individuals’ wind energy investment preferences and the properties of the investments they have already made could give interesting insights into their investment decision process. As this study provides no evidence as to why respondents expect a return in the range of 8.50–11.50%, future research should aim to identify retail investor’s past investments in order to elicit their ROI expectations and the benchmark they use for assessing the expected ROI from wind energy investments.

Appendix B. (See Table A1) Table A1 Non-demographic characteristics of the sample. Variable

Measurement scale

Mean value

Environmental attitude Financial knowledge Asset valuation

5-point Likert scale 4-point Likert scale 5-point Likert scale

4.02 2.80 2.23

References [1] [2] [3] [4]

[5] [6] [7]

[8]

[9] [10] [11]

[12] [13] [14]

IEA, Energy Technology Perspectives 2008. Paris. (2008). IEA, World Energy Investment Outlook 2014 Special Report. Paris. (2014). IEA, Key World Energy Statistics 2013. Paris. (2013). E. Menichetti, Renewable Energy Policy Risk and Investor Behaviour: An Analysis of Investment Decisions and Investment Performance, Dissertation. St. Gallen, 2010. OECD, OECD Environmental Outlook to 2030. Paris. (2008). UNFCCC, Investment and Financial Flows to Address Climate Change. Bonn. (2007). A. Masini, E. Menichetti, The impact of behavioural factors in the renewable energy investment decision making process: conceptual framework and empirical findings, Energy Policy 40 (2012) 28–38. R. Wüstenhagen, E. Menichetti, Strategic choices for renewable energy investment: conceptual framework and opportunities for further research, Energy Policy 40 (2012) 1–10. F.X. Aguilar, Z. Cai, Exploratory analysis of prospects for renewable energy private investment in the U.S. Energy, Economics 32 (2010) 1245–1252. Climate Policy Initiative. The Global Landscape of Climate Finance 2013. (2013). K. Gehles, Bürger finanzieren mit. Praxisbeispiele für Anlageprodukte wie Klimasparbriefe, Nachrangdarlehen oder stille Beteiligungen, Informationen zur Raumentwicklung (2012) 535–542. Ö. Yildiz, Financing renewable energy infrastructures via financial citizen participation –the case of Germany, Renew. Energy 68 (2014) 677–685. Trend: research. Marktakteure Erneuerbare-Energie-Anlagen in der Stromerzeugung. Köln (2011). D. Kahneman, A. Tversky, Prospect theory: an analysis of decision under risk, Econometrica 26 (1979) 263–291.

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32 [15] J. Jordan, K.P. Kaas, Advertising in the mutual fund business: the role of judgmental heuristics in private investors’ evaluation of risk and return, J. Finance Serv. Mark. 7 (2002) 129–140. [16] V.L. Bajtelsmit, A. Bernasek, Why do women invest differently than men? Finance Couns. Plann. 6 (1996) 1–10. [17] E. Drury, M. Miller, C.M. Macal, D.J. Graziano, D. Heimiller, J. Ozik, et al., The transformation of southern California’s residential photovoltaics market through third-party ownership, Energy Policy 42 (2012) 681–690. [18] J. Zarnikau, Consumer demand for ‘green power’ and energy efficiency, Energy Policy 31 (2003) 1661–1672. [19] N. Capon, G. Fitzsimons, R. Alan Prince, An individual level analysis of the mutual fund investment decision, J. Finance Serv. Res. 10 (1996) 59–82. [20] T. Harrison, Segmenting the market for retail financial services. The international review of retail, Distrib. Consum. Res. 5 (1995) 271–286. [21] A. Diamantopoulos, B.B. Schlegelmilch, R.R. Sinkovics, G.M. Bohlen, Can socio-demographics still play a role in profiling green consumers? A review of the evidence and an empirical investigation, J. Bus. Res. 56 (2003) 465–480. [22] M. J. Bürer, Public Policy and Clean Energy Private Equity Investment. Dissertation. Zürich, (2008). [23] M. Loock, Going beyond best technology and lowest price: on renewable energy investors’ preference for service-driven business models, Energy Policy 40 (2012) 21–27. [24] N. Kottke, Entscheidungs- und Anlageverhalten von Privatinvestoren: Psychologische Aspekte der Wertpapieranlage, 1st ed., Dt. Univ.-Verl, Wiesbaden, 2005. [25] K.L. Lim, G.N. Soutar, J.A. Lee, Factors affecting investment intentions: a consumer behaviour perspective, J. Financ. Serv. Mark. 18 (2013) 301–315. [26] N. Murovec, R.S. Erker, I. Prodan, Determinants of environmental investments: testing the structural model, J. Clean. Prod. 37 (2012) 265–277. [27] F. Distler, Bürgerwindräder—Motive privater Investoren. Bachelorarbeit. Fresing, (2013). [28] K. Granoszewsk, A. Spiller, C. Reise, O. Mußhoff. Die Diffusion regenerativer Energien in der deutschen Landwirtschaft—Investitionsverhalten in einem politisch induzierten Markt, (2011). [29] J. Nilsson, Investment with a conscience: examining the impact of pro-social attitudes and perceived financial performance on socially responsible investment behavior, J. Bus. Ethics 83 (2008) 307–325. [30] B. Bollinger, K. Gillingham, Peer effects in the diffusion of solar photovoltaic panels, Mark. Sci. 31 (2012) 900–912. [31] G. Walter, Sozialsychologische Akzeptanz von Windkraftprojekten an potentiellen Standorten: Eine quasiexperimentelle Untersuchung. Zürich, (2012). [32] S. Lüthi, R. Wüstenhagen, The price of policy risk—empirical insights from choice experiments with European photovoltaic project developers, Energy Econ. 34 (2012) 1001–1011. [33] K.J. Lancaster, A new approach to consumer theory, J. Polit. Econ. 74 (1966) 132. [34] K. Backhaus, Backhaus-Erichson-Plinke-Weiber, B. Erichson, W. Plinke, R. Weiber, Multivariate Analysemethoden: Eine anwendungsorientierte Einführung, 11th ed., Springer, Berlin [u.a.], 2006. [35] B.K. Orme, Which conjoint method should I use? Sequium, WA, 2007. [36] R.L. Priem, D.A. Harrison, Exploring strategic judgment: methods for testing the assumptions of prescriptive contingency theories, Strat. Manage. J. 15 (1994) 311–324. [37] B. Skiera, S. Gensler, Berechnung von Nutzenfunktionen und Marktsimulationen mit Hilfe der Conjoint-Analyse. Teil II. Wirtschaftswissenschaftliches Studium WiSt, Zeitschrist für Studium und Forschung 2002 (2016) 258–263. [38] Sawtooth Software. The ACA/Web v6.0 Technical Paper. Sequium, WA, (2007). [39] Software, ACBC Technical Paper. Orem, UT, (2014). [40] S. Kaufmann, K. Künzel, M. Loock, Customer value of smart metering: explorative evidence from a choice-based conjoint study in Switzerland, Energy Policy 53 (2013) 229–239. [41] A. Burkhalter, J. Kaenzig, R. Wüstenhagen, Kundenpräferenzen für leistungsrelevante Attribute von Stromprodukten, ZS Energy Wirtsch. 33 (2009) 161–172. [42] K. Sammer, R. Wüstenhagen, The influence of eco-labelling on consumer behaviour –results of a discrete choice analysis for washing machines, Bus. Strat. Environ. 15 (2006) 185–199. [43] J. Ahn, G. Jeong, Y. Kim, A forecast of household ownership and use of alternative fuel vehicles: a multiple discrete-continuous choice approach, Energy Econ. 30 (2008) 2091–2104. [44] J.F. Casey, J.R. Kahn, A.A. Rivas, Willingness to accept compensation for the environmental risks of oil transport on the Amazon: a choice modeling experiment, Ecol. Econ. 67 (2008) 552–559. [45] S. Chattopadhyay, The random expenditure function approach to welfare in RUM: the case of hazardous waste clean-up, Resour. Energy Econ. 31 (2009) 58–74. [46] S. Farber, B. Griner, Valuing watershed quality improvements using conjoint analysis, Ecol. Econ. 34 (2000) 63–76. [47] H. Glenn, P. Wattage, S. Mardle, T. van Rensburg, A. Grehan, N. Foley, Marine protected areas—substantiating their worth, Mar. Policy 34 (2010) 421–430. [48] D. Chew, J. Graham, T. Harris, A. Hutton, C. Kantor, T. King, et al., Roundtable on Corporate Disclosure, J. Appl. Corporate Finance 16 (2004) 36–62. [49] B. Orme, ACBC Technical Paper. Sequium, Washington, (2009).

31

[50] B.K. Orme, Getting Started with Conjoint Analysis: Strategies for Product Design and Pricing Research, Research Publishers LLC, Glendale, CA, USA, 2014. [51] R.M. Johnson, B.K. Orme, A New Approach to Adaptive CBC. Sequium, Washington, (2007). [52] R. Bauer, K. Menrad, T. Decker, Adaptive hybrid methods for choice-based conjoint analysis: a comparative study, IJMS 7 (2015) 1–14. [53] C.N. Chapman, J.L. Alford, C. Johnson, R. Weidemann, M. Lahav. CBC vs. ACBC: comparing Results With Real Product Selection. in: Sawtooth Software I, (ed.). Sawtooth Software Conference Proceeding. Sequium, WA, p. 199–206, (2009). [54] R. Wuebker, N. Hampl, R. Wüstenhagen, The Strength of strong ties in an emerging industry: experimental evidence of the effects of status hierarchies and personal ties in venture capitalist decision making, Strate. Entrepreneurship J. 9 (2015) 167–187. [55] R. Johnson, A Perspective on Adaptive CBC: (What Can We Expect from Respondents?). Sequium, Washington, (2008). [56] A. Bergmann, N. Hanley, R. Wright, Valuing the attributes of renewable energy investments, Energy Policy 34 (2006) 1004–1014. [57] A. Dimitropoulos, A. Kontoleon, Assessing the determinants of local acceptability of wind-farm investment: a choice experiment in the Greek Aegean Islands, Energy Policy 37 (2009) 1842–1854. [58] M. Getzner, S. Grabner-Kräuter, Consumer preferences and marketing strategies for green shares, Int. J. Bank Mark. 22 (2004) 260–278. [59] S. Ku, S. Yoo, Willingness to pay for renewable energy investment in Korea: a choice experiment study, Renew. Sustain. Energy Rev. 14 (2010) 2196–2201. [60] R.A. Nagy, R.W. Obenberger, Factors influencing individual investor behavior, Financial Anal. J. 50 (1994) 63–68. [61] M. Achtnicht, Do environmental benefits matter? Evidence from a choice experiment among house owners in Germany, Ecol. Econ. 70 (2011) 2191–2200. [62] M. Clark-Murphy, G.N. Soutar, What individual investors value: some Australian evidence, J. Econ. Psychol. 25 (2004) 539–555. [63] K. Ek, L. Persson, 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 (2014) 193–203. [64] W.O. Bearden, R.G. Netemeyer, K.L. Haws, Handbook of Marketing Scales: Multi-item Measures for Marketing and Consumer Behavior Research, 3rd ed., C alif SAGE, Thousand Oaks, 2011. [65] N. Hanley, C. Nevin, Appraising renewable energy developments in remote communities: the case of the North Assynt Estate, Scotland, Energy Policy 27 (1999) 527–547. [66] Sawtooth Software. The CBC/HB System for Hierarchical Bayes Estimation Version 5.0 Technical Paper. Sequium, WA, (2009). [67] DekaBank, Anlegerprofil-Check. Available from: https://www.deka.de/decontent/rechner/anlegerprofilcheck.jsp. (last accessed 01.22.15). [68] Sawtooth Software. SSI Web v8.3. Orem, UT, (2014). [69] K. Ek, Public and private attitudes towards green electricity: the case of Swedish wind power, Energy Policy 33 (2005) 1677–1689. [70] K. Ek, P. Söderholm, Norms and economic motivation in the Swedish green electricity market, Ecol. Econ. 68 (2008) 169–182. [71] Axel Springer SE. MDS online. Available from: http://online.mdsmediaplanung.de/vakm/fd66366068839cf8457dfadab34242e9/client#start (last accessed: 01.26.15.). [72] Statistisches Bundesamt, Einkommen, Einnahmen & Ausgaben. Available from: https://www.destatis.de/DE/ZahlenFakten/GesellschaftStaat/ EinkommenKonsumLebensbedingungen/EinkommenEinnahmenAusgaben/ Tabellen/Haushaltsnettoeinkommen.html. (lat accessed: 01. 26.15.). [73] M.K. Oschlies, A Behavioral Finance Perspective on Sustainable Energy Investment Decisions, Masterarbeit. St. Gallen, 2007. [74] B. Orme, J. Howell, Application of Covariates Within Sawtooth Software’s CBC/HB Program: Theory and Practical Example. Sequium, WA, (2009). [75] M.A. Petrova, NIMBYism revisited: public acceptance of wind energy in the United States, WIREs Clim. Change 4 (2013) 575–601. [76] M. Wolsink, Wind power implementation: the nature of public attitudes: equity and fairness instead of ‘backyard motives’, Renew. Sustain. Energy Rev. 11 (2007) 1188–1207. [77] Van. der Loo, A. Frans, Mediating Windpower in the Netherlands: The Task Force Windpower Implementation, Novem, Utrecht, 2001. [78] A. Masini, E. Menichetti, Investment decisions in the renewable energy sector: an analysis of non-financial drivers, Technol. Forecasting Soc. Change 80 (2013) 510–524. [79] M.D. Kaplowitz, T.D. Hadlock, R. Levine, A comparison of web and mail survey response rates, Public Opin. Q. 68 (2004) 94–101. [80] Statistisches Bundesamt. Anteil der Haushalte in Deutschland mit Internetzugang von 2006 bis. Available from: http://de.statista.com/statistik/ daten/studie/72660/umfrage/anteil-der-haushalte-in-deutschland-mitinternetzugang/. (last accessed: 02.03.15.) (2014). [81] AGOF., internet facts 2014–01. Available from: http://www.agof.de/internetfacts/. (last accessed: 04.14.10.). [82] D.A. Dillman, J.D. Smyth, L.M. Christian, Internet, mail, and Mixed-mode Surveys: The Tailored Design Method, 3rd ed., Wiley & Sons, Hoboken, N. J, 2009. [83] S. Chassot, N. Hampl, R. Wüstenhagen, When energy policy meets free-market capitalists: the moderating influence of worldviews on risk

32

J. Gamel et al. / Energy Research & Social Science 14 (2016) 22–32

perception and renewable energy investment decisions, Energy Res. Soc. Sci. 3 (2014) 143–151. [84] D. Muzyka, S. Birley, B. Leleux, Trade-offs in the investment decisons of European venture capitalists, J. Bus. Venturing 11 (1996) 273–287. [85] B.K. Sovacool, What are we doing here? Analyzing fifteen years of energy scholarship and proposing a social science research agenda, Energy Res. Soc. Sci. 1 (2014) 1–29.

[86] Deutsche Bundesbank, Panel on houshold finances, Available from: https:// www.bundesbank.de/Navigation/EN/Bundesbank/Research/Panel on household finances/panel on household finances.html.