Determinants of Southern Italian households’ intention to adopt energy efficiency measures in residential buildings

Journal of Cleaner Production 153 (2017) 83e91

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Determinants of Southern Italian households’ intention to adopt energy efficiency measures in residential buildings* M. Irene Prete a, Luigi Piper a, Cristian Rizzo a, Giovanni Pino a, Mauro Capestro a, Antonio Mileti a, Marco Pichierri c, Cesare Amatulli b, Alessandro M. Peluso a, Gianluigi Guido a, * a b c

Department of Management, Economics, Mathematics and Statistics, University of Salento, Ecotekne Campus, Via per Monteroni, 73100 Lecce, Italy Ionian Department of Law, Economics and Environment, University of Bari, Via Duomo, n. 259, 74123 Taranto, Italy Department of Management, Alma Mater Studiorum University of Bologna, Via Capo di Lucca 34, 40126 Bologna, Italy

a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 August 2016 Received in revised form 13 February 2017 Accepted 22 March 2017 Available online 23 March 2017

Implementing Energy Efficiency Measures (EEMs) in residential buildings is one of the most effective ways of decreasing household energy consumption. The present research adopts an extended version of the Theory of Planned Behavior e which includes attitude, subjective norms, perceived behavioral control, and environmental concern e to investigate the antecedents of households’ intention to adopt and willingness to pay for EEMs. The research was conducted in a Southern Italian region that has intensively promoted the adoption of renewable and energy-saving technologies. The results show that attitude is the main determinant of households’ intention to adopt and willingness to pay for EEMs. However, subjective norms, perceived behavioral control, and environmental concern have their own positive effects based on the income level, education, and age of household subgroups. The results have practical implications for companies and policy-makers interested in promoting EEM adoption and encouraging sustainable development. © 2017 Elsevier Ltd. All rights reserved.

^ as de Handling Editor: Cecilia Maria Villas Bo Almeida Keywords: Energy efficiency measures Energy savings Theory of planned behavior Environmental concern Willingness to pay

1. Introduction As the world population expands, energy conservation has risen to the forefront of public discourse, with governments focusing on how to instill the public with greater energy awareness. In Europe, commercial and residential buildings account for almost 40% of citizens’ energy consumption (Allouhi et al., 2015). The use of Energy Efficiency Measures (EEMs) e which entails providing the same level of service while using less energy e is one of the most

* This paper was part of the project “The identification of a Territorial Marketing Model for the Management of an Apulian Industrial District” 2013, financed by CUIS (Consorzio Universitario Interprovinciale Salentino), the Municipality of Galatina, Area di Sviluppo Industriale (ASI), and the Department of Management, Economics, Mathematics, and Statistics (University of Salento), under the direction of Prof. G. Guido. * Corresponding author. Department of Management, Economics, Mathematics and Statistics, University of Salento, Ecotekne Campus, Via per Monteroni, 73100 Lecce, Italy. E-mail address: [email protected] (G. Guido).

http://dx.doi.org/10.1016/j.jclepro.2017.03.157 0959-6526/© 2017 Elsevier Ltd. All rights reserved.

effective ways to decrease household energy consumption, therefore, ensuring more sustainable cities and regions. The present study focuses on residential buildings, as they represent 75% of the total building stock (Eurostat, 2010). In contrast to energy-saving measures, EEMs require the use of innovative materials and construction methods, as well as the installation of technologies such as solar photovoltaic, micro-wind, solar thermal, heat pumps, biomass boilers, and pellet stoves. They constitute a relatively new system able to foster energy conservation without restricting people’s lifestyles, and have seen a rapid diffusion in residential buildings during the past few decades, partly due to the various initiatives promoted by public and private organizations. Also in Italy several municipalities and regions have launched programs and incentive operations to foster the adoption of technologies aimed at saving energy and facilitating sustainable production. Against this background, this study focuses on Apulia (see Fig. 1), a Southern Italy region which can be considered one of the best examples in the adoption of innovations and EEMs in this area. For instance, the Act No. 13/2008 and the Regional Act No. 14/2009

84

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91

have allowed households to increase the volume of existing buildings in exchange for adopting energy efficient technologies. Furthermore, the recent PAN (Puglia Active Network) Project, promoted by the Apulia Regional Government and Enel (an Italian distributor of electricity and gas), has provided large investments aimed to foster the development of energetically sustainable solutions. The diffusion of EEMs in residential buildings represents one of the most sensitive policy issues in this area and Apulia has been the first Italian region able to activate, according to the 2014/ 2020 European financial programs, an investment of 2 billion Euro for energetically efficient purposes. Indeed, the local government has allocated 33.6% of its funds toward increasing environmental protection through local development (i.e., Regional Operative Programme, 2014e2020) and has provided several incentives e in the form of subsidies, tax exemption or tax reduction e to stimulate the adoption of EEMs. However, these incentives have not been as effective at fostering the adoption and diffusion of EEMs as policymakers had hoped (Berardi, 2013). Thus, there is a clear need to design more effective policies, which requires a better understanding of what prompts households to adopt EEMs. However, scholarship has not widely investigated the factors that determine households’ intention to use EEMs, nor has it evaluated how companies and policymakers could leverage these factors in their communication strategies. To help close this gap, this research examines the determinants of households’ intention to adopt and willingness to pay for EEMs by employing an extended version of the well-known Theory of Planned Behavior, which has transcended its social psychological origins (Ajzen, 1991) to become a reference framework in several environmentally relevant research settings. Using this theory, scholars can better understand and predict different human behaviors, such as farmers’ intention to grow biofuel crops (Guido, 2009; Peluso, 2015), planters’ intention to supply oil palm residues (Chin et al., 2016), as well as energy-saving behaviors (Chen,

2016). The Theory of Planned Behavior (Ajzen, 1991) postulates that three main determinants influence individuals’ intention to perform a given behavior: the attitude toward the behavior e a subjective positive or negative evaluation of the behavior based on the perceived advantages or disadvantages deriving from that behavior; the social norms e the perceived social pressure to perform or not perform the behavior based on the subjective perception that others might approve or disapprove of that behavior; and the perceived behavioral control e the subjective evaluation of how easy or difficult it will be to perform the behavior based on perceived facilitators of or obstacles to that behavior. This study extends the basic framework of the Theory of Planned Behavior (Ajzen, 1991) by including another determinant: namely, the environmental concern, defined as the degree to which people are aware of problems regarding the environment and support activities aimed to solve them or even engage personally in such activities (Dunlap and Jones, 2002; Fransson and Garling, 1999). Worldwide, an increasing number of people is becoming conscious of the environmental impact of their consumption behavior, and are interested in reducing it (Paul et al., 2016). Because environmental concern was found to significantly influence people’s intention to enact sustainable consumption behavior it seems reasonable to hypothesize that it may also determine households’ intention to adopt energy efficiency measures. Therefore, the present research tests a model in which households’ intention to adopt and willingness to pay for EEMs are expressed as a function of their attitude toward adoption, subjective norms, perceived behavioral control, and environmental concern. Unlike the adoption of unconstrained behaviors (e.g., energy savings), the use of EEMs requires an investment decision; thus, the inclusion of households’ willingness to pay might considerably improve the diagnostic and predictive validity of the tested model. In addition, it has been examined whether and how the relationships between these four antecedents and households’ intention to adopt and willingness to pay vary across different levels of income, education, and age (Fig. 2). 2. Determinants of households’ intention to adopt and willingness to pay for energy efficiency measures (EEMs)

Fig. 1. Location of the Apulia region in Southern Italy.

A number of studies have used the Theory of Planned Behavior to examine the intention to adopt EEMs (e.g., Abrahamse and Steg, 2009, 2011; Harland et al., 1999). Most of them demonstrate that households’ attitudes toward EEMs e and, particularly, the advantages they perceive e drive their decision to adopt these technologies in their residential buildings (Ek, 2005; Ma et al., 2013; Wang et al., 2014; Whitmarsh and O’Neill, 2010). However, prior investigations into specific EEMs (i.e., insulation activity, improved woodstoves, light bulbs, and unbleached papers) have shown that environmental concern (Vlek, 2000) together with subjective norms and perceived behavioral control (Khorasanizadeh et al., 2016; Nyrud et al., 2008) may also exert an influence on individuals’ intentions. Previous studies have revealed that environmental concern positively influences consumers’ attitudes and intention to purchase energy-saving branded products (Hartmann ~ ez, 2012) and, moreover, it can be deemed to and Apaolaza-Ib an be a relevant determinant of people’s intention to keep electricity €derholm, 2010; Pothitou et al., 2016). By down (e.g., Ek and So enlarging the basic framework of the Theory of Planned Behavior (Ajzen, 1991), Chen (2016) found that additional factors regarding the self, in general, and one’s moral obligation, in particular (i.e., one’s sense of responsibility to act in ways that are morally correct for the self, others or the environment), can help shape individuals’ intentions in ethically relevant situations. However, research on the impact of environmental concern on households’ intention to adopt

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91

Fig. 2. Conceptual model.

EEMs remains scarce (e.g., Wenshun et al., 2011) and, to date, very few studies (e.g., Abrahamse and Steg, 2009; Bamberg, 2003) have attempted to extend Ajzen’s (1991) model by considering this variable. Another stream of research has examined households’ willingness to pay for EEMs, particularly in relation to different types of technologies, such as energy retrofits (Achtnicht and Madlener, 2014), insulation of windows and facades (Banfi et al., 2008), biomass energy (Borchers et al., 2007), and photovoltaic systems (Jager, 2006). For instance, Hansla et al. (2008) showed that households’ willingness to pay for green electricity depends on their positive attitude toward it. Other scholars focused on specific situations able to hinder the adoption of EEMs have found that households’ decision to adopt EEMs can be impeded by their limited knowledge about prices and €kkinen and Belloni, 2011; Jakob, 2006). Other technologies (Ha studies highlighted economic factors as possible predictors of willingness to pay, including subsides and time horizon (Alberini et al., 2011), operational costs (Kennedy and Basu, 2013), installation costs, installation and maintenance costs (Mundaca et al., 2010), long payback time (Achtnicht, 2011), capital costs (Scarpa and Willis, 2010), and also lack of information (Stieb and Dunkelberg, 2013). Another stream of research has related households’ energy consumption to socio-demographic variables, such as income, age, and educational level. The literature is particularly divided on the specific influence of income level. Some studies have shown that households’ income affects their degree of investment in such technologies (Bartiaux et al., 2006), while others have found no significant correlation (Barr et al., 2005; Ürge-Vorsatz and Hauff, 2001). To address this disparity, Yang and Zhao (2015) have showed the moderating effect of family income on the relationship between attitude toward and the intention to adopt renewable energy equipment. The level of education influences the adoption of energyefficient technologies (Ürge-Vorsatz and Hauff, 2001), with higher-education households being more inclined to adopt EEMs (Poortinga et al., 2003). However, many authors have only demonstrated a positive relationship between education and different environmental attitudes (e.g., Amador et al., 2013; Roberts, 1996; Hunter and Toney, 2005); thus, this factor’s specific role is not yet clearly defined. By contrast, past research has extensively explored the relationship between a household’s age and its pro-environmental behavior, finding mixed results when comparing groups of different ages. Hunter and Toney (2005), for instance, found that elderly people are less inclined to save energy, whereas Roberts (1996) and Olli et al. (2001) found that elderly households have a greater propensity to act environmentally. With regard to EEMs, Mahapatra and Gustavsson (2008) found that elderly people are less likely to adopt energy-efficient technologies. More recently,

85

Wai and Bojei (2015) demonstrated the moderating effect of age on the intention to engage in environmentally friendly purchasing behaviors. Although research points to attitude as the most relevant determinant of households’ intention to adopt and willingness to pay for EEMs (e.g., Hansla et al., 2008; Wang et al., 2014), a number of studies suggest that the specific determinants vary according to regional differences. Consumers’ energy-related choices are often socially embedded and influenced by institutional constraints, so that any analysis of consumers’ beliefs needs to consider the specific territorial context (Gaps, 1998). This study tries to fill this gap and is aimed to investigate the antecedents of households’ intention to adopt and willingness to pay for EEMs on one area e the Apulia region in Southern Italy e in which EEM adoption has yet to reach adequate diffusion, despite a broad range of policy efforts offering financial incentives. 3. Methodology 3.1. Pilot study The pilot study aimed to assess residents’ beliefs via an openended questionnaire, which was administered to a sample of 20 households living in the Apulia region. In particular, respondents were asked to indicate: (i) the principal advantages and disadvantages deriving from the adoption of EEMs (i.e., behavioral beliefs); (ii) the categories of people (so-called “important others”) that they believed could encourage or discourage the adoption of EEMs (i.e., normative beliefs); and (iii) the situations that they believed could facilitate or hinder the adoption of EEMs (i.e., control beliefs). The answers to the open-ended questionnaires were contentanalyzed in order to identify residents’ behavioral, normative, and control beliefs. Table 1 shows: (i) the principal perceived advantages/disadvantages derived from implementing EEMs; (ii) the categories of people who hold the most influence over the decision to use EEMs; and (iii) the events or situations that may facilitate or impede the intention to use EEMs. Only items that were mentioned more than twice by respondents were considered for the main study. 3.2. Main study 3.2.1. Sample The main study was carried out on a random sample of 128 Apulian households. Respondents who had the responsibility to decide the adoption of EEMs were considered as the target population of the study. Respondents were mainly male (55.5%), middleaged (Mage ¼ 44.64; SD ¼ 16.63) and married people (60.2%), with an annual income less than 20,000 V (78.1%) (see Table 2 for sample descriptive statistics). 3.2.2. Procedure and measures In order to assess their behavioral beliefs, the main questionnaire asked respondents to indicate 1) the probability that the previously identified advantages/disadvantages might occur (on a seven-point scale: 1 ¼ “Not at all likely”, 7 ¼ “Extremely likely”), and 2) the perceived importance of each advantage/disadvantage (on a seven-point scale: 1 ¼ “Not at all important”, 7 ¼ “Very important”). Regarding normative beliefs, respondents were asked to evaluate 1) the likelihood that important others would exert an influence on their intention to adopt EEMs (on a seven-point scale: 1 ¼ “Not at all likely”, 7 ¼ “Extremely likely”), and 2) the perceived importance attributed to others’ opinions (on a seven-point scale: 1 ¼ “Not at all important”, 7 ¼ “Very important”). Respondents

86

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91

Table 1 Behavioral, normative, and control beliefs. Behavioral beliefs

N (%)

Normative beliefs

N (%)

Control beliefs

N (%)

1

A reduction of energy consumption

15 (75%)

Family members

18 (90%)

18 (90%)

2 3

13 (65%) 10 (50%)

Friends Partner

15 (75%) 9 (45%)

4

High installation costs Possibility to become energy-independent from traditional energy providers A better quality of life

Access to financial support by banks and credit institutions Subsidies or tax reduction/exemptions Difficulties in the disposal of installation

5 (25%)

Young people

7 (35%)

5 6

Maintenance costs A growth of real estate prices

2 (10%) 1 (5%)

Elderly people Environmental associations

2 (10%) 2 (10%)

The possibility to use more aesthetically pleasing installations The possibility to sell energy to power company The lack of information

16 (80%) 13 (65%) 7 (35%) 5 (25%) 1 (5%)

Notes: n ¼ 20.

Table 2 Sample profile of the respondents. Items

Classification

N

Frequency (%)

Gender

Male Female 20e29 30e40 40e50 50e60 More than 60 Single Married 0 1 2 3 4 Elementary/Secondary school High school Bachelor Master/PhD Professional Merchant Clerk Worker Student Retired House person Other Less than 5000 V 5000e10,000 V 10,000e20,000 V 20,000e30,000 V 30,000e50,000 V 50,000e100,000 V More than 100,000 V

71 57 24 36 24 19 25 49 77 43 24 32 16 3 21 58 36 13 29 8 28 9 15 17 9 13 38 26 36 21 5 1 1

55.47 44.53 18.75 28.12 18.75 14.84 19.53 38.28 60.15 33.59 18.75 25.00 12.50 2.34 16.40 45.31 28.12 10.16 22.66 6.25 21.87 7.03 11.72 13.28 7.03 10.15 29.69 20.31 28.12 16.41 3.91 0.78 0.78

Age

Marital Status Number of children

Education

Job

Annual Income

indicated control beliefs in a similar fashion, evaluating 1) the probability that the previously identified situations could affect their intention to adopt EEMs (on a seven-point scale: 1 ¼ “Not at all likely”, 7 ¼ “Extremely likely”), and 2) the importance they assigned to such situations (on a seven-point scale: 1 ¼ “Not at all important”, 7 ¼ “Very important”). To evaluate households’ environmental concern, respondents were asked to indicate 1) the probability that using EEMs would reduce environmental negative consequences (on a seven-point scale: 1 ¼ “Not at all likely”, 7 ¼ “Extremely likely”), and 2) the importance they attributed to this issue (on a seven-point scale: 1 ¼ “Not at all important”, 7 ¼ “Very important”). Next, respondents indicated both the likelihood that they would enact such behavior(s) (on a seven-point scale: 1 ¼ “Not at all likely”, 7 ¼ “Extremely likely”), and the strength of their intention (on a seven-point scale: 1 ¼ “Not at all”, 7 ¼ Extremely”), as well as how much they would be willing to pay for such EEMs. Finally, the questionnaire collected sociodemographic information.

All participants were assured their answers were anonymous, they could withdraw at any time, there were no right or wrong responses, and they were told to answer questions as honestly as possible. 4. Analyses and results In order to operationalize Ajzen’s (1991) determinants (i.e., attitude, subjective norms, and perceived behavioral control) and households’ intention to adopt EEMs, the score given to each determinant was multiplied by the respective probability, and then all these values were averaged (a procedure widely used in expectancy-values models). In line with other studies (Ürge-Vorsatz and Hauff, 2001), a positive relation between households’ level of education and the intention to adopt EEMs was found (r ¼ 0.25, p < 0.01), whereas the correlation with households’ income appeared to be marginally significant (r ¼ 0.17, p ¼ 0.06). Meanwhile, elderly people seem to be less likely to adopt such technologies (r ¼ 0.27, p < 0.01), confirming Mahapatra and Gustavsson’s (2008) findings. In the next step, a Confirmatory Factor Analysis (CFA) was performed to validate the measures used to assess households’ intention to adopt and willingness to pay (i.e., the traditional determinants of Ajzen’s model e attitude, subjective norms, and perceived behavioral control e along with our new one, environmental concern). Next, a path analysis was conducted setting our four aforementioned measures as the independent variables; households’ willingness to pay as the dependent variable, and households’ intention to adopt EEMs as the mediator. Finally, the path analysis also considered the potential moderating role of level of income, education and age, by means of different multi-group analyses. 4.1. Confirmatory Factor Analysis (CFA) The internal consistency of each scale was first evaluated by means of Cronbach’s a coefficient. The scales achieved reliability, as Cronbach’s a coefficient resulted higher than the recommended level of 0.7 (Nunnally, 1978). The Confirmatory Factors Analysis (CFA) was also conducted to assess the adequacy of the measurement model by examining the constructs reliability (Anderson and Gerbing, 1988). The results of the CFA, summarized in Table 3, returned acceptable fit statistics: c2/df ¼ 1.38; p ¼ 0.012; Goodness of Fit Index (GFI) ¼ 0.889; Comparative Fit Index (CFI) ¼ 0.841; Root Mean Square Residual (RMR) ¼ 0.066. 4.2. Path analysis This research extended the Theory of Planned Behavior (Ajzen, 1991) by including households’ environmental concern as another potentially crucial determinant of their intention to adopt

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91

87

Table 3 Description of observed variables and results of CFA. Type of item Type of measurement M Attitudes A reduction of energy consumption BB Possibility to become energy- independent from traditional energy providers BB A better quality of life BB High installation costs BB Subjective norms Family members NB Friends NB Partner NB Young people NB Perceived Behavioral Control Access to financial support by banks and credit institutions CB Subsidies or tax reduction/exemptions CB The possibility to sell energy to power company CB The use of more aesthetically pleasing installations CB Difficulties in the in the disposal of installation CB Intention to adopt EEMs Strength of the intention to adopt EEMs IV Likelihood to adopt EEMs IE

SD

Standardized estimates a 0.730

Multiplicative Multiplicative Multiplicative Multiplicative

31.594 27.773 29.273 25.992

13.358 12.523 12.570 14.3750

0.801** 0.608** 0.654** 0.510**

Multiplicative Multiplicative Multiplicative Multiplicative

29.820 22.070 31.023 27.727

14.262 11.451 14.315 13.622

0.680** 0.568** 0.855** 0.560**

Multiplicative Multiplicative Multiplicative Multiplicative Multiplicative

24.031 25.813 23.367 18.250 17.219

13.950 14.027 14.278 14.217 12.023

0.903** 0.887** 0.713** 0.478** 0.233*

Direct Direct

5.640 5.650

1.541 1.535

0.756** 0.872**

0.750

0.790

0.790

Notes: n ¼ 128; * ¼ p < 0.05; ** ¼ p < 0.01; BB ¼ Behavioral Belief; NB ¼ Normative Belief; CB ¼ Control Belief; IV ¼ Intention-as-Volition item; IE ¼ Intention-as-Expectation item.

and willingness to pay for EEMs. By using AMOS 18.0, the itemparceling procedure was followed in order to obtain more accurate estimates, achieve a better model fit (Bandalos, 2002; Nasser and Wisenbaker, 2003), and maintain the model’s parsimony in light of the sample size. Therefore, attitude, subjective norms, and perceived behavioral control were treated as latent variables, were measured by using combining multiplicative scores obtained on each belief item into a single, observed variable. Results from a maximum likelihood estimation returned acceptable fit statistics (c2/df ¼ 0.036; p ¼ 0.782; GFI ¼ 0.997; CFI ¼ 1; RMR ¼ 0.013). There appeared to be a significant direct effect of attitude toward EEMs (b ¼ 0.369; p < 0.001), perceived behavior control (b ¼ 0.194; p < 0.05), and environmental concern (b ¼ 0.173; p < 0.05) on the intention to adopt EEMs, as well as an indirect effect of attitude (b ¼ 0.191; p < 0.05) and perceived behavioral control (b ¼ 0.101; p < 0.05) on households’ willingness to pay via behavioral intention. In order to show how the effects of Ajzen’s determinants vary according to households’ income level, education, and age, three separate multi-group analyses have been conducted. In particular, three different median-splits were performed based on each of these three socio-demographic variables, which produced two subsamples (below or above the median value) for each variable. Therefore, the median-split procedure returned two groups of respondents with either a lower or a higher level of income; two groups of respondents with either a lower or a higher level of education, and two groups of respondents with either a lower or a higher level of age. For each multi-group analysis, an unconstrained model (no equality constraints were considered) against a constrained model (structural parameters were set to be equal across the two examined groups) was compared. 4.3. The moderating role of income For the model concerning households’ level of income, the Chisquare difference test between the unconstrained and constrained models returned a non-significant result (Dc2 ¼ 1.774, Ddf ¼ 5, p > 0.050). This finding suggests that the proposed model is metrically invariant across the examined samples and, therefore, differences across each path have been tested. First, the model has been estimated on each of the two subgroups separately, to verify whether it has acceptable fit for each

group (Low-income Model: c2/df ¼ 0.163, p ¼ 0.921; GFI ¼ 0.997; Adjusted Goodness of Fit Index (AGFI) ¼ 0.982; CFI ¼ 1; BentlerBonnet Normed Fit Index (NFI) ¼ 0.997; RMR ¼ 0.010; HighIncome Model: c2/df ¼ 1.016, p ¼ 0.384; GFI ¼ 0.985; AGFI ¼ 0.892; CFI ¼ 1; NFI ¼ 0.978; RMR ¼ 0.033). After establishing the validity of the separate models, a multi-group analysis was conducted in order to examine the possible moderating effect of income on households’ decision to adopt EEMs. The model fit proved to be adequate in this case (c2/df ¼ 0.59; p ¼ 0.739; GFI ¼ 0.991; CFI ¼ 1; RMR ¼ 0.010). Following Byrne (2010), income’s moderating effect was examined by evaluating the critical ratio differences between each path. As shown in Table 4, all tvalues resulted lower than the critical value of 1.96, suggesting that a household’s level of income does not moderate these relationships. However, the results revealed that attitude and environmental concern influence the intention to adopt EEMs for lowincome households, while the only significant linkage for wealthy households was perceived behavioral control. 4.4. The moderating role of education For the model concerning households’ level of education, the Chi-square difference test between the unconstrained and constrained models returned a non-significant result (Dc2 ¼ 1.774, Table 4 The moderating effects of income. Paths

Direct Effects ATT / INT SN / INT PBC / INT EC / INT INT / WTP Indirect Effects ATT / WTP SN / WTP PBC / WTP EC / WTP

Level of income Low

High

t

0.428*** 0.072 0.161 0.193* 0.696***

0.246 0.201 0.257* 0.153 0.329

0.901 0.757 0.130 0.264 0.350

0.298*** 0.050 0.112 0.134

0.081 0.066 0.085* 0.050

N(Pooled sample) ¼ 128; n(low level of income) ¼ 64; n(high level of income) ¼ 64; * ¼ p < 0.05; *** ¼ p < 0.001.

88

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91

Ddf ¼ 5, p > 0.050). This result suggests that the proposed model is metrically invariant across the examined samples and, as previously, differences across each paths have been tested. Like before, the model has been estimated on each of the two sub-groups separately, to verify whether it has acceptable fit for each group (Low-Education Model: c2/df ¼ 0.82, p ¼ 0.483; GFI ¼ 0.990; AGFI ¼ 0.928; CFI ¼ 1; NFI ¼ 0.987; RMR ¼ 0.021; High-Education Model: c2/df ¼ 2.280, p ¼ 0.077; GFI ¼ 0.958; AGFI ¼ 0.703; CFI ¼ 0.952; NFI ¼ 0.928; RMR ¼ 0.046). After establishing the validity of the separate models, a multi-group analysis was conducted in order to examine the possible moderating effect of education on households’ decision to adopt EEMs. As before, the model fit proved adequate (c2/df ¼ 1.556; p ¼ 0.156; GFI ¼ 0.977; CFI ¼ 0.987; RMR ¼ 0.021), so the critical ratio differences between each path have been examined. All t-values resulted lower than the critical value of 1.96, suggesting that a household’s level of education does not moderate these relationships (Table 5). Despite the non-significant differences across each path, the results revealed that attitude and perceived behavioral control influence the intention to adopt EEMs among loweducation households, while attitude and environmental concern arose as significant linkages for highly educated households.

(b ¼ 0.183; p < 0.05) exert a significant influence on younger households’ willingness to pay via intention to adopt EEMs, while perceived behavioral control (b ¼ 0.183; p < 0.001), attitude (b ¼ 0.163; p < 0.05) and subjective norms (b ¼ 0.099; p < 0.05) are the variables that indirectly affect older households’ willingness to pay. Next, the critical ratio differences between paths have been examined by evaluating the t-values. In contrast to income level and education, age moderates not only the effects that subjective norms and environmental concern exert on the intention to adopt EEMs, but also the relationship between intention and willingness to pay (Table 6). Specifically, the effect of subjective norms on the intention to adopt EEMs is significantly higher for elderly households; conversely, the effect of environmental concern on the intention to adopt is significantly greater for young people. 4.6. Interpretation of findings

For the model concerning households’ age, the Chi-square difference test between the unconstrained and constrained model returned a significant result (Dc2 ¼ 16.019, Ddf ¼ 5, p < 0.01); therefore, the models differ according to households’ age. Like before, the model has been estimated on each of the two sub-groups separately, to verify whether it has acceptable fit for each group (Low-Age Model: c2/df ¼ 0.530, p ¼ 0.663; GFI ¼ 0.992; AGFI ¼ 0.943; CFI ¼ 1; NFI ¼ 0.989; RMR ¼ 0.02; High-Age Model: c2/df ¼ 0.821, p ¼ 0.482; GFI ¼ 0.987; AGFI ¼ 0.910; CFI ¼ 1; NFI ¼ 0.986; RMR ¼ 0.025). After establishing the validity of the separate models, a multi-group analysis was conducted in order to examine the possible moderating effect of age on households’ decision to adopt EEMs. The model fit in this case was acceptable (c2/ df ¼ 0.675; p ¼ 0.67; GFI ¼ 0.99; CFI ¼ 1; RMR ¼ 0.02). According to the results, the intention to adopt EEMs for young households is mostly affected by attitude towards EEMs (b ¼ 0.294; p < 0.05), perceived behavioral control (b ¼ 0.257; p < 0.05) and environmental concern (b ¼ 0.289; p < 0.05), while for elderly households, the significant linkages included attitude towards EEMs (b ¼ 0.378; p < 0.01), perceived behavioral control (b ¼ 0.336; p < 0.01), and subjective norms (b ¼ 0.204; p < 0.05). Meanwhile, attitude (b ¼ 0.186; p < 0.05) and environmental concern

The goal of this research was to establish which determinants of Ajzen’s (1991) model most significantly influence the intention of Apulian residents to adopt energy-efficient technologies. In contrast to other studies that merely identified which determinants most influence the decision to adopt EEMs, this research extends the Theory of Planned Behavior by including households’ environmental concern and willingness to pay. Results e obtained by adopting a structural equation modeling approach e were first analyzed in relation to the entire sample of respondents. The application of the Ajzen’s Model revealed that the main antecedents of the intention to adopt energy efficiency measures are attitudes, perceived behavioral control and environmental concern. The indirect effects of attitudes and perceived behavioral control e via behavioral intention e on households’ willingness to pay were also assessed. This study aimed also at exploring the possible moderating role of socio-demographic variables (i.e., age, level of income and education). To this end, three different multi-group analysis have been performed. The Chi-square difference test between the unconstrained and constrained models returned a non-significant result for households’ level of income and education, and a significant one for households’ age. Such results provided evidence for a moderating effect of age on the tested model. Specifically, it has been demonstrated that age moderates the effect of subjective norms and environmental concern on the intention to adopt EEMs, as well as the relationship between intention and willingness to pay. With regards to the level of education and income, it has been showed the presence of any moderating effect for the tested model. However, despite the non-significant differences across each path,

Table 5 The moderating effects of education.

Table 6 The moderating effects of age.

4.5. The moderating role of age

Paths

Level of education Low

Direct Effects ATT / INT SN / INT PBC / INT EC / INT INT / WTP Indirect Effects ATT / WTP SN / WTP PBC / WTP EC / WTP

Paths High

t

0.366*** 0.124 0.24* 0.123 0.466***

0.282* 0.108 0.170 0.296* 0.554***

0.170*** 0.058 0.112* 0.057

0.156 0.060 0.094 0.164*

N(Pooled sample) ¼ 128; n(low level * ¼ p < 0.05; *** ¼ p < 0.001.

of education)

¼ 79; n(high

0.443 0.068 0.319 1.151 1.055

level of education)

¼ 49;

Direct Effects ATT / INT SN / INT PBC / INT EC / INT INT / WTP Indirect Effects ATT / WTP SN / WTP PBC / WTP EC / WTP N(Pooled sample) ¼ 128; n(low *** ¼ p < 0.001.

Level of age Low

High

t

0.294* 0.005 0.257* 0.289* 0.632***

0.378*** 0.336*** 0.204* 0.052 0.485***

0.640 2.434 0.256 2.165 2.363

0.186* 0.030 0.162 0.183*

0.183*** 0.163* 0.099* 0.025

level of age)

¼ 65; n(high

level of age)

¼ 63; * ¼ p < 0.05;

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91

Fig. 3. The tested model’s results.

attitudes and perceived behavioral control resulted the antecedents mostly affecting the intention to adopt EEMs among low-education households, while attitudes and environmental resulted significant for well-schooled households. Besides, attitudes and households’ environmental concern influence the intention to adopt EEMs for low-income households, whereas the only significant linkage for wealthy households was perceived behavioral control (Fig. 3). 5. Discussion The energy efficiency of residential buildings remains a highly debated theme among researchers and policymakers. Different authors have studied the diffusion of energy efficiency measures (EEMs) from a firm-level perspective, identifying the factors that favor or impede the adoption of such environmental friendlier technologies (e.g., Liu et al., 2013; Suk et al., 2013). However, research has not deeply explored the determinants behind households’ decision to adopt EEMs, despite their potential contribution to a more sustainable and cleaner society. To fill this gap, this paper proposed an extension of the Theory of Planned Behavior (Ajzen, 1991) in order to generate useful insights for marketing and sustainable development. Specifically, this study shed light on households’ beliefs regarding the adoption of EEMs by examining the Apulia region of Southern Italy, an area that has seen recent economic development thanks largely to tourism (Guido et al., 2012; Pino et al., 2014, 2015). Apulia represents a useful case study due to local policymakers’ considerable efforts at promoting the adoption of pro-environmental behavior and fostering sustainable development. By extending the Theory of Planned Behavior (Ajzen, 1991), this research revealed that attitude, perceived behavioral control, and environmental concern exert a significant effect on households’ intention to adopt and willingness to pay for EEMs. The positive relationship between environmental concern and intention to adopt EEMs suggests that people concerned with the environment are more likely to embrace such measures, which is consistent with previous findings (Kennedy and Basu, 2013; Stieb and Dulkenberg, 2013). Moreover, the possible moderating role of specific sociodemographic variables has been evaluated and, as expected, these causal associations significantly change according to households’ level of income, education, and age. With regard to income level, it has been found that high-income households are more influenced by specific situations or events (PBC), whereas low-income households focuses on the advantages or disadvantages (ATT) of using EEMs and on environmental concerns. The latter finding is coherent with previous research that found a negative relationship between income and environmental concern (Olli et al., 2001), although the link between these two variables is yet to be clarified

89

(Fairbrother, 2013; Franzen, 2003). Regarding education level, results revealed that less educated households are more influenced by their attitudes toward EEMs and perceived behavioral control, whereas highly educated households are more influenced by their environmental concern. This latter result is in line with prior studies (e.g., Hunter and Toney, 2005; Amador et al., 2013) according to which highly educated people are more concerned with the environment and usually undertake actions to defend it. Together with income and education, also age exerts a positive influence on the intention to adopt EEMs. Results showed that elderly households base their decision more than on environmental concern on attitude, perceived behavioral control, and subjective norms, consistent with the fact that elderly people tend to trust more other people’s opinion (Guido, 2014). In contrast, young people base their own decision on the environmental concerns and the advantages and disadvantages of using EEMs. This finding corroborates results of previous studies according to which younger people are highly worried about environmental issues (Mohd Suki, 2013), and such concerns determine their intention to adopt environmentally-friendly behaviors (e.g. Royne et al., 2011). A number of studies have analyzed the determinants of households’ intention to adopt EEMs by focusing on advantages and perceived barriers as main predictors of intention and/or willingness to pay (Achtnicht, 2011; Alberini et al., 2011; Kennedy and Basu, 2013; Mundaca et al., 2010; Scarpa and Willis, 2010; Stieb and Dunkelberg, 2013). Conversely, this study e embracing the Theory of Planned Behavior (Ajzen, 1991) as a framework reference e not only specifically identifies the advantages or disadvantages to adopt EEMs, but also the events or situations that may facilitate or hinder their adoption, as well as the categories of people able to influence such a choice. Besides, other research has applied the Ajzen’s (1991) model to study the adoption of specific energy-efficient technologies. For instance, Harland et al. (1999) found that households’ adoption of energy saving light bulbs depend on past behavior, attitude, and perceived behavioral control. Nyrud et al. (2008) analyzed consumers’ adoption of improved woodstoves applying an extended version of the Theory of Planned Behavior (Ajzen, 1991). The present study adds to literature and, albeit similar at first glance to some other researches applying the Theory of Planned Behavior to analyze energy-saving behaviors and, in particular, to the study proposed by Chen (2016), it substantially differs from this recent study in at least four aspects. A first crucial difference lies in the target behavior examined. Indeed, while Chen (2016) focused on a general set of energy-saving behaviors (e.g., drive less), this research analyzes the specific behavior regarding the adoption of EEMs. Second, differently from Chen (2016)’s work, the present research operationalizes the variables involved in the model on the basis of actual beliefs of households. Third, while Chen (2016)’s study only focused on main effects, this research goes a step further and explores the moderating role of socio-demographic variables such as income, education, and age. Finally, the two studies also differ in the geographical setting in that, while Chen (2016) examined energy-saving behaviors in Taiwan, this research investigates the adoption of EEMs in a Western country (namely, Italy). 5.1. Policy implications Policy efforts at the national and regional level have always involved economic and fiscal incentives, while local governments have mainly focused on legal and administrative aspects (Berardi, 2013). For instance, the National Building Codes have been continuously updated to favor the adoption of energy-saving technologies (Laws n. 192 in 2005, n. 311 in 2006, n. 28 in 2011,

90

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91

and Directive EPBD recast in 2010). Similarly, the Apulia region has instituted different measures to foster the adoption of EEMs. For instance, through the so-called “Conto Energia” (DM May 5th 2011), it has been given to residents the possibility to sell solar or wind power energy produced in excess to energy companies. Moreover, the detaxation of energy-saving retrofitting costs allowed households to compensate for 55% of costs resulting from taxes (Laws n. 296/06, and n. 214/11). The Apulia region provided further incentives with the Law n. 13 in 2008 (“Norme per l’abitare Sostenibile”) and Regional Law n. 14 in 2009 (“Piano Casa”). This study adds complexity to the notion regarding households’ decision to adopt EEMs. It points to the richness of behavioral drivers able to motivate individual conducts and the need to consider all of them depending on the different market segments. Moreover, the mix of policy actions which are necessary in each case is likely to be different and identifiable only by considering the specific context. Given this level of complexity, different policies, promotions strategies, and advertising campaigns should be carried out according to targets’ different age, education, and environmental concern. Indeed, the results of this study suggest that future policy initiatives should combine economic incentives with dissemination campaigns, with the latter aimed at increasing households’ knowledge about the benefits of adopting such measures. For instance, messages enforcing young people sensibility towards environmental problems should be used to underline possible advantages in adopting correct energy lifestyle; whereas, credible testimonials or opinion leaders should be involved to persuade elderly households. 5.2. Limitations and future research This study features limitations that may facilitate future research. Firstly, this study does not test the “real” behavior, namely the adoption of EEMs, but the mere intention to adopt them. Secondly, in contrast to other findings that examined specific energyefficient technologies (e.g., Banfi et al., 2008; Borchers et al., 2007; Ek, 2005), analyzed EEMs in general and, thus, did not differentiate the analysis for each type of technology. Future studies could repeat the analysis for specific energy-efficient technologies (e.g., compressed air systems, lighting systems, etc.). This might help companies to better calibrate their marketing and business strategies. 6. Conclusions Decreasing household energy consumption is one of the main objectives to be pursued in the next decades to significantly improve environmental sustainability of buildings. This study has investigated the antecedents of households’ intention to adopt and willingness to pay for EEMs, offering useful insights in the realm of reducing environmental impact (Piper et al., 2013). An extended version of the Theory of Planned Behavior (Ajzen, 1991) e a reference framework in several environmentally relevant research settings e was adopted and data were collected in Apulia e a Southern Italy region where the diffusion of EEMs in residential buildings represents a sensitive policy issue and where local governments have intensively promoted the adoption of energy-saving technologies. Findings revealed that attitude is the main determinant of households’ intention to adopt EEMs, and subjective norms, perceived behavioral control, and environmental concern have their own positive effects based on the income level, education, and age of household subgroups. This paper helps to advance knowledge about environmental sustainable development and, from a practical point of view, it provides useful insights to both marketing managers of energy suppliers and communication managers of local governments.

References Abrahamse, W., Steg, L., 2009. How do socio-demographic and psychological factors relate to households’ direct and indirect energy use and savings? J. Econ. Psychol. 30 (5), 711e720. Abrahamse, W., Steg, L., 2011. Factors related to household energy use and intention to reduce it: the role of psychological and socio-demographic variables. Hum. Ecol. Rev. 18 (1), 30e40. Achtnicht, M., 2011. Do environmental benefits matter? Evidence from a choice experiment among house owners in Germany. Ecol. Econ. 70 (11), 2191e2200. Achtnicht, M., Madlener, R., 2014. Factors influencing German house owners’ preferences on energy retrofits. Energy Policy 68, 254e263. Ajzen, I., 1991. The theory of planned behavior. Organ. Behav. Hum. Dec. 50, 179e211. Alberini, A., Banfi, S., Ramseier, C., 2011. Energy Efficiency Investments in the Home: Swiss HOUSEHOLDS and Expectations about Future Energy Prices. Working paper series 11e80. Center for Energy Policy and Economics (CEPE), ETH Zurich. Allouhi, A., El Fouih, Y., Kousksou, T., Jamil, A., Zeraouli, Y., Mourad, Y., 2015. Energy consumption and efficiency in buildings: current status and future trends. J. Clean. Prod. 109, 118e130. Anderson, J.C., Gerbing, D.W., 1988. Structural equation modeling in practice: a review and recommended two-step approach. Psychol. Bull. 103, 411e423. Amador, F.J., Gonz alez, R.M., Ramos-Real, J., 2013 November. Supplier choice and WTP for electricity attributes in an emerging market: the role of perceived past experience, environmental concern, and energy saving behavior. Energy Econ. 40, 953e966. Bandalos, D.L., 2002. The effects of item parceling on goodness-of-fit and parameter estimate bias in structural equation modeling. Struct. Equ. Model. 9 (1), 78e102. Bamberg, S., 2003. How does environmental concern influence specific environmentally related behaviors? A new answer to an old question. J. Environ. Psychol. 23, 21e32. Banfi, S., Farsi, M., Filippini, M., Jakob, M., 2008. Willingness to pay for energysaving measures in residential buildings. Energy Econ. 30 (2), 503e516. Barr, S., Gilg, A.W., Ford, N., 2005. The household energy gap: examining the divide between habitual and purchase-related conservation behaviours. Energy Policy 33 (11), 1425e1444. Bartiaux, F., Vekemans, G., Gram-Hanssen, K., Maes, D., Cantaert, M., Spies, B., Jensen, O.M., 2006. Socio-Technical Factors Influencing Residential Energy Consumption SEREC: Final report.. Berardi, U., 2013. Stakeholders’ influence on the adoption of energy-saving technologies in Italian homes. Energy Policy 60, 520e530. Borchers, A.M., Duke, J.M., Parsons, G.R., 2007. Does willingness to pay for green energy differ by source? Energy Policy 35 (6), 3327e3334. Byrne, B.M., 2010. Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming. Routledge. Chen, M.F., 2016. Extending the theory of planned behavior model to explain people’s energy savings and carbon reduction behavioral intentions to mitigate climate change in Taiwanemoral obligation matters. J. Clean. Prod. 112 (2), 1746e1753. Chin, H.C., Choong, W.W., Alwi, S.R.W., Mohammed, A.H., 2016. Using Theory of Planned Behaviour to explore oil palm smallholder planters’ intention to supply oil palm residues. J. Clean. Prod. 126, 428e439. Dunlap, R.E., Jones, R.E., 2002. Environmental concern: conceptual and measurement issues. In: Dunlap, R.E., Michelson, W.W. (Eds.), Handbook of Environmental Sociology. Greenwood Press, Westport (U.S.), pp. 482e524. Ek, K., 2005. Public and private attitudes towards “green” electricity: the case of Swedish wind power. Energy Policy 33 (13), 1677e1689. € derholm, P., 2010. The devil is in the details: household electricity saving Ek, K., So behavior and the role of information. Energy Policy 38 (3), 1578e1587. Eurostat, 2010. Housing Statistics. European Commission. Fairbrother, M., 2013. Rich people, poor people, and environmental concern: evidence across nations and time. Eur. Sociol. Rev. 29 (5), 382e910. Fransson, N., G€ arling, T., 1999. Environmental concern: conceptual definitions, measurement methods, and research findings. J. Environ. Psychol. 19 (4), 369e382. Franzen, A., 2003. Environmental attitudes in international comparison: an analysis of the ISSP surveys 1993 and 2000. Soc. Sci. Q. 84 (2), 297e308. Gaps, S.E., 1998. Barriers and conceptual chasms: theories of technology transfer and energy in buildings. Energy Policy 26 (15), 1105e1112. Guido, G., 2009. Behind Ethical Consumption. Peter Lang, Bern, Switzerland. Guido, G., 2014. Il Comportamento di Consumo degli Anziani [The Consumer Behavior of the Elderly] (Bologna, Italy: Il Mulino). Guido, G., Peluso, A.M., Prete, M.I., Amatulli, C., Pino, G., Pace, C., 2012. Customercentric strategies in place marketing: an analysis of places’ identities and perceived images. In: Kauffman, H.R., Khan Panni, M.F.A. (Eds.), Customercentric Marketing Strategies: Tools for Building Organizational Performance. IGI Global, Hershey, PA, pp. 435e452. € Hakkinen, T., Belloni, K., 2011. Barriers and drivers for sustainable building. Build. Res. Inf. 39 (3), 239e255. €rling, T., 2008. Psychological determinants of Hansla, A., Gamble, A., Juliusson, A., Ga attitude towards and willingness to pay for green electricity. Energy Policy 36 (2), 768e774. Harland, P., Staats, H., Wilke, H.A., 1999. Explaining proenvironmental intention and behavior by personal norms and the theory of planned behavior. J. Appl. Soc.

M.I. Prete et al. / Journal of Cleaner Production 153 (2017) 83e91 Psychol. 29 (12), 2505e2528. n ~ ez, V., 2012. Consumer attitude and purchase intention Hartmann, P., Apaolaza-Iba toward green energy brands: the roles of psychological benefits and environmental concern. J. Bus. Res. 65 (9), 1254e1263. Hunter, L.M., Toney, M.B., 2005. Religion and attitudes toward the environment: a comparison of Mormons and the general U.S. population. Soc. Sci. J. 42 (1), 25e38. Jager, W., 2006. Stimulating the diffusion of photovoltaic systems: a behavioural perspective. Energy Policy 34 (14), 1935e1943. Jakob, M., 2006. Marginal costs and co-benefits of energy efficiency investments: the case of the Swiss residential sector. Energy Policy 34 (2), 172e187. Kennedy, M., Basu, B., 2013. A study on the implementation of renewable heating technologies in the domestic sector in Ireland with implications on consumers’ decision-making. J. Clean. Prod. 44 (1), 133e142. Khorasanizadeh, H., Honarpour, A., Park, M.S.A., Parkkinen, J., Parthiban, R., 2016. Adoption factors of cleaner production technology in a developing country: energy efficient lighting in Malaysia. J. Clean. Prod. 131, 97e106. Liu, X., Niu, D., Bao, C., Suk, S., Sudo, K., 2013. Awareness and acceptability of companies on market-based instruments for energy saving: a survey study in Taicang, China. J. Clean. Prod. 39, 231e241. Ma, G., Andrews-Speed, P., Zhang, J., 2013. Chinese consumer attitudes towards energy saving: the case of household electrical appliances in Chongqing. Energy Policy 56, 591e602. Mahapatra, K., Gustavsson, L., 2008. An adopter-centric approach to analyze the diffusion patterns of innovative residential heating systems in Sweden. Energy Policy 36 (2), 577e590. Mohd Suki, N., 2013. Young consumer ecological behaviour: the effects of environmental knowledge, healthy food, and healthy way of life with the moderation of gender and age. Manag. Environ. Qual. 24 (6), 726e737. Mundaca, L., Neij, L., Worrell, E., McNeil, M., 2010. Evaluating energy efficiency policies with energy-economy models. Annu. Rev. Environ. Resour. 35, 305e344. Nasser, F., Wisenbaker, J., 2003. A Monte Carlo study investigating the impact of item parceling on measures of fit in confirmatory factor analysis. Educ. Psychol. Meas. 63 (5), 729e757. Nunnally, J.C., 1978. Psychometric Theory, second ed. McGraw-Hill, New York. Nyrud, A.Q., Roos, A., Sande, J.B., 2008. Residential bioenergy heating: a study of consumer perceptions of improved woodstoves. Energy Policy 36 (8), 3169e3176. Olli, E., Grendstad, G., Wollebaek, D., 2001. Correlates of environmental behaviors bringing back social context. Environ. Behav. 33 (2), 181e208. Paul, J., Modi, A., Patel, J., 2016. Predicting green product consumption using theory of planned behavior and reasoned action. J. Retail. Consum. Serv 29, 123e134. Peluso, A.M., 2015. Psychological drivers in the adoption of morally controversial innovations: the moderating role of ethical self-identity. Bus. Ethics A Eur. Rev.

91

24 (3), 252e263. Pino, G., Prete, M.I., Guido, G., 2014. The dominant rationality of local stakeholder networks: the case of a Southern Italian province. Local Econ. 29 (6e7), 687e707. Pino, G., Guido, G., Peluso, A.M., 2015. Perceived images and vocations of local territorial systems: implications for place branding strategies. J. Prod. Brand Manag. 24 (3), 287e302. Piper, L., Pelillo, V., De Franchis, E., Lay-Ekuakille, A., 2013. Cost minimization of waste recycling for lowering environmental impact. In: Proceedings 4th Imeko TC19 Symposium (Lecce, Italy). Poortinga, W., Steg, L., Vlek, C., Wiersma, G., 2003. Household preferences for energy-saving measures: a conjoint analysis. J. Econ. Psychol. 24 (1), 49e64. Pothitou, M., Hanna, R.F., Chalvatzis, K.J., 2016. Environmental knowledge, proenvironmental behaviour and energy savings in households: an empirical study. Appl. Energy 184, 1217e1229. Roberts, J.A., 1996. Green consumers in the 1990s: profile and implications for advertising. J. Bus. Res. 36 (3), 217e231. Royne, M.B., Levy, M., Martinez, J., 2011. The public health implications of consumers’ environmental concern and their willingness to pay for an eco-friendly product. J. Consum. Aff. 45 (2), 329e343. Scarpa, R., Willis, K., 2010. Willingness-to-pay for renewable energy: primary and discretionary choice of British households’ for micro-generation technologies. Energy Econ. 32 (1), 129e136. Stieß, I., Dunkelberg, E., 2013. Objectives, barriers and occasions for energy efficient refurbishment by private households. J. Clean. Prod. 48, 250e259. Suk, S., Liu, X., Sudo, K., 2013. A survey study of energy saving activities of industrial companies in the Republic of Korea. J. Clean. Prod. 41, 301e311. Ürge-Vorsatz, D., Hauff, J., 2001. Drivers of market transformation: analysis of the Hungarian lighting success story. Energy Policy 29 (10), 801e810. Vlek, C., 2000. Essential psychology for environmental policy making. Int. J. Psychol. 35 (2), 153e167. Wai, Y.S., Bojei, J., 2015. Cohesiveness of social and individual forces in strengthening attitude-pro-environment behavior relationship e a concept paper. Int. J. Bus. Econ. Laws 7 (2), 1e6. Wang, P., Liu, Q., Qi, Y., 2014. Factors influencing sustainable consumption behaviors: a survey of the rural residents in China. J. Clean. Prod. 63, 152e165. Wenshun, W., Xiaohua, L., Hualong, L., 2011. Empirical research of the environmental responsibility affected on the urban residential housing energy saving investment behavior. Energy Procedia 5, 991e997. Whitmarsh, L., O’Neill, S., 2010. Green identity, green living? The role of proenvironmental self-identity in determining consistency across diverse proenvironmental behaviours. J. Environ. Psychol. 30 (3), 305e314. Yang, S., Zhao, D., 2015. Do subsidies work better in low-income than in highincome families? Survey on domestic energy-efficient and renewable energy equipment purchase in China. J. Clean. Prod. 108, 841e851.