Energy Policy 115 (2018) 171–180

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

Exploring the effects of non-cognitive and emotional factors on household T

electricity saving behavior
⁎
Shanyong Wangb, Shoufu Lina,b, , Jun Lib
a
School of Economics, Fujian Normal University, Fuzhou, Fujian Province 350007, PR China
b
School of Management, University of Science and Technology of China, Hefei, Anhui Province, 230026, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: Understanding and promoting household electricity saving behavior is vital to reduce electricity consumption
Theory of planned behavior and carbon emissions. The main purpose of this research is to analyze the effects of non-cognitive (personal
Electricity saving behavior moral norm and habit) and emotional factors (positive anticipated emotion) on household electricity saving
Personal moral norm behavior using a comprehensive model integrating the theory of planned behavior and the theory of inter-
Habit
personal behavior. The model is empirically tested using questionnaire survey data collected from 320 house-
Positive anticipated emotion
holds. The results indicate that personal moral norm, habit and positive anticipated emotion are important
determinants of residents’ intention to save electricity. Furthermore, it is find that habit is also positively as-
sociated with electricity saving behavior. However, positive anticipated emotion is negatively associated with
electricity saving behavior, which means that residents who present more positive anticipated emotion about
electricity conservation intend to save electricity in their daily lives but actually end up consuming more
electricity. In addition, the results also suggest that residents with positive attitude towards saving electricity and
greater senses of control have stronger intention to save electricity, while the role of subjective norm is not
significant. Based on these results, policy implications for research and practice and suggestions for further
research are discussed.

1. Introduction electricity consumption pattern and promoting electricity saving be-

With the rapid development of economy and the improvement of
living standards in the past decades, the electricity consumption in
Chinese household sector is increasing in an incredible speed (Du et al.,
2017; Ding et al., 2017). According to a report issued by National En-
ergy Administration of China,1 the average household electricity con-
sumption in 2000 is only 232.8 kWh, but by 2015 it increased to
698.3 kWh in China. The annual growth rate is nearly about 7.60%.
There is no doubt that the household electricity consumption in China
will continue increase in the next few years because the electricity
consumption and economic development exhibit a significant positive
causal nexus (Yuan et al., 2007). The increased household electricity
consumption intensifies the energy crisis in China and brings huge
number of carbon emissions and other toxic gases, which may damage
the environment and individuals’ health (Pothitou et al., 2016; Zhang
et al., 2013). Meanwhile, it is worth noting that household electricity
consumption has a large saving potential (Ouyang and Hokao, 2009).
Murata et al. (2008) indicated that 28% reduction in household elec-
tricity consumption in China could be achieved through changing
havior by the year of 2020. Thus, considering the negative con-
sequences and the large saving potential, several measures and relevant
researches should be taken to reduce household electricity consumption
and promote electricity saving behavior.
Currently, the measures and research efforts focus on household
electricity consumption behavior can be divided into three major ca-
tegories, namely economic oriented perspective, technological oriented
perspective and psychological behavior oriented perspective (Zhou and
Yang, 2016; Arawomo, 2017; da Silva and Cerqueira, 2017). Economic
oriented perspective mainly focuses on price-based measures and fi-
nancial incentive-based measures (Zhou and Yang, 2016). Theoreti-
cally, it seems that price-based measures such as multi-part tariffs and
peak-load pricing can reduce household electricity consumption dra-
matically and promote electricity saving behavior. However, in prac-
tice, most studies have found that the effects of these measures is re-
latively limited (Reiss and White, 2005; Zhou and Teng, 2013). This is
because that compared with other household consumption ex-
penditures, the expenditure on electricity consumption is rather small
and electricity consumption usually exhibits low price elasticity (Zhou

⁎
Correspondence to: School of Economics, Fujian Normal University, No. 32, Shangsan Road, Fuzhou, Fujian Province 350007, PR China.
E-mail address: linshf2003@126.com (S. Lin).
1
See detail at http://data.stats.gov.cn/easyquery.htm?cn=C01.

https://doi.org/10.1016/j.enpol.2018.01.012
Received 22 August 2017; Received in revised form 26 December 2017; Accepted 6 January 2018
0301-4215/ © 2018 Elsevier Ltd. All rights reserved.
S. Wang et al.
and Teng, 2013; Yu and Guo, 2016). Most households are insensitive to
electricity price change and unwilling to save electricity to sacrifice
their comfortable living style and welfare in their daily lives (Cheung
et al., 2017). Financial incentive-based measures aim to provide sub-
sidies to households to compensate for their sacrifices (Handgraaf et al.,
2013). Previous studies have suggested that financial incentive-based
measures can only be effective in a short period of time unless the
measures are long lasting (Handgraaf et al., 2013; Frederiks et al.,
2015). Hence, financial incentive-based measures may not encourage
electricity saving behavior effectively in the long run. Technological
oriented perspective aims to improve electricity consumption efficiency
and promote energy-efficient technology and appliance to reduce and
save electricity consumption (Zhou and Yang, 2016). However, relying
on energy-efficient technology and appliance are not sufficient to re-
duce electricity consumption due to the ‘rebound effect2’ (Sorrell and
Dimitropoulos, 2008; Georges et al., 2017). For example, Nilsson et al.
(2014a) noted that the energy-efficient technology has improved
greatly and technical energy efficiency of appliance has improved
substantially over the past decades, while in the same period the
household electricity consumption has also increased.
Considering the limitations of economic oriented perspective and
technological oriented perspective, scholars have been more and more
recognized that psychological behavior oriented perspective is of great
significance in achieving electricity saving goals (Martinsson et al.,
2011; Hori et al., 2013; De Leeuw et al., 2015; Bertoldo and Castro,
2016; Shi et al., 2017). Psychological behavior oriented perspective
aims to achieve household electricity conservation by underscoring
some psychological factors, such as attitude, social norms and en-
vironmental awareness to promote electricity saving behavior
(Martinsson et al., 2011; Fornara et al., 2016; Zhou and Yang, 2016;
Ding et al., 2017). In previous studies, scholars have paid attention to
psychological factors and many psychological behavior models and
theories have been developed to understand household electricity
consumption behavior and explore the influencing factors (Barr et al.,
2005; De and Steg, 2009; Hori et al., 2013; Nilsson et al., 2014b;
Fornara et al., 2016; Yu and Guo, 2016; Ding et al., 2017). Among these
models and theories, theory of planned behavior (TPB) is widely used.
Rivis et al. (2009) indicated that TPB is probably the most influential
theory in explaining environmentally relevant behavior. In this re-
search, we also try to adopt TPB to understand household electricity
behavior from the psychological behavior oriented perspective.
In fact, in the pro-environmental behavior domain, theory of rea-
soned action (TRA) and norm activation model (NAM) have also often
been used to study pro-environmental behavior, such as recycling be-
havior, green buying behavior and energy conservation behavior
(Davies et al., 2002; Wahid et al., 2011). However, they have often been
questioned due to their own limitations (Shi et al., 2017). TRA has been
questioned since this theory only focuses on volitional control factors
and ignores the non-volitional factors, such as time, opportunities and
resources (Lam and Hsu, 2004). NAM has been criticized since that it
only focuses on internal factors and ignores the external factors, such as
social environment, time and resources (Shi et al., 2017). Compared
with TRA and NAM, TPB not only considers the non-volitional factors
(e.g., perceived behavioral control) but also external factors (e.g.,
subjective norm). Hence, it is appropriate to select TPB as the basic
theoretical framework in this research to understand household elec-
tricity saving behavior. In addition, several scholars have noted that
additional psychological factors and variables can also be added to TPB
to improve the model explanatory power (Kaiser and Scheuthle, 2003;
Shi et al., 2017; Gao et al., 2017). In this research, two additional
factors, namely non-cognitive factors (e.g., personal moral norm and
2
Rebound effect refers to an increase in energy use efficiency by 1% will cause a re-
duction in energy consumption that is far below 1% or, sometimes, it can even cause an
increase in energy consumption.
172
Energy Policy 115 (2018) 171–180
habit) and emotional factors (e.g., positive anticipated emotion) are
incorporated into TPB model to better understand household electricity
saving behavior (Hassin et al., 2009; Webb et al., 2013; Fornara et al.,
2016). The main goals of this research are to explore whether these
factors and variables significantly affect household electricity saving
behavior and what measures can be taken to promote household elec-
tricity saving behavior.
There are several theoretical and applied contributions of this re-
search. Theoretically, to the best of our knowledge, this is the first time
to incorporate non-cognitive factors and emotional factors together into
TPB to understand household electricity saving behavior, which en-
riches the existing understanding of household electricity saving be-
havior. Meanwhile, this research provides a paradigm for under-
standing household electricity saving behavior in pro-environmental
research filed. This paradigm can help refine and strengthen future
research on electricity saving behavior. In addition, this research
highlights the importance of non-cognitive and emotional factors in
promoting household electricity saving behavior and provides new di-
rections for future research in pro-environmental behavior field. In
practice, based on the research findings, measures and intervention
strategies can be taken to promote household electricity saving beha-
vior.
The remainder of this research is organized as follows. Section 2
focuses on literature review. Section 3 proposes the conceptual frame-
work and hypotheses. Section 4 focuses on data and the research
method. Data analysis and results are presented in Section 5. In Section
6, we discuss the results and implications. In Section 7, we conclude the
research and address the research limitations.
2. Literature review
TPB is first proposed by Ajzen in 1991 and now it is the most
popular theory to predict and explain individual's behavior in a wide
range of fields, especially in pro-environmental domain (Ajzen, 1991).
Klöckner (2013) noted that approximately 40% of all papers published
in environmental psychology domain have employed TPB as their basic
theoretical framework. In fact, TPB has been successfully applied to
explore household environmentally friendly behavior, such as energy
conservation behavior, green purchasing behavior and other sustain-
able consumption behavior (Kaiser and Scheuthle, 2003; Chen and
Tung, 2014; Wang et al., 2014; De Leeuw et al., 2015; Yadav and
Pathak, 2016).
Though TPB has received strong support in explaining en-
vironmentally friendly behavior and own several advantages compared
with TRA and NAM, it also has several shortcomings and these short-
comings can be solved by adding additional factors and variables
(Bamberg, 2003; Gao et al., 2017). One of the shortcomings of TPB is
that as for the social norm, TPB overestimates the effect of it and under
represents the contribution of internal moral norm, particularly per-
sonal moral norm (Webb et al., 2013; Cheung et al., 2017). Abrahamse
and Steg (2009) and Bertoldo and Castro (2016) argued that most pro-
environmental behaviors are motivated by personal moral norm rather
than social norm. Meanwhile, unlike other pro-environmental behavior
(e.g., green buying behavior or green travel behavior), household
electricity saving behavior has less visibility to other people (e.g.,
friends, relatives or neighbors). Intuitively, it can be predicted that the
effect of social norm on household electricity saving behavior is likely
to be of less importance.
Another shortcoming is that the variables in TPB are cognitive and
rational predictors and TPB largely relies on the assumption that in-
dividuals make rational choices (Abrahamse and Steg, 2009; Demarque
et al., 2015). As we know, to some degree, environmentally friendly
behavior cannot be just considered as a result of rational choice (Kals
et al., 1999). Many behaviors are guided more by automatic, repeated
and positive affective processes, which means that some non-cognitive
and emotional factors, such as personal moral norm, habit and emotion
S. Wang et al.
should be considered (Steg and Vlek, 2009; Quested et al., 2013).
Klöckner and Blöbaum (2010) and Klöckner (2013) noted that habit or
past behavior plays an important role in forming behaviors that are
repeated over time, such as environmentally friendly behavior. Kals
et al. (1999) noted that emotion about environmental issues is parti-
cularly important in explaining environmentally friendly behavior. In
addition, the theory of interpersonal behavior also argues that emotion
is likely to play an important role in conservation behavior (Russell
et al., 2017). Failing to explore and explain the role of emotion hampers
the understanding of individuals’ pro-environmental behavior (Morris
et al., 2002).
In this research, non-cognitive factors are defined as individuals’
patterns of thought and feelings towards household electricity saving
behavior and the thought and feelings are depended on their own
characteristics such as conscientiousness, skill and experience.
Considering that personal moral norm is originated from one's own
judgment and conscientiousness, habit is originated from one's past
experience, and they are all belong to the variables of personal char-
acteristics, this research regards personal moral norm and habit as non-
cognitive factors. Taking these reasons and viewpoints into account,
this research tries to add non-cognitive factors (e.g., habit and personal
moral norm) and emotional factors into the basic theoretical framework
to further and better understand household electricity saving behavior.
3. Research hypotheses and conceptual framework
3.1. Basic variables in TPB
TPB predicts that behavior is directly determined by behavioral
intention and behavioral intention is predicted by attitude towards the
behavior, subjective norm and perceived behavioral control (Yadav and
Pathak, 2016). Attitude refers to positive or negative evaluations about
certain specific behavior (Gao et al., 2017). When individuals hold a
positive attitude towards the behavior, their intentions to engage in a
specific behavior are increased and vice versa. Subjective norm refers to
perceptions of social pressures from significant others to perform or not
perform the behavior (Cheung et al., 2017). Individuals tend to comply
with the expectations or viewpoints of these important people (e.g.
relatives or friends). If individuals think that the important other people
expect them to engage in a specific behavior, they are more likely to
perform the behavior. Perceived behavioral control refers to the per-
ceived ease or difficulty of performing the behavior (Fornara et al.,
2016). If individuals perceive that they are easy to perform the beha-
vior, they are more intended to perform the behavior (Shi et al., 2017).
Consistent with TPB, in the context of electricity conservation, it can
be inferred that residents’ intention to save electricity will be increased
when residents hold a positive attitude towards saving electricity, if
residents think that important people (e.g., friends or neighbors) expect
them to save electricity, and if residents perceive that they have an
adequate level of control, such as ability, time, skills and opportunities
to be able to perform the electricity saving behavior. In this research,
attitude, subjective norm and perceived behavioral control is con-
sidered as the positive predictors of residents’ intention to save elec-
tricity. Hence, we hypothesize that:
H1a:. Attitude towards saving electricity is positively related to
resident's intention to save electricity.
H1b:. Subjective norm is positively related to resident's intention to
save electricity.
H1c:. Perceived behavioral control is positively related to resident's
intention to save electricity.
Based on the same logic, in line with TPB, we also predict that re-
sident's intention to save electricity is a significant and positive
173
Energy Policy 115 (2018) 171–180
predictor of household electricity saving behavior. The greater the in-
tention to save electricity, the higher the electricity saving behavior will
be performed. Thus, it is hypothesized that:
H2:. Resident's intention to save electricity is positively related to
household electricity saving behavior.
3.2. Non-cognitive and emotional factors
This research responds to the shortcomings of TPB and advances the
research on household electricity saving behavior by examining both
non-cognitive and emotional factors. Two non-cognitive factors (per-
sonal moral norm and habit) and one emotional factor (positive an-
ticipated emotion) are added into TPB.
3.2.1. Personal moral norm
Personal moral norm refers to individuals perform a particular be-
havior based on their own moral responsibilities or obligations (Stern,
2000; Kaiser and Scheuthle, 2003; Bertoldo and Castro, 2016). In fact,
personal moral norm is the core factor in norm activation model (Gao
et al., 2017; Shi et al., 2017). According to this model, individual's pro-
environmental behavior is determined not only by social norms but also
personal moral norm. High personal moral norm can motivate in-
dividuals to perform pro-environment behavior, while low personal
moral norm can impede the behavior. A number of previous studies
have found that personal moral norm positively affects individual's pro-
environmental behavior. For example, Bamberg et al. (2007) noted that
individuals with high personal moral norm are more likely to choose
public transportation for traffic than those with low personal moral
norm. Fornara et al. (2016) found that personal moral norm positively
affects household energy consumption behavior.
Similarity, in this research, it is reasonable to predict that residents
with high personal moral norm will have more moral obligations and
responsibilities to save electricity. Wasting electricity or not saving
electricity will disobey their moral norms and obligations which may
lead them to feel guilty and uncomfortable. Taken together, it can be
speculated that residents with high personal moral norm are more
likely to form intention to save electricity. Thus, it is hypothesized that:
H3:. Personal moral norm is positively related to resident's intention to
save electricity.
3.2.2. Habit
Psychological research increasing suggests that the behavior which
is repeated or habitual is more easily affected by past behavior or habit
(Hassin et al., 2009). When behavior is repeatedly performed, it re-
quires less deliberate consideration so that it becomes habit (Russell
et al., 2017). According to Verplanken and Holland (2002) and Russell
et al. (2017), habit refers to relatively stable behavioral patterns, which
can be triggered by past behavior and executed without rational eva-
luation. Habit is important and once individuals form the habit, they
will be more likely to perform the behavior.
In relation to electricity conservation, we argue that electricity
saving behavior is likely to have a strong habitual element. Electricity
saving behavior is usually performed automatically and unconsciously.
Given that the electricity saving behavior is a repeated and habitual
behavior, we therefore consider that residents’ electricity saving habit
will play an important role in saving electricity. In daily lives, once the
residents own the electricity saving habit, they are more intended to
save electricity and perform the electricity saving behavior naturally
and continually. Hence, we expect that the greater electricity saving
habit residents own, the more likely they are willing to save electricity
and report high levels of current electricity saving behavior. The hy-
potheses are:
S. Wang et al.
H4a. : Habit is positively related to resident's intention to save
electricity.
H4b. : Habit is positively related to household electricity saving
behavior.
3.2.3. Emotion
Emotion refers to the feelings (e.g., positive or negative) or reac-
tions to an event or issue (Russell et al., 2017). An individual is likely to
engage in an event or issue when he presents a positive anticipated
emotion on this event or issue, otherwise he is unwilling to engage in.
In this research, positive anticipated emotion refers to the positive
psychological states when performing a certain behavior (e.g., house-
hold electricity saving behavior). Positive anticipated emotion consists
of expected feelings of pride, exciting and confidence (Han and Hyun,
2018). Han and Hyun (2018) stressed that positive anticipated emotion
firstly signals the importance and significance of an event or issue and
then provides stimulus for performing the behavior. According to the
theory of interpersonal behavior, it is known that positive anticipated
emotion plays an important role in performing environmentally
friendly behavior, particularly conservation behavior (Russell et al.,
2017). Several past researches have shown the relationship between
positive anticipated emotion and environmentally friendly behavior.
Bamberg et al. (2007) observed that individuals are willing to take
public transportation for traffic when they present positive anticipated
emotion about this behavior. Rivis et al. (2009) conducted a meta-
analysis and found that positive anticipated emotion is positively re-
lated to individuals’ intention to perform environmentally friendly be-
havior. Webb et al. (2013) showed that positive anticipated emotion
has a positive effect on intention to engage in energy saving behavior.
These studies have identified positive anticipated emotion as a de-
terminant of intention to perform environmentally friendly behavior.
Based on the same logic, in the context of electricity conservation, it can
be predicted that when residents think that saving electricity in their
daily lives is beneficial to the environment and they present a positive
anticipated emotion about saving electricity, they are more intended to
save electricity. In addition, according to affective events theory, we
know that positive anticipated emotion is a direct predictor of actual
behavior because it provides a motivational impetus for action (Weiss
and Beal, 2005; Han and Hyun, 2018). Consistent with affective events
theory, this research therefore postulates that positive anticipated
emotion has a direct effect on household electricity saving behavior and
we expect that the more positive anticipated emotion residents present
the more likely they perform the actual electricity saving behavior.
Given these points of view, we hypothesize that:
H5a:. Positive anticipated emotion is positively related to resident's
intention to save electricity.
H5b:. Positive anticipated emotion is positively related to household
electricity saving behavior.
Based on above analysis, the research framework is depicted in
Fig. 1.
4. Data and research method
4.1. Sample and data collection
The research hypotheses are tested using questionnaire survey data
collected from residents in Hefei, China from mid-July to mid-
December of 2016. Hefei is the provincial capital of Anhui province and
is also the political, economic and cultural center of the province. With
the reform and opening up policy, the speed of economy development is
amazing. According to “The Economist”, Hefei is the fastest growing
cities in China in 2015. With the development of economy, the human
population has increased rapidly. At the end of 2015, the size of the
174
Energy Policy 115 (2018) 171–180
permanent resident population in Hefei is more than 7 million. Along
with the population growth and the improvement of residents’ living
standards, the household electricity consumption has increased sharply.
For example, the total household electricity consumption in 2015 is
2.95 billion kWh, which has increased by 10% over 2014.3 In addition,
the electricity resource is insufficient in Hefei and it needs other cities’
electricity support. Hence, it is urgent and necessary for Hefei to mo-
tivate household to save electricity in their daily lives.
In general, it is difficult for researchers to collect data directly from
households in China because many residents are reluctant to cooperate
with the research team. Thus, to conduct the survey smoothly and ea-
sily, the research team sough assistance from the local neighborhood
committee, which is the community management organization of re-
sidential quarter in China (Wang et al., 2011). To obtain a re-
presentative sample, the research team randomly selected 30 re-
sidential quarters in Hefei to conduct the survey. The research team
sent 30 copies of printed questionnaires to the contact person of each
neighborhood committee, and then the contact person random selected
30 households in residential quarter and distributed these ques-
tionnaires to the householders. The questionnaire is consisted of 5 se-
parate modules: demographic information (e.g., family size, educa-
tional level and age), three basic constructs of TPB (attitude towards
saving electricity, subjective norm and perceived behavioral control),
three additional constructs (personal moral norm, habit and positive
anticipated emotion), intention to save electricity, and electricity
saving behavior.
Moreover, some scholars have noted that relies on cross-section
design may lead to common method bias which can inflate the re-
lationships between constructs (Podsakoff et al., 2003). By measuring
the independent and dependent constructs at different points in time
can overcome this limitation. Following this logic, a four-stage ques-
tionnaire survey is conducted. In the first stage (conducted in August
2016), the contact person asked the participants to answer modules 1
and 2. In the second stage (conducted in September 2016), participants
are required to answer model 3. In the third (conducted in October
2016) and fourth stage (conducted in November 2016), participants are
required to answer modules 4 and 5 separately. Participants are in-
cluded in the sample if they respond to all four-stage surveys.
Furthermore, to avoid data bias, participants are told that their re-
sponses are strictly anonymity and confidentiality and never be leaked
out, and there is no right or wrong answers. To improve the response
rate, participants are told that they can obtain 40 RMB as remuneration
upon finishing the questionnaire survey. In total, 492 finished ques-
tionnaires are received. The research team reviewed the questionnaires
and discarded the incomplete and missing data questionnaires. Finally,
320 useable questionnaires are obtained, with a response rate of
35.56% (320/900). The demographic profiles of participants are pre-
sented in Table 1. As shown in Table 1, among the participants, nearly
half (53.44%) of the participants are female and 60% are aged between
25 and 45 years old. The participants are relatively well-educated, and
70% earns an associate degree or a bachelor's degree. About 48.44%
earns a monthly income ranging from 4000RMB to 8000RMB, and
35.31% earns a monthly income ranging from 8001RMB to 12000RMB.
More than half (60.31%) of the participants have medium family size
within 3–4 members.
In addition, to test the potential non-response bias, T-tests are
conducted to compare the early responses (those who returned the
completed questionnaires within 10 days) and late responses (those
who returned the completed questionnaires during the last 10 days)
(Armstrong and Overton, 1977). The results show no significant dif-
ferences between these responses at 0.05 level. Thus, it can be con-
cluded that non-response bias is not a major problem in this research.
3
See detail at http://www.ahtjj.gov.cn/tjj/web/index.jsp.
S. Wang et al. Energy Policy 115 (2018) 171–180

Fig. 1. Research framework of household electricity

saving behavior.
Habit
Habit
Attitude

H1a H4a H4b

Subjective norm
Subjective norm
H1b
Intention to save Electricity
Intention to save H2 Electricity
electricity saving behavior
electricity saving behavior
H1c
Perceived behavioral
Perceived behavioral
control
control H5b
H5a
H3

Personal moral norm Positive anticipate

Personal moral norm Positive anticipate
emotion
emotion

Table 1 adapted from Nilsson et al. (2014a) and Cheung et al. (2017).
Demographic profiles of participants. Considering the survey is conducted in China, we firstly developed
an English version questionnaire and then translated it into Chinese
Demographics Frequencies Percentages (%)
version. In addition, to verify and improve the adapted survey items, a
Gender pilot survey is conducted. Based on the feedbacks, several modifications
0. Female 171 53.44% and refinements of the items are made such as deleting some items,
1. Male 149 46.56% changing wordings of items and adjusting the format of the ques-
Age
1. Under 25 51 15.94%
tionnaire.
2.25–35 111 34.69%
3.36–45 83 25.94%
4.46–55 46 14.38% 4.3. Descriptive statistics
5.56 and over 29 9.06%
Educational level
1. High school or below 20 6.25% The results of descriptive statistics for all the constructs are shown
2. Associate degree 125 39.06% in Table 3. The mean of construct reflects the level of participants’
3. Bachelor's degree 102 31.88% perception to the constructs studied (see Fig. 1). The participants show
4. Master's degree 58 18.13% higher levels of subjective norm (Mean = 3.96), habit (Mean = 3.75),
5. PhD 15 4.69%
Monthly income
perceived behavioral control (Mean = 3.63), personal moral norm
1. Less than4000 RMB 20 6.25% (Mean = 3.60), attitude towards saving electricity (Mean = 3.51) and
2.4000–8000 RMB 155 48.44% electricity saving behavior (Mean = 3.51). The levels of positive an-
3.8001–12000 RMB 113 35.31% ticipated emotion (Mean = 3.16) and intention to save electricity
4. More than12000 RMB 32 10.00%
(Mean = 3.01) are relatively smaller and near to the midpoint of the
Family size
1.1–2 79 24.69% scale (namely 3). In addition, as presented in Table 3, there exist as-
2. 3–4 193 60.31% sociations between each of the construct and all the correlations among
3. More than 4 48 15.00% the constructs are lower than 0.60, which means that the multi-
Total 320 100 collinearity is not a serious problem and it is appropriate to conduct
further analysis (Gao et al., 2017).
4.2. Measures

4.4. Data analysis method

Constructs in this research are latent variables and they are mea-
sured with multiple item scales. All the items are developed from prior
Structural equation modeling (SEM) technique is a commonly used
research papers. Some of the wordings of the items are slightly modified
method to test the models with latent variables (Chin, 1998). Thus, in
to fit the context of this research. The items are measured on a five-
this research, SEM technique is adopted to analyze the collected data.
point Likert scale ranging from 1 (“strongly disagree”) to 5 (“strongly
There are two kinds of models in SEM analysis, namely measurement
agree”). The constructs and items are shown in Table A1.
model and structural model (Anderson and Gerbing, 1988). The mea-
The measurement items of attitude towards saving electricity, sub-
surement model is used to test the relationships between latent vari-
jective norm, perceived behavioral control and intention to save elec-
ables and the corresponding items, and the structural model is used to
tricity are developed based on Wang et al. (2014), Cheung et al. (2017)
examine the relationships between these latent variables.
and Ding et al. (2017). The measurement items of electricity saving
In the following data analysis, Amos 22.0 software packages are
behavior are developed based on Barr et al. (2005) and Hori et al.
used to estimate the measurement model to assess the reliability and
(2013). The measurement items of personal moral norm are adapted
validity of the constructs, and then assess the structural model to test
from Bamberg et al. (2007) and Fornara et al. (2016). The measurement
the relationships among the hypothesized constructs.
items of habit are adapted from Zhang et al. (2013) and Russell et al.
(2017). The measurement items of positive anticipated emotion are

175
S. Wang et al. Energy Policy 115 (2018) 171–180

Table 2
Results of confirmatory factor analysis.

Construct Item Standardized factor Loading Cronbach's alpha Composite reliability AVE

Attitude (AT) AT1 0.86*** 0.88 0.93 0.81

AT2 0.90***
AT3 0.93***
Subjective norm (SN) SN1 0.81*** 0.80 0.88 0.71
SN2 0.85***
SN3 0.86***
Perceived behavioral control (PBC) PBC1 0.84*** 0.79 0.87 0.70
PBC2 0.85***
PBC3 0.82***
Personal moral norm (PMN) PMN1 0.85*** 0.83 0.90 0.74
PMN2 0.88***
PMN3 0.86***
Habit (HAT) HAT1 0.82*** 0.83 0.89 0.74
HAT2 0.85***
HAT3 0.90***
Positive anticipated emotion (EMO) EMO1 0.92*** 0.83 0.89 0.74
EMO2 0.93***
EMO3 0.71***
Intention to save electricity (INT) INT1 0.85*** 0.80 0.88 0.71
INT2 0.85***
INT3 0.83***
Electricity saving behavior (BEH) BEH1 0.83*** 0.85 0.91 0.77
BEH2 0.90***
BEH3 0.90***

Note: Significant at *** p < 0.001.

5. Data analysis and results
5.1. Measurement model analysis
Confirmatory factor analysis (CFA) is performed to examine each
construct's measurement properties such as reliability, uni-
dimensionality, convergent validity and discriminant validity. The CFA
results indicate an acceptable fit between the measurement model and
the data set. The model fit indices are listed as follows: the ration of Chi-
square to degree of freedom (χ2/df) is 1.69, Comparative fit index (CFI)
is 0.96, Incremental fit index (IFI) is 0.96, Normed fit index (NFI) is
0.91, Tucker-Lewis index (TLI) is 0.94, Goodness of fit index (GFI) is
0.93, and Root mean square error of approximation (RMSEA) is 0.04.
In general, Cronbach's alpha values and composite reliability values
are used to assess the reliability of the constructs (Fornell and Larcker,
1981). As shown in Table 2, all Cronbach's alpha values range from
0.79 to 0.88 and composite reliability values range from 0.87 to 0.93.
They are all higher than the recommended threshold value of 0.70 and
support for reliability (Fornell and Larcker, 1981). To evaluate the
convergent validity of the constructs, individual item loading and
average variance extracted (AVE) are used (Fornell and Larcker, 1981).
As shown in Table 2, the CFA results indicate that all items are sig-
nificantly loaded on the constructs. The loadings are significant at
p < 0.001 level and greater than the threshold value of 0.70. All the
AVE values range from 0.70 to 0.81, both of which exceed the bench-
mark value of 0.50 (Fornell and Larcker, 1981). These results provide a
strong support for unidimensionality and convergent validity.
Furthermore, we need to test the discriminant validity of the con-
structs. According to Paulraj et al. (2008), the relationships between
shared variances among constructs and AVE values are compared. As
shown in Table 3, all the correlations between constructs are less than
the square roots of AVE values, which support the discriminant validity
of the constructs. Therefore, it can be concluded that the reliability and
validity of the latent constructs are acceptable.
5.2. Structural model and hypothesis analysis
Though a four-stage questionnaire survey is conducted to alleviate
the common method bias, it is necessary to further assess the common
method bias in structural model analysis. According to Harman (1976),
Harman's one-factor test is performed. The results indicate that all the
measurement items could be divided into seven factors, with eigenva-
lues greater than 1.0 and accounting for 71.80% of the variance. The
first factor only explains 23.33% of the variance and less than the
benchmark value of 50.0% (Harman, 1976). The results indicate that
the common method bias is not a major concern in this research.
In order to test the proposed research hypotheses, the structural
model is analyzed. Overall, the model fit for the structural model is

Table 3
Means, Standard Deviations and Correlations (n = 320).

Construct Mean SD AT SN PBC PMN HAT EMO INT BEH

AT 3.51 0.78 0.90

SN 3.96 0.75 0.29** 0.84
PBC 3.63 0.74 0.19** 0.48** 0.84
PMN 3.60 0.78 0.07 0.08 0.09 0.86
HAT 3.75 0.68 0.03 0.06 0.02 0.11* 0.86
EMO 3.16 0.65 −0.06 −0.06 0.04 0.06 0.01 0.86
INT 3.01 0.86 0.24** 0.30** 0.45** 0.21** 0.11 0.11 0.84
BEH 3.51 0.88 0.09 0.34** 0.55** 0.18** 0.19** 0.04 0.53** 0.88

Note: (1) The diagonal (bold) elements are the square roots of AVE values and the off-diagonal elements are the correlations among the constructs. (2) Significant at * p < 0.05 and **
p < 0.01.

176
S. Wang et al.
Attitude
0.15** 0.07*
Subjective norm
Subjective norm
0.08
Intention to save
Intention to save
electricity
electricity
0.14**
Perceived behavioral
Perceived behavioral
control R2=0.26
control
0.11*
0.37***
Personal moral norm Positive anticipate
Personal moral norm
good (χ2/df = 1.84, CFI = 0.94, IFI = 0.95, NFI = 0.91, TLI = 0.93,
GFI = 0.93, RMSEA = 0.05). The model explains 26% of the variance
in intention to save electricity and 30% of the variance in electricity
saving behavior. Fig. 2 presents the results of hypothesis testing. The
results suggest that, in line with TPB, attitude towards saving electricity
(β = 0.15, p < 0.01) and perceived behavioral control (β = 0.14,
p < 0.01) are significantly related to intention to save electricity, thus
providing support for Hypotheses 1a and 1c. The path coefficient from
subjective norm to intention to save electricity is not significant (β =
0.08, p > 0.05), thus Hypothesis 1b is not supported. Consistent with
TPB, intention to save electricity is significantly and positively related
to electricity saving behavior (β = 0.53, p < 0.001), thus providing
support for Hypothesis 2.
Hypothesis 3 posits that personal moral norm positively affects the
intention to save electricity. According to Fig. 2, it can be observed that
the path coefficient is 0.37 (p < 0.001), thus supporting Hypothesis 3.
Hypotheses 4a and 4b discuss the relationships between habit and in-
tention to save electricity (4a) and electricity saving behavior (4b). The
results in Fig. 2 suggest that habit positively affects the intention to save
electricity (β = 0.07, p < 0.05) and electricity saving behavior (β =
0.14, p < 0.01), thus, supporting Hypotheses 4a and 4b. Hypotheses 5a
and 5b posit that positive anticipated emotion positively affects inten-
tion to save electricity (5a) and electricity saving behavior (5b). Ac-
cording to Fig. 2, Hypothesis 5a is supported (β = 0.11, p < 0.05) and
Hypothesis 5b is not supported (β = −0.03, p < 0.05). The results
suggest that the experience of more positive anticipated emotion is
associated with higher intention to save electricity but contrary to our
predictions, is associated with lower actual electricity saving behavior.
6. Discussion and implication
The present research uses a comprehensive model which in-
corporates not only traditional cognitive factors (attitude, subjective
norm and perceived behavioral control) but also the neglected non-
cognitive and emotional factors (personal moral norm, habit and po-
sitive anticipated emotion) to understand household electricity saving
behavior. This research highlights the importance of non-cognitive and
emotional factors as drivers of electricity saving behavior.
The results suggest that personal moral norm, habit and positive
anticipated emotion are positively and significantly related to residents’
intention to save electricity. Meanwhile, we find that habit is also po-
sitively associated with electricity saving behavior. These results are
consistent with previous studies in understanding pro-environmental
behavior such as energy conservation behavior, green purchasing
Energy Policy 115 (2018) 171–180
Fig. 2. Results of hypothesis testing.
Habit
Habit
0.14**
Electricity
0.53*** Electricity
saving behavior
saving behavior
R2=0.30
-0.03*
Positive anticipate
emotion
emotion
behavior and recycling behavior (Klöckner, 2013; Webb et al., 2013;
Cheung et al., 2017; Russell et al., 2017). However, contrary to our
expectations, positive anticipated emotion is negatively associated with
electricity saving behavior. This is the inverse of the expected re-
lationship and indicates that the more positive anticipated emotion
residents present, the less electricity saving behavior they will perform
in practice. This is particularly surprising because the results also de-
monstrate that positive anticipated emotion has a positive effect on
intention to save electricity. These results suggest that positive antici-
pated emotion does not translate into actual electricity saving behavior,
despite residents intend to do so.
These findings are confusing but interesting to explain. The reasons
can be explained as follows. As we know, some scholars such as Chen
and Tung (2014), Wang et al. (2014), Russell et al. (2017) and Gao
et al. (2017) have noted that there exit difference between behavioral
intention and actual behavior. Behavioral intention means that the
specific behavior will occur sometime in the future, while the actual
behavior means that the specific behavior will occur immediately or has
already occurred. According to Lerner and Keltner (2001), the beha-
vioral responses of positive anticipated emotion are quite different.
Individuals are more likely to state that they are willing to perform the
behavior that occurs in the future and reluctant to perform the behavior
that occurs immediately. Based on the same logic, in this research, it is
known that residents’ intention to save electricity occurs in the future
and therefore the positive anticipated emotion can lead to strong and
positive intention. However, considering electricity saving behavior is
imminent, residents may unwilling to sacrifice their comfortable living
style to perform the electricity saving behavior at once. What’ more,
some residents may have avoidance-oriented beliefs and think that they
can save electricity in the future so not perform the electricity saving
behavior right now is forgivable. Thus, the positive anticipated emotion
can lead to reduction in performing actual electricity saving behavior.
Another reason for why this result differs from other studies’ results is
due to the difference in data collection process. Other studies such as
Webb et al. (2013) and Han and Hyun (2018) rely on cross-section
design and collect data at the same point in time. Under this situation,
participants tend to give the same answers when they evaluate inten-
tion to save electricity and actual electricity saving behavior simulta-
neously. However, in this research, a four-stage questionnaire survey is
conducted and the participants evaluate intention to save electricity
and actual electricity saving behavior at different points in time. Par-
ticipants’ perceptions to these two constructs are quite different. Thus,
positive anticipated emotion has different effects on intention to save
electricity and actual electricity saving behavior.
177
S. Wang et al.
Furthermore, consistent with TPB, attitude and perceived beha-
vioral control are positively related to intention to save electricity.
However, the effect of subjective norm on intention to save electricity is
not significant, which means that important others (e.g., relatives or
neighbors) do not influence residents’ intention to save electricity. This
result is inconsistent with previous studies (Yadav and Pathak, 2016;
Fornara et al., 2016; Cheung et al., 2017; Shi et al., 2017). One rea-
sonable explanation is that the household electricity saving behavior is
almost invisible to friends, relatives or neighbors. Residents’ electricity
saving behavior almost relies on themselves and is not affected by
others. Hence, the effect of subjective norm is not significant. In addi-
tion, it is worth noting that personal moral nom has the largest impact
on intention to save electricity, which means that residents with high
levels of personal moral norm are more inclined to save electricity. This
result is consistent with previous studies, which reported that the per-
sonal moral norm is associated with a stronger preference for per-
forming pro-environmental behavior (De Groot and Steg, 2009; Webb
et al., 2013; Fornara et al., 2016).
The findings of this research have several implications for scholars
and practitioners. Theoretically, this research enriches and contributes
to household electricity saving literatures by proposing a comprehen-
sive model which incorporates cognitive, non-cognitive and emotional
factors. Previous research neglects to explore the role of non-cognitive
and emotional factors in household electricity saving (Cheung et al.,
2017; Gao et al., 2017). This research is a timely response and under-
scores the importance of non-cognitive and emotional factors. Fur-
thermore, this research suggests that behavioral intention cannot be
considered as a proxy measure for behavior. Indeed, if we only measure
behavioral intention it will draw the conclusion that the positive an-
ticipated emotion is an effective way for future interventions designed
to save electricity. However, in fact, this research presents a negative
relationship between positive anticipated emotion and the self-reported
behavior. Thus, in the following studies, scholars should be careful to
conflate behavioral intention and behavior. In addition, a four-stage
questionnaire survey is conducted in this research to collect data at
different time points to reduce the common method bias. The ques-
tionnaire survey method in this research is beneficial and referential to
scholars in their following studies.
In practice, based on the research findings, this research also pro-
vides important implications for saving electricity. Firstly, considering
the importance of personal moral norm and positive anticipated emo-
tion, national and local education and publicity campaigns about
household electricity saving can be set up for residents (Gao et al.,
2017). These education and publicity campaigns can make residents
realize that every citizen has the responsibility and obligation to reduce
electricity consumption and save electricity which is beneficial to im-
prove their personal moral norm (Webb et al., 2013). Meanwhile, re-
levant government agencies can provide certain reports to residents to
highlight the negative consequences of wasting electricity (e.g., in-
tensifying energy crisis and increasing carbon emissions) and explain
the social benefits of saving electricity to foster their positive antici-
pated emotion about electricity conservation (Fornara et al., 2016).
Secondly, considering the importance of perceived behavioral control
and habit, electricity saving manuals can be provided to residents and
electricity saving lectures and training programs can be also launched
by residential committees (Shi et al., 2017). These measures can en-
hance residents’ skills and knowledge to save electricity and thus to
improve their perceived behavioral control. In addition, these measures
can also make residents realize that it is not difficult to save electricity
when they follow the guidelines and help them to develop a good habit
to save electricity.
7. Conclusions and limitations
In this research, a comprehensive model is developed to explain
178
Energy Policy 115 (2018) 171–180
household electricity saving behavior. A four-staged questionnaire
survey is conducted to collect data from 320 households in Hefei,
China. SEM is adopted to analyze the collected data and test the re-
search hypotheses. The results indicate that attitude, perceived beha-
vioral control, personal moral norm, habit and positive anticipated
emotion are significantly related to intention to save electricity, while
subjective norm has no significant effect on intention to save electricity.
Meanwhile, it is worth noting that positive anticipated emotion is ne-
gatively associated with actual electricity saving behavior. Our findings
underscore the importance of non-cognitive and emotional factors as
drivers of electricity saving behavior and suggest that these non-cog-
nitive and emotional drivers deserve much more attention. This re-
search adds to our understanding of household electricity saving be-
havior and provides important implications for future researchers and
practitioners.
Although this research offers some interesting findings and makes
significant theoretical and practical implications, it has several limita-
tions which must be acknowledged and possibly overcome in future
research. Firstly, the data are only collected from one city, namely
Hefei, China, which may restrict the generalizability of results. Though
Hefei shares some common characteristics with other areas and regions,
it also has some characteristics of its own. Since the 21st century, along
with the deepening of political and economic system reforms, the speed
of economy development in Hefei is amazing. According to “The
Economist”, Hefei is the fastest growing cities in China in 2015.
Meanwhile, the human population has also increased rapidly. At the
end of 2015, the size of the permanent resident population in Hefei is
more than 7 million. With the fast development of economy and the
population increase, and with the improvement of residents’ living
standards, the household electricity consumption has increased sharply.
In addition, the natural resource endowment is poor in Hefei and the
electricity resource is insufficient and it needs other cities’ electricity
support. Electricity saving has drawn wide attention and residents’
environmental awareness is relatively high and they are planning to
saving electricity in their daily lives. The findings may be effective in
specific contexts and may not be suitable to other contexts. In general,
the findings of this research are suitable for areas which have the si-
milar characteristics with Hefei, such as economic development level,
institutional and economic environment, natural resource endowment,
population, and residents’ environmental awareness. The findings of
this research are not suitable for other areas which have different
characteristics with Hefei. Thus, in future research, we should attempt
to collect data from more areas and regions to make the findings more
general and universal. Secondly, this research relies merely on self-re-
ported electricity saving behavior. Participants are inclined to over-
estimate their electricity saving behavior and thus affect the accuracy of
research conclusions. In the following studies, scholars can conduct
observations to measure electricity saving behavior. Finally, this re-
search only considers the effects of psychological factors on household
electricity saving behavior. Other factors such as information feedback
and financial incentive scheme are not considered. This is what we set
out to do in future research.
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (Grant number 71601174 and 71571172), China
Postdoctoral Science Foundation (Grant number 2016M590583),
Fundamental Research Funds for the Central Universities, National
Social Science Foundation of China (Grant number 16CJL020) and
Fujian Fund of Soft Science Research (Grant number 2017R0034).
S. Wang et al. Energy Policy 115 (2018) 171–180

Appendix A

See Appendix Table A1

Table A1
Constructs and measurement items.

Constructs Measurement items

Attitude Saving electricity in my home is significant to reduce carbon emissions.

Saving electricity in my home is valuable to improve the air quality.
Saving electricity in my home is a wise move.
Subjective norm My neighbors think that I should save electricity in daily lives.
My relatives think that I should save electricity in daily lives.
Other people who are important to me think I should save electricity in daily lives.
Perceived behavioral control I believe that I am capable of saving electricity in my home.
I have knowledge and skills to save electricity in my home.
It is mostly up to me whether I save electricity in my home.
Personal moral norm Waste electricity in my home would go against my environmental protection principles.
I have a moral obligation to save electricity to contribute to the improvement of the environment.
I would feel guilty about not saving electricity in my home.
Habit Saving electricity in my home has become automatic to me.
Saving electricity in my home has become a habit to me.
Saving electricity in my home is natural to me.
Positive anticipated emotion If I take measures to save electricity in my home, I will feel excited.
If I take measures to save electricity in my home, I will feel relaxed.
If I take measures to save electricity in my home, I will feel proud.
Intention to save electricity I intend to save electricity in my home.
I plan to save electricity in my home.
I will make an effort to save electricity in my home.
Electricity saving behavior I always turn off electrical appliances when no one else is left in the room.
I always turn electrical appliances off completely rather than to a standby mode.
I have purchased energy efficient electrical appliances in the past few years.

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), 711–720.
Ajzen, I., 1991. The theory of planned behavior. Organ. Behav. Hum. Decis. Process. 50
(2), 179–211.
Anderson, J.C., Gerbing, D.W., 1988. Structural equation modeling in practice: a review
and recommended two-step approach. Psychol. Bull. 103 (3), 411.
Arawomo, D.F., 2017. Electricity billing systems and household electricity use behavior in
Ibadan, Nigeria. Energy Sustain. Dev. 40, 77–84.
Armstrong, J.S., Overton, T.S., 1977. Estimating nonresponse bias in mail surveys. J.
Mark. Res. 396–402.
Bamberg, S., 2003. How does environmental concern influence specific environmentally
related behaviors? A new answer to an old question. J. Environ. Psychol. 23 (1),
21–32.
Bamberg, S., Hunecke, M., Blöbaum, A., 2007. Social context, personal norms and the use
of public transportation: two field studies. J. Environ. Psychol. 27 (3), 190–203.
Barr, S., Gilg, A.W., Ford, N., 2005. The household energy gap: examining the divide
between habitual-and purchase-related conservation behaviors. Energy Policy 33
(11), 1425–1444.
Bertoldo, R., Castro, P., 2016. The outer influence inside us: exploring the relation be-
tween social and personal norms. Resour. Conserv. Recycl. 112, 45–53.
Chen, M.F., Tung, P.J., 2014. Developing an extended theory of planned behavior model
to predict consumers' intention to visit green hotels. Int. J. Hosp. Manag. 36,
221–230.
Cheung, L.T., Chow, A.S., Fok, L., Yu, K.M., Chou, K.L., 2017. The effect of self-de-
termined motivation on household energy consumption behavior in a metropolitan
area in southern China. Energy Effic. 10 (3), 549–561.
Chin, W.W., 1998. The partial least squares approach to structural equation modeling.
Mod. Methods Bus. Res. 295 (2), 295–336.
Davies, J., Foxall, G.R., Pallister, J., 2002. Beyond the intention-behavior mythology: an
integrated model of recycling. Mark. Theory 2 (1), 29–113.
De Groot, J.I., Steg, L., 2009. Morality and pro-social behavior: the role of awareness,
responsibility, and norms in the norm activation model. J. Soc. Psychol. 149 (4),
425–449.
De Leeuw, A., Valois, P., Ajzen, I., Schmidt, P., 2015. Using the theory of planned be-
havior to identify key beliefs underlying pro-environmental behavior in high-school
students: implications for educational interventions. J. Environ. Psychol. 42,
128–138.
Demarque, C., Charalambides, L., Hilton, D.J., Waroquier, L., 2015. Nudging sustainable
consumption: the use of descriptive norms to promote a minority behavior in a
realistic online shopping environment. J. Environ. Psychol. 43, 166–174.
Ding, Z., Wang, G., Liu, Z., Long, R., 2017. Research on differences in the factors
influencing the energy-saving behavior of urban and rural residents in China-A case
study of Jiangsu Province. Energy Policy 100, 252–259.
Du, L., Guo, J., Wei, C., 2017. Impact of information feedback on residential electricity
demand in China. Resour. Conserv. Recycl. 125, 324–334.
Fornara, F., Pattitoni, P., Mura, M., Strazzera, E., 2016. Predicting intention to improve
household energy efficiency: the role of value-belief-norm theory, normative and
informational influence, and specific attitude. J. Environ. Psychol. 45, 1–10.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable
variables and measurement error. J. Mark. Res. 18, 39–50.
Frederiks, E.R., Stenner, K., Hobman, E.V., 2015. Household energy use: applying be-
havioral economics to understand consumer decision-making and behavior. Renew.
Sustain. Energy Rev. 41, 1385–1394.
Gao, L., Wang, S., Li, J., Li, H., 2017. Application of the extended theory of planned
behavior to understand individual's energy saving behavior in workplaces. Resour.
Conserv. Recycl. 127, 107–113.
Georges, E., Cornélusse, B., Ernst, D., Lemort, V., Mathieu, S., 2017. Residential heat
pump as flexible load for direct control service with parametrized duration and re-
bound effect. Appl. Energy 187, 140–153.
Han, H., Hyun, S.S., 2018. What influences water conservation and towel reuse practices
of hotel guests? Tour. Manag. 64, 87–97.
Handgraaf, M.J., de Jeude, M.A.V.L., Appelt, K.C., 2013. Public praise vs. private pay:
effects of rewards on energy conservation in the workplace. Ecol. Econ. 86, 86–92.
Harman, H.H., 1976. Modern Factor Analysis. University of Chicago Press, Chicago IL.
Hassin, R.R., Bargh, J.A., Zimerman, S., 2009. Automatic and flexible: the case of non-
conscious goal pursuit. Soc. Cogn. 27 (1), 20–36.
Hori, S., Kondo, K., Nogata, D., Ben, H., 2013. The determinants of household energy-
saving behavior: survey and comparison in five major Asian cities. Energy Policy 52,
354–362.
Kaiser, F.G., Scheuthle, H., 2003. Two challenges to a moral extension of the theory of
planned behavior: moral norms and just world beliefs in conservationism. Personal.
Individ. Differ. 35 (5), 1033–1048.
Kals, E., Schumacher, D., Montada, L., 1999. Emotional affinity toward nature as a mo-
tivational basis to protect nature. Environ. Behav. 31 (2), 178–202.
Klöckner, C.A., 2013. A comprehensive model of the psychology of environmental be-
havior-a meta-analysis. Glob. Environ. Change 23 (5), 1028–1038.
Klöckner, C.A., Blöbaum, A., 2010. A comprehensive action determination model: toward
a broader understanding of ecological behavior using the example of travel mode
choice. J. Environ. Psychol. 30 (4), 574–586.
Lam, T., Hsu, C.H., 2004. Theory of planned behavior: potential travelers from China. J.
Hosp. Tour. Res. 28 (4), 463–482.
Lerner, J.S., Keltner, D., 2001. Fear, anger, and risk. J. Personal. Soc. Psychol. 81 (1), 146.
Martinsson, J., Lundqvist, L.J., Sundström, A., 2011. Energy saving in Swedish house-
holds. The (relative) importance of environmental attitudes. Energy Policy 39 (9),
5182–5191.
Morris, J.D., Woo, C., Geason, J.A., Kim, J., 2002. The power of affect: predicting

179
S. Wang et al.
intention. J. Advert. Res. 42 (3), 7–17.
Murata, A., Kondou, Y., Hailin, M., Weisheng, Z., 2008. Electricity demand in the Chinese
urban household-sector. Appl. Energy 85 (12), 1113–1125.
Nilsson, A., Hansla, A., Biel, A., 2014b. Feeling the green? Value orientation as a mod-
erator of emotional response to green electricity. J. Appl. Soc. Psychol. 44 (10),
672–680.
Nilsson, A., Bergstad, C.J., Thuvander, L., Andersson, D., Andersson, K., Meiling, P.,
2014a. Effects of continuous feedback on households' electricity consumption: po-
tentials and barriers. Appl. Energy 122, 17–23.
Ouyang, J., Hokao, K., 2009. Energy-saving potential by improving occupants' behavior
in urban residential sector in Hangzhou City, China. Energy Build. 41 (7), 711–720.
Paulraj, A., Lado, A.A., Chen, I.J., 2008. Inter-organizational communication as a rela-
tional competency: antecedents and performance outcomes in collaborative buyer-
supplier relationships. J. Oper. Manag. 26 (1), 45–64.
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., Podsakoff, N.P., 2003. Common method
biases in behavioral research: a critical review of the literature and recommended
remedies. J. Appl. Psychol. 88 (5), 879.
Pothitou, M., Hanna, R.F., Chalvatzis, K.J., 2016. Environmental knowledge, pro-en-
vironmental behavior and energy savings in households: an empirical study. Appl.
Energy 184, 1217–1229.
Quested, T.E., Marsh, E., Stunell, D., Parry, A.D., 2013. Spaghetti soup: the complex world
of food waste behaviors. Resour. Conserv. Recycl. 79, 43–51.
Reiss, P.C., White, M.W., 2005. Household electricity demand, revisited. Rev. Econ. Stud.
72 (3), 853–883.
Rivis, A., Sheeran, P., Armitage, C.J., 2009. Expanding the affective and normative
components of the theory of planned behavior: a meta-analysis of anticipated affect
and moral norms. J. Appl. Soc. Psychol. 39 (12), 2985–3019.
Russell, S., Young, C.W., Hardin, K., Robinson, C., 2017. Bringing habits and emotions
into food waste behavior. Resour. Conserv. Recycl. 125, 107–114.
da Silva, P.P., Cerqueira, P.A., 2017. Assessing the determinants of household electricity
prices in the EU: a system-GMM panel data approach. Renew. Sustain. Energy Rev.
73, 1131–1137.
Shi, H., Fan, J., Zhao, D., 2017. Predicting household PM 2.5-reduction behavior in
Chinese urban areas: an integrative model of theory of planned behavior and norm
activation theory. J. Clean. Prod. 145, 64–73.
Sorrell, S., Dimitropoulos, J., 2008. The rebound effect: microeconomic definitions,
180
Energy Policy 115 (2018) 171–180
limitations and extensions. Ecol. Econ. 65 (3), 636–649.
Steg, L., Vlek, C., 2009. Encouraging pro-environmental behavior: an integrative review
and research agenda. J. Environ. Psychol. 29 (3), 309–317.
Stern, P.C., 2000. Towards a coherent theory of environmentally significant behavior. J.
Soc. Issues 56, 407–424.
Verplanken, B., Holland, R.W., 2002. Motivated decision making: effects of activation and
self-centrality of values on choices and behavior. J. Personal. Soc. Psychol. 82 (3),
434.
Wahid, N.A., Rahbar, E., Shyan, T.S., 2011. Factors influencing the green purchase be-
havior of Penang environmental volunteers. Int. Bus. Manag. 5 (1), 38–49.
Wang, Z., Zhang, B., Yin, J., Zhang, Y., 2011. Determinants and policy implications for
household electricity-saving behavior: evidence from Beijing, China. Energy Policy
39 (6), 3550–3557.
Wang, Z., Zhang, B., Li, G., 2014. Determinants of energy-saving behavioral intention
among residents in Beijing: extending the theory of planned behavior. J. Renew.
Sustain. Energy 6 (5), 053127.
Webb, D., Soutar, G.N., Mazzarol, T., Saldaris, P., 2013. Self-determination theory and
consumer behavioral change: evidence from a household energy-saving behavior
study. J. Environ. Psychol. 35, 59–66.
Weiss, H.M., Beal, D.J., 2005. Reflections on Affective Events Theory. In The effect of
affect in Organizational Settings. Emerald Group Publishing Limited, Bingley, UK, pp.
1–21.
Yadav, R., Pathak, G.S., 2016. Young consumers' intention towards buying green products
in a developing nation: extending the theory of planned behavior. J. Clean. Prod. 135,
732–739.
Yu, Y., Guo, J., 2016. Identifying electricity-saving potential in rural China: empirical
evidence from a household survey. Energy Policy 94, 1–9.
Yuan, J., Zhao, C., Yu, S., Hu, Z., 2007. Electricity consumption and economic growth in
China: cointegration and co-feature analysis. Energy Econ. 29 (6), 1179–1191.
Zhang, Y., Wang, Z., Zhou, G., 2013. Determinants and implications of employee elec-
tricity saving habit: an empirical study in China. Appl. Energy 112, 1529–1535.
Zhou, K., Yang, S., 2016. Understanding household energy consumption behavior: the
contribution of energy big data analytics. Renew. Sustain. Energy Rev. 56, 810–819.
Zhou, S., Teng, F., 2013. Estimation of urban residential electricity demand in China using
household survey data. Energy Policy 61, 394–402.