Accepted Manuscript

Impact Factors of Household Energy-Saving Behavior: An Empirical Study of

Shandong Province in China

Cheng-Yao Zhang, Biying Yu, Jin-Wei Wang, Yi-

Ming Wei PII: S0959-6526(18)30641-3

DOI: 10.1016/j.jclepro.2018.02.303
Reference: JCLP 12250

To appear in: Journal of Cleaner Production

Received Date: 18 May 2017

Revised Date: 26 February 2018
Accepted Date: 28 February 2018

Please cite this article as: Cheng-Yao Zhang, Biying Yu, Jin-Wei Wang, Yi-Ming Wei,
Impact Factors of Household Energy-Saving Behavior: An Empirical Study of Shandong
Province in China, Journal of Cleaner Production (2018), doi:
10.1016/j.jclepro.2018.02.303

This is a PDF file of an unedited manuscript that has been accepted for publication. As a
service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form.
Please note that during the production process errors may be discovered which could
affect the content, and all legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Impact Factors of Household Energy-Saving Behavior: An Empirical Study of

Shandong Province in China

Cheng-Yao Zhang 1,2,3, Biying Yu 1,2,3,4,, Jin-Wei Wang 1,2,3, Yi-Ming Wei1,2,3

1 Centerfor Energy and Environment Policy Research, Beijing Institute of Technology, Beijing,
100181, China.
2 School of Management and Economics, Beijing Institute of Technology, Beijing, 100181, China.
3 Beijing Key Lab of Energy Economics and Environmental Management, Beijing 100081, China.
4 Sustainable Development Research Institute for Economy and Society of Beijing, Beijing 100081,
China.

Abstract: As China gradually completes the process of industrialization, its industrial energy
consumption growth is now slowing and even decreasing. Meanwhile, household energy
consumption in the residential sector has seen steady growth. This paper aims to explore the
mechanisms of factors affecting urban household energy-saving behavior including the habitual
energy-saving behaviors and purchasing energy-saving behaviors. A structural equation model is
built to analyze the influencing routes and effects of individual objective and subjective
characteristic factors, external influencing factors, and energy-saving intentions on shaping
energy- saving behaviors. The empirical results drawing on the survey data collected in Shandong
province show that external influencing factors that have two kinds of mediating effects through
energy- saving intentions and individual subjective characteristic factors are the most crucial
factors to energy-saving behaviors. Values included in individual subjective factors and quality of
energy- saving products included in external influencing factors have greatest effects on energy-
saving behaviors.

Key words: Individual subjective factors; external influencing factors; energy-saving intentions;

energy-saving behaviors; Structural equation model

1. Introduction
Attributing to the fast-growing economy and high growth of per capita income in the last


Corresponding author. Phone & Fax: +86-10-68918651;
E-mail address: yubiying_bj@bit.edu.cn; yubiying_bj@hotmail.com.
2
 ACCEPTED MANUSCRIPT

decade in China (annual growth being 10.4% for urban residents and 12.9% for rural residents)

(Chinese Statistic Yearbook, 2017), household domestic energy consumption (i.e., residential

energy consumption) is becoming more and more significant. Since 2012, Chinese residential

energy consumption has become the second largest energy consumption sector behind only the

industrial sector. In 2015, residential energy consumption reached 50,099 tons of standard coal

equivalent (tce), which comprised as much as 11.7% of the total Chinese energy consumption

(Chinese Energy Statistic Yearbook, 2016). Consequently, China’s 13th Five-Year (2016-2020)

Plan and the 2016 Energy Work Guidance have paid great attention to household energy

consumption.

The household energy consumption is engendered by people’s daily behaviors including the

ownership behavior of energy-consuming end uses and end-use usage behavior, which is more

heterogeneous and difficult to regulate than of other sectors. As a result, the effectiveness of

energy conservation policies on household sector become much uncertain. Therefore, to reduce

household energy consumption, it seems more reasonable to start from the behavioral perspective,

in other words, the sustainability in the household sector to a large extent relies on whether

energy-saving behavior is adopted or not.

Energy-saving behavior can be defined as the behaviors through which people try to reduce

overall energy use (Barr et al., 2005). Generally, it is divided into two categories (Barr et al.,

2005; Black et al., 1985; Dillman et al., 1983; Raaij and Verhallen, 1983; Stern, 1992) : (1)

habitual energy-saving behaviors with repetitive efforts to reduce energy use by curtailment

measures (e.g., turning the power off when not using the appliances, choosing a shower instead of

a bath, using daily public transports or on foot instead of private cars, and avoiding or reducing

the usage of air conditioners); and (2) one-shot purchasing behaviors targeting the purchase of

more-efficient technology and the replacement of the old technology (e.g., buying energy-saving

lamps, choosing electric vehicle instead of gasoline cars, and purchasing simple packaging and

recyclable products). In recent years, research on energy-saving behaviors has gradually

caught more and more attention of scholars throughout the world (Barr et al., 2005; Hori et al.,

2013; Jackson, 2005; Kang et al., 2012; Wang et al., 2014a; Webb et al., 2013; Yue et al., 2013).

Household energy-saving behaviors are revealed to be affected by many factors, which can be

summarized into three main sources: individual characteristic factors, external influencing factors
3
 ACCEPTED MANUSCRIPT

and energy-saving intentions

(Bartkus et al., 1999; Chan and Yam, 1995; Haron et al., 2005; Reiss and White, 2006; Wang et al.,

4
 ACCEPTED MANUSCRIPT

2014b). The individual characteristic factors can be further classified into objective and subjective

factors. The former is also known as sociodemographic characteristics, such as gender, age,

income and education background (Alibeli, 2009; Schwepker and Cornwell, 1991). The individual

subjective characteristic factors indicate the factors mainly reflecting individual behavioral

attitude, preference, subjective norm and perceived behavioral control, such as environmental

awareness, sense of responsibility, values and energy-saving knowledge (Bartkus et al., 1999;

Chan and Yam, 1995; Gärling et al., 2003; Haron et al., 2005; Reiss and White, 2006; Wang et al.,

2014b). Regarding the external influencing factors related to the market, society and government,

for example, persuasion and demonstration, social norms, prices, policies and regulations and

other extrinsic factors, it is highlighted that their influences cannot be neglected (Carlsson-

Kanyama et al., 2005; Egmond et al., 2005; Lindén and Carlsson-Kanyama, 2002; Sardianou,

2007; Steg, 2008). Energy-saving intention is a kind of behavioral intention. Some studies

mentioned behavioral intention was the most direct antecedent variable of behavior, which

represented the degree to which a person is ready to perform a certain behavior (Ajzen, 1991,

2002; Chen, 2013; Fishbein and Ajzen, 1975). This study defines the energy-saving intention as

the tendency to implement the energy- saving behavior.

Based on the previous literatures, this paper will consider the impacts of all these three

sources: the individual subjective and objective characteristic factors (e.g., gender, age,

occupation, degree of concern, sense of responsibility, values) , external influencing factors (e.g.,

the quality of energy- saving products, social norms, publicity and education) and energy-saving

intention, on shaping household energy-saving behaviors including habitual energy-saving

behaviors and purchase behaviors. A unified model, including all the elements we have already

mentioned, will be set up to analyze the mechanisms of motions and interrelationships of the

various factors influencing energy- saving behaviors. This study draws on the empirical context of

Jinan city, the capital of Shandong Province that is one of the provinces with the most developed

economy and highest energy consumption in China. The population, GDP and energy

consumption of Shandong province is about 7.2%, 9.1% and 9% in the total of China in 2016,

respectively. We handed out 350 questionnaires to residents in Jinan City, and 297 valid responses

were finally obtained for supporting this analysis.

The remaining part of this paper is structured as follows. A literature review in the second
5
 ACCEPTED MANUSCRIPT

section summarizes the articles on household energy-saving behaviors and analyzes factors that

influence those behaviors. Section 3 introduces our structural equation model (SEM), determines

variables, raises hypotheses and establishes the research model. Section 4 explains the survey

design and the collected data. Section 5 shows the model results and discussions for the empirical

study. This study ends with conclusions, policy implications and limitation in Section 6.

2. Literature review

2.1 Summary of influencing factors

More and more scholars have realized the importance of analyzing household energy
consumption from the behavioral perspective. Identifying the factors that influence household
energy consumption is the basis for exploring the effects, paths and mechanisms of shaping
individual energy-saving behaviors. Table 1 lists some relevant literature. In summary, existing
research shed light on the influence of demographic variables (e.g., age, gender), individual
subjective characteristic factors (e.g., values, energy knowledge) and external influencing factors
(e.g., social norms, policies) on the adoption of energy-saving behaviors, by employing the survey
data collected in some specific cities or regions. For example, Kang et al. (2012) examined the
impacts of individual subjective characteristic factors (e.g., energy-saving consciousness and
energy-saving knowledge) on residents’ energy-saving behavior, and significantly positive effect
were revealed. Feng and Reisner (2011) adopted regression analysis to analyze the impacts of
demographic and individual subjective characteristic factors. They found that individuals who had
more environmental knowledge, treated environmental issues more seriously and support
environmental resource conservation were more likely to implement energy-efficient behaviors,
and females did a better job than males. Several studies further paid attention to the impacts of
external influencing factors on the energy-saving behaviors in addition to the influence of
demographic variables and individual subjective characteristic factors (Gadenne et al.2011).
Moreover, some researchers argued that individual objective and subjective factors affected
energy-saving behaviors through acting on their energy-saving intentions (Wang et al., 2014b;
Webb et al., 2013), and the behavioral intention played the most important role in explaining the
behaviors (Wang et al.,2014b). Concerning the effect paths of the influencing factors on the
energy-saving behavior, majority of existing studies analyzed the direct effects of influencing
factors (Feng and Reisner, 2011; Kang et al., 2012; Nordlund and Garvill, 2003; Webb et al.,
2013), while only limited scholars further demonstrated the indirect or mediating effects between
influencing factors and energy-saving behaviors. For instance, Wang et al. (2014b) suggested that
6
 ACCEPTED MANUSCRIPT

energy knowledge exerted an indirect

7
 ACCEPTED MANUSCRIPT

influence on energy-saving behaviors via environmental attitudes. We can conclude that though
many studies have examined the mechanisms of shaping energy-saving behavior, most of them
only accommodate one or several dimensions of the above-mentioned factors, which are not
comprehensive enough at this moment. Besides, the influential pathways need to be expanded to
dig more into the behavioral mechanism for shaping energy-saving behavior.

2.2 Individual characteristic factors

In considering affective energy-saving behaviors, the existing research on individual factors

concentrates on both objective and subjective characteristic factors.

2.2.1 Individual objective characteristic factors

Many researchers treat individual objective characteristic factors as sociodemographic

variables, such as gender, education background, income or whether there is a child in the

household. Some studies held the view that the individual objective characteristic factors had

significant effect on energy-saving behaviors (Feng and Reisner, 2011; Martinsson et al., 2011;

Poortinga et al., 2003; Yue et al., 2013). For example, Poortinga et al. (2003) found that education

background was significantly related to energy-saving behaviors and was strongly correlated with

the adoption of different energy-saving measures. Yu et al. (2013) found that household income

and household size had significant influence on the purchase and usage of energy-saving

appliances. However, some studies got the opposite results, and they found that occupation had

little to do with behaviors, and the household’s education background or family size did not affect

energy-saving activities (Curtis et al., 1984; Sardianou, 2007; Wang et al., 2014b). In addition,

some studies cannot conclude a significant correlation between socio-demographic variables (e.g.,

gender, occupation, family members’ age, and education level) and energy-saving behavior

(Gatersleben et al., 2002; Olli et al., 2001); Therefore, the impacts of individual objective

characteristics on energy-saving behavior are diverse and no consensus has been reached for some

variables probably due to the differences of empirical contexts and data.

8
Table 1 Relevant literature related to energy-saving behavior analysis (Source: Author analysis)
Valid Influencing factors Considering mediating
Author
data Survey area Method Demograph Individual External Intention Behavior effects or indirect effects
(year)
sample subjective ic variables influencing factors
Yue et al. Nanjing and other 5 Regression characteristic factors
638 √ √ √ Energy-saving behavior
(2013) cities analysis
Wang et al. 50 villages among these Sustainable Consumption
1403 SEM √ √ √
(2014a) 35 selected regions Behavior
Webb et al. Online surveys in a
200 SEM √ √ Energy-saving behaviors
(2013) major Australian city
Gadenne et al. Three ‘environmentally MANOVA
218 √ √ √ Environmental behaviors
(2011) friendly’ firms analysis
Yang et al. Energy curtailment
526 Hefei , Anyang, Beijing SEM √ √
(2016) behavior
Hori et al. Regression Energy-saving behavior;
215 Dalian and other 4 cities √ √ √
(2013) analysis environmental behavior
Wang et al. Energy knowledge→energy-
276 Beijing SEM √ √ √ Energy-saving behavior
(2014b) saving behaviors via attitudes
Liu et al.
336 Suzhou SEM √ √ √ Green purchasing behavior
(2012)
Feng and Reisner. Regression Individual behaviors；
347 Shaanxi Province √ √
(2011) model public behaviors
Kang et al. Regression
280 Busan √ Energy-saving behavior
(2012) analysis
logit
Wang et al.
816 Beijing regression √ √ Electricity saving behavior
(2011)
analysis
Nordlund (2003) 2500 Sweden SEM √ Reducing personal car use

7
 ACCEPTED MANUSCRIPT

2.2.2 Individual subjective characteristic factors

Another group of research focusing on the relationship between individual subjective

characteristic factors and energy-saving behaviors is available (Hori et al., 2013; Kang et al.,

2012; Liu et al., 2012; Wang et al., 2014b; Wang et al., 2011; Webb et al., 2013; Yang et al., 2016;

Yue et al., 2013). For example, some researchers shed light on the impacts of environmental

concerns and energy knowledge, and found they were critical in affecting households’ adoption of

energy- saving behaviors (Han et al., 2013; Steg, 2008). Some studies on energy-saving behavior

showed that moral sense and responsibility were important variables that affect environmental

behavior (Liu et al., 2012; Wang et al., 2014b; Yang et al., 2016). Previous studies also have

demonstrated that individuals with strong confidence of control ability had a positive effect on

energy-saving behaviors (Baker et al., 2007; Cheng et al., 2006; Conner and Abraham, 2001).

Scott et al. (2000) observed that environmental responsibility and personal norms had a significant

impact on conscious energy-saving behavior. Webster and Zhang (2004) investigated the impacts

of social responsibility on ecological consumption and found that socially responsible people were

more likely to actively participate in community activities and purchase green-packaged products.

Han et al. (2013) found that the energy-saving behavior of Dutch households was related to their

knowledge about saving energy. The lack of energy-saving knowledge would impede household

energy saving (Harland et al., 2007). Arı and Yılmaz (2016) modeled family recycling behavior

and explored the effects of attitudes to recovery, recovery subjective norms and recovery

perceptional behavior- control factors were decisive in affecting recycling behaviors. Based on

these studies, it is necessary to take individual subjective factors into consideration. Finally, we

select degree of concern, sense of responsibility, values, energy knowledge and control view as the

representative for the individual subjective factors in this study.

2.3 External influencing factors

External influencing factors such as policies and regulations have also been investigated, but

by only few researchers (Dillman et al., 1983; Ertz et al., 2016; Gadenne et al., 2011; Stern, 1992;

Yue et al., 2013). Yue et al. (2013) argued that situational factors such as social norms had

significant positive moderating effects on energy-saving behavior via the influence of behavioral

ability and energy-saving awareness. Ertz et al. (2016) examined the direct impact of contextual

8
 ACCEPTED MANUSCRIPT

factors (perceived busyness, perceived wealth, perceived power) on energy-saving behaviors and

the indirect effects of attitude variables on these behaviors. Sardianou (2007) administered a

questionnaire survey on Greek household conditions and found that publicity and education factors

were positively related to household energy-saving behaviors. Wang et al. (2011) found that

economic benefits, policy and social norms had the positive influence on electricity saving

behaviors. Rui and Antunes (2011) studied the purchases of efficient energy and energy-efficient

equipment by European consumers, and found that quality was the most important factor

influencing their behavior. Wang et al. (2014b) suggested financial incentives could be adopted to

help promote environmental behaviors among Beijing’s residents. Schultz et al. (2007)

demonstrated that proper combination of descriptive and imperative norms were best able to

promote the implementation of energy-saving behavior. Given the significant impact of external

influencing factors on energy- saving behaviors, this study select quality of energy-saving

products, publicity and education, subjective norms, and policies and regulations as the

representative external influencing factors, while the infrastructural or social external factors that

might influence the energy-saving behaviors are not considered here.

Drawing on the existing literature, we find that different researchers select different

influencing factors to analyze energy-saving behaviors by considering different behavioral

mechanisms. However, most of them only selected one or several factors, which were not

comprehensive enough. Hence, in order to more fully consider the impact of various factors, this

study goes beyond the existing research by shedding light on the impacts of all four types of

influencing factors mentioned above on shaping household energy-saving behavior, including

individual subjective characteristic factors, demographic variables, external influencing factors

and energy-saving intention. Besides, not only were these selected influencing factors less

comprehensive in previous research, but they also had dissimilar influential paths. Majority of

research have investigated the direct influential relationship among individual characteristic

factors, external influencing factors, energy-saving intentions, and energy-saving behaviors. While

limited articles have explored the mediating effects (indirect effects), and this might overestimate

or underestimate the impact of factors. Therefore, this study will further discuss the mediating

effects in addition to the direct effects between influencing factors and energy-saving behaviors,

hoping to provide more in-depth insights on the mechanism

9
 ACCEPTED MANUSCRIPT

and policy suggestion for promoting energy-saving behaviors.

10
 ACCEPTED MANUSCRIPT

3. Methods
In order to clarify the direct and indirect relationships between main factors and household

energy-saving behaviors, a structural equation model (SEM) with latent variables is built

considering SEM cannot only deal with multiple dependent variables simultaneously, but also can

estimate the fit coherence of the whole model.

3.1 Structural equation model

SEM is a multivariate statistical model that can analyze the causal relationships between

variables based on the covariance matrix of variables; thus, it is also known as covariance

structure analysis (Reisinger et al., 1999). Structural Equation Model can establish a unified model

to consider multiple dependent variables simultaneously and explore the internal mechanisms

between influencing factors and energy-saving behaviors. Furthermore, it can also estimate not

only the direct effects between different factors but also the indirect effects so as to analyze the

mediating effects based on one model and can allow independent and dependent variables to

contain measurement errors. We conduct data analysis and simulation verification using AMOS

23.0 software.

SEM can be divided into two categories: the measurement model, and the structural model.

(1) Measurement model

Because the assumed hypothesis cannot be measured directly, the observed, recorded, or

measured variables are combined to construct latent variables using the measurement model. The

relationship between the latent variable and the observed variable is usually written as follows:

X = Λxξ + (1)

δ Y = Λyη (2)

+ε
where,  is an exogenous latent variable;  is an endogenous latent variable; X is a vector

composed by the exogenous observed variable;  is an error term of the exogenous index X; Y is a

vector composed by the endogenous observed variable;  is an error term of the endogenous index

Y; x indicates the relationship between exogenous observed variable and exogenous latent

variable; y indicates the relationship between endogenous observed variable and endogenous

latent variable.

11
 ACCEPTED MANUSCRIPT

(2) Structural model

The structural model depicts the causal relationships between latent variables, usually in a

linear form. The measurement model enables the latent variables of the structural model to be

estimated, albeit indirectly. The structural equation is usually written as follows:

η = Βη + Γξ + ζ (3)

where,  is an endogenous latent variable ,  is an exogenous latent variable ; B indicates the

relationship between endogenous latent variables;  indicates the influence of the exogenous

latent variable on the endogenous latent variable ;  is an error term of the structural equation.

3.2 Model framework and hypotheses

Energy-saving
intention
Age

H1
Gender
Individual subjective
characteristic factors
H3
Occupation

H5
Highest education
External influencing H2
factors
H4 Family member

Latent variable
Monthly household
Energy-saving
disposable income
behavior
Control variable

Figure 1 Simplified urban household energy-saving behavior model

Figure 1 shows a simplified model for explaining the hypothesized relationships among

individual subjective characteristic factors, external influencing factors, energy-saving intentions,

and energy-saving behaviors. Individual objective characteristic factors including age, gender,

occupation, highest education level for household members and monthly household disposable

income are added to the model as control variables. We assume that the individual subjective

characteristic factors and external influencing factors are directly influential to energy-saving

intentions (H1, H3) and energy-saving behaviors (H2, H4). Energy-saving intentions are assumed

to have significant influence on energy-saving behaviors (H5). The proposed model structure can

also investigate whether the individual and external factors indirectly affects energy-saving
12
 ACCEPTED MANUSCRIPT

behaviors via energy-saving intentions. In addition, the external influencing factors might affect

individual subjective characteristic factors and then indirectly influence energy-saving behaviors.

These assumptions can be traced in Figures 1 and 2 by the arrows in the path.

3.3 Conceptual variable definition

The main variables are given in Figure 2. Based on the review results, five factors were used

to represent the influences of individual subjective characteristics, including degree of concern,

sense of responsibility, environmental values, energy knowledge and control view (Alibeli, 2009;

Arcury et al., 1985; Bartkus et al., 1999; Gärling et al., 2003; Haron et al., 2005; Kim Y., 2005;

Scott and Willits, 1994). Specifically, degree of concern, which is similar to environmental

concerns, environmental attitudes, means how much individuals care about energy and related

issues (Steg, 2008). Sense of responsibility is defined as moral or duty sense when individuals

consume energy (Gärling et al., 2003).The concept of environmental values is drawing on the NEP

(New Environmental Paradigm) concept proposed by Dunlap (2002), which means the general and

fundamental attitudes of individuals towards the environment and environmental issues. Energy

knowledge refers to the knowledge and skills needed for rational choice of energy-saving

behaviors (Schahn and Holzer, 1990). Following Sun (2006) and Chen (2009), this study takes

control view as a research variable and explains it as the individual’s own idea about whether their

energy-saving behaviors can change the status quo and help tackle energy problems. The external

influencing factors can be regarded as situation factors affecting household energy consumption

behaviors. We selected four variables: quality of energy-saving products, publicity and education,

social norms, and policies and regulations (Carlsson-Kanyama et al., 2005; Egmond et al., 2005;

Lindén and Carlsson-Kanyama, 2002; Sardianou, 2007) to explore their impact on energy-saving

intentions and energy-saving behaviors, and their relationships with individual characteristics. The

full model is shown in Figure 2.

13
 ACCEPTED MANUSCRIPT

Sense of responsibility
Energy-saving
Degree of concern intention Age

H1
Values Gender
Individual subjective
characteristic factors
Energy knowledge H3
Occupation

Control view
H5
Highest education
Quality of energy- H2
saving product
Family member
Publicity and
education External influencing
factors H4 Monthly household
Energy-saving disposable income
Social norms behavior

Policies and
regulations Latent variable

Control variable

Figure 2 Urban household energy consumption behavior model

4. Survey and data

To fulfill this study, we designed and carried out a questionnaire survey (shown in Appendix

Table A.1). The questionnaire in this study consists of four parts: demographic factors, energy-

saving behaviors, energy-saving intentions and influencing factors for energy-saving behavior.

Drawing on questions used in previous research, we specifically added questions for this analysis.

4.1 Survey design and sampling

There are two subscales of energy-saving behaviors: the purchasing behaviors and habitual

energy-saving behaviors. The habitual energy-saving behaviors were assessed using questions

from the existing studies (Barr et al., 2005; Chen, 2009; Lindén and Klintman, 2003; Stern, 1992).

We select six habitual energy-saving behaviors which are always used (B1-B6 shown in Appendix

Table A.1), such as taking the power off when do not use the appliances, choosing a shower

instead of a bath, using daily public transport or on foot than using private cars, avoiding or

reducing usage of air conditioners and turning off the lights when leaving room. The purchasing

behaviors targeting more-efficient technology are measured by three frequently focused behaviors

in the previous studies (B7-B9 shown in Appendix Table A.1) (Chen, 2009; Gyberg and Palm,

2009; Lindén and Klintman, 2003; Rui and Antunes, 2011), including choosing energy-saving

lamps, preferring energy-efficient vehicle when buying a car, choosing simple packaging and
14
 ACCEPTED MANUSCRIPT

recyclable products

15
 ACCEPTED MANUSCRIPT

when purchasing similar products. However, B1-B4 are excluded because they do not pass the

validity test after the pilot survey, and B5-B9 are left. Energy-saving intention is designed based

on Chen (2013) and explained by four questions like willing to spend more money buying energy-

saving products, intending to pay more attention to low-carbon dynamic news, willing to use

driving skills to reduce fuel consumption and wanting to change the lifestyles of high-carbon

consumer behavior

Fifteen items measuring individual subjective characteristic factors are borrowed from Hsu

and Roth (1998), Chen (2009) and Tanner and Kast (2003), which can be grouped into five first-

order latent variables, including degree of concern, sense of responsibility, values, control view

and energy knowledge. Each factor is measured by three or four questions, for instance, degree of

concern is measured by if the respondents concerned about the energy problems reported by media,

whether felt worried or nervous when hearing or seeing issues about energy and paid little attention

to energy consumption or not. The measurements of external influencing factors are adapted from

Stern (2000), Sun (2006) and Chen (2009) that composed by nine items which can be grouped into

four first-order latent variables, including quality of energy-saving products, publicity and

education, social norms and policies and regulations. For example, social norms are measured by

questions such as whether the respondents’ behavior are affected by the surrounding persons like

families and friends or not, and whether the respondents care about other people’s adoption of

energy-saving behaviors or not. Policies and regulations are measured by two questions related to

whether respondents agree that the relevant policies and regulations have effect on energy-saving

behaviors or play an important role in promoting energy-saving behaviors.

In this study, energy-saving behavior, energy-saving intentions and influencing factors were

all measured with the Likert scale. We divided energy-saving behaviors into “ never do,” “

occasionally do,” “sometimes do,” “always do” or “ every time do” where the score reflected the

respondents’ frequency of behavior . Energy-saving intentions and influencing factors were

divided into “strongly disagree,” “do not agree,” “uncertain,” “agree” or “strongly agree”. We

assigned 1 to 5 points to these actions and attitudes. In principle, the higher the score is, the

stronger household intention to energy-saving becomes, and the more frequent energy-saving

behaviors are.

Jinan city, whose annual GDP reached 98.76 billion dollars in 2016, is selected as the empirical
16
 ACCEPTED MANUSCRIPT

context as it is a typical city under fast urbanization in China. It is the capital of Shandong province,

17
 ACCEPTED MANUSCRIPT

one of the biggest industrial provinces in China. A large increment of household energy

consumption can be foreseen if no interventions are implemented for Jinan. To ensure the

reliability and validity of the questionnaire, we selected 50 Jinan citizens to conduct pilot survey

in 2016. The questionnaire was then revised and a formal version formed. We conducted surveys

in commercial and residential neighborhoods, by randomly visiting the government departments,

schools, insurance companies, shopping malls and residential streets during October–December

2016. The candidate respondents in these areas were invited based on a convenient sampling

method and those who agreed to participate in the survey were asked to fill in the questionnaire by

a face-to-face interview. The sample is selected with gender and age proportions following the

distribution of national and regional population. A total of 350 interviews were completed and 297

valid responses were obtained.

4.2 Data

Table 2 shows the descriptive statistics of the survey data. It can be seen that the randomly

selected respondents have diverse age, educational level, occupation, family type, and household

income. The proportion of male is 51.2% and employment rate is 70.9%, which are consistent

with the statistical data of the Shandong province (50.8% and 67.3%, respectively) (Shandong

Statistical

Yearbook, 2015).
Table 2 Descriptive statistics of survey sample
Demographic variables Category Number of people Percentage
Gender Male 152 51.2
Female 145 48.8
Age Under 18 years old 3 1.0
18–30 years old 84 28.3
31–40 years old 104 35
41–50 years old 75 25.3
51–60 years old 24 8.1
More than 60 years old 7 2.4
Education background Junior high school and below 17 17.7
High school or secondary school 56 34.9
College 76 23.6
Bachelor degree 121 15.7
Bachelor degree or above 27 8.1
Family style Living alone 23 7.7
Married, no children or 56 18.9
not living with their
children
Two-generation family 169 56.9
Three- or four- generation family
18
 ACCEPTED MANUSCRIPT

49 16.5
Occupation Retirement and housewives 16 6.9
Student 13 5.4

19
 ACCEPTED MANUSCRIPT

Government department staff 26 8.8

General worker or service 79 28.1
person
Teacher 69 18.2
Engineer 3 1.0
General office staff 16 5.4
Manager 28 9.4
Others 47 16.8
Average monthly Under USD 302.24 17 5.7
household income USD 302.24–755.61 121 40.7
USD 755.61–1511.21 132 44.4
USD 1511.21–4533.64 19 6.4
Above USD 4533.64 8 2.7
Source: Author analysis.

5. Results and discussions

5.1. Reliability, validity test, and normality test

First, the study needs to check the reliability and validity of the measurement model because

they are the prerequisite for obtaining valid results for the structural model. The internal

consistency reliability is mainly verified by the Cronbach’s alpha coefficient. As shown in Table

3, the Cronbach’s alpha coefficients of the variables including degree of concern, sense of

responsibility, values, energy knowledge, control view, quality of energy-saving products,

publicity and education, social norms and policies are lying between 0.721~0.906, which are

greater than 0.7. This supports the reliability of latent variables. In this study, all of the involved

individual objective characteristic factors are not significant, probably because different

individuals with the same socio-demographic variable attributes might have diverse choices.

However, including them as the control variables increased the accuracy of the model. Therefore,

we still keep them in the model, but exclude them in the follow-up result analysis and relevant

figures and tables. The validity of the test model can be reflected by the average variance extracted

from the structure (AVE). The validity of the measurement model is confirmed because all AVE

values are greater than 0.5. Regarding the structural model, all the fitness indexes are qualified

according to the test criterion (see Table 4), indicating an acceptable model accuracy (Raykov and

Marcoulides, 2006). Finally, the model results are given in Figure 3.

20
 ACCEPTED MANUSCRIPT

Table 3 Model accuracy indexes of the measurement model

Latent variable Observed variable Cronbach’s  Factor
AVE
loading
Degree of concern X11 I am very concerned about the energy 0.844 0.93 0.654
problems reported by media.
X12 I feel worried or nervous when I hear or see 0.78
issues about energy.
X13 I usually pay little attention to energy 0.70
consumption and details of its use.
Sense of X21 I am willing to sacrifice personal interests to 0.838 0.91 0.653
responsibility save energy and for environmental protection.
X22 In order to save energy and protect 0.80
environment I do not hesitate to offend some
people.
X23 Energy issues are related to the whole society, 0.70
and everyone has the responsibility to save
energy.
Values X31 Human beings should respect nature and live 0.906 0.89 0.719
in harmony with nature.
X32 Nature and the environment have the same 0.90
value as human beings.
X33 We should not exploit new natural resources 0.72
to protect nature and the environment.
X34 Economic development is the most important, 0.87
and environmental issues can be considered
later.
Energy knowledge X41 How much do you know about the purchasing 0.747 0.88 0.618
energy-saving behaviors in questions 1–3?
X42 How much do you know about habitual 0.68
energy-saving behaviors in questions 4–9?
Control view X51 We can improve and solve some 0.845 0.89 0.675
environmental problems if we work hard.
X52 It will help to improve and solve some 0.90
environmental problems if we take some
actions.
X53 It is impossible for ordinary people to 0.65
improve and solve environmental problems.
Quality of energy- Y11 The public praise of energy-efficient products 0.721 0.83 0.575
saving product is an important factor in determining whether
I should buy them or not.
Y12 The quality of energy-saving products is an 0.68
important factor in determining whether I
should buy them or not.
Publicity and Y21 We should improve the strength of publicity 0.796 0.81 0.656
education and education, and guide more people to
implement energy-saving behaviors.
Y22 My energy-saving behavior will be affected 0.81
by the information I get from newspapers,
television and other media.
Social norms Y31 My energy-saving behavior will be affected 0.767 0.71 0.576
by my families, friends and teachers.
Y32 If people around me are engaging in energy- 0.64
saving behavior, I will also engage in more
energy-saving behaviors.
Y33 There are few people around me who are 0.83
concerned about energy conservation, and I
need more help and co-participation.
Policies and Y41 I undertake energy-saving behavior because 0.753 0.75 0.601
regulations of relevant policies and regulations.
Y42 Policies and regulations play an important 0.80
role in promoting and encouraging me to
engage in energy-saving behaviors.
Source: Author analysis

21
 ACCEPTED MANUSCRIPT

Table 4 Energy-saving behavior SEM fitting index

Index CMIN DF CMIN/DF GFI AGFI NFI IFI TLI CFI RMSEA
Model 1014.082 682 1.487 0.858 0.837 0.826 0.935 0.926 0.935 0.041
results
Standard ＜3 ＞0.8 ＞0.8 ＞0.8 ＞0.9 ＞0.8 ＞0.9 ＜0.05
Model Excellent Good Good Good Excellent Excellent Excellent Excellent
accuracy
Source: Author analysis.
X11 0.93***
0.78*** Degree of
X12
concern Z1 Z2 Z3 Z4 Quality of 0.69*** Y11
X13 0.70***
energy-saving
X21 0.91*** 0.73*** 0.75*** 0.90*** 0.79***
product 0.46*** Y12
Sense of 0.68***
X22 0.80***
responsibility Energy-saving 0.76***
0.70*** 0.28** intentions 0.39*** 0.66*** Y21
X23 Publicity and
0.64*** education
X31 0.89*** 0.66*** Y22
Individual 0.30*** 0.73***
X32 0.90*** External
subjective
Values 0.78*** 0.65*** influencing
X33 0.72*** characteristic
factors 0.52***
factors 0.51*** Y31
X34 0.87***
0.69*** Social norms 0.42*** Y32
X41 0.88***
Energy 0.34*** 0.26**
X42 0.68*** Energy-saving 0.68*** Y33
knowledge 0.70***
behaviors
X51 0.89*** 0.67***

0.90*** Control Policies and 0.56*** Y41

X52 0.56*** 0.54*** 0.62*** 0.53*** 0.61***
view regulations
X53 0.65*** 0.64*** Y42
B1 B2 B3 B4 B5

Latent variable

Observed variable

Figure 3 Final energy-saving behavior model results

Note: 1. *p < 0.05; **p < 0.01; ***p < 0.001
2. Results for the control variables are not shown here since they are all statistically insignificant.

5.2 Effects of individual subjective characteristic factors, external influencing factors and

energy-saving intentions

The proposed model in Figure 3 explains 59.7% of the variance of energy-saving behaviors

and 37.7% of the variance of energy-saving intentions. Table 5 shows the results of the finally

determined structural model. The results of the five hypotheses are statistically significant and H1,

H2, H3, H4, H5 are found to have positive effects on energy-saving intentions and/or energy-

saving behaviors. The paths and effects are given in Table 5. Individual subjective characteristic

factors and external influencing factors have a positive effect on energy-saving intentions and

energy- saving behaviors, and energy-saving intentions have a positive effect on energy-saving

behaviors. Table 6 shows the total effect of external influencing factors on energy-saving

22
 ACCEPTED MANUSCRIPT

behaviors is the

greatest. And we found the direct effect of individual subjective characteristic factors on energy-

23
 ACCEPTED MANUSCRIPT

saving behaviors are greater than that of external influencing factors, in contrast, the indirect effect

is far less than that of external influencing factors. Therefore, it can be seen that the effects of the

external influencing factors on energy-saving behaviors such as taking the power off when do not

use the appliances for a long time or choosing energy-saving lamps are mainly derived from

indirect

effects.
Table 5 Latent variable path coefficients
UnStd.
Hypothesis Path
coefficient
Z-value P Result

Energy -saving Individual

H1 intentions ← subjective 0.409 2.278 **
Positive
correlation
characteristic factor
Energy-saving Individual
H2 behavior Positive
← subjective
characteristic factor
0.673 3.783 ***
correlation
Energy -saving External Positive
H3 intentions 0.402 3.584 ***

H4
Energy-saving ← influencing
factors
0.367 2.783 **
correlation
Positive
behavior

H5
Energy-saving
behavior
← External
influencing
0.406 4.079 ***
correlation
Positive
factors correlation
← Energy -saving
intentions
Note: *p < 0.05; **p < 0.01; ***p < 0.001
Source: Author analysis.

Table 6 The effects of influencing factors

Action pathways Direct effect Indirect effect Total effect
Individual subjective characteristic factors 0.339*** 0.084*** 0.422***
→Energy-saving behaviors
External influencing factors
0.264*** 0.391*** 0.655***
→Energy-saving behaviors
Note: *p < 0.05; **p < 0.01; ***p < 0.001
Source: Author analysis.

5.3 Analysis of the mediating effect of external influencing factors

In order to further clarify and decompose the indirect effects of external influencing factors

on energy-saving behaviors, we build three independent mediation models for individual

subjective characteristic factors, external influencing factors, energy-saving intentions and energy-

saving behaviors (Figures 4-1, 4-2, 4-3). The direct, indirect, and total effects between variables

24
 ACCEPTED MANUSCRIPT

are obtained from the path analysis, which can provide a better understanding of the mediating

effects (Table 7-1, 7-2, 7-3). As mentioned above, the impact of external influencing factors on

energy-

25
 ACCEPTED MANUSCRIPT

saving behavior is mainly due to indirect effects, and the specific sources and size of the indirect

effects can be revealed by models 4-1 and 4-2. External influencing factors exert an indirect effect

on energy-saving behaviors through energy-saving intentions and individual subjective

characteristic factors. And the individual subjective factors further directly act on energy-saving

behavior and indirectly act on it through energy-saving intentions (see model 4-3). This implies

that the external influencing factors affect energy-saving behaviors via three channels, leading to

greater overall effects than the individual subjective characteristic factors. Moreover, we could

also find that energy-saving intention is an important factor that not only has positive direct

impact on energy- saving behaviors, but also plays an important role in forwarding the mediating

effects from individual subjective characteristic factors and external influencing factors to

energy-saving

Energy-saving Individual
intentions subjective
characteristic factors
0.58*** 0.38*** 0.66*** 0.45***

External influencing Energy-saving External influencing Energy-saving

factors 0.42*** behaviors 0.35***
factors behaviors

Note: *p < 0.05; **p < 0.01; ***p < 0.001 Note: *p < 0.05; **p < 0.01; ***p < 0.001

Figure 4-1 Mediation Model 1 Figure 4-2 Mediation Model 2

Energy-saving
intentions

0.53*** 0.37***

Individual
subjective Energy-saving
0.47*** behaviors
characteristic fa ctors

Note: *p < 0.05; **p < 0.01; ***p < 0.001

Figure 4-3 Mediation Model 3

26
 ACCEPTED MANUSCRIPT

Table 7-1 Analysis of the mediating effects between variables (Mediation Model 1)
Action pathways Direct Indirect Total
effect effect effect
External influencing factors
→Energy-saving behaviors 0.425*** 0.219*** 0.643***
Mediation
External influencing factors
Model 1 0.575*** --- 0.575***
→Energy-saving intentions
(Figure 4-1)
Energy-saving intentions --- 0.380***
→Energy-saving behaviors 0.380***
Note: *p < 0.05; **p < 0.01; ***p < 0.001

Table 7-2 Analysis of the mediating effects between variables (Mediation Model 2)
Action pathways Direct Indirect Total
effect effect effect

External influencing factors

Mediation 0.347*** 0.295*** 0.641***
→Energy-saving behaviors
Model 2 External influencing factors
(Figure 4-2) →Individual subjective characteristic factors 0.656*** --- 0.656***
Individual subjective characteristic factors
→Energy-saving behaviors 0.449*** --- 0.449***
Note: *p < 0.05; **p < 0.01; ***p < 0.001

Table 7-3 Analysis of the mediating effects between variables (Mediation Model 3)
Action pathways Direct Indirect Total
effect effect effect

Individual subjective characteristic factors

Mediation 0.470*** 0.200*** 0.670***
→Energy-saving behaviors
Model 3 Individual subjective characteristic factors
(Figure 4-3) 0.534*** --- 0.534***
→Energy-saving intentions
Energy-saving intentions
→Energy-saving behaviors 0.375*** --- 0.375***
Note: *p < 0.05; **p < 0.01; ***p < 0.001
Source: Author analysis

5.4 Effects of first-order latent variables

Table 8 The influence coefficients of the first-order latent variables
Latent variables Second-order latent variables Std. coefficient
Individual subjective Degree of concern 0.68
characteristic factors Sense of responsibility 0.64
Environmental values 0.78
Energy knowledge 0.69
Control view 0.67
External influencing Quality of energy-saving product 0.76
factors Publicity and education 0.73
Social norms 0.52
Policies and regulations 0.70
Source: Author analysis。

27
 ACCEPTED MANUSCRIPT

The specific effects of first-order latent variables on energy-saving behaviors can be

measured by analyzing how much they can explain individual subjective characteristic factors and

external influencing factors, and then affecting energy-saving intentions and energy-saving

behaviors. Table

8 lists the influence coefficients between first-order latent variables and second-order latent

variables including individual subjective characteristic factors and external influencing factors. It

can be seen that the coefficients of all first-order latent variables are significantly positive.

Combined with the coefficients between second-order variables, the effects of first-order latent

variables on energy-saving behaviors can be derived.

(1) Effects of different individual subjective characteristic factors

Concerning the first-order latent variables for shaping individual subjective characteristic

factors, values have the greatest impact. And the stronger household environmental values are, the

more possible for individuals or households to adopt energy-saving behaviors. Values towards

energy-saving and green consumption may affect households’ motivation to buy energy-saving

products or implement energy-saving behaviors such as avoiding or reducing usage of air

conditioners. This is consistent with some scholars’ findings that environmental and ethical

consumer values, especially those such as universalism and altruism, will actively promote moral

obligations and personal norms (Chen and Chang, 2012; Eze and Ndubisi, 2013). The second

factor is energy knowledge whose effect on explaining individual subjective characteristic factors

is 0.69, indicating that knowledge plays an important role on stimulating individuals to adopt

energy-saving behaviors like taking the power off when do not use the appliances for a long time

in daily life. This result is consistent with the case of Dutch household found by Han et al. (2013).

In other words, the lack of energy-saving knowledge will cause obstacles for energy saving

(Harland et al., 2007; Tanner and Kast, 2003; Vermeir and Verbeke, 2006). The control view has a

positive effect on energy-saving behaviors. Individuals with strong self-confidence and ability will

have a positive impact on intention and behaviors, thereby improving household self-confidence in

implementing energy-saving behaviors. The degree of concern (coefficient is 0.68) and the sense

of responsibility (0.64) have an important contribution for explaining individual subjective

characteristic factors and then affect the choice of energy-saving behaviors and products like

choosing simple packaging and recyclable products when purchasing similar products. The greater
28
 ACCEPTED MANUSCRIPT

the degree of concern and the sense of responsibility are, the greater the positive impacts on

energy-saving behaviors are. These

29
 ACCEPTED MANUSCRIPT

are consistent with existing findings (Gärling, 2003; Reiss and White, 2006; Wang et al., 2014).

Those suggest that improving household degree of concern and sense of responsibility for the

environment could encourage energy-saving behaviors.

(2) Effects of different external influencing factors

The external influencing factors related to market, society and government also have

significant impacts on energy-saving behaviors. Regarding the first-order latent variables that are

used to explain the external influencing factors, the contribution of quality of energy-saving

products is the largest (0.76), implying the quality of energy-saving products is more influential to

household decisions on purchasing energy-saving products compared to other first-order latent

variables like publicity and education, social norms, and policies and regulations. This result is

consistent with the findings that awareness of poor product quality is an important barrier to

green- buying activity (Smith et al., 2013). The explanation effect of social norms on the external

influencing factors is the least compared to other three variables, indicating that social norms are

less influential for promoting energy-saving behaviors. Publicity and education have an important

impact on the choice of energy-saving behaviors. Positive publicity and education through media,

governments, schools and other sectors could change individual’s attitudes and then guiding

households to implement more energy-saving behaviors such as using daily public transports or on

foot rather than private cars. The choice of energy-saving behaviors is closely related to relevant

government laws and regulations. In other words, the government can formulate policies and

regulations on energy conservation to encourage household energy-saving behaviors.

6. Conclusions and policy implications

6.1 Conclusions

This study provides more comprehensive insights on the mechanisms of how individual

subjective, external influencing factors and energy-saving intentions act on energy-saving

behaviors, by establishing a unified structural equation model to accommodate all elements

together. In addition to the direct effects, the mediating effects or indirect effects of the influential

factors on energy-saving behaviors are also discussed. This study takes Jinan city in Shandong

province as the empirical context and a questionnaire survey was conducted to support this
30
 ACCEPTED MANUSCRIPT

analysis. Several

31
 ACCEPTED MANUSCRIPT

conclusions are summarized as follows.

Individual subjective characteristic factors including degree of concern, sense of

responsibility, environmental values, energy knowledge and control view, as well as the external

influencing factors like quality of energy-saving products, publicity and education, social norms

and policies and regulations have significant impacts on shaping and changing energy-saving

behaviors. While individual objective characteristic factors are found statistically insignificant in

this empirical analysis as the control variables, probably due to diverse effects for different socio-

demographic groups. The total effects of external influencing factors on energy-saving behaviors

are greater than that of the individual subjective characteristic factors. The reason is that in

addition to the direct effect, the external influencing factors affect energy-saving behaviors via

energy-saving intentions and the individual subjective characteristic factors.

Values and energy knowledge belonging to individual subjective characteristic factors have

the greatest impacts on shaping household energy-saving behaviors. Quality of energy-saving

products is the most important factors among external influencing factors on affecting energy-

saving behaviors. Therefore, multifaceted policies are necessary for promoting household energy-

saving behaviors.

6.2 Policy implications

This study has important implications for household energy-saving behaviors. Specifically,

(1) Individual subjective characteristic factors are very important for changing energy-saving

behaviors, therefore, policies should be designed to help individuals establish environmental

values, sense of responsibility for energy saving and environmental protection behaviors. We can

learn from the existing advanced or successful measures so as to encourage people to engage in

energy-saving behaviors and buy energy-saving products, for example, the national emission

reduction policy1 implemented by Japan, energy-saving publicity and creative competitions

created by a domestic

1Japan implemented a practical activity to reduce emissions called “One person, One day and One kilogram,”
which advocated reducing the frequency of showering, using reusable shopping bags and other methods to achieve
energy- saving behaviors. By increasing the participation of the whole society and strengthening the influence of
32
 ACCEPTED MANUSCRIPT

degree of concern, values, sense of responsibilities and social norms, households can be guided to participate more
in energy saving and environmental protection activities.

33
 ACCEPTED MANUSCRIPT

school2, and a low-carbon generalized system of preferences (GSP) promotion platform 3 used in

Guangdong Provincial.

(2) It is found that external influencing factors not only directly impact energy-saving

behaviors, but also affect the individual subjective characteristic factors firstly, and then have a

further effect on energy-saving behaviors. Therefore, energy planners or policy makers could

influence energy-saving behaviors by awareness campaign targeting the general public and

meanwhile consciously guide households to improve their energy-saving concerns, values, energy

knowledge and control view to implement energy-saving behaviors. We can borrow the

experience from the developed countries, for example, Japan promoted “visualization,”4 California

launched a specific “showerhead standard” as a water-saving measure5, and the Singapore

government encouraged households to buy energy-efficient appliances and imposed punitive taxes

on high- energy cars6. In addition, some studies have shown that economic policies (e.g.,

government subsidy policies and preferential tax policies) can significantly affect household

energy-saving behaviors, and could also encourage households to buy energy efficient appliances

(Lee, 2009; Sardianou, 2007).

It is obvious that there are many kinds of measures can shape household energy-saving

behaviors. Government should guarantee the execution of the proposed policies and regulations.

At the same time, enterprises should strengthen their research abilities to produce more efficient

and high-quality energy-saving products.

2 Wuhou District, Sichuan Province held an energy saving and innovation competition for primary school students
in the region. The events included activities such as energy-saving creative painting, energy-saving gold ideas, an
energy-saving essay subject and energy-saving handwork. The games help the students form energy-saving values,
enhance their sense of responsibility and embed energy-saving knowledge.
3 In June 2016, Guangdong Province produced a low-carbon and the Internet model to promote green

consumption. The provincial government had not only developed a provincial-level carbon GSP promotion
platform, but also created websites and apps that conveyed low-carbon knowledge, energy-saving products and
technologies to help households increase their environmental values, degree of concern and energy-saving
knowledge.
4 By using various types of electric meters, gas meters and other kinds of meters, the energy consumption and CO
2
emissions of a single machine can be measured at any time, which would be convenient for people to observe their
daily energy-consuming activities and enhance their social norms gradually. They also implement an energy-
saving consumption labeling system, which was established according to different energy consumption levels and
play the role of policy standard.
5 California introduced two water-saving measures, one of which required the nation’s most water-efficient

showerhead to use less water than the current model. According to this policy, starting from July 2018, the
showerheads sold in California should be limited to no more than 1.8 gallons of water per minute. This indicates
that the government makes positive energy-saving policy and regulation for energy-saving products to help save
water. 6The Singapore government advocates that Singaporeans bought energy-saving appliances autonomously. In
addition, they claimed that government should impose punitive taxes on high-energy consumption vehicles and
34
 ACCEPTED MANUSCRIPT

meanwhile encourage people to use public transport to guide people to save energy.

35
 ACCEPTED MANUSCRIPT

6.3 Limitation and future research

This study takes a statistical approach to explore the interrelationships between the various

influencing factors and their effects on energy-saving behaviors. Although this study has made

some progress on considering more comprehensive mechanisms of shaping energy-saving

behavior, it has some limitations. For example, to improve the representativeness of the survey

data, larger sample would be better. The influencing factors can be expanded in the future study

by taking the influence of, for instance, energy price, infrastructure, and technology development

into account. In addition, further study can be considered based on the rules found in our study,

using a behavior-oriented approach to describe the formation process of energy-saving behavior,

then assess the concrete effect of policies on behavior changing.

Acknowledgements
The authors acknowledge financial support received through National Key R&D Program of
China (2016YFA0602603), and from the National Natural Science Foundation of China (Nos.
71603020, 71521002, and 71642004), and the support from the Joint Development Program of
Beijing Municipal Commission of Education.

References
Ajzen, I., 1991. The theory of planned behavior. Organizational Behavior & Human Decision Processes
50, 179-211.
Ajzen, I., 2002. Perceived Behavioral Control, Self‐Efficacy, Locus of Control, and the Theory of
Planned Behavior. Journal of Applied Social Psychology 32, 665-683.
Alibeli, M.A., 2009. Environmental Concern: A Cross National Analysis. Mansoura University Journal
of Agricultural Sciences.
Arcury, T.A., Johnson, T.P., Scollay, S.J., 1985. Ecological Worldview and Environmental Knowledge:
The “New Environmental Paradigm”. The Journal of Environmental Education 17, 35-40.
Arı, E., Yılmaz, V., 2016. A proposed structural model for housewives' recycling behavior: A case study
from Turkey. Ecological Economics 129, 132-142.
Baker, E.W., Al‐Gahtani, S.S., Hubona, G.S., 2007. The effects of gender and age on new technology
implementation in a developing country. Information Technology & People 20, 352-375.
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, 1425-1444.
Bartkus, K.R., Hartman, C.L., Howell, R.D., 1999. The measurement of consumer environmental
knowledge: Revisions and extensions. Journal of Social Behavior & Personality 14, 129-146.
Black, J.S., Stern, P.C., Elworth, J.T., 1985. Personal and contextual influences on househould energy
adaptations. Journal of Applied Psychology 70, 3-21.

36
 ACCEPTED MANUSCRIPT

Carlsson-Kanyama, A., Lindén, A.L., Eriksson, B., 2005. Residential energy behaviour: does
generation matter? International Journal of Consumer Studies 29, 239–253.
Chan, Y.K., Yam, E., 1995. Green movement in a newly industrializing area: A survey on the attitudes
and behaviour of the Hong Kong citizens. Journal of Community & Applied Social Psychology 5,
273–284.
Chen, 2009. Study on Urban Residential Energy Consumption Behavior. Dalian University of
Technology.
Chen, 2013. Study on the Low-carbon Consumption Behavior and its Influencing Factors of City
Residents. Jiangnan University.
Chen, Y.S., Chang, C.H., 2012. Enhance Green Purchase Intentions: The Roles of Green Perceived
Value, Green Perceived Risk, and Green Trust. Management Decision 50, 502-520.
Cheng, S., Lam, T., Hsu, C.H.C., 2006. Negative Word-of-Mouth Communication Intention: An
Application of the Theory of Planned Behavior. Journal of Hospitality & Tourism Research 30, 95-
116.
Conner, M., Abraham, C., 2001. Conscientiousness and the Theory of Planned Behavior: Toward a
more Complete Model of the Antecedents of Intentions and Behavior. Personality and Social
Psychology Bulletin 27, 1547-1561.
Curtis, F.A., Simpsonhousley, P., Drever, S., 1984. Household Energy Conservation. Energy Policy.
Dillman, D.A., Rosa, E.A., Dillman, J.J., 1983. Lifestyle and home energy conservation in the United
States: the poor accept lifestyle cutbacks while the wealthy invest in conservation Journal of
Economic Psychology 3, 299-315.
Dunlap, R.E., 2002. Environmental Sociology: A Personal Perspective on Its First Quarter Century.
Organization & Environment 15, 10-29.
Egmond, C., Jonkers, R., Kok, G., 2005. A strategy to encourage housing associations to invest in
energy conservation. Energy Policy 33, 2374-2384.
Ertz, M., Karakas, F., Sarigöllü, E., 2016. Exploring pro-environmental behaviors of consumers: An
analysis of contextual factors, attitude, and behaviors. Journal of Business Research 69, 3971-3980.
Eze, U.C., Ndubisi, N.O., 2013. Green Buyer Behavior: Evidence from Asia Consumers. Journal of
Asian and African Studies 48, 413-426.
Feng, W., Reisner, A., 2011. Factors influencing private and public environmental protection
behaviors: results from a survey of residents in Shaanxi, China. Journal of Environmental
Management 92, 429- 436.
Fishbein, M., Ajzen, I., 1975. Belief, Attitude, Intention and Behaviour: an introduction to theory and
research. Philosophy & Rhetoric 41, 842-844.
Gärling, T., Fujii, S., Gärling, A., Jakobsson, C., 2003. Moderating effects of social value orientation on
determinants of proenvironmental behavior intention. Journal of Environmental Psychology 23, 1-9.
Gadenne, D., Sharma, B., Kerr, D., Smith, T., 2011. The influence of consumers' environmental beliefs
and attitudes on energy saving behaviours. Energy Policy 39, 7684-7694.
Gatersleben, B., Steg, L., Vlek, C., 2002. Measurement and determinants of environmentally significant
consumer behavior. Environment & Behavior 34, 335-362.
Gyberg, P., Palm, J., 2009. Influencing households’ energy behaviour—how is this done and on what
premises? Energy Policy 37, 2807-2813.
Han, Q., Nieuwenhijsen, I., Vries, B.D., Blokhuis, E., Schaefer, W., 2013. Intervention strategy to
stimulate energy-saving behavior of local residents. Energy Policy 52, 706-715.

37
 ACCEPTED MANUSCRIPT

Harland, P., Staats, H., Wilke, H.A.M., 2007. Situational and personality factors as direct or personal
norm mediated predictors of pro-environmental behavior: Questions derived from norm-activation
theory. Basic and Applied Social Psychology 29, 323-334.
Haron, S.A., Paim, L., Yahaya, N., 2005. Towards sustainable consumption: an examination of
environmental knowledge among Malaysians. International Journal of Consumer Studies 29, 426-
436.
Hori, S., Kondo, K., Nogata, D., Han, B., 2013. The determinants of household energy-saving
behavior: Survey and comparison in five major Asian cities. Energy Policy 52, 354-362.
Hsu, S.J., Roth, R.E., 1998. An Assessment of Environmental Literacy and Analysis of Predictors of
Responsible Environmental Behaviour Held by Secondary Teachers in the Hualien Area of Taiwan.
Environmental Education Research 4, 229-249.
Jackson, T., 2005. Motivating Sustainable Consumption: A Review of Evidence on Consumer
Behaviour and Behavioural Change. Policy Studies Institute.
Kang, N.N., Cho, S.H., Kim, J.T., 2012. The energy-saving effects of apartment residents’ awareness
and behavior. Energy & Buildings 46, 112-122.
Kim Y., C.S., 2005. Antecedents of Green Purchase Behavior: An examination of collectivism,
environmental concern and PCE. Advances in Consumer Research 32, 592-599.
Lee, K., 2009. Gender differences in Hong Kong adolescent consumers' green purchasing behavior.
Journal of Consumer Marketing volume 26, 87-96(10).
Lindén, A.L., Carlsson-Kanyama, A., 2002. Voluntary agreements—a measure for energy-efficiency in
industry? Lessons from a Swedish programme. Energy Policy 30, 897-905.
Lindén, A.L., Klintman, M., 2003. The Formation of Green Identities - Consumers and Providers.
Liu, X., Wang, C., Shishime, T., Fujitsuka, T., 2012. Sustainable consumption: Green purchasing
behaviours of urban residents in China. Sustainable Development 20, 293–308.
Martinsson, J., Lundqvist, L.J., Sundström, A., 2011. Energy saving in Swedish households. The
(relative) importance of environmental attitudes. Energy Policy 39, 5182-5191.
Nordlund, A.M., Garvill, J., 2003. Effects of values, problem awareness, and personal norm on
willingness to reduce personal car use. Journal of Environmental Psychology 23, 339-347.
Olli, E., Grendstad, G., Wollebaek, D., 2001. Correlates of environmental behaviors: Bringing back
social context. Environment & Behavior 33, 181-208.
Poortinga, W., Steg, L., Vlek, C., Wiersma, G., 2003. Household preferences for energy-saving
measures: A conjoint analysis. Journal of Economic Psychology 24, 49-64.
Raaij, W.F.V., Verhallen, T.M.M., 1983. A behavioral model of residential energy use. Journal of
Economic Psychology 3, 39-63.
Raykov, T., Marcoulides, G.A., 2006. A First Course in Structural Equation Modeling (2nd Eds).
Reisinger, Y., Turner, L., Ryan, C., Faulkner, B., 1999. Structural equation modeling with Lisrel:
application in tourism. Tourism Management 20, 71–88.
Reiss, P.C., White, M.W., 2006. Evaluating Welfare with Nonlinear Prices. Nber Working Papers.
Rui, G., Antunes, D., 2011. Energy efficiency and appliance purchases in Europe: Consumer profiles
and choice determinants. Energy Policy 39, 7335-7346.
Sardianou, E., 2007. Estimating energy conservation patterns of Greek households. Energy Policy 35,
3778-3791.
Schahn, J., Holzer, E., 1990. Studies of Individual Environmental ConcernThe Role of Knowledge,
Gender, and Background Variables. Environment and Behavior 22, 767-786.

38
 ACCEPTED MANUSCRIPT

Schultz, P.W., Nolan, J.M., Cialdini, R.B., Goldstein, N.J., Griskevicius, V., 2007. The Constructive,
Destructive, and Reconstructive Power of Social Norms. Psychological Science 18, 429.
Schwepker, C.H., Cornwell, T.B., 1991. An Examination of Ecologically Concerned Consumers and
Their Intention to Purchase Ecologically Packaged Products. Journal of Public Policy & Marketing
10, 77-101.
Scott, Parker, Rowlands, 2000. Determinants of energy efficiency choices in the home: A case study of
Waterloo Region. Environments 28, 75-100.
Scott, D., Willits, F.K., 1994. Environmental Attitudes and Behavior: A Pennsylvania Survey.
Environment and Behavior 26, 239-260.
Smith, J.S., Andrews, D., Jr, J.J.C., Gleim, M.R., 2013. Against the Green: A Multi-method
Examination of the Barriers to Green Consumption. Journal of Retailing 89, 44–61.
Steg, L., 2008. Promoting household energy conservation Energy Policy 36, 4449-4453.
Stern, P.C., 1992. Psychological Dimensions of Global Environmental Change. Karger.
Stern, P.C., 2000. New Environmental Theories: Toward a Coherent Theory of Environmentally
Significant Behavior. Journal of Social Issues 56, 407-424.
Sun, Y., 2006. Study on Residents’ Environmental Behavior and Its Influencing Factors. Chinese
Journal of Management 9, 144-150.
Tanner, C., Kast, S.W., 2003. Promoting sustainable consumption: Determinants of green purchases by
Swiss consumers. Psychology & Marketing 20, 883–902.
Vermeir, I., Verbeke, W., 2006. Sustainable food consumption: exploring the consumer "attitude -
behavioral intention" gap. Journal of Agricultural and Environmental Ethics 19, 169-194.
Wang, P., Liu, Q., Qi, Y., 2014a. Factors influencing sustainable consumption behaviors: a survey of
the rural residents in China. Journal of Cleaner Production 63, 152-165.
Wang, Z., Zhang, B., Li, G., 2014b. Determinants of energy-saving behavioral intention among
residents in Beijing: Extending the theory of planned behavior. Journal of Renewable & Sustainable
Energy 6, 711-720.
Wang, Z., Zhang, B., Yin, J., Zhang, Y., 2011. Determinants and policy implications for household
electricity-saving behaviour: Evidence from Beijing, China. Energy Policy 39, 3550-3557.
Webb, D., Soutar, G.N., Mazzarol, T., Saldaris, P., 2013. Self-determination theory and consumer
behavioural change: Evidence fromahousehold energy-saving behaviour study. Journal of
Environmental Psychology 35, 59-66.
Webster, Cynthia, Zhang, L., 2004. Exploring the utility of the fishbein-ajzen model and social learning
theory in predicting unhealthy consumption behavior, AMA Winter Educators' Conference
Proceedings, p. 148.
Yang, S., Zhang, Y., Zhao, D., 2016. Who exhibits more energy-saving behavior in direct and indirect
ways in china? The role of psychological factors and socio-demographics. Energy Policy 93, 196-
205.
Yu, B., Zhang, J., Fujiwara, A., 2013. Evaluating the direct and indirect rebound effects in household
energy consumption behavior: A case study of Beijing. Energy Policy 57, 441-453.
Yue, T., Long, R., Chen, H., 2013. Factors influencing energy-saving behavior of urban households in
Jiangsu Province. Energy Policy 62, 665-675.

39
 ACCEPTED MANUSCRIPT

Appendix
Table A.1 Questions included in the questionnaire
Constructs Items
Degree of concern X11 I am very concerned about the energy problems reported by media.
X12 I feel worried or nervous when I hear or see issues about energy.
X13 I usually pay little attention to energy consumption and details of its use.
Sense of X21 I am willing to sacrifice personal interests to save energy and for environmental
responsibility protection.
X22 In order to save energy and protect environment I do not hesitate to offend some
people.
X23 Energy issues are related to the whole society, and everyone has the responsibility
to save energy.
Values X31 Human beings should respect nature and live in harmony with nature.
X32 Nature and the environment have the same value as human beings.
X33 We should not exploit new natural resources to protect nature and the environment.
X34 Economic development is the most important, and environmental issues can be
considered later.
Energy X41 How much do you know about the purchasing energy-saving behavior in questions
knowledge 7–9?
X42 How much do you know about habitual energy-saving behaviors in questions 1–6?
Control view X51 We can improve and solve some environmental problems if we work hard.
X52 It will help to improve and solve some environmental problems if we take
some actions.
X53 It is impossible for ordinary people to improve and solve environmental problems.
Quality of Y11 The public praise of energy-efficient products is an important factor in determining
energy- saving whether I should buy them or not.
product Y12 The quality of energy-saving products is an important factor in determining
whether I should buy them or not.
Y21 We should improve the strength of publicity and education, and guide more people
Publicity and to implement energy-saving behaviors.
education Y22 My energy-saving behavior will be affected by the information I get from
newspapers, television and other media.
Y31 My energy-saving behavior will be affected by my families, friends and teachers.
Social norms Y32 If people around me are engaging in energy-saving behavior, I will also engage in
more energy-saving behaviors.
Y33 There are few people around me who are concerned about energy conservation,
and I need more help and co-participation.
Y41 I undertake energy-saving behavior because of relevant policies and regulations.
Policies and
regulations Y42 Policies and regulations play an important role in promoting and encouraging me
to engage in energy-saving behaviors.
Energy-saving Z1 I am willing to spend more money buying energy-saving products.
intention Z2 I would like to pay more attention to low-carbon dynamic news, and actively
respond to low-carbon consumption.
Z3 When I driving I am willing to use driving skills to reduce fuel consumption.
Z4 I am willing to sacrifice some of the convenience to change the lifestyle of high-
carbon consumer behaviors.
Energy-saving B1 When you do not use the appliances for a long time, you will take the power off to
behavior reduce the TV, air conditioners, computers, drinking fountains, microwave ovens
and other household appliances standby power consumption.
B2 You will choose a shower instead of a bath.
B3 Use daily public transports, such as buses, bikes or on foot.
B4 You will save water when bathing in public baths
B5 When you leave the room, you will turn off the lights
B6 When you are in a public workplace, you will avoid or reduce using air
conditioners
B7 When purchasing lamps, you will choose energy-saving lamps.
B8 When you buy or plan to buy a car, you will prefer a small displacement car (Fuel
saving vehicle).
B9 When purchasing similar products, you prefer choosing simple packaging and
recyclable products.

Source: Author analysis

40
 ACCEPTED MANUSCRIPT

Note: B1-B4 are excluded because they do not pass the validity test after the pilot survey, and B5-B9
are left.

41
 ACCEPTED MANUSCRIPT

Highlights

 The mechanism of factors affecting household energy-saving behavior are explored.

 A structural equation model is built to analyze the influencing routes and effects.
 External influencing factors are the most crucial factors to energy-saving behaviors.
 Individual subjective factors play evident role on shaping energy-saving behaviors.
 Direct and intermediary effects are found which have special policy implications.