Gap Latihan 3

environmental science & policy 39 (2014) 77–94
Available online at www.sciencedirect.com
ScienceDirect
journal homepage: www.elsevier.com/locate/envsci
Examining the determinants of public

environmental concern: Evidence from national
public surveys
Xinsheng Liu *,1, Arnold Vedlitz, Liu Shi

Institute for Science, Technology and Public Policy, The Bush School of Government and Public Service Texas A&M
University, College Station, TX 77845-4340 USA
article info abstract
Article history: Early research showed that citizens’ environmental concern in the United States was linked
Received 21 October 2013 to three individual-level factors: socio-demographic variables, political orientations, and
Received in revised form personal beliefs or worldviews about human-nature relations. Given many changes in the
21 February 2014 American society over the last several decades, one important, yet unanswered question is
Accepted 22 February 2014 whether these factors still drive public environmental concern in the United States today,
Available online 22 March 2014 and if so, to what extent. This study, drawing from extant theoretical and empirical studies,
aims to reinvestigate the determinants of citizens’ environmental concern by employing
Keywords: three national public surveys conducted in 2004, 2007, and 2013. Our data analyses confirm
Environmental concern and expand the findings of previous research on the significance and importance of political
Environmental policy ideology, fundamental beliefs about human-nature relations, and certain socioeconomic
New Ecological Paradigm (NEP) factors such as gender and race in explaining citizens’ environmental concern. More
Political ideology specifically, political liberals, people with higher New Ecological Paradigm values, females,
socio-demographics and Non-Whites tend to be more concerned about environmental problems than their
United States counterparts are. Our data analyses also reveal some interesting findings when compared to
many previous studies: first, our data indicate a positive relationship between age and
environmental concern, suggesting that older people in the United States are more con-
cerned about the environment than younger adults; second, unlike most past research
showing a positive Education-Environmental Concern relationship, our study suggests that
education level seems to have little effect in explaining citizens’ environmental concern
measured in this study. Key implications for environmental policymaking and recommen-
dations for future research are discussed in the conclusion.
# 2014 Elsevier Ltd. All rights reserved.
1. Introduction and Rationale conditions to make sound public policy and promote citizens’
engagement in proenvironment behaviors. Over the last
Understanding individual citizens’ environmental concern several decades, a large body of literature has examined
(EC) and various underlying factors is one of the key necessary various driving forces of individual citizens’ EC in the United
* Corresponding author. Tel.: +1 979 845 4120; fax: +1 979 862 8856.
E-mail address: x.liu@tamu.edu (X. Liu).
1
Xinsheng Liu is also a visiting professor of Peking University-Fudan University-Jilin University Co-Innovation Center for State
Governance in China.
http://dx.doi.org/10.1016/j.envsci.2014.02.006
1462-9011/# 2014 Elsevier Ltd. All rights reserved.
78 environmental science & policy 39 (2014) 77–94
States and around the world. Numerous studies have found 2002, p. 485). This definition correctly indicates that EC is a
that among many other factors, the level of citizens’ EC are very broad concept covering a wide range of phenomena with
frequently associated with individuals’ socio-demographic multiple aspects and multiple dimensions (see also Xiao and
characteristics (for examples, see Van Liere and Dunlap, 1980; Dunlap, 2007; Alibeli and White, 2011). To avoid ambiguity in
Ester and Van Der Meer, 1982; Stern et al., 1983; Arcury and concept definition and variations or errors in variable
Christianson, 1990; Jones and Dunlap, 1992; Howell and Laska, measurement, Dunlap and Jones suggest that researchers
1992; Guth et al., 1995; Eckberg and Blocker, 1996; Dietz et al., ‘‘need to think clearly at the outset about what aspects or
2007; McCright and Dunlap, 2011b; Clements, 2012), political facets of environmental concern they want to measure, and
orientations (see for examples, Dunlap, 1975; Van Liere and then carefully conceptualize them prior to attempting to
Dunlap, 1981; Stern et al., 1983; Howell and Laska, 1992; Gamba measure them’’ (Dunlap and Jones, 2002, p. 515).
and Oskamp, 1994; Dietz et al., 1998; Dunlap et al., 2001; The extant EC research is also limited in several other
Olofsson and Ohman, 2006; Dietz et al., 2007; Wood and aspects. For instance, most existing studies rely on a snapshot
Vedlitz, 2007; Clements, 2012), and variations in people’s analysis derived from one single public survey, and the
ecological value and belief systems (Dunlap and Van Liere, findings typically cannot be compared and cross-validated due
1978, 1984; Stern et al., 1995; Dalton et al., 1999; Pierce et al., to the absence of comparable surveys/datasets across time.
1999; Dunlap et al., 2000; Clements, 2012). Moreover, many previous studies primarily focused on
The extant studies have greatly contributed to our citizens’ concern about certain environmental issues that
understanding of what factors affect citizens’ EC. However, were highly salient several decades ago (e.g., air and water
there are some gaps in the EC literature. First, most of the best pollution). In recent years, some new environmental issues
empirical findings from previous EC studies were derived from such as global warming and climate change have emerged and
public survey data collected in the 1970s, 1980s, and 1990s, and started to draw public attention. While it is important to
very few nation-wide empirical studies have been conducted continue examining the driving forces of citizens’ concern
to re-examine the bases of individuals’ EC by employing the about air and water pollutions, it is also important to
polling data collected in recent years. Over the last several investigate how these driving factors affect the levels of
decades, many aspects of American society, culture, and citizens’ concern about the newly emerged environmental
demographic composition have changed. It is reasonable to issues.
ask if the social, political, and belief-system bases of In this study, we attempt to fill some of the gaps in the EC
individuals’ EC have also changed and if the key empirical literature and make a few contributions. First, we employ US
findings from those studies are still valid in the new national public survey data to re-examine several key
millennium. While these important questions cannot be individual-level sources of citizens’ EC identified in previous
answered without new empirical data, some studies imply EC literature and to analyze whether and how these sources
that such changes might have occurred. For instance, nearly affect people’s concern about the environment in recent years.
all previous studies found that younger people or younger In doing so, we focus on three possible sources that were
generations were more concerned about the environment frequently examined in the EC literature: individuals’ socio-
than older people or older cohorts. However, a recent study by demographic characteristics (such as gender, age, race,
Twenge et al. (2012) finds significant generational differences. education, and income), respondents’ political ideology, and
They show that the younger Millennials, or Me Generation their beliefs and values about the fundamental relations
(born after 1982), are more concerned about materialistic between human and the environment. Second, following the
values and less interested in caring about others and saving suggestions of Dunlap and Jones (2002) and others, we attempt
the environment when compared to Baby Boomers (born 1946– to focus on one specific aspect of the broad environmental
1961) and to Generation Xers (born 1962–1981) at the same age. concern concept: the perceived seriousness of certain environ-
In another study on citizens’ support for climate change mental problems facing the United States. This focus allows us
policies, Dietz et al. (2007) find that younger people express to reduce the ambiguity in concept definition; it also allows
less support for climate change mitigation policies than older future comparisons between our study and other research
adults do. These studies, in addition to other recent studies on using similar EC definition and measurement. Third, we
historical shifts and declining trends in youth’s environmental attempt to measure people’s EC for both long-existing
attitudes and behaviors (for example, see Wray-Lake et al., environmental issues (e.g., pollution) and newly emerged
2010), suggest that there may be some underlying changes in environmental problems (e.g., global warming and climate
society and some of the findings derived from past EC studies change); we also consider three levels of abstractness–
probably need to be re-examined. concreteness when measuring citizens’ concern for different
Second, many scholars note that there are varying, mixed, environmental objects. Fourth, we use datasets derived from
and somewhat inconsistent or conflicting research findings in three nation-wide public surveys, conducted in 2004, 2007, and
each area of the EC literature. One possible cause of this, as 2013, respectively. Using data from these surveys conducted in
pointed out by Dunlap and Jones (2002) and Klineberg et al. the past decade allows us to cross-examine the findings from
(1998), is that the concept environmental concern has been multiple representative samples and to make relatively more
defined unclearly and measured differently across studies. As generalized observations over time.
a complex concept, EC refers to ‘‘the degree to which people In the following sections, we first review the main findings
are aware of problems regarding the environment and support of existing EC studies on three individual-level driving forces:
efforts to solve them and/or indicate a willingness to sociodemographics, political ideology, and belief/value sys-
contribute personally to their solution’’ (Dunlap and Jones, tem about human-nature relations. We then introduce our
environmental science & policy 39 (2014) 77–94 79
survey methodologies and describe the questionnaire items regarding why women were more environmentally concerned
and variable measurements. Next, we present the main than men were. While most studies found that women were
results and findings of our analyses. In the final section, we more environmentally concerned than men were, a few
discuss key implications and recommend some strategies for studies did yield inconsistent evidence about the gender
future research. differences. One study found no significant gender difference
between women and men on EC (Arcury and Christianson,
1993). Another study revealed that while women incorporate
2. Determinants of Individual Environmental more environmentally friendly behaviors into their domestic
Concern routine activities (e.g., saving energy around the house,
recycling reusable items), they were no more likely than
In the past several decades, many scholars have examined men to be engaged in society-oriented environmental activ-
public concern about environmental problems in the United ism (Tindall et al., 2003; see also Hunter et al., 2004 for a cross-
States (e.g., Dunlap, 1991; Dunlap and Scarce, 1991; Jones and cultural comparative analysis involving 22 nations). Given
Dunlap, 1992; Kanagy et al., 1994; Kempton et al., 1995; Yeager these inconsistent findings, Zelezny et al. (2000) surveyed a
et al., 2011) and around the world (e.g., Abramson and decade of research from 1988 to 1998 on this topic and
Inglehart, 1995; Inglehart, 1995; Krosnick, 2000; Clements, concluded that the majority of previous studies reported that
2012). Among many other possible sources, three factors at the women were significantly more concerned about the envi-
individual level—sociodemographics, political orientations ronment relative to men.
and certain fundamental beliefs, values, or worldviews about Previous studies have probed the effect of age on EC. Many
human-nature relations, are frequently identified to be studies conducted in the 1980s and 1990s showed that young
associated with citizens’ EC, though the findings on the people were more concerned about environmental issues than
direction and strength of these correlates vary across studies. older adults were (Dietz et al., 1998; Kanagy et al., 1994; Jones
and Dunlap, 1992; Howell and Laska, 1992; Arcury and
2.1. Sociodemographic factors and environmental concern Christianson, 1990; Van Liere and Dunlap, 1980). In one study,
Jones and Dunlap (1992) demonstrated that, among 11 socio-
Research examining the relationship between sociodemo- demographic and sociopolitical factors, age was the best
graphic factors and citizens’ EC mainly focuses on identifying predictor of EC, with younger people being more environmen-
the types of individuals who tend to be most concerned about tally concerned relative to older people. Van Liere and Dunlap
environmental problems. This approach typically focuses on (1980, p. 183) attributed this age-EC correlation to ‘‘‘young
standard individual socio-demographic characteristics such people are less integrated into the American economic system
as gender, age, education, income, and race, and a large or, more generally, the dominant social order,’’ and to
number of studies have contributed to this topic.2 ‘‘solutions to environmental problems often are viewed as
For the gender-EC relationship, most empirical studies threatening the existing social order, . . . [so] it is logical to
revealed modest to strong gender differences between women expect youth to support environmental reform and accept
and men, suggesting that women tend to have higher levels of proenvironmental ideologies more readily than their elders.’’
EC than men do (McStay and Dunlap, 1983; Bord and O’Connor, There are also other theories to explain the age effect on EC for
1997; Blocker and Eckberg, 1987, 1989; Lai and Tao, 2003; Biel the youth: different generational change and experience
and Nilsson, 2005). There are three interrelated theories on the compared to the older generations (Kanagy et al., 1994; Mohai
gender gap. The first theory argues that the gender gap in EC and Twight, 1987; Van Liere and Dunlap, 1980) and increased
may be caused by different expectations for females and education about environmental issues among the youth
males during parenthood and socialization processes (Hamil- (Howell and Laska, 1992).
ton, 1985; McStay and Dunlap, 1983; Blocker and Eckberg, 1989; Previous studies have also hypothesized and investigated
Zelezny et al., 2000). The second theory contends that the the relationship between social class status (as measured by
gender gap may result from gender-based divisions in the income, education, and occupational prestige in Van Liere and
labor market and home (Blocker and Eckberg, 1997). The third Dunlap, 1980) and the level of EC. One hypothesis, rooted in
theory attributes the gender gap to different value formation Maslow’s theory of hierarchy of human needs (Maslow, 1970),
processes between women and men (Stern et al., 1995). In their expects that those with higher social class should be more
widely cited article, Davidson and Freudenburg (1996) provid- concerned about environmental problems than those with
ed an extensive review and examined several hypotheses lower social class; they have satisfied their basic material
needs and thus become more concerned about higher level
2
It is important to note that some EC studies also include other needs such as the need for better environmental quality (see
sociodemographic status variables such as homeownership (e.g., Van Liere and Dunlap, 1980). Another explanation indicates
Whittaker et al., 2005), employment (e.g., Van Liere and Dunlap, that people with higher income and higher education tend to
1980; Jones and Dunlap, 1992; Whittaker et al,, 2005), religion assimilate environmental information more quickly, have
identity (e.g., White, 1967; Hand and Van Liere, 1984; Naess, more knowledge about environmental problems, and possess
1989; Whitney, 1993; Guth et al., 1995; Eckberg and Blocker,
more skills in analyzing the true conditions of environmental
1996), and place of residence (e.g., Tremblay and Dunlap, 1978;
issues. Furthermore, there are studies suggesting that high-
Fransson and Garling, 1999; Arcury and Christianson, 1990; Howell
and Laska, 1992). Our review here is restricted only to the litera- income and well-educated people are more likely to have post-
ture that examined effects of gender, age, education, income, and materialist views emphasizing quality of life and environ-
race on citizens’ EC. mental sustainability instead of economic growth and
material possessions (Inglehart, 1995; Van Liere and Dunlap, would expand government regulations and impose more costs
1980). to business, researchers predicted that the cleavages in
This ‘higher social class–higher EC’ hypothesis received political ideology would differentiate liberals from conserva-
mixed evidence. In a number of empirical studies, income and tives in their concern and support for environmental issues
occupation were found as weak, nonstatistically significant, or (Constantiti and Hanff, 1972; Dunlap and Gale, 1972; Dunlap,
inconsistent predictors of individuals’ EC, but educational 1975). Empirical evidence from many EC studies revealed that
level was slightly or moderately associated with EC, with the people with liberal political ideology were more concerned
well-educated displaying more concern about environmental about environmental issues than were people with conserva-
problems than their counterparts (Dietz et al., 1998; Kanagy tive ideology (Dunlap, 1975; Van Liere and Dunlap, 1981;
et al., 1994; Jones and Dunlap1992; Howell and Laska1992; Samdahl and Robertson, 1989; Hine and Gifford, 1991; Howell
Arcury and Christianson, 1990; Van Liere and Dunlap, 1980). and Laska, 1992; Gamba and Oskamp, 1994; Dietz et al., 1998;
Race is also a commonly used sociodemographic variable in Whittaker et al., 2005). Several recent studies using national
explaining individual citizens’ EC. The race-EC relationship public survey data also demonstrate that citizens with
varies across studies. Some of the early EC literature indicated stronger liberal political ideology tend to be more concerned
a ‘‘concern gap’’ between Blacks and Whites, suggesting that about the environment (Wood and Vedlitz, 2007; Hinich et al.,
Black/Non-Whites were less likely to perceive environmental 2013).
risks and express EC than Whites (for reviews of racial EC gap
literature, see Taylor, 1989; Mohai, 1990; Jones, 1998, 2007). 2.3. New Ecological Paradigm and environmental concern
However, this view has been challenged by other studies
(Jones, 1998, 2002; Mohai, 1990, 2003; Mohai and Bryant, 1998), Another important factor that many scholars have suggested
showing that the racial gap between Blacks and Whites does to incorporate into the analysis of public EC is individuals’
not exist. On the contrary, the best available evidence appears general value orientations, beliefs or norms (see, for examples,
to suggest that Non-Whites/Blacks tend to be more concerned Stern et al., 1983; Stern et al., 1995), particularly individuals’
for the environment than Whites are. For instance, one study ecological worldviews and fundamental beliefs about human-
by Jones and Dunlap (1992) found that Whites were less environment relationships.
concerned about the environment than Non-Whites were. In the late 1970s and early 1980s, Dunlap and Van Liere first
Similarly, Dietz et al. (1998) found evidence, though limited, developed a set of constructs to measure different ecological
that Blacks were more proenvironmental than Whites were, worldviews and beliefs about human-nature relationships
and Hohai and Bryant (1998) found that Blacks expressed (Dunlap and Van Liere, 1978, 1984). Their constructs, originally
substantially greater concern than did Whites, particularly known as the New Environmental Paradigm Scale, and then
over local environmental problems. More recent studies revised to the New Ecological Paradigm (NEP) Scale (Dunlap
further demonstrate that Blacks/Non-White Latinos have et al., 2000), were the most widely used measure for ecological
become more environmentally sensitive and more concerned worldviews in environmental studies. For Dunlap and his
than their white counterparts (Whittaker et al., 2005; Jones and colleagues, the NEP Scale accesses individuals’ ‘‘primitive
Rainey, 2006). This seems reasonable since minorities are beliefs’’ about the environment and the nature-humanity
often more vulnerable to the negative effects of environmental relationship (Dunlap et al., 2000). Stern et al. (1995) indicate
stresses. that NEP measures ‘‘folk ecology’’—a set of generalized beliefs
about human-nature interactions, which exert influences on
2.2. Political ideology and environmental concern individuals’ perceptions, attitudes, and behavioral intentions
regarding specific environmental problems. Similarly, Cluck’s
In addition to the standard sociodemographic variables, EC study (1998) concludes that NEP represents a general ‘envi-
scholars also frequently incorporate certain political orienta- ronmental worldview’ dimension with subdimensions such as
tion factors such as party affiliation and political ideology into ‘balance of nature’ and ‘humans over nature.’ For these
their studies of citizens’ EC. scholars, a higher score on the NEP Scale reflects a stronger
Party affiliation was believed to influence the level of proecological worldview and stronger belief about human-
individuals’ EC. While there were some studies showing the nature interdependency, and that orientation is thought, and
influence of party affiliation on people’s concern over issues demonstrated in various studies, to lead to higher concern
like climate change (see, for example, McCright and Dunlap, about environmental issues and stronger proenvironmental
2011a), many past studies indicated that EC was not attitudes and behaviors (Stern et al., 1995; Dalton et al., 1999;
significantly associated with individuals’ political party Pierce et al., 1999).
identification (Dillman and Christensen, 1972; Buttel and The NEP Scale originally developed by Dunlap et al. (2000)
Flinn, 1974). includes 15 questions, but many researchers have used a
Political ideology was also believed to affect EC. Political revised or shortened version of NEP in order to keep their
ideology generally reflects a set of basic beliefs and values surveys to a user-friendly length (for a detailed review, see
about the relationship between government and business. Hawcroft and Milfont, 2010). La Trobe and Acott (2000)
Political liberals, compared to political conservatives, believe maintain that abbreviated version of the NEP Scale can be
that government should play a bigger and more active role in used to measure fundamental ecological worldviews if
regulating business behaviors in order to reduce socioeco- individual factors or dimensions of the NEP Scale form
nomic inequality among citizens and protect the collective coherent factors with adequate reliability. Cordano et al.
well-being of the society. Because environmental protection (2003) evaluate the predictive validity of the original and
revised versions of the NEP Scale as well as some abbreviated 21, 2013 and March 14, 2013. KnowledgePanel is built on a
NEP-derived scales, and conclude that all of these scales representative, random sample of US households using a
(original, revised, or abbreviated NEP measures) are valid process called ‘‘address-based sampling’’ (ABS) that involves
measures in explaining a significant amount of the variance in probability-based sampling of addresses from the US Postal
individuals’ intention to engage in proenvironmental behav- Service’s Computerized Delivery Sequence (CDSF).3 The
ior. As indicated below in our surveys and data analyses, we survey included 1,313 complete interviews.
employ an abbreviated version of NEP Scale to measure All three surveys began with a set of general questions on
individual citizens’ ecological worldviews and their beliefs how individual Americans were concerned about a variety of
about the biophysical environment and human-ecological public issues, including a few questions about environmental
relationships. issues (see Appendix A for the question items and wording).
The orderings of the issue-concern questions were random-
ized for each respondent. The surveys also included the
3. Survey Method and Variable Measurement respondents’ sociodemographic information, self-assessed
political ideology, as well as their beliefs about the human-
According to extant research discussed above, three individ- nature relationships as measured by the New Ecological
ual-level factors—socio-demographic characteristics, politi- Paradigm questions (see Appendix A).
cal ideological orientations, and ecological beliefs about
human-nature relationships—typically influence EC. Our 3.2. Dependent variable
review also indicates that the specific effects of these factors
vary somewhat across the large bodies of literature in each The dependent variable in this study is the level of
area. respondents’ EC. Since EC is a broad concept, it has been
While the best available empirical evidence from past argued that researchers should carefully design the survey
research seems to suggest that women, younger people, more questionnaires that will be used to measure citizens’ EC
educated, Non-Whites, people with stronger liberal political (Dunlap and Jones, 2002; see also Van Liere and Dunlap, 1981;
ideology, and those with stronger beliefs in the New Ecological Klineberg et al., 1998; Bergnguer et al., 2005). Following the
Paradigm (NEP) are more environmentally concerned com- suggestion by Dunlap and Jones (2002) that research should
pared to their counterparts, it is important to note that some of clarify what specific facets of EC are being examined, we define
the evidence are varying, mixed, and inconsistent across EC in this study as the degree to which a person perceives the
studies. Because of this, many findings in the EC studies are seriousness of environmental problems (see also Ester and
rather inconclusive and thereby require continuous re- Van Der Meer, 1982 for a similar definition).
examinations with new designs and new data. As discussed In environmental studies, individuals’ concern over envi-
above in Introduction and Rationale, there are several other ronmental problems covers a wide array of objects, ranging
gaps or limitations in the existing EC research. To address from the most abstract concepts (e.g., the ecosystem, the
some of the limitations, in this study we use data obtained natural environment), to moderately abstract issues (such
from three national public surveys of random-sampled adults as watershed conservation, deforestation, biodiversity, global
in the United States to re-examine these individual-level warming, and climate change), to very specific environmental
determinants of citizens’ EC. Drawing from past research, we problems (such as recycling, drinking water quality, nuclear
are particularly interested in examining the effects of the three waste disposal, toxic chemical regulation)4. In addition,
individual-level determinants on respondents’ EC. some environmental problems (such as air and water
pollution) have been there for decades while others (such
3.1. Survey method and procedure as global warming/climate change or hydraulic fracturing/
fracking) have just emerged as new environmental problems
Our research team at Texas A&M University designed the in recent years. Conceptually, there are different objects in
three US nation-wide public surveys. The 2004 and 2007 terms of abstractness–concreteness levels and different
surveys were conducted as part of a larger research project temporal connotations of people’s concern about the envi-
supported by the National Oceanic and Atmospheric Admin- ronment.
istration (NOAA). Both surveys, conducted through a Comput-
er-Assisted Telephone Interview (CATI) system and 3
Using ABS provides a statistically valid representation of the
implemented by Public Policy Research Institute at Texas
US populations as well as many difficult to survey populations
A&M University, obtained national random samples provided that include cell phone only households, African Americans, and
by Survey Sampling International (SSI). The 2004 survey was Latinos.
conducted from July 13 to August 10, 2004, and the 2007 survey 4
For a review on measuring concern over various environmen-
was in the field from April 3 to July 18, 2007. The 2004 and 2007 tal objects at different levels, see Klineberg et al., 1998, p. 736–738.
surveys included 1,093 and 833 complete interviews, respec- In their research, Klineberg et al. (1998) argue that one of the major
tively. limitations and possible causes for inconsistent findings in past
studies was that EC was measured differently across studies:
The 2013 survey was part of another larger research project
some studies used only abstract level environmental objects in
supported by both Texas Sea Grant through NOAA and Texas
questionnaire items while others measured EC exclusively based
A&M University. It was administered by professional polling on specific environmental questions. See Dunlap and Jones (2002)
provider GFK Custom Research, LLC (formerly Knowledge for an extensive discussion of various definitions in the EC liter-
Networks) using their ‘‘KnowledgePanel’’ between February ature.
To measure respondents’ EC, we incorporated a few and EC, we have no prior expectation about whether or how
question items into our first 2004 survey to capture how household income may affect people’s EC.
concerned respondents are about: (a) ‘‘the Environment’’—an Respondents’ ethnicity was self-identified as either Black
object at the most abstract level, (b) ‘‘Global Warming and (or African American), White, Asian, American Indian, or
Climate Change’’—an object at the moderately abstract level Native Hawaiian and other Pacific Islander), and we recoded
and a newly emerged environmental issue, and (c) ‘‘Pollu- and dichotomized respondents’ ethnicity. In our analysis, Race
tion’’—an object at a very specific level and a long existing White is a dummy variable with 1 representing for White and 0
environmental issue. We repeated the same questions on ‘‘the representing for all other races. Although there are still
Environment’’ and ‘‘Global warming and climate change’’ in theoretical and empirical debates about the relationship
the 2007 and 2013 surveys, though we did not ask respondents between race and EC, the best available findings suggest that
to rate their concern about ‘‘Pollution’’ in the 2007 survey. The Whites tend to be less concerned for the environment than
manners of the questions asked about the respondents’ Non-Whites are (e.g., Jones and Dunlap, 1992). Thus, we expect
concern in all the surveys were similar (see Appendix A for a negative relationship between Race White and EC.
question wording). For Political Ideology, respondents were asked to identify
Based on the surveys, eight separate measures for themselves on a Liberal-Conservative scale as one of the
respondents’ EC were constructed. Three were from the following categories that best described their political views:
2004 survey, including Concern about the Environment 2004 strongly liberal (=1), liberal (=2), slightly liberal (=3), middle of
(Concern_En_2004), Concern about Global Warming and Climate the road (=4), slightly conservative (=5), conservative (=6),
Change 2004 (Concern_GWCC_2004), and Concern about Pollution strongly conservative (=7), and other (excluded from this
2004 (Concern_Pollution_2004); two were from the 2007 study). Almost all previous studies indicated that political
survey, including Concern about the Environment 2007 (Con- conservatives are less environmentally concerned than
cern_En_2007) and Concern about Global Warming and Climate political liberals are. Thus, it is expected in our analysis that
Change 2007 (Concern_GWCC_2007); and three from the 2013 the self-assessed, left-right ideological variable is negatively
survey, including Concern about the Environment 2013 (Con- associated with people’s EC.
cern_En_2013), Concern about Global Warming and Climate Change For the NEP Scale, we used an abbreviated version of the
2013 (Concern_GWCC_2013), and Concern about Pollution 2013 Revised NEP Scale developed and tested by Dunlap, Liere,
(Concern_Pollution_2013). These measures capture respon- Mertig, and Jones (2000). The Revised NEP Scale consisted of
dents’ concern about both a long-existing pollution issue and fifteen Likert-scale question items; our abbreviated version of
the new global warming and climate change problem. These the Revised NEP Scale used seven of the fifteen questions,
constructs also reflect individuals’ EC better when compared asking respondents to indicate their opinion (strongly
to the measures derived from either very abstract or very agree = 4, agree = 3, disagree = 2, or strongly disagree = 1)
specific environmental objects questions. about the relationships between human beings and the
physical environment (See Appendix A for the NEP questions).
3.3. Independent variables We used the mean score of the seven items to measure
respondents’ NEP Scale. Cronbach’s Alpha scores (0.755 for the
Independent variables in this study are standard sociodemo- 2004 survey, 0.837 for the 2007 survey, and 0.853 for the 2013
graphic variables (gender, age, education, household income, survey) derived from reliability tests indicate that these seven
and race), respondents’ self-assessed political ideology, and question items are highly inter-correlated and internally
their NEP Scale. The Appendix A contains a list of questions we consistent, suggesting that they primarily measure the same
used in our survey to measure these variables. latent construct—NEP Scale. A recent study (Amburgey and
Gender_Female was a dummy variable with 1 representing Thoman, 2012) recommends that research with the NEP
female and 0 male. Based on the theories of different gender employs confirmatory factor analysis (CFA), and if CFA is
socialization processes and the empirical findings from past unavailable, treats the NEP Scale as five correlated subscales.
research, we expect that women would show more EC For the purpose of our study, we believe that it is appropriate
compared to men. to use the mean score of the seven items to measure
Age was respondents’ actual age. Because most existing respondents’ ecological beliefs. Based on previous studies,
studies found an inverse relationship between age and EC, we we expect that the respondents’ NEP Scales be positively
hypothesize that age is negatively associated with the level of related to the measures of their EC.
people’s EC. Descriptive statistics of all the dependent and independent
Respondents’ Education was reported at six levels, ranging variables are listed in Appendix B. Bivariate correlation matrix
from the lowest level of education (elementary to some high tables are attached in Appendix C.
school = 1) through the highest category (postgraduate
degree = 6). Based on the best evidence from past studies,
the level of education is expected to be positively related with 4. Data Analyses and Results
EC.
Household Income recorded respondents’ annual household We used OLS regressions to conduct our data analysis. Each of
income in the previous year. There were eleven different levels the eight EC measures was regressed on respondents’
with 1 being < $10,000 and 11 being > $100,000). Since previ- sociodemographic factors (Gender-Female, Age, Education,
ous theories and empirical findings did not provide clear Household Income, and Race White), their Self-Assessed Political
evidence about the relationship between household income Ideology, and their NEP Scale. To discern further the specific
Table 1 – Determinants of Concern about the Environment 2004.

Concern_En_2004 MODEL 1 MODEL 2 MODEL 3
Sociodemographics Sociodemographics + Sociodemographics +
Political Ideology Political Ideology + NEP
Constant 7.684*** 9.388*** 2.480***
(0.385) (0.414) (0.703)
Sociodemographic Variables
Gender _Female 0.683*** 0.468*** 0.370**
(0.156) (0.153) (0.148)
Age -0.003 -0.000 0.002
(0.005) (0.005) (0.005)
Education -0.023 -0.104* -0.058
(0.060) (0.059) (0.059)
Household Income -0.011 0.008 0.010
(0.027) (0.026) (0.025)
Race _White -0.457** -0.253 -0.384*
(0.215) (0.208) (0.202)
Political Ideology: Self-Assessed
Liberal-Conservative -0.406*** -0.201***
(0.044) (0.045)
New Ecological Paradigm
NEP Scale 2.139***
(0.176)
Adj. R Square 0.025 0.119 0.270
F 5.24 19.01 38.75
Sig. F 0.000*** 0.000*** 0.000***
N 825 803 715
* p < 0.1; ** p < 0.05; *** p < 0.01. Cell entries are unstandardized regression coefficients. Standard errors are reported in parentheses.
effects of the independent variables, three regression models Scale to Model 2 to include all three blocks of independent
were performed for each EC measure: Model 1 included variables.
respondents’ sociodemographic variables only, while Model 2 Tables 1–8 report the regression results on each of the eight
added political ideology to Model 1, and Model 3 added NEP dependent variables respectively. Since survey data tend to
Table 2 – Determinants of Concern about Global Warming and Climate Change 2004.
Concern_GWCC_2004 MODEL 1 MODEL 2 MODEL 3
Constant 6.916*** 9.514*** 0.732
(0.480) (0.503) (0.853)
Socio-Demographic Variables
Gender _Female 0.918*** 0.584*** 0.317*
(0.195) (0.186) (0.179)
Age 0.001 0.003 0.008
(0.006) (0.006) (0.006)
Education -0.062 -0.170** -0.108
(0.075) (0.072) (0.071)
(0.033) (0.031) (0.031)
Race _White -0.959*** -0.609** -0.725***
(0.267) (0.253) (0.245)
(0.053) (0.055)
NEP Scale 2.687***
(0.215)
Adj. R Square 0.039 0.174 0.321
F 7.71 29.05 49.11
Sig. F 0.000*** 0.000*** 0.000***
N 822 800 714
Table 3 – Determinants of Concern for Pollution 2004.

Concern_Pollution_2004 MODEL 1 MODEL 2 MODEL 3
Constant 7.555*** 9.014*** 2.614***
(0.382) (0.415) (0.701)
Gender _Female 0.651*** 0.506*** 0.359**
(0.155) (0.153) (0.147)
Age -0.002 0.000 0.003
(0.005) (0.005) (0.005)
Education -0.043 -0.114* -0.124**
(0.060) (0.059) (0.058)
(0.027) (0.026) (0.025)
Race_ White -0.413* -0.278 -0.436**
(0.213) (0.209) (0.202)
(0.044) (0.045)
NEP Scale 2.049***
(0.176)
Adj. R Square 0.022 0.092 0.244
F 4.76 14.62 33.85
Sig. F 0.000*** 0.000*** 0.000***
N 825 803 715
have heteroscedasticity problem (i.e., the variance of the error A large chi-square from the Breusch–Pagan/Cook–Weisberg
term is not constant), we used the Breusch–Pagan/Cook– test indicated that heteroscedasticity was present in all
Weisberg test to detect whether the error variances are a regressions. Thus, Huber–White robust estimation of the
multiplicative function of one or more variables in the model. standard errors was used to address the heteroscedasticity

Political Ideology Political Ideology +NEP
Constant 8.276*** 10.156*** 3.674***
(0.498) (0.501) (0.711)
Gender_Female 1.021*** 0.769*** 0.779***
(0.179) (0.169) (0.161)
Age 0.012* 0.017*** 0.022***
(0.006) (0.006) (0.006)
Education -0.161** -0.163** -0.045
(0.070) (0.066) (0.063)
Household Income -0.044 -0.045 -0.028
(0.030) (0.028) (0.027)
Race _White -0.727*** -0.637*** -0.791***
(0.253) (0.239) (0.226)
(0.047) (0.050)
NEP Scale 1.639***
(0.143)
Adj. R Square 0.085 0.227 0.381
F 12.63 30.73 48.95
Sig. F 0.000*** 0.000*** 0.000***
N 629 610 547
Constant 7.034*** 10.148*** 0.807
(0.673) (0.646) (0.894)
Gender_Female 1.207*** 0.780*** 0.661***
(0.242) (0.217) (0.202)
Age 0.016* 0.026*** 0.030***
(0.008) (0.008) (0.007)
Education -0.085 -0.103 0.029
(0.095) (0.085) (0.080)
Household Income -0.074* 0.074** -0.051
(0.041) (0.036) (0.034)
Race_White -0.729** -0.591* -0.710**
(0.341) (0.307) (0.284)
(0.061) (0.063)
NEP Scale 2.469***
(0.180)
Adj. R Square 0.059 0.282 0.467
F 8.88 40.73 69.43
Sig. F 0.000*** 0.000*** 0.000***
N 628 609 547
problem (Huber, 1967; White, 1980). In addition, we also variables, and in no regression was the variance inflation
conducted diagnostic tests for multicolinearity. There was factor (VIF) found to be above 1.32 (VIFs above 3 indicate a
little indication of multicolinearity among the independent multicolinearity issue. See Fox, 1991).

Constant 5.800*** 7.489*** .857*
(0.316) (0.353) (0.502)
Gender _Female 0.436*** 0.386*** 0.277**
(0.143) (0.140) (0.127)
Age 0.018*** 0.022*** 0.022***
(0.004) (0.004) (0.004)
Education 0.097 0.013 0.059
(0.065) (0.064) (0.058)
(0.025) (0.024) (0.022)
Race _White -0.407** -0.181 -0.268*
(0.173) (0.169) (0.154)
(0.044) (0.043)
NEP Scale 1.596***
(0.091)
Adj. R Square 0.023 0.092 0.282
F 6.968 22.343 68.647
Sig. F 0.000*** 0.000*** 0.000***
N 1296 1263 1207
Constant 5.631*** 8.616*** 0.029
(0.376) (0.404) (0.557)
Gender _Female 0.676*** 0.510*** 0.356**
(0.170) (0.160) (0.141)
Age 0.009* 0.015*** 0.014***
(0.0056) (0.005) (0.004)
Education 0.103 -0.035 0.019
(0.077) (0.073) (0.065)
(0.030) (0.028) (0.025)
Race _White -0.910*** -0.551*** -0.615***
(0.206) (0.193) (0.171)
(0.051) (0.048)
NEP Scale 2.089***
(0.100)
Adj. R Square 0.025 0.169 0.392
F 75891 44.264 113.079
Sig. F 0.000*** 0.000*** 0.000***
N 1309 1273 1216
Similar to the findings from previous studies, our regres- statistically significant, and the signs are all in the positive
sion results demonstrate that Gender Female is a strong and direction, strongly indicating that women are more environ-
consistent predictor of individuals’ EC. The coefficients for mentally concerned than men—whether the concern was
Gender Female in all 24 regression models in Tables 1–8 are about the general environment or specific environmental
Table 8 – Determinants of Concern for Pollution 2013.

Concern_Pollution_2004 MODEL 1 MODEL 2 MODEL 3
Constant 5.540*** 6.985*** 1.210**
(0.382) (0.343) (0.498)
Gender _Female 0.588*** 0.547*** 0.435***
(0.137) (0.136) (0.126)
Age 0.022*** 0.025*** 0.024***
(0.005) (0.004) (0.004)
Education 0.090 0.014 0.044
(0.062) (0.062) (0.058)
(0.024) (0.024) (0.022)
Race_ White -0.529*** -0.352** -0.414***
(0.213) (0.164) (0.153)
(0.043) (0.043)
NEP Scale 1.410***
(0.090)
Adj. R Square 0.039 0.089 0.244
F 11.618 21.530 59.904
Sig. F 0.000*** 0.000*** 0.000***
N 1303 1267 1204
problems, or whether the concern was about the long existing regression results in Tables 1–5 demonstrate that political
pollution issue or the new challenge of global warming and conservatives are less concerned about the environment
climate change. (across all the five EC measures) than self-reported political
In many previous studies, Age was evidently a predictor of liberals are.
individuals’ EC with an inverse relationship showing that As expected, NEP Scale is also an important and significant
younger people were more concerned about the environ- predictor in explaining individual citizens’ EC. As shown in our
ment. However, in our data analysis derived from the 2004 regression results, NEP is positively contributing to higher EC;
survey (Tables 1–3), Age does not appear to be a statistically the coefficients are consistently significant with all five types
significant factor in determining any of the three EC of EC measured in Tables 1–5.
measures. Age does show statistically significant influence Consistent with previous findings that the standard socio-
in the regression models using the more recent data from the demographic-economic variables usually only explain no
2007 survey (Tables 4 and 5) and the 2013 survey (Tables 6–8), more than 15% of the variations in publics’ EC (Van Liere
but surprisingly it shows a positive relationship with and Dunlap, 1980; Klineberg et al., 1998), our Model 1
respondents’ EC, suggesting that older citizens are more regressions using only sociodemographics as predictors in
concerned about the environment than younger people. This Tables 1–8 also show a limited explanatory power: all the
positive relationship between Age and EC is contrary to the adjusted R squares are lower than 15%, ranging from 2.3% to
findings from many earlier studies, but seems to be consistent 8.5%. However, in the expanded regression models (i.e.,
with the recent studies discussed earlier on possible genera- Models 2 and 3) where Political Ideology and NEP variables
tional changes and declining trends of youth’s attitude are added as additional explanatory variables, the explanatory
toward environmental issues. We shall further discuss this power is improved noticeably—all the adjusted R squares are
in the concluding section. significantly higher than that of the earlier models. For
In contrast to many past studies, the effect of Education is instance, for Concern about the Environment-2004 in Table
not consistent in our regression analyses. An overwhelming 1, the adjusted R square increases noticeably from Model 1
majority of the regression models shows that Education exerts using only sociodemographics as predictors (0.025) to Model 2
little or no influence on citizens’ EC. Among all 24 regressions by adding Political Ideology (0.115) to Model 3 by further adding
in Tables 1–8, 18 models indicate that Education is not a both Political Ideology and NEP (0.270). For both Concern about
significant factor. For those regressions with Education Global Warming and Climate Change-2004 and Concern about
showing statistically significant, the effect of Education Pollution-2004, the adjusted R squares are also much higher
generally disappears when the NEP variable is added to the when Political Ideology variable is added in Model 2 in Tables 2
regression models (see, for examples, the results in Table 1, and 3. The adjusted R squares are further improved when both
Table 2, and Table 4). The bivariate correlations in Appendix C Political Ideology and NEP variables are included in the Model 3
also confirm that there are basically little or no correlations regressions.
between Education and the three measures of citizens’ EC in This pattern of increased explanatory power from Model 1
our surveys. This finding, suggesting that Education is perhaps to Model 2 to Model 3 is also true for the regression analyses
no longer a strong predictor in explaining people’s EC, is using the 2007 survey data: for Concern about the Environ-
unexpected as many previous studies conducted in the 1970s ment-2007 in Table 4, the adjusted R squares are improved
through 1990s suggested that a higher level of education from 0.085 in Model 1 to 0.227 in Model 2 and further to 0.381 in
contributes to a higher level of EC. Model 3; for Concern about Global Warming and Climate
With regard to other sociodemographics, Household Change-2007 in Table 5, the adjusted R squares are improved
Income is not a significant factor in predicting individuals’ even more significantly from 0.059 in Model 1 to 0.282 in Model
EC in nearly all regression models throughout all the 2, and 0.467 in Model 3. Similar patterns of increased adjusted
three surveys. Although Household Income is statistically R squares can also be found in Tables 6–8 using the 2013 survey
significant in Models 1 and 2 of Table 5 in explaining data. The substantial improvement of the adjusted R squares
respondents’ concern about global warming and climate in Model 2 and Model 3 indicate that political ideology and NEP
change, its effect disappears when the NEP Scale is added to values are far stronger factors affecting individuals’ EC
the regression (i.e., Model 3 in Table 5). Like many previous compared to the influence of socio-demographic character-
studies, our results generally demonstrate that the rela- istics.
tionship between household income and EC is unstable and
insignificant.
Consistent with previous studies, our findings show strong 5. Conclusion, Policy Implications, and Future
evidence that people’s EC is affected by Race, with Whites Research
being less likely to be environmentally concerned than Non-
Whites are. In Tables 1–8, Race White is negatively associated In their research article published two decades ago, Jones and
with all the EC measures in all the regression models, and Dunlap (1992) used data from General Social Surveys (1973–
among the 24 regression models, 21 models show that Race 1990) and attempted to answer an important research
White is statistically significant. question in environmental attitude study: have the social
Throughout all the models with the Liberal-Conservative bases of individual citizens’ EC changed over time? While
position included, the self-assessed, left-right political ideolo- Jones and Dunlap (1992, p. 28) concluded that the social
gy exerts a statistically significant influence in determining determinants of EC (such as gender, age, and education) had
people’s EC. Consistent with the finding of past research, our remained remarkably and consistently stable from 1973 to
1990, it is necessary to re-examine the bases and determinants with and supportive of other recent studies that we discussed
of citizens’ EC today, as over the last several decades, there earlier in Introduction and Rationale (Twenge et al., 2012;
have been many drastic changes and significant fluctuations Campbell, 2012; Dietz et al., 2007; Wray-Lake et al., 2010; see
in America’s social, economic, political, cultural, and physi- also Clements, 2012). Possible reasons for why the age–EC
cal–environmental conditions. relationship may have shifted are beyond this research, but
Drawing on extant research, we attempted to reinvestigate several scholars provide some explanations for this possible
how the key sociodemographic variables, political ideology change. Twenge et al. (2012) suggest that the rising impor-
positions, and belief/value systems affect individual citizens’ tance of extrinsic values (such as money) in American society
EC in the new millennium. We employed public opinion may have contributed to declining EC in younger generations.
datasets obtained from three nation-wide public surveys Louv (2005) argues that this shift may be caused by what he
conducted in 2004, 2007, and 2013, respectively, measured called Nature-Deficit Disorder in young Americans, which
individual respondents’ EC in different ways (including their refers to a widespread and growing trend that young people
concerns about the general environment and more specific tend to spend more time indoors, and thereby disrupting their
environmental problems as well as their concerns about long ability to connect to nature and the environment. Addition-
existing and newly emerged environmental issues), and used ally, another potential contributing factor to this change may
regression models to analyze the effects of the three possible be the pervasive Technofix worldview in young people, which
sources of EC. is espoused and enhanced by the growing technological
Consistent with findings from past research, our data competency in younger generations. This technological
show that individual political ideology positions and their optimism suggests that technology can and will solve all
fundamental beliefs about human-nature relationships are problems, including environmental problems, and thereby
substantially significant factors that affect citizens’ concern reducing youth’s perception of various environmental risks
about the environment—political conservatives are less (Kilbourne et al., 2001). Of course, these are just theoretical
concerned about the environment than political liberals, speculations, all of which are subject to empirical tests in
and people with higher NEP values tend to have higher EC. future research.
The effects of political ideology and NEP values are In addition to the surprising findings on age effect, we
consistently found across all three surveys conducted in also find new and interesting effects of education on
different times of recent years, across all different types of citizens’ EC. In contrast to most previous findings that a
EC measures, and when controlling socio-demographic higher level of education promotes a higher level of EC, we
variables in the regressions. Moreover, as shown in our find evidence showing that education has no effect at all in
regression results, political ideology and NEP values dem- most of our regression models, and little or very weak effect
onstrate much stronger powers in explaining people’s EC in the few remaining models. Based on the evidence, we
when compared to the overall influence of individuals’ think that, overall, education is probably no longer a
socio-demographic variables. statistically significant predictor in explaining people’s EC
While the influence is not as substantial as that of political today.
ideology or NEP values, some socio-demographic variables, In recent years, the theory and practice of environmental
such as gender and race, show statistically significant effects policy have been increasingly concerned with placing the
on respondents’ EC. Similar to previous findings, our data citizen at the center of policy making. On the one hand, public
analyses demonstrate that there is still a gender gap, opinions and individual citizens are one of the primary
indicating that men are less environmentally concerned than drivers and agents for policy making and policy change. A
women are on all types of EC measured in this study. Race clear and updated understanding of the composition and
affects EC in most of the regression models included in this underlying factors of citizens’ EC will inform policy makers
study, showing that Whites are less environmentally con- and help formulate fair and efficient policy responses to the
cerned than Non-Whites are. These findings are supportive concerns expressed by citizens. On the other hand, individual
and similar to ‘the white men effect’ found in other recent citizens are also the target of environmental policies. A
studies on American attitudes toward environmental issues thorough empirical examination of the factors that influence
(see McCright and Dunlap, 2011a). individual citizens’ EC is a prerequisite to making sound and
When compared with previous studies, our data analyses feasible environmental policies that motivate citizens to
reveal some new interesting, and even surprising findings, change their behaviors and promote citizens’ engagement in
about the effects of age on citizens’ EC. While most previous pro-environment activities. While our study generally con-
studies indicated that younger people were more concerned tributes to the accumulation of our knowledge and provides
about the environment, our regression analyses using the 2004 some nuanced and updated information about American
survey data show that age is not a predictor for people’s EC. citizens’ EC and its underlying factors, the results of our data
More surprisingly, contrary to previous findings, age positively analysis also have some specific implications for our
affects the two EC measures in our regression models using environmental policy, particularly for our environmental
the 2007 and 2013 surveys. This positive age effect in the 2007 education policy.
and 2013 surveys is even more statistically significant when The future of the nation and the planet greatly relies on a
political ideology and NEP variables are added to the well-educated population to be wise stewards of the environ-
regression models (see Tables 4, 5 and 7). ment and natural resources. Environmental education pro-
While our finding on the positive age–EC relationship is grams and policies can help our citizens understand the
limited to our 2007 and 2013 data, it seems to be consistent complex connections between economic prosperity and
environmental quality and balance their short-term consid- and expands the findings of previous research on the
erations and long-term concerns. While environmental significance of political ideology and NEP values as important
education must involve everyone, our findings support the determinants of individual citizens’ EC; our study also
use of segmentation strategy based on the differences in suggests that there still seems to be a sociodemographic base
individual citizens’ EC predicted by various socialdemo- for individuals’ EC, as gender and race continue to play
graphics. The gender gap and race gap found in our study important roles in affecting people’s environmental attitudes
suggest that future environmental education policies and and possibly their behaviors. Future research should continue
programs should allot more resources and direct more efforts to explore the influences of gender, race, political orientations,
toward the groups of people who are less environmentally and fundamental values/worldviews on citizens’ EC.
concerned, such as white and male citizens. An important Second, although some of our findings on how age and
objective of this strategy would be to engineer a shift in their education affect people’s EC are different from, and even
environmental attitudes from less concerned to more con- contrary to, many findings from previous research, we do
cerned. not wish to imply that the data, analyses, and interpreta-
In addition to the white-male citizens, another target group tions offered here are definitive or conclusive. However, we
of environmental education policy is young Americans. do believe these findings are informative and open to
Although our finding on the positive age–EC relationship is further empirical tests. Of course, one possible source of the
only limited to the 2007 and 2013 data, it raises alarms for our differences between our study and the past research might
policymakers and educators. No matter what causes the be that we examined only one aspect of people’s EC in this
younger Americans to be less concerned about the environ- study—that is, their perceived seriousness of environmental
ment than the older citizens, future environmental education problems, while many past studies used a relatively broader
policies and programs should pay closer attention to the definition of citizens’ EC. Future research should continue to
young generations, educating them with proenvironmental explore how education and age affect people’s EC, which
values, raising their environmental awareness, and engaging should be defined as clearly as possible, and provide
them with more nature-oriented and outdoor activities. This more empirical evidence about the relationships between
focus on the young generations is particularly important given them.
the trends of growing materialism, increasing disconnection Third, our study cannot (and was not designed to) explain
from nature, and spreading ‘technofix’ perspective among why age has reversed its role in shaping people’s EC, but we
young Americans. Raising environmentally conscious gen- find several potential explanations plausible. There are
erations will help ensure that tomorrow’s problem solvers and possibly broader trends in American society that might have
decision makers are well prepared for the challenges they will caused this shift. Future research should pay closer attention
likely face. to the theoretical conjectures discussed above and explore
We also recommend that our environmental education other possible explanations.
policy focus on basic environmental values and core beliefs. Fourth, individual citizens’ EC may grow out of and
Our study indicates that citizens’ EC is split along the reflect as much the circumstances of one’s community as
dimension of political ideology and basic value-belief one’s personal circumstances. We included many personal
systems. Social scholars have observed that the contempo- level variables to explain citizens’ EC, but omitted contex-
rary environmental movement is based on a transformation tual level factors in our models. A fuller theory or model of
of social, political, and ecological beliefs (see, for example, citizens’ EC must attend to contextual effects (such as
Inglehart, 1995; Abramson and Inglehart, 1995). While the indicators of local or community’s environmental risks) that
new values and beliefs on political economy and environ- may influence individual level variables and individual
mental issues have been gradually woven into the main- citizens’ EC.
stream of American culture over the last several decades, the Finally, future research should continue to examine and
conflicts between different ideologies and between old and re-examine the determinants of citizens’ EC on a regular
new ecological paradigms still play a significant role in driving basis. As shown in our study, some factors that were
citizens’ EC in different directions. Sometimes these conflicts, found substantially and consistently influential years or
along with short-term political currents (e.g., the prevailing decades ago do not necessarily exert the same impacts on
view of placing economic costs over environmental benefits today’s EC; and more importantly, signs or directions of
in the White House throughout Reagan’s Administration), these effects may change over time—a negative relationship
can cause immense setbacks in environmental policies and between two variables can reverse to positive association
result in long-lasting negative policy consequences. While over time, and vice versa. Our knowledge and scientific
policy efforts to establish personal environmental norms understanding about people’s concern for the environment
seem to be much more difficult in the short term than raising needs to be updated and re-examined, and new explana-
citizens’ environmental awareness and enhancing their tions and theories need to be built based on new empirical
environmental knowledge, successful environmental educa- data.
tion policy must be accomplished by attempting to encourage
appropriate values and crucial beliefs toward the environ-
ment. Conflict of interest
There are also several important implications for future
research on individual citizens’ EC. First, our study confirms The authors declare that there are no conflicts of interest.
Appendix A. Survey Question Wording and Coding

Variables Question Coding
Dependent Variables
Concern_En_2004 Now, I’d like to list some specific issues that concern people. On a 0–10 scale
scale from 0 to 10, with 0 indicating completely unconcerned and 10
indicating extremely concerned, rate these issues on how worried you
are about them right now.
Concern_GWCC_2004
Concern_ Pollution_2004 The Environment (in all three surveys)
Concern_En_2007 Global Warming and Climate Change (in all three surveys)
Concer_GWCC_2007 Pollution (in 2004 and 2013 surveys only)
Concern_En_2013
Concern_GWCC_2013
Concern_ Pollution_2013
Independent Variables
Age How old are you? Number in actual years
Gender_Female As part of the survey, I am required to ask: are you male or female? 1 = female; 0 = male
Education What is the highest level of education you have completed? 1–6 scale
Elementary or some high school; High school graduate/GED; Trade
or vocational certification; Some college/Associates degree; College
graduate or Post grad degree
Household Income What was the estimated annual income last year for your household? 1–11 scale
Less than $10,000,10 to $20,000, 21 to $30,000, 31 to $40,000, 41 to
$50,000, 51 to $60,000, 61 to $70,000, 71 to $80,000, 81 to $90,000, 91 to
$100,000, More than $100,000
Race_White From the following options, do you consider yourself to be: 1 Black, or 1 = White (Non-Hispanic);
African American; 2 White; 3 Asian/American Indian; 4 Native 0 = Non-White
Hawaiian or other Pacific Islander
Liberal-Conservative Which of the following categories best describes your political views? 1–7 scale: 1 = strongly liber-
Would you say that you are: Strongly Liberal, Liberal, Slightly Liberal, al to 7 = strongly conserva-
Middle of the Road, Slightly Conservative, Conservative, Strongly tive. Category 8 = Other was
Conservative, Other(Specify) treated as missing value
NEP Scale I am going to read you some statements about human beings and the Average NEP Scale based on
physical environment. For each statement, please indicate whether the answers to the seven
you: strongly agree, agree, disagree, or strongly disagree. items:
We are approaching the limit of the number of people the earth can 4 = strongly agree
support.
When humans interfere with nature it often produces disastrous 3 = agree
consequences.
Plants and animals have as much right as humans to exist. 2 = disagree
The earth is like a spaceship with limited room and resources. 1 = strongly disagree
The balance of nature is very delicate and easily upset.
Humans were meant to rule over the rest of nature (reversely coded)
If things continue on their present course, we will soon experience a
major ecological catastrophe.
Appendix B. Descriptive Statistics
Descriptive Statistics: 2004 Survey
N Minimum Maximum Mean Std. Deviation
Concern_En_2004 1087 0 10 7.34 2.313

Concern_GWCC_2004 1083 0 10 6.24 2.892
Concern_Pollution_2004 1091 0 10 7.22 2.278
Gender_Female 1093 0 1 .56 .497
Age 1061 18 90 47.31 16.399
Education 1081 1 6 4.16 1.408
Household Income 849 1 11 6.31 3.160
Race_White 1056 0 1 .84 .366
Liberal-Conservative 1047 1 7 4.21 1.759
NEP Scale 936 1 4 2.77 .458
Valid N (list wise) 712

Concern_En_2007 833 0 Descriptive Statistics: 2007 Survey 7.80 2.323
Concern_GWCC_2007 831 0 10 6.97 3.103
Gender_Female 826 0 1 .53 .499
Age 808 18 99 52.76 15.095
Education 826 1 6 4.30 1.391
Race_White 788 0 1 .86 .349
NEP Scale 741 1 4 2.82 .617
Valid N (listwise) 547
Concern_EN_2013 1297 0 Descriptive Statistics: 2013 Survey 6.66 2.60
Concern_GWCC_2013 1310 0 10 5.81 3.11
Concern_Pollution_2013 1304 0 10 6.56 2.52
Gender_Female 1313 0 1 .50 .50
Age 1313 18 92 49.45 17.16
Education 1313 1 6 3.02 1.20
Race_White 1313 0 1 .77 .42
NEP Scale 1239 1 5 3.44 .75
Valid N (list wise) 1201
Appendix C. Correlation Matrix
Correlations: 2004 Survey
Concern_ Concern_ Concern_ Age Gender_ Education Income Race Liberal- NEP
En_2004 GWCC_ Pollution_ Female White Conservative Scale
2004 2004
Concern_En_2004 1
Concern_GWCC_ 0.6832*** 1
2004
Concern_Pollution_ 0.7413*** 0.6643*** 1
2004
Age -0.0419 -0.0285 -0.0395 1
Gender_Female 0.1643*** 0.1507*** 0.1605*** -0.0230 1
Education -0.0218 -0.0273 -0.0673* 0.0454 -0.0118 1
Household Income -0.0411 -0.0442 -0.0378 0.0147 -0.0701* 0.3615*** 1
Race_White -0.0919** -0.1250*** -0.0941** 0.1952*** -0.0004 -0.0035 0.1155*** 1
Liberal- -0.3253*** -0.3965*** -0.2806*** 0.0740** -0.1518*** -0.1324*** 0.0529 0.1351*** 1
Conservative
NEP Scale 0.4893*** 0.5163*** 0.4702*** -0.0756** 0.1357*** -0.0305 -0.0472 -0.0363 -0.3691*** 1
* p < 0.1; ** p < 0.05; *** p < 0.01
Concern_ Concern_ Age Gender_ Education Income Race White Liberal- NEP
En_2007 GWCC_ Female Conservative Scale
2007
Concern_En_2007 1 Correlations:
2007 Survey
Concern_GWCC_2007 0.6970*** 1
Age 0.0718* 0.0717* 1
Gender Female 0.2633*** 0.2188*** -0.0396 1
Education -0.1034* -0.0692 0.0043 -0.0764* 1
Household Income -0.1494*** -0.1466*** -0.1433*** -0.1202*** 0.3442*** 1
Race White -0.1196*** -0.0835* 0.1434*** -0.0141 0.0361 0.0670 1
Liberal- -0.4024*** -0.4889*** 0.1170*** -0.1699*** 0.0018 0.0162 0.0804* 1
Conservative
NEP Scale 0.5479*** 0.6215*** -0.0452 0.1429*** -0.1084* -0.1202*** 0.0038 -0.4630*** 1
* p < 0.1; ** p < 0.05; *** p < 0.01

Correlations: 2013 Survey
Concern_ Concern_ Concern_ Age Gender_ Education Income Race Liberal- NEP
En_2013 GWCC_ Pollution_ Female White Conservative Scale
2013 2013
Concern_En_2013 1
Concern_GWCC_ 0.783*** 1
2013
Concern_Pollution_ 0.855*** 0.743*** 1
2013
Age 0.112*** 0.028 0.139*** 1
Gender_Female 0.083*** 0.106*** 0.115*** -0.009 1
Education 0.025 0.019 0.021 0.005 0.035 1
Household -0.043 -0.033 -0.050* 0.070** -0.029 0.400*** 1
Income
Race_White -0.044 -0.111*** -0.060** 0.161*** 0.038 0.101*** 0.102*** 1
Liberal- -0.263*** -0.396*** -0.225*** 0.121** -0.086*** -0.110*** 0.018 0.128*** 1
Conservative
NEP Scale 0.504*** 0.585*** 0.457*** -0.046 0.064** 0.008 -0.031 -0.032 -0.367*** 1
* p < 0.1; ** p < 0.05; *** p < 0.01
references Cordano, M., Welcomer, S., Scherer, R., 2003. An analysis of the
predictive validity of the New Ecological Paradigm Scale. The
Journal of Environmental Education 34, 22–28.
Dalton, R.J., Gontmacher, Y., Lovrich, N.P., Pierce, J.C., 1999.
Abramson, P.R., Inglehart, R., 1995. Value Change in Global
Environmental attitudes and the new environmental
Perspective. University of Michigan Press, Ann Arbor,
paradigm. In: Dalton, R.J., Garb, P., Lovrich, N.P., Pierce,
Michigan. J.C., Whitely, J.M. (Eds.), Critical masses: Citizens, nuclear
Alibeli, M.A., White, N.R., 2011. The Structure of Environmental weapons production, and environmental destruction in the
Concern. International Journal of Business and Social
United States and Russia. MIT Press, Cambridge, MA, pp.
Science. 2 (4) 1–8. 195–230.
Amburgey, J.W., Thoman, D.B., 2012. Dimensionality of the New Davidson, D.J., Freudenburg, W.R., 1996. Gender and
Ecological Paradigm: Issues of Factor Structure and
Environmental Risk Concerns: A Review and Analysis of
Measurement. Environment and Behavior 44 (2) 235–
Available Research. Environment and Behavior 28, 302–339.
256. Dietz, T.A., Dan, A., Shwom, R., 2007. Support for Climate
Arcury, T.A., Christianson, E.H., 1990. Environmental worldview Change Policy: Social Psychological and Social Structural
in response to environmental problems: Kentucky 1984 and
Influences. Rural Sociology 72 (2) 185–214.
1988 compared. Environment and Behavior 22, 387–407.
Dietz, T., Stern, P.C., Guagnano, G.A., 1998. Social Structural and
Arcury, T.A., Christianson, E.H., 1993. Rural & urban differences Social Psychological Bases of Environmental Concern.
in environmental knowledge and actions. Journal of Environment and Behavior 30, 450–471.
Environmental Education 25, 19–25.
Dillman, D.A., Christensen, J.A., 1972. The public value for
Biel, A., Nilsson, A., 2005. Religions Values and Environmental pollution control. In: Burch, W., et, al. (Eds.), Social Behavior,
Concern: Harmony and Detachment. Social Science Natural Resources and the Environment. Harper & Row, New
Quarterly. 86, 178–191.
York, pp. 237–256.
Blocker, T.J., Eckberg, D.L., 1989. Environmental issues as
Dunlap, R.E., 1975. The impact of political orientation on
women’s issues: General concerns and local hazards. Social environmental attitudes and actions. Environment and
Science Quarterly 70, 586–593. Behavior 7, 428–454.
Blocker, T.J., Eckberg, D.L., 1997. Gender and environmentalism:
Dunlap, R.E., Gale, R.P., 1972. Party membership and
Results from the 1993 General Social Survey. Social Science
environmental politics: a legislative roll-call analysis. Social
Quarterly 78, 841–858. Science Quarterly 55 (3) 670–690.
Bord, R.J., O’Connor, R.E., 1997. The Gender Gap in Dunlap, R.E., Van Liere, K.D., 1978. The ‘new environmental
Environmental Attitudes: The Case of Perceived
paradigm’: A proposed measuring instrument and
Vulnerability to Risk. Social Science Quarterly 78, 830–840. preliminary results. Journal of Environmental Education 9,
Buttel, F.H., Flinn, W.L., 1974. The structure of support for the 10–19.
environmental movement, 1968-1970. Rural Sociology 39 (1) Dunlap, R.E., Van Liere, K.D., 1984. Commitment to the
56–69.
dominant social paradigm and concern for environmental
Clements, B., 2012. The sociological and attitudinal bases of quality. Social Science Quarterly 65, 1013–1028.
environmentally-related beliefs and behaviour in Britain. Dunlap, R.E., Van Liere, K.D., Mertig, A.G., Jones, R.E., 2000.
Environmental Politics 21 (6) 901–921.
Measuring Endorsement of the New Ecological Paradigm: A
Cluck, R., 1998. The Multi-dimensional basis of Environmental
Revised NEP Scale. Journal of Social Issues 56 (3) 425–445.
Attitudes. Ph. D Dissertation. Mississippi State University. Dunlap, R.E., 1991. Trends in Public Opinion toward
Constantiti, E., Hanff, K., 1972. Environmental concern and Lake Environmental Issues: 1965-1990. Society and Natural
Tahoe: a study of elite perceptions, backgrounds and
Resources 4, 285–312.
attitudes. Environment and Behavior 4 (2) 209–242.
Dunlap, R.E., Jones, R.E., 2002. Environmental Concern: Kanagy, C.L., Humphrey, C.R., Firebaugh, G., 1994. Surging
Conceptual and Measurement Issues. In: Dunlap, R.E., Environmentalism: Changing Public Opinion or Changing
Michelson, W. (Eds.), Handbook of Environmental Sociology Publics? Social Science Quarterly 75, 804–819.
<SE>Please check whether page numbers are to be set here Kempton, W., Boster, J., Hartley, J., 1995. Environmental values
‘‘(p’’.<SE>(p. Greenwood Publishing Group, Westport, CT, in American culture. MIT Press, Cambridge, MA.
pp. 484–524. Kilbourne, W.E., Beckman, S.C., Lewis, A., Van Dam, Y., 2001. A
Dunlap, R.E., Scarce, R., 1991. Poll trends: Environmental multinational examination of the role of the dominant social
problems and protection. Public Opinion Quarterly 55 (4) paradigm in environmental attitudes of University students.
651–672. Environment and Behavior 33, 209–228.
Eckberg, D.L., Blocker, T.J., 1996. Christianity, environmentalism Klineberg, S.L., McKeever, M., Rothenbach, B., 1998.
and the theoretical Problem of fundamentalism. Journal for Demographic Predictors of Environmental Concern: It Does
the Scientific Study of Religion 35, 343–355. Make a Difference How It’s Measured. Social Science
Ester, P., Van Der Meer, F., 1982. Determinants of Individual Quarterly 79, 734–753.
Environmental Behavior: An Outline of a Behavioral Model La Trobe, H.L., Acott, T.G., 2000. A modified NEP/DSP
and Some Research Findings. The Netherlands’ Journal of environmental attitudes scale. The Journal of Environmental
Sociology 18 (15) 54–94. Education 32 (1) 12–20.
Fox, J., 1991. Regression Diagnostics. Sage Publications, Lai, J.C., Tao, J., 2003. Perception of Environmental Hazards in
Thousand Oaks, CA. Hong Kong Chinese. Risk Analysis 23, 669–684.
Fransson, N., Garling, T., 1999. Environmental Concern: Louv, R., 2005. Last child in the woods: Saving our children from
Conceptual Definitions, Measurement, Methods, and nature-deficit disorder. Algonquin, Chapel Hill, NC.
Research Findings. Journal of Environmental Psychology. 19, McCright, A.M., Dunlap, R.E., 2011a. The politicization of climate
369–382. change and polarization in the American public’s views of
Gamba, R.J., Oskamp, S., 1994. Factors influencing community global warming, 2001-2010. The Sociological Quarterly 52,
residents’ participation in commingled curbside recycling 155–194.
programs. Environment and Behavior 26, 587–612. McCright, A.M., Dunlap, R.E., 2011b. Cool dudes: The denial of
Guth, J.L., Green, J.C., Kellstedt, L.A., Smidt, C.E., 1995. Faith and climate change among conservative white males in the
the environment: Religious beliefs and attitudes on United States. Global Environmental Change 21 (4) 1163–
environmental policy. American Journal of Political Science 1172.
39, 364–382. Maslow, A.H., 1970. Motivation and personality. Viking Press,
Hamilton, L.C., 1985. Who cares about water pollution? New York.
Opinions in a small-town crisis. Sociological Inquiry 55, 170– McStay, J., Dunlap, R.E., 1983. Male-Female Differences in
181. Concern for Environmental Quality. International Journal of
Hand, C.M., Van Liere, K.D., 1984. Religion, Mastery-over- Women’s Studies 6, 291–301.
Nature, and Environmental Concern. Social Forces 63, 555– Mohai, P., Bryant, B., 1998. Is There a ‘‘Race’’ Effect on Concern
570. for Environmental Quality? The Public Opinion Quarterly 62
Hawcroft, L.J., Milfont, T.L., 2010. The use (and abuse) of the (4) 475–505.
New Environmental Paradigm Scale over the last 30 years: a Mohai, P., Twight, B.W., 1987. Age and Environmentalism: An
meta-analysis. Journal of Environmental Psychology 30, 143– Elaboration of the Buttel Model Using National Survey
158. Evidence. Social Science Quarterly 68, 798–815.
Hine, D.W., Gifford, R., 1991. Fear appeals, individual Mohai, P., 1990. Black Environmentalism. Social Science
differences, and environmental concern. Journal of Quarterly 71 (4) 744–765.
Environmental Education 23, 36–41. Mohai, P., 2003. Dispelling Old Myths: African American
Hinich, M., Liu, X., Vedlitz, A., Lindsey, C., 2013. Beyond the Left- Concern for the Environment. Environment 45 (5)
Right Cleavage: Exploring American Political Choice Space. 11–26.
Journal of Theoretical Politics. 25 (1) 75–104. Naess, A., 1989. Ecology, Community, and Lifestyle: An Outline
Howell, S.E., Laska, S.B., 1992. The changing face of the of an Ecosophy. Cambridge University Press, Cambridge.
environmental coalition: A research note. Environment and Olofsson, A., Ohman, S., 2006. General Beliefs and
Behavior 24, 134–144. Environmental Concern: Transatlantic Comparison.
Huber, P.J., 1967. The behavior of maximum likelihood Environmental and Behavior 38 (6) 768–790.
estimates under nonstandard conditions. In Proceedings of Pierce, J.C., Dalton, R.J., Zaitsev, A., 1999. Public perceptions of
the Fifth Berkeley Symposium on Mathematical Statistics environmental conditions. In: Dalton, R.J., Garb, P., Lovrich,
and Probability. Berkeley, CA: University of California Press N.P., Pierce, J.C., Whitely, J.M. (Eds.), Critical masses:
vol. 1, 221–223. Citizens, nuclear weapons production, and environmental
Hunter, L.M., Hatch, A., Johnson, A., 2004. Cross-National destruction in the United States and Russia. MIT Press,
Gender Variation in Environmental Behaviors. Social Science Cambridge, MA, pp. 97–129.
Quarterly 85 (3) 677–694. Samdahl, D.M., Robertson, R., 1989. Social Determinants of
Inglehart, R., 1995. Public support for environmental protection: Environmental Concern – Specification and Test of the
objective problems and subjective values in 43 societies. PS: Model. Environment and Behavior 21, 57–81.
Political Science & Politics 28 (1) 57–72. Stern, P.C., Dietz, T., Kalof, L., Guagnano, G.A., 1995. Values,
Jones, R.E., Dunlap, R.E., 1992. The Social Bases of beliefs and pro-environmental action: Attitude formation
Environmental Concern: Have They Changed Over Time? toward emergent attitude objects. Journal of Applied Social
Rural Sociology 57 (1) 28–47. Psychology 25, 1611–1636.
Jones, R.E., 1998. Black Concern for the Environment: Myth Taylor, D.E., 1989. Blacks and the Environment: Toward an
versus Reality. Society & Natural Resources: An International Explanation of the Concern Gap between Blacks and Whites.
Journal 11 (3) 209–228. Environment and Behavior 21 (2) 175–205.
Jones, R.E., 2002. Blacks Just Don’t Care: Unmasking Popular Tindall, D.B., Davies, S., Mauboule, C., 2003. Activism and
Stereotypes about Concern for the Environment among Conservation Behavior in an Environmental Movement: The
African-Americans. International Journal of Public Contradictory Effects of Gender. Society and Natural
Administration 25 (2–3) 221–251. Resources 16, 909–932.
Tremblay, K.R., Dunlap, R.E., 1978. Rural-urban residence and California: Is ‘Environmentalism’ Still a White Phenomenon?
concern with environmental quality: A replication and Political Research Quarterly 58 (3) 435–447.
extension. Rural Sociology 43, 474–491. Wood, D.B., Vedlitz, A., 2007. Definition, Information Processing,
Twenge, J.M., Campbell, K.W., Freeman, E.C., 2012. Generational the Politics and of Global Warming. American Journal of
Differences in Young Adults’ Life Goals, Concern for Others, Political Science 51 (3) 552–568.
and Civic Orientation, 1966-2009. Journal of Personality and Wray-Lake, L., Flanagan, C.A., Osgood, D.W., 2010. Examining
Social Psychology 102 (5) 1045–1062. Trends in Adolescent Environmental Attitudes, Beliefs, and
Van Liere, K.D., Dunlap, R.E., 1981. Environmental concern: does Behaviors across Three Decades. Environment and Behavior
it make a difference how it’s measured? Environment and 42 (1) 61–85.
Behavior 13 (6) 651–676. Xiao, C., Dunlap, R.E., 2007. Validating a Comprehensive Model
Van Liere, K.D., Dunlap, R.E., 1980. The Social Bases of of Environmental Concern Cross-Nationally: A U.S, -
Environmental Concern: A Review of Hypotheses, Canadian Comparison. Social Science Quarterly 88 (2) 471–
Explanations and Empirical Evidence. The Public Opinion 493.
Quarterly Vol. 44 (2) 181–197. Yeager, D.S., Larson, S.B., Krosnick, J., Tompson, T., 2011.
White, H., 1980. A heteroskedasticity-consistent covariance Measuring Americans’ issue priorities: A new version of the
matrix estimator and a direct test for heteroskedasticity. most important problem question reveals more concern
Econometrica 48, 817–830. about global warming and the environment. Public Opinion
Whitney, E., 1993. Lynn White, Ecotheology, and History. Quarterly 75, 125–138.
Environmental Ethics 15, 151–169. Zelezny, L.C., Chua, P., Aldrich, C., 2000. Elaborating on Gender
Whittaker, M., Segura, G.M., Bowler, S., 2005. Racial/Ethnic Differences in.
Group Attitudes toward Environmental Protection in

Gap Latihan 3

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gap Latihan 3

Uploaded by

Copyright:

Available Formats

environmental science & policy 39 (2014) 77–94

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/envsci

Examining the determinants of public

Xinsheng Liu *,1, Arnold Vedlitz, Liu Shi

article info abstract

Table 1 – Determinants of Concern about the Environment 2004.

Table 3 – Determinants of Concern for Pollution 2004.

Table 4 – Determinants of Concern about the Environment 2007.

Table 6 – Determinants of Concern about the Environment 2013.

Table 8 – Determinants of Concern for Pollution 2013.

Appendix A. Survey Question Wording and Coding

Appendix B. Descriptive Statistics

Descriptive Statistics: 2004 Survey

N Minimum Maximum Mean Std. Deviation

Concern_En_2004 1087 0 10 7.34 2.313

N Minimum Maximum Mean Std. Deviation

Appendix C. Correlation Matrix

Correlations: 2004 Survey

* p < 0.1; ** p < 0.05; *** p < 0.01

* p < 0.1; ** p < 0.05; *** p < 0.01

Correlations: 2013 Survey

* p < 0.1; ** p < 0.05; *** p < 0.01

You might also like

* p < 0.1; p < 0.05; * p < 0.01

* p < 0.1; p < 0.05; * p < 0.01

* p < 0.1; p < 0.05; * p < 0.01