The effect of social network sites usage in climate change awareness in Latin America

Abstract

Using data from the Latinobarómetro (Latin Barometer) survey of 2017 to analyze the effect of social network site usage on climate change awareness in 18 Latin American countries, this article makes three contributions. First, it offers results on the socioeconomic determinants of climate awareness in a region of the world where there is scant published evidence in this regard. Second, it shows the effect of social media consumption on climate change awareness by assessing the role of each of the most popular sites: YouTube, Facebook, Instagram, Twitter, LinkedIn, WhatsApp, Snapchat, and Tumblr. Third, it assesses the effects of multi-platform consumption. The results show that YouTube has the strongest and most robust positive and statistically significant effect on climate change awareness, followed by Instagram, Twitter, and WhatsApp, while being a multi-platform user also has a positive and statistically significant effect on climate change awareness. The implications of these findings for understanding the role of social media in the development of environmental awareness are discussed.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11111-023-00417-4.

Introduction

Climate change is one of the main challenges that humankind is facing today (International Panel on Climate Change, 2018; United Nations Environment Programme, 2021). Although the scientific community, followed by significant sectors of society, widely agrees on its existence and its anthropogenic roots, this phenomenon is still denied by many people around the world (Ashe & Poberezhskaya, 2022; Fagan & Huang, 2019; McCright & Dunlap, 2011; Weintrobe, 2013). Interestingly, this situation overlaps with the ongoing heated debate as to whether social network site usage merely produces slacktivism (Smith et al., 2019) and reinforces environmental skepticism (Dunlap, 2013), or if it helps to promote acknowledgement of the problem in creative ways (Parham, 2016) and fosters green citizenship (Anderson, 2017; Boykoff, 2020; Gil de Zúñiga et al., 2012; Lindgren, 2015; Rotman et al., 2011; Takahashi et al., 2017).

It has been argued that the media is key to raising awareness about “the environment,” which is a quite abstract object and thus difficult to apprehend (Dryzek, 2005). Given the long-term, worldwide impact of the environmental crisis, the social media could play an important part in informing the public about the multiple issues and aspects that they cannot directly perceive, as suggested by several message-centered or small-scale audience studies (Hautea et al., 2021; Höijer, 2010; Olausson, 2009; Terracina-Hartman et al., 2013). Large-scale, direct measures of public perception have also confirmed the positive relationship between individual environmental concern and media consumption in general terms or, alternatively, news media use in particular (Harring et al., 2011; Ahern, 2012; Takahashi et al., 2017; Udalov & Welsens, 2021). The positive effect is a stronger trend when considering environment-specific media consumption (Good, 2006; Holbert et al., 2003; Huang, 2016; James et al., 1997; Lee, 2011; Östman, 2013; Shaheen Ali, 2021; Wang & Lin, 2018).

However, less is known about the influence of social media platforms, which have led to a major shift in the communicational paradigm (Boyd, 2010; Boyd & Ellison, 2008; van Dijk, 2006) to the point of propelling a “network society” (Castells, 1996). While the impact of Internet use on climate change awareness has increasingly been explored (Good, 2006; Kahlor & Rosenthal, 2009; Ma et al., 2022; Taddicken, 2013; Zhao, 2009), there are few studies addressing the effect of social network sites usage in these attitudes (Tuitjer & Dirksmeier, 2021). The multiple affordances and the unprecedented content-carrying capacity of sites like YouTube, Facebook, and Instagram puzzle researchers who are trying to ascertain their overall effect on either promoting climate change awareness or furthering its denial.

In this paper, we aim to investigate the most popular social network sites in Latin America and to identify the influence of each one in raising climate change awareness and acceptance. We explore the effect on climate awareness of using any platform and study the effect of each of the following eight sites: YouTube, Facebook, Instagram, Twitter, LinkedIn, WhatsApp, Snapchat, and Tumblr. We focus this paper on to what extent the Latin American population acknowledge the existence of the problem of climate change. Thus, we intend to account for the relationship between social network usage and climate change awareness in its narrow conceptualization as a purely cognitive construct indicating whether the individual is aware of the endangered environment, with acceptance or denial of the phenomenon as possible outcomes. As previous cross-national studies did (i.e. Knight, 2016; Lewis et al., 2019; Zhou, 2015), we focus on one central question for the dependent variable measuring environmental awareness. Considering that this topic is scarcely addressed in general and in particular in Latin America (with the exceptions of Evans, 2015; Evans & Zechmeister, 2018; and Azócar et al., 2021), we intend to account for the effect of individual characteristics and more specifically, individual’s social network usage, on climate change awareness, which constitutes a first step that allows future research analyzing country-level heterogeneity within the Latin American region. To this end, we analyze the primary variable to approach this issue: climate change awareness, more specifically, to what extent climate change is portrayed as a problem. We carry out logistic regression analyses using data from the annual representative survey Latinobarómetro (2017) conducted in 18 Latin American countries in 2017. The total analytical sample comprises 18,917 individuals.

In doing so, we provide empirical evidence of these potential effects from a region of the world that is barely addressed in this debate, namely, Latin America, which is of interest for four main reasons. First, the recently released report International Public Opinion on Climate Change (Leiserowitz et al., 2022) conducted in 192 countries shows that, by comparison with other world regions, climate change concern is high among Latin Americans. This finding challenges Inglehart’s well-established postmaterialist interpretation of environmental values (1995), asserting that meeting basic material needs allows the emergence of postmaterialist values as environmental protection. Second, by comparison with other peripheral regions of the world system, high Internet use levels (Global Change Data Lab & University of Oxford, 2017) situate Latin America closer to North America and Europe regarding potential access to online media. Third, government restrictions on access to online content are fewer than in other peripheral regions (Freedom House, 2019). Fourth, and most importantly, we address Latin American population’s attitudes as this region is expected to suffer some of the worst effects of climate change (Jafino et al., 2020; Ryan & Gorfinkiel, 2016).

Climate change awareness

Climate change awareness can be defined as perceived concerns over threats to human society and natural ecosystems resulting from climate change (Kim & Hall, 2020; Lee et al., 2015). The phenomenon has been the object of extensive empirical research since the 1980s, along with its determinants at the individual and contextual levels (Brügger et al., 2021; Douenne & Fabre, 2020; Knight, 2016; Lee et al., 2015; Nauges et al., 2021; Nistor, 2022; Poortinga et al., 2019; Zhou, 2015). Its counterforce, climate denial or skepticism, have also received scientific attention (Ashe & Poberezhskaya, 2022; Capstick et al., 2015; Whitmarsh, 2011).

By comparison with the initial debate about climate change, public opinion has progressively aligned with scientific consensus (Nisbet & Myers, 2007; Poushter & Huang, 2019), and levels of environmental concern remain generally high (Leiserowitz et al., 2022) even after the appearance of other global priorities like of COVID-19 pandemic (Brügger et al., 2021; Drews et al., 2022; Van Ootegem et al., 2022). However, while the environmental crisis and climate change, in particular, are global phenomena (Beck, 1992, 2009; Crutzen, 2002; Giddens, 1990; Urry, 2011), studies of environmental concern at the individual level have traditionally focused on the so-called developed countries—mostly the USA and Western Europe (Chryst et al., 2018; Harring et al., 2011; Pisano & Lubell, 2017; Tranter & Booth, 2015; Tuitjer & Dirksmeier, 2021). The exceptions are research using worldwide surveys from about a decade ago (Kim, 2011; Knight, 2016; Lee et al., 2015; Tranter & Booth, 2015; Zhou, 2015), which include at least some peripheral countries of the global system, and the Facebook-based global survey by Leiserowitz et al. (2022).

In the case of Latin America, the recently released results based on the international survey that studies Facebook users’ responses (Leiserowitz et al., 2022) show that, by comparison with the rest of the world, a higher proportion of people in Latin America report that they are very or somewhat worried about climate change and that it will harm future generations “a great deal.” After signing the Paris Agreement, Latin American countries have committed to reduce greenhouse effect gases through mitigation and adopting policy involving public participation (Samaniego et al., 2019). Although these national processes around climate policy did not necessarily engage with the wider society through information dissemination in the media or public consultations (Florez, 2016), some countries—Chile, Brazil, Costa Rica, Mexico, Panama, and Uruguay (Samaniego et al., 2019)—did establish highly participative processes of deliberation around the national contribution to the Paris Agreement. Accordingly, levels of concern about climate change in Mexico and Chile (together with Portugal) rank as the highest worldwide (Leiserowitz et al., 2022). Costa Rica, one of the Latin American countries that is most committed to protecting the environment through biodiversity conservation policies and limiting extractive industries like open-pit mining (Kirby & O’Mahony, 2018; Steinberg & Kraft, 2001), is the country where respondents are the most likely to think that climate change is happening in Latin America (94%). In fact, they rank third globally according to the latest data (Leiserowitz et al., 2022). On the other hand, Haiti, for example, is at the opposite pole of accepting that climate change is happening and, globally speaking, belongs to the group with the highest numbers of respondents stating that they have never heard about climate change (32%) (Leiserowitz et al., 2022). In the case of Ecuador, Eisenstadt and West (2017) show that environmental awareness is low in regions where oil extraction is occurring, while in others with plans for future oil extraction environmental concern is higher.

Moving from a contextual to the individual level, Inglehart’s postmaterialist thesis postulates that more affluent individuals would show increased levels of concern (Brügger et al., 2021; Franzen & Meyer, 2010; Franzen & Vogl, 2013; Gelissen, 2007; Kemmelmeier et al., 2002; Nauges et al., 2021), while the review of previous findings conducted by McCright et al. (2016) shows that income is often not significant. Nevertheless, some studies that tested Inglehart’s thesis at the individual level found that there was usually an income effect in environmental concern (Brügger et al., 2021; Franzen & Meyer, 2010; Franzen & Vogl, 2013; Gelissen, 2007; Kemmelmeier et al., 2002; Nauges et al., 2021).

As for other sociodemographic factors, several studies, including cross-national research, show that educational level is a predictor of environmental awareness (Franzen & Vogl, 2013; Zhou, 2015; Knight, 2016; Poortinga et al., 2019; Poushter & Huang, 2019; Brügger et al., 2021; Nistor, 2022). Indeed, two of the most stable predictors of environmental awareness across a worldwide representative sample were individual educational attainment and environmentally related risk perception (Lee et al., 2015), a finding that was confirmed by Evans and Zechmeister (2018) for the Latin American region. The review by McCright et al. (2016) also concluded that highly educated individuals tend to be more concerned about climate change.

Gender has been found to be closely related to climate change awareness, with women ranking higher than men in terms of environmental awareness (Davidson & Freudenburg, 1996; Bord & O’Connor, 1997; Zelezny et al., 2000; McCright, 2010; McCright et al., 2016; Poortinga et al., 2019). Indeed, the study by McCright et al. (2016) found that gender is a strong predictor of climate change awareness. However, other cross-country studies found no statistically significant results of women’s higher awareness (Franzen & Vogl, 2013) or have shown considerable variation across countries (Knight, 2019). The relevance of distinguishing measures of awareness through knowledge from other indicators of concern is suggested by findings showing that, while men have higher levels of knowledge than women, their perception of the severity of climate change is virtually the same (Douenne & Fabre, 2020).

Meanwhile, studies show mixed findings regarding age effects on climate change awareness. According to McCright et al. (2016), 45 out of 89 studies show no statistically significant effects, while 36 out of 89 show that younger adults are more concerned about climate change. A review of multilevel studies across 33 countries found increased concerns at younger ages, which then decreased with age (concave effect) (Franzen & Vogl, 2013). A longitudinal study in New Zealand revealed that one of the only two variables significantly related to climate concern was age: younger people and those who endorse environmental values exhibit significantly greater concern about climate change than other respondents (Milfont, 2012). This age pattern was also found in France (Douenne & Fabre, 2020) and across Europe (Poortinga et al., 2019).

It was only in this century that media use was analyzed as a potential determinant of environmental interest. Previous studies have tried to provide answers to the question of how media use relates to environmental attitudes through diverse measures of media usage: the accumulation of different media used by individuals as indicative of their level of access to communication (Lee et al., 2015), binary use-not use variables for diverse type of media (Zhang & Zhong, 2020), and also measures that consider both frequency and type of use (Diehl et al., 2019; Takahashi et al., 2017; Udalov & Welfens, 2021; Zhang & Skoric, 2018), thus providing more detailed data on media potential influence in attitudes. In the analysis by Lee et al. (2015) of an unprecedentedly extensive global poll of 119 countries representing over 90% of the world’s adult population, respondents’ “access to electronic communication” of any kind had no statistically significant effects on climate change awareness worldwide, confirming previous findings from Germany with regard to the effect on individuals showing traces of skepticism (Taddicken, 2013). In contrast, according to several studies, media use (newspapers, print magazines, TV news, radio news, mobile phone, email, and the Internet) positively predicted environmentally related concerns (Ahern, 2012; Harring et al., 2011; Östman, 2013; Takahashi et al., 2017; Thaker et al., 2017; Udalov & Welfens, 2021). The analysis of survey data in 20 countries suggests that use of the social media for news is associated with less climate skepticism (Diehl et al., 2019).

In summary, education and gender appear to be the strongest predictors of climate awareness, while mixed findings result from the analysis of age and wealth. Unless reduced to the question of material access to online communication, the role of media has been given little attention in the models of previous cross-national studies on climate change awareness or climate change skepticism. The literature review shows that media use has a positive effect on pro-environmental attitudes, while there are mixed findings regarding the effect of Internet use; research on the effect of social network sites’ specifically is addressed in the next section. Regions like Latin America are poorly represented in this field of research, although it constitutes a region of great interest due to relatively higher levels of climate awareness in this region.

Use of social networking sites

Research on social media use and environmental concern has a more limited record than Internet access or general media because of the relative novelty of platforms like Facebook, Twitter, YouTube, and Instagram, although it is increasing with the migration of public conversation to the digital world.

As the Internet has become the main battlefield of opposing views about climate change, scholars have increasingly studied this phenomenon following the seminal contributions of Good (2006), who empirically confirmed how exposure to environment-related Internet content influences environmental concerns. In the last decade, case-based audience studies have been proliferating (Arendt & Matthes, 2014; Brereton, 2018; Brereton & Gómez, 2020; Hunting & Hinck, 2017; Parham, 2016; Shapiro & Park, 2015) more than research evaluating the generalized use of social media technology on environmental sustainability outcomes at large scale (Pearson et al., 2016).

Among the many possibilities of Internet-based digital activity, social network sites like Facebook, Twitter, Instagram, or YouTube provide a unique arena for nurturing public opinion on any issue. Reviews have exposed how the phenomenon involves environmental communication in particular (Anderson, 2017) and how climate change has a life of its own in social media (Pearce et al., 2019). Social network sites influence takes place through dynamics related to information, self-expression, sociality, and artificial intelligence mediations. First, individuals engaging with these sites are potentially exposed to media content which resembles that traditionally provided by newspapers, TV, or radio, and this could result in better informed audiences. However, while former informational contents were related to specific formats like newscasts, nowadays they are not necessarily created and fact-checked by means of journalistic practices ensuring trustworthiness, which has led to the spread of “fake news” (Quandt et al., 2019). Second, social network platforms also promote audience expression and content creation. While the traditional media provide few opportunities in reduced spaces for audience participation, social network platforms encourage a new paradigm of active users who become prod-users (Bruns, 2009) of multimedia posts and comments covering private and public issues. Not only individuals but also groups, institutions, companies, and other stakeholders participate in this process. As a result, amateur contributions and grass-roots campaigns now share this arena with “social” content produced by professionals for commercial brands, and entertainment and news organizations. Third, the social component of these network sites is a breeding ground for the generation of public opinion through peer pressure and all the kinds of social dynamics attached to individuals creating communities of friends or followers, share content, and exchange opinions. Nonetheless, these socially rooted processes of opinion formation are mediated by algorithms leading to unprecedented, complex processes whose outcomes in terms of reaching majorities supporting or opposing highly controversial issues are difficult to predict. According to Sunstein (2007), platform algorithms, along with personal choices, promote the creation of an “echo chamber” by connecting users with like-minded people whose posts construct the online environment experienced in social networking sites. Conversely, it is possible to come across content that is not actively searched for, as is the case of incidental news exposure on social media sites like Facebook (Gottfried & Shearer, 2016), which ends up shaping public perceptions of important issues (Anderson et al., 2014, 2016).

These scenarios reveal a complex process and raise the question of whether the use of social network sites propels the formation of informed public opinion or boosts climate change denial. Apart from the potential uses and dynamics listed above, frequency of use and time spent on social media by users could be crucial to assess the effect in climate attitudes, while presenting a reliability challenge for self-reported data.

Last and crucial as the field of research advances, the particularities of each social network platform may constitute an adequate field for the dissemination of either pro- or anti-environmental discourse. For instance, social media sites like Facebook were born and raised in the spirit of social networking sites, while other platforms like Twitter offer microblogging services, and YouTube and Instagram are seen as mainly content-sharing sites. Platforms like YouTube work within the logic of video search engines or video databases and have users with an online profile that sign in and actively participate in different communities around specific contents.

As for the impact of social network sites on climate change awareness, Twitter is the platform that has mostly attracted scholars’ attention (Ghermandi & Sinclair, 2019; Pearce et al., 2019), probably because of the possibility of large-scale content and interaction analysis which other social network sites prevent by design. In the case of Twitter, climate change issues were found to be discussed among like-minded users thus promoting polarization (Williams et al., 2015). Nevertheless, big data analysis of tweets reveals that the credibility of the sources of information weakens the positive effect of echo chambers in reinforcing skeptical views on climate change (Samantray & Pin, 2019). The review by Pearse et al. (2019) has identified relevant Twitter trends with regard to climate change: use of mainstream information sources, discussion of “settled science,” polarization, and responses to temperature anomalies. More recent research points to Twitter as a place of diverse narratives on climate change, where emotions fuel denial discourse (Eslen-Ziya, 2022), and interactions about sustainability and climate change are predominantly negative in terms of sentiments (Mouronte-López & Subirán, 2022). Furthermore, deniers and believers have been found to choose diverse terms, topics, and tones in their tweets of the last 13 years, while climate change remains a highly politicized issue (Effrosynidis et al., 2022).

In the case of Facebook, a study on issue-specific (energy in this case) engagement in the USA found that it depended on perceptions of personal influence in political action related to climate change (Vraga et al., 2015). As for content led effects, the engagement rate in Facebook fan pages focused on climate awareness is higher when posts refer to disasters, expressing sadness, and suggesting preventive measures (Deo & Prasad, 2020). Another study, conducted in Chile, having analyzed the impact of both YouTube and Facebook, found that users who consume information and discussion about public affairs in these social media are prompted to become much more involved in actions against projects that harm the environment than less active social media users (Halpern et al., 2013). According to experimental research (Lutzke et al., 2019), the influence of fake news about climate change in Facebook was found to be weakened by simple media education interventions. Finally, Facebook use at the country level negatively correlates with the perceived efficacy of information about climate change suggested by respondents to the European Social Survey (Tuitjer & Dirksmeier, 2021).

Inquiring into the effects of YouTube on climate concern, Anderson’s comprehensive review of the social media found only a couple of studies (Anderson, 2017), which would be consistent with the little research that has been done on environmental communication in this platform (León & Bourke, 2018). Most studies offer an indirect assessment of how climate awareness might be affected by use of this social network site. For instance, echo chambers have been identified among scientific news users (Bessi et al., 2016) and the general public (Porter & Hellsten, 2014; Williams et al., 2015) when diverse aspects of climate change were analyzed. In the latter study, which is focused on YouTube comments, it seems that the issue is discussed as scientific fraud, political conspiracy, and even as a media-altered reality, depending on the communities that have formed around certain climate events. Other recent studies add to the hypothesis of the platform’s skeptical effect through “strategically distorted information” (Allgaier, 2019, p. 1). Several studies by Shapiro et al. should be added to this brief list. Their findings suggest that YouTube users are likely to respond to claims about the science of climate change in ways that politicize the issue (Shapiro & Park, 2015), that videos about the climate encourage post-viewing discussion in the comments feed (Shapiro & Park, 2018), and that emotions dominate users’ reactions (Meza et al., 2018). Apart from content, an experimental study has shown that the number of views influences participant perceptions of the importance of the issue (Spartz et al., 2017). Recent YouTube audience studies focusing on engagement with climate change narratives have highlighted the efficacy of emotional appeals, especially hope (Brereton & Gómez, 2020; Shriver-Rice et al., 2022).

Despite the popularity of Instagram (Auxier & Anderson, 2021), this platform has barely been studied along with climate awareness. One exception is the study by Vraga et al. (2020), which shows that fact-based correction after misinformation on climate change contributes to reducing misperception. If research on Instagram consumption in this regard is scarce, to our knowledge, there are no studies addressing the role of WhatsApp, Snapchat, or Tumblr from the perspective of environmental concern.

In short, social network sites in general and specific social media platforms have not been the object of empirical research aiming to generalize findings about the impact of social network sites on climate awareness. Only a few very scattered studies shed some light on specific aspects of social media affordances and dynamics potentially shaping public opinion on climate change besides information access. However, they lack substantive power of generalization, and it is not possible to draw conclusions about the overall effect of social network sites or the isolated effect of specific social media consumption on individual awareness of climate change.

Two opposing hypotheses derived from this debate on the relationship between social network sites and climate change awareness are summarized in the “Climate change awareness” and “Use of social networking sites” sections. On the one hand, a first hypothesis indicates a positive effect of social network site usage on climate change awareness as a result of speculating that social network sites would emulate the extensively researched role of legacy media and news consumption in the development of environmental awareness. On the other hand, a negative effect of social network site usage on climate change awareness could be expected when considering the available evidence that portrays the social media as a breeding ground for skepticism.

Materials and methods

We use data from Latinobarómetro (Latin Barometer), an annual representative public opinion survey, carried out in Latin American countries and, specifically, material from that of 2017 which included questions on usage of social network sites and climate change awareness. This survey consisted of a total sample of 20,200 individuals in face-to-face interviews carried out between June 22 and August 28 in 18 Latin American countries (1000 to 1200 per country). We excluded 98 underaged individuals that were only interviewed in Brazil and Nicaragua. An account of the operationalization and descriptives, including missing values, of all the variables is provided in Table 1 (descriptives provided for the sample after excluding underage individuals; n = 20,102). The methodological information of the survey does not explain how non-response to the survey is treated, and no information was provided when we inquired about this by e-mail. We provide further information about the survey and data quality assessment in the Supplementary Materials. The total analytical sample consists of 18,917 individuals, after we excluded the missing values of the dependent variable (no answer or “do not know” = 1185). The socioeconomic status category showed missing values, which we recoded as an additional category (“do not know” = 3.01% and no answer = 0.89%).

Table 1.

Definitions of measures used in the analyses and descriptive statistics for climate change acknowledgement (n = 20,102; unweighted)

Variable	Definition (n)	Categories	Mean	SD	Minimum	Maximum
Climate change awareness	Based on the following question, recorded: Do you strongly agree (n = 1478; 7.35%), agree (n = 4894; 24.35%), disagree (n = 8707; 43.31%), or strongly disagree (n = 3838; 19.09%) with the following statement: “Climate change does not exist.” No answer or do not know (n = 1185; 5.89%)	0 = Climate change denial (strongly agree; agree) 1 = Climate change awareness (strongly disagree; disagree)	0.66	0.47	0	1
Gender	Two categories: Male (n = 9611; 47.81%); Female (n = 10,491; 52.19%)	1 = Male 2 = Female	1.52	0.50	1	2
Age	Four categories: 18–25 (n = 4527; 22.52%); 2. 26–40 (n = 6759; 33.62%); 3. 41–60 (n = 5962; 29.66%); 4. More than 60 (n = 2854; 14.20%)	1 = 18–25 2 = 26–40 3 = 41–60 4 = More than 60	2.35	0.98	1	4
Socio-economic status	Recoded from self-perceived socio-economic status: high (n = 657; 3.27%); medium high (n = 1207; 6.00%); medium (n = 8450; 42.04%); medium low (n = 5001; 24.88%); low (n = 4002; 19.91%; do not know (n = 606; 3.01%); no answer (n = 179; 0.89%)	1 = High (high; medium high) 2 = Medium (medium) 3 = Low (medium low; low) 4 = Missing values (do not know; no answer)	2.43	0.71	1	4
Educational background	Recoded from illiterate (n = 1460; 7.26%); incomplete primary (n = 3055; 15.20%); complete primary (n = 3232; 16.08%); incomplete secondary (n = 3349; 16.66%); complete secondary (n = 4650; 23.13%); incomplete tertiary (n = 2003; 9.96%), complete tertiary (n = 2353; 11.71%)	1 = Illiterate (illiterate) 2 = Primary (incomplete primary; complete primary) 3 = Secondary (incomplete secondary; complete secondary) 4 = Tertiary (incomplete tertiary, complete tertiary)	2.76	0.87	1	4
Country	Country where the survey was conducted (more information available in the supplementary materials)	1 = Argentina 2 = Bolivia 3 = Brazil 4 = Chile 5 = Colombia 6 = Costa Rica 7 = Dominican Republic 8 = Ecuador 9 = El Salvador 10 = Guatemala 11 = Honduras 12 = Mexico 13 = Nicaragua 14 = Panama 15 = Paraguay 16 = Peru 17 = Uruguay 18 = Venezuela	9.47	5.31	1	18
Social networks sites usage	Based on the following question: Do you use the following social networks?: YouTube (Yes: n = 6679; 33.23%; No: n = 13,423; 66.77%); Facebook (Yes: n = 11,659; 58.00%; No: n = 8443; 42.00%); Instagram (Yes: n = 3877; 19.29%; No: n = 16,225; 80.71%), Twitter (Yes: n = 2329; 11.59%; No: n = 17,773;88.41%); LinkedIn; WhatsApp (Yes: n = 12,104; 60.21%; No: n = 7998; 39.79%); Snapchat (Yes: n = 6679; 33.23%; No n = 13,423; 66.44%); Tumblr (Yes: n = 238; 1.18%; No: n = 19,864; 98.82%)	YouTube (1 = Yes; 0 = No)	0.33	0.47	0	1
		Facebook (1 = Yes; 0 = No)	0.58	0.49	0	1
		Instagram (1 = Yes; 0 = No)	0.19	0.39	0	1
		Twitter (1 = Yes; 0 = No)	0.12	0.32	0	1
		LinkedIn (1 = Yes; 0 = No)	0.02	0.14	0	1
		WhatsApp (1 = Yes; 0 = No)	0.60	0.49	0	1
		Snapchat (1 = Yes; 0 = No)	0.06	0.24	0	1
		Tumblr (1 = Yes; 0 = No)	0.01	0.11	0	1
Number of social networks sites used	Based on the following question: Do you use the following social networks?: YouTube; Facebook; Instagram, Twitter; LinkedIn; WhatsApp; Snapchat; Tumblr; other	0 = None: 29.63% 1 = One: 13.30% 2 = Two: 20.48% 3 = Three or more: 36.59%	1.59	1.26	0	3

Source: Authors’ elaboration from Latinbarómetro, 2017

Climate change awareness is the dependent variable with a binary outcome indicating whether the respondents are aware of or in denial about this environmental problem. Climate change skepticism is ascertained among respondents who state that they agree or strongly agree with the statement, “Climate change does not exist,” while those who are aware of climate change disagree or strongly disagree with this statement. In Fig. 1, we present the level of awareness per country. Apart from the question on climate change acknowledgement, in 2017, the Latinobarómetro included a few other questions on climate change-related attitudes that could help to understand the phenomenon in Latin American countries. However, our research interest focuses on climate change acknowledgement for two reasons. First, our paper intends to account for climate denial. According to the meta-study by Rode et al. (2021), belief in climate change is more easily weakened than strengthened as a result of media exposure; therefore, a focus on variables related to denial for a study aimed at exploring the effect of media use in climate awareness seems entirely appropriate. Also, previous research suggests that policy support attitudes—as measured through supplementary questions focused on who should act and if countries should give priority to climate change, for the case of Latinobarómetro—are less influenced by media exposure than beliefs (Rode et al., 2021). Second, the study followed the path of previous studies that chose to work with the one question that better suits the main research objective (Knight, 2016; Lewis et al., 2019; Zhou, 2015). The first attitudinal step in the full acceptance of the severity of the problem is acknowledging the existence of climate change, while other variables like the one regarding the anthropogenic roots of climate change stem from this main one and could be useful to assess the extent of individual acceptance of climate science in future analysis.

Fig. 1.

Climate change awareness in Latin American countries (% of the population aged 18 or over) (n = 20,102). Source: Authors’ elaboration from Latinbarómetro, 2017

The independent variable of interest is social network sites usage for each platform (Facebook, WhatsApp, Snapchat, Instagram, YouTube, Tumblr, Twitter, and LinkedIn). As highlighted in the “Use of social networking sites” section, both frequency and type of usage matter in understanding the potential influence of social network usage on climate change awareness. However, this survey provides a quite simple independent variable that just distinguishes between usage and not usage, thus not providing further nuances such as frequency and type of usage. This constitutes a limitation discussed in the “Discussion” section. We explore the effect of being exposed to each of these platforms, and that of being a user of one, two, three, or more of these network sites. Although this survey constitutes an improvement due to the distinction of each social network site instead of a less precise variable such as Internet use or access to electronic media, it would benefit from redesigning this question to account for at least the intensity of social media use. The group of independent variables is related to the respondent’s socio-economic background: gender, age, educational background, socioeconomic status, and country of residence.

Hypotheses

Following the literature review, we state here two hypotheses that indicate opposite effects. The assessment will reject or fail to reject the hypotheses based on whether the results are statistically significant or not.

Hypothesis 1. A positive effect of social network site usage on climate change awareness. Therefore: each platform (H 1.1) and multi-platform usage (H 1.2) are expected to have a positive effect on climate change awareness.

Hypothesis 2. A negative effect of social network site usage on climate change awareness. Therefore: each platform (H 2.1) and multi-platform usage (H 2.2) are expected to have a negative effect on climate change awareness.

Analytical approach

We carry out multivariate analyses using logistic regression models. Given that the original dependent variable is ordinal, we considered performing an ordered logistic regression. However, having found that the assumption of proportional odds did not hold, we performed a logistic regression. Since there are individual effects and country-specific effects in our data, multilevel modeling seems to be the best option. However, running multilevel modeling among 18 countries raises some methodological concerns and even the question of whether multilevel modeling analysis should be performed. There is a long-running discussion on the minimum number of level 2 units needed to perform multilevel modeling analyses, with opinions ranging from 10 to 50 (for an in-depth review, see Bryan and Jenkins (2016)). However, these authors conclude that, for logit models, 30 countries should be the minimum. Following this methodological contribution, we conduct logistic regression analysis, controlling for the individual’s country of origin. As the focus of our research is the effect of social network usage in climate change awareness, introducing the country of residence is the best approach for considering the overall country-specific effects that it covers but it is not restricted to, for example, economically, politically, socio-culturally, and environmentally related aspects at a national level.

We conduct stepwise logistic regression analyses, as presented in Table 2, where:

Table 2.

Logistic regression estimating climate change awareness for Latin America (n = 18,917)

	Model 1			Model 2			Model 3
Variable	$\beta$	SE	OR	$\beta$	SE	OR	$\beta$	SE	OR
Sex (Ref: men)	− 0.035	0.031	0.965	− 0.017	0.032	0.983	− 0.026	0.031	0.974
Age (Ref: 18–25)
26–40	− 0.023	0.043	0.978	0.032	0.046	1.032	0.038	0.046	1.039
41–60	− 0.059	0.043	0.942	0.069	0.053	1.071	0.081	0.053	1.085
61 +	− 0.056	0.055	0.945	0.132*	0.073	1.141	0.149*	0.074	1.160
Socio-economic status (Ref: high)
Medium	0.271***	0.074	1.312	0.267***	0.074	1.306	0.263***	0.073	1.301
Low	0.366***	0.081	1.442	0.389***	0.083	1.475	0.383***	0.083	1.467
Missing	0.398***	0.155	1.488	0.423***	0.159	1.527	0.427***	0.160	1.533
Educational background (Ref: illiterate)
Primary	0.072	0.071	1.074	0.046	0.070	1.047	0.040	0.070	1.041
Secondary	0.284***	0.092	1.328	0.176*	0.085	1.192	0.168*	0.085	1.183
Tertiary	0.567***	0.132	1.763	0.368***	0.114	1.444	0.373***	0.115	1.452
Country (Ref: Costa Rica)
Argentina	0.648***	0.197	1.912	0.644***	0.197	1.904	0.652***	0.198	1.920
Bolivia	− 0.040	0.091	0.960	0.045	0.100	1.046	0.048	0.100	1.049
Brazil	0.524***	0.172	1.688	0.527***	0.173	1.694	0.530***	0.174	1.699
Chile	− 0.095	0.089	0.910	− 0.099	0.089	0.906	− 0.094	0.089	0.911
Colombia	0.465***	0.158	1.592	0.470***	0.159	1.600	0.488***	0.162	1.630
Dominican Republic	− 0.783***	0.044	0.457	− 0.776***	0.044	0.460	− 0.785***	0.044	0.456
Ecuador	− 1.084***	0.031	0.338	− 1.015***	0.034	0.362	− 1.021***	0.033	0.360
El Salvador	− 0.480***	0.060	0.619	− 0.460***	0.062	0.631	− 0.457***	0.062	0.633
Guatemala	− 0.410***	0.065	0.663	− 0.309**	0.073	0.734	− 0.308**	0.073	0.735
Honduras	− 0.382***	0.066	0.683	− 0.323***	0.070	0.724	− 0.326***	0.070	0.722
Mexico	0.239*	0.123	1.270	0.266**	0.127	1.305	0.277**	0.129	1.320
Nicaragua	− 0.508***	0.059	0.602	− 0.379***	0.069	0.685	− 0.383***	0.068	0.682
Panama	− 0.410***	0.064	0.664	− 0.370***	0.068	0.691	− 0.365***	0.068	0.694
Paraguay	0.603***	0.190	1.828	0.645***	0.199	1.906	0.653***	0.200	1.921
Peru	− 0.043	0.091	0.958	0.040	0.101	1.040	0.040	0.100	1.041
Uruguay	1.309***	0.431	3.701	1.285***	0.422	3.615	1.300***	0.428	3.670
Venezuela	− 0.106	0.086	0.899	− 0.035	0.094	0.966	− 0.036	0.093	0.965
Social network site use (Ref: no use)
YouTube				0.263***	0.058	1.301
Facebook				0.006	0.047	1.006
Instagram				0.130*	0.061	1.138
Twitter				0.129*	0.071	1.138
LinkedIn				0.163	0.158	1.177
WhatsApp				0.102*	0.051	1.107
Snapchat				− 0.108	0.069	0.898
Tumblr				− 0.211	0.130	0.809
Number of social network sites (Ref: none)
1							0.093	0.060	1.097
2							0.127*	0.059	1.135
3 +							0.443***	0.082	1.557
Constant	0.241*	0.141	1.272	0.012	0.117	1.012	0.003	0.116	1.003
Pseudo R2	0.0615			0.0657			0.0651

*p < 0.05; ** p < 0.01; *** p < 0.001

Source: Authors’ elaboration from Latinbarómetro, 2017

Model 1: Tests the effect of socioeconomic background (sex, age, socioeconomic status, and educational background) on climate change awareness. We control for country effects, taking Costa Rica as the reference category, as it is known as a pro-environment country (Kirby & O’Mahony, 2018; Steinberg & Kraft, 2001).

Model 2: Tests the effect of the specific social network sites usage (YouTube, WhatsApp, Facebook, Twitter, Instagram, Tumblr, LinkedIn, and Snapchat), controlling by socioeconomic background and country of residence.

Model 3: Shows the effect of using one, two, three, or more sites, controlling by socioeconomic background and country of residence. We conduct robustness checks to confirm our results, performing the same analysis but excluding the six countries—Argentina, Brazil, Chile, Colombia, Honduras, and Venezuela—in which the distribution per gender and age is not consistent with the results of the Census data (see Supplementary Materials).

Results

Figure 1 presents the descriptive results on climate change awareness. Uruguay is the country with higher levels of climate change awareness among the adult population (87.0%), followed by Argentina (75.9%), Brazil (73.2%), and Colombia (73.2%). On the other hand, Ecuador is the country with lower levels of awareness (42.3%), followed by the Dominican Republic (46.1%) and Nicaragua (49.3%). Table 2 presents the results for the logistic regression analyses, showing the regression coefficient ($\beta$), the standard errors (SE), and the odds ratio (OR). The results indicate that there are no statistically significant effects of gender in climate change awareness. Regarding age, although the first model shows a negative and non-statistically significant effect of increased ages in climate change awareness, when introducing variables related with social media consumption, we find that, by comparison with the youngest groups, there is a positive and statistically significant effect for ages of 61 and above in climate change awareness. Both gender and age seem to have a specific pattern in Latin America, which we discuss in the following section.

Socioeconomic status shows consistent, statistically significant results indicating that upper-class individuals are less likely to be aware of the climate change threat than middle- and lower-class individuals as well as those who do not place themselves in either of the latter two categories. The effect is strong and statistically significant (p-value < 0.001) across models. Also, more educated individuals show a positive and statistically significant effect on climate change awareness across models (p-value < 0.001 for those who access tertiary education).

The results by country indicate that individuals living in Argentina, Brazil, Colombia, Mexico, Paraguay, and Uruguay are more aware of climate change than the inhabitants of Costa Rica. If we consider that Model 1 accounts for country differences, we find that climate change awareness is markedly higher among the Uruguayan population (OR: 3.701; p-value < 0.001) followed by Argentinians (OR: 1.912; p-value < 0.001), and Paraguayans (OR: 1.828; p-value < 0.001). Conversely, people living in the Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Nicaragua, and Panama are more skeptical when compared to citizens of Costa Rica. If we consider Model 1 accounts for country differences, we find that climate change denial is higher in the Dominican Republic population (OR: 0.457; p-value < 0.001), Nicaragua (OR: 0.602; p-value < 0.001), and El Salvador (OR: 0.619; p-value < 0.001).

Model 2 introduces consumption of specific social network sites as a determinant of climate awareness. Among the different social network sites, we find that YouTube has the strongest effect on climate change awareness comparing with other social network site usage (OR: 1.301; p-value < 0.001). It is followed by Twitter and Instagram, which have a similar effect (OR: 1.138; p-value < 0.05), and WhatsApp (OR: 1.107; p-value < 0.05). Facebook and LinkedIn show a positive non-statistically significant effect on awareness. Conversely, Snapchat and Tumblr have a negative non-statistically significant effect on climate change. Model 3 shows that being a user of one or more social network sites has a positive effect on climate change awareness. This positive effect is significant when two sites are used (OR: 1.135; p-value < 0.05), and also three sites (OR: 1.557; p-value < 0.001). Pseudo R² is rather low, about 0.060 across models, which indicates these models’ limitation in explaining climate change awareness in Latin America. Indeed, climate change awareness is a very complex phenomenon, affected by multiple aspects that are not possible to grasp through this survey (i.e., how climate change is affecting the immediate context of the respondent). Despite this complexity, we provide reliable results as our models indicate that most of the independent variables included have a statistically significant effect on climate change awareness. Certainly, the large sample size allows us to capture these statistically significant correlations that we might not capture with a small sample size. The robustness tests show that results are quite stable with exception of WhatsApp and using two platforms, which’s results are not statistically significant (see Supplementary Materials).

Discussion

Given the generally scant evidence around determinants of climate change awareness in Latin America, it is worth discussing the results of socioeconomic variables before moving to hypotheses testing. Although climate change awareness among women is slightly lower than among men and not statistically significant, this finding is puzzling given that most studies show statistically significant results for increased awareness among women (Franzen & Vogl, 2013; McCright, 2016; Poortinga et al., 2019). A regional effect might be in place here, as previous worldwide studies (Franzen & Vogl, 2013; Knight, 2019) were less prone to find significant effects of gender in environmental awareness, as happens with this case. Interestingly, in the first model, when results are not controlled by social network usage, there is a non-statistically significant effect of age. However, when usage of social network sites is introduced into the models, the oldest age group (population over 60 years old) shows higher levels of awareness when compared to the youngest age group. This finding is against expectations following McCright et al. (2016) review stating that most studies show no statistically significant effects of age or increased concern among young adults. Our finding suggests that there are different ways in which platforms shape individuals’ awareness across generations or phases in the life cycle, which could be related to different generational patterns in media consumption. Besides, two different explanations can be offered regarding how generations cope with climate change-related issues. On the one hand, the popular explanation refers to a bias among young people which is developed to cope with “bad news” that threatens their way of life and prospects (Eckersley, 1999; Threadgold, 2012). On the other hand, the “ecology of the poor” (Martínez Alier, 1991; 2011) might offer an explanation from the standpoint of global environmental justice that might explain why older individuals in Latin America, who have been witnessing for much longer the environmental degradation associated with the extractive activities traditionally carried out in world system peripheries, might be more environmentally aware than younger generations. More updated empirical research is needed to examine how the contextual factor of belonging to the Global South interacts with an individual’s background in shaping environmental concerns.

The results are consistent across models indicating that upper-class individuals are more prone to denying climate change in Latin America, which surprisingly contradicts Inglehart’s interpretation of the impact of wealth on environmental values (Inglehart, 1990, 1995, 1997), which is shown by numerous studies (i.e., Brügger et al., 2021; Franzen & Vogl, 2013; Nauges et al., 2021), while also challenging those that found no significant effects (McCright et al., 2016). Upper-class individuals might deny the existence of climate change as a psychological mechanism for bridging the dissonance with their high-carbon lifestyles (Stoknes, 2014). It is also worth considering that groups of higher socioeconomic status, especially in peripheral economies like those of Latin America, often play a major role in environmental degradation because of the impact of economic activities that maintain their group privileges. In fact, findings from Ecuador suggest that individuals that have not enjoyed the benefits of industrial income and those who are more vulnerable to environmental damage would be more concerned about the environmental impact of this kind of project (Eisenstadt & West, 2017). However, the limitations of using self-assessed categories of social economic status should be borne in mind when weighing these results. Our results on educational background are consistent with previous findings (Franzen & Vogl, 2013; Lee et al., 2015; McCright et al., 2016; Evans & Zechmeister, 2018; Pousthter & Huang, 2019; Brügger et al., 2021; Nistor, 2022), indicating that better educated individuals tend to be more aware of climate change.

Literature highlighting Costa Rica’s commitment to and success in protecting the environment (Kirby & O’Mahony, 2018; Steinberg & Kraft, 2001) was the reason why we take this country as our reference category for exploring country-level heterogeneity in Latin America. It is interesting to note that despite Costa Rica’s reputation as an eco-friendly nation, it shows an average awareness of climate change according to our data. These country results confirm previous findings regarding greater levels of climate change awareness in Brazil (Poushter & Huang, 2019), but it does not fully align with recent findings from a Facebook-based global survey (Leiserowitz et al., 2022). Moreover, other countries like Paraguay, Argentina, Colombia, Mexico, and, more remarkably, Uruguay are among those that are more concerned about climate change issues. Mexico and Uruguay—along with Costa Rica and other countries—could have been expected to show high levels of awareness given their encouragement of public participation when constructing national climate policies following the 2016 Paris Agreement (Samaniego et al., 2019). The Caribbean area, with the exception of Costa Rica, presents the lowest levels of awareness in our study, which is not in line with recent findings showing high climate change awareness among countries like Puerto Rico and Nicaragua (Leiserowitz et al., 2022). However, we should be cautious in comparing our results with a Facebook-based survey of Latin Americans because of the different approaches. Latin American Facebook users represent 57.1% (results weighted per population size) of the population, thus a Facebook-based survey considerably limits the sample frame, especially when considering that just users were approached (Leiserowitz et al., 2022). Regarding the higher rates of climate change awareness in Uruguay, different hypotheses that could be posited would suggest the need for future research and also test if any of these hypotheses could hold to other countries showing increased awareness. These would include considering, firstly, a sort of Inglehart postmaterialist effect (1995) related to relatively increased levels of quality of life; second, since Uruguay’s economy is not based on extractivism, it could be showing the opposite of Eisenstadt and West’s (2017) hypothesis; third, the effect of emergent movements pointing to the high levels of methane emission resulting from industrial beef production; fourth, the greater political commitment to fight climate change. On the other hand, among the countries showing lower levels of concern, we highlight the case of Ecuador, which could fit Eisenstadt and West’s (2017) extractivist hypothesis that populations that obtain benefits from oil extraction are less concerned about environmental issues.

More importantly, by highlighting the role of each of the digital platforms, the present study contributes to gain deeper understanding of the effect of social media consumption in climate change awareness. Moreover, it allows assessment of the extent to which being a user of a variety of platforms has an impact on awareness, regardless of the type of usage, which could vary in intensity and level of engagement with the platform. While previous studies highlight the role of specific campaigns or denial discourse in shaping environmental attitudes or problematizing the issues by promoting climate change denial, our research can help in drawing conclusions about the overall effect on the population’s attitudes, beyond the specific content being addressed.

As for the hypotheses, we confirm hypothesis 1 stating that increased exposure to social media contents—holding in mind the limitation of this in the framework of this study—has an effect on climate change awareness. Given the positive and statistically significant results, this hypothesis holds for YouTube, Instagram, Twitter, and WhatsApp, and for users of two or three or more social networks. Our findings on the strong, positive effects of YouTube in shaping pro-environment attitudes are groundbreaking and prove that the climate skepticism proliferating in YouTube which was pointed out by Porter and Hellsten (2014) or Allgaier (2019) does not have a significant effect among Latin American population in general. The results for Instagram call for further research into how, in the domain of Latin America, environmental issues are addressed in this platform. This paper’s results on Twitter follow the expected direction in accordance with what was pointed out by Cody et al. (2015), namely that climate change activists are more active than deniers in this platform, despite traces of polarization (Williams et al., 2015).

Results on WhatsApp are especially intriguing since it has been designed as a private messenger app, although it is possible to disseminate content from other websites, which raises questions about innovative uses of this application. Approaching individuals as multi-platform users makes it possible to see that sites are independent but, at the same time, bridged by users through their online activity. This insight opens up space for supplementary audience-based research.

This study has several data-related limitations. First, it lacks information regarding the methodology of the survey, especially with regard to how non-response is treated and other limitations derived from the comparison of the Latinobarómetro with Census data (see Supplementary Materials). Second, since this study relies on cross-sectional analyses instead of longitudinal methods, its main drawback is not being able to account for cause-and-effect relationships. Third, it relies on subjective measures of socioeconomic status, an approach that can muddy the waters when compared with the effect of an objective focus. Fourth, social network site users are deemed as such if they answer that they use the analyzed social networks. However, this information does not allow us to differentiate forms of usage and frequency. As a result, this study mixes sporadic and intensive users. As a wide array of informative, educational, and entertainment activities might occur when individuals use social network sites, it would be interesting to control the results by other media consumption, for example, of newspapers, print magazines, TV news, and radio news, to obtain a more complete picture of the media repertoires of individuals that might have an impact on their climate awareness. These issues should be considered in designing future surveys and also in the methodological development of this emerging field of study. As a supplementary question, knowing the degree to which the reported social media uses are passive, active, or interactive would help to better assess and reconstruct audience reception processes that might be involved in developing climate awareness or denial. With the same agenda of enriching the understanding of the phenomenon in the region, it would also be interesting to analyze the relationship between social network usage and other variables related to climate change attitudes, such as awareness of its anthropogenic roots or, including supplementary political variables, the assessment of who is supposed to act in response. Fifth, although the use of social network apps like Snapchat has declined in recent years, and popular new ones like TikTok have emerged, they are not considered in this study. Finally, this study intended to approach climate change awareness in Latin America at an individual level while not focusing our attention on country-level heterogeneity, which is an interesting and needed approach to be developed in the future.

Conclusions

This study offers the first results on the influence of social network sites on climate change awareness in Latin America, a region where the phenomenon has barely been analyzed. While previous cross-national studies have considered Internet use without distinguishing between platforms, we offer new material on the impact of using each of the most popular social network sites in 18 Latin American countries. We show that certain social networks site consumption correlates with climate change awareness. We also find that when two or more networks are used, there is a positive, statistically significant effect on climate change awareness.

One of the most notable findings is the considerable effect of YouTube on climate change awareness in Latin America, which could be seen as a triumph of content over sociability when it comes to developing environmental awareness through social media. Despite previous studies reporting the prominence of denial discourse on Twitter (Eslen-Ziya, 2022; Samantray & Pin, 2019), our results indicate that the platform seems to play a role in reproducing a message of climate change awareness. However, we would have expected a greater influence considering that this social network is preferred by scientists (Walter et al., 2019) and scientific institutions (León et al., 2021). The results concerning WhatsApp are particularly intriguing given its origin as a private messenger service. More research is needed to explain the role of all these platforms, but especially of Instagram and WhatsApp, and to ascertain how a widely used messenger service that is mainly associated with the spread of fake news during political campaigns might play a role in shaping climate change attitudes. Additionally, this study accounts for the use of multiple social network sites, which is an interesting aspect as it shows that the more sites used the greater the concern.

Increasingly widespread access to the media and the concomitant changes in communication paradigms have led to a growing role of social networking sites in shaping audience awareness and potentially facilitating a global eco-citizenship. In addressing this issue of climate awareness, our article also sheds light on the ongoing discussion as to whether or not usage of social network sites plays a detrimental role in producing knowledgeable, responsible citizens (Anderson, 2017; Boykoff, 2020; Gil de Zúñiga et al., 2012; Lindgren, 2015; Rotman et al., 2011; Takahashi et al., 2017). With intense use of the Internet as one of the consequences of the COVID-19 lockdown, the debate has become increasingly heated. Hence, these results are relevant and timely for scholars, policymakers, and world citizens who are concerned about one of the greatest challenges now being faced by humankind.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

VG identified the knowledge gap in the field of environmental communication and performed the literature review. AGC designed the methodology and analyzed the data. AGC and VG wrote and approved the final manuscript.

Funding

This work was supported by the National Agency for Research and Innovation of Uruguay, ANII (Grant Identifier POS_EXT_2017_1_146694), the Ministry of Environment of Uruguay and the United Nations Development Programme (UNDP), within the framework of the research project “Communication and Responses to Climate Change” (Grant Identifier ICC_X_2021_1_171438); the Irish Research Council (IRC), within the framework of the project “Stories for Sustainability in the Semi-peripheries of the World System: Online Audiovisual Media Influence on Young People's Attitudes in Ireland and Uruguay” (Grant Identifier GOIPG/2018/2963); CERCA Programme, Generalitat de Catalunya.

Data Availability

Data used in this paper are available open-access on Corporación Latinobarómetro's website. We provide information about the data source, including the link, in the references. During the review process the Editor provided guidelines to present the data source this way. If other information is needed, please, let us know.Corporación Latinobarómetro [Latin Barometer Corporation] (2017). Latinobarómetro 2017 [Latin Barometer 2017]. Corporación Latinobarómetro. https://www.latinobarometro.org/latContents.jsp

Declarations

Conflict of interest

The authors declare no competing interests.