Spektor Et Al, Climate Change Beliefs (2023)

Article https://doi.org/10.
1038/s41467-023-42729-x
Climate change beliefs and their correlates in

Latin America
1 2
Received: 12 October 2022 Matias Spektor , Guilherme N. Fasolin & Juliana Camargo1
Accepted: 20 October 2023
The ability of climate skeptics to block climate action depends on prevailing

beliefs among the public. Research in advanced democracies has shown
Check for updates
skepticism about the existence, the causes, and the consequences of climate
change to be associated with socio-demographic features and political ideol-
1234567890():,;
1234567890():,;
ogy. Yet, little is known about climate-related beliefs elsewhere. We address

this gap by mapping beliefs in climate change and their correlates in Latin
America. We show skepticism over the existence and anthropogenic origins of
climate change to be limited, but identify a high number of skeptics around the
severity of its consequences. Furthermore, we show skepticism to be corre-
lated with psychological rather than socio-political factors: individualistic
worldviews in particular drive disbelief in the severe consequences of climate
change, a worrying finding in contexts where social trust is low. These findings
offer a starting point for better addressing the constraining effects of climate
skepticism in the Global South.
Climate skepticism has driven scholars to investigate the range of Opinion Project), leaving out a range of theoretically relevant
factors influencing public perceptions of climate change1,2. Research individual-level variables.
suggests that political ideology and demography play an outsized role This gap in knowledge is problematic because public beliefs in
in shaping people’s beliefs about climate change, when compared to climate change condition climate action17,18, and Latin America is likely
other social and psychological factors3,4. However, since the bulk of to be disproportionately affected by climate change19. Understanding
existing work focuses on advanced democracies like United States5,6, the factors that are associated with public perceptions of climate
the United Kingdom7,8, Germany9, and Australia10, the question change in developing countries20–22 is necessary if scholars and policy
remains so to whether similar patterns occur in the developing world11. makers are to be able to mount effective strategies to push back
In the case of Latin America, home to some of the most diverse against climate skepticism23 and tailor effective communication stra-
and threatened biomes on the planet, the study of climate psychology tegies to promote positive change in the context of the current climate
has focused on climate change concern12 and climate risk perception13 crisis.
rather than climate change beliefs, defined as propositional cognitions Here, we test the individual-level factors that shape people’s
about the existence, causes, and consequences of climate change. A perceptions of the existence, the anthropogenic causes, and the severe
single study has inspected the factors associated with the belief in the consequences of climate change, three different but complementary
anthropogenic origins of climate change, but not the critical question mental constructs that scholars have shown to constitute climate
of beliefs in its existence and its consequences14. Other studies have change beliefs24–26. By fielding surveys to nationally diverse samples in
explored the effect of country-level variables (e.g., prosperity and Argentina, Brazil, Colombia, Chile, Ecuador, Peru, and Mexico
democracy) on climate change beliefs, but not the individual-level (N = 5338), we are able to study a group of countries that account for
variables shaping such public perceptions15,16. These limitations are more than 80% of regional carbon emissions in Latin America. We
compounded by the fact that these studies draw on general surveys investigate key factors that previous literatures have shown to corre-
that lack dedicated climate sections in their questionnaires (e.g., The late with climate change beliefs. We first test the role for political
Gallup World Poll, Latinobarometer, and Latin American Public ideology, a variable known to matter across advanced
1
School of International Relations, Fundação Getulio Vargas, Avenida Paulista 542, São Paulo 01310-000, Brazil. 2Department of Political Science, Vanderbilt
University, 230 Appleton PI, Nashville, TN 37203, USA. e-mail: [email protected]
Nature Communications | (2023)14:7241 1

Article https://doi.org/10.1038/s41467-023-42729-x
democracies24,27,28. We then turn to socio-demographic variables such Belief in the existence of climate change
as sex, age, education, religion, income, and race, correlates which Considering Latin America as a whole, psychological factors are the
previous studies have shown to be relevant in some settings but no most associated with belief in the existence of climate change. The
others3,29. In addition, we test psychological variables such as indivi- strongest correlate is perceived scientific consensus (respondents
dualism and egalitarianism worldviews, which previous literature has who hold this perception scored, on average, 0.49 points more on
found to be impactful globally27 but not in Latin America14. We also the climate change existence scale (0-8) than those who do not,
assess the New Ecological Paradigm (NEP), a set of values that is holding all else constant (p-value < 0.01)), followed by NEP (β = 0.305,
strongly associated with climate change perceptions in countries as p-value < 0.01) and personal experience with extreme weather events
diverse as U.S., China, Germany and, Netherlands30–32. Finally, in order (β = 0.232, p-value < 0.01) (Supplementary Table 12, column viii).
to ensure the comparability between our findings and other important These results are consistent across almost all countries at the 5%
works in the field, we assess the association between climate change significance level (Supplementary Table 12, columns i to vii). How-
beliefs and other psychological factors theorized as antecedents of ever, these results should be taken cautiously because we find the
climate change perceptions, namely; objective33 and subjective34 reliability of the NEP scale to be low (α = 0.53 fails to meet the con-
knowledge, trust in scientists35, scientific consensus36, and personal ventional 0.60 criterion for an acceptable Cronbach’s alpha).
experience of extreme weather events37,38. Other psychological antecedents matter to a lesser extent (Fig. 1).
For example, at the regional level belief in the existence of climate
Results change has a positive correlation with people’s subjective and objec-
The overwhelming majority of people in Latin America perceive cli- tive knowledge (β = 0.140, p-value < 0.01 and β = 0.165, p-value < 0.01,
mate change to be happening (over 90% in all countries) (Supple- respectively) (Supplementary Table 12, column viii). Yet, at the country
mentary Fig. 1) as a result of human activity (93% on average) level subjective knowledge is only statistically significant in Ecuador,
(Supplementary Fig. 3). Yet, public opinion in the region is more Mexico, and Peru (Supplementary Table 12, columns v, vi and vii) and
divided about the severity of its consequences (an average of 65% of objective knowledge is only statistically significant in Colombia,
respondents believe climate change’s impact to be negative) (Sup- Ecuador and Peru (Supplementary Table 12, columns iv, v and vii).
plementary Fig. 5). The remainder of this section summarizes our Furthermore, Latin Americans who trust scientists tend to believe cli-
findings about the association between political and socio- mate change is happening more than those who do not, a positive
demographic and psychological variables with beliefs about the exis- association that we only observe for the overall sample at the 10%
tence, origins, and consequences of climate change. Supplementary significance level (Supplementary Table 12, column viii). In addition,
Tables 1 and 4 present mean and standard deviations within each we find a positive association between egalitarian values and belief in
dependent and independent variable across the sample, while Sup- the existence of climate change but it is small in magnitude and weakly
plementary Tables 2 and 6 display a summary of correlations among consistent across countries (β = 0.067, p-value < 0.05, Supplementary
the measures in this study. Table 12, column viii). Finally, we find no statistically significant
B. Political ideology and

A. Psychological variables Socio−Demographic variables
0.02 [−0.05, 0.09]
Political ideology: Left
0.14 [ 0.08, 0.21]
Subjective knowledge 0.10 [ 0.04, 0.16]
Political ideology: Progressive
0.01 [−0.06, 0.07]
Female
0.16 [ 0.09, 0.24]
Objective knowledge 0.01 [−0.08, 0.11]
Education: High school or equivalent
0.03 [−0.07, 0.13]
Education: Undergraduate or more
0.49 [ 0.35, 0.62] 0.01 [−0.13, 0.16]
Scientific consensus Religion: Catholic
0.14 [−0.08, 0.35]
Religion: Evangelical Pentecostal or other evangelical
0.12 [−0.02, 0.27] 0.09 [−0.09, 0.26]
Trust in scientists Religion: Evangelical Traditional
0.00 [−0.15, 0.16]
Religion: Others/No Relig.
0.31 [ 0.21, 0.40] 0.12 [ 0.03, 0.21]
The New Ecological Paradigm (NEP) Income: 1 − 2 minimum wages
0.00 [−0.12, 0.12]
Income: 2 − 3 minimum wages
−0.02 [−0.07, 0.03] −0.04 [−0.17, 0.09]
Individualism worldview Income: 3 − 5 minimum wages
0.01 [−0.15, 0.17]
0.04 [−0.09, 0.18]
0.07 [ 0.01, 0.13] Income: 10 minimum wages or more
Egalitarianism worldview
0.03 [−0.09, 0.15]
Income: Do not know/ Prefer to not answer
0.00 [ 0.00, 0.00]
0.23 [ 0.12, 0.35] Age (Years)
Personal experience (extreme weather events)
−0.24 [−0.44,−0.03]
Race: Black
−0.50 −0.25 0.00 0.25 0.50 −0.50 −0.25 0.00 0.25 0.50
Estimate Estimate
Fig. 1 | Correlates of Belief in the Existence of Climate Change in Latin America. existence scale (i.e., 0–8) given a unit increase in the covariates. The higher the
Results from an ordinary least squares model regressing climate change existence scale, the greater the confidence that climate change is happening. The width of the
on a set of independent variables (A) and socio-demographic characteristics (B). confidence intervals for each coefficient is 95% with heteroskedasticity-robust
Respondents from all countries (n = 2887 observations) are included in the model. standard errors. Reference baseline for Education is “Elementary (Primary) or less”,
For each independent variable, point estimates of the coefficients and confidence Religion is “Atheist” and Income is “0–1 minimum wages”.
intervals, in brackets, are reported. Coefficients are changes in the climate change

correlation between holding individualistic worldviews and belief in estimation approach (Supplementary Tables 25 and 26, column i). This
the reality of climate change. This result is consistent across countries. said, race seems to be a potentially relevant factor: Fig. 1 reveals that
Turning to socio-political factors, we find that political ideology is black Latin Americans are more skeptical about the reality of climate
weakly associated with the belief in the existence of climate change in change than their non-black co-nationals. We read these results cau-
the overwhelming majority of countries (Fig. 1). We define political tiously, however, because the negative association is statistically sig-
ideology alongside right-left and conservative-progressive scales nificant in Argentina alone (Supplementary Table 12, column i).
which highlight contrasting individual beliefs over social and political
change (see Methods). Taken as a whole, participants who self-identify Belief in anthropogenic climate change
as progressives tend to express less skepticism about the existence of Figure 2 displays the main correlates of the belief that human activity is
climate change than their conservative co-nationals (β = 0.102, the major cause of climate change. Among the psychological variables,
p-value < 0.05, Supplementary Table 12, column viii). At the country individualistic worldviews stand out as the only negative, statistically
level, however, the relationship is statistically significant only for Brazil significant correlate of skepticism about the anthropogenic origins of
and Chile (Supplementary Table 12, columns ii and iii). Furthermore, climate change. Considering Latin America as a whole, a one-unit
results show that the left-right divide has no statistically significant increase in the individualistic values scale is associated, on average,
correlation with the belief in the existence of climate change with a reduction of 3.8 percentage points in the probability of believing
(β = 0.019, p-value > 0.1, Supplementary Table 12, column viii). in the anthropogenic nature of climate change (p-value < 0.01). With
Socio-demographic variables too are weak correlates of the per- the exception of Ecuador and Mexico, the association is statistically
ception that climate change is happening (Fig. 1). For instance, when significant across the board (Supplementary Table 13, columns v
comparing respondents with no formal schooling or primary educa- and vi).
tion to those with secondary (e.g., high school) and post-secondary Other psychological variables are positively associated with the
education (e.g., undergraduate), there is not significant difference in belief in the anthropogenic nature of climate change: objective
their likelihood of believing in the existence of climate change. In knowledge (β = 0.054, p-value < 0.01), perceived scientific consensus
addition, age (in years) fails to be a significant factor, as each additional (β = 0.070, p-value < 0.01), and trust in scientists (β = 0.056,
year of age is not associated with a stronger belief in the reality of p-value < 0.05) (Supplementary Table 13, column viii). However,
climate change. Moreover, our findings indicate that an increase in results are not consistent across all countries. Furthermore, endor-
income, measured by each additional increment in minimum wages sement of the NEP scale is positively correlated with belief in
(ranging from 1 to 10 or more), is not correlated to a higher likelihood anthropogenic climate change in the entire sample (β = 0.045,
of believing in climate change. Similarly, the results demonstrate that p-value < 0.01, Supplementary Table 13, column viii), although we
female participants are not less likely to be skeptical about the exis- recall that the low internal validity of this scale prevents us from
tence of climate change than male co-nationals, dispelling any gender- making reliable inferences. Overall, in terms of magnitude, these
based skepticism about this issue. These results hold regardless of our psychological variables have similar correlation coefficients with the

0.01 [−0.01, 0.04]
0.00 [−0.01, 0.02]
Subjective knowledge −0.03 [−0.05, 0.00]
0.00 [−0.02, 0.02]
Female
0.05 [ 0.03, 0.08]
Objective knowledge 0.04 [ 0.01, 0.07]
0.06 [ 0.03, 0.09]
0.07 [ 0.03, 0.11] 0.00 [−0.03, 0.03]
−0.08 [−0.16, 0.00]
0.06 [ 0.00, 0.11] 0.03 [−0.01, 0.07]
−0.01 [−0.05, 0.02]
0.05 [ 0.02, 0.07] 0.01 [−0.02, 0.04]
0.02 [−0.02, 0.05]
−0.04 [−0.05,−0.02] 0.00 [−0.04, 0.04]
0.00 [−0.05, 0.04]
0.01 [−0.03, 0.05]
0.01 [−0.01, 0.03] Income: 10 minimum wages or more
0.01 [−0.03, 0.04]
0.00 [ 0.00, 0.00]
0.01 [−0.01, 0.04] Age (Years)
−0.01 [−0.06, 0.05]
Race: Black
−0.2 −0.1 0.0 0.1 −0.2 −0.1 0.0 0.1

Estimate Estimate
Fig. 2 | Correlates of Belief in the Anthropogenic Causes of Climate Change in in brackets, are reported. Coefficients multiplied by one hundred are percentage
Latin America. Results from an ordinary least squares (linear probability) model point changes in the probability of believing climate change is mainly caused by
regressing the perception of climate change anthropogenic causes on a set of human activity given a unit increase in the covariates. The width of the confidence
independent variables (A) and socio-demographic characteristics (B). Respondents intervals for each coefficient is 95% with heteroskedasticity-robust standard errors.
from all countries (n = 2,883 observations) are included in the model. For each Reference baseline for Education is “Elementary (Primary) or less”, Religion is
independent variable, point estimates of the coefficients and confidence intervals, “Atheist” and Income is “0–1 minimum wages”.

perception of anthropogenic climate change in the model including beliefs about the consequences of climate change: people who score
all countries in the sample. The linear combination assessment of higher on individualistic values tend to see climate change’s impact as
coefficients reveals they are not significantly different from each less negative than their co-nationals who score lower on that scale
other at the 5% level and, thus, no psychological variable has a (Fig. 3). Specifically, considering the participating countries as a whole,
stronger association than the others (Supplementary Table 23). a one-unit increase in the individualistic value scale is associated, on
We find political ideology to be weakly associated with skepticism average, with an 11.5 percentage points (p-value < 0.01) reduction in
about the anthropogenic origins of climate change (Fig. 2). The posi- the probability of an individual believing climate change will have
tive association between left-right and anthropogenic climate change negative impacts. This relationship is statistically significant, similar in
is not statistically significant when we estimate our model for Latin size, and consistent across all countries in the sample (Supplementary
America as a whole (β = 0.015, p-value > 0.1), (Supplementary Table 13, Table 14, columns i through vii), suggesting that individualistic values
column viii). Analyzing each country separately, only Ecuador and shape public perceptions of climate impact across the region.
Mexico have a positive statistically significant association between We find that objective knowledge also correlates with the per-
these variables (β = 0.072, p-value < 0.1 and β = 0.068, p-value < 0.05, ception of the consequences of climate change. Our analysis of the
respectively) (Supplementary Table 13, columns v and vii). We find that entire sample shows that the perception of more negative impacts of
progressivism (as compared to conservatism) is negatively associated climate change is positively and statistically significantly associated
with skepticism in anthropogenic climate change (β = −0.025, with objective knowledge (β = 0.067, p-value < 0.01, Supplementary
p-value < 0.05) (Supplementary Table 13, column viii). The association Table 14, column viii), a result observed in 4 out of the 7 countries in
is not statistically significant in 6 out of the 7 participating countries the sample (Supplementary Table 14, columns i, ii, iv and vii). While
(Supplementary Table 13, columns i, ii, iv, v, vi and vii). similar in magnitude, the relationships of other psychological variables
Finally, results show socio-demographic variables to be weakly are more heterogeneous across countries. For example, considering
correlated with the belief in anthropogenic climate change. For our sample as a whole, perceived scientific consensus has a positive
example, we find that sex, religion, income, age and race are not sta- and statistically significant association with belief in the severity of
tistically correlated with the belief that climate change is caused by climate change (β = 0.062, p-value < 0.05, Supplementary Table 14,
humans (Fig. 2). The exception—albeit one that is weakly uniform column viii), but, analyzing each country separately, it is statistically
across countries—is our finding that more highly educated respon- significant only for Chile (Supplementary Table 14, column iii). Fur-
dents are less skeptical about the causal role of human activity in thermore, when assessing the entire sample, the association between
causing climate change (Supplementary Table 13, column viii). the NEP scale and the perception of serious consequences is positive
and statistically significant (β = 0.081, p-value < 0.01, Supplementary
Belief in the consequences of climate change Table 14, column viii), but once again we take these results with caution
Among the set of psychological variables, individualistic worldviews due to our concerns about the scale’s reliability. Finally, we find that at
has one of the strongest and statistically significant associations with the 5% significance level, subjective knowledge, trust in scientists,

0.08 [ 0.05, 0.12]
0.01 [−0.02, 0.04]
Subjective knowledge 0.01 [−0.03, 0.05]
−0.03 [−0.07, 0.01]
Female
0.07 [ 0.04, 0.10]
Objective knowledge 0.08 [ 0.03, 0.13]
0.15 [ 0.09, 0.20]
0.06 [ 0.01, 0.11] −0.07 [−0.13, 0.00]
−0.16 [−0.26,−0.05]
0.04 [−0.03, 0.11] −0.08 [−0.17, 0.00]
−0.06 [−0.13, 0.01]
0.08 [ 0.04, 0.12] 0.06 [ 0.00, 0.12]
0.08 [ 0.02, 0.15]
−0.11 [−0.14,−0.09] 0.07 [ 0.01, 0.13]
0.07 [−0.01, 0.15]
0.08 [ 0.01, 0.15]
0.00 [−0.03, 0.04] Income: 10 minimum wages or more
0.01 [−0.05, 0.07]
0.00 [ 0.00, 0.00]
0.05 [−0.01, 0.10] Age (Years)
0.05 [−0.04, 0.13]
Race: Black
−0.3 −0.2 −0.1 0.0 0.1 0.2 −0.3 −0.2 −0.1 0.0 0.1 0.2
Estimate Estimate
Fig. 3 | Correlates of Belief in the Consequences of Climate Change in Latin are reported. Coefficients multiplied by one hundred are percentage point changes
America. Results from an ordinary least squares (linear probability) model in the probability of believing climate change consequences will be negative given a
regressing the perception of climate change consequences on a set of independent unit increase in the covariates. The width of the confidence intervals for each
variables (A) and socio-demographic characteristics (B). Respondents from all coefficient is 95% with heteroskedasticity-robust standard errors. Reference base-
countries (n = 2887 observations) are included in the model. For each independent line for Education is “Elementary (Primary) or less”, Religion is “Atheist” and Income
variable, point estimates of the coefficients and confidence intervals, in brackets, is “0–1 minimum wages”.

perceived personal experience with extreme weather events, and minimize climate risk scenarios because accepting them could lead to
egalitarian values do not show a statistically significant association limitations on activities they tend to strongly support, such as com-
with the perception of climate change impacts (Fig. 3 and Supple- merce, industry, and free enterprise40. Since members of the public
mentary Table 14, column viii). ranking high on individualistic values tend to be motivated reasoners41
Furthermore, results for the entire region show the association (i.e., they are more likely to seek out information that reinforces their
between political ideology and perceptions of climate change impact existing beliefs), an important question rising from this study that
to be positive and statistically significant (Fig. 3). Specifically, we find future research could explore is how best to persuade them to
that individuals on the left are more likely to perceive the impact of recognize the severity of climate change without triggering
climate change as negative comparted to their co-nationals on the backlash42,43.
right (β = 0.085, p-value < 0.01) (Supplementary Table 14, column viii). Empirically, the outsized role of individualistic values as deter-
Yet, at the country level, we find this relationship to be statistically minant of climate skepticism mirrors wider social trends in Latin
significant only in 3 out of the 7 countries surveyed (Supplementary America. The region has seen a surge in independent self-construal,
Table 14, columns ii, ii, and vi). In addition, we find no statistically personal autonomy, self-expression, and achievement orientation,
significant association between conservatism and skepticism about trends that may be contributing to the activation and intensification of
the impact of climate change (p-value > 0.1) (Fig. 3). individualism over reliance on state institutions44. Furthermore, Latin
Last but not least, the analysis of socio-demographic variables America has experienced rapid erosion in levels of social trust45,
reveals two patterns (Fig. 3). First, education is the most consistent meaning the decline of dense community interactions around shared
demographic correlate of perceived impacts of climate change. We norms46. This erosion may in turn reinforce individualistic tendencies,
find that individuals with secondary education (e.g., high school or as people become less likely to rely on collective action and institu-
equivalent), and particularly those with undergraduate education, are tions than on their own resources to resolve social problems. Future
less doubtful about the negative impacts of climate change compared research could empirically test these ideas, in particular the relation-
to individuals with only primary education or no formal education, ship between social trust and individualistic values in shaping climate-
suggesting that level of education may moderate the individual’s related attitudes47.
ability to concretize the effects of climate change (Supplementary Our finding that individualistic values do not correlate with the
Table 14, column viii). Second, in terms of statistical significance, belief in the existence of climate change also requires explanation.
demographic variables such as age, sex, race, income, and religion Although the present study cannot adjudicate why this is the case,
either have a moderate relationship with the perceived impacts of existing theoretical literatures provide useful foundations for a ten-
climate change or none at all. In a minority of countries, we find reli- tative interpretation. We know that beliefs in the causes and con-
gious denominations such as Catholic, Evangelical Pentecostal, and sequences of climate change can highlight the need for mitigation and
Evangelical traditional to be associated with minimizing the severity of adaptation policies more intensely than the belief in the mere exis-
climate change (Supplementary Table 14, columns iii, iv and vii). tence of climate change, thereby raising the prospect of government
intervention to cope with the effects of climate change34. In turn, this
Discussion may trigger a negative reaction from highly individualistic people, who
This study presents unique data on the individual-level factors that normally push back against big-government intrusion in their way of
shape climate change perceptions in Latin America. Overall, mass life48. These findings are consistent with new literature suggesting that
publics across the region are highly concerned about climate change. climate skeptics are intuitively aware of the distinct dimensions of
We find only limited room for skepticism over the existence of climate climate change beliefs, and have intentionally ceased to target belief in
change or its anthropogenic origins, but identify a high number of the existence of climate change (where scientific evidence is plentiful
skeptics around the severity of its consequences. These results raise and increasingly hard to rebuke), questioning instead its human causes
questions as to what factors might be behind these trends. We identify and the severity of its consequences23,32.
psychological variables—particularly individualistic values—as the Two other psychological factors in this study also play a role in
most influential correlates of climate skepticism across the region, shaping climate change perceptions, although their influence proved
while we find political ideology and socio-demographic factors not to to be limited and inconsistent across the sample: scientific consensus
be as impactful. Taken together, these findings underscore the rele- and objective knowledge. Scientific consensus is the most influential
vance of psychological factors in understanding climate change per- correlate of the belief in the existence of climate change, but it weakly
ceptions in Latin America and hold significant implications for the correlates with beliefs in causes and consequences. One plausible
crafting of effective pro-climate communication strategies aimed at explanation is that in the low-trust societies of Latin America the per-
populations across this region of the Global South. ception of consensus among the scientific community might in itself
In the present study, individualistic values are the most powerful be reassuring, counterbalancing personal doubts about the credibility
determinant of climate skepticism over the origins and consequences of climate science. It remains unclear, however, why this same logic
of climate change, but not its existence. These findings challenge does not apply to beliefs in causes and consequences. The lack of
previous literatures, which found no role for individualistic worldviews correlation between scientific consensus and beliefs on causes and
on the human origins of climate change in Latin America14. Further- consequences therefore casts a shadow of doubt on the notion
more, our results offer novel evidence that such worldviews are cor- established in existing literatures that scientific consensus should be
related with public perceptions of climate change consequences, an seen as a ‘gateway belief’ that opens the door to other related beliefs,
association of high political salience due to the known links between such as the anthropogenic causes and worrisome consequences of
individual-level perceptions about the consequences of climate climate change49,50. Furthermore, it is worth noting that while scientific
change and support for climate action32. consensus in our study somewhat shapes beliefs in climate change,
There are theoretical and empirical reasons for the outsized role trust in scientists does not. The lack of association may indicate that
of individualistic values as a driver of climate skepticism. Theoretically, people view consensus among scientists as more tangible evidence of
it has been long established that people holding individualistic values the truth of climate change, compared to more abstract notions of
downplay the human causes and the severity of climate change, since trustworthiness.
acknowledging them would demand changes in personal lifestyles that Objective knowledge in this study is significant across all dimen-
individualistic people typically resist39. Furthermore, research has sions of climate change beliefs, albeit in a minority of sampled coun-
shown that people holding more individualistic values are inclined to tries. This result confirms previous research showing positive

relationships between climate beliefs and objective knowledge34. It America in anthropogenic climate change and its consequences are
also dovetails with existing literature noting that domain-specific influenced by individualistic values, messages are likely to be more
knowledge is more useful than subjective knowledge measures in effective when they are respectful of these values and align with these
explaining climate change perceptions51. Although the effect of types of motivations60. For example, one strategy might be to advance
objective knowledge on beliefs varies across countries, investing in climate solutions by highlighting the economic benefits (e.g., pro-
climate communication centered around it may be valuable to moting green jobs) that arise from fighting climate change61. Empha-
increase levels of belief in the existence, causes and consequences of sizing benefits may be particularly effective in garnering support for
climate change and, by implication, facilitating climate-friendly poli- pro-environmental action not only because individualistic people are
cies support and behavioral change52. more likely to embrace solutions consistent with pro-market values,
Other psychological variables in this study only had small or null but also because these solutions help address broader development
impact in shaping climate change perceptions. For example, subjective challenges in Latin America.
knowledge, a factor that ranks consistently high in other studies on Second, the sources of pro-climate messages can play an impor-
climate change perceptions, exhibited significant variation in our tant role in providing effective communication to the general public.
study on Latin America27. Moreover, mirroring empirical research Since individualistic people tend to be more influenced by their in-
conducted in the Global North53, we found limited and inconsistent group, calls for climate action coming from in-group members should
associations between personal experience with extreme weather be prioritized at the expense of messages coming from sources who
events and climate change beliefs. Overall, these heterogeneous emphasize collective action and the common good62,63. This may
results suggest that climate change beliefs in Latin America may be include pro-climate messages coming from sources that are not typi-
associated with a unique set of variables, in line with existing multi- cally seen as environment-friendly but who have in recent years
countries studies on the topic13,15. A proper examination of climate nonetheless become invested in good governance in the climate
change dynamics in the region requires more within-case studies and sphere (e.g., bankers, business leaders, agribusiness). Future research
in-depth comparisons to understand why these factors vary cross- should explore the effects of such voices on the public in general and
nationally. Future research should explore the contextual factors that on those members of the public who rank high in individualistic values
contribute to differences in individual-level effects and their interac- and tend to exhibit a strong disdain or mistrust towards government
tion with country-level conditions to explain these variations. interventions. Research to date in this field has been scant and pri-
Political ideology and socio-demographic variables have limited marily centered on advanced democracies alone.
influence on climate change perceptions in our sample. Unlike it is the Third, tailored communication techniques that leverage scientific
case in the US27,54 and Western Europe55, in Latin America the divide consensus around the existence of climate change may nudge mass
between climate change skeptics and non-skeptics is not primarily publics to engage in pro-climate action. While the existing literature
driven by political ideology. Instead, the pattern here is more similar to emphasizes the positive downstream influence of scientific consensus
findings from Central and Eastern Europe56, where political cleavages on climate-friendly policies and behavior64, the challenge lies in
play a minor role in shaping climate change perceptions. One possible effectively conveying this consensus to the public in societies char-
explanation is that people in less developed countries find it hard to acterized by low literacy rates and limited access to information on
see climate change through the prism of left-right or progressive- climate change. Across Latin America, the costs of accessing infor-
conservative divides, since the issue has been mostly restricted to mation are high and climate change is not politically salient. Taking
intra-elite debates due to its relatively low salience compared to other these factors into account, communicating the existence of scientific
pressing social problems57. These results bode well for the emergence consensus should be done through media that is easily accessible and
of broad pro-climate coalitions in Latin America, since pro-climate seen as largely reliable, as it is the case of radio and television65. As
action can garner public support from individuals across the political recent research has shown, if communicators using these media were
spectrum. to be able to show how individuals can contribute to solving the cli-
Our study also found weak associations between socio- mate challenge, their messages may help drive public climate action
demographic factors and perceptions of climate change. This result (e.g., voting, donating, protesting)66.
contradicts the theoretical literature suggesting that younger, more Last but not least, the current study is not without limitations.
educated, higher income, and female individuals are less likely to be First, the cross-sectional nature of this analysis does not allow for the
climate skeptics29. Empirically, the results also diverge from findings identification of causal relationships between variables, and there may
from the U.S58. and Western Europe (and to a lesser extent Eastern be endogenous relationships that were not accounted for. Further
Europe)3, where demographics are important factors shaping climate research using experimental designs is necessary to establish causality
change beliefs. The limited effects of certain demographic variables and fully explore the implications of the findings presented here.
(e.g., education and sex) even differ from findings from other devel- Second, we acknowledge that our definition of individualistic values as
oping areas of the world, in particular Africa, where education and personal attitudes towards the role of government in society is narrow,
gender are important correlates of the belief in anthropogenic climate even if it is widely used in the field4,67. Many other aspects of this
change59. Although future research needs to investigate what explains complex construct may influence climate change perceptions. Future
the weak socio-demographic associations in Latin America, these studies should therefore explore broader conceptualizations of indi-
results highlight that demographic factors can carry different mean- vidualistic values, such as the extent to which individuals prioritize
ings in different contexts, with important implications for under- their personal interests over the interests of society as a whole. This
standing beliefs about climate change. Finally, we are confident that approach may shed valuable light on the specific psychological
our results are not driven by our convenience sample, as they align mechanisms that underlie the relationship between individualism and
with studies using nationally representative samples in Latin America climate skepticism, as well as inform the development of more tar-
to assess other climate dimensions such as climate risk perception13 geted interventions to promote pro-climate behaviors. Third, drawing
and climate change concern12. on the insight from the present study that individual perceptions of the
Taken together, these results about the determinants of climate role of government in society shape climate change beliefs, future
change beliefs can potentially help improve communication strategies work should explore the role of ideological liberalism, adding an extra
aimed at promoting climate action. Although our findings only allow layer to the study of political orientations in climate psychology.
us to be speculative, existing theories in the scholarly literature can Fourth, our finding showing that personal experience with extreme
illuminate the path forward. First and foremost, since beliefs in Latin weather is weakly associated with climate change beliefs may result

from the fact that our question did not attribute weather events to Materials
climate change. As noted in Supplementary Information, we have Questionnaires were first elaborated in Portuguese and then translated
some reason to believe individuals make this attribution intuitively in into Spanish by a native speaker and taking into account translation/
Latin America, but future work should follow recent research68 and back-translation procedures. Concerns about comprehension and
explicitly test the consistency of these results. translatability were addressed by testing and piloting of ques-
In conclusion, this study has shed light on the correlates of climate tionnaires in all seven countries contemplated in the survey.
change perceptions in Latin America. Now, armed with a clearer
understanding of the variables associated with climate change per- Dependent variable measures (climate change perceptions)
ceptions in the region, pro-climate actors can develop more effective Building on previous research, we measure climate change percep-
policies, strategies, and campaigns to confront climate skepticism tions along three theoretically distinct but complementary
head-on, a crucial step towards mitigating the impacts of climate dimensions24,32. The distinction is relevant because scholars have
change and promoting a more sustainable future. Ultimately, by shown individuals to hold inconsistent cognitions around the exis-
building upon the findings of this study and continuing to investigate tence, the causes, and the consequences of climate change32, and
the correlates of climate change perceptions in diverse contexts, we evidence suggests each belief may be influenced by different
can deepen our understanding of the social and psychological pro- antecedents32,34. To measure belief in the existence of climate change17[,72
cesses that shape beliefs not only in this particular region but across participants were asked, “You may have heard the idea that the world’s
the Global South. climate is changing due to increases in average temperatures over the
past 150 years. What’s your personal opinion on this? Do you think that
Methods the world’s climate is changing?” Respondents could then choose one
Participants of the following responses: “Yes (1), No (0), Don’t know.” Respondents
The study included nationally diverse participants from seven Latin who answered “yes” or “no” were then asked: “How sure are you that
American countries (Argentina, Brazil, Chile, Colombia, Ecuador, climate change is [not] happening?” (0 = not sure at all, 3 = extremely
Peru, and Mexico), surveyed by the Netquest, an internet polling sure). Responses to these questions were recoded to create an eight-
company, between October and November 2021. Together, these point certainty scale (0 = extremely sure global warming is not hap-
countries account for over 80% of greenhouse gas emissions (GHF) in pening, 4 = don’t know, 8 = extremely sure global warming is hap-
Latin America. Netquest builds its online panels through an opt-in pening). Belief in the anthropogenic causes of climate change was
recruitment method, where respondents are randomly selected for assessed with the question, “Assuming climate change is happening,
survey invitation, using population quotas to produce nationally do you think it is….” Respondents selected one of the four options:
diverse samples. Netquest is certified with ISO 26362, an “Caused mostly by human activities, Caused by human activities and
international high-quality standard for online panels, and complies natural changes, Caused mostly by natural changes in the environ-
with the European Society for Opinion and Market Research ment, Neither because global warming isn’t happening.” Following
(ESOMAR). In general, this means that Netquest engages in a number standard practice, respondents who answered “caused mostly by
of quality control steps, including the exclusion of speeders or human activities” were coded as 1, while all others as 031,73. Finally, we
preventing duplicate responders. Additional information on data measure belief in the consequences of climate change with the standard
quality and sampling is available in Netquest technical information question used in the literature3: “How good or bad do you think the
reports69. impact of climate change will be on people across the world?” The
We chose Netquest because it provides comprehensive national question was measured using an 11-point scale, ranging from 0
panels in Latin America, which allow us to relatively approximate the (extremely bad) to 10 (extremely good).
national representativeness of the participating countries. In this
study, respondents were recruited to match the demographic com- Independent variables
position (e.g., quotas), particularly of sex, age, and education, laid Knowledge about climate change was measured using one question
out by the national census of each country surveyed. The final sample that embraces participants’ subjective perception of their own exper-
comprises 5400 participants, with all respondents above 18 years of tise (subjective knowledge), and another that focuses on respondents’
age. Each country surveyed had roughly 830 respondents. The only knowledge about the human causes of climate change (objective
exception was Ecuador, where just 421 respondents were interviewed knowledge). We measured subjective knowledge by asking respon-
due to panel coverage constraints. For demographic information dents the following question: “How much do you feel you know about
broken for each of the countries, see Supplementary Table 8. Overall, climate change?” Participants rated their perceived amount of
the sample of each country is balanced with respect to sex and age, knowledge on a four-point scale (1 = nothing, 2 = a little, 3 = a moderate
approximating closely to their respective official statistics (Supple- amount, 4 = a lot) but with a fifth option for people who “don’t know”.
mentary Table 11). Moreover, one main concern of online samples is Objective knowledge was measured by asking participants, “Indicate
that their respondents tend to be more educated than the general whether you think each of the following is a major cause of climate
population, a tendency that can be exacerbated in both lower change, or not a cause at all.” Six items were adopted from Guy et al.34,
and upper and middle-income countries70. In general terms, our which include three true causes (“Pollution/emissions from business
sample is well-distributed in terms of education, with the over- and industry”, “People driving their cars”, “Destruction of tropical
representation of educated participants more pronounced in Ecua- forests”) and three false causes (“Use of aerosol spray cans”, “Use of
dor and Peru (Supplementary Table 10). Nonetheless, we are chemicals to destroy insect pests”, and “Nuclear power generation”).
cautious and all the models estimated in the study include sample Responses of items were coded as correct (1) or incorrect (0) and
weights to adjust for sample representativeness. Finally, our country summed to create a total score ranging from 0 to 6.
samples are also broadly diverse in other important demographic Beliefs about science were measured through a question about
indicators, including race, religion, and income level (Supplementary scientific consensus around climate change (“consensus heuristic”),
Table 8), which are known to correlate with important dimensions of and another about trust in scientists (“source heuristic”). These two
climate change perceptions. More generally, online surveys in Latin measures comprise the two main heuristics about science and climate
America have been shown to by and large replicate results from change that have been implicated in the existing literature27. In the case
nationally representative samples71, reinforcing the reliability of our of scientific consensus about climate change, participants were asked,
sample. “Which comes closest to your own view?” The response options were:

“Most scientists think global warming is happening”, “A lot of dis- country. Following previous studies, items were measured on a five-
agreement among scientists about whether or not global warming is point scale (1 = never, 2 = once, 3 = twice, 4 = more than three, 5 = can’t
happening”, “Most scientists think global warming is not happening”, remember) and responses were combined and dichotomized to form a
and “I do not know enough to say.” Respondents who answered “Most dummy describing personal experience (0 = no experience, 1 =
scientists think global warming is happening” were coded as 1, and all experience)77.
other answers as 073. We measured trust in scientists by asking, “How Political ideology was measured using two different questions.
much do you trust scientists as a source of information about climate First, respondents were asked to rate how right or left they are on a ten-
change?” Participants rate their level of trust on a four-point scale (1 = point scale (1 = left, 10 = right). Right-wing individuals generally
strongly distrust; 4 = strongly trust)35. advocate for limited government intervention in the economy, and
To measure environmental values, we used a 4-item revised ver- traditional social values, while left-wing individuals typically support
sion from the New Ecological Paradigm scale based on previous government intervention to address economic inequality, social jus-
research74. The items include “Humans are severely abusing the tice issues, and expanded social welfare programs. Second, respon-
environment”, “The so-called ‘ecological crisis’ facing humankind has dents were asked to choose the option that better characterizes their
been greatly exaggerated”, “The earth is like a spaceship with very political values in a conservative-progressive dimension. The response
limited room and resources”, and “If things continue on their present options were rated on a five-point scale (1 = very progressive, 5 = very
course, we will soon experience a major ecological catastrophe.” Items conservative). Conservatism can be defined as a political ideology that
were responded to on a four-point scale (1 = strongly disagree; 4 = advocates for traditional values, limited government intervention, and
strongly agree). There was also a fifth option for people who “don’t the preservation of established institutions and social norms. By con-
know”. The participants who chose the “don’t know” option were trast, progressivism is a political ideology that promotes social and
treated as missing values on this scale. The NEP scale was constructed political change, often through government intervention, to address
by taking the mean scores across these four items. The original scale social injustices and improve the well-being of society. We opted to use
showed poor internal reliability across the whole sample (α = 0.37), the word “progressive” rather than “liberal” because “liberal” in Latin
which a closer analysis revealed to be driven by the answers to the America can be associated with the orthodox economic policy pre-
second item, “The so-called ‘ecological crisis’ facing humankind has ferences of the political right. “Progressive” (progresista in Spanish or
been greatly exaggerated”. Consequently, we conducted the analyses progressist in Portuguese) provides a better characterization of what
on the three positively worded items only, which makes the scale more the existing literature labels as “liberal” in this context.
reliable. This said, NEP continues to carry limitations, given its alpha The demographic variables include sex (binary: male or female),
(0.53) did not exceed the conventional 0.60 criteria. For this reason, age (in years), education level (ordinal: elementary (primary) or less;
we suggest caution in interpreting the results based on the NEP scale. high school or equivalent; and undergraduate or more), religion
Values were measured using the individualism and egalitarianism (Evangelical Christian/Traditional; Evangelical Protestant; Evangelical
cultural worldviews derived from cultural cognition theory. We follow Pentecostal; Evangelical Neo-Pentecostal; Other Evangelical denomi-
the standard definition of worldviews as mental constructs pertaining nations; Catholic; Kardecist/Spiritualist; Jewish; Agnostic; Atheist;
to the role of the state in society rather than more expansive under- Other Religion.), race (White; Black or Pardo); Indigenous; Other. In all
standings of the relationship between individuals and society as a countries (except for Brazil), we also include “Mestizo” as an option
whole4,67,75. We operationalized our individualism scale on the basis of choice, given that it is a racial classification present in these countries),
five items76, including the following: “If the government spent less time income-based on minimum wages (from 1 to 10 minimum wages
trying to fix everyone’s problems, we’d all be a lot better off”, “Our or more).
government tries to do too many things for too many people. We
should just let people take care of themselves”, “The government Statistical analysis
interferes too much in our everyday lives”, “Government regulation of To evaluate what factors determine climate change beliefs in Latin
business usually does more harm than good”, and “People should be America we estimated the following equation by ordinary least squares
allowed to make as much money as they can, even if it means some for each country and our whole sample:
make millions while others live in poverty.” All items were responded
0 0
to a four point-scale (1 = strongly disagree; 4 = strongly agree). The yi = β0 + Psychological Variablesi γ + Political Ideology Variablesi α
individualism scale is constructed by taking the mean scores across 0
+ SocioDemographic Variablesi μ + εi
these 5 items, which together showed acceptable internal reliability
ð1Þ
(α = 0.67). Note that the individualism scale has been reported in other
studies in the Global South with similar internal consistency, under- where yi is one of our three dependent variables (belief in the
scoring the reliability of this measure in our study67. existence of climate change, belief in the anthropogenic causes of
Egalitarianism worldview is also a composite measure based on climate change, and belief in the consequences of climate change)
five items76. Items include: “The world would be a more peaceful place for individual i; Psychological Variablesi is a vector of eight psy-
if its wealth were divided more equally among nations,” “In my ideal chological variables of individual i; Political Ideology Variablesi is a
society, all basic needs (food, housing, health care, education) would vector of two political ideology variables of individual i; and
be guaranteed by the government for everyone,” “I support govern- SocioDemographic Variablesi is a vector of six population-related
ment programs to get rid of poverty,” “Discrimination against mino- characteristics of individual i. Reference baseline for education is
rities is still a very serious problem in our society.” All items were “elementary (primary) or less”, for religion is “atheist” and for
responded to a four point-scale (1 = strongly disagree; 4 = strongly income is “0–1 minimum wages”. Standard errors presented in the
agree). As in the case of the individualism scale, the scale of egalitar- results are robust to heteroscedasticity and all estimations include
ianism was created by taking the mean scores across these five items. sample weights.
The scale showed relatively strong internal reliability across the overall For belief in the existence of climate change, the vector of para-
sample (α = 0.72). meters of interest, γ, α and μ, can be interpreted as changes in the
Personal experience with extreme weather events was assessed by climate change belief scale (0–8) given a unit increase in the covari-
asking respondents to recall how often in the last five years they ates. For belief in the anthropogenic causes of climate change, each
experienced extreme weather events (e.g., severe heat waves, parameter of the vectors γ, α and μ multiplied by one hundred can be
droughts, freak storms, flooding etc.) while residing in their home interpreted as percentage point changes in the probability of believing

climate change is mainly caused by human activity given a unit respondents with points they accumulate and then exchange for
increase in the binary covariates. For belief in the consequences of goods or cellphone minutes.
climate change, γ, α and μ multiplied by one hundred can be inter-
preted as percentage point changes in the probability of believing Reporting summary
climate change impacts will be negative given a unit increase in the Further information on research design is available in the Nature
binary covariates. Portfolio Reporting Summary linked to this article.
To gauge the robustness of our ordinary least squares estimates
we performed several tests by employing alternative methods to Data availability
estimate the parameters of our main equation. First, we estimated The processed data used in this study are available at the Harvard
three multilevel (hierarchical) models, one for each dependent vari- Dataverse as Spektor, Matias; Fasolin, Guilherme; Camargo, Juliana,
able, with a random intercept specified at the country-level (Supple- 2022, “Replication Data for: Climate Change Beliefs and their Corre-
mentary Table 25). Second, we estimated three fixed-effects models, lates in Latin America”, https://doi.org/10.7910/DVN/F4KNNS, Har-
one for each dependent variable, by including country-level effects vard Dataverse, V2. The following software has been used to process
(Supplementary Table 26). Both of these approaches allow us to con- the data: the R software (version 4.2.2) The integrated development
trol for possible time-invariant unobserved heterogeneity between environment (IDE) of choice was RStudio.
respondents of each country. Third, we used an ordinal logistic model
to regress belief in the existence of climate change on the three vectors Code availability
of correlates (Supplementary Table 27). Finally, we estimated two The code used in this study are available at the Harvard Dataverse as
binary logistics models regressing the belief in the anthropogenic Spektor, Matias; Fasolin, Guilherme; Camargo, Juliana, 2022, “Repli-
causes of climate change and belief in the consequences of climate cation Data for: Climate Change Beliefs and their Correlates in Latin
change also on the three vectors (Supplementary Tables 28 and 29). America”, https://doi.org/10.7910/DVN/F4KNNS, Harvard Data-
Overall, the results are substantially similar irrespective of the alter- verse, V2.
native specifications of our main equation.
References
Robustness checks 1. Capstick, S. & Whitmarsh, L. Perceptions of climate change. In (eds
Besides employing alternative methods to estimate the parameters of Clayton, S. & Manning, C.) Psychology and Climate Change: Human
our main equation (see Methods section), several complementary tests Perceptions, Impacts, and Responses. 13–33 (Elsevier Academic
were performed to gauge the robustness of our findings. First, we Press, 2018).
provide evidence of the absence of multicollinearity. Supplementary 2. Clayton, S. et al. Psychological research and global climate change.
Table 16 reveals that the variance inflation factor (VIF)—a measure of Nat. Clim. Chang. 5, 640–646 (2015).
the amount of multicollinearity in regression analysis—is low for all our 3. Poortinga, W., Whitmarsh, L., Steng, L., Bohm, G. & Fisher, S.
independent variables. All the VIPs are below 1.5, which relieves our Climate change perceptions and their individual-level determi-
concern about the potential correlation between the main variables. nants: a cross-European analysis. Glob. Environ. Change 55,
Second, even after finding correlation among the independent 25–35 (2019).
variables to be low, we performed F-tests to mitigate concerns that the 4. Hornsey, M. J. The role of worldviews in shaping how people
independent variables might not be jointly statistically significant. appraise climate change. Curr. Opin. Behav. Sci. 42, 36–41 (2021).
Supplementary Table 17 shows the results of the estimations of the 5. Howe, P., Mildenberger, M., Marlon, J. & Leiserowitz, T. Geographic
joint significance test for all the models we ran in the study, con- variation in opinion on climate change at state and local scales in
sidering all independent variables. We find that all independent vari- the USA. Nat. Clim. Chang. 5, 569–503 (2015).
ables are jointly significant in each of the four main OLS models 6. Leiserowitz, A. et al. Climate change in the American mind, Sep-
estimated for the whole sample. We also ran the F-test of joint sig- tember 2021. (Yale University and George Mason University. Yale
nificance for different groups of independent variables separately Program on Climate Change Communication, New Haven,
(Supplementary Tables 18–20). For each different group of indepen- CT, 2021).
dent variables, we found that they are jointly significant in each of the 7. Capstick, S. B. & Pidgeon, N. F. What is climate change scepticism?
four main OLS models estimated for the whole sample. These results Examination of the concept using a mixed methods study of the UK
indicate the models fit the data well and suggest that the independent public. Glob. Environ. Change 24, 389–401 (2014).
variables included in the model are not irrelevant. 8. Whitmarsh, L. Scepticism and uncertainty about climate change:
Finally, to attenuate concerns that some of our results might be dimensions, determinants and change over time. Glob. Environ.
statistically significant by chance, we ran a Benjamini-Hochberg test to Change 21, 690–700 (2011).
adjust p-values for multiple hypothesis tests78. As shown in Supple- 9. Engels, A., Huther, O., Schafer, M. & Held, H. Public climate-change
mentary Table 21, the overall results of the estimations using the scepticism, energy preferences and political participation. Glob.
adjusted p-values corroborate the results using standard p-values. All Environ. Change 23, 1018–1027 (2013).
the statistical interpretations remain roughly the same, even after 10. Leviston, Z., Leitch, A., Greenhill, M., Leonard, R. & Walker, I. Aus-
accounting for the multiple hypothesis tests. This suggests that our tralians' Views of Climate Change (Commonwealth Scientific and
inferences about statistical significance are not an artifact of the Industrial Research Organisation, 2011).
number of predictors in our regression models. 11. Capstick, S., Whitmarsh, L., Poortinga, W., Pidgeon, N. & Upham, P.
International trends in public perceptions of climate change over
Ethics statement the past quarter century. WIREs Clim. Change 6, 35–61 (2015).
This study was approved by the Ethical Review Committee of the 12. Lewis, G. B., Palm, R. & Feng, B. Cross-national variation in deter-
Fundação Getulio Vargas (FGV) (Ethics no. 053/2021). All participants minants of climate change concern. Environ. Politics 28,
informed voluntary consent with an IRB-approved consent protocol 793–821 (2019).
before being allowed to proceed to the full questionnaire. The survey 13. Lee, T. M., Markowitz, E. M., Howe, P. D., Ko, C. Y. & Leiserowitz, A. A.
did not collect identifying information about respondents and/or use Predictors of public climate change awareness and risk perception
any type of deception. The panel provider modestly compensated the around the world. Nat. Clim. Chang. 5, 1014–1020 (2015).

14. Hornsey, M. J., Harris, E. A. & Fielding, K. S. Relationships among perceptions of global warming. Public Underst. Sci. 23,
conspiratorial beliefs, conservatism and climate scepticism across 866–883 (2014).
nations. Nat. Clim. Chang. 8, 614–620 (2018). 36. McCright, A. M., Dunlap, R. E. & Xiao, C. Perceived scientific
15. Levi, S. Country-level conditions like prosperity, democracy, and agreement and support for government action on climate change
regulatory culture predict individual climate change belief. Com- in the USA. Climatic Change 119, 511–518 (2013).
mun. Earth Environ. 2, 1–10 (2021). 37. Demski, C., Capstick, S., Pidgeon, N., Sposato, R. G. & Spence, A.
16. Azócar, G. et al. Climate change perception, vulnerability, and Experience of extreme weather affects climate change mitigation
readiness: inter-country variability and emerging patterns in Latin and adaptation responses. Climatic Change 140, 149–64 (2017).
America. J. Environ. Stud. Sci. 11, 23–36 (2021). 38. Myers, T. A., Maibach, E. W., Roser-Renouf, C., Akerlof, K. & Leiser-
17. Goldberg, M., Gustafson, A., Ballew, M., Rosenthal, S. & Leiserowitz, owitz, A. A. The relationship between personal experience and
A. Identifying the most important predictors of support for climate belief in the reality of global warming. Nat. Clim. Chang. 3,
policy in the United States. Behav. Public Policy 5, 480–502 (2021). 343–347 (2013).
18. Spence, A., Poortinga, W., Butler, C. & Pidgeon, N. S. Perceptions of 39. Kahan, D. Fixing the communication failure. Nature 463,
climate change and willingness to save energy related to flood 296–297 (2010).
experience. Nat. Clim. Chang. 1, 46–49 (2011). 40. Kahan, D. M. in Handbook of Risk Theory (eds Roeser, S. et al.)
19. IPCC. Summary for Policymakers (eds Pörtner, H.-O. et al.) in Cli- 725–759 (Springer, Dordrecht, 2012).
mate Change 2022: Impacts, Adaptation, and Vulnerability. Con- 41. Kahan, D. M. et al. The Tragedy of the Risk-Perception Commons:
tribution of Working Group II to the Sixth Assessment Report of the Culture Conflict, Rationality Conflict, and Climate Change. Temple
Intergovernmental Panel on Climate Change. (Cambridge Uni- University Legal Studies Research Paper No. 2011-26 (2011).
versity Press, Cambridge, UK and New York, USA, 2022). 42. Druckman, J. N. & McGrath, M. C. The evidence for motivated rea-
20. Hopkins, D. Country comparisons. Nat. Clim. Chang. 5, 975 (2015). soning in climate change preference formation. Nat. Clim. Chang.
21. Crona, B., Wutich, A., Brewis, A. & Garint, M. Perceptions of climate 9, 111–119 (2019).
change: linking local and global perceptions through a cultural 43. Bayes, R., Druckman, J. N., Goods, A. & Molden, D. C. When and how
knowledge approach. Climatic Change 119, 519–531 (2013). different motives can drive motivated political reasoning. Political
22. Editorial. Replication studies hold the key to generalization. Nat. Psychol. 41, 1031–1052 (2020).
Commun. 13, 7004 (2022). 44. Krys, K., Vignoles, V. L., de Almeida, I. & Uchida, Y. Outside the
23. Hornsey, M. J. & Fielding, K. L. Understanding (and reducing) inac- “cultural binary”: understanding why latin american collectivist
tion on climate change. Soc. Issue Policy Rev. 14, 3–35 (2020). societies foster independent selves. Perspect. Psychol. Sci. 17,
24. Poortinga, W., Spence, A., Whitmarsh, L., Capstick, S. & Pidgeon, N. 1166–1187 (2022).
F. Uncertain climate: an investigation into public scepticism about 45. Keefer, P. & Scartascini, C. Trust: the key to social cohesion and
anthropogenic climate change. Glob. Environ. Change 21, growth in Latin America and the Caribbean. Inter American Devel-
1015–1024 (2011). opment Bank, (2022).
25. Heath, Y. & Gifford, R. Free-market ideology and environmental 46. Fukuyama, F. Trust: The Social Virtues and the Creation of Pros-
degradation: the case of belief in global climate change. Environ. perity. (Free Press, 1995).
Behav. 38, 48–71 (2006). 47. Bodor, A., Varjú, V. & Grunhut, Z. The effect of trust on the various
26. Bostrom, A. et al. Causal thinking and support for climate change dimensions of climate change attitudes. Sustainability 12,
policies: International survey findings. Glob. Environ. Change 22, 1–19 (2020).
210–222 (2012). 48. Kahan, D. M., Jenkins-Smith, H. & Braman, D. Cultural cognition of
27. Hornsey, M. J., Harris, E. A., Bain, P. G. & Fielding, K. S. Meta-analyses scientific consensus. J. Risk Res. 14, 147–174 (2011).
of the determinants and outcomes of belief in climate change. Nat. 49. Lewandowsky, S., Gignac, G. E. & Vaughan, S. The pivotal role of
Clim. Chang. 6, 622–626 (2016). perceived scientific consensus in acceptance of science. Nat. Clim.
28. McCright, A. & Dunlap, R. E. The politization of climate change and Change 3, 399–404 (2013).
polarization in the American public’s views of global warming, 50. Cook, J. et al. Consensus on consensus: a synthesis of consensus
2001–2010. Sociological Q. 52, 155–194 (2011). estimates on human-caused global warming. Environ. Res. Lett. 11,
29. Milfont, T. L., Milojev, P., Greaves, L. M. & Sibley, C. G. Socio- 048002 (2016).
structural and psychological foundations of climate change beliefs. 51. Shi, J. et al. Knowledge as a driver of public perceptions about
N.Z. J. Psychol. 44, 17–30 (2015). climate change reassessed. Nat. Clim. Change 6, 759–762 (2016).
30. Dunlap, R. E., Van Liere, K. D., Mertig, A. G. & Jones, R. E. Measuring 52. Ajzen, I., Joyce, N., Sheikh, S. & Cote, N. G. Knowledge and the
endorsement of the new ecological paradigm: a revised NEP scale. prediction of behavior: the role of information accuracy in the
J. Soc. Issues 56, 425–442 (2000). theory of planned behavior. Basic Appl. Soc. Psychol. 33,
31. Ziegler, A. Political orientation, environmental values, and climate 101–117 (2011).
change beliefs and attitudes: an empirical cross-country analysis. 53. Howe, P. D., Marlon, J. D., Mildenberger, M. & Shield, B. S. How will
Energy Econ. 63, 144–153 (2017). climate change shape perceptions? Environ. Res. Lett. 14,
32. Van Valkengoed, A. M., Steg, L. & Perlaviciute, G. Development and 113001 (2019).
validation of a climate change perceptions scale. J. Environ. Psy- 54. Campbell, E., Kotcher, J., Maibach, E., Rosenthal, S. & Leiserowitz,
chol. 76, 101652 (2021). A. Predicting the importance of global warming as a voting issue
33. Bord, R. J., O’Connor, R. E. & Fisher, A. In what sense does the public among the registered voters in the United States. Curr. Res. Ecol.
need to understand global climate change? Public Underst. Sci. 9, Soc. Psychol. 2, 100008 (2021).
205–218 (2000). 55. McCright, A. M., Dunlap, R. E. & Marquart-Pyatt, S. T. Political
34. Guy, S., Kashima, Y., Walker, I. & O’Neill, S. Investigating the effects ideology and views about climate change in the European Union.
of knowledge and ideology on climate change beliefs. Eur. J. Soc. Environ. Politics 25, 338–358 (2016).
Psychol. 44, 421–429 (2014). 56. Chaisty, P. & Whiteeld, S. Attitudes towards the environment: Are
35. Hmielowski, J. D., Feldman, L., Myers, T. A., Leiserowitz, A. & Mai- post-communist societies (still) different? Environ. Politics 24,
bach, E. An attack on science? Media use, trust in scientists, and 598–616 (2015).

57. Zechmeister, E. & Corral, M. Individual and contextual constraints 76. Leiserowitz, A. Climate change risk perception and policy pre-
on ideological labels in Latin America. Comp. Political Stud. 46, ferences: the role of affect, imagery, and values. Climatic Change
675–701 (2013). 77, 45–72 (2006).
58. McCright, A. M. & Dunlap, R. E. Cool dudes: the denial of climate 77. Van der Linden, S. The social-psychological determinants of cli-
change among conservative white males in the United States. Glob. mate change risk perceptions: towards a comprehensive model. J.
Environ. Change 24, 1163–1172 (2011). Environ. Psychol. 41, 112–124 (2015).
59. Simpson, N. P. et al. ClimateChange Literacy Africa. Nat. Clim. 78. Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a
Chang. 11, 937–944 (2021). practical and powerful approach to multiple testing. J. R. Stat. Soc.:
60. Hornsey, M. J. & Lewandowsky, S. A toolkit for understanding Ser. B (Methodol.) 57, 289–300 (1995).
and addressing climate scepticism. Nat. Hum. Behav. 6,
1454–1464 (2022). Acknowledgements
61. Bain, P. G. et al. Co-benefits of addressing climate change can The authors would like to acknowledge support from the Applied
motivate action around the world. Nat. Clim. Change 6, Research Network at FGV. M.S. would also like to thank Brazil’s
154–157 (2016). National Research Council for research support. We thank Renan
62. Boon-Falleur, M. et al. Leveraging social cognition to promote Chicarelli Marques and Vitória Totti Salgado for valuable research
effective climate change mitigation. Nat. Clim. Chang. 12, assistance.
332–338 (2022).
63. Goldberg, M. H., Gustafson, A., Rosenthal, S. A. & Leiserowitz, A. Author contributions
Shifting Republican views on climate change through targeted M.S. and G.N.F. designed and fielded the survey. J.C. conducted the
advertising. Nat. Clim. Change 11, 573–577 (2021). statistical analyses. M.S. and G.N.F. wrote the final manuscript.
64. Bayes, R., Bolsen, T. & Druckman, J. N. A research agenda for cli-
mate change communication and public opinion: the role of sci- Competing interests
entific consensus messaging and beyond. Environ. Commun. 17, The authors declare no competing interests.
16–34 (2020).
65. Goldberg, M. H. et al. The experience of consensus: video as an Additional information
effective medium to communicate scientific agreement on climate Supplementary information The online version contains
change. Sci. Commun. 41, 659–673 (2019). supplementary material available at
66. Doherty, K. & Webler, T. Social norms and efficacy beliefs drive the https://doi.org/10.1038/s41467-023-42729-x.
Alarmed segment’s public-sphere climate actions. Nat. Clim.
Change 6, 879–884 (2016). Correspondence and requests for materials should be addressed to
67. Thaker, J., Smith, N. & Leiserowitz, A. Global Warming Risk Per- Matias Spektor.
ceptions in India. Risk Anal. 40, 2481–2897 (2020).
68. Ogunbode, C. A., Demski, C., Capstick, S. B. & Gennaro, R. Attri- Peer review information Nature Communications thanks Jack Zhou and
bution matters: revisiting the link between extreme weather the other, anonymous, reviewers for their contribution to the peer review
experience and climate change mitigation responses. Glob. of this work.
Environ. Change 54, 31–39 (2019).
69. Netquest answers to ESOMAR 28 questions to help online research Reprints and permissions information is available at
buyers (Netquest, 2015). http://www.nature.com/reprints
70. The World Bank. World Bank Country and Lend Groups. https://
datahelpdesk.worldbank.org/knowledgebase/articles/906519- Publisher’s note Springer Nature remains neutral with regard to jur-
world-bank-country-and-lending-groups#:~:text=For%20the% isdictional claims in published maps and institutional affiliations.
20current%202024%20fiscal,those%20with%20a%20GNI%
20per. (2023). Open Access This article is licensed under a Creative Commons
71. Bush, S. S., & Clayton, A. Facing Gender: Gender and Climate Attribution 4.0 International License, which permits use, sharing,
Change Attitudes Worldwide. Am. Political Sci. Rev. First View, 1–18 adaptation, distribution and reproduction in any medium or format, as
(Forthcoming). long as you give appropriate credit to the original author(s) and the
72. Bergquist, P., Mildernberg, M. & Stokes, L. C. Combining climate, source, provide a link to the Creative Commons license, and indicate if
economic, and social policy builds up support for climate action in changes were made. The images or other third party material in this
the US. Environ. Res. Lett. 15, 055019 (2020). article are included in the article’s Creative Commons license, unless
73. Howe, P., Mildenberger, M., Marlon, J. R. & Leiserowitz, A. Geo- indicated otherwise in a credit line to the material. If material is not
graphic variation in opinions on climate change at state and local included in the article’s Creative Commons license and your intended
scales in the USA. Nat. Clim. Chang. 5, 596–603 (2015). use is not permitted by statutory regulation or exceeds the permitted
74. Budescu, D., Por, H. H., Broomell, S. B. & Smithson, M. The inter- use, you will need to obtain permission directly from the copyright
pretation of IPCC probabilistic statements around the world. Nat. holder. To view a copy of this license, visit http://creativecommons.org/
Clim. Chang. 4, 508–512 (2014). licenses/by/4.0/.
75. Kahan, D. M. et al. The polarizing impact of science literacy and
numeracy on perceived climate change risks. Nat. Clim. Chang. 2, © The Author(s) 2023
732–735 (2012).

Spektor Et Al, Climate Change Beliefs (2023)

Uploaded by

Copyright:

Available Formats

Spektor Et Al, Climate Change Beliefs (2023)

Uploaded by

Document Information

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

Spektor Et Al, Climate Change Beliefs (2023)

Uploaded by

Copyright:

Available Formats

Article https://doi.org/10.

Climate change beliefs and their correlates in

Accepted: 20 October 2023

The ability of climate skeptics to block climate action depends on prevailing

ogy. Yet, little is known about climate-related beliefs elsewhere. We address

Nature Communications | (2023)14:7241 1

B. Political ideology and

Nature Communications | (2023)14:7241 2

B. Political ideology and

−0.2 −0.1 0.0 0.1 −0.2 −0.1 0.0 0.1

Nature Communications | (2023)14:7241 3

B. Political ideology and

Nature Communications | (2023)14:7241 4

Nature Communications | (2023)14:7241 5

Nature Communications | (2023)14:7241 6

Nature Communications | (2023)14:7241 7

Nature Communications | (2023)14:7241 8

Nature Communications | (2023)14:7241 9

Nature Communications | (2023)14:7241 10

Nature Communications | (2023)14:7241 11

You might also like