Epistemic Communities and Public Support for the Paris Agreement on Climate Change

• DOI: 10.1177/1065912920946400
• Published date: 01 December 2021
• Date: 01 December 2021
• Subject Matter: Articles

2021, Vol. 74(4) 866 –881

https://doi.org/10.1177/1065912920946400

Political Research Quarterly

© 2020 University of Utah

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1065912920946400

journals.sagepub.com/home/prq

What role can policy experts play in shaping support for

international cooperation? Is public opinion on interna-

tional cooperation sensitive to the views of policy experts

on issues related to international cooperation? Historically,

the public and political leaders have been thought to yield

to the views of subject matter experts (Haas 1992b).

Currently, when the “death of expertise” (Nichols 2017)

and “fake news” are highlighted both as a threat to demo-

cratic norms and wielded as a political cudgel, the effec-

tiveness of cues from policy experts may be muted.

Furthermore, as concerns about “backlash effects” or

shifts in opinion away from counter-attitudinal informa-

tion (Nyhan and Reifler 2010) rise, experts on the policy

problems that global cooperation is meant to address may

question the usefulness and be concerned about the

potentially counterproductive effects of sharing their

expertise with the public and policy makers.

We therefore ask, can policy experts or epistemic

communities (ECs)—groups of individuals with an

“authoritative claim to policy relevant knowledge” in a

given policy domain (Haas 1992b)—affect public sup-

port for particular policy alternatives? And to what extent

does the level of consensus among EC members affect

their ability to do so? The answers to these questions,

especially in the context of ascendant anti-globalization

political movements, are still undefined. This is the case

despite a large literature on the role of political and soci-

etal elites in shaping public opinion. There is strong evi-

dence, for example, that a long list of elites, including

religious leaders (Adkins et al. 2013), elected political

leaders (Bolsen, Druckman, and Cook 2014), journalists

(Groeling and Baum 2009), political party leaders (Lenz

2013), and celebrities (Marsh, Hart, and Tindall 2010),

can shape and mobilize public opinion on issues for

which they may have little or no domain-specific knowl-

edge. However, we are on less sure footing when it

comes to understanding public reactions to the views of

policy experts and ECs on some of society’s most press-

ing problems. This is, ironic since scholars have long

argued that ECs are uniquely situated to provide credible

946400

PRQXXX10.1177/1065912920946400Political Research QuarterlyMaliniak et al.

research-article2020

1William & Mary, Williamsburg, VA, USA

2The University of North Carolina at Chapel Hill, USA

3University of Georgia, Athens, USA

Corresponding Author:

Ryan Powers, School of Public and International Affairs, University of

Georgia, 202 Herty Drive, Athens, GA 30602-1492, USA.

Email: ryan.powers@uga.edu. Web: http://ryanpowers.net

Epistemic Communities and Public

Support for the Paris Agreement

on Climate Change

Daniel Maliniak1, Eric Parajon2, and Ryan Powers3

Abstract

We study how informing the public about the views of international policy experts shapes public support for

international cooperation. Using survey experiments, we test whether variation in levels of support among experts with

differing types of domain-specific knowledge can shape public support for a recent and politically salient international

treaty: the UNFCCC COP21 Paris Climate Agreement. Our results show that the public is, under certain conditions,

deferential to the views of experts, with respondents reporting increasingly higher levels of support for the COP21

agreement as support among experts increased. In addition, we provide suggestive evidence that domain-specific

expertise matters: When it comes to support for the COP21 agreement, the public is most sensitive to the views of

climate scientists, while exposure to the views of international relations and international economics experts have

less dramatic and less consistent effects. Despite these results, we find that it is exposing the public to information

about opposition to a proposed treaty among members of relevant epistemic communities that has greatest and most

consistent effects. Our findings thus provide new insight into the conditions under which epistemic communities can

shape public support for particular policy alternatives.

Keywords

international cooperation, public opinion, environmental politics, survey experiments

Article

Maliniak et al. 867

2 Political Research Quarterly 00(0)

and actionable insight into the causes and consequences

of—and solu tions to—emerging global public goods

problems like reducing nuclear proliferation (Adler 1992),

protecting the ozone layer (Haas 1992a), and forestalling

anthropogenic climate change (Allan 2017).

In this paper, we draw on both the elite cuing and ECs

literatures to develop expectations about how, in sharing

their views publicly, groups of elites with specialized

and policy-relevant knowledge can affect mass support

for particular policy alternatives. Our expectation is that

the influence of ECs is conditional both on the level of

domain-specific knowledge that the public believes the

group to have and on the group’s prevailing policy

preferences.

We test these expectations using an original survey

experiment designed to measure the effect of the views

of different ECs on public opinion in the context of the

highly politicized policy debate over continued U.S.

participation in the COP21 climate agreement. We focus

on this case because the agreement was politically

salient and had a genuinely uncertain future when we

were designing and fielding our survey. While President

Obama had signed the agreement, there was some

debate (in the media, if not in legal circles) over whether

the agreement needed Congressional approval. The

2016 Presidential election added to uncertainty about

the agreement’s future because while Hillary Clinton

supported the agreement, Donald Trump vehemently

opposed it. This case thus allows us to gauge the effect

of exposing the public to information about the views of

different types of knowledge experts on support for a

politically salient and strongly contested international

agreement.

Our study relies on a experimental manipulation in

which we present respondents with results from a fictional

survey of scholars of climate science, international rela-

tions (IR), or international economics (IE) about their sup-

port for the United States joining the COP21 agreement.1

We randomly vary the level of support among scholars so

that some respondents learn that scholars overwhelmingly

support the agreement, while other respondents learn that

scholars are overwhelmingly opposed to it. In a third treat-

ment group, respondents learn that scholars are split. We

then ask respondents to report their own level of support

for the agreement. Following the experiment, we provided

respondents with a full debrief which provided informa-

tion on the Paris agreement, anthropogenic climate

change, and, where available, the actual views of experts

on these issues.

The results that we report below show that relative to

a control group that received no information about the

views of experts, respondents who learned that experts

are opposed or split on the agreement were less likely to

support it, with those in the opposed treatment being least

supportive. While those learning that scholars support the

agreement were more supportive of the agreement than

those in the control group, the estimated effect size was

small and not distinguishable from zero. This stepwise

pattern of results is consistent with a public that is defer-

ential to the expertise of scholars, but not unconditionally

so. We also find suggestive evidence that domain exper-

tise matters: The public is most receptive to information

about the views of climate scientists.

We make three contributions. First, we test theories of

ECs and policy change with implications for public opin-

ion at the individual level. We argue that endorsements

from ECs contain important “knowledge cues” that signal

to the public the level of specialized knowledge that

members of the EC have about a specific policy.

Consistent with past work on ECs, we argue that a given

EC will be most effective in moving the public when the

public perceives them as having domain-specific and

policy-relevant knowledge on a given issue and when

there is broad agreement among the community on the

relative merit of a particular policy option.

Second, we show that efforts to educate the public

about the views of policy experts can have important

effects on support for particular policy alternatives.

Increasingly, social scientists are asked to “bridge the

gap” between the academy and the public by writing op-

eds, blogging, or otherwise engaging with the public and

policy makers. In general, these calls are premised on the

belief that these audiences will respond productively to

learning about the wisdom or folly of particular policy

alternatives from experts. The specialized issue domain

knowledge that scholars possess and the relative profes-

sional independence that they enjoy ostensibly make their

views and recommendations on public policy questions

more reliable and credible than those of partisan actors,

industry lobbyists, or other interested parties. There is

concern, however, that perceptions of scholars as ivory

tower elites who are out of touch with the practical reali-

ties of policy implementation may cause the views of

scholars to be discounted or ignored (Nichols 2017; Walt

2012). In contrast and consistent with other recent work,

we find that informing the public about the views of ECs

toward a given policy proposal generally moves public

opinion in the direction of the views of experts. On aver-

age, those exposed to a treatment condition in which they

learned that an EC opposes a given policy are less likely

to support the policy than those assigned to a control con-

dition which provided no information about EC views.

This effect is most dramatic and consistent when the EC

in question has domain-specific knowledge about the

issue at hand.

Third, we provide evidence that even for a politically

salient issue like the COP21 climate agreement in the

midst of a presidential election, there is little evidence of