Correcting misinformation about climate change: the impact of partisanship in an experimental setting

Abstract

Misperceptions of the scientific consensus on climate change are an important problem in environmental policy. These misperceptions stem from a combination of ideological polarization and statements from prominent politicians who endorse information contradicting or misrepresenting the scientific consensus on climate change. Our study tests a source credibility theory of correction using different partisan sources of information in a survey experiment. We find that corrections from Republicans speaking against their partisan interest are most likely to persuade respondents to acknowledge and agree with the scientific consensus on anthropogenic climate change. The extent of these effects vary by the partisanship of the recipient. Our results suggest that the partisan gap on climate change can be reduced by highlighting the views of elite Republicans who acknowledge the scientific consensus on anthropogenic climate change.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2

Notes

  1. 1.

    All replication materials for this study are available at http://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/KV6S5V

  2. 2.

    While the results from the pilot and the experiment reported in this paper are consistent in terms of the significance of effects, this pilot did not control for the content of the correction message.

  3. 3.

    AMT participant pools offer more demographic diversity than student or in-person convenience samples (Buhrmester et al. 2011, Paolacci and Chandler 2014). Although this population is more liberal, irreligious, and interested in politics than the average American, AMT samples have been used to replicate well-established findings in social psychology and political science (e.g., Berinsky et al. 2012). Recognizing that sources like AMT are different from conventional techniques for estimating dynamics of public opinion, our study seeks to gauge differences in responses to randomized treatments applied to an array of participants, rather than inferring differences in observed population frequencies from a sample.

  4. 4.

    McCright et al. (2016) find that an “ACC denial” treatment significantly reduces MTurk respondents’ reported agreement with climate science. Our denial statement is of similar length and phrasing (155 words, compared to 142 words in McCright et al.) to expose respondents to the misinformation.

  5. 5.

    In our pilot study used actual quotes from real politicians from the two parties (or a climate scientist). We found similar results and comparable effects sizes as those we report here. However, because the content and source of the messages co-varied, the pilot could not isolate the source effect.

  6. 6.

    All results discussed include leaners as party identifiers.

  7. 7.

    The coefficient differences and their 95% confidence intervals are estimated using the “margins, contrast” procedure in Stata 14.2, and the (default) delta method of standard error estimation.

  8. 8.

    This is one possible limitation of the study: the control group received a substantially smaller treatment (~150 words) while respondents in other groups received a substantially longer treatment including the misinformation statement and the correction.

References

  1. Achen CH, Bartels LM (2016) Democracy for realists: why elections do not produce responsive government. Princeton University Press, Princeton

    Google Scholar 

  2. Ahler DJ (2014) Self-fulfilling misperceptions of public polarization. J Politics 76(3):607–620

    Article  Google Scholar 

  3. Bastardi A, Uhlmann EL, Ross L (2011) Wishful thinking: belief, desire, and the motivated evaluation of scientific evidence. Psychol Sci 22(6):731

    Article  Google Scholar 

  4. Berinsky, A. (2017). Rumors and health care reform: experiments in political misinformation. British Journal of Political Science. doi:https://doi.org/10.1017/S0007123415000186

  5. Berinsky AJ, Huber GA, Lenz GS (2012) Evaluating online labor markets for experimental research: Amazon.com’s mechanical turk. Polit Anal 20(3):351–368

    Article  Google Scholar 

  6. Boussalis C, Coan TG (2016) Text-mining the signals of climate change doubt. Glob Environ Chang 36:89–100

    Article  Google Scholar 

  7. Brulle RJ (2014) Institutionalizing delay: foundation funding and the creation of US climate change counter-movement organizations. Clim Chang 122(4):681–694

    Article  Google Scholar 

  8. Buhrmester M, Kwang T, Gosling SD (2011) Amazon’s Mechanical Turk a new source of inexpensive, yet high-quality, data? Perspect Psychol Sci 6(1):3–5

    Article  Google Scholar 

  9. Calvert RL (1985) The value of biased information: a rational choice model of political advice. J Politics 47(2):530–555

    Article  Google Scholar 

  10. Chong D, Druckman JN (2007) Framing theory. Annu Rev Polit Sci 10:103–126

    Article  Google Scholar 

  11. Cook J, Oreskes N, Doran PT, Anderegg WR, Verheggen B, Maibach EW et al (2016) Consensus on consensus: a synthesis of consensus estimates on human-caused global warming. Environ Res Lett 11(4):048002

    Article  Google Scholar 

  12. Coppock, A. (2018). Generalizing from survey experiments conducted on Mechanical Turk: a replication approach. Political Science Research and Methods. Forthcoming

  13. Ding D, Maibach EW, Zhao X, Roser-Renouf C, Leiserowitz A (2011) Support for climate policy and societal action are linked to perceptions about scientific agreement. Nat Clim Chang 1(9):462–466

    Article  Google Scholar 

  14. Ditto PH, Lopez DF (1992) Motivated skepticism: use of differential decision criteria for preferred and nonpreferred conclusions. J Pers Soc Psychol 63(4):568

    Article  Google Scholar 

  15. Druckman JN (2001) On the limits of framing effects: who can frame? J Polit 63(4):1041–1066

    Article  Google Scholar 

  16. Druckman, J. N., & Lupia, A. (2016). Preference change in competitive political environments. Annual Review Political Science, 19

  17. Dunlap, R. E., & Jacques, P. J. (2013). Climate change denial books and conservative think tanks: exploring the connection. Am Behav Sci, 0002764213477096

  18. Dunlap RE, McCright AM, Yarosh JH (2016) The political divide on climate change: partisan polarization widens in the US. Environ: Sci Policy Sustainable Development 58(5):4–23

    Google Scholar 

  19. Egan, P. J., & Mullin, M. (2017). Climate change: US public opinion. Annual Review of Political Science, 20(1)

  20. Gauchat G (2012) Politicization of science in the public sphere: a study of public trust in the United States, 1974 to 2010. Am Sociol Rev 77(2):167–187

    Article  Google Scholar 

  21. Guber DL (2013) A cooling climate for change? Party polarization and the politics of global warming. Am Behav Sci 57(1):93–115

    Article  Google Scholar 

  22. Kahan DM (2015) Climate-science communication and the measurement problem. Political Psychology 36(S1):1–43

    Article  Google Scholar 

  23. Kahan DM, Jenkins-Smith H, Braman D (2011) Cultural cognition of scientific consensus. J Risk Res 14(2):147–174

    Article  Google Scholar 

  24. Kahneman D, Tversky A (1996) On the reality of cognitive illusion. Psychol Rev 103(3):582–591

    Article  Google Scholar 

  25. Krosnick JA, Holbrook AL, Lowe L, Visser PS (2006) The origins and consequences of democratic citizens’ policy agendas: a study of popular concern about global warming. Clim Chang 77(1–2):7–43

    Article  Google Scholar 

  26. Kuklinski JH, Quirk PJ, Jerit J, Schwieder D, Rich RF (2000) Misinformation and the currency of democratic citizenship. J Polit 62(3):790–816

    Article  Google Scholar 

  27. Kunda, Z. (1990). The case for motivated reasoning. Psychological bulletin, 108(3), 480

  28. Layzer, J. A. (2012). Open for business: conservatives’ opposition to environmental regulation. MIT Press

  29. Levendusky M, Malhotra N (2016) Does media coverage of partisan polarization affect political attitudes? Political Communication 33(2):283–301

    Article  Google Scholar 

  30. Lewandowsky S, Gignac GE, Vaughan S (2013) The pivotal role of perceived scientific consensus in acceptance of science. Nat Clim Chang 3(4):399–404

    Article  Google Scholar 

  31. Lodge, M., & Taber, C. S. (2013). The rationalizing voter. Cambridge University Press

  32. Lupia A, McCubbins MD (1998) The democratic dilemma: can citizens learn what they need to know? Cambridge University Press, Cambridge

  33. McCright AM, Charters M, Dentzman K, Dietz T (2016) Examining the effectiveness of climate change frames in the face of a climate change denial counter-frame. Top Cogn Sci 8:76–97

    Article  Google Scholar 

  34. Motta, M. (2017). The dynamics and political implications of anti-intellectualism in the United States. American Politics Research, 1532673X17719507

  35. Mullinix KJ, Leeper TJ, Druckman JN, Freese J (2015) The generalizability of survey experiments. J Experimental Political Sci 2(2):109–138

    Article  Google Scholar 

  36. Nyhan B, Reifler J, Ubel PA (2013) The hazards of correcting myths about health care reform. Med Care 51(2):127–132

    Article  Google Scholar 

  37. Nyhan B, Reifler J, Porter E, Wood T (2018) Taking corrections literally but not seriously? The effects of information on factual beliefs and candidate favorability. Working paper accessed at https://www.dartmouth.edu/~nyhan/trump-corrections.pdf

  38. Oreskes N, Conway EM (2011) Merchants of doubt: how a handful of scientists obscured the truth on issues from tobacco smoke to global warming. Bloomsbury Publishing, New York

  39. Paolacci G, Chandler J (2014) Inside the Turk: understanding Mechanical Turk as a participant pool. Curr Dir Psychol Sci 23(3):184–188

    Article  Google Scholar 

  40. Petty RE, Cacioppo JT (1986) The elaboration likelihood model of persuasion. Adv Exp Soc Psychol 19:123–205

    Article  Google Scholar 

  41. Pew Research Center (2015) Global concern about climate change, Broad Support for Limiting Emissions. November

  42. Priester JR, Petty RE (1995) Source attributions and persuasion: perceived honesty as a determinant of message scrutiny. Personal Soc Psychol Bull 21(6):637–654

    Article  Google Scholar 

  43. Shipan CR, Lowry WR (2001) Environmental policy and party divergence in congress. Political Res Quarterly 54(2):245–263

    Article  Google Scholar 

  44. Skocpol T, Hertel-Fernandez A (2016) The Koch network and republican party extremism. Perspectives on Politics 14(3):681–699

    Article  Google Scholar 

  45. Thorson E (2016) Belief echoes: the persistent effects of corrected misinformation. Political Communication 33(3):460–480

    Article  Google Scholar 

  46. Uscinski, J. E., & Parent, J. M. (2014). American conspiracy theories. Oxford University Press

  47. Van der Linden, S. L., Leiserowitz, A.A., Feinberg, G. D., & Maibach, E.W. (2015). The scientific consensus on climate change as a gateway belief: experimental evidence. PLOS One https://doi.org/10.1371/journal.pone.0118489

  48. Zaller, J. (1992). The nature and origins of mass opinion. Cambridge University Press

Download references

Acknowledgements

The authors thank Adam Berinsky, Thomas Hayes, Paul Herrnson, Blair Johnson, Matto Mildenberger, Megan Mullin, Brendan Nyhan, Mike Shor, Matthew Singer, Gabriela Tafoya, Steven Webster, participants in the UConn Political Economy Workshop, and the editors and reviewers at Climatic Change for their constructive feedback on this study.

Replication materials

Replication materials for this study are available at https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/KV6S5V

Author information

Affiliations

Authors

Corresponding author

Correspondence to Salil D. Benegal.

Appendix

Appendix

Appendix 1: Treatment wording

Shown to all treatment groups:

Senator James Inhofe (R-OK), Chairman of the Senate Environment Committee, stated in an interview last month that environmental regulations aimed at addressing climate change were “alarmist” and that the science over climate change remains unsettled.

“It's important to question whether climate change is even a problem for human existence,” Inhofe said. “Thus far no one has seriously demonstrated any scientific proof that increased global temperatures would lead to the catastrophes predicted by alarmists. The claim that global warming is caused by man-made emissions is simply untrue and not based on sound science.”

Inhofe has made similar comments in the past, most notably during a Senate speech in 2015 when he argued that the earth is currently in a cooling period. Inhofe has been a strong critic of recent federal bills aiming to cut carbon pollution by regulating power plant emissions and oil drilling. The regulations pose a serious threat to the country’s economy, he said.

Treatment 1: Only received misinformation article with no correction.Footnote 8

Treatment 2: Misinformation article with the following (scientist source) conclusion:

However, several scientists have been critical of Inhofe’s remarks on climate change. Members of NASA pointed to recent reports from the Intergovernmental Panel on Climate Change (IPCC) that recently declared that human activities have caused most of earth’s temperature rise since 1950 and will continue to do so in the future. The IPCC, which is a non-partisan panel of scientists from over 100 countries, estimates that global temperatures may increase by as much as 4.8 °C over pre-industrial levels by 2100. Dr. James Hansen, a climatologist who worked at NASA for over 40 years said, “When you have over 90 percent of the world’s scientists who have studied this stating that climate change is occurring and that humans play a contributing role, it's time to defer to the experts.”

Treatment 3: Misinformation article with the following (Democrat source) conclusion:

However, several Democrats have been critical of Inhofe’s remarks on climate change. Senators Brian Schatz (D-HI) and Maria Cantwell (D-WA) pointed to recent reports from the Intergovernmental Panel on Climate Change (IPCC) that recently declared that human activities have caused most of earth’s temperature rise since 1950 and will continue to do so in the future. The IPCC, which is a non-partisan panel of scientists from over 100 countries, estimates that global temperatures may increase by as much as 4.8 °C over pre-industrial levels by 2100. Democrat Senator Sheldon Whitehouse (D-RI) said, “When you have over 90 percent of the world’s scientists who have studied this stating that climate change is occurring and that humans play a contributing role, it's time to defer to the experts.”

Treatment 4: Misinformation article with the following (Republican source) conclusion:

However, several Republicans have recently been critical of Inhofe’s remarks on climate change. Senators John Thune (R-SD) and Kelly Ayotte (R-NH) pointed to recent reports from the Intergovernmental Panel on Climate Change (IPCC) that recently declared that human activities have caused most of earth’s temperature rise since 1950 and will continue to do so in the future. The IPCC, which is a non-partisan panel of scientists from over 100 countries, estimates that global temperatures may increase by as much as 4.8 °C over pre-industrial levels by 2100. Republican Congressman Chris Smith (R-NJ) said, “When you have over 90 percent of the world’s scientists who have studied this stating that climate change is occurring and that humans play a contributing role, it's time to defer to the experts.”

Appendix 2: Question wording

Survey questions

Outcome variables:

Please state your level of agreement with the following statements on a scale of 1–10, 10 indicating the strongest level of agreement.

Acknowledgement of scientific consensus on climate change

There is a general consensus among scientists that the average temperature on earth has been getting warmer over the past few decades.

  • Strongly disagree... 10. Strongly agree

Human activity affects climate change

The problem of climate change is mainly due to human activity such as burning fossil fuels.

  • Strongly disagree... 10. Strongly agree

Seriousness of climate change

How important a problem do you think climate change is at this moment, on a scale of 1–10?

  • 1: Not important at all ... 10: Very important

Attention check

Which of the following statements best summarizes the article that you just read?

  1. a.

    Senator James Inhofe believes that climate change is an urgent problem

  2. b.

    Senator James Inhofe is skeptical about climate change [correct answer]

  3. c.

    Senator James Inhofe is optimistic about the 2016 election

  4. d.

    Senator James Inhofe wants bipartisan policy on climate change

Respondents who did not answer (b) to the above question were considered to have not adequately read the assigned treatment and were removed from the sample.

Appendix 3: Demographics

Variable Mechanical Turk sample
Sex Male 51%
  Female 49%
Race White 81%
  Asian 7%
  Hispanic 5%
  Black 6%
Education Up to high school 12%
  Some college studies, no degree 36%
  Four year college degree 37%
  Postgraduate degree 15%
Partisanship Republican 20%
  Democrat 40%
  Independent 31%
N   1338

Demographics across treatments

  Treatment type
Variable Null (denial) Expert Democrat Republican
Male 53% 52% 52% 47%
White 83% 81% 76% 83%
College or higher 50% 57% 48% 52%
Republican 17% 18% 22% 20%
Democrat 38% 39% 36% 47%

Sample Ideology breakdown (percentage by party identification)

  Republican Democrat Independent Total
Very conservative 16.8 0.3 1.3 66
Conservative 54.3 2.7 6.0 230
Moderate 25.3 19.8 71.1 399
Liberal 3.0 52.9 16.6 424
Very liberal 0.6 24.3 5.1 186
  100 100 100 1305

Appendix 4: Summary of pilot study

A preliminary/pilot study was conducted on Amazon Mechanical Turk in April 2015. Participants in this study were randomly assigned one of the following treatments, with a post-treatment survey. While the results from the two experiments are largely consistent in terms of the significance of effects, this design failed to control for (a) the content of the correction message, and (b) inadequate isolation of personal affect or credibility for the correction sources dues to national name recognition.

Shown to all treatment groups:

In an interview with the Cincinnati Enquirer, Senate Majority Leader Mitch McConnell told the editorial board that he does not believe in climate change, and that the science over climate change remains unsettled. McConnell has raised doubts about climate change in the past, but never as directly.

“I don't buy that climate is changing. In 1970s, we were all concerned about the ice age coming,” said the Republican senator, who comes from Kentucky, and was unanimously elected to be Senate Majority Leader last month.

McConnell said that the science remains unsettled on the problem of climate change. “Pick your period and data will support your bias,” said McConnell, indicating that the earth has gone through cooling and warming periods in the past. “Each side has their scientists, and they can all go in and argue,” he responded, when asked whether he believed if human activity contributes to global warming. McConnell also criticized newly proposed Environmental Protection Agency rules that would cut carbon pollution from coal power plants. The regulations pose a serious threat to Kentucky’s coal industry, he said.

Treatment 1: Only received denial article with no correction.

Treatment 2: Denial article with the following conclusion:

McConnell’s statements contradict findings from the most recent report published by the Intergovernmental Panel on Climate Change (IPCC.) The panel, which consists of thousands of scientists and reviewers from more than 100 countries, confirmed with 95% certainty that human activities have caused most of earth’s temperature rise since 1950, and will continue to do so in the future. The IPCC’s most recent report, released on 2 November 2014 in Copenhagen, estimates that global temperatures may increase by as much as 4.8 °C over pre-industrial levels by 2100.

Treatment 3: Denial article with the following conclusion:

McConnell’s statements were refuted by several members of the Democrat Party, including Senators Elizabeth Warren, Brian Schatz, and Minority Leader Harry Reid. In an earlier statement, Schatz (D-Hawaii) had called climate change “the most pressing issue of our time,” declaring that there was overwhelming evidence that “it is real, it is caused by humans, it is happening now, and it is solvable.”

Treatment 4: Denial article with the following conclusion:

Although many other Republican senators have been skeptical of climate change, a growing number have become more accepting of the issue and its risks in the past year. Marco Rubio (R-FL) and John Thune (R-SD) have in recent months publicly acknowledged that global temperatures are increasing, while Chris Christie (R-NJ) said two years ago that “When you have over 90 percent of the world’s scientists who have studied this stating that climate change is occurring and that humans play a contributing role, it's time to defer to the experts.”

Respondents were asked similarly worded questions about whether climate change is occurring and whether it is affected by human activity. A similar attention check was included to ensure respondents had read the treatment. A total of 969 respondents completed this survey. Results showed that a Republican correction was effective in persuading other Republicans to update their reported attitudes on climate change closer to the scientific consensus, and had moderate effects on Independents and Democrats. In an effort to rely on actual quotes from real politicians, the details of corrective messages differed in this pilot study. This sacrificed some degree of internal validity for external validity, as the intensity and phrasing of messages differed by group. We then developed similar corrective treatments for the main study discussed in the manuscript, but with a greater focus on maintaining internal validity.

Mean opinions by treatment group and partisanship from pilot

Partisanship Corrective treatments:
  None (denial) Expert (IPCC) Democrat Republican
Belief in evidence that climate change exists
Republican
(std. deviation)
(N)
6.392
(2.293)
56
7.195
(2.595)
46
6.510
(2.550)
49
7.609
(2.157)
64
Democrat 8.705
(1.711)
129
8.877
(1.588)
155
8.644
(1.681)
135
8.895
(1.646)
124
Independent 7.615
(1.914)
39
8.588
(1.760)
34
7.560
(2.083)
25
8.170
(2.301)
41
Combined 7.970
(2.098)
241
8.480
(1.975)
254
8.000
(2.187)
223
8.360
(1.979)
250
Belief that human activity affects climate change
Republican
(std. deviation)
(N)
5.446
(2.507)
56
5.586
(2.963)
46
5.489
(2.534)
49
6.234
(2.758)
64
Democrat 8.364
(1.672)
129
8.380
(1.762)
155
7.925
(1.835)
135
7.967
(1.929)
124
Independent 5.846
(2.508)
39
6.941
(2.282)
34
6.360
(2.885)
25
7.073
(2.327)
41
Combined 7.149
(2.451)
241
7.610
(2.422)
254
7.179
(2.411)
223
7.308
(2.339)
250

Two-tailed t tests from this study:

Partisanship Corrective treatments (compared to the null/denial group)
  Expert (IPCC) Democrat Republican
Belief in evidence that climate change exists
Republican ∆ mean .802*
(1.657)
.117
(0.248)
1.216***
(2.992)
Democrat ∆ mean .171
(.877)
−.060
(0.292)
.189
(0.226)
Independent ∆ mean .972***
(2.248)
−.055
(1.070)
.555
(0.389)
Belief that human activity affects climate change
Republican ∆ mean .140
(0.259)
.0433
(0.088)
.787**
(1.628)
Democrat ∆ mean .016
(0.079)
−.438**
(2.026)
−.396*
(1.749)
Independent ∆ mean 1.095**
(1.939)
.513
(0.753)
1.227**
(2.269)

T-statistics in parentheses, * P < 0.10, ** P < 0.05, *** P < 0.01

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Benegal, S.D., Scruggs, L.A. Correcting misinformation about climate change: the impact of partisanship in an experimental setting. Climatic Change 148, 61–80 (2018). https://doi.org/10.1007/s10584-018-2192-4

Download citation