Telling or told? The role of implicit and explicit media cues in the formation of perceived affective polarization in the debate about climate change mitigation in Germany

Slides presented at the 10th European Communication Conference (ECC), in Ljubljana (Slovenia), 27 September 2024.

This slide deck combines two separate conference submissions presented together, both drawing on the same underlying dataset but analyzing different dependent variables.

Conference paper I: "Telling or Told? The role of implicit and explicit media cues in the formation of perceived affective polarization in the debate about climate change mitigation" (Quirin Ryffel, Thomas Zerback, Alina Jakob, Nayla Fawzi)

In this study, we investigated whether exposure to implicit (i.e., conflict coverage or mentions of extremity) and explicit polarization cues (e.g., explicit mentions of polarization/divides) in climate change media coverage shapes perceptions of affective polarization, both in media coverage and in society. We conducted a linkage study, combining a quantitative content analysis of N = 652 articles from 13 German news media outlets (TV, newspapers, online magazines, alternative) with a representative two-wave panel survey (N = 2051) conducted immediately before and after the content analysis in June and August 2023. The content analysis covered the period between the two panel waves and identified implicit and explicit polarization cues at the article level. For each outlet, a slant score was calculated by subtracting the number of articles containing polarization cues from those with de-polarization cues. We then multiplied the outlet-specific scores with respondents’ self-reported frequency of outlet use (our independent dosage variables). Perceptions of affective polarization in media coverage[2] and society[3] were both measured with three items on a 7-point Likert scale (e.g., "In the climate debate, the media portray two camps that are increasingly hostile towards each other"). Age, gender, education level, political orientation, issue-specific interest and knowledge, attitude extremity, social media use and interpersonal discussion (all issue-specific), as well as the levels of perceived affective polarization in wave 1 served as controls. Results of a path analysis show neither a direct effect of the dosage of explicit nor implicit polarization on perceived affective media polarization. Rather, issue-specific attitude extremity as well as interpersonal discussion about climate change increase perceived affective media polarization. As predicted, however, perceived affective polarization of media coverage on climate change mitigation has a substantial effect on perceived affective polarization of society regarding the issue, underlining the importance of perceptions of media coverage on climate change mitigation as well as the role of interpersonal communication in shaping such perceptions.

[1] Reliability scores of the content-analytic variables (Brennan Prediger’s Kappa) ranged between .71 and 1.00.
[2] αW1 = .85, MW1 = 4.77, SDW1 = 1.39; αW2 = .86, MW2 = 4.87, SDW2 = 1.36
[3] αW1 = .73, MW1 = 5.24, SDW1 = 1.21; αW2 = .77, MW2 = 5.28, SDW2 = 1.22

Conference paper II: "Between Headlines and Reality? News Media's Impact on False Polarization in the Debate about Climate Change Measures in Germany" (Alina Jakob, Nayla Fawzi, Quirin Ryffel, Thomas Zerback)

In this study, we tested whether media coverage of the German climate policy debate produces false polarization, meaning a perceived opinion gap between supporters and opponents that is larger than the actual gap using the same study design and data.

For our independent variables, we then calculated an exposure variable that captures the prevalence of exposure to implicit and explicit (de)polarization cues on an affective and ideological level in the media coverage, weighted by the respective media use. Our dependent measures include actual ideological polarization (= difference between mean positions of opponents and supporters), perceived ideological polarization (= difference between individually assessed positions of “a typical opponent” and “a typical supporter”) and false polarization (= difference between perceived and actual polarization) regarding four measures.
Results show a pattern of false polarization and a tendency to overestimate the gap between opinion camps. This is evident for four out of five climate change-related measures, such as banning the installation of new gas and oil heating systems or restricting consumption for future generations. The only exception is the degree of polarization regarding the reactivation of nuclear power plants, which is almost correctly assessed. Controlling for false polarization perceptions in the first wave, sociodemographics and political attitudes, linear regression results indicate that particularly the dose of polarization that private, tabloid or alternative media evoke in their reporting have a significant impact on their audience’ misperceptions. Results underline news media's responsibility when presenting polarized issues as this can indeed make ideological polarization appear greater than it actually is which, in turn, might hinder the possibility of reaching an agreement within society on how to address climate change.

Transcript

1. Telling or told? The role of implicit and explicit media cues in the formation of perceived affective polarization in the debate about climate change mitigation in Germany Between headlines and reality? News media’s impact on false polarization in the debate about climate change mitigation in Germany Quirin Ryffel University of Zurich Alina Jakob Johannes Gutenberg University Mainz Thomas Zerback Heinrich Heine University Dusseldorf Nayla Fawzi Johannes Gutenberg University Mainz
2. Polarization in Germany. (Deutscher Bundestag, n.d.) 2
3. Polarization in Germany. German society is not divided in its entirety, but within certain “trigger points” or “conflict arenas” (Mau et al., 2023) 3
4. Polarization in Germany. identity-political recognition migration, territorial access & inclusion socio-economic distribution environmental policy (Mau et al., 2023, p. 25) 4
5. Polarization in Germany. Consensus on the existence of anthropogenic climate change and the need for climate action in Germany (Mau et al., 2023) The extreme weather events make it clear that climate change is already happening. Climate change is one of the greatest threats to humankind. Climate change is not as dangerous as is claimed. (inverted) 0% 20 disagree (1,2) 40 undecided (3) 60 80 agree (4,5) Note. N = 2577; Fawzi et al. (2024) 5
6. Polarization in Germany. Consensus on the existence of anthropogenic climate change and the need for climate action (Mau et al., 2023) Almost 75% say that they perceive (very) strong conflicts between those who care and those who do not care so much about the environment (Mau et al., 2023) The extreme weather events make it clear that climate change is already happening. Climate change is one of the greatest threats to humankind. Climate change is not as dangerous as is claimed. (inverted) 0% 20 disagree (1,2) 40 undecided (3) 60 80 agree (4,5) Note. N = 2577; Fawzi et al. (2024) 6
7. Polarization in Germany. Actual opinions on the concrete implementation of stricter measures are rather diverse (Wolf et al., 2023) Climate protection measures restrict my individual freedom too much. Climate protection measures are damaging the German economy. Climate protection should be given priority by law. 0% 20 disagree (1,2) 40 undecided (3) 60 agree (4,5) Note. N = 2577; Fawzi et al. (2024) 7
8. Polarization & the media. News media can have a significant impact on citizen’s perception of public opinion This perception of public opinion can: affect citizens’ attitudes, (other) perceptions, cognitions & behaviors (Levendusky & Malhotra, 2016) (Gunther & Storey, 2000; Noelle-Neumann, 1974; Yang et al., 2016) 8
9. Polarization & the media. News media can have a significant impact on citizen’s perception of public opinion (Levendusky & Malhotra, 2016) Medias coverage can lead to overestimations of opinion differences and extremity within society - resulting in a stronger perceived than actual ideological polarization (= false polarization) This perception of public opinion can: − affect citizens’ attitudes, (other) perceptions, cognitions & behaviors (Gunther & Storey, 2000; Noelle-Neumann, 1974; Yang et al., 2016) This misperception can: − affect citizens' own political opinions (Ahler, 2014) − increase actual ideological and affective polarization (Fernbach & Van Boven, 2021; Levendusky & Malhotra, 2013) (Westfall et al., 2015; Wilson et al., 2020) 9
10. Overarching question. What influence does the use of news media and exposure to polarization cues have… 1. on perceived affective polarization in media coverage & within society (affective level) 2. on perceived and false polarization (ideological level) …in the context of the climate action debate in Germany? 10
11. Linkage study: Design. 1 two-wave panel survey − w1: Jun 23, n = 3140 − w2: Aug 23, n = 2051 − Panel provider: Bilendi 2 content analysis Articles/items from 13 German news media outlets/news programs, n = 625 related to climate change − TV programs, newspapers, online magazines, alternative media − published between the waves (Jul 23) − thematically related to climate change and climate action − Database: Factiva 11
12. Linkage study: Dosage. Polarization dosage = Media use𝑜𝑢𝑡𝑙𝑒𝑡 w1 × Pol cueoutlet − Depol cueoutlet measured in survey on a 7-point scale, then weighted: 0 (1–2), 0.5 (3–5), 1 (6–7) (Wirz et al., 2017) TV news: Tagesschau, Heute, RTL Aktuell, :newstime (Sat.1) Newspapers: Bild, FAZ, Welt, SZ, FR, taz, Die Zeit News magazines: Der Spiegel, Focus Online Alternative media: Achtung, Reichelt! (YouTube) for a more detailed picture of the coverage on climate change: − Balance score: (pol – depol cues) − depolarization and polarization cues presented a) explicitly and b) implicitly in media coverage 12
13. Media cues: Telling or told? Implicit polarization cues Explicit polarization cues ◉ Depiction of extreme behavior of ◉ Explicit mentions of polarization or synonyms (e.g., division) group members (e.g., violence) ◉ Explicit mentions of extremity of ◉ Depiction of conflicts between opinions or synonyms (e.g., radical) two or more groups (Zerback et al., 2015) 13
14. Media effects on polarization perceptions. Implicit polarization cues Explicit polarization cues ◉ Depiction of extreme behavior of ◉ Explicit mentions of polarization or synonyms (e.g., division) group members (e.g., violence) ◉ Explicit mentions of extremity of ◉ Depiction of conflicts between opinions or synonyms (e.g., radical) two or more groups (Zerback et al., 2015) 14
15. Media effects on polarization perceptions. Implicit polarization cues Explicit polarization cues ◉ Depiction of extreme behavior of ◉ Explicit mentions of polarization or synonyms (e.g., division) group members (e.g., violence) ◉ Explicit mentions of extremity of ◉ Depiction of conflicts between opinions or synonyms (e.g., radical) two or more groups (Zerback et al., 2015) Polarization cues in media coverage have been shown to increase affective polarization in the U.S. (Malhotra & Levendusky, 2016a) 15
16. Implicit & explicit cues. Implicit polarization cue Explicit polarization cue Trial begins after attack on climate activist in Altmark (own translation) Climate movement in Munich Are activists becoming more radical? (own translation) 16
17. Content analysis: intercoder reliability. Implicit cues Brennan & Prediger‘s Kappa Holsti Explicit cues Brennan & Prediger‘s Kappa Holsti Extreme behaviour .60 .82 Explicit polarization .84 .92 Conflict .71 .88 Explicit extremity .84 .94 Consensus .92 .97 Explicit depolarization .79 .92 Explicit nonextremity .88 .95 Note: Three intercoder reliability tests were conducted with five trained student coders until satisfactory levels of reliability were achieved. N = 26 (third test). 17
18. Results of content analysis: Prevalence of cues (N = 652). Prevalence of explicit pol. cues: Prevalence of explicit depol. cues: ◉ Polarization: 4.1 % (n = 27) ◉ Depolarization: 1.4 % (n = 9) ◉ Extremity: 3.7 % (n = 24) ◉ Non-extremity: 0.9 % (n = 6) 18
19. Results of content analysis: Prevalence of cues (N = 652). 19
20. Results of content analysis: Prevalence of cues (N = 652). Prevalence of implicit pol. cues: Prevalence of depol. cues: ◉ Conflict: 28.7 % (n = 187) ◉ Consensus: 3.2 % (n = 21) ◉ Extreme behaviour: 9.0 % (n = 59) 20
21. Content Analysis: Prevalence of Implicit Cues in Sample (N = 652). 21
22. Research question 1. What influence does the received dosage of implicit and explicit polarization cues have… 1. on perceived affective polarization in media coverage & of the society (affective level) 2. on perceived and false polarization (ideological level) …in the context of the climate action debate in Germany? 22
23. Theoretical model & measures. H1a H2 H1b 23
24. Theoretical model & measures. Sum index: number of implicit cues per outlet x weight of outlet use (Min = -1, Max = 225, M = 31.52, SD = 43.40) H1a H2 H1b 5.24.,SD SD==1.34, 1.21,αα==.86). .77).Example: Example: Four items (7-point Likert scales, M = 4.84, Supporters and the opponents are becoming increasingly hostile towards each other. The media give impression that supporters and opponents have less and less to say to each other. Sum index: number of explicit cues per outlet x weight of outlet use (Min = -3.5, Max = 37.5, M = 4.91, SD = 7.46) 24
25. Telling or told? R2 .30*** .53*** H1a .08 .55*** -.06 H1b .41*** H2 R2 .51*** .45*** Note. Model fit: χ2 = 6450.163, df = 152, p < .001; SRMR = .044; Robust RMSEA = .041; Robust CFI = .951 ; n = 1476. All anaylses were performed with the package lavaan in RStudio. All significance tests were calculated using maximum likelihood estimation with robust (Huber-White) standard errors and a scaled test statistic that is (asymptotically) equal to the Yuan-Bentler test statistic. SRMR = standardized root mean square residual; RMSEA = root mean square error of approximation; CFI = comparative fit index. PMed = Perceived affective polarization in media coverage. PSoc = Perceived affective polarization in society. Significant coefficients are bolded. Control variables (included in model but not shown in the figure): Age, Gender** (female), education, issue-specific involvement, issue-specific attitude extemity*, and interpersonal discussion about climate change, political orientation, frequency of social media use*. Standardized path coefficients (β): *p < .05. **p < .01. ***p < .001. 25
26. Conclusion. Content analysis: − Overall: Over a third of news items contained (at least one) polarization cue(s) in climate coverage in July 2023! − Higher share of implicit (especially conflict) vs. explicit polarization cues − Depolarization cues in media coverage almost inexistent Linkage analysis − Null effect of dosage of explicit and implicit polarization cues on the perceived affective polarization depicted in media coverage − Positive correlation between perceived affective polarization depicted in media and perceived affective polarization in society 26
27. Discussion. Other drivers of perceived affective polarization in media coverage: Alternative explanations? 1. Too short time span of one month for effects to unfold? 2. Only a selection of journalistic outlets that are used were analyzed 3. Images? 4. Social media content or interpersonal discussion? Lastly: Underestimation of effects in linkage analyses due to measurement error (Scharkow & Bachl, 2017) 27
28. Overarching Question. What influence does the use of news media and exposure to polarization cues have… 1. on perceived affective polarization in media coverage & of the society (affective level) 2. on perceived and false polarization (ideological level) …in the context of the climate change debate in Germany? Alina Jakob | JGU Mainz 28
29. Mind the Gap. Media’s influence on the gap between actual and perceived ideological polarization is often overlooked in polarization research and only documented along party lines (Levendusky & Malhotra, 2016) Research on false polarization on the issue level can be found in social psychology, whereby the groups studied were identified within a specific thematic conflict (e.g. Chambers et al., 2002; Robinson et al., 1995; Sherman et al., 2003) party level issue level Distinction at the issue level makes it possible to link polarization research with research on the climate of opinion, with perceived polarization acting as a bridge and as specific form of perception of social collectives (Mutz, 1998) 29
30. We have our differences. false polarization = perceived - actual polarization actual ideological polarization = mean position of supporters – opponents perceived ideological polarization = individually assessed position of “typical supporter” –“typical opponent” (Levendusky & Malhotra, 2016; Sherman et al., 2003) 30
31. Groups Studied. I am in favor of stricter climate protection measures. I am against stricter climate protection measures. w1: n = 712 n = 1742 w2: n = 583 n = 1265 Distribution of opinion is consistent with similar surveys of the German population (e.g. infratest dimap, 2023; Mau et al., 2023; Wolf et al., 2023) 31
32. Actual Polarization. opponents () supporters () Note. n = 1640. 32
33. Perceived Polarization. typical opponent () typical supporter () Note. n = 2051. 33
34. False Polarization. Note. Actual Pol. n = 1640; Perceived Pol. n = 2051. 34
35. Media & False Polarization. While dose of polarization that public, private, tabloid or alternative media evokes negative effects, we find positive effects for left-wing media TV Newspapers Online Magazines Alternative Media Tagesschau (-) Heute RTL Aktuell (-) :newstime (Sat.1) Bild FAZ Welt (-) SZ FR taz (+) Die Zeit Der Spiegel Focus Online Achtung, Reichelt! (-) 35
36. Media & False Polarization. False Polarization: Restriction of consumption (w2) B SE Constant .543 .282 False Polarization (w1) .381 .036 .288*** Involvement .030 .036 .019 Age .000 .004 -.002 Education .151 .114 .031 Gender (male) -.402 .103 -.082*** Social Media Use -.009 .024 -.008 Exposure RTL Aktuell -.076 .035 -.060* Exposure Welt -.005 .002 -.069** Note. n = 2051, R2 = 0.192, R2 adj. = 0.184, F(6, 1975) = 24.623, p < .001. ***p < .001 ** p <. 01 * p < .05. ß Negative effects = assessment of polarization improves with dosage received by these outlets 36
37. Media & False Polarization. False Polarization: Ban on new gas and oil heating systems (w2) B SE Constant 1.171 .381 False Polarization (w1) .329 .023 .323*** Involvement -.066 .146 .061* Age .000 .005 -.001 Education .362 .146 .061* Gender (male) .014 .132 .002 Social Media Use -.078 .030 -.062* Exposure Taz .008 .003 .098* Note. n = 2051, R2 = 0.131, R2 adj. = 0.122, F(19, 1834) = 14.502, p < .001. ***p < .001 ** p <. 01 * p < .05. ß Positive effects = assessment of polarization worsens with the exposure of those outlets 37
38. Final Toughts. Polarization of the climate mitigation debate is (slightly) overestimated. However, groups are not generally perceived as more extreme than they actually are. Media use has a weak influence on false polarization perceptions, the exact pattern needs to be further researched (longer term, across several conflicts, different samples) Latent Profile Analysis indicates that a lower false polarization, i.e., the negative effects, occurs with those media outlets that are primarily only received in hyper usage types (exception: public media “Tagesschau”) 38
39. References. Ahler, D. J. (2014). Self-fulfilling misperceptions of public polarization. Journal of Politics, 76(3), 607–620. https://doi.org/10.1017/S0022381614000085 Andre, P., Boneva, T., Chopra, F., & Falk, A. (2024). Globally representative evidence on the actual and perceived support for climate action. Nature Climate Change, 1–7. https://doi.org/10.1038/s41558-024-01925-3 Deutscher Bundestag. (n.d.). Sitzverteilung des 20. Deutschen Bundestages. Deutscher Bundestag. https://www.bundestag.de/parlament/plenum/sitzverteilung_20wp Chambers, J. R., Baron, R. S., & Inman, M. L. (2006). Misperceptions in intergroup conflict: Disagreeing about what we disagree about. Psychological Science, 17(1), 38–45. https://doi.org/10.1111/j.1467-9280.2005.01662.x Fernbach, P. M., & van Boven, L. (2022). False polarization: Cognitive mechanisms and potential solutions. Current Opinion in Psychology, 43, 1–6. https://doi.org/10.1016/j.copsyc.2021.06.005 Gunther, A. C., & Storey, J. D. (2003).The influence of presumed influence. Journal of Communication, 53(2), 199–215. https://doi.org/10.1111/j.1460-2466.2003.tb02586.x Levendusky, M., & Malhotra, N. (2016a). Does Media Coverage of Partisan Polarization Affect Political Attitudes? Political Communication, 33(2), 283–301. https://doi.org/10.1080/10584609.2015.1038455 Levendusky, M. S., & Malhotra, N. (2016b). (Mis) perceptions of partisan polarization in the American public. Public Opinion Quarterly, 80(S1), 378–391. https://doi.org/10.1093/poq/nfv045 39
40. References. Mau, S., Lux, T., & Westheuser, L. (2023). Triggerpunkte: Konsens und Konflikt in der Gegenwartsgesellschaft. Suhrkamp Verlag. Mutz, D. C. (1998). Impersonal influence: How perceptions of mass collectives affect political attitudes. Cambridge University Press. Robinson, R. J., Keltner, D., Ward, A., & Ross, L. (1995). Actual versus assumed differences in construal:" Naive realism" in intergroup perception and conflict. Journal of Personality and Social Psychology, 68(3), 404. https://doi.org/10.1037/00223514.68.3.404 Scharkow, M., & Bachl, M. (2017). How Measurement Error in Content Analysis and Self-Reported Media Use Leads to Minimal Media Effect Findings in Linkage Analyses: A Simulation Study. Political Communication, 34(3), 323–343. https://doi.org/10.1080/10584609.2016.1235640 Sherman, D. K., Nelson, L. D., & Ross, L. D. (2003). Naï realism and affirmative action: Adversaries are more similar than they think. Basic and Applied Social Psychology, 25(4), 275-289. https://doi.org/10.1207/S15324834BASP2504_2 Vliegenthart, R., Stromback, J., Boomgaarden, H., Broda, E., Damstra, A., Lindgren, E., ... & Van Remoortere, A. (2023). Taking political alternative media into account: Investigating the linkage between media repertoires and (mis) perceptions. Mass Communication and Society, 1-25. https://doi.org/10.1080/15205436.2023.2251444 Westfall, J., van Boven, L., Chambers, J. R., & Judd, C. M. (2015). Perceiving political polarization in the United States: Party identity strength and attitude extremity exacerbate the perceived partisan divide. Perspectives on Psychological Science, 10(2), 145–158. https://doi.org/10.1177/1745691615569849 40
41. References. Wilson, A. E., Parker, V. A., & Feinberg, M. (2020). Polarization in the contemporary political and media landscape. Current Opinion in Behavioral Sciences, 34, 223–228. https://doi.org/10.1016/j.cobeha.2020.07.005 Wojcieszak, M. E., & Price, V. (2012). Perceived versus actual disagreement: Which influences deliberative experiences? Journal of Communication, 62(3), 418–436. https://doi.org/10.1111/j.1460-2466.2012.01645.x Wolf, I., Ebersbach, B. & Huttarsch, J. H. (2022). Soziales Nachhaltigkeitsbarometer der Energie- und Verkehrswende 2023: Was die Menschen in Deutschland bewegt - Ergebnisse einer Panelstudie zu den Themen Energie und Verkehr. https://ariadneprojekt.de/publikation/soziales-nachhaltigkeitsbarometer-2023/ Zerback, T., Koch, T., & Krämer, B. (2015). Thinking of Others: Effects of Implicit and Explicit Media Cues on Climate of Opinion Perceptions. Journalism & Mass Communication Quarterly, 92(2), 421-443. https://doi.org/10.1177/1077699015574481 41