![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
Are the United States (US) and China destined to fall into a ‘Thucydides trap’ of power transitions leading to great power conflict? This study explores how the intersubjective perception of media-disseminated narratives of US–China interdependence may shape the likelihood of war. In two randomized online experiments, we manipulated ordinary Americans’ perceptions of US–China relations with real CNN video clips that narrated a US–China power transition as either positive or zero sum. Across both experiments, more zero-sum narratives boosted perceived US–China competition, increasing intergroup mistrust, anger and subsequent desires for a tougher China policy. The second study also revealed that individual differences in nationalism and uncertainty avoidance moderated the effects of the perception of media narratives on mistrust and anger. Viewers actively interpret media they are exposed to. These findings empirically demonstrate the power of narratives: specifically, they reveal the psychological mechanisms linking structural changes in the balance of power to the individual-level processes that may determine great power war and peace.
Acknowledgements
Both authors would like to thank Harold and Ruth Newman for their generous gift to the University of Oklahoma (OU), which made this research possible. The first author would also like to thank the OU students who recruited the participants for study 1 in autumn 2015.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on Contributors
Peter Gries is the Lee Kai Hung Chair and Director of the Manchester China Institute at the University of Manchester, where he is also Professor of Chinese politics. He studies the political psychology of international affairs, with a focus on China and the United States. He is author of The Politics of American Foreign Policy: How Ideology Divides Liberals and Conservatives over Foreign Affairs (Stanford, 2014) and China’s New Nationalism: Pride, Politics, and Diplomacy (California, 2004), and scores of peer-reviewed academic journal articles. He also co-edited State and Society in 21st Century China (Routledge 2004) and Chinese Politics (Routledge 2010). Email: [email protected]
Yiming Jing is an assistant professor at the Institute of Psychology, Chinese Academy of Sciences. He does research on social psychology and teaches cultural psychology. He is interested in understanding the sociocultural context of human behavior especially pro-sociality and trust. His work has appeared in various journals including Journal of Experimental Psychology: General, Social Cognitive and Affective Neuroscience, Journal of Experimental Social Psychology, and Journal of Cross-Cultural Psychology. Email: [email protected]
Notes
1 <https://www.belfercenter.org/thucydides-trap/book/thucydides-praise> accessed 10 July 2019.
2 Glaser <https://www.youtube.com/v/1Sw6W5NG73g>, Kiernan <https://www.youtube.com/v/vRmXNOIEbP0>, Powell <https://www.youtube.com/v/1UDS3_4jwAI>, Paulson <https://www.youtube.com/v/rLKDpB7ORyk>.
3 F (3, 254) = 30.17, p < .001, partial η2 = .26. Following statistical convention (Cohen Citation1988), a ‘partial eta squared’ of .01 is considered small, .06 medium, and .14 large.
4 Credible: M = 5.62, SD = 1.16 versus M = 4.42, SD = 1.31; likeable M = 5.02, SD = 1.22 versus M = 3.93, SD = 1.27.
5 F (4, 253) = 20.61, p < .001, partial η2 = .25.
6 Professional: M = 5.72, SD = 1.18 versus M = 4.76, SD = 1.38; balanced: M = 4.53, SD = 1.43 versus M = 3.63, SD = 1.41.
7 F (1, 249) = 101.79, p < .001, partial η2 = .29.
8 M = 5.86, SD = 1.06 versus M = 3.58, SD = 1.45.
9 F (1, 249) = 3.03, p = .083.
10 F (1, 249) = 229.68, p < .001, partial η2 = .48.
11 M = 5.26, SD = 1.62 versus M = 2.68, SD = 1.44.
12 M = 6.39, SD = 1.03 versus M = 4.07, SD = 1.24, p < .001, 95% CI for the mean difference = [1.32, 2.53].
13 M = 2.88, SD = 1.62 versus M = 2.45, SD = 1.18, p > .250, 95% CI for the mean difference = [−.22, .70]. For a plot of this interaction, see Figure S2 in the supplemental materials.
14 F (1, 249) = 7.33, p = .007, partial η2 = .03.
15 M = 4.70, SD = 1.45 versus M = 4.15, SD = 1.44.
16 F (1, 249) = 12.53, p < .001, partial η2 = .05.
17 M = 4.13, SD = 1.50 versus M = 3.31, SD = 1.52.
18 Indirect effects: (1) b = .04, SE = .02, 95% CI = [.01, .09]; (2) b = .02, SE = .01, 95% CI = [.00, .07] ; and (3) b = .05, SE = .02, 95% CI = [.02, .10].
19 Indirect effect: b = .09, SE = .03, 95%CI = [.04, .15].
20 Unless mentioned otherwise, all measures in study 2 were identical to those used in study 1.
21 F (3, 125) = 31.02, p < .001, partial η2 = .43.
22 M = 5.73, SD = 1.20 versus M = 4.94, SD = 1.68.
23 M = 4.75, SD = 1.25 versus M = 3.20, SD = 1.49.
24 M = 3.74, SD = 1.58 versus M = 3.08, SD = 1.24.
25 F (4, 124) = 7.45, p < .001, partial η2 = .19.
26 Professional: M = 6.05, SD = 1.19 versus M = 5.03, SD = 1.68; balanced: M = 5.07, SD = 1.35 versus M = 4.10, SD = 1.75.
27 F (1, 120) = 89.41, p < .001, partial η2 = .43.
28 M = 6.27, SD = 1.05 versus M = 3.39, SD = 1.34.
29 F (1, 120) = 12.34, p = .001, partial η2 = .09.
30 M = 4.54, SD = 1.53 versus M = 3.56, SD = 1.25.
31 F (1, 120) = .54, p > .250.
32 F [1, 120] = 6.93, p = .010, partial η2 = .05.
33 F [1, 120] = 5.80, p = .018, partial η2 = .05.
34 Russia: M = 4.46 versus M = 3.95; Japan: M = 3.36 versus M = 2.93.
35 F [1, 120] = 8.48, p = .004, partial η2 = .07; M = 3.44 versus M = 2.32.
36 F [1, 120] = 33.50, p < .001, partial η2 = .22; M = 2.44 versus M = 3.81.
37 F [1, 120] = .40, p > .250, partial η2 = .00; M = 3.44 versus M = 3.76.
38 F [1, 120] = 2.55, p = .113, partial η2 = .02; M = 2.37 versus M = 2.80.
39 Indirect effects: (1) b = .08, SE = .04, 95% CI = [.02, .17]), and (2) b = .14, SE = .04, 95% CI = [.07, .24].
40 Indirect effect: b = .03, SE = .03, 95% CI = [−.01, .07].
41 b = .09, SE = .04, 95% CI = [.05, .18].
42 One SD above the mean: b = .50, p < .001 versus b = .22, p = .004.
43 One SD below the mean of nationalism: indirect effect = .08, SE = .04, 95% CI = [.01, .16]; one SD above the mean of nationalism: indirect effect = .19, SE = .06, 95% CI = [.08, .33].
44 b = −.11, SE = .04, 95% CI = [−.19, −.02].
45 One SD above the mean of uncertainty avoidance: b = −.35, p < .001 versus one SD below the mean: b = −.07, p = .195.
46 One SD below the mean of uncertainty avoidance: indirect effect = .03, SE = .03, 95% CI = [−.04, .13]; one SD above the mean: indirect effect = .14, SE = .04, 95% CI = [.06, .25].
47 Indirect effect: b = .06, SE = .04, 95% CI = [.00, .15].
48 Indirect effect: b = −.01, SE = .05, 95% CI = [−.13, .08].
49 Indirect effect: b = −.02, SE = .02, 95% CI = [−.08, .00].
50 Indirect effect: b = .03, SE = .03, 95% CI = [−.02, .12].
51 Overall, study 1’s participants were more conservative (M = 4.03, SD = 1.72) and more nationalistic (M = 5.29, SD = 1.23) than study 2’s participants (M = 3.23, SD = 1.57, and M = 4.45, SD = 1.58, respectively).
52 A regression analysis revealed that nationalism moderated the effect of the media manipulation (more positive-sum-framed Paulson versus more negative-sum-framed Kiernan) on perceived US–China economic competition from the news report, β = −.28, p = .002. Media manipulations had a greater impact on perceived bilateral economic competition among highly nationalistic participants (one SD above the mean: Mpositive-sum-framed = 2.34 versus Mnegative-sum-framed = 6.46) than among participants low on nationalism (one SD above the mean: Mpositive-sum-framed = 1.26 versus Mnegative-sum-framed = 6.67).
53 Because we did not manipulate our mediators independently from the independent variables, however, it is plausible that both the mediators and the dependent variables were shaped by an unmeasured third variable. We therefore need to be modest about this causal claim (see Imai et al [Citation2011] on the sequential ignorability assumption).
