Skip to main content

Who Believes Propaganda? Media Effects during the Anti-Japanese Protests in Beijing*


The Chinese media have undergone commercial liberalization during the reform era. Interviews with media practitioners reveal that media reform has brought about three different types of newspapers that differ with respect to their degree of commercial liberalization. Based on a natural experiment during the anti-Japanese protests in Beijing in 2005, this article shows that urban residents found more strongly commercialized newspapers more persuasive than less commercialized newspapers. Provided that the state can enforce press restrictions when needed, commercial liberalization promotes the ability of the state to influence public opinion through the means of the news media.

Hide All

1 See, for example, the blog by Phoenix TV journalist Luqiu Luwei at, accessed 5 July 2008; see also Sina BBS at, accessed 19 July 2008. On slogans see “Protests of the West spread in China,” New York Times, 21 April 2008.

2 Conversation with journalist of official paper in Beijing, 19 December 2008.

3 See, for example, CCTV, 16 March 2008; People's Daily, 17 March 2008; Beijing Youth Daily, 22 March 2008; Caijing, 24 March 2008; Sina, 15 April 2008.

4 Therefore, some observers suspected that “Anti CNN” messages were part of a larger propaganda effort to discredit reports that contradicted the official line of the state. See New York Times, 25 March 2008; Sueddeutsche Zeitung, 28 March 2008. Foreign Ministry spokesperson Qin Gang denied any direct links between the website and the Chinese government., accessed 20 July 2008.

5 See, for example, Li C.-c., Voices of China: The Interplay of Politics and Journalism (New York: Guilford Press, 1990); Lynch D.C., After the Propaganda State: Media, Politics, and “Thought Work” in Reformed China (Stanford: Stanford University Press, 1999); Esarey A., “Cornering the market: state strategies for controlling China's commercial media,” Asian Perspective, Vol. 29, No. 4 (2005), pp. 3783; Zhao Y., Communication in China: Political Economy, Power, and Conflict (Lanham: Rowman & Littlefield, 2008); Polumbaum J. and Lei X., China Ink: The Changing Face of Chinese Journalism (Lanham: Rowman & Littlefield, 2008).

6 J.J. Kennedy, “Maintaining popular support for the Chinese Communist Party: the influence of education and the state-controlled media,” Political Studies, Vol. 57, No. 3 (2009), pp. 517–36; Chen X. and Shi T., “Media effects on political confidence and trust in the People's Republic of China in the post-Tiananmen period,” East Asia: An International Quarterly, Vol. 19, No. 3 (2001), pp. 84118.

7 See, for example, J. Zhang, “Cong meijie fazhan jiaodu lun guojia yulun anquan” (“A discussion of national public opinion security based on the development of the mass media”), PhD thesis, School of Journalism and Communication, Peking University, 2006.

8 See Houn F.W., “Chinese communist control of the press,” Public Opinion Quarterly, No. 22 (1958–59), pp. 435–48.

9 For example, Xinhua became 40% government and 60% self-supported in 1991. Xin X., “A developing market in news: Xinhua News Agency and Chinese newspapers,” Media, Culture & Society, Vol. 28, No. 1 (2006), pp. 4566.

10 “Guanyu guifan xinwen chubanye rongzi huodong de shishi yijian” (“Opinion on the implementation of regulations on financial activities of the publication and press industry”), General Administration of Press and Publications, 25 July 2003.

11 Official papers can largely rely on subscriptions by government offices at all levels of government (interviews no. 9, 10, 13, 22, 26, 27, 33, 34, 37, 40, 43, 46), but official papers can now also be bought at newspaper stands (interviews no. 8, 34).

12 Translations of the Chinese terms kaifang and baoshou.

13 In 2001 the central government transformed state assets into Party-owned assets. Hu Z., “The post-WTO restructuring of the Chinese media industries and the consequences of capitalization,” Javnost/The Public, Vol. 20, No. 4 (2003), pp. 1936.

14 Interviews no. 5, 7, 8, 14, 16, 17, 21, 22, 25. See also Brady A.-M., Marketing Dictatorship: Propaganda and Thought Work in Contemporary China (Lanham: Rowman & Littlefield, 2008).

15 Slogans of the Xinjingbao and the Fazhi wanbao, respectively.

16 Expertise is defined as a source's “presumed knowledge and ability to provide accurate information.” See Petty R. and Wegener D., Attitude Change: Multiple Roles for Attitude Change (Boston: McGraw-Hill, 1998), p. 344. Objectivity refers to perceptions of media sources to be unbiased, accurate, fair and “to tell the whole story.” See Iyengar S. and Kinder D.R., Psychological Accounts of Agenda-Setting (Beverly Hills: Sage, 1985); Miller J. and Krosnick J., “News media impact on the ingredients of presidential evaluations: politically knowledgeable citizens are guided by a trusted source,” American Journal of Political Science, Vol. 44, No. 2 (2000), pp. 301–15.

17 Interviews no. 2, 8, 27, 43.

18 Interviews no. 8, 43.

19 Media scholar Chen Lidan defines xuanchuan as “Using various symbols to communicate a certain concept in order to influence people's thought and their actions.” Chen L., “Yong shishi shuo hua shi xuanchuan fangfa er bu shi xinwen xiezuo guilu” (“Using facts to write news is a propaganda method and not a rule to write news reports”), Renmin wang (People's Net) (2003).

20 In this context the goal of ensuring social stability draws together people who have divergent opinions and possibly hold incompatible positions on media control. See Latham K., “Nothing but the truth: news media, power and hegemony in south China,” The China Quarterly, No. 163 (2000), pp. 633–54 at pp. 650–51.

21 Interviews no. 38, 27, 4. See also Kang Y., Xinwen yu zhengzhi yaolue (Summary of News and Politics) (Beijing: Beijing guangbo xueyuan chubanshe, 2001).

22 Interviews no. 38, 27, 4. See also Z. Zhao and F. Cai, “Maohe er shenli: cong chuanbo neirong de jiaodu kan xinwen yu xuanchuan de chayi” (“Apparently harmonious but actually different: difference between news and propaganda from the perspective of communication content”), Eighth National Conference on Communication Studies, Tsinghua University, Beijing, 2004.

23 Da bian qiu refers to an excellent shot in ping pong that hits the very edge of the table, making it extremely difficult for the opponent to return it.

24 See, for example, Eagly A., Wood W. and Chaiken S., “Causal inferences about communicators and their effects on opinion change,” Journal of Personality and Social Psychology, Vol. 36, No. 4 (1978), pp. 424–35.

25 D. Stockmann, “Media trust in China,” paper presented at the Annual Meeting of the Midwest Political Science Association, 2006.

26 Interviews in one district were delayed for reasons unrelated to the protests. Results remain stable when interviews in this district are controlled for.

27 See, for example, Cook T.D. and Campbell D.T., Quasi-experimentation: Design & Analysis Issues for Field Settings (Chicago: Rand McNally, 1979).

28 Table A1 in the Appendix displays differences between control and treatment groups. Tables A2 to A4 show that results remain stable when holding control variables constant.

29 Wan M., Sino-Japanese Relations: Interaction, Logic, and Transformation (Stanford: Stanford University Press, 2006); Reilly J., “China's history activism and Sino-Japanese relations,” China: An International Journal, Vol. 4, No. 2 (2006), pp. 189216; J.C. Weiss, “Powerful patriots: nationalism, diplomacy and the strategic logic of anti-foreign protest,” PhD thesis, University of California, San Diego, 2008.

30 The issue was brought up at a meeting between Koizumi and Hu at the Asian–African summit in May. See People's Daily Online, 2 May 2005.

31 New York Times, 9, 10 and 20 April 2005, Agence France Presse 14 April 2005, TASS, 6 March 2005.

32 In other cities protests had taken place in the first week of April. See China Daily, 1 April 2005, Deutsche Presse Agentur, 7 April 2005.

33 Personal observation by the author.

34 None of my interviewees asked about issue sensitivity in international news reporting before 9 April mentioned Japan. In handbooks used at newspapers in Beijing in order to train journalists Japan was not mentioned as a sensitive issue.

35 Interviews no. 13, 20, 21, 25. See also Deutsche Presse Agentur, 11 April 2005, Agence France Presse, 10 April 2005, Herald Tribune, 20 April 2005, New York Times, 20 and 25 April 2005.

36 Interviews no. 20, 21, 25.

37 For details see D. Stockmann, “Information overload? Collecting, managing, and analyzing Chinese media content,” in A. Carlson, M. Gallagher, K. Lieberthal and M. Manion (eds.), Chinese Politics: New Sources, Methods, and Field Strategies (New York: Cambridge University Press, forthcoming).

38 A t-test indicated that this change in the tone of news reporting was statistically significant at the 99% confidence level (p = 0.006). Results do not change when controlling for article length or number of references towards Japan. Results are included in the online appendix available at:

39 Today, Xinhua articles are rarely published as must-carry news. Interviews no. 13, 8, 21, 22.

40 An exception was the People's Daily, which relied more heavily than usual on reports by its Japanese correspondent.

41 Xinjingbao, 25 March 2005 and 6 April 2005.

42 Beijing Youth Daily, 5 April 2005, Xinjingbao, 6 April 2005.

43 See, for example, additional articles in Beijing Youth Daily, 5 April 2005, Jinghua shibao, 6 April 2005.

44 People's Daily, 6 April 2005. Similar reports were published in the China Daily.

45 See, for example,, accessed 4 August 2008.

46 According to the demonstration law from 31 October 1989, Chinese citizens have the right to demonstrate, but have to apply for permission at the Public Security Bureau first. See, accessed 23 July 2008.

47 On 10 and 12 April, media briefings with Foreign Ministry spokesman Qin Gang were convened. See, accessed 31 May 2007. For reports see Beijing Youth Daily, 13 April 2008, Jinghua shibao, 13 April 2005, China Daily, 13 April 2005. Similar announcements by the spokesperson of the Public Security Bureau followed. See People's Daily and Beijing Youth Daily, 22 April 2005.

48 Beijing Youth Daily, 24 April 2005.

49 Pen names that sound similar to respective state and Party units are sometimes adopted by officials when publishing in the Chinese media. Lingdao juece sinxi (Leadership Decision News), 4 February 2008.

50 People's Daily, 26 April 2005.

51 In the 1990s villagers preferred television and radio over newspapers. Zhang X., “Wo guo nongcun xinwen chuanbo xiantai yanjiu” (“A study of media communication in China's rural region”), in Chen C. and Mi X. (eds.), Zhongguo chuanbo xiaoguo toushi (A Perspective on Media Effects in China) (Shenyang: Shenyang chubanshe, 1989), pp. 146–66.

52 95.3% of Beijingers watched television, 72.5% read newspapers and 26% listened to the radio in 2000. Ke H., Meijie yu aoyün: yige chuanbo xiaoguo de shizheng yanjiu (The Media and the Olympics: A Quantitative Study of Media Effects) (Beijing: Zhongguo chuanmei daxue chubanshe, 2004). The pattern was similar in the BAS 2004.

53 In 1996, about 30% of Beijing residents subscribed to newspapers using their personal funds as opposed to public funds. Yu G., Meijie de shichang dingwei: yige chuanbo xuezhe de shizheng yanjiu (The Position of the Media Market: A Quantitative Approach to the Study of Communications) (Beijing: Beijing guangbo xueyuan chubanshe, 2000).

54 Ibid. This was consistent with the BAS 2004 data.

55 Most Beijingers read newspapers from two different newspaper types.

56 95% confidence interval ranged between 31 and 42%.

57 95% confidence interval ranged between 53 and 85%.

58 95% confidence interval ranged between 36 and 86%.

59 95% confidence interval ranged between 55 and 90%. Old Wang is only 33% likely to read commercialized papers (95% confidence ranged between 24 and 42%). A dummy variable for having travelled to countries in Europe or North America was dropped from the analysis, because it predicted the use of commercialized papers perfectly. Results can be retrieved from the author upon request.

60 For further explanations of how utility can induce selective exposure to information that conflicts with pre-held beliefs see Festinger L., A Theory of Cognitive Dissonance (Evanston: Row, 1957); Frey D., “Recent research on selective exposure to information,” Advances in Experimental Social Psychology, Vol. 19 (1986), pp. 4180.

61 After 9 April those Beijingers with the most negative views of Japan avoided newspapers. See D. Stockmann, “What kind of information does the public demand? Getting the news during the 2005 anti-Japanese protests,” in S. Shirk (ed.), Changing Media, Changing China, forthcoming. Since non-readers were excluded from the statistical analysis, avoidance did not influence the empirical results presented here.

62 The assumption here is that readers who were interested in Japan were also interested in the outcome of the visits. Indeed, according to the BAS sample, Beijingers who worry about Japanese imperialism tend to also feel threatened by Taiwanese independence (correlation was 0.6).

63 After both leaders had left the mainland official papers continued to feature articles related to their visits throughout May. For example, the People's Daily published 12% of all articles on Soong and 17% on Lien after their respective visits.

64 Moderately aware citizens tend to be most easily persuaded by news media messages, because poorly aware citizens do not receive media messages and the highly aware are more resistant to change their pre-held attitudes. McGuire W. (ed.), Personality and Susceptibility to Social Influence (Chicago: Rand McNally, 1968), pp. 1130–87; Converse P.E., “The nature of belief in mass publics,” in Apter D. (ed.), Ideology and Discontent (New York: Free Press, 1964), pp. 206–61; Zaller J., The Nature and Origins of Mass Opinion (New York: Cambridge University Press, 1992).

65 The coefficient of the interaction term was not statistically significant because of the small n of the treatment group. This indicates that we cannot be 95% certain that we would retrieve similar results over repeated samples. However, the dynamics are similar when comparing Beijingers' use of the internet and newspapers, thus further providing evidence that “new” media are more effective than “old” media in appeasing citizens. See Stockmann, “What kind of information does the public demand?”

66 Detailed statistical results are included in Table OA2, Figures OA1 and OA2 in the online appendix. Results show that the alternative explanation that people in a state of crisis will seek more information outlets was not confirmed. Readers tended to consume fewer rather than more newspapers after 9 April.

67 The side with higher audience costs is less likely to back down in a foreign crisis and therefore able to signal its intentions to other states more credibly than states with lower audience costs. Fearon J.D., “Domestic political audiences and the escalation of international disputes,” American Political Science Review, Vol. 88, No. 3 (1994), pp. 577–92. Weiss, “Powerful patriots: nationalism, diplomacy and the strategic logic of anti-foreign protest.”

* This article is part of a book project examining the impact of media commercialization on news content and public opinion in China, tentatively titled Propaganda for Sale. For fruitful research collaboration I would like to thank Iain Johnston, Shen Mingming and the members of the Research Center for Contemporary China. I am also grateful for financial support provided by the Harvard-Yenching Institute and the Center for Chinese Studies at the University of Michigan. Many thanks as well to the participants of the Chinese politics workshop at the University of Wisconsin, Madison for helpful comments and to Wang Mingde for research assistance.

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

The China Quarterly
  • ISSN: 0305-7410
  • EISSN: 1468-2648
  • URL: /core/journals/china-quarterly
Please enter your name
Please enter a valid email address
Who would you like to send this to? *


Altmetric attention score

Full text views

Total number of HTML views: 25
Total number of PDF views: 238 *
Loading metrics...

Abstract views

Total abstract views: 769 *
Loading metrics...

* Views captured on Cambridge Core between September 2016 - 16th January 2018. This data will be updated every 24 hours.