Africa Media Review Vol. 8 No. 2 1994 ©African Council for Communication Education The Agenda-Setting Function of the International Mass Media: The Case of Newsweek in Nigeria Enoh Tanjong* and Gary D. Gaddy** Abstract The agenda-setting function of the media in an international context is investigated here by examining the relationship between exposure to Newsweek by educated urban Nigerians and the importance they place on various world problems (controlled by demographic factors, cosmopolitanism and national media use variables). A combination of two research methods was employed: a secondary analysis of a sample of 1,213 "better educated" adults in three urban areas of Nigeria in November, 1981 commissioned by the United States Information Agency (USIA), and a content analysis of the international edition of Newsweek for October and November, 1981, seven weeks before and during the survey. The investigation yielded inconclusive results in its comparison of the Newsweek agenda to the magnitude of the relationships between reading Newsweek and selecting each of the same problems as important. Specifically, when comparing the Newsweek's coverage of 12 international problems to Newsweek readers' agenda (as compared to non-readers, controlling for background factors), measured from the estimated effects of Newsweek readership and reliance on the selection of each of these problems, a modest positive but non significant rank-order correlation (r = 0.475) was found. * Enoh Tanjong teaches in the Department of Journalism and Mass Communication, University of Buea, Cameroon. ** Gary D. Gaddy is a statistical consultant with the Institute for Research in Social Science at the University of North Carolina, USA. Le Role du Programme d'Action des Mass Media Internationaux: Le Cas de Newsweek au Nigeria par Enoch Tanjong* et Gary D. Gaddy** R6sum& Cet article etudie le role des media dans la determination d'importance des nouvelles dans un contxte international, en examinant le rapport entre l'exposition a Newsweekdes Nigerians instruits et cosmopolites et rimportance que ceux-ci placent sur de divers problemes dans le monde (le temoin de cette etude est les variables des facteurs demographiques, le cosmopolitanisme et l'utilisation des mfedia nationaux). On a utilise deux methodes de recherche: une analyse secondaire d'un echantillon de 1,213 adultes 'mieux instruits' et provenant de trois regions urbaines au Nigeria, en novembre 1981; cette recherche a ete dirigee par l'Agence d'Information des Etats Unis. La deuxieme methode etait une analyse du contenu de l'edition internationale de Newsweek pour les mois d'octobre et de novembre 1981, sept semaines avant et au cours de cette etude.Cette etude a donne des resultats peu concluant dans sa cbmparaison du role de Newsweek au liens etroits qui existent entre la lecture de Newsweek et la selection des problemes presentes dans celui-ci comme etant les problemes les plus importants. Pour dormer un exemple, une comparaison du reportage de Newsweek de 12 problemes internationaux avec la selection par les lecteurs de cette revue des problemes qu'ils considerent les plus importants a decouvert que l'eflfet de la lecture et la selection de ces problemes a une correlation modeste et positive, maisnonsigniflcatifder=0, 475. •Enoch Tanjong enseigne au Departement de Journalisme et de Mass Communication a l'Universite de Buea au Cameroun. •• Gary D. Gaddy est specialiste des statistiques a l'lnstitut des Recherches en Sciences Humaines a l'Universite de North Carolina aux Etats Unis. Introduction Following Goldlng and Murdock's (1978) assessment of cross-national media studies, research on the Third World should not simply extrapolate findings about media in the First World to circumstances elsewhere which are presumed to be embryonic microcosms of Western systems. One hypothesis about mass communication and public opinion rooted in an international function of the mass media is agenda setting. As formulated by Cohen (1963) in The Press and Foreign Policy, agenda setting was portrayed in the context of the foreign policy aspect of international relations. The major empirical and theoretical works on agenda-setting, however, have been conducted mostly in Western industrialized societies, particularly the United States, national media and often on domestic political issues. (McCombs and Shaw, 1972; McLeod, et aL, 1974; Becker, 1975; Becker, 1982; Siune and Kline; 1975, and Iyengar and Kinder, 1985). This study attempts to expand the investigation of the shaping of international public opinion to a Third World setting by examining the relationship between the priority an international news medium (Newsweek) placed on various world problems and the priority its educated Nigerian readers give them. Put in another way, the agenda- setting hypothesis is tested here by examining whether Newsweeks Nigerian audience's perceptions of the seriousness of international problems are related to the amount of coverage Newsweek devotes to them. If, as has been contended, the agenda-setting function of the media is strongest among unobtrusive issues and problems - that is, ones that are far from the direct contact and personal experience of the audience - the agenda of international problems tested here provide just such a set of seemingly unobtrusive issues. Further, this test allows for a naturalistic comparison to studies done in the First World without the possible confounding factor of a non-comparable agenda- setting mass medium. This is possible because the international edition of Newsweek distributed in Africa is generally the same in format and substance as the American national edition - the kind of medium used in many previous studies of agenda-setting in the United States. The study of international agenda setting is of more than academic interest; it is a matter of significant practical and political importance. At the international level, agenda setting has wide ranging implications - both for the arguments relating to cultural imperialism and for the understanding of the development of international relations and the subsequent development of foreign policy. If the "invasion" of foreign media brings with it the cultivation of "foreign" political priorities, these could certainly have greater Importance than the cultivation of foreign tastes in clothing or music. Nigeria as a Context for Study As a major Third World country, Nigeria is a good place to study beyond the Western world, especially given the scarcity of communication research in Africa. Nigerians comprise one of four Africans south of the Sahara. In addition, Nigerian political and media systems are among the most developed on the continent with a multi-party political system, substantial freedom of the press, and a broad diversity of media. By many Western democratic standards, most if not all of the independent black African countries south of the Sahara can only be said to have authoritarian press systems (Siebert et. aL, 1956). But as Hachten (1971) has noted, the Nigerian press has a firmer tradition of free expression than the majority of African countries. The political system of Nigeria at the time of the survey In 1981 was a multi-party, American-style presidential democracy. One consequence of this open political system is a diverse and relatively open media system. For instance. In 1981, Nigeria had 15 daily newspapers, 12 political weeklies, 25 radio and 20 television stations, the largest number of mass media outlets in all of black Africa (Mytton, 1983). Interestingly, the primary motivation for producing a newspaper in Nigeria, according to Mytton, was political rather than commercial. Saliency of International Problems World problems are many and widely scattered: some within the experience of the public but most far out of its reach. The mass media with their power to create Immediacy regardless of natural or man- made barriers can bring events happening In distant lands to the sitting rooms of the audiences. Considering the high level of abstraction sometimes involved in International issues, the concrete images of the mass media may create a salience for Issues otherwise Irrelevant. It would also be reasonable to assume that this salience created will be In proportion to the amount of coverage given these issues and will be strongest for international problems that are not within the direct experience of the audience. Method Two research methods were Integrated in the empirical test presented in this study: a secondary analysis of a sample of "better educated" adults In three urban areas of Nigeria in November of 1981 and a content analysis of seven weeks of the international edition of Newsweek for the weeks before and during the survey. Survey Data The secondary analysis was of a survey of national media audiences in Nigeria, commissioned by the United States Information Agency (USIA). A local survey research organization (Office of Research Bureau Nigeria), supervised by USIA specialists, conducted the field work. The sample included personal interviews with 1,213 adults aged 16 years or older with at least four years of secondary-school education living in three Nigerian cities. Sampling followed a multi-stage area probability design within these three purposefully selected Nigerian urban areas: Lagos in the West, Enugu in the East, and Kano in the North. Those with at least three years of university education were oversampled so as to comprise one third of the sample. Although the three urban areas differed substantially In size, roughly equal samples were selected from each. Analyses were performed on the unweighted total sample. Content Analysis To examine the potential relationship between Newsweek readership and the priorities assigned international problems, the international news content of seven weekly issues of Newsweek coverage was defined by column inches, position of the story as indicated by two binary (or dummy) variables representing "cover story" and "top of the week" billing and the number of Illustrations and pictures that accompanied the story. A confirmatory factor analysis of the four measures yielded one factor explaining 72% of the variance, so the four Indicators were then combined into a single measure of coverage using factor scoring coefficients as weights. These coefficients gave substantial weight to size because of its greater variance (0.36 being number of column inches), and somewhat less to the number of illustrations (0.50 beingnumber of illustrations) and minor weight to cover stories (0.10) and stories that were "top of the week" (0.09). A log was kept of all International Issues covered by Newsweek during this period to Identify all international problems covered and not just the 12 cited in the survey. Three international issues given notable coverage during this period but not among the 12 problems in the survey were (In order of coverage): the assassination of President Anwar Sadat of Egypt on October 6, 1981, the modernization programmes of mainland communist China and the proposed sale of AWACS planes to Saudi Arabia. Operational Definitions of Variables Education Education was measured as a two category variable: secondary school and higher education. (Due to selection, no one had less than four years of secondary school education and, because of direct over sampling, 35% of the sample had some higher education). Gender Interviewers coded the sex as male or female. Arising from this about 63% were male, primarily because of the selection of only the "better educated". Ethnic and Religious Groups While Nigeria is comprised of a variety of ethnic groups (more than ten major tribes speaking well over two hundred different languages), three tribal groups dominate: Hausa (15%), Yoruba (25%), Ibo (42%) with others making up the rest (18%). The two major religions represented in the sample were Christianity (75%) and Islam (24%). Since two major ethnic groups consisted predominantly of one religion (Hausas being mostly Moslems and Ibos Christians), the three major ethnic groups were coded as four binary variables (with two separate variables for the Christian and Moslem Yorubas) with "other" as the reference category in the regression analyses. Age Respondents were asked to report their ages, which were collapsed in six categories from ages 15 to 20, up to 51 years and above. (Approximately 77% were under age 30). Occupational Elites Five occupational groups were considered elite among this educated sample: academic leaders (11%), communication and media professionals (1%), business professionals (15%), government leaders (3%) and university and secondary school students (23%). Elite occupation was coded as a binary variable with 1 for elite and 0 for all others. Cosmopolitanism Dye (1966) defined as cosmopolitan, individuals whose scales of social experience, primary interest and involvement are international in scope. Cosmopolitan individuals perceive themselves primarily as members of a social organization larger than their local community. Three items were used as indicators of respondents' cosmopolitanism: how often they talk about international affairs with other people, whether they have to keep informed about international affairs for their work, and whether they have travelled or studied in other countries. National Media Exposure Exposure was measured for three national media: radio, television and newspaper. National radio exposure was a composite index of listening to three national radio stations (Radio Nigeria, Lagos; Radio Nigeria, Kaduna; the Voice of Nigeria, Lagos), with one point given for each station listened to. Television exposure was measured by asking the respondents to indicate how many days a week they spent watching television. A composite index of four national newspapers [Daily Times, Punch, New Nigerian and National Concord) constituted newspaper exposure, with one point for each paper read. Exposure to Newsweek For readership of international print media, respondents were also asked if they read Newsweek, Time, and Africa. For each, they were also asked if they relied on it as a source of international affairs information. Both readership and reliance were coded as binary variables. Because of the possible redundancy in coverage between news magazines, only Newsweekwas included in the agenda-setting analysis using an additive index of readership and reliance which both increases reliability and reduces multicollinearity, since the two measures are highly correlated. Saliency of International Problems The importance the audience placed on various international problems was measured via a battery of 12 international problems current during that period. Respondents were told: "Here is a list of international problems. Read over this list and tell me which of these you personally think are the two most serious problems at the present time?" • South Africa/Namibia/Angola fighting • Libyan involvement in Chad • Conflicting claims to Western Sahara • Fighting in El Salvador • Labour unions in Poland • OPEC and the price of oil • Iran/Iraq war • International terrorism • U.S. intervention in other countries • Soviet intervention in other countries • Cameroon border dispute • Economic gap between rich and poor countries Each problem was coded 2 if selected first, 1, if selected second and 0 if not selected. Sample Characteristics Due to the selection of only those with four years of secondary education or more and the over sampling of those with higher education, the demographic characteristics of sex, religion and ethnic groups did not accurately match the population of Nigeria as a whole. Male respondents were almost twice (63%) as female respondents (37%) in a population where females equal or outnumber their male counterparts. Some additional bias in the sample towards men may have been a function of the urban sample, since men dominate the urban work force. For similar reasons, Christians and Ibos are over- represented in comparison to the population estimates. Results Even though the international news magazines examined in this survey were published in English, readership among the educated in Nigerian urban areas Is quite high. About 45% of the respondents said they read Newsweek: 26% Time; and 12% Africa. These numbers were comparable to the numbers who reported listening to the major international radio broadcasts (53% for the BBC and 25% for the Voice of America). Evidence of Agenda-setting A comparison of the agenda of Newsweek readers and non-readers (as indicated by the proportion selecting each of the 12 international problems listed in the survey) indicated a close correspondence between the two, (Table 1 columns 2 and 3). The first six problems (although slightly re-ordered) were the same for both groups: South Africa/ Namibia/Angola lighting, OPEC and the price of oil, economic gap between rich and poor countries, Libyan involvement in Chad, the Iran/Iraq war and the Cameroon border dispute. All of these problems had the common thread of being related to the daily lives and near to the dally experience of the audience than the problems less frequently selected: the Iran/Iraq war and OPEC and the price of oil being near to Nigeria as an oil exporting nation member state and a substantially Moslem country. The six less frequently chosen problems were identical in order for both groups: U.S. intervention in other countries, labour unions in Poland, conflicting claims to the Western Sahara and the fighting In El Salvador. All these problems tended to be more abstract and further from the audience's immediate experience. When readers' and non-readers' priorities were compared statistically, casual observation was confirmed: there was no significant difference between the two groups (Spearman rank order correlation coefficient r = 0.04; p < 0.05, df = 11). A comparison of Newsweek readers' priorities for the 12 international problems and Newsweek coverage indicates little relationship (r = 0.13). For non-readers, the correlation was 0.02, which is clearly not significant (Table 1, columns 1,2 and 3). These data may seem to suggest that agenda-setting was not at work here. These comparisons, however, did not present clear evidence of the agenda-setting hypothesis, since uncontrolled correlations between media coverage and audience priorities do not take into account pre-existing differences between readers and non-readers. To control these differences, regressions were run predicting the selecting of each of the 12 international problems as the first or second most important, controlling for demographic, cosmopolitan and national media exposure variables, and Newsweek readership/ reliance. The betas for Newsweek readership/reliance were used as indicators of the magnitude of the effect of reading Newsweek on the selection of each problem as important. The betas of Newsweek readership/reliance predicting selection of the international problems were rank ordered then tested against the proportion of coverage given by Newsweek to these problems (Table 1, columns 1 and 4). If there is evidence of Newsweek influencing Table 1: Agendas Based on Newsweek Coverage, Readers and Non-Readers Seiectto. International Problems Economic gap between rich and poor countries U.S intervention in other countries Labour unions in Poland International terrorism Fighting, in South Africa, Namibia and Angola Fighting in ElSalvador Cameroon border dispute OPEC oil crisis Libyan Involvement in Chad Soviet intervention In other countries Conflicting claims to Western Sahara Iran/Iraq war Newsweek Coverage (rank) Readers' Selection (rank) Nonreaders' Selection (rank) Newsweek Beta* (rank) 39.2% (1) 13.2% (3) 10.5% (5) 26.7 (2) 9.0 (3) 6.7 (4) 4.9 (5) 4.2 (6) 4.1 (7) 3.5 (8) 1.8 (9) 0.0 (11) 0.0 (11) 0.0 (11) N = 31 5.4 (7) 1.7 (10) 5.0 (8) 33.5 (1) 0.6 (12) 5.7 (6) 15.0 (2) 8.0 (4) 2.6 (9) 1.4 (11) 7.9 (5) 540 4.6 (7) 1.6 (10) 3.1 (8) 30.9 (1) 0.9 (12) 11.1 (4) 12.3 (2) 9.4 (6) 2.4 (9) 1.3 (11) 11.9 (3) 6 73 .01 (5) .04 (2) -.03 (10) .01 (4) .07 (1) .01 (3) -.08 (12) -.00 (7) -.02 (9) -.05 (11) .00 (6) -.02 (8) 1,213 Betas for Newsweek Readership Predicting Selection of 12 problems as important. * Betas of Newsweek readership/reliance controlling for sex, age, education, elite occupation, ethnic and religious status, talking about international affairs, job requiring exposure to international affairs, foreign travel and study and exposure to national radio, television and newspapers. None of the betas are significant at 5% level of significance. 10 its readers in the importance they assign to these problems, then larger betas should be observed among the international problems given most coverage (labour unions in Poland, the economic gap between rich and poor countries, U.S intervention in other countries and international terrorism) compared to the least covered international problems (conflicting claims to the Western Sahara, Soviet intervention in other countries, OPEC and oil crisis and the Cameroon border dispute). In other words, an agenda-setting effect should manifest itself as a correlation between the coverage given an issue and likelihood that a reader (in contrast to a nonreader) will select that issue as important, taking pre-existing characteristics into account. While the rank-order correlation between Newsweek coverage of international problems and audience priorities of 0.475 does suggest some relationship, it is not strong enough to be statistically significant (r = 0.475; p > 0.05.) In addition, none of the individual betas measuring the relationship of Newsweek readership/reliance on selection of any of the problems is either modest in size or statistically significant. Further, there is little suggestion that the priority assigned to more distant problems, which were given substantial coverage, was changed any more than that given to closer problems. The smaller rank-order comparison of the Newsweek agenda for the six weeks after the survey (using regression to correct coverage after the survey period for the inertia in coverage from the weeks before and during it) with the betas for readership (r = .322) again suggest but fails to significantly support an agenda-setting effect. Thus, although the positive relationship after controlling for differences between readers and nonreaders is suggestive (as is the comparison between "readership effects" and Newsweek before and after the survey), the results fail to clearly confirm the agenda-setting hypothesis. Discussion The tendency of many researchers in international communication has been to see Third World audiences as passive receivers of international influence, whether in the form of information, opinions or priorities (Izcaray & McNelly, 1987). The evidence presented here of one such audience, however, suggests some immunity to such effects. While the distance of "international problems" from the everyday life of ordinary people (whether In the First, Second or Third World) would seem to recommend this as fertile area for the cultivation of agenda by the media, there are some reasons why such an effect might not show up in any substantial way. 11 One reason this audience might be resistant to such Influence Is apparent in the demographic characteristic of these Newsweek readers. They are educated people who are, relatively speaking, more widely travelled and widely read, and more interested in International affairs. Hence, people likely to have their own Interests, political leanings and points of view on international affairs as well as multiple sources of information - both within and without the media. So, while they read Newsweek, they would not necessarily blindly follow the Newsweek agenda. This suggests, perhaps, that if such agenda-setting effects are to be found, international radio may be a better medium on which to focus, since the audience of international radio is not as educated or sophisticated as the International print audience. Further, the argument that Hale (1975) has made generally in favour of the greater impact of international communication In the Third World In general, may explain the lesser impact found in this study. Hale contends that the tight controls over domestic media systems exercised in much of the Third World reinforce the impact of foreign media, since the lower the quality of the local media, the higher the credibility of foreign media. Thus, the relative freedom and diversity of the Nigerian media may lead to diminished effects from the international media. Other countries with more authoritarian media systems may be better placed to look for such effects. Another set of reasons for the weak effects found here may be methodological. One is the measure of Newsweek exposure which consists here of the sum of two dichotomies. Future studies should include at least estimates of frequency and time of exposure if not measures of attention to specific content as well as the purpose behind that exposure. Further more, the statistical power of a rank-order correlation is limited by the number of items ranked. Thus, despite the survey sample of 1,213, the power of the test presented here was constrained by the use of 12 issues. Future studies should offer a larger number of issues, and measure their salience via a more sensitive technique (such as rating each in importance on a scale of 0 to 100) rather than selecting the "most important" and "next most important". Finally, future studies should examine the agenda-setting hypothesis with the use of over-time data modelling the dynamic aspect directly rather than by proxy as done here. Conclusion Despite the multiplicity of explanations that may be raised to dismiss the null results found here, the simplest remains the most plausible: Newsweek has little, if any, impact, on the agenda of international 12 problems of its "better educated" Nigerian readership. Even if the rank-order correlation found were significant, the consistently modest size of the betas across the 12 issues suggest the impact of Newsweek coverage on individual reader's issue salience is limited. If this is so, it puts a dent both in the conception of Third World audiences as passive and malleable, and the revisionist view of media effects the agenda-setting hypothesis represents. While the evidence of one study cannot deny previous demonstrations of agenda setting (Iyengar et aL, 1985), it does raise a question about the limits of even this theoretical response to the limited effects model of the media. References Becker, Lee B. (1982). The Mass Media and Citizen Assessment of Issue Importance: A Reflection on Agenda-Setting Research.' In D.C Whitney, E. Wartella and S. Windahl (Eds.). Mass Communication. Review Yearbook, Vol.3, Beverly Hills, CA: Sage. Becker, Lee. B., Maxwell McCombs and Jack McLeod. (1975). The Development of Political Cognitions.1 In S.H Chaffee (Eds.). Political Communication: Issues and Strategies for Research. Beverly Hills, CA: Sage, pp. 21-630. Cohen, Bernard C. (1963). The Press and Foreign Policy. Princeton, NJ: Princeton University Press. Dye, Thomas R. (1966). The Local-Cosmopolitan Dimension and the Study of Urban Polities'. Social Forces, Vol. 41, No.3 pp. 239-246. Golding, Peter, and Graham Murdock,(1978). Theories of Communication and Theories of Society'. Communication Research, Vol. 5, No. 3 pp. 339- 356. Hachten, William A. and Harva S. Hachten (1971). Muffled Drums. Ames, IA: Iowa State University Press. Hale, Julian (1975). Radio Power: Propagandaand International Broadcasting. Philadelphia: Temple University Press. Iyengar, Shanto and Donald R. Kinder (1985). 'Psychological Accounts of Agenda Setting'. In Sidney Kraus and Richard M. Perloff (Eds.). Mass Media and Political Thought: An information-Processing Approach. Bervely Hills CA: Sage, pp.117-140. Izcaray, Fausto and John T. McNelly (1987). 'Selective Media Use by Venezuelans: The Passing of the Passive Audience in Rapidly Developing Society'. Studies in Latin American Popular Culture. Vol. 6 pp. 27-41. McCombs, Maxwell E. and Donald L. Shaw. (1972). The Agenda-Setting Function of Mass Media'. Public Opinion Quarterly. Vol. 36, pp. 176-87. 13 McLeod, Jack M., Lee B. Becker and James E. Byrnes (1974). 'Another Look at the Agenda-Setting Function of the Press'. Communication Research Vol. 1, pp. 131-166. Mytton, Graham (1983). Mass Communication in Africa. London: Edward Arnold Publishers Ltd. Siebert, Fred; Theodore Peterson and Wilbur Schramm (1956). FourTheories of the Press. Urbana: University of Illinois Press. Siune, Karen and Gerald Kline (1975). 'Communication, Mass Political Behaviour, and Mass Society'. In S.H Chaffee (Eds.). Political Communication: Issues and Strategies for Research Beverly Hills, CA Sage. 14