Public attitudes toward media control and incitement of conflicts in Eastern Africa

Abstract

Public opinion research in two post-conflict African countries, Ethiopia and Rwanda, offers new insights into how audiences view the trustworthiness and performance of different media and different ownership sectors. This survey examines how citizens of Rwanda and Ethiopia describe the role of mass media in covering conflicts and in stoking the attitudes that can lead to or worsen conflicts. State-owned media are widely used as a source of news, yet are also widely distrusted, particularly when covering conflicts. They are also seen as less reliable than the privately-owned media. Still, different levels of use and confidence in state media suggest that top-down reconciliation efforts can have measurable results. Media can play a role in damping the flames of ethnic conflict, much as they can play a role in fanning those flames.

Keywords

Africa conflict media peace journalism surveys

The media have played a notable role in mobilizing mass action in developing countries, particularly in Africa. This mobilizing power helped to underpin the postcolonial hopes of the era of development journalism, but it was also associated with disastrous consequences. State propaganda on national media escalated the ethnic civil conflicts that culminated in genocide in 1994, during Rwanda’s guerrilla war, and Ethiopia’s border conflicts with Eritrea and Somalia in 1993 and 1994. Although previous explorations have advanced our understanding of how people use media in conflicts and how press performance affects democratization, security, and other important factors, much terrain remains to be examined on how citizens evaluate their media in times of conflicts/post-conflicts in Eastern Africa. The pressure to privatize and liberalize state media and the growing influence of self-directed interactive media (e.g. Aginam, 2005), underscore the importance of understanding how the changing media landscape affects public attitudes. This study specifically compares attitudes toward the performance of privately-owned and state-owned media in Rwanda and Ethiopia, two Eastern African countries that experienced some of the worst conflicts of the 1990s in Africa – both fueled in large part by the national media (Snyder and Ballentine, 1996).

These two countries offer important examples of the news media’s role in informing citizens about state affairs in ways that can be detrimental to democratic values. The government and the military in Rwanda used the national radio to spread fear, rumors, and panic through kill-or-be-killed rhetoric, which incited hatred and led to ethnic cleansing. Radio and television propaganda instigated the killing of more than half a million Tutsi people and also gave instructions how to do it (Mamdani, 1996). In Ethiopia, the broadcast media were used by the government to spread hate propaganda, which helped trigger political violence that led to the death of more than 300,000 people in a span of 20 years (Fessehatzion, 2003). These attributes underscore the utility of this study to such central concepts as media control and mass mobilization in times of conflicts. While these countries should not be assumed to be representative, and while generalizations about the roots of conflict will always remain problematic, this examination of media influence there offers ideas and a structure for understanding the interplay of media, policy, and publics in other countries where conflict is a recent memory or a looming risk.

Understanding public attitudes toward such influences could hold lessons for governments, educators and trainers, and the media themselves. The implicit promise of studying media freedom as a conflict inhibitor, and of more specific propositions like peace journalism, is that the same media that mobilize for war can also demobilize – or at least promote the creation of room to talk peace or negotiate. The peace journalism model offers a contrast to a traditional model of impartial journalism, in which news reports simply focus on the newsworthiness of the conflict, regardless of whether that decision itself represents a series of ideological decisions that play a role in whether or how a conflict continues.

Even critics of such concepts (e.g. Hanitzsch, 2004) have pointed to the importance of further study of the effects that media practices, systems, messages, and contexts can have on conflict at many levels. Thus, this study seeks to expand understanding of African audiences and their views of how media cover conflicts and inflame the attitudes that can lead to or worsen them. A stratified multistage random sample conducted from late 2008 into early 2009 in Rwanda and Ethiopia allows for comparisons of media use and media trust across social and occupational categories.

Literature review

The pen, the microphone and the sword

Recent research has consistently pointed to the importance of press freedom in restraining or suppressing interstate conflict (e.g. Choi and James, 2007, 2008; Van Belle, 1997, 2000) – as long as press freedom is shared by the potential parties to an armed conflict. This is primarily conceived as a structural factor. Free-press states treat others’ news as legitimate, thus constraining either side’s ability to dominate a narrative (Van Belle, 1997). Open media make bluffing more difficult at home and reduce uncertainty in international signaling (Choi and James, 2007), and independent reporting makes it harder for leaders to hide the costs of a potential conflict from their audiences (Choi and James, 2008). As with the democratic peace, in which context those studies arose, this is an effect of shared press freedom, rather than press freedom itself. The existence of a normative effect – that a free press, like democracy, is an indicator that a country is more peaceable in general – remains tantalizing but unsupported. Still, this growing body of research underscores the importance of considering the media in studying conflicts.

Not all effects of the press are so benign, and an even broader tradition of study points to potentially harmful results of the news process. The needs of news, which include immediacy, simplicity, and the sort of news values epitomized by the primacy of conflict, are often fundamentally at odds with the needs of peace (Wolfsfeld et al., 2008). Indeed, ‘the standard definitions of what is considered news generally ensure a steady flow of negative and threatening information about the other side’ (p. 2). Nor is this solely a function of state-controlled or state-influenced media. As criticisms of US press performance in the run-up to war with Iraq in 2003 have underscored, an independent press is also free to beat the war drums by its own choosing.

That freedom is reflected in the underlying concepts of peace journalism: if the media can choose stories, words, and images in ways that propagate and inflame conflict, they can equally choose stories, words, and images in ways that defuse tension and promote negotiation (Fawcett, 2002; Lee and Maslog, 2005; Ross and Tehranian, 2009). This intuitively attractive notion has been questioned for its perceived inattention to the complexities of news routines (Wolfsfeld, 2001) and for its theoretical reliance on large media effects and passive audiences (Hanitzsch, 2004). Still, peace journalism remains a provocative addition to the study of media in conflict. For this article’s examination of attitudes toward media control, it is also a reminder that if media can be encouraged from above to incite violence, they can also be encouraged to counter it.

Ethnic conflicts like the ones discussed here can be distinguished from the larger category of civil wars in general. An ethnic civil war is a ‘war in which the key issues at stake – that is, the express reasons political power is being contested – involve either ethnic markers such as language or religion or the status of ethnic groups themselves’ (Kaufman, 2006: 202). Thus, ethnic conflicts are considered less susceptible to efforts to address other factors involved in civil war, such as information shortages or commitment problems. The very web of myths and symbols that creates ethnic identity can also justify hostility against the ethnic other. If this appears particularly true of intrastate and asymmetrical conflicts, it could be because most conflicts in recent decades deal more with such status issues – the ‘very structure of the relationship between the parties’ (Wolfsfeld et al., 2008: 376) – than with the claims traditionally at stake in more symmetric conflicts.

Kaufman’s (2006) symbolic politics theory holds that ‘ethnic wars are driven by hostile popular emotions toward out-groups, emotions harnessed by political leaders wielding emotive ethnic symbols.’ This can backfire into a ‘symbolic politics trap’ that leaders are unable to escape even if they want to. Disputes among ethnic groups can produce a security dilemma of their own, the crux of which ‘is that actors, through efforts to enhance their own security, provoke fear and countermeasures, with the result that less rather than more security is obtained in the end’ (Melander, 2009: 99). To slow or avoid such conflicts, elites are encouraged to tone down not only their own speech but those of the media they control. Kaufman’s (2006) prescription reflects the goals of peace journalism: The media can be encouraged to humanize the enemy, and journalists can be trained to address the inherent bias that results from addressing a narrow audience. In less normative terms, this concept reflects Wolfsfeld et al.’s (2008: 376) understanding that the social construct of news assumes that journalists hold a set of ‘working assumptions about the nature of the political world’.

At the system level, ‘the transparency resulting from independent media coverage’ (Melander, 2009: 106) is an important factor in reducing intergroup uncertainty. Peace-building in post-conflict situations, then, should not only keep an eye on what the potentially inflammatory established media do but should encourage the growth and development of more professional, socially responsible media outlets (Paris, 2004). These considerations underscore the importance of multiple media sectors, those owned but not managed by the state and those in private hands but independent from state influence, as well as those published or broadcast from abroad. That complicated maze of ownership and medium is particularly relevant in Africa and, while these two countries do not constitute a generalizable sample, they suggest the variety of issues that need to be borne in mind.

Broadcasting has been one of Africa’s major forces for social and political mobilization, though the role of media in the democratization process has been slow to develop. Radio and television were often used to promote development and public health goals. Newspapers were primarily a source of information for elites, and radio, though seen as ‘the people’s medium’, has only rarely ‘fulfilled the democratic potential embedded in these thoughts’ (Van der Veur, 2002: 81). In autocratic regimes with limited press freedom, this trend is reflected in higher levels of regime support and anti-democratic attitudes in heavy users of broadcast media, compared with newspaper or internet users (Norris, 2008). It is also reflected in an emphasis on the privatization of radio without a parallel emphasis on independent public service broadcasting (Aginam, 2005; Davidson, 1998). Such transitional regions as the former Yugoslavia have seen that a decentralized or privatized press is not always a more peaceable press (Kurspahic, 2003). Paris (2004: 76) suggests that a ‘vibrant but irresponsible’ press sector in Rwanda contributed to the polarization that helped make the 1994 genocide possible.

Major General Paul Kagame, who has presided over the Government of Rwanda since the 1994 genocide, has gradually transformed and reformed the media both as a public and a private press (Karnell, 2002). The majority of Rwandese, particularly in rural Rwanda, depend on radio for information and current affairs. In both Rwanda and Ethiopia, radio news presenters also review the major headlines and main stories in newspapers every morning. A few Rwandese and Ethiopians can afford to buy newspapers, and some readers ‘rent’ the papers from the vendors every late afternoon to peruse the pages at a fraction of the cost. Television is a comparative newcomer, and the distinct attitudes this study uncovers toward older and newer media underscore the complex patterns created by different forms of ownership and control.

Conflict and media in Rwanda and Ethiopia

Rwanda and Ethiopia provide an appropriate initial comparison for a number of reasons. Both countries are part of greater Eastern Africa and both peoples come from the Bantu clan. In the last 20 years, the governments of Rwanda and Ethiopia have both been overthrown by their military and have also suffered attempted coups d’état (Herbst, 2004). Both have been governed by generals, have a strong military presence and representation in their legislative process, and have used the media to set off ethnic violence. Rwanda and Ethiopia have also implemented constitutional reforms that have been widely seen as a way of consolidating the president and military regime’s grip on political power (Hyden and Lanegran, 2005).

Further, Rwanda and Ethiopia stand at parallel ends of a political continuum in terms of governance, military affairs and relationships, and institutional polity. At one end, the two countries have gone down in history as the epitome of an African holocaust (Herbst, 2004; Mazrui, 1980). Barely 40 years ago, both countries were governed by kings and emperors, yet both monarchies experienced defiance from some groups of their people, who opposed customary rule and hereditary politics to forge a powerful spirit of unity. In that context, the two countries also present different patterns of political development in recent decades, and the distinct forms their conflicts took allow for broader insights from a single set of measurements of media attitudes.

The media and Rwanda’s genocide

Throughout the 1980s, Rwanda’s national radio and television were used to promote public health and safety and modern farming, reflecting traditional goals of development journalism. France provided financial aid to the Rwandan Government to buy thousands of cheap radios and batteries to disseminate civic education and to promote modern farming techniques (Uvin, 1998). Rwanda also built stronger radio transmitters and repeater transmission towers to cover the entire country in the late 1980s and early 1990s. Yet for all those hopeful goals, at a critical point in Rwanda’s development, the media fed the ethno-political tensions that led to genocide, rather than slowing them (Rotberg and Weiss, 1996; Sonwalkar, 2004).

From the early 1990s, the Rwanda Government owned the only television station and local radio stations in the country under the umbrella of the National Information Office (ORINFOR). The only non-state broadcast media were international outlets (Hachten, 2004). In early efforts to fuel hate between the ethnic groups, state agencies, including the army, used their authority over the state-owned media to unite the Hutu against the Tutsi (Karnell, 2002; Olorunnisola, 1995). When the Rwandan Patriotic Front (RPF) invaded Rwanda in early 1990, Radio Rwanda began to promote the killing of Tutsis in regions surrounding the capital, Kigali (Chalk, 1999; Karnell, 2002).

When the RPF rebels took over some towns in the north, the government used Radio Rwanda to urge Hutu solidarity against what broadcasters called the Tutsi intruders and traitors (Des Forges, 2002; Reyntjens, 1999). Pro-government politicians used the airwaves in 1991 to incite mass killings of those they called Tutsi supporters of the RPF and Hutu moderates who wanted to take power from the Hutu President, Juvénal Habyarimana (Gatwa, 1995).

Meanwhile, opposition politicians forced Habyarimana to appoint them to various agencies and eventually formed a coalition with his government. Some of these moderate Hutus were liberal and socially inclusive (Prunier, 1995). Among the ministries they controlled was the newly created ministry of information, responsible for ORINFOR (Des Forges, 2002).

The moderate Hutus managed to enact a new press law in 1991 (Fardon and Furniss, 2000), and the following year, Radio Rwanda was freed from state control. But the president and his allies maneuvered their way back into control of ORINFOR, and the newly established private radio RTLM (Radio-Tèlèvision Libre des Mille Collines) was taken over by a Hutu militia called the Interahamwe that served the Rwandan Government. Programs on RTLM and ORINFOR shattered the bonds of solidarity between Hutu and Tutsi (Schabas, 2000). The Interahamwe carried out a series of targeted massacres on Tutsi minorities and Hutu moderates, even before the genocide of 1994 (Des Forges, 2002).

When a plane carrying President Habyarimana was shot down in April 1994 as it approached Kigali, RTLM and ORINFOR mobilized Hutu civilians, the army, the Interahamwe, and the Presidential Guard to hunt down Tutsis and moderate Hutus (Schabas, 2000). That signaled the beginning of the genocide. RTLM and ORINFOR proclaimed that the Tutsi were to blame for Habyarimana’s death and the RPF invasion (Chalk, 1999; Prunier, 1995).

Over the next three months, hundreds of thousands of unarmed Rwandese died. Students killed teachers and vice versa, Hutu parents killed children from Tutsi spouses, neighbors slaughtered neighbors, and priests lobbed grenades at congregants who had sought shelter in churches, as broadcasts urged them on, even providing instructions (Frohardt and Temin, 2003; Gourevitch, 2000; Mamdani, 1996; Prunier, 1995). The dead included 52 Rwandese journalists who were identified as Tutsis or moderate Hutus (Kalyango and Eckler, 2010).

The media after the genocide

Rehabilitating the media presented an enormous challenge to the Rwanda Patriotic Front (RPF) Government that overthrew Habyarimana’s allies to end the genocide. The government secured funds from international agencies such as the IMF, UNESCO, and others to renovate equipment. These and other organizations also provided emergency funds to recruit and train young journalists (Karnell, 2002). Local and international training efforts lasted for more than a decade, and the new generation of reporters was charged with sensitizing citizens to respect human dignity, human rights and the rule of law, and to foster peace, reconciliation and tolerance. Despite these developments, the Rwandan Government had its own agenda.

A media law approved in 2002 established the High Press Council to guarantee freedom of the press. Yet this body regulates electronic broadcasting and imposes censorship on the press with a legal guideline of preventing a repeat of the 1994 hate speech (Reyntjens, 2004). The first privately-owned independent radio station was established in 2004, but all private broadcast and print media face government restrictions and generally exercise self-censorship. Some proprietors of the private press are members of Rwanda’s governing council.

Critics argue that journalists who investigate government corruption, authoritarianism and abuse of public office face harsh restrictions and intimidation (Franklin and Love, 1998). A total of 17 well-known journalists have been incarcerated in the last 10 years, some without trial. Some topics are forbidden, such as covering Rwanda’s military, the justice department, the State House, and any tyrannical acts committed by the RPF.

The media in Ethiopia

The first major tensions that involved a crackdown on the press escalated when Ethiopia annexed Eritrea in 1962 (Allen and Seaton, 1999). From 1962 until the early 1990s, Ethiopia banned Eritrean political parties, closed some of the vocal gutter press, and arrested journalists who raised concerns about Ethiopia’s hegemony (Jacquin-Berdal and Plaut, 2005). Throughout the period under President Haile Mariam Mengistu (1974 to 1991), a protracted civil conflict cost an estimated 100,000 people their lives, and more than 50 journalists were imprisoned or disappeared inside the country (Allen and Seaton, 1999; Fessehatzion, 2003).

In 1991, the Ethiopian People’s Revolutionary Democratic Front (EPRDF) overthrew Mengistu, which escalated the conflict with Eritrean separatists and weakened the military’s supremacy (Abbink, 1994). Eritrea managed to gain its independence in 1993, but discord between the neighboring states has continued since then.

During a 1998 conflict with Eritrea, Ethiopian Prime Minister Meles Zenawi’s government appointed some politicians to manage the national broadcaster in order to effectively control its ideological mass appeal for bloodshed (Lyons, 2007). Zenawi’s government used state- and privately-owned media to spread propaganda against the Eritrean people and government (Simon, 2006). Sources in newscasts called on citizens to rise up to protect national sovereignty at all costs (Simon, 2006). Consequently, media watchdogs like Reporters without Borders (2007) argue that Ethiopian media lack an independent voice when it comes to conflict.

Broadcasts that stoked anti-Eritrean sentiments also provided cover to unknown assailants to commit atrocities against both Ethiopian and Eritrean civilians because the state had a common enemy, the Tigrai guerrilla movement, to blame (Fessehatzion, 2003). With adult literacy at an estimated 42 percent, citizens rely on broadcasting for information, and according to Simon (2006), most media consumers accept inflammatory broadcasts as absolute facts.

As of December 2008, Ethiopia had one privately-owned daily newspaper, six main radio news stations (two privately owned), and one national/main television station. Just like Rwanda, Ethiopia launched its first truly privately-owned commercial radio stations in early 2007. Two stations, Sheger Radio and ZAMI FM, went on the air in mid-2007. The government imposed stringent restrictions, especially for ZAMI FM, which was licensed to provide news and talk shows. So, for the new ZAMI FM and other privately-owned press, the government required them to hire news managers from the state-owned national broadcasts (International Federation of Journalists: IFJ, 2007).

Like Rwanda, Ethiopia lacks a strong industrial base, which would support private and independent media with advertising revenue. As a result, commercial media are also forced to depend on government for advertisements and tax breaks or waivers. Radio programming, a key source of news and current affairs, is often weak because personnel are poorly trained and underpaid (Amogne, 2005). Draconian media laws have been used to close news outlets and arrest editors and journalists (Amogne, 2005; Simon, 2006), and journalists covering conflicts have suffered punitive sentences for exposing indiscriminate killings of civilians and violation of human rights (Amogne, 2005). Privately-owned media are restricted in conflict coverage, while state-owned media are given privileged treatment – and even allowed to sell information to the privately-owned media.

Despite similarities in media use, media development, and legal or economic restrictions on independent journalism, these conflicts played out in different ways: in Rwanda’s case the descent into genocide, in Ethiopia’s, a conflict that exacted a large toll in human life yet remained more restrained. Public attitudes toward media performance, reliability, and incitement, then, can help illuminate the role of media in conflicts and in maintaining peace.

After reporting survey results on media consumption, media trust, and occupational characteristics of the population, the study will explore the following research questions:

RQ1: How do citizens of Rwanda and Ethiopia describe the role of mass media in ethnic conflicts?

RQ2: Which media do different respondent groups in Rwanda and Ethiopia trust with coverage of conflicts and crises?

Methods

Survey research

Cross-sectional surveys were conducted in Rwanda and Ethiopia between 6 December 2008 and 31 January 2009. The goal was to survey 500 citizens of voting age from each country. A total of 467 respondents from Ethiopia completed the survey instrument out of 500 that were randomly handed out, and a total of 403 respondents from Rwanda completed the surveys out of 500 that were randomly handed out. Citizens were selected using a stratified multistage cluster random sampling. A ‘citizen’ was defined as an individual over 18 years of age who resided in that country. Survey coordinators also informally screened respondents to ensure that those chosen for the survey were able to articulate their interests and attitudes clearly. Public opinion and attitudes were sought on issues ranging from media use and influence, coverage of conflicts, and attitudes towards government actions.

Measures of media use were meant to take into account the demographic groups in which different media are most popular. Radio is very popular among both rural and urban dwellers. Television is restricted largely to urban dwellers who can afford a TV set and to areas with electricity and lots of bars or other social centers. Newspapers are read by literate citizens from urban areas who can afford them, but they are also a shared product through privately-owned radio in urban and rural communities.

The stratified random sampling ensured that the samples from each country were proportionately categorized for comparative purposes. Since both countries are ethno-geographically heterogeneous, an equitable sample of respondents that represents the opinions, attitudes, and demographics of citizens was taken. The following stratification was followed in both countries: 50 unemployed workers randomly selected from the latest records from civil society organizations such as churches, mosques, and job-training or vocational centers; 50 students; 50 civil servants randomly selected from a list provided by the Ministry of Public Service; and 50 members of the business community randomly selected from a list of each country’s Chamber of Commerce and Industries as well as the private sector associations. Also included in the data were 50 professionals, such as teachers, nurses, lawyers and others, and 50 entertainers, including artists¹ and sports personnel who were randomly selected from a list provided by the Departments of Youth, Culture and Sports. Also randomly selected were 50 housewives from urban areas and townships, and 150 farmers and peasants from farmlands, upcountry villages, and townships from a list provided by district officers. Farmers and peasants represent over 65 percent of the population in Rwanda and Ethiopia.

In the multistage cluster sampling, the first stage was to randomly select a cluster of farmlands, upcountry villages, and townships. Then a second cluster of homes was randomly selected. Finally, a respondent was randomly picked in a household. For ease of analysis in this pilot study, respondents’ occupations were categorized as white-collar (teachers, professionals, civil servants, journalists, and those registered as seeking work) and blue-collar (businesspeople, farmers, artist/entertainers, traders, laborers, housewives and law enforcement). These categories were determined from the lead researcher’s experience in East African business and journalism.

An approximate representation of the population was thus assured from house to house, at workplaces, and also at academic and public/state institutions. A total of 12 field survey coordinators were recruited from each country to administer the survey questionnaires. Rwandese and Ethiopians are often less than willing to participate in academic projects like this one due to personal security reasons, especially with sensitive questions concerning their political leaders. All field survey coordinators had to demonstrate good communication skills for eliciting cooperation. They were matched with a region of their ethnicity or residence where their presence would create trust and comfort. Matching field survey coordinators to particular strata considering gender, language, religion, and appropriate caste was meant to minimize concerns about political entrapment A single survey instrument with identical questions was used in both countries.

The field coordinators gave the survey instrument to respondents in hard copy (paper format). Most respondents preferred to fill out an English version, although translated versions in Kinywarwanda, Amharic, Oromo, and Tigrinya were at hand. Two primary reasons support the need for a face-to-face survey: more than 80 percent of the population in both countries has no access to either a conventional land telephone line or cellular phone, and an estimated 70 percent of residences do not have access to a physical-address infrastructure where a mail survey or self-administered survey can be effectively conducted. The questionnaire was piloted on five Rwandese and five Ethiopians to ascertain the reliability of the items used to construct the scales.

The instrument measured media use and exposure and attitudes toward coverage of conflicts and governance to assess degree of agreement or disagreement on matters of trust, interest, influence, coverage, government actions, and other attributes. Other items sought opinion about news coverage, the privately-owned and state-owned media, and media influence. Attitudes about media coverage of conflicts and other government actions were also sought.

Results

Descriptive findings

The survey yielded 870 complete responses, 467 (53.7%) in Ethiopia and 403 (46.3%) in Rwanda. Of the total (excluding students), 352 (43.8%) reported white-collar occupations and 451 (56.2%) reported blue-collar occupations. In Ethiopia, the proportion was 189 (42.8%) white-collar, 253 (57.2%) blue-collar; in Rwanda, it was 163 (45.2%) white-collar, 198 (54.8%) blue-collar (see Table 1).

Table 1.

Self-reported occupation categories for all respondents (%).

	Both countries	Ethiopia	Rwanda
Businessperson	52 (6.0)	29 (6.2)	23 (5.7)
Farmer	53 (6.1)	30 (6.4)	23 (5.7)
Artist/entertainer/sports	62 (7.1)	39 (8.4)	23 (5.7)
Teacher/professor	72 (8.3)	39 (8.4)	33 (8.2)
Trader/hawker	89 (10.2)	47 (10.1)	42 (10.4)
Professional	70 (8.0)	37 (7.9)	33 (8.2)
Laborer	86 (9.9)	43 (9.2)	43 (10.7)
Civil servant	68 (7.8)	36 (7.7)	32 (7.9)
Journalism/media	55 (6.3)	31 (6.6)	24 (6.0)
Unemployed	87 (10.0)	46 (9.9)	41 (10.2)
Student	67 (7.7)	25 (5.4)	42 (10.4)
Housewife	70 (8.0)	43 (9.2)	27 (6.7)
Law enforcement	39 (4.5)	22 (4.7)	17 (4.2)
Total	N = 870	467	403

Among all respondents in both countries, privately-owned radio is most frequently given as the main source of news on current affairs (28%), followed by state-owned radio (22%), state-owned television (15%), and privately-owned television (14%). In Rwanda, state-owned radio is the leading source of news (31%), followed by privately-owned radio (21%) and privately-owned television (6%) and the internet (9%). In Ethiopia, privately-owned radio (34%) is followed by privately-owned television (20%), state-owned radio (14%) and privately-owned newspapers (13%) (see Table 2). Overall, respondents in blue-collar occupations (N = 451) were more likely to get news from privately-owned media, and those in white-collar occupations (N = 352) were more likely to get news from state-owned media.

Table 2.

Main sources of news on current affairs, by country (%).

	Both countries	Ethiopia	Rwanda
Private radio	246 (28.3)	160 (34.3)	86 (21.3)
State radio	190 (21.8)	67 (14.3)	123 (30.5)
Private television	119 (13.7)	95 (20.3)	24 (6.0)
State television	130 (14.9)	44 (9.4)	86 (21.3)
Private newspaper	91 (10.5)	60 (12.8)	31 (7.7)
State newspaper	30 (3.4)	15 (3.2)	15 (3.7)
Friends/family	3 (0.3)	2 (0.4)	1 (0.2)
Internet	61 (7.0)	24 (5.1)	37 (9.2)

Important distinctions emerge when respondents describe their trust in individual media. It is no surprise that the state-run media are less trusted overall, but not all distrust is equal. In both countries, trust in state radio and newspapers is linear, with about two-thirds of respondents at the lowest two points on a 7-point scale and about 80 percent in the overall ‘distrust’ category. State television, on the other hand, finds about 50 percent in the lowest two categories but nearly a fourth of respondents (23.8% overall, 23.6% in Ethiopia, 24.1% in Rwanda) reporting that they ‘somewhat trust’ that medium.

There are no statistically significant differences between countries in levels of trust in the three state-run media, but trust in all three private media is significantly higher in Ethiopia than in Rwanda (see Table 3).

Table 3.

Trust in private and state media (1–7, lowest–highest) (M).

	Ethiopia	Rwanda	t (df = 868)	sig.
Private radio	4.67	4.35	2.768	.006
State radio	2.49	2.46	.151	.763
Private television	4.87	4.43	3.520	.001
State television	2.88	2.93	–.464	.643
Private newspaper	4.22	4.00	1.874	.061
State newspaper	2.39	2.37	.151	.880

Media incitement and reliability

The first research question asks about respondents’ attitudes toward the media’s role in ethnic conflict (see Table 4). Counting both countries, respondents in general are more likely to agree than to disagree that the media incite ethnic conflicts (M = 4.60) and that the media are divisive and dangerous to national security (M = 4.47). Those who rely on state-owned media for information about public affairs are significantly more likely than those who rely on privately-owned media to agree that the media are divisive (χ² (df = 4) 13.42, p = .01). Interestingly, the Rwandese and Ethiopians who rely on neither are about evenly divided, and nearly all these respondents pointed to the internet as their main source of information. Rwandese (M = 4.97) were significantly more likely to agree than Ethiopians (M = 4.04) that the media are divisive.

Table 4.

Media performance (M).

	All respondents	Ethiopia	Rwanda
Do media incite conflicts?	4.60	4.33	4.91***
Are media divisive/dangerous?	4.47	4.04	4.97***
Regime truthful on war?	3.30	3.28	3.31
State media incomplete on war?	5.32	5.20	5.46*
Trust private media on war?	4.01	4.05	3.96

Notes: For mean differences between countries: * = p < .05; *** = p < .001

Scale: (1 = low, 7 = high)

Similarly, respondents in both countries were more likely than not (56%) to agree that the media incite ethnic conflicts. Rwandese respondents (M = 4.91) were significantly more likely to agree than Ethiopians (M = 4.33). Those who rely on state-owned media for information are slightly more likely to agree than those who rely on privately-owned media that the media incite ethnic conflicts, but the difference is not statistically significant (p = .192). Again, those who rely on the internet are more evenly divided.

The second research question asks about which media are trusted by different groups. In both countries, respondents were more likely to say privately-owned media provide the most reliable news about conflicts and crises. This response was much higher in Ethiopia (72.6%) than in Rwanda (57.6%). In Rwanda, about a quarter of respondents (25.8%) said state-owned media provided the most reliable conflict news (compared with 15% in Ethiopia), and 16.6 percent said the internet provided the most reliable conflict news (compared with 12.4% in Ethiopia) (see Table 5). There were no significant differences by occupational category.

Table 5.

Most accurate news about conflicts and crises (%).

	Private	State	Internet
Ethiopia	339 (73)	70 (15)	8 (12)
Rwanda	232 (58)	104 (26)	67 (17)
White-collar	225 (63.9)	76 (21.6)	51 (14.5)
Blue-collar	302 (67)	85 (18.8)	64 (14.2)
All	571 (66)	174 (20)	125 (14)

Notes: The difference between countries is significant: χ ²(df = 2) 22.76, p < .001.

All categories of media users deemed the privately-owned media to be the most reliable source of conflict news. Those who get most of their news from privately-owned media are the least likely to consider state-owned media their best source of conflict news. Those who get their news from the internet are most likely to trust conflict news from state-owned media – nearly 30 percent give this response – but the overall differences are not significant.

Respondents in both countries are about neutral on whether they trust what they read in privately-owned media about conflicts (overall, M = 4.01; Ethiopia, 4.05; Rwanda, 3.96). They slightly disagree that the regime can be counted on to tell the truth during a conflict (overall M = 3.30; Ethiopia, 3.28, Rwanda, 3.31). Respondents agree that state-owned media are unlikely to tell the whole story during a conflict (overall M = 5.32), and that distrust is significantly higher in Rwanda (M = 5.46) than in Ethiopia (M = 5.20).

Those in blue-collar occupations are slightly more likely to either agree or disagree that media incite ethnic conflicts than those in white-collar occupations, but that difference is not statistically significant (p = .094). There were no notable differences by occupational category on whether the media are divisive. Blue-collar workers (58.8%) are significantly more likely than white-collar workers (45.5%) to rely on private media for news and slightly more likely to rely in the internet (see Table 6).

Table 6.

Main source of news by occupational category (%).

	Both		Ethiopia		Rwanda
	White-collar	Blue-collar	White-collar	Blue-collar	White-collar	Blue-collar
Private	160 (45.5)	265 (58.8)	114 (60.3)	185 (73.1)	46 (28.2)	80 (40)
State	173 (49.1)	151 (33.5)	68 (36)	51 (20.2)	105 (64.4)	100 (51)
Internet	19 (5.4)	35 (7.8)	7 (3.7)	17 (6.7)	12 (7.4)	18 (9)
	χ ² (df = 2) 20.28, p < .001		χ ² (df = 2) 14.49, p = .001		χ ² (df = 2) 7.17, p = .028

Those who rely on state-run media for information about current affairs are the most likely to agree that the media incite ethnic conflicts (M = 4.75) and are divisive and dangerous to national security (M = 4.74). Those who rely on privately-owned media (4.52) or the internet (M = 4.28) are less likely to say the media incite ethnic conflicts, and the difference approaches significance (p = .082). Users of privately-owned media (M = 4.33) or the internet (M = 4.08) are significantly less likely to agree that the media are divisive (p = .005). There were no significant differences by media use in either country on whether the media were divisive or prone to inciting conflict.

Discussion

The central objective of this study was to compare how citizens of Rwanda and Ethiopia describe the role of mass media in covering ethnic conflicts and to determine public attitudes towards media performance on matters of trust, reliability, and incitement. The findings indicate some encouraging and some less encouraging news for both the privately-owned media and the state-owned media in Rwanda and Ethiopia.

First, this study supports the conjecture that the electronic media, especially radio, continue to be a major force for social and political mobilization in the two countries. They are identified as the main source of news on current affairs, and their broad appeal could help spread information and education that prevents or mitigates conflicts. But this potential is moderated by negative evaluations of both privately-owned and state-owned media. In both countries, the media are still perceived as likely to spread the sort of hate and incitement that have fueled ethnic conflict in the past. This response is stronger in Rwanda, perhaps reflecting the scale of the genocide there, but it is visible in both. Respondents in Rwanda are significantly more likely to see the media as dangerous to national security, whereas Ethiopians were about neutral on that issue.

State-owned media in general are less trusted, and neither the regimes nor the media they own are much trusted to provide complete, truthful coverage in conflicts. The media in general, it seems, are still perceived as perpetrators of political factionalism fueled by messages of ethno-sectarianism. Previous research has described this tendency as the cause of public despair, which has given way to animosity toward the media in general (Fessehatzion, 2003; Takirambudde, 1995). Yet evidently, not all media are equal, as shown by the previously mentioned patterns of trust toward state-owned television. The evidence here suggests that media channels perceived as more modern, or at least less connected with the past, are better received.

The data from Rwanda present an interesting counter-trend as well: a significantly higher public attitude that state-owned media will be the most accurate in a crisis. This could reflect Rwanda’s efforts since the genocide to use the media as a force for reconciliation. If that is the case, post-transition governments, as well as international agencies, regional security communities, and donor groups, have reason to be encouraged. Media that were used to stoke a crisis can apparently be used to dampen it as well.

It is also worth noting that the now-infamous Radio Mille-Collines (RTLM) began as a privately-owned station and was later taken over by allies of the regime. Despite the benefits usually associated with decentralizing or privatizing the media, not every media proprietor will be a responsible one. Rwandese’s relatively lower expectations for the privately-owned media could reflect this recent memory, and that provides a plausible explanation for these results.

The Rwandese use the internet more and trust it more as a source of conflict news. That development also suggests a continued role for international media in providing comprehensive, professional news as a potential alternative to the inflammatory product that can arise within a state’s borders. Access to information can be as important a part of press freedom as access to the press itself.

There seems to be little difference by occupational category in attitudes toward media, although those in white-collar occupations seem slightly more inclined to see the media in the middle ground on divisiveness and incitement. The difference between the countries when occupational categories are compared is striking. In Ethiopia, both sectors strongly prefer the private media; in Rwanda, both categories prefer the state media, albeit white-collar workers more so than blue-collar workers. This could suggest that under Rwanda’s particular conditions, those with presumably better access to education and media are more amenable to state-owned media’s efforts to present a picture of reconciliation. Conversely, those at the other end of the social spectrum appear to be more suspicious and more active in selecting ‘none of the above’ between state-owned and privately-owned media.

Greater overall trust in the privately-owned media probably reflects the dedication of journalists in both countries. Independent journalists who work with privately-owned media have fought hard for their rights through a combination of defiance, legal measures, and parliamentary lobbying (Lush, 1998). It is hardly a surprise that most respondents were more likely to say the privately-owned media provide the most reliable news about conflicts and crises. However, there is still the absence of favorable media policy, law, and regulation, which undermine media freedom and the chances for the privately-owned media to function professionally. Consequently, as Amogne (2005) pointed out, more independent local journalists have been charged with engaging in seditious or treasonable acts for reporting stories that highlight the plight or human suffering of the opposite side.

The history of media and politics in Eastern Africa has not been erased: state-owned broadcasters have been used by government to manipulate the public by stirring ethno-political divisions and by selectively labeling state opponents as terrorists and opportunists (Ocitti, 1999; Rubongoya, 2007), and state radio bears the brunt of public disdain toward those media. Still, the results presented here suggest two theoretically based avenues of interest. From a perspective rooted in news routines, a change in how the ‘enemy’ looks could arise from a deliberate effort by journalists themselves (Lee and Maslog, 2005; Obonyo, 2010) or from a change in policy that is then reflected in what ‘news’ looks like. In Wolfsfeld et al.’s (2008) ‘politics–media–politics’ model, journalists ‘not only reflect political realities but they also actively translate them into news stories that are both interesting and culturally resonant’ (p. 377). Thus, a change in patterns of media use or trust like the ones suggested by these results could come about as a result of business as usual. From a more normative perspective, one branch of peace journalism theory (Keeble, 2010) rejects the idea of journalistic professionalism in favor of explicit acknowledgement of journalism as ‘political practice’. In this case, those ‘political realities’ would be acknowledged as a part of the media system, not kept at arm’s length until they could be translated. Professional suspicion of such a contrarian idea could be allayed by Keeble’s comparison to the ‘citizen journalism’ practiced in mainstream US news outlets in the 1990s.

Conclusion

Responses to this survey almost certainly reflect an array of current and post hoc impressions of media, making it more difficult to suggest confidently what causes and effects are associated. Still, this study provides a starting point for governments, educators, and media stakeholders to better understand what type of media the Rwandese and Ethiopians prefer, their media use, and their attitudes toward media conflict incitement, reliability, and trust. Future research should test some of these conclusions more directly, perhaps complementing a survey like this one with the qualitative texture offered by depth interviews or focus group studies. Up-to-date analyses of media content would also provide important benchmarks.

The state-owned news media in Rwanda and Ethiopia have been susceptible to censorship and direct political control. Their reliance on official sources can present a daunting challenge to an unsophisticated audience that is often inadequately informed to critically evaluate what is right or not right – a concern that is assumed to make it easier for broadcast messages to generate so much violence and ethnic cleansing in the past. Results from Rwanda, though, suggest that citizens are being cautious gatekeepers – a hopeful argument against the idea that the broadcast media can be wielded at will by autocrats.

These findings support the idea that if the Rwandan and Ethiopian media have previously succeeded in fueling conflicts yet remain widely used among citizens, the same media can be used to prevent or mitigate conflicts. Those media might have to demonstrate the sort of efforts toward harmony suggested in Rwanda in recent years to regain lost trust.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

Author biographies

Yusuf Kalyango Jr. is Director of the Institute for International Journalism and Assistant Professor in the E.W. Scripps School of Journalism at Ohio University. He teaches international journalism, media and conflicts, mass communication research, and electronic media, and is the author of African Media and Democratization: Public Opinion, Ownership, and Rule of Law (Peter Lang Publishing, 2011).

Fred Vultee is Assistant Professor of Journalism in the Department of Communication at Wayne State University. He teaches news editing, political communication, research methods, and media and conflicts. His research focuses on media performance in conflicts and crises. Address: Department of Communication, Wayne State University, 567 Manoogian, Detroit, MI 48201, USA. [email: vulteef@wayne.edu]

References

Abbink

(1994) Refractions of revolution in Ethiopian ‘Surmic’ societies: An analysis of cultural response. In: Marcus

(ed.) New trends in Ethiopian studies. Lawrenceville, NJ: Red Sea Press, 734–755.

Aginam

(2005) Media in ‘globalizing’ Africa: What prospect for democratic communication? In Hackett

Zhao

(eds) Democratizing Global Media: One World, Many Struggles. Lanham, MD: Rowman & Littlefield.

Allen

Seaton

(1999) The Media of Conflict: War Reporting and Representations of Ethnic Violence. London: Zed Books.

Amogne

(2005) Public access to government information in Ethiopia: An assessment of public information practice in selected government institutions. Unpublished MA thesis, Addis Ababa University, Addis Ababa.

Chalk

(1999) Hate radio in Rwanda. In: Adelman

Suhrke

(eds) The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire. New Brunswick, NJ: Transaction.

Choi

James

(2007) Media openness, democracy, and militarized interstate disputes. British Journal of Political Science 37(1): 23–46.

Choi

James

(2008) Civil-military structure, political communication, and the democratic peace. Journal of Peace Research 45(1): 37–53.

Davidson

(1998) The price of African press freedom. In: Dennis

Snyder

(eds) Media and Democracy. New Brunswick, NJ: Transaction, 51–58.

Des Forges

(2002) Silencing the voices of hate radio in Rwanda. In: Price

Thompson

(eds) Forging Peace: Intervention, Human Rights, and the Management of Media Space. Edinburgh: Edinburgh University Press, 236–258.

10.

Fardon

Furniss

(2000) African Broadcast Cultures: Radio in Transition. Oxford: James Currey Publishers.

11.

Fawcett

(2002) Why peace journalism isn’t news. Journalism Studies 3(2): 213–223.

12.

Fessehatzion

(2003) Shattered Illusion, Broken Promise: Essays on the Eritrea–Ethiopia Conflict (1998–2000). Lawrenceville, NJ: Red Sea Press.

13.

Franklin

Love

(1998) Whose news? Control of the media in Africa. Review of African political economy 78(25): 545–590.

14.

Frohardt

Temin

(2003) Use and Abuse of Media in Vulnerable Societies (special report 110). United States Institute of Peace, Washington, DC. Available at: www.usip.org/pubs/specialreports/sr110.html (accessed 2 March 2009).

15.

Gatwa

(1995) Ethnic conflict and the media: The case of Rwanda. Media development 42(3): 18–20.

16.

Gourevitch

(2000) We Wish to Inform You that Tomorrow We Will Be Killed with Our Families. New York: Picador.

17.

Hachten

(2004) Reporting Africa’s problems. In: Okigbo

Eribo

(eds) Development and Communication in Africa. Lanham, MD: Rowman & Littlefield, 79–87.

18.

Hanitzsch

(2004) Journalists as peacekeeping force? Peace journalism and mass communication theory. Journalism Studies 5(4): 183–195.

19.

Herbst

(2004) African militaries and rebellion: The political economy of threat and combat effectiveness. Journal of Peace Research 41(3): 357–393.

20.

Hyden

Lanegran

(2005) Aids, policy and politics: East Africa in comparative perspective. Review of Policy Research 12(2): 47–65.

21.

IFJ (2007) Status of Journalists Report, 2007. In: Bagblo

Thomasi

(eds) International Federation of Journalists Publications. Dakar: Africa Regional Office.

22.

Jacquin-Berdal

Plaut

(2005) Unfinished Business: Ethiopia and Eritrea at War. Lawrenceville, NJ: Red Sea Press.

23.

Kalyango

Jr. Eckler

(2010) Media performance, agenda building, and democratization in East Africa. In: Communication Yearbook 34: 355–389.

24.

Karnell

(2002) Counteracting hate radio in Africa’s Great Lakes region: responses and lessons. Journal of International Communications 8(1): 111–137.

25.

Kaufman

(2006) Escaping the symbolic politics trap: Reconciliation initiatives and conflict resolution in ethnic wars. Journal of Conflict Resolution 43(2): 201–218.

26.

Keeble

(2010) Peace journalism as political practice: A new, radical look at the theory. In: Keeble

. (eds) Peace Journalism, War and Conflict Resolution. New York: Peter Lang, 49–68.

27.

Kurspahic

(2003) Prime-Time Crime: Balkan Media in War and Peace. Washington, DC: United States Institute of Peace.

28.

Lee

Maslog

(2005) War or peace journalism? Asian newspaper coverage of conflicts. Journal of Communication 55(2): 311–329.

29.

Lush

(1998) The role of the African media in the promotion of democracy and human rights. In: Kayizzi-Mugerwa

. (eds) Towards a Partnership with Africa: Challenges and Opportunities. Uppsala: Nordic African Institute, 42–65.

30.

Lyons

(2007) Avoiding Conflict in the Horn of Africa: Ethiopia and Eritrea. London: Council on Foreign Relations.

31.

Mamdani

(1996) Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism. Princeton, NJ: Princeton University Press.

32.

Mazrui

(1980) Africa Condition: A Political Diagnosis. Cambridge, MA: Cambridge University Press.

33.

Melander

(2009) The geography of fear: Regional ethnic diversity, the security dilemma and ethnic war. European Journal of International Relations 15(1): 95–124.

34.

Norris

(2008) Silencing dissent: The impact of restrictive media environments on regime support. Paper presented at the annual meeting of the American Political Science Association, Boston, MA.

35.

Obonyo

(2010) Peace journalism training in East Africa. Media Development 57(3): 59–63.

36.

Ocitti

(1999) Media and Democracy in Africa. Cambridge, MA: Weatherhead Center for International Affairs.

37.

Olorunnisola

(1995) When tribal wars are mass mediated: Re-evaluating the policy of ‘non interference’. Gazette 56(2): 123–138.

38.

Paris

(2004) At War’s End: Building Peace after Civil Conflict. Cambridge: Cambridge University Press.

39.

Prunier

(1995) The Rwanda Crisis: History of a Genocide. New York: Columbia University Press.

40.

Reporters without Borders (2007) Reporters without Borders Annual Report on Press Freedom – Ethiopia. Division of External Relations, Geneva: RSF-Refworld Publication, 1–11.

41.

Reyntjens

(1999) A dubious discourse on Rwanda. African Affairs 98(1): 119–141.

42.

Reyntjens

(2004) Rwanda, ten years on: From genocide to dictatorship. African Affairs 103(441): 177–191.

43.

Ross

Tehranian

(eds) (2009) Peace journalism in times of war. New Brunswick, NJ: Transaction.

44.

Rotberg

Weiss

(1996) From Massacres to Genocide. Cambridge. MA: World Peace Foundation.

45.

Rubongoya

(2007) Regime Hegemony in Museveni’s Uganda: Pax Musevenica. New York: Palgrave Macmillan.

46.

Schabas

(2000). Hate speech in Rwanda: The road to genocide. McGill Law Journal 46(144): 141–171.

47.

Simon

(2006) Media coverage of the 2005 Ethiopian elections. Presented at a workshop on Media and Good Governance, Addis Ababa University and UNESCO.

48.

Snyder

Ballentine

(1996) Nationalism and the marketplace of ideas. International Security 21(2): 5–40.

49.

Sonwalkar

(2004) Out of sight, out of mind: The non-reporting of small wars and insurgencies. In: Allan

Zelizer

(eds) Reporting War: Journalism in Wartime. New York: Routledge, 206–223.

50.

Takirambudde

(1995). Protection of labour rights in the age of democratization and economic restructuring in Southern Africa. Journal of African Law 39: 39–63.

51.

Uvin

(1998) Aiding Violence: The Development Enterprise in Rwanda. West Hartford, CT: Kumarian Press.

52.

Van Belle

(1997) Press freedom and the democratic peace. Journal of Peace Research 34(4): 405–414.

53.

Van Belle

(2000) Press Freedom and Global Politics. Westport, CT: Praeger.

54.

Van der Veur

(2002) Broadcasting and political reform. In Hyden

. (eds) Media and Democracy in Africa. New Brunswick, NJ: Transaction, 81–105.

55.

Wolfsfeld

(2001) The News Media and Peace Processes. Washington, DC: United States Institute of Peace.

56.

Wolfsfeld

Alimi

Kailani

(2008) News media and peace building in asymmetrical conflicts: The flow of news between Jordan and Israel. Political Studies 56(2): 374–398.