Terje Skjerdal and Mulatu Alemayehu Moges

The ethnifi cation of the Ethiopian media Th e ethnicifi cation of the Ethiopian media

A research report

By Terje Skjerdal and Mulatu Alemayehu Moges

Addis Ababa, November 2020

Th is report was commissioned by Fojo Media Institute and International Media Support.

Th e authors are responsible for the content.

Th ey can be contacted at [email protected] and [email protected]. Contents

Acronyms ...... 4 Executive summary ...... 5 Research approach and methodology ...... 6

Terje Skjerdal: Media and ethnicity in in historical perspective ...... 9

Terje Skjerdal: Media and ethnicity during the current Ethiopian leadership ...... 14

Mulatu Alemayehu: A framing analysis of stories in selected media channels ...... 35

References ...... 56

Figures and tables

List of informants for qualitative research interviews ...... 15 Figure 1: Occurrences of infl ammatory terms on Twitter...... 17 List of researched cases for framing analysis ...... 35 List of researched media channels ...... 36

3 Acronyms

AAPO All-’s Organization APD Amhara Democratic Party AJA Amhara Journalist Association AMMA Amhara Mass Media Agency (cf. ATV) ASRAT Amhara Satellite Radio and Television ATV Amhara Television (belongs to AMMA) CARD Center for Advancement of Rights and Democracy DW Dimtsi Weyane Tigray EBA Ethiopian Broadcasting Authority EBC Ethiopian Broadcasting Corporation (belongs to ERTA) EFFORT Endowment Fund for the Rehabilitation of Tigray EJA Ethiopian Journalist Associaction ENA Ethiopian EPLF Eritrean People’s Liberation Front EPRDF Ethiopian People’s Revolutionary Democratic Front ERTA Ethiopian Radio and Television Agency (cf. EBC) ESAT Ethiopian Satellite Television and Radio ETV Ethiopian Television (now: EBC) FBC Fana Broadcasting Corporate NGO Non-governmental organization OBN Broadcasting Network (belongs to ORTO) OBS Oromia Broadcasting Service OFC Oromo Federalist Congress OJA Oromia Journalist Association OLF OMN ORTO Oromia Radio and Television Organization (cf. OBN) PP SNNPR Southern Nations, Nationalities, and Peoples’ Region TJA Tigray Journalist Association TMMA Tigray Mass Media Agency (cf. TTV) TPLF Tigray People’s Liberation Front TTV Tigray Television (belongs to TMMA) WJS Worlds of Journalism Study

4 Executive summary

Th is report looks into the issue of media and frames of reference in the current Ethiopi- ethnicity in contemporary Ethiopia. As any- an media discourse. Ethnicity is being em- one can witness, there have been immense ployed as a key to interpret media messages, changes in the local media situation sin- and it outlines a frame for classifying media ce Dr. assumed leadership in channels and personalities. We identify two April 2018. In just two years, the country has major frames, annihilation and othering, improved its ranking on Reporter Without which are indicative of the tendency of jour- Borders’ Press Freedom Index by 51 places, nalists to align with their ethnic background. which is the greatest leap of any country We also fi nd that many media practitioners anytime – by far. Th ere is a lot to celebrate. are deeply concerned about the develop- At the same time, the press situation in the ments. Research data point towards a more country is deeply disconcerting. Th e reason pluralistic media society in Ethiopia (albeit for this is that the media landscape is as po- segmented pluralism), but also signifi cantly larized as never before, and that parts of the more polarization. Th e polarization is fuel- media are being blamed for playing a pivotal led by ethno-nationalistic media of diff erent role in the turbulence which have erupted in origin and ownership, vis-à-vis pan-Ethio- various parts of the country since 2019. Th e pianist channels. Th e previous stronghold of suspicion of many people is that the media the federal state media (EPRDF-supportive) are being exploited to propagate an ethnic has been overtaken by a fragmented state agenda. media structure with growing signifi cance Towards this backdrop, the intention of of regional mass media agencies. Journalists the study is to examine the role of the media are beginning to form alliances along regi- in ethnic tensions in contemporary Ethio- onal and ethnic fault lines, as illustrated by pia. Th e methodology is twofold, combining the launch of Amhara Journalists Associati- in-depth interviews with framing analysis. on, Oromia Journalists Association, Tigray For the interview part, we have met appro- Journalists Association, and so forth. Jour- ximately 25 persons with key positions in all nalists are highly inclined to use sources major parts of the media in the country. For which support their own ideological inte- the framing analysis, we have scrutinized rest, and avoid quoting sources from other the media coverage of eight crucial incidents ethnicities which could balance the story. between 2018 and 2020 in ten diff erent me- When it comes to professional ideals, vario- dia outlets. us channels practice a hybrid reporting style We argue that the tendencies we’re ob- which blends acknowledged standards with serving in the media can be described as an interventionist or activist agenda. We ar- an intensifi ed ethnifi cation process. By this gue that ethnic and political agendas are at we mean that ethnic belonging and identity risk of being at odds with media professio- politics are gaining signifi cance as central nalism.

5 Research approach and methodology

Th is research study consists of two main parts: an analysis of the media environment and an analysis of media content. Both parts make use of qualitative rese- arch methodologies.

Th e analysis of the media environment is part of the study are visited in the framing based on observations plus interviews with analysis, plus a few print publications. Th e key actors in the local media. Approximate- methodological approach for the framing ly 25 in-depth interviews were conducted. analysis is explained more in detail in the 20 of the informants agreed to be named in fi rst part of that analysis. this report, and are listed on page 15. A few Part one of the study is conducted by Ter- more informants from diff erent parts of the je Skjerdal, while part two is conducted by media sector (federal, regional and private) Mulatu Alemayehu. However, the two parts were interviewed on condition of anony- should be seen in conjunction. Together, mity. Th ose who appear with name are iden- they intend to give insights into the increa- tifi ed in the discussion only in cases where singly important role of ethnicity in the Et- it has been explicitly agreed that they will be hiopian media confi guration and in the co- quoted. For the most part, the information untry’s media debate. in the report is not attributed to a specifi c Th e report also includes an introductory person. Th is is to secure openness in the in- discussion of ‘Media and ethnicity in Et- terview situation and to underline that the hiopia in historical perceptive’ which shows interpretation of the research data is entirely how some of the current tendencies have the responsibility of the researchers. deeper roots in the history of the media in Th e interviews were mainly conducted in the country (p. 9). Th e review is primarily March 2020 (in Ethiopia), with follow-up based on known literature, but it also con- interviews until November 2020 (online). tains extracts of research fi ndings of ethni- Several of the informants were arrested in city in Ethiopian newsrooms which have not July 2020 following the unrest aft er the as- been published before. sassination of , and are Th e study has needed to make various still detained at the time of completion of choices regarding which terminology to use, this report (November 2020). Th e intervi- for example in relation to the term ‘ethnic ews with these informants were conducted media’, which is oft en seen in the current de- prior to the recent turbulence. bate. Th us, the following section will contain Th e second part of the research consists of a discussion of key terms. analysis of selected media stories published between 2018 and 2020. Th e methodology The term ‘ethnic media’ for this part is framing analysis. Altogeth- Th e term ‘ethnic media’ is oft en used to re- er eight stories have been studied, each of fer to the newly established media channels which illustrates a particular perspective on with ethnic affi liation (such as Asrat TV and ethnicity in the media debate. All the major OMN), but the phrase can be misleading 6 broadcasting networks discussed in the fi rst and might give the wrong connotations. Th e Ethiopian Broad- In the global research literature, ‘ethnic as ethnic media. ESAT, for example, which casting Corporation media’ is generally used to refer to a diff e- was established as a diaspora channel with (EBC) is offi cially a rent type of media than some of the most studios in Amsterdam, and Was- public broadcaster, vocal and popular outlets on Ethiopia’s me- hington DC, aspired to oppose the Ethiopi- but in global research dia scene. ‘Ethnic media’ is a category which an government’s offi cial narratives through terms the company denotes media outlets targeting immigrant satellite broadcasts reaching local audiences would more appropria- communities in societies where such groups back in Ethiopia and had global distribution. tely be defi ned as a stand out as a minority vis-à-vis the rest of Th e station has never had the limited geo- state broadcaster. the population (Deuze, 2006; Matsaganis et graphical outreach which is typically associ- (Photo by Terje al., 2011). Th e term is mostly used to des- ated with ethnic media. Skjerdal) cribe immigrant media cultures in the US On this basis we prefer not to use ‘ethnic and Western Europe, and has given birth to media’ to point to the current media chan- a distinct reporting style, ‘ethnic journalism’ nels operating in Ethiopia, even if they are (Yu, 2019). It is a rather blameless activity clearly propagating ethnic interests. Prefer- which comes out of the need to share in- red terms are descriptive phrases such as formation and maintain cultural cohesion ‘ethnically-affi liated media’, “ethnically-ba- among people who have settled in a foreign sed media’, and so forth. country. Salon Ethiopia, a bi-weekly news- paper published in Amharic for the Ethiopi- ’State media’ vs. ’public media’ an community in Seattle, would thus qualify Further on terminology, we use ‘state media’ as ‘ethnic media’. instead of ‘public media’, even if the latter is Diaspora media aimed at audiences back the correct term to describe the offi cial Et- home, on the other hand, are of a diff e- hiopian media according to existing media rent category and would not be regarded legislation1. Th is is because in the general

7 1 BroadcasƟ ng Service ProclamaƟ on No. 533/2007 research literature, ‘public media’ points The term ’mainstream media’ to channels which are not associated with Lastly, it is necessary to address the term particular political interests, while there is ‘mainstream media’, which seems to be in- no doubt that the offi cial media channels in creasingly used in Ethiopian media analysis. Ethiopia, both on federal and regional level, Th e term is used diff erently by diff erent pe- have a special mandate to speak the interests ople and in diff erent contexts. In the global of the government. Even though offi cially literature, for the most part, the name tends established companies such as Ethiopian to refer to widely distributed and common Broadcasting Corporation (EBC) are less of media outlets representing middle-of-the- a mouthpiece for the authorities than they road content, as opposed to alternative me- used to be, they are still operating as an ex- dia which present a diff erent perspective by tension of the public administration and are challenging the majority view. However, on led by persons who are government par- other occasions ‘mainstream media’ is em- ty-affi liated. Th e CEO of EBC is accountable ployed rather as a generic term to signify a to the Parliament. media platform other than online media (i.e. A similar loyalty to the authorities are fo- newspapers, radio and television). Th e latter und in companies such as Fana Broacasting usage is similar to ‘legacy media’, which is Corporate and Walta Media and Commu- yet another increasingly appearing phrase nication Corporate, which both continue (cf. ‘traditional media’). to back the communication strategy of the ‘Mainstream media’ can be confusing ter- central government and the Prosperity Par- minology because it is sometimes unclear ty (PP). At the same time, these companies whether it relates to content or technology, are not formally governed by the Parliament but even more importantly in this context, or the cabinet through organizational stru- if it does denote a certain style of content, ctures, thereby earning a degree of indepen- which is the most common understanding dence, although perhaps in name only. Th ey of the term globally, it would be debatable are therefore referred to as private channels, which type of media represents the avera- formally. ge consumption in Ethiopia. Is it moderate Th e Ethiopian legislation indeed does not outlets such as Reporter and Sheger FM? operate with the term ‘private’, but refers to Large channels such as EBC and Fana, alt- ‘commercial’ media (besides ‘public’ and hough they are state-affi liated? Because of ‘community’2). We prefer to speak of ‘private this, and due to the transitional situation media’ in this report, so as to emphasize that which the country is in, we avoid the use of they are independent from the authorities ‘mainstream media’ in this study, unless it is in comparison with the state media. A term clear from the context what it refers to. We which could be considered for future usage, generally prefer to label diff erent media gen- however, is ‘non-state media’. res by descriptive terms.

2 BroadcasƟ ng Service ProclamaƟ on No. 533/2007

8 Media and ethnicity in Ethiopia in historical perspective

By Terje Skjerdal

When reviewing the current developments in the Ethiopian society, one could easily get the impression that ethnicity has become the new mainspring of the media. Th e matter of the fact is, however, that ethnicity has always been an undercurrent in Ethiopian media and journalism.

Th rough out the diff erent regimes, ethnic 1991, p. 49; Smith, 2008, p. 219; MoI, 1969, interests have played an important role as a p. 140). Notably, however, Oromiff a was motivator for daily reporting, and the media banned from the offi cial media for most of have been used by the authorities to propa- the imperial period (Fellman, 1992; Me- gate political as well as ethnic causes. Th is is kuria, 1997). only to be expected in a society where, ever Indeed, the fi rst time that the Oromos since the fi rst print publications came on the could listen to radio broadcasts in their own market in the early 1900s, media have been language was during the Italian occupation used for national and strategic purposes. 1936–41, but the decision by the Fascists to During the Ethiopian Empire, Haile Se- use Oromo on Ethiopian Radio was rever- lassie treated the media as a mouthpiece for sed once the Emperor returned from exile3. Ethiopian unity based on Amhara supre- In 1962, the ban on Oromo was challenged macy. Th is was refl ected in for example the when Radio Mogadishu started broadcas- language policy of the media, which meant ting in Oromiff a from Somalia, targeting that Amharic became widely dominant both the Oromo population in Ethiopia. Th e in the print and broadcasting media. Fore- programme, called Raadiyoo Afaan Qottu, ign languages, however, were never a thre- became immensely popular, much to the at to the empire and have a long history in distress of . It has been clai- various print publications (such as English, med that the Emperor was so furious with French and Swahili). the broadcasts that he arranged the assas- By 1970, Ethiopian Radio served seve- sination of the programme founder Ayub ral local groups by off ering programmes Abubakar in Mogadishu in 1965 (Gow, in Afar, Somali, Tigre and Tigrigna (Janas, 2004, p. 310).

3 ‘Events in Oromo history during the reign of Haile Selassie I’. Available at: hƩ p://www.gadaa.com/haile- 9 selassie.html Dimtse Woyane Tigray began as a resistance radio broadcasting from caves in the northern part of the country. (Photo by Terje Skjerdal)

Th is only led to further resistance from the fort to build the Tigray resistance movement Oromo population, and fi nally the regime (i.e. Tigray People’s Liberation Front, TPLF) gave in and allowed Oromo programming (Amanuel G, 2019; Danial, 2019). Th e on a regional station, Radio, in 1972. had equal concerns for Oromo separatism. Immediately aft er the fall of the Empire, Sustained discrimination the Oromo cultural resistance movement during the Derg had great hopes for their people when Oro- Th e Derg regime (1974–91) continued to miff a was allowed on state radio, although view the media as a potential battleground it lasted only for a couple of years (Kuwee, for ethnic confl ict. A one-language policy 1998). In 1976, Bariisaa, the Oromo news- was enforced in the state media in favour of paper which still comes out on a weekly ba- Amharic (Menychle, 2020). Th e use of other sis, was launched by the Oromo Liberation languages was seen as a vehicle for regional Front (OLF). Th e publication was nationali- separatism, and Tigrigna in particular faded zed by the Derg the year aft er. Illustratively, from the offi cial media scene. whereas Bariisaa was initially printed with From remote caves in the , Latin script, the Derg forced it to use Am- however, clandestine radio broadcasts emer- haric fi dels, as writing Oromo with Latin let- 10 ged, becoming a major instrument in the ef- ters was forbidden (Gow, 2004, p. 208; Ke- bede, 2003, p. 479). Th e newspaper reverted Oromo journalists at the time expressed that to Latin script only aft er EPRDF’s liberation the reorganizing was speculative and resul- in 1991. ted in a weakening of Oromo interests in the state media (Skjerdal, 2013a, p. 198). Tigray The coming of EPRDF and intro- interests, on the other hand, continued since duction of regional programmes many leadership positions were still fi lled by Post-1991, EPRDF’s ethnic federalism phi- TPLF affi liates. losophy was soon integrated in its media Among reporters, it was never doubted policy. Diff erent regions and ethnic groups that Tigrayans were in power. Th is aff ected began to be acknowledged by the media, the everyday interaction in the newsroom. If and over time several languages got their a Tigrayan was passing by, it would eff ective- own slots on the national broadcaster. Oro- ly put an end to any discussion since it was mo programmes commenced on Ethiopian as if a government representative entered the Television (ETV) on 23 June 1992. room. Th e introduction of programmes aimed at Th e ethnic radar was actively in use by a certain ethnic group was not without con- any employee in the organization. It implied troversy. Several ETV veterans were skepti- that if a newcomer arrived in the newsroom, cal. “What do television broadcasts do for reporters and editors would, consciously or whose majority of populati- unconsciously, begin to detect the backgro- on are poor farmers with no access to tele- und of the new employee. It was common vision? Is it not a waste amidst the scarcity for reporters to assume that they would be of ETV resources?” said one ETV employee supported by management persons of the (Kejela, 2009). Within ETV, there was shor- same ethnicity on diff erent occasions, for tage of technical personnel and producers example if they applied for an educational who worked for the new regional language sponsorship. programmes. For the government, however, it was a deli- Downplaying ethno- berate strategy in its regional empowerment political tensions package to introduce Oromo broadcasts, For at least two decades, ETV and later EBC together with Tigrigna programmes. At this had a special role in promoting ethnic fede- time (1990s), there were no regional media ralism and multiethnicity for the populati- companies in Ethiopia, and the introduction on. Th e narrative was told in an educative/ of ethnic-based content occurred within the advocacy frame (Senait, 2011). Potential two monopolized state broadcasters in Ad- problems with the ethno-political system dis Ababa, that is, Ethiopian Television and were downplayed. Th e sources in news-ori- Ethiopian Radio. ented reports were chiefl y government offi - During Meles Zenawi (r. 1991–2012) and cials and supporters, as with most ETV re- well into the leadership of Hailemariam De- porting at the time. Journalists in the salegn (r. 2012–18), TPLF dominated the Some of the private outlets allowed critical Ethiopian media, leadership positions in the federal state me- views on ethnic politics, although limited. both state and dia. In terms of the production, however, the Studies of the media framing of the Nati- private, have Amharic divisions and departments were ons, Nationalities and Peoples’ Day – cele- developed a fi ne more powerful – and even more so today brated every year to commemorate the ra- judgement in rela- (2020). Both Amhara and Tigray employees tifi cation of the Ethiopian Constitution on 8 tion to potentially were associated with the government, parti- December 1994 – illustrate this. While the sensitive content. cularly in ETV. state media portrayed the event as a day of A restructuring process which started in celebration, the private weekly Addis Admas the state broadcaster around 2009, howe- questioned the purpose of the day and pro- ver, divided the diff erent language depart- posed that it has a divisive role, jeopardizing ments in a new manner and implicated that unity among diff erent people groups (Dessa- the Tigray and Oromo departments were legn, 2015; Hailu, 2015). relocated to regional mass media agencies Journalists in the Ethiopian media, both in and , respectively. Some state and private, have developed a fi ne jud- 11 gement in relation to potentially sensitive dents actually gave details about their ethn- content. Stories dealing with religion and icity when they had the option to remain ethnicity require particulary careful treat- silent on the issue. ment. Many such stories have not been run Among the proportion who did respond, at all if they contain latent confl ict. 16 diff erent ethnic groups were identifi ed, Internal confl ict, despite representing a of which three were vastly dominant: Oro- continuous challenge for Ethiopian govern- mo (31.3%), Amhara (20.7%) and Tigray ments, has had minimal coverage in both (36.3%). All other groups had a represen- state-owned and private newspapers. It has tation of 3% or less. Th e fi gures should not oft en been characterized by an annihilati- be regarded as fully generalizable for the on frame, meaning that actors, events and composition of Ethiopian journalists, as the explanations are simply absent in the co- sampling of media outlets (27 companies in verage (Mulatu, 2017). total) were not entirely representative. For Ethnicity is a strong Although much of the marginalization example, Oromia Mass Media Agency was predictor of process has occurred through self-censors- included in the study while Amhara Mass political sympathy hip performed by individual journalists in Media Agency was not. among Ethiopian the newsroom (Skjerdal, 2010), certain con- However, it is striking that about one- journalists. tent has been entirely banned from treat- third of the respondents identifi ed them- ment. Th is could include not just hard news selves as Tigray, while less than 10% of the coverage, but music and culture as well. By interviews were conducted in the Tigray re- means of example, Ethiopian Radio prohi- gion (Mekelle). Th e clear indication is that bited the song ‘Lalibela’ by the artist Gosaye the major media houses in , not Tesfaye from being played on the station least EBC, still had a considerable proporti- aft er complaints from the Hamina people on of employees hailing from Tigray at the (Gezahegn, 2007). Th e problem with the time when the data collection was underta- song was the description of a ‘Lalibela’ wo- ken (2013 to 2015). man who wander around early in the mor- Th e Worlds of Journalism Study (WJS) ning singing and begging, which is also a also queried about political sympathies, stereotypical portrayal of the Hamina. Th e which yielded some interesting though per- depiction of a minority group as poor and haps not surprising fi ndings vis-à-vis ethnic defi cient is where the trouble comes in. background. Th e most strongly expressed Ethnic issues are therefore delicate not pro-government stance (i.e. supporting only in relation to the larger political con- EPRDF at the time) was found among fl icts of the country, but on minority and Tigray journalists. Only 3.1% of the respon- community level as well. dents in this group identifi ed themselves as oppositional. Withholding of ethnic Th e opposite tendency was found among identity among journalists Oromo journalists, where a mere 12.8% sig- Ethnic identity remains a sensitive issue for nalled pro-government sympathies. Amha- journalists, as for other groups in society. ra journalists fell somewhere in-between, Many reporters and editors avoid expressing though very few expressed an anti-govern- their identity, although their affi liation with ment sentiment. It should be added that a particular language department or other most respondents regardless of ethnicity identifi ers would unescapably attach them placed themselves in the middle of the spe- to a particular background. ctrum when asked about political sympat- In the Worlds of Journalism Study (Ha- hies, meaning that they neither wanted to nitzsch et al., 2019), where 350 Ethiopian express a pro-government nor oppositional journalists were interviewed individually stance, perhaps to avoid any controversy. about their professional identity, only half What the WJS fi ndings indicate, neverthe- (or more precisely 51.9%) agreed to expose less, is that ethnicity is a strong predictor of their ethnic belonging. Another way to des- political sympathy among Ethiopian journa- cribe the response rate, of course, is that it lists. In the Oromo journalism fraternity in 12 was surprising that so many of the respon- particular, there seems to have been a critical attitude towards the central government for among the practicing journalists in the pre- long, both in privately-owned and state-run viously mentioned survey. media houses (Skjerdal, 20174). It is diffi cult to tell from these surveys alo- ne whether a unitarian identity is gaining ’Non-ethnic’ journalism students ground among young people, but the high Th at ethnicity is sensitive has also emerged proportion of the journalism students who in research conducted among journalism did not want to identify with a particular students. In 2015, 350 graduate and under- ethnicity is striking. graduate journalism students from fi ve uni- Th e question of possible impact of perso- versities (Addis Ababa, Bahir Dar, Mekelle, nal background in actual reporting is more Jigjiga and Wollega) were surveyed about diffi cult to settle, but the overall evidence their professional aspirations, personal qua- points in the direction of a strong relation lifi cations and so forth as part of the Journa- between identity and journalistic priori- lism Students Across the Globe project. ties, both in the global research and in as- A number of the respondents were stron- sessments of local reporting practices. Th e gly off ended by the question about ethnic Head of the Journalism programme at Ad- belonging. Characteristic responses were, dis Ababa University, Mekuria Mekasha, for “I hate this question”, “Is this necessary?” one, observes that professional journalists and “I don’t care about my ethnicity” (see have a tendency to report on their own peo- examples at the bottom of this page). A high ple group positively, while other groups are number expressed a pan-Ethiopian iden- covered more critically. Th e impact of own tity (“Th ere’s no ethnicity, I’m Ethiopian”). ethnicity is particularly perceptible in the Th ere was also a higher proportion of the coverage of confl ict issues (Mekuria et al., journalism students who informed of a dual 2018, p. 40). ethnicity (identifying with two groups) than

4 The WJS data regarding ethnicity have not been published before.

Responses from a survey among Ethiopian journalism students in 2015. Many of the students were off ended when asked about their ethnicity. 1: ”I am not ethno-centric (I see people equal).” 2: ”Ethiopian (don’t ask bastard questions).”

13 Media and ethnicity during the current Ethiopian leadership

By Terje Skjerdal

Th e media reforms aft er 2018 represent the most sweeping game-changer for the Ethiopian media since 1991. Th e ethnifi cation processes which have taken place in the media alongside liberalization have received less attention.

Outwardly, the public can today, for the fi rst unfamiliar in the Ethiopian media discour- time in Ethiopia’s media history, experien- se. Even liberal-minded politicians struggle ce genuine media pluralism. Th is implies a a little to get used to the new atmosphere. realistic choice between diff erent channels One of our informants from a state media and voices, in which audiences are free to channel tells about an incident where he consume oppositional media without fear of challenged the Prime Minister to respond reprisals. Th is contrasts the situation during to the claim that there are two governments the 2016 demonstrations in Oromia when in Ethiopia; one in Addis Ababa and one in TV watchers risked being detained in their Tigray. Dr. Abiy was irritated by the questi- homes if they tuned in to the wrong dias- on, but he gave an answer and the interview pora channels using satellite dishes. Th e was broadcast. In the past, it is unlikely that following analysis will begin by describing an interview where the premier was visibly such changes, felt by both audiences and annoyed would have passed the editorial media professionals, before turning to a de- desk, if the question was at all asked in the eper examination of the more challenging fi rst place. developments in the local media. Media houses still experience that govern- Internally, the most fundamental chan- ment offi cials call them to have a say on edi- ge for the media themselves is perhaps the torial decisions, but while such a call in the removal of fear mechanisms in the news- past meant a one-way delivery of directives, room. Th is has previously lead to wide- these days editors are able to discuss the case spread self-censorship in both state-owned with the offi cials. It is increasingly accepted and private media (Skjerdal, 2010). Not only that the editorial decision should rest with are the media now able to cover just about the media, not with a government offi ce. any topic, they are also free to invite and Another signifi cant change is non-discri- confront politicians who disagree with the minatory treatment of private and state-ow- station’s perspective. ned media channels. More than 50 outlet Overnight, public fi gures have been pus- are habitually invited to press conferences hed to adopt an attitude of transparency and and they realize equal access to public in- 14 accountability which traditionally has been formation. Th e Freedom of the Mass Media and Access to Information Proclamation 2007 was one such move; the scrapping of (590/2008) is under revision, and so are all the old press law and introduction of a new other laws related to journalism and media media law with an accompanying access to operations. information section in 2008 another. Nota- Representatives from all parts of the me- bly, licencing of private TV stations began in dia sector emphasize that the liberalization 2017, a year before Abiy entered offi ce. of the media is a positive move. It is hard to Th e problem of the past, nevertheless, fi nd any editor or journalist who defend the was that formal liberalization involved a previous government’s more restrictive line large degree of informal interpretation. For on issues such as media licencing or with- example, a journalistic tool such as access to holding of information. information was largely reserved for indivi- At the same time, it is important to em- duals with the right connections. Likewise, phasize that the liberalization of the Ethio- private broadcasting licences were kept for pian media did not commence with Abiy applicants close to the government (Skjer- Ahmed. Th ose who have followed Ethiopian dal, 2013b). media governance over time will have ob- Curiously, much media regulation enfor- served that the sector gradually has moved ced by the current Ethiopian leadership still towards more independence and greater depend on informal judgement. A number diversity. Th e allowing of private radios in of laws related to media practice have been

Informants

1. Girma Gutama ...... Oromia Media Network (OMN) 2. Tolera Fikru ...... Oromia Media Network (OMN) 3. ...... Oromia Media Network (OMN)/Oromo Federalist Congress (OFC) 4. Belay Manaye ...... Asrat TV 5. Gera Getachew ...... Ethiopian Satellite Television and Radio (ESAT) 6. Eskinder Nega ...... Ethiopis/Balderas for True Democracy 7. Daniel Berhane ...... Horn Aff airs 8. Gebrehiwot Tesfa ...... Dimtse Woyane Tigray (DWET) 9. Tibebu Belete ...... Ahadu Radio 10. Amare Aregawi ...... The Reporter 11. Tamrat G/giorgis ...... Fortune 12. Gashew Fentahun ...... Amhara Mass Media Agency (AMMA) 13. Zelalem Genemo ...... Southern Region Radio and Television Agency 14. Hagos Niguse ...... Tigray Mass Media Agency (TMMA) 15. Asfa Ahmed ...... Fana Broadcasting Corporate (FBC) 16. Ayele Addis ...... Amhara Journalist Association (AJA) 17. Tsegaw Tadesse ...... Tigray Journalist Association (TJA) 18. Nazif Jemal ...... Oromia Journalist Association (OJA) 19. Befekadu Hailu ...... Center for Advancement of Rights and Democracy (CARD) 20. Atnafu Brhane ...... Center for Advancement of Rights and Democracy (CARD)

Various other informants were interviewed on condition of anonymity, representing diff erent media channels in the capital city and the regions, both from the state-owned and privately owned media.

The research interviews were conducted between March and September 2020. 15 repealed, but they are yet to be replaced with bed in political terms, oft en following a pri- new legislation. An awaited Mass Media vate vs. state media axis, the new polariza- Policy Bill was discussed by the Council of tion is overwelmingly informed by ethnic Ministers in August 2020, but it is uncertain interest. It is now common to describe me- when it will be authorized. dia outlets primarily out of their ethnic sym- Many new media businesses in Ethio- pathies. Th is is most sharply pronounced in pia are therefore partially operating in a newly returned companies such as Oromia legal vacuum. Some of them are registered Media Network (OMN, Oromo-affi liated) as commercial enterprises, but they lack and Amhara Satellite Radio and Television proper broadcasting permission and must (ASRAT, Amhara-affi liated), but a num- transmit programmes from abroad even if ber of other media institutions have similar It is now common much of the content production is taking commitments to one region or one specifi c to describe media place in Ethiopia. Th eir formal legal position ethnicity. outlets primarily is therefore little diff erent than it was during Th e state media are subject to a similar out of their ethnic the previous government. tendency. In the past, offi cial media outlets What makes a big diff erence, nevertheless, were largely centralized and would repre- sympathies. is that the current government has welco- sent the EPRDF coalition as a whole, but med diaspora channels to the country and with the growth in regional mass media is not interested in clamping down on their agencies, each company is now nurturing its activities. Behind this observation there is an own identity, resulting in increased polari- important lesson in Ethiopian media gover- zation with agencies in other regions or with nance, namely that the informal aspects still local companies carrying a diff erent ethnic play an important role when defi ning the belonging. Ownership has become subordi- room for media freedom and permissable nate to ethnic commitment. For example, in public deliberation. Hence, when describing the Tigray region, privately-owned Dimtsi the ‘homecoming’ of the diaspora media Weyane Tigray joins forces with state-ow- in 2018, an informant uses the expression, ned Tigray Mass Media Agency to counter “Abiy allowed Jawar to return”, as if the fate the media from other regions. of Jawar Mohammed’s media company was Th e observed polarization is also of a dif- essentially a matter of personal judgement ferent intensity than before. Veteran media by the Prime Minister rather than the conse- personalities describe it as the fi ercest po- quence of formal media liberalization. And larization ever seen in the Ethiopian me- perhaps, to a large extent, it was. dia landscape, spurred by an ‘amplifi cation of the ethnic agenda’ (anonymous veteran Intensifi ed polarization journalist). And while the divergence bet- One of the unsolicited secondary eff ects of ween diff erent media outlets is greater than the newborn media pluralism in 2018 is in- before, it is also being played out at multiple creased polarization between media outlets. levels. On the media type level, the diff erent It should be underlined that polarization is outlets gravite towards activist types, mode- not a new phenomenon in Ethiopian me- rate types, and loyalist types. On the politi- dia logics. In fact, it was one of the foremost cal level, outlets polarize around ethno-na- characteristics of the media environment in tionalist versus pan-Ethiopian positions. the 1990s when radical print outlets with On the ethnic level, media companies more sympathies for the previous Derg regime openly than before express allegiance to an challenged the government through rebel Oromo, Amharic, or Tigray identity and so journalism. Strong polarization lasted at forth. least until the watershed elections in 2005, An exemplar way to detect polarization making the Ethiopian media more divided is through a scrutiny of the diff erent chan- than comparative media environments in nels’ presentation of the same stories, as East Africa (Stremlau, 2018). demonstrated by this study’s accompanying However, the post-2018 polarization is of framing analysis (see the latter part of the a diff erent type and character than before. report). On the most fundamental level, the 16 While previous dividing lines were descri- choice of which stories to cover and which ones to ignore (annihilate) is a sign of diver- International Women’s Day 8 March 2020. gence between the channels. When the same As part of the live transmission, the channel events are covered, the choice of sources aired a statement by an Oromo woman who tends to make a signifi cant diff erence, as do warned against interethnic marriage, calling the selected narrative and the interpretation for Oromo men to divorce their habesha of events. (Amhara and Tigray) wives. Th e testimoni- A focused study of crisis coverage in EBC al was viewed as a provocation to set diff e- vs. OMN is revealing. While EBC calls at- rent people groups up against each other. In tention to stability, OMN brings provocati- fact, when asked about the incident, OMN ons for mass rallying. While EBC forwards agrees that the broadcasting of the statement offi cial sources, OMN allows their own re- was unfortunate and claims that the station porters to speak. Where EBC seems to un- would not have transmitted the incident if derestimate the cases and their consequen- they knew it was coming (Girma Gutema, ces, OMN exaggerates. And, notably, the personal interview, 13 March 2020). Audi- two channels never use the other channel ences, nevertheless, perceived the statement as a source, according to the study (Endal- as a message not just coming from a woman kachew, 2018). who was given the chance to speak on a live In this study, we found that in the few transmission, but from OMN itself. cases where journalists quote stories from Studies indicate strong diff erences betwe- Expressions of other channels, they do it solely to show so- en the social media and the edited media hatred and animo- lidarity with people referred in the story or when it comes to hate speech. Expressions sity are much more to show how a story reported in another me- of hatred and animosity are much more pre- prevalent on social dia outlet is dangerous to their own ethnic valent on social networking sites than in the networking sites group. In other words, identity politics ap- professional media. Th is is evident in reports than in the profes- pears to override professional obligations in from other parts of the East African region sional media. the daily priorities in the media. Th e outco- (Dolan and Ferroggiaro, 2019; York, 2019), me is fragmentation in the media landscape, and is documented in research from Ethio- where diff erent channels operate in the same pia as well. One study found that 95.4% of environment but polarize in keeping with entries on Facebook, Twitter and YouTube political and ethnic agendas. contained hate speech utterances when the topic was ethnicity or religion (Mulugeta, Mounting hate speech 2019). Th e situation was diff erent during Many – if not most – journalists and editors the 2015 elections when there was found view hate speech as growing and a major relatively little hate speech in the online problem for the Ethiopian media. Th e ope- debate. According to a study of election-re- rational defi nition of hate speech here is lated Facebook postings, only 0,4% of the described broadly, referring to advocacy of statements could be classifi ed as hate spe- hatred against a group which constitutes in- ech (Gagliardone et al., 2016). It should be citement to discrimination, hostility or vio- emphasized that objective measures of hate lence – although the demaraction between speech are diffi cult, and results will to a large acceptable and unacceptable provocation extent depend on choices made in the sam- depends on individual perspectives (Somer- pling as well as the defi ning criteria of what ville, 2011). Th at hate speech prevails in the constitutes hate speech. However, based on social media has been an established fact for observation over time and the overall rese- long. What is new, according to media pro- arch evidence, the indication is that online fessionals, is that hate speech utterances and hate speech has increased. incitement to violence have moved to the Th e Center for Advancement of Rights and mainstream media as well (i.e. the journalis- Democracy (CARD), a local NGO, has do- tically edited media). Th e type of hate speech cumented a clear correlation between peaks the informants have in mind is particularly of online hate speech and outburst of violen- that which contains ethnic animosity. ce. One example is the use of the infl amma- One incident several informants mention tory term neft egna (derogatory for Amhara was a live event broadcast by OMN on the people) on Twitter, which peaked when 17 violence occurred in October 2019 and July mation Proclamation in February 2020, as 2020 (fi gure 1). It is also found that locally an exacerbation of the problem (Téwodros, posted hate speech decreases whenever the 2019). government activates Internet blockouts. Among journalists and editors consulted Th is happened for example in early 2020, in this study, there are also diff erent opini- when the Internet was blocked in western ons. Somewhat surprisingly, perhaps, edi- Oromia, including Ambo University, where tors with an otherwise liberal track record many of the protest voices came from. call for more active engagement by the go- According to CARD, most of the utteran- vernment when it comes to dealing with ces which might be defi ned as hate speech tendencies of extremism. Th ey associate originate in the diaspora. Th is they know the current attitude of the government with because entries from netizens abroad ty- incapability and powerlessness. “It is as if Much of the hate pically lack context and are less connected someone is stealing my car when the police- speech can be with the actual events on the ground. Equal- man is standing next to it, just watching,” traced to online ly alarming, much of the hate speech can be one editor explains. Others view hate speech platforms at traced to online platforms at Western uni- as a natural consequence of the new media Western versities, indicating that Ethiopian diaspo- policy regime where free media are allowed, universities. ra scholars and students are active posting and they see no rational way for interven- infl ammatory messages in diff erent online tion unless liberal principles are overruled. media channels to infl uence the population Although hate speech is mainly associated back in Ethiopia. Local participants in the with user-generated online content, it is also debate are more cautious (cf. Abrham, 2019; becoming a problem in the professional me- Meseret, 2020; Sileshie, 2014). dia, according to informants. Th e part of the Th e use of legal action to curb hate speech media which is fi rst singled out is the ‘ethno- is controversial. Some are of the opinion that nationalist’ media, but also regional mass ethnic hatred in the media has come to such media agencies are mentioned. an extent that it could be equated with Radio When referring to the ethno-nationalist des Mille Collines’ role during the Rwandan media, mainly two channels are mentioned, genocide in 1994 and that action therefo- OMN and Asrat Media. CARD, in its nuan- re needs to be taken (Ayele, 2020). Others ced analysis of hate speech, would not fully propagate a more liberal stance and view agree that the two channels are comparable intervention by authorities, such as the in- with the social media when it comes to ethn- troduction of the Hate Speech and Disinfor- ic provocation and incitement to violence.

Figure 1: Occurrences of infl ammatory terms on Twitter, July 2018 to July 2020. Source: Cen- 18 ter for Advancement of Rights and Democracy, CARD (used with permission). Girma Gutema took over According to the NGO, ethnically-motiva- of these three companies pinpoints cases of as Executive Director of ted channels such as OMN know quite well ethno-nationalistic media polarization. OMN from Jawar Mo- how far they can go without overstepping hammed in 2020. Oromo the limit for acceptable speech. However, OMN and the infl uential nationalism is visible in monitoring agencies suggest that the chan- role of Jawar Mohammed his offi ce by means of the nels are constantly guilty of ‘indirect incite- Oromia Media Network (OMN) began as OLF fl ag to the left and ment’, in the sense that they let invited guests a satellite TV station among the Oromo di- the fl ag of the Oromia propagate a one-sided ethnic agenda which aspora in Minneapolis, Minnesota in 2014. regional state to the right. feeds activists and demonstrators with ide- Th e station is still registered as a charity in (Photo by Terje Skjerdal) as which in turn escalate in the social media the US, since such an arrangement is not al- as an element of ‘outrage communication’ lowed in Ethiopia. (Téwodros, 2020). Th us, fairly or unfairly, a Th e fi nancing of the station’s activities is few newcomers on the Ethiopian media sce- somewhat opaque. It is clear that US audien- ne are frequently blamed for the increased ces do fundraising activities for OMN. In Et- aggression observed in the professional part hiopia, income is generated through adverti- of the media. Who are they, and how do they sing and sponsorship from banks and other perceive their own identity? sources, but the fi nancing is demanding be- Th e following section will do a closer cause businesses steer away from the station scrutiny of the most characteristic ethno- due to its anti-government reputation. Th e centric channels in three diff erent directi- network has 41 employees in Addis Ababa, ons: Oromo, Amhara and Tigray. OMN and of whom 28 are journalists. Asrat Media represent Oromo and Amhara OMN has an outspoken aim to advocate identity, respectively. Th ey are both young for the Oromo people. Th e rationale for this, media companies based in Addis Ababa and in the words of the station’s director, is that are to a certain extent staff ed with returned the Ethiopian media landscape is skewed diaspora activists. Dimtsi Weyane Tigray and “eliminates a part of the country” in its (DW), on the other hand, has a four-decade coverage (Girma Gutema, personal intervi- long history as a Tigray-affi liated media out- ew, 13 March 2020). At the same time, OMN let, for many years based in Mekelle. Each claims to provide a platform for diff erent pe- 19 Above: OMN began as a ’struggle medium’ in 2014, but now claims to take on a new profi le. (Photo by Terje Skjerdal)

Right: A painting in Jawar Mohammed’s reception room portrays him as the personifi cation of OMN. (Photo by Terje Skjerdal)

ople groups, demonstrated by the fact that coverage, while the station took advantage it broadcasts in seven diff erent languages, from the constant sharing of fi rst-hand in- allegedly more than any other compara- formation about political developments and ble media channel. However, all languages events. Th e integration of the social media except Amharic turn out to be confi ned to in the communication strategy was crucial. the Oromia region, or to SNNPR zones close Th us, in Jawar’s own words, OMN is more of to Oromia. Th e station’s management esti- a social movement than a media channel. mates that 90% of its employees are Oromo. Two years into the new government, Politically speaking, OMN supports ‘mul- OMN sees itself as entering a new role. Whi- In Jawar’s own tinational federalism’, which for outsiders le the station was established as an activist words, OMN is would seem similar to ethnic federalism, but media channel, it now pretends to weaken more of a social the latter is a term that the station defi es. The its political linkages. Jawar has entered po- movement than a perspective contrasts unitarist and secessio- litics (Oromo Federalist Congress, OFC) media channel. nist positions, which are the other two main and is no longer the company’s director. He political stances found in Oromo politics claims to be fully out of the media operation (Østebø and Tronvoll, 2020). and only has an informal advisory functi- OMN played a key role in staging the qe- on vis-à-vis OMN. However, it is revealing erroo protests from 2016 onwards (Mulua- that when he talks about OMN’s activities, it lem, 2019). Th e station utilized on-the-site does not go long until he begins referring to reports from volunteers, who took on a role ‘we’ instead of ‘they’. So, in his own percepti- as citizen journalists. According to the stati- on, he is still an integrated part of the media on’s founder Jawar Mohammed, the integra- house’s profi le and engagement. Offi cially, tion of media and politics is a key to OMN’s he stepped down from the company in Janu- success. Th e leadership of the station and ary 2020 when Girma Gutema took over as 20 the leadership of the movement were inter- director. vowen. Th e protesters benefi ted from media Th at OMN is assuming a new role, is not something that representatives from other would already be at fi re”. eTh problem, acco- media companies seem to discern. For them, rding to him, is that people confuse strong OMN continues to be an activist channel for opinion with hate speech. His perspective ethnic interests, as intense as ever. OMN’s on the matter is contested by various other role during the death of Hachalu Hundessa actors in the media environment. in June 2020 and the subsequent turbulence OMN has competitors on both regional served to confi rm this perception. One week and national levels. Th e regional competiti- ahead of the killing, OMN aired an intervi- on is largely about audiences and fi nances, ew with the popular Oromo singer where while the national competition is more about he commented on a number of contentious ideology and ethnic belonging. Within the political, ethnic and historial issues, besides Oromo population, a private diaspora-ori- Tigray-affi liated being challenged on his alleged support for ginated company like Oromia Broadcasting media are a defi ned the Prime Minister (Endalkachew, 2020). Service (OBS) represents a competitor for enemy for OMN. Th e following days and weeks demonstrated OMN in terms of income generation. Ho- the channel’s proximity to Jawar as they fol- wever, the content of OBS is more entertain- lowed his arrest and the battle over Hachalu’s ment-oriented than OMN and the two chan- burial. nels are not directly rivals on the news arena. OMN is conscious about its political role. Th e regional mass media agency OBN, on Jawar admits that they push their agenda the other hand, is seen as a real competitor “really aggressively”, but within the confi nes for OMN in terms of audience shares, as it of the law. Th e station’s self-perception is in- represents a pro-government political stan- formed by a liberal free speech mentality co- ce. Th e most fi erce competition for OMN on rresponding to American First Amendment the national level is with channels of diff e- principles. Th e mindset is easily at odds with rent ethnic belonging. Tigray-affi liated me- Th e popular Oromo artist an Ethiopian loyalist, nation-building repor- dia are a defi ned enemy for OMN, as OMN Hachalu Hundessa (to the ting philosophy (Skjerdal, 2017). and its political patron OFC advocate a nar- left ) was interrogated by When it comes to hate speech, OMN deni- rative where Tigrayans since since 1991 have OMN’s journalist Guyo es any allegation that it forwards ethnic ha- represented the ‘hardliners’ in Ethiopian po- Wariyo on political and tred or engages in incitement to violence. litics, continuously threatening Oromo in- ethnic issues one week Jawar claims that if he did post anything terests (Jawar, 2020). Similarly, nationalistic before he was shot dead that came close to hate speech, “this country Amhara-aligned outlets are rivals for OMN. in Addis Ababa. (Capture from OMN 22 June 2020)

21 Belay Manaye is editor in Asrat TV. Asrat Media is Discussed in the following section, Asrat while there are smaller branches in Bahir expanding and moved to Media has a special position in this regard. Dar and Gonder (Amhara region). Th ere new offi ces in Addis Ababa are concrete plans to open two new bran- in 2020. Belay was detai- Amhara ethno-nationalism ch offi ces. Th e expansion is currently at a ned together with three and Asrat Media standstill, since Asrat’s activities have been other journalists from Asrat Media is in many ways the equivalent temporarily scaled down aft er several of its the station in July 2020, of OMN on the Amhara side. Like OMN, key personnel were arrested in July 2020 in accused of incitement to Asrat Media has an unconcealed intention connection with political turmoil and are communal violence. Belay to speak with an ethnic mandate. Its per- still in detention awaiting a court process (as is here photographed out- ception of the media situation in Ethiopia is of November 2020). side of Asrat Media’s news that Amhara voices are neglected, and that Th e production is only on television so far, building in March 2020. the Amhara region and its people are falsely transmitted through satellite, YouTube and (Photo by Terje Skjerdal) blamed for wrongdoings in the country, his- Facebook. A radio channel is in the plan- torically as well as today. ning, pending licence from EBA. Th e inten- Th e acronym ASRAT literally stands for tion of Asrat Media is to cover all of Ethio- Amhara Satellite Radio and Television, but pia, although the Amhara region is its main the name is double-barrelled and is usual- focus. Language-wise, the station so far only ly spelled with small caps (Asrat), referring broadcasts in Amharic, but it has an intenti- to Asrat Woldeyes, the personal doctor of on to start programmes in Afan Oromo and Emperor Haile Selassie who later became Tigrigna as well (plus English). Th e televisi- professor of medicine at Addis Ababa Uni- on production mainly consists of news (40 versity and who founded the All-Amhara minutes per day) and documentaries. Th e People’s Organization (AAPO). Like the dis- channel will broaden its profi le by includi- tinguished surgeon, Asrat Media aspires to ng entertainment programmes in the near be the voice of the voiceless – and the voice- future. less that Asrat Media pretends to speak on One of Asrat Media’s aims is to stand up behalf of are the negelected Amhara. for the Amhara people and expose injustice Asrat Media is licensed as a public media which other media channels neglect either company in Washington DC, but most of on purpose or by design. Th e coverage of its staff work from Ethiopia. 12 journalists the abducted Amhara students in Western 22 belong to the main offi ce in Addis Ababa, Oromia in December 2019 is a case in point. Th e whereabouts of the students, at least 16 trean People’s Liberation Front (EPLF) in in number, are still unknown at the time of the fi ght against the Derg. Aft er ideological writing (November 2020), but in July 2020, battle between the two insurgence groups, the Federal High Court charged 17 members Voice of Woyane became a purely TPLF-dri- of the Oromo Liberation Front (OLF) for ven radio in 1985 (Stremlau, 2012). Since kidnapping the students5. Indicatively, the 1992, when the radio station no longer was fi rst channel to release the story about the forced to do underground operations, DW disappearance of the students was Asrat Me- has transmitted from Mekelle (Meseret, dia. It followed up closely in the next weeks 2013). and months. Family members of the abduc- DW’s commitment has consistently been ted students were interviewed, and govern- towards Tigray, but its political role has ment representatives were confronted. As changed throughout its four decades of exis- revealed by the framing analysis later in this tence, echoing the major turns in Ethiopian report, the federal state media and media politics. In the 1980s it was a resistance ra- channels with a diff erent ethnic alignment dio; from 1991 to 2018 a pro-government were largely silent about the incident, or fra- radio; and since 2018 an emerging opposi- med it quite diff erently than Asrat Media. tion radio, culminating with TPLF’s depar- OMN treated it primarily as a drama. ture (or, in TPLF’s view, exclusion) from the Media people and commentators general- government coalition in November 2019. ly portray Asrat Media as ethnically incli- Th e long-lasting symbiosis between the ned and one-sided. Th ey claim that invited media outlet and regional politics is under- guests in the channel are oft en ‘Amharic scored by the fact that several of TPLF’s cur- extremists’ who reiterate a similar narrative. rent leadership fi gures were the ones who The recent realign- Th e channel is found to create a narrative operated the radio during the struggle in ment of DWET where the federal government is believed to the 1980s. Th e party’s chairman since 2017, implies that the secretely interfere with and destabilize the Debretsion Gebremichael, led the team who station has Amhara region (Daniel, 2020). initially set up the radio and served as Voice become a conduit OMN claims that the diff erence between of Woyane’s technician during those years for Tigray’s them and Asrat Media is that while OMN is (Gagliardone, 2016). grievances vis-à-vis open about its ethnic affi liation, Asrat Media Th e recent realignment of DW implies that central Ethiopia. beclouds their identify behind a ‘non-ethn- the station has become a conduit for Tigray’s ic’ company name. An OMN representative grievances vis-à-vis central Ethiopia. Th e condemns Asrat Media for being elite-ori- media company’s loyalty to the federal go- ented. Other informants suggest that the vernment has eroded entirely. Illustratively, reason Asrat Media attracts support from when Tigray went ahead and held regional businesses and people is that they, represen- elections in September 2020 against war- ting a group which has lost power to Oromo nings from the federal government, DW interests in the top leadership, feel increas- covered the elections diligently. By contrast, ingly under pressure in the current political Ethiopian Radio did not mention the event landscape. by a word on the election day, following in- structions from the Ethiopian Broadcasting The realignment of Authority. More over, DW closed its Ad- Dimtsi Weyane Tigray dis Ababa branch aft er the assassination of Dimtsi Weyane Tigray (DW) has a diff erent Hachalu Hundessa and the reporters were background than both OMN and Asrat Me- sent back to Mekelle. dia, but the company is equally ethnic-based What we make of this is that govern- and is closely tied in with political interests. ment-loyal media outlets which used to be Established in 1979, the clandestine radio in close partnership have chosen diff erent was fi rst a joint initiative by the Tigray Pe- directions and have been separated as a re- ople’s Liberation Front (TPLF) and the Eri- sult of the political changes since 2018. Th e

5 ‘Ethiopian court charges 17 individuals for kidnapping Dembi Dolo university students’, Ezega News, 18 July 2020. Available at: hƩ ps://www.ezega.com/News/NewsDetails/8024/Ethiopian-Court-Charges-17-in- 23 dividuals-for-Kidnapping-Dembi-Dolo-University-Students Tigray region has become more nationalis- company still retains studios in Amsterdam, tic, and so has its media. (Further discussion London and Washington DC, although of DW and Tigray media politics is provided most of its production now takes place in Et- The Tigray region later in the analysis.) hiopia. Th e company has 23 staff members has become more in its headquarters in Addis Ababa, half of nationalistic, and Pan-Ethiopianism and ESAT whom are journalists. so has its media. A number of other newcomers on the Et- ESAT has a large following in Ethiopia hiopian media scene, many of them retur- and is highly depending on its network of ned diaspora channels, have come a little in informants. Th e informants trust the chan- the shadow in the debate surrounding the nel to speak up against the authorities and markedly ethno-nationalistic channels. Th e for pan-Ethiopianism, although the new most prominent of these is arguably Ethio- political line of Abiy Ahmed has made the pian Satellite Television and Radio (ESAT), channel relatively less sceptical to the go- which since its inception in 2010 has been vernment than when it reported from the one of the most vocal networks in oppositi- diaspora. With a certain base in Amhara on to the Ethiopian government. It was ban- nationalism, ESAT has always used Amha- ned by the authorities until 2018 as a result ric as its main language. However, the chan- of its affi liation with Ginbot 7, a group which nel also uses Afaan Oromo and recently the Parliament had defi ned as a terrorist or- introduced Tigrigna in its news broadcasts, ganization. Th e channel has been less wil- which perhaps testifi es to its aspiration to ling to admit its association with Ginbot 7, become more all-inclusive. but its closeness to leaders of the party such as Berhanu Nega is rather evident (Mesfi n, The resurrection of Ethiopis 2019). Several journalists and managers in Another example of a channel with Amha- the media company have served in various ra roots and pan-Ethiopian sympathy is the capacities in the party. Illustratively, ESAT weekly newspaper Ethiopis. Th e publication played a key role in the welcoming of Ginbot has a long and chequered history. It was fi rst 7 back to Ethiopia on September 8, 20186. launched as a EPRDF-critical outlet by Es- ESAT continues to be critical of the go- kinder Nega in 1993. Eskinder was later sen- vernment, but its style has changed from tenced for support to terrorism (July 2012), when it was purely a diaspora channel. Th e but the newspaper was re-launched in a new

6 ‘ESAT patrioƟ c Ginbot 7 welcome ceremony at Addis Ababa InternaƟ onal Airport’. Available at: hƩ ps:// www.youtube.com/watch?v=WPvSbU44gi0

Deputy editor Gera Getachew in ESAT’s news studio in Addis Ababa. Th e channel stands for pan-Ethiopianism and is regarded as less contro- versial today than when it was operated from the diaspora. (Photo by Terje Skjerdal)

24 form aft er his release from prison in 2018. or ethnicity. Today, Eskinder has entered politics as lea- Th e growth of such companies is not a re- der of Balderas for Genuine Democracy. sult of the liberalized media policy that came He is still the owner of Ethiopis, but has with Abiy Ahmed. Reforms started already pulled out of the daily news production. In during Hailemariam Dessalegn, especially July 2020, Eskinder was again arrested, this aft er 2017 when private TV broadcasting time for allegedly organizing youth protest licences began to be issued (Busch, 2017). groups in the aft ermath of the assissination Th e commercial media market in Ethiopia is of Hachalu Hundessa. still growing, and this is a development far Th e oscillating role of Eskinder Nega is in- less contested than the proliferation of poli- dicative of the hybrid identity of many priva- tically and ethnically based media. The oscillating role te actors in the local media market. Th ey are of Eskinder Nega committed to professional journalism, but Regionalization of is indicative of the they also propagate a political agenda. Es- Ethiopian journalism hybrid identity kinder’s own response to this claim is that he It is no longer enough to review media hou- of many private sees no problem with the shift ing roles. To ses in Addis Ababa to grasp the currents in actors in the local the contrary, he regards the proximity bet- Ethiopian journalism. More and more of the media market. ween media and politics as fruitful for both journalistic production is taking place in the parties. regions. Th is tendency has emerged over a Every time Eskinder is detained, however, number of years, little by little. Th e most im- it appears as somewhat unclear whether he portant edifi ce in this regard is the regional- is arrested in capacity of being a politician ly-owned ‘mass media agencies’, which are or a media practitioner. International press operated by the regional states and aimed free dom bulletins habitually labels him jour- at the local population. Th e vital activity of nalist or editor, but local authorities prefer to the mass media agencies are TV and radio see him as a political activist. Furthermore, broadcasts, but many of the agencies also in the local debate, it is hard for any renow- publish print products on a weekly basis, ned media personality like Eskinder Nega oft en in diff erent languages. It should be no- to escape ethnic labelling. In online discus- ted that the mass media agencies, although sions, he has been called both Oromopho- organized under the regional states, are not bic and anti-Tigrayan (Hewan, 2019). Th is, as such an arm of the public relations of the again, is symtomatic of the increased ethnic regional state. Each regional state also has interpretation that we claim to observe in its own information bureau within the sta- the Ethiopian mediatized debate. te administration which takes care of public relations vis-à-vis its citizens, which might The popularity of include its own periodicals (e.g. weekly non-political channels newspapers). Th ese should not be confused When discussing private media outlets and with the outlets produced by the mass media the role of the returned diaspora, it should agencies. Th e regional mass media agencies not be forgotten that much of the activity is has a journalistic mandate, although the ties non-political. A number of the TV compa- to the regional state are perceptible in vario- nies which operate from Addis Ababa today, us ways. although they technically broadcast the con- All regions, including the city administra- tent via satellite from foreign cities such as tions of Addis Ababa and Dire Dawa, have Dubai, focus on leisure and entertainment. their own mass media agency. Most agencies Ethiopian Broadcasting Service (EBS) is were established from 2005 onwards (Ward, one of the largest companies of this kind. It 2011). In the fi rst phase, they depended a lot broadcasts sports, comedies, cooking shows, on the federal media, especially Ethiopian church services, and so forth. Current aff airs Radio and Television Agency (ERTA, now are part of an ‘Africa Report’ programme, but EBA), for technical support and transmis- news is not a priority – in fact, disregarded. sion. In recent years, however, the diff erent EBS aims to promote ‘Ethiopian and African regional agencies have invested substanti- values’ without favouring a particular region ally in their own equipment and terrestrial 25 transmitter stations. On top of that, many among specifi c ethnic cohorts, and their agencies also broadcast on satellite, both as distinctiveness challenges a pan-Ethiopian a means to reach their own rural population agenda. Th ree of the agencies in particular and to give a service to audiences abroad. – Oromia, Amhara and Tigray – are also TV Oromia (also known as OBN), for exam- highly conscious that they compete with ple, utilizes six satellites to serve Oromos the federal state media, both with regard to living in East and North Africa, the Midd- audiences and political narratives. Amhara le East, Europe, North America, Australia Mass Media Agency (AMMA), for example, and large parts of Asia (Busch, 2017, p. 46). has a designated reporter who follows every Th e fi nancing of the agencies mainly comes move of Abiy Ahmed, aiming to break news from the administration of the regional sta- about the PM and the government before te, although each agency is perpetually en- EBC does so. couraged to collect more of the funding by In 2016, AMMA opened its own branch in own means, i.e. through advertising. Addis Ababa, from where ‘correspondents’ Th e number of journalists working in the will report back to Bahir Dar and the rest regional media is overwhelming. Many of of the Amhara region about national issues. the mass media agencies have several hun- Th e Addis Ababa branch of AMMA views dred journalists in their ranks. Th e largest, itself as a conduit for Amhara perspectives Oromia Radio and Television Organization in the capital city, and they invite persons of (ORTO), employs 530 reporters and editors; their ethnicity, particularly politicians and Amhara Mass Media Agency has around other elites, to comment on important issues 400; while the agencies of the Southern Re- and to bring forth alternative solutions and gion and Addis Ababa City Administration refl ections. isTh includes sources who may have around 300 journalists each (the total not necessarily agree with offi cial govern- In the past, EPRDF number of employees is around twice as ment policy. For example, Eskinder Nega, may perhaps have much for each company). What this means leader of Balderas for Genuine Democracy, viewed the mass is that the regional state media have surpas- has been given ample time to speak about media agencies as sed the federal state media in terms of man- his political views. Occasionally, representa- regional extensions power. Th eir content production does not tives of other ethnicities, like OLF leaders, quite reach up to the expectations in relation are used as sources. Jawar Mohammed has of the federal to the number of journalists employed, but been invited too, but he has so far not prio- government’s several of the agencies broadcast at least 16 ritzed giving interviews to AMMA. communication hours per day on television and perhaps 24 Th e coverage of AMMA, as for the other strategy, but this is hours on radio. mass media agencies, is nevertheless clearly no longer how the Th e use of languages varies a lot, with in favour of the respective regional govern- state media South Region Radio and TV Agency in the ment. Th is could also mean challenging the function in the most extreme position, airing programmes policy of the federal government. In the country. in 49 languages each day (Fekede, 2019, past, EPRDF may perhaps have viewed the p. 44). Only one other media company in mass media agencies as regional extensions Sub-Saharan Africa can boast a similar lan- of the federal government’s communication guage diversity – a media business in Nige- strategy, but this is no longer how the state ria. media function in the country. Th e regional media have developed their own identity, The regional media as an ally and their political commitment is fi rst of all against the federal government to the regional government. Th at said, when Th e regional mass media agencies increa- peace and stability are jeopardized, the re- singly nurture a profi leindependent from gional mass media agencies have a history that of the federal state media. Th ey are con- of turning loyal to the central government. scious that their audience are people living During the Oromo protests that broke out in the regional state and, increasingly, peo- in 2016 and the subsequent state of emer- ple of their ethnicity residing in other parts gency periods, ORTO was faithful to the of the country and abroad. As such, the re- government and largely followed the line of 26 gional mass media serve to bring cohesion EBC (Busch, 2017, p. 49). Th e latter does not Above: Amhara Mass Media Agency in Bahir Dar has highly advan- ced studios and a helicopter landing pad on its premises.

Left : Th e Southern Nations, Nationalities and Peoples’ Region Radio and Television Agency in Hawassa, popularly called South Radio and TV, produces content in no less than 49 languages every day.

Below: A reporter’s vehicle parked outside of the headquarters of Ethiopian Somali Television (ESTV) in Jigjiga. ESTV is part of the Somali Regional State Mass Media Agency. (All photos by Terje Skjerdal)

27 necessarily represent the regional agency’s primary volition; an element of felt coerci- on from the federal government could also have infl uenced the editorial line. In the regions, In the regions, people view the mass me- people view the dia agency as an ally who speaks up for them mass media against the federal government. EBC, by agency as an ally contrast, is regarded as a mouthpiece of the who speaks up for government and habitually triggers anger by ETV (EBC) and FBC were targeted in the protests in Gonder in 2016. them against the the local population on contentious issues. federal govern- During the mass protests in the Amhara re- ment. gion in 2016, the state broadcaster became before the central media began to adopt a one of the targets. Th e message of the protest new political line where ‘ethnic federalism’ was ‘Stop mass killing of Amhara people’, disappeared from the vocabulary and regio- but among the protesters one could also spot nal autonomy was modulated (cf. Svensson, posters with the slogan ‘Shame on ETV & 2019). Equally important, Tigray power was FBC’ (Fana Broadcasting Corporate). Peo- probed, which had been unheard of in the ple in the area in the fi rst place had diffi culty state media ever since EPRDF stepped in. with EBC’s portrayal of the Welkait confl ict, A major provocation came a few months but as events unfolded, the allegedly biased into Abiy Ahmed’s tenure, when EBC bro- coverage of the subsequent protests became adcast a documentary giving details about a provocation in itself. Indeed, it went to systematic torture commanded by security the extent that EBC had to cancel reporting offi cers (‘Yefi tih Sekoka’/‘Agony of Justice’). trips to Gonder because they were threate- Human Rights Watch characterized the do- ned by local citizens who warned them that cumentary as ‘refreshing’ (Horne, 2018), but they would be chased if they dare arrive. Th e to TPLF and audiences in Mekelle, the hour- regional broadcaster, however, was most long programme was greatly disturbing, as welcome and covered the demonstrations it framed Tigray offi cials and the TPLF lea- on a day-to-day basis. dership as chief perpetrators (Amanuel Y, Th e place where regionalization of Et- 2019). A few weeks earlier, EBC had expo- hiopian journalism has been most sharply sed a largely Tigray-driven network invol- pronounced since 2018, however, is in the ved in industry corruption. northernmost part of the country, in Tigray. At that juncture EBC lost its confi dence Th e circumstances in that region require a among Tigray audiences, if not before. Pe- distinct treatment. ople and leaders in the region claimed that the perspective of the broadcaster was bia- Consolidation of the sed and that it had abandoned its peaceful Tigray-based media intentions. According to Tigray informants Amidst the attention given to the reconfi gu- in this study, people of the region today ration of the media landscape in the central straightforwardly ‘hate’ EBC and have no parts of the country, the changes in the me- interest in following the federal state me- dia situation in the far north could easily be dia. Th e animosity towards the state media disregarded. However, the development in involves Fana and Walta as well, since the- the Tigray-affi liated media since 2018 quite se channels toe the line of the government, markedly captures some of the major ten- strengthened by the fact that they too chose dencies in local media governance. It also to distribute the ‘Yefi tih Sekoka’ documen- demonstrates how media, politics and ethn- tary. icity are deeply intertwined in the Ethiopian Th e political processes during the fi rst two society. years of Abiy Ahmed’s rule have resulted Nearly three decades of TPLF dominance in a full consolidation of the Tigray-based in the federal media were eff ectively termi- media. TPLF can no longer trust the fede- nated when the reformist government as- ral media to speak for its interests, and is 28 sumed offi ce in 2018. It did not take long thus relegated to the region’s own media. Advisor to the President of One of the two largest media houses in the broadcasting only; it produces content in the Tigray region, Geta- regional capital city Mekelle is Tigray Mass 2-3 local languages in addition to Tigrigna chew Reda, talking con- Media Agency (TMMA), although the insti- and Amharic; and like TMMA, the compa- fi dently to DW about the tution actually has a rather modest structure ny has around 175 journalists. And more armed confl ict in Tigray compared with the mass media agencies in important for this analysis: Th e two media on 14 November 2020. Th e some of the other regions. Th e number of houses have a nearly identical political pro- interview was willingly journalists is 175, and the total number of fi le. Both are generously informed by TPLF transmitted by Tigray Te- employees around 300. In contrast to some in source selection and reporting approach. levision (TMMA) as well, of the other regional agencies, TMMA does Th e mission of the Tigray-based media, in testifying to the closeness not have a print publication, and the online their own words, is to balance the narratives between the two Mekel- activity is minimal. Th e fi rst radio studio fl owing in from the federal government. le-based media companies. was built only in 2020. However, the agen- According to employees, DW enjoys re- Th e interview demonstra- cy is preparing major investment, and the latively more freedom than TMMA when tes the congruence betwe- number of employees is expected to double it comes to potentially uncover mismana- en media and ethnicity in the coming years. Mekelle recently got its gement in the regional administration and in today’s Ethiopia. It fi rst decent printing press which makes pu- to do critical reporting. But the critical co- also refl ects contempora- blication and distribution of regional news- verage is ‘semi-investigative’ at best (Eden, ry political realignment, papers easier. 2020). No one doubts that DW is politically as Getachew served as Th e revitalization of the Tigray-based me- affi liated, and it would be hard for the station Minister for Government dia draws little distinction between the regi- to do negative coverage of TPLF, which they Communication Aff airs onal state media and outlets which offi cially feel is the agenda of the federal media these during Hailemariam De- are defi ned as commercial media. eTh major days. salegn. (Capture from DW company on the commercial or private side Th e ongoing consolidation of the Tigray 14 November 2020) is Dimtsi Weyane Tigray (DW), owned by media is a textbook example of political the Endowment Fund for the Rehabilitation parallelism in the media sector. Political of Tigray (EFFORT). It was born out of the parallelism points to the extent to which anti-Derg struggle in the 1980s and is still the media in a society refl ects political de- popularly dubbed ‘Th e voice of TPLF’. Th e velopments and the party landscape (Hal- company has many similarities with TMMA lin and Mancini, 2004). Previous research – it is based in Mekelle; it has a small bran- has detected high degrees of political pa- ch offi ce in Addis Ababa; it is operational in rallelism in the Ethiopian media landscape 29 during EPRDF in terms of ideological lin- main rationale given for the initiatives is kages (Menychle, 2017, 2019). Th e current professional enhancement, but the need for developments in Tigray are a manifestation unity along ethnic lines is quite perceptible. The current de- of continued political parallelism where eth- To this end, it is striking that when the velopments in no-national commitment overrides owners- Oromia Journalists Association (OJA) was Tigray are a mani- hip structures. about to be established, founding members festation of con- Th e perception of both TMMA and DW proposed to name the organization ‘Oromo’ tinued political journalists, and certainly journalists from Journalists Association, presumingly becau- parallelism where most other media companies in the region se it would gather members of one specifi c ethno-national would be of the same opinion, is that they ethnicity. Today the perception is that the commitment over- need to join forces to counter an unsolicited organization represents media workers in rides ownership new political line from what they describe as the Oromia region, regardless of ethnicity. an ‘anti-federalist’ government. TMMA and However, it is still a requirement for mem- structures. DW do compete in the daily news producti- bership that members are competent in the on for audiences and attention, but the real Afaan . match is taking place on another arena, and Th e largest of the regional organizations is this is a game in which they play together on the Amhara Journalists Association (AJA), the same team against a perceived external which has 315 members, a hundred more enemy. than its counterpart in Oromia. Both associ- ations are registered with the administration Emergence of regional in their regional state. journalists associations Th e organizations arrange skills-oriented Another sign of the regionalization of Ethio- training sessions on topics such as agri- pian journalism is the establishment of re- culture, environment and confl ict repor- gional journalists associations. Th e fi rst two ting. Th ey also engage in more contentious unions were started in Amhara and Oromia activities like debates on human rights and in 2014, and their activities have intensifi ed. democratic transition. In particular areas, By 2020 there are associations in progress in it becomes clear that the organizations are the Tigray and Somali regions as well. Th e advocates for specifi c regional interests. For Reiterations of ethnic discourses in the international media

While ethnicity has been a strong backdrop for much local media interaction, it has also proven to be important in the framing of Ethiopia induced by the international media. In a study which compared Al Jazeera Online’s coverage with The Ethiopian Herald and Reporter in the fi rst eight months after Abiy Ahmed came to power, it was found that Al Jazeera at- tached ethnic attribution (‘Oromo’) to the Prime Minister in all the sampled stories (N=13) (Natinael, 2019). By contrast, the two local outlets chose ethnic atribution in only one- fourth of the stories (N=62).

For The Ethiopian Herald and Reporter, attributions such as ‘change agent’, ‘open-minded’ and ‘peacemaker’ were far more important when describing Abiy Ahmed than his ethnic identity. To some extent, this is natural because the local media do not need to explain the Prime Minister’s ethnic background in every news story. But equally important, being mo- derate channels, the two newspapers are careful not to overplay the ethnic narrative. For the international media on the other hand, ethnic tensions appear to be the major frame which can explain the challenges of contemporary Ethiopia.

30 This way, the current description of Ethiopia as a hotbed of internal dispute is adopted and reinforced by the global media as well. example, one of the missions of the Tigray association was not there primarily to report Journalists Association (TJA; not yet formal- on the event, but as a supporter of regional ly established) is to defend Tigray reporters and political interests alongside other parti- who sense exclusion. On several occasions, cipating organizations. media workers from the region have been Th e regional journalists associations at- hindered in their practice, like when journa- tracts members from both the private and lists from Dimtsi Weyane Tigray and Tigray the state media. OJA, for example, gathers Television (TMMA) were denied access to a journalists from media companies as diver- conference in Addis Ababa organized by the se as EBC (the Oromo section) and OMN. National Electoral Board. TJA interprets this Th is could indicate that members, at least as victimizing on ethnic grounds. Further- to some extent, meet on professional rath- more, the association registers that the fede- er than political grounds, but it could also The Tigray Jour- ral media in the country show little interest denote a strong desire to foster a fellowship nalists Association in events taking place in Tigray. Right aft er where the common denominator is regional registers that the Abiy came to power, the central media gave and ethnic belonging. For OJA, one of its federal media show attention to the region and broadcasted in- aim is to promote Afaan Oromo in the pu- little interest in terviews with TPLF leaders, but this is not blic discourse. Although spoken by the lar- events taking place happening anymore, according to TJA. Th e gest ethnic group in the country, Afaan Oro- in Tigray. armed confl ict which broke out in Novem- mo has not been fully developed as a media ber 2020 eff ectively put an end to any posi- language. Th is too is mentioned as one of the tive coverage of Tigray by the federal state concerns by OJA. media. Th e spread of regional associations comes Th e outcome of the felt exclusion by at a time when national journalists associa- Tigray journalists is further cohesion among tions have been idle for a number of years. the region’s media workers, and between Th e oldest of the associations, the Ethiopian the media and non-media environments, Journalists Association (EJA), celebrated its including the regional government, TPLF. 50th anniversary in March 2019, but only Illustratively, when TPLF celebrated its 45th two members and the president showed up anniversary in February 2020, TJA was in- at the event7. As such, the developments of vited to observe the events through the re- the professional associations compare with gion’s Forum for Political Parties and Civil an overall trend in the Ethiopian media: Th e Society Organizations. Th is meant that the previously strong federal state organizations

7 ‘Understanding journalist associaƟ ons in Ethiopia’, Mersa Media InsƟ tute. Available at: hƩ ps://mersa- media.org/chapter-one/ Th e government-loyal Ethiopian Journalists Association is struggling while new regional associations are being formed. (Photo by Terje Skjerdal)

31 EBC journalists feel that their private counterpart are struggling, while the regional organizati- on than before. Th e broadcaster no longer FBC is better resourced ons are in a state of growth. censors stories or events on orders from the to meet the challenges of political leadership. However, its pro-go- the new media situation. A weakened position for vernment standpoint is perceptible. For (Photo by Terje Skjerdal) the federal state media example, echoing the government’s attitu- Th e federal media occupy a gradually shrin- de, EBC chose to largely ignore the regio- king position in the overall Ethiopian media nal elections in Tigray in September 2020. environment. When referring to the federal Th e channel portrays itself as fulfi lling a state media, we include fi rst and foremost moderate position in a polarized media en- Ethiopian Broadcasting Corporation (EBC), vironment. Th e state broadcaster still has EBC chose to large- in addition to Ethiopian News Agency diff erent language departments (Amharic, ly ignore the regi- (ENA) and newspapers published by Ethio- Afaan Oromo, Tingrinya, Somali, Afar and onal elections in pian Press Agency (Addis Zemen, al-Alem, English), but its intention is to downplay Tigray in Septem- Bariisaa and Th e Ethiopian Herald). Around rather than accentuate ethnic diff erences. ber 2020. 2010, Ethiopian Radio and Ethiopian Tele- Th e general tendency, nevertheless, is that vision (since 2015: EBC) were the most con- EBC is losing its previously dominant posi- sumed channels in both radio and television tion in the Ethiopian media market. It loses (Ward and Selam, 2011). Recent surveys to regional competitiors and is not welcome indicate that private stations have three to in various regional cities on diff erent oc- four times as many listeners and viewers as casions. It is subordinate to other govern- the state-owned ones (Mekuria et al., 2018). ment-affi liated media companies, in parti- Th e private stations are particularly popular cular Fana Broadcasting Corporate, which in the urban areas, but they are increasingly is perceived by EBC employees as better reaching the rural population too. equipped and more journalistically ambiti- EBC continues to stay close to the go- ous than the offi cial state broadcaster. And, 32 vernment, but has a somewhat freer positi- vis-à-vis Addis Ababa’s new generation of energetic media businesses, EBC is slow channels grew exponentially and became a and outdated. Illustratively, the broadcaster major site for advancing ethnic-based poli- currently does not even have a functioning tics. As long as the Ethiopian online sphere website (November 2020). was restricted, the local media market was relatively less infl uenced by the exchanges in Conclusion – fl oating professio- the diaspora environments. Th is all changed nalism and liberal dilemmas in 2018, not only because of the unblocking Recent political developments have led to of some 200 websites, but more importantly, signifi cant changes in the Ethiopian media. because the diaspora was allowed to physi- Some of the changes could at fi rst appear cally return to the homeland. contradictory. While media policy is libera- Th e media personalities who returned lized, the media environment itself has tur- were not uniform. Some had a strong iden- ned more polarized and fragmented. Actual tity as professional media workers, others reporting is characterized by contestation of – a higher number – were activists who vie- narratives along ethnic lines. An ethno-na- wed the media as an instrument for political tionalistic narrative emerging from a feeling achievements. A media ‘mogul’ like Jawar of suppression and exclusion contrasts a Mohammed has never claimed to do pro- unitarist narrative built on a conviction of fessional journalism, but is still behind one a common Ethiopian identity. Th e diff erent of the most infl uential media ventures in channels function as extensions of political Ethiopia today. Characteristically, OMN has positions, testifying to political parallelism built its following through the social media in the Ethiopian media. Even more prevalent and in online channels where the diff eren- than political parallelism, one might suggest tiation between professionals and amateurs that the current Ethiopian media situation is is de-emphasized. At the same time, some of characterized by ethnic parallelism. Th is is the diaspora notables are among the stron- underscored by the fact that diff erent media gest advocates of free speech and related Many newsrooms channels show commitment to a certain re- liberal principles. A case in point, again, is emerge as professi- gion or group of people. Many media houses Jawar Mohammed, who in the research in- onal monocultures and newsrooms emerge as professional mo- terview for this study tells of an incident where one identity nocultures where one identity or ethnicity is where OMN apparently had to educate Et- or ethnicity is dominant over others. hiopian Broadcasting Authority about the dominant over Furthermore, the new media situation is diff erence between defamation against a others. characterized by fl oatingprofessio nalism. higher government offi cial and a regular ci- Several of the most famous media personali- tizen. “Th ey [EBA] were so fascinated that ties in Ethiopia have an indeterminate relati- they came to our offi ce to learn more about onship with the journalistic profession. Th ey how to handle these issues,” Jawar claims. vacillate between diff erent stances – betwe- Th e details of the particular incident are not en activist, political and journalistic roles. important here, but the episode illustrates A number of these are ex-diaspora charac- how some of the diaspora returnees have ters. Historically, it is well documented that internalized a rhetoric which resonates very many journalists who left the country went well with globally recognized media free- into exile because of political sympathies dom principles. and antipathies (Skjerdal, 2011). Th ey beca- Liberal principles have, perhaps surpri- me part of a manifaceted constituency dub- singly, less to say about journalistic profes- bed the ‘confl ict-generated diaspora’ (Lyons, sionalism, apart from affi rming elementary 2007). Some had sound journalistic aspirati- standards such as the denunciation of dan- ons (Solomon, 2019), while others exploited gerous hate speech. A trained social media media opportunities abroad, taking up for activist like Jawar Mohammed knows how example engagements in language depart- far one can stretch provocative speech to ments in established media organizations avoid repercussions, hence appearing as a li- such as Voice of America and Deutsche Wel- beral-minded and professional communica- le, to nurture ethnic agendas in a freer en- tor. In the current Ethiopian media sphere, vironment. In the meanwhile, social media some social media communicators appear 33 as insiders to modern media logics, while Tigray is currently the foremost example. traditional media professionals who call for In a politically paralleled media society, responsible reporting and restraints are de- the diff erent channels are particularly alert emed illiberal and out-of-date. Th is poses a when their own group is threatened. Th is re- An emerging challenge to media regulators, self-accoun- sults in a federal media faithfully defending (semi-)liberal tability bodies and media assistance agen- the military intervention in Tigray, with system, where cies alike, which all aim to foster a profes- backing from otherwise competing media stabilizing sional and responsible media environment, institutions such as the regional media in mechanisms are while at the same time realizing that calling the Amhara region and the Oromo-aligned not in place, is a for any measure which would be perceived media, though for a diff erent set of reasons. fertilizer for one- as a restriction in the media sphere could On the other edge the Tigray-based media sided and radical easily be mistaken for a return to coercion turn all the more loyal to TPLF and aggres- expressions. and control. Th is is, perhaps, some of the sive towards the federal government. Th e reason for EBA’s reluctance to speed up the hostility was only further fuelled when the licencing process and policy development. government decided to shut down all digital Th e Authority is entitled to follow up media and telephone communication in the Tigray liberalization to fulfi l the Prime Minister region in November 2020. For veterans in ambitions, but if licences are issued and an the Tigray media, the situation curiously re- applicant is left out, it would appear as if me- sembles the conditions under which clande- dia liberalization is in reverse. stine TPLF radio broadcasts emerged in the Th e liberal dilemma is further accentuat- 1980s. ed by the fact that a progressive policy does Another condition for increased polari- not restrict media activity on the basis of zation stems from the liberal media logics content orientation and advocacy alone. In discussed above. An emerging (semi-)libe- other words, the circumstance that a media ral system, where stabilizing mechanisms channel chooses to focus on one ethnicity are not in place, is a fertilizer for one-sided and advocates for a specifi c region is not dis- and radical expressions. Th e diff erent chan- qualifying according to liberal media philo- nels serve to aggravate each other, resulting sophy, especially since pluralism is in itself in further polarization. A third condition an ideal. A functioning liberal media en- for increased polarization is the steadily vironment, however, requires balance bet- growing use of social media in Ethiopia, ween diff erent channels and support mecha- devoid of editorial guidelines and respons- nisms where self-accountability is key. Th e bility. Lastly, national elections, which are latter is not in place in Ethiopia, although a planned for May/June 2021, represent typi- promising media council has been establis- cal events which exacerbate ethnic diff eren- hed (formally registered 2019); a code of ces (Stremlau and Price, 2009). conduct is formulated; and an editors’ guild Along this argument, it should not come has been formed. Many questions remain as a surprise if Ethiopia slides back on the unanswered towards genuine self-accounta- next press freedom index, to be publicized bility, such as membership criteria, whether in April 2021. Th e question is rather how to open up for complaints from the public, long it will take before the local media will and how to institute sanctions. consolidate their activities as an industry In the meanwhile, polarization among not mirrored by hostility and divisionism. media channels is likely to grow stronger. Central to this development is professional Diff erent developments point in this dire- enhancement and a reliable self-accountabi- ction. One condition is the increase of con- lity system. fl ict in the regions, of which the clashes in

34 A framing analysis of stories in selected media channels

By Mulatu Alemayehu Moges

Th is part of the report focuses on how ethnicity is used in media texts to promote or protect a certain ethnic group, political ideology or political narrative.

Since the coming of EPRDF into power in untry should focus on politics, politicians 1991, ethnicity has been at the center of and their ideological diff erences. For in- Ethiopian politics, government structures, stance, Van Dijk (1998, p. 264) emphasizes and even the social system. Most of the Et- that the political ideology and the political hiopian political groups are divided ethn- discourse in a given country are processed o-linguistically and deal with issues in line through texts. Th e text cannot be produced with ethnic interests (Abbink, 2011; Bekalu, without political ideology. According to Van 2017; Dereje, 2018). However, the interplay Dijk, it is important to understand the per- between media discourse and ethnic iden- ceived communication context of the text tity in Ethiopia has not been comprehensi- and the discourse (ideas) in the stories. It vely studied. Hence, it is relevant to assess is therefore vital to put a lens on the texts/ how ethnicity shapes the media discourse stories to assess the link between media and about social and political issues. For this ethnicity in Ethiopia. purpose, the study dominantly applies a One can observe diff erent opinions/stan- qualitative framing analysis through textual ces and refl ections based on people’s politi- analysis. Textual analysis can help scrutini- cal or ethnic affi liation when sensitive politi- zing the text for intertextual and presuppo- cal and economic issues arise in the country. sitions, and analyze media stories from the Th ese diff erences and taking sides with re- current context of the country (politically, spect to the aforementioned affi liations historical ly and culturally) (Fü rsich, 2009; are refl ected in the media which are used McKee, 2003; Frey, Botan and Kreps, 2000). by political parties or groups to get across To show the overall trend of the media in their stances and political ideologies to their dealing with cases of ethnicity, a simple qu- supporters as well as their opponents. Along antitative data analysis is used. this line, Van Dijk (1998) underscores that By means of method, the study fi rst pur- ideological struggle in the media can be re- posely identifi ed the most sensitive cases fl ected during the rise of serious and con- with political implications. Th e selected ca- troversial confl ict cases. us,Th the study ses are believed to be at the center for dif- has purposely selected critical cases that ferences among Ethiopian political groups recently happened in the country. Among and their affi liates. Scholars propose that the many cases that demonstrate ideologi- studying media and ethnicity in a given co- cal diff erences between political elites and 35 The researched cases

Abducted students: in connection with the crises in universities in Amhara and Oromia regions, some 21 students, 17 of whom were female, of Dembi Dollo univer- sity (Oromia region) were reported missing in December 2019 while heading to their families. Th is created an outcry of the Amhara people who raised concerns about the whereabouts and security of the abducted students, all of whom were Amhara.

Clashes between Gondar Amhara and Qimant: Th e Qimant com- munity resides in Central Gondar zone, Amhara region. By setting a committee, the community has raised a self-administration demand. Th ough self-administration was settled by a referendum in 2017, the situation has aggravated and people have lost their lives and property. It has also become one of the causes of the political dis- pute between the Tigray and Amhara regional states1. At the end of September 2019, a confl ict arose between the Qimant and Amhara in Central Gondar.

Welkait case: Th is case is another cause for the rift between the Amhara and Tigray regional states. Th e Welkait Identity and Self-Administration Committee claims that the Welkait belong to the Amhara region, as people in the area are mainly Amhara. However, the Tigray regional state administers the area as one of its zones, claiming that the demographic composition is largely Tigray. Others maintain that Tigray dominance has been achieved through social engineering by depopulating the Amhara who originally lived there. Th e controversy over Welkait and its historical ownership is an ongoing tension between the Amhara and the Tigray regional states.

Irreecha case: Irreecha, which is a ritual ceremony of the Oromo community, has traditionally been celebrated in Bishoft u, 40 kilometers away from Addis Ababa. In 2019, it was celebrated in Addis Ababa for the fi rst time in 150 years. eTh celebration and the accompanying speeches unearthed serious political diff erences.

Jawar’s message on Facebook: In October 2019 the politician and activist Jawar Mohammed posted a text message on his Facebook page that called for support to take steps to protect what he claims is “a siege of the government security forces”. Subsequent to his call for help, his supporters went out to the street and took action. Th e government confi rmed that 87 people were killed in connection to his call, and the case was followed by heightened political disputes.

June 22, 2019 assassinations: Th is incident has been described as one of the most serious ones since Dr. Abiy Ahmed came to power. Four top offi cials, including the regional president of the Amhara region and the Chief of the Army as well as a retired general, were assassinated. Th e circumstances of the case are still not clear.

Sidama self-administration: Local youth describing themselves as ‘Ejeto’ went out to the street to claim Sidama as a regional state by demanding a referendum before the day 11.11.11 in the Ethiopian calendar (July 19, 2019). Unless the federal government conceded, the people said they would claim that Sidama is a regional state. As a result, people were killed, displaced and property was distracted.

Establishment of the Prosperity Party: In December 2019, the former EPRDF changed its structure, ideology, and included other regional parties as core 36 members of the party. Th e restructuring has fueled many political controversies.

1 hƩ ps://www.ethiopia-insight.com/2018/12/16/violent-qemant-dispute-fueling-explosive-amha- ra-Ɵ gray-divide/ activists, eight were chosen for analysis (see outlet which attracts many people. Today, box). sensitive political issues are narrated and All cases were covered and contemplated discussed on television channels owned by by diff erent parts of the media, both in the diverse groups. Th e other medium which is social media and by the professional media. studied is the newspaper. Th ough the num- For the sake of research coherence, this stu- ber of newspapers published in the country dy is limited to the established (professio- is limited, they are infl uential and continue nal) media. Th ese channels, although some to set agendas in the public debate. Altoget- of them were established quite recently in her ten broadcast and print media outlets Ethiopia, represent a continuation of the af- were identifi ed for analysis (see box below). fi liations between media and politics in the Th ese media outlets represent diversity in country (Menychle, 2019). Th ey operate terms of language (Amharic, Oromiff a, and under editorial decision-making principles Tigrigna); they represent both regional and and traditionally play a gate-keeping role federal media outlets; they are both priva- in the public deliberation. To consider their tely owned and state-owned; and they deal ownership and target audience is therefo- with diverse political inclinations in relation re of great interest. Th e researcher is aware to ethnic identity. Most of the outlets have of the signifi cant role of non-professional vibrant online editions too. Hence, although channels these days, but a comprehensive the studied content is derived from televisi- discussion of the wide range of such out- on broadcasts and print publications, most lets in addition to the professional channels of the content is distributed online as well. would run the danger of becoming superfi - Th e last step in the sampling was to deter- cial since the fl ows and logics of the social mine the stories. While each medium has media are rather diff erent from those of the a variety of genres, the focus here is news. professionally edited media. However, the- News represents the interests of both sour- re is a growing body of research on the role ces, reporters and owners, and it has poten- of the social media in the Ethiopian setting tially high infl uence on audiences. Th e re- which could complement the fi ndings in this searcher reviewed what the selected media research (e.g. Skjerdal and Sintayehu, 2020; reported about the chosen cases during the Téwodros, 2020). period when the incidents occurred. Th e For this framing analysis, media selecti- stories related to the identifi ed cases which on focused on two major media platforms: were reported during the chosen periods television channels and newspapers. Tele- were purposively collected for review. In the vision is prioritized owing to the fact that search of the cases, 133 stories from the se- it plays a dominant role in the Ethiopian lected media outlets were reviewed: DW (19 political discourse. Th is is related to the stories); Tigray TV (9); Amhara TV (17); current expansion of television in terms of Asrat (26); OBN (10); OMN (12); EBC (11); ownership, number of channels and increase Addis Zemen (12); Ethiopis (8); and Berera in program production. It is also a powerful (9). Since the study applies a qualitative ap-

The studied media outlets Broadcast media Newspapers Oromia Broadcasting Network (OBN) Addis Zemen Oromo Media Network (OMN) Ethiopis Dimtse Weyane Tigray (DWET) Berera Tigray TV (TTV, belongs to TMMA) Amhara TV (ATV, belongs to AMMA) Amhara Satellite Radio and Television (Asrat) Ethiopian Broadcasting Corporation (EBC) 37 proach and sampling is limited to specifi c Furthermore, the study investigated how incidents, the number of stories picked from sources have been used to report the selected each medium is not necessarily equal. Th e cases. It was found that for the 133 stories, stories were traced out for the qualitative the selected media used a total of 145 sour- analysis. Th e unit of analysis was the story ces of information. 101 of these were politi- and its components (words, phrases, senten- cal party members (politicians and activists) ces, paragraphs and images). and party leaders of the opposition and ruling party. Only fi ve sources were univer- Analytical approach sity offi cials or experts, presented without Th is study is exploratory in nature. Howe- a particular political affi liation. Th e rest of ver, with an interest in showing the scope the sources (39) were other groups of peo- of the stories and their nature, the study ple, such as university staff and students. In includes brief quantitative information. Th e other words, the majority of the sources in quantitative data include how many stories the ethnic-based stories were political party were reviewed, the size of the stories, place- members and political activists. ment, sources, and tone of the stories. Th ese One important point that was also iden- variables can help the readers get the overall tifi ed in the quantitative assessment is that picture of the stories analyzed in the study. the majority of the stories were reported While the lead researcher, as well as two from a single source. Th is follows from the research assistants who are native speakers fact that 145 sources were quoted to report of Tigrigna and Afaan Oromo, used a tally 133 stories. Diverse ideas did not have much method to see the comprehensive trends place. Obviously, the sources of the stories of media and ethnicity, the following ma- are a major determining element that can jor trends were identifi ed. eTh fi rst one re- shape the story’s message as well as its fra- gards the size of the stories. Among the 133 me. stories, 102 were lengthy, which is defi ned as As noted earlier, the approach of this stu- either more than three minutes on television dy is to analyze the text by using framing or more than 300 words in a newspaper. 15 and othering theories. For instance, framing stories were medium in size. Th ey were bet- theory, which is vastly used in analyzing ween one minute and ten seconds to three texts, could help the study in structuring the minutes on television, and 150–300 words meaning of the text by taking into account in the newspaper. Th e rest of the stories were sources, texts, audiences, and cultures, as short, i.e. less than one minute and ten se- framing can be produced in all these aspe- conds, or less than 150 words. By its nature, cts (De Vreese, 2005). Also, framing is an lengthy stories have the chance to deal with understanding of the perceived reality of a issues in detail. According to agenda-setting text (Entman, 1993). Similarly, the theory theory, media that report cases in length are of otherness (othering) is an important fra- likely to set an agenda for those particular mework to assess the interplay of media and cases (McCombs and Shaw, 1972). ethnicity. In the theory of otherness, one can Similarly, placement of the stories is also focus on highlighting how the media play a an indication of the tendency of the media role in the creation of otherness, which con- towards ethnicity or political identifi cation. sists of two diff erent groups – ‘Us’ vs. ‘Th em’ In the tally, stories were fi rst classifi ed into (Staszak, 2008; Said, 1978). Using otherness three categories according to placement. as a backdrop could help the study to capi- Th e fi rst three stories in a broadcast were talize on how the media categorize ideas in a classifi ed as prime; stories four to seven as form of ‘Us’ and ‘Th em’, how the media use medium; and the rest of the stories as ‘nor- sources to strengthen their ideas, and how mal’. For newspapers, stories were classifi ed meaning and identity are constructed. as front-page, inside pages, or back page. In assessing the dichotomy of ‘Us’ and Of the total 133 stories, 93 were classifi ed ‘Th em’ and the overall framing of the stories as prime stories, and 21 as medium stories. posted in the sampled media outlets, the Th is indicates that the chosen media outlets study looked at four important parts of the 38 give emphasis to stories with ethnic content. text. Entman (1993) proposes that where framing is processed, the narration tends to show how the media isolate, hide and pro- focus on how the problem is reported, how crastinate to conceal either negative sides the causes are identifi ed, what moral judge- of its own ethnic group or positive achie- ment is given, and which solutions and re- vements of another. Annihilation frames medies are suggested. Th ese four framing are expressed through diff erent techniques, approaches are important for the study by such as silence, blurring, slanting, charming, systematically discussing ‘Us’ and ‘Th em’ in and non-ethnic motives. particular, and media and ethnicity in gene- ral. Also, the ‘ideological square’ of Van Dijk A. Silence (omission) (1998) has been used to analyze the media Silence is one of the techniques that the me- texts in defi ning or categorizing ‘Us’ and dia can use to ignore important aspects of an ‘Th em’. Th e four elements of the ideological ethnic or political group which is perceived square are: express/emphasize information as an enemy. Silence is also used to protect that is positive about ‘Us’; express/emphasi- either its own ethnic group from criticism ze that is negative about ‘Th em’; suppress/ or ignore the other’s positive achievements. While action was de-emphasize information that is positive Media such as OBN, DW and Tigray Tele- taken in the Am- about ‘Th em’; and, suppress/de-emphasize vision are examples of this in the case of the hara region to put information that is negative about ‘Us’ (Van abducted Amhara students in Oromia. Whi- pressure on the go- Dijk, 1998, p. 268). le action was taken in the Amhara region to vernment to work put pressure on the government to work on on the release of Data analysis the release of the abducted students, these the abducted stu- By using these conceptual tools, the analysis media channels (OBN, DW and Tigray TV) dents, OBN, DWET focuses on how ethnicity infl uences positi- did not report the case. Similarly, previewing and Tigray TV did ve or negative framing of the stories in the its news channel, EBC was hesitant to report not report the case. media. As mentioned earlier, important and the case too. controversial cases were identifi ed. eseTh One can argue that these media should not cases can be a departure point among poli- be criticized for not reporting this case since tical and ethnic groups in the country. Diff e- the news selection criteria might not meet rent media channels will have diff erent ways the interest of a particular station. But the of articulations. A thorough analysis of texts case is not a simple issue that can be easily can help us understand how the media fra- ignored. As abduction of female students is me the cases in the current complex political a new experience in Ethiopia, it attracts both and ethnic narratives. Th e study attempts the national and the international media9. to zoom in and out on these selected texts Both federal and regional media channels to look deeply into how the ‘Us’ and ‘Th em’ such as EBC, OBN, DW and Tigray TV were dichotomies of ethnicity are refl ected. By expected to report the case. However, they using these approaches, two major frames shied away from it. were inductively identifi ed. eseTh are anni- For OBN in particular, one could argue hilation and othering frames. that the case should have acquired better co- verage. Th e abduction happened in the Oro- I. ANNIHILATION FRAMING mia region which is the core area for OBN, Th e description annihilation frame8 is meaning that it would be easy for its journa- adapted from symbolic annihilation theory lists to access the information. Th e case was which seeks to explain trends of undermi- sensitive and had humanitarian aspects, but ning, underrepresentation and trivialization OBN was hesitant to report on it. of objects, originally focusing on women Taking the above-mentioned reasons, th- (Tuchman, 1978). In this study, it is used to ere seems to be a deliberate silence on the

8 This is not the fi rst Ɵ me to idenƟ fy annihilaƟ on frames in the Ethiopian media discourse, cf. Mulatu A. Moges’ (2017) study ‘Why Silence: Internal Confl ict ReporƟ ng in Ethiopian Newspapers’. For an example outside of Ethiopia, see Gavriely-Nuri and Balas (2010). 9 E.g. ‘Growing outcry in Ethiopia over abducted university students’, , 27 January 2020. Available at: hƩ ps://www.voanews.com/africa/growing-outcry-ethiopia-over-abducted-university-stu- 39 dents issue. Th e exclusion of the case to be aired yield any stories related to the confl ict bet- on OBN, DW and Tigray Television can ween the Qimant and the Gondar Amhara be reasoned out at least in two ways. First, people during the course of the incident. OBN, which is a state-owned medium, may Th us, silence, a known technique in anni- not have an interest in bringing forth an is- hilation framing, is detected in the selected sue which could create a rift between ethnic media. It can serve to minimize criticism of groups (Amhara and Oromo). Journalists of the particular media’s ‘own’ ethnic group, or OBN may ignore the case to minimize ethn- to de-emphasize the positive aspects of ot- ic extremism. Th e second reason is directly hers. related to ethnic entities. As the abducted students were originally from Amhara, the B. Blurring issue may not be a priority for OBN, DW, Unlike silence, blurring is a form of making and Tigray TV. the sensitive parts of an issue vague and in- Another case of silence was observed in stead deal with other aspects that are related Media such as some media channels in relation to the June to the main case. Th is is particularly done to Asrat and Amhara 22, 2019 assassination of offi cials. For in- defi ne the cause and suggest a solution to the Television, which stance, DW and Tigray Television were not problem. For instance, blurring is observed are mostly repor- much interested in broadcasting the case of in the case of Qimant. Media such as Asrat ting stories from the killing of the Amhara regional offi cials and Amhara Television, which are mostly Amhara, tend to in Bahir Dar. While giving much emphasis reporting stories from the Amhara region, report the case of to the two top military offi cials originally tend to report the case of the Qimant in a Qimant in a from the Tigray region, a scrutiny of the two blurring way. A story reported on October blurring way. stations indicates that they did not pay tri- 18, 201910 on Asrat television, and a story bute to the Amhara offi cials who were killed reported on October 22, 201911 on Amha- in the same incident. One can link the silen- ra Television attempts to present the case of ce with the channels’ ethnic affi liation. Qimant from a diff erent angle. Th e former Th e case of the Qimant confl ict is anot- deals with the arrest of 32 suspected youth her example of media silence. Th e clashes who were traveling to Humera, Tigray regi- between Qimant and Amhara people in the on. In the story, an offi cial notes that these fi rst week of October 2019 in Central Gon- individuals who have more than one iden- dar zone claimed the lives of many people tifi cation card were going to Humera, and on both sides. Journalistically speaking, the police suspected them to be travelling to ob- issue was defi nitely worth reporting. Howe- tain military training by TPLF, a perceived ver, EBC was silent. As a national medium, enemy, to cripple the Amhara region, par- it can be reasoned out that the station has ticularly Central Gondar zone. Th e second many issues which can be more newsworthy story deals with the arrest of cattle which for the channel than this one. Nevertheless, were stolen by some groups during the con- taking the sensitivity of the issue, and the fl ict. By quoting a military offi cer who is in number of people aff ected by this confl ict, charge of security in the area, the story no- including the death toll, by any criteria, the tes that people were in confl ict in connecti- case would be a priority for the station. Un- on with the Qimant and the Amhara. But like other media, such as OMN, DW and the story does not address the questions of Tigray Television, which arguably reported where the confl ict occurred, what the cau- the case with exaggeration, EBC shied away. ses were, and what the consequences were. It may be reasoned that the station prefer- Th e story only focuses on the commitment red to refrain from intervening in a complex of the security forces in protecting people’s ethnic case. As a result, the case was ignored properties from theft . by the station. Th e same was true for Addis In a similar manner, Addis Zemen, EBC, Zemen. A review of stories in the state-run and Amhara Television did not mention the newspapers during this period does not chaos in connection with the case of Jawar

40 10 hƩ ps://www.youtube.com/watch?v=XXy46m_tJ0M 11 hƩ ps://www.youtube.com/watch?v=Ō hxOqdvQ9A in October 2019, and the subsequent killings and destruction of property in the Oromia region. Th ese media preferred to focus on the police report which was shallow and wanted to conceal the worst consequences of the incident. Th e police report describes the incident in a blurring manner. For instance, it was clear that roads were closed, and peo- ple were atrociously killed and beaten by the gangs who identify themselves as , literally meaning ‘youth’. But the report did not focus on that crisis, except calling people to calm down. Th e technique of blurring, as well as silen- cing, are used mainly to hide negative aspe- cts of a story. Th is is, according to Van Dijk (1998), a technique employed by the media has a tendency to slant the case to a diff erent Th e Prime Minister’s Press to suppress negative information of ‘Us’. Th e theme. By using sources who are politically Secretary Nigusu Tilahun same is true for the sampled media in that affi liated with the party (TPLF), the stories is being interviewed by they attempted to either conceal or oblique maintain that the people in Welkait are pure- EBC on June 22, 2019 in the negative sides of their own ethnic group. ly Tigres and want to stay in the government relation to the assassinati- OBN, DW and Tigray Television in silen- structure of the Tigray regional state. It se- on of the Amhara regional cing the story of the abducted students, and ems that the two media attempt to deviate state president and other Amhara Television and Asrat in blurring the the attention of the audiences by reporting top offi cials in the Amha- Qimant case, as well as Ethiopian Television political tension as fabricated. Th e stories ra region. and Amhara Television in blurring the Jawar report that the current tension in the area is case are typically not hiding the negative si- not the interest of the Welkait people, but in des of their own ethnic groups’ interest but the interest of the federal and the Amhara also undermining the situation or the people regional governments. of the other sides. In fact, in the case of EBC In a similar way, a story reported on Octo- and Addis Zemen, the suppression may not ber 20, 2019 on DW refl ects a deviation be directly related to the dichotomy of ‘Us’ from the main issue. While 87 people were and ‘Th em’ based on ethnic identity. Th is confi rmed dead and property was damaged can be related to either the interests of the in the Oromia regional state and stability in media in maintaining the tension or fear of the capital was in question, DW resisted to inciting further tension and confl ict among report the whole scenario. Instead of repor- the communities. Instead of dealing with the ting the devastating nature of the situation, issues and bring solutions for the cases by which was claimed to be instigated by Jawar identifying causes and confl ict lines, the sta- as he called support from his group (the qe- te media prefer to silence the circumstances erroo), the story used the incident to critici- In connection with (Mulatu, 2017). ze the federal government as weak since it the Welkait case, failed to protect its people and the system. the story repor- C. Deviating from the main issue To quote a sentence from the story repor- ted by DWET and (slanting) ted on October 20, 2019, “youth in Addis Tigray Television A third technique within annihilation that Ababa and Oromia went out to demonstrate had a tendency to has been observed in the Ethiopian media against Abiy’s government”. Th e reporting of slant the case to a is the attempt to divert the main causes and the above two cases is an example of how the different theme. consequences of the cases. Th e study fi nds media tend to impose a diff erent meaning of various cases that can show the deviating the issues. Instead of reporting the incidence interests of the media. A few examples from as accurately as possible, the channels gave the selected media can serve as an illustrati- much attention to the political interest of on. In connection with the Welkait case, the their own ethnic group, directing criticism story reported by DW and Tigray Television at the federal government in which their re- 41 sual, however, is that while ignoring cases of the students’ crises in the universities in the Amhara and Oromo regions, the stati- on (Tigray Television) pays more attention to the political interests of the regional state, which is under the rule of TPLF. Th ough the crises were happening far from the regional state, the media can report the stories as if they were new incidents, consisted of plenty of cases, and had human interest. EBC and Addis Zemen also used slanting techniques in the case of the assassinati- on of the offi cials in the Amhara region in June 2019. Th e two media channels reported the cases as a coup d'état. But the incidents did not fulfi l the criteria of a coup, as many commented. Th e two media however pre- A report on DWET lationship cannot be described as smooth. sented the cases in a slant manner to shift showing how Mekelle Th e two media outlets (Tigray Television the political tension in the region as well as University is a convenient and DW) also did not pay attention to the in the federal state. Th is is because the two place for students coming crises of university students in the Amhara outlets are state-owned and are supposed to from other parts of the and Oromia regions. What is important here focus on the interests of the state. country. is to note how the media attempted to arti- In summary, the slanting technique is culate the political narratives of the Tigray used by the media channels to send politi- regional state through their reports. Some of cally motivated messages that can favour their stories had a slant in the other directi- their own ethnic group. Instead of dealing on. For instance, Tigray Television systema- with the cases, identifying the true causes tically refl ected the interests of the regional and consequence of the incidents, the media state in relation to the then university crises tend to report to meet the political interest in the country. To quote an extract from the of the ethnic group they serve. story aired on October 30, 2019, D. Charming interest Th e federal government is not working Another important technique used in the in ensuring peace and making the uni- annihilation frame is charming the oth- versities safe places for all students who er. Th is is observed as a form of deviation are coming from the federal states.12 from the main case but used to show how the groups are smart and best for others. Th e DW and Tigray Th is quote implies that this medium has charming interest of the media can be ob- Television tend to extra interests. For instance, the phrase “all served for example in the case of students’ portray the Tigray students who are coming from the federal crises in the universities. It is clear that uni- region as a heaven states” was used to undermine the legiti- versities in the country, particularly in the for all students mate power of the current government in Amhara and Oromia regions, have been in from other regions serving the federal structure. Th e regional crisis. Th e teaching-learning process has not by highlighting state of Tigray frequently complains that been smooth, and students have not been at attractive events Abiy’s government is tending to destroy or peace. Th ese cases were reported by parts of from the universiti- demolish the federal states that TPLF claims the media. DW and Tigray Television, ho- es and the region. it has been fi ghting for a longer period. Such wever, gave a diff erent dimension. Th ey did kinds of political interests were refl ected not report the whole crisis, and tended to in the story. One cannot deny that the me- portray the Tigray region as a heaven for all dia oft en refl ect the political interest of the students from other regions by highlighting owners. What makes this reporting unu- attractive events from the universities and 42 12 All translaƟ ons are made by the lead researcher and research assistants. the region. For instance, Tigray Television reported about University students who visited tourist attractions in the area. A similar report on the same channel showed how students of Mekelle University were go- ing to the fi eld to help farmers in harvesting the crops. DW reported a story with quite similar themes. On November 10, 2019, the chan- nel reports a story about a fruitful discussion between religious leaders and elders of Raya and students of Raya University13. It appears to be a good attempt at a discussion with the students in the time of crisis. Another story shows how Tigray region is a welco- ming place for studying by inviting students who come from other parts of the country14. Th e interviewed students say, “Mekelle is a Tigray born students placed in Deber- DWET is still regarded as peaceful place for the students. Th e town tabor University stated that since they a strong affi liate of TPLF is not like we heard before we came here.” know the honesty and hospitality of the and covered the 45th Although it is fair to report on those events, Amhara people they dare to come here anniversary of the party the airing comes while there is serious cri- to study. Students noted that what they duly in February 2020. sis at other universities in the Amhara and have heard and what they have been Th e screenshot above is Oromia regions. Th e intention of the stories experiencing in the Amhara region are from the celebrations in seems to be to show how the regional state quite diff erent. Welkait. is peaceful and working in bringing mutu- al understanding among the students in the Similarly, stories on Amhara Television university. It is arguably an attempt to charm show that universities inside the region are others by portraying their own ethnic group safe. For instance, one story reports how as kindhearted. the Tigray students in Bahir Dar University Th e same channel (DW) appealed to are well treated in the university. Th e jour- charm when reporting TPLF’s 45 years of nalist interviews a couple of students who establishment at Welkait. By quoting offi cials are originally from the Tigray region. Th e from the regional state, the channel wanted story quotes them as saying, “we are conti- to show how the regional ruling party pays nuing our studies is a very smooth manner. much attention to the people of Welkait, and We live in harmony with other students who celebrating the day with “its people” is one of come from diff erent parts of the country”16. the signs. Th e story seemingly did not want Another student says, “I think TPLF used to mention the clear cause of the political this opportunity to divide the two peoples. tension between Tigray and Amhara over We are now in a safe place where many pe- Welkait. ople live together peacefully. In a dorm, we More examples of reports that can show are eight and we live together peacefully that “my ethnic group or region is best” and harmoniously”17. Th e two stations tend might be derived from Asrat and Amha- to not only counterbalance the warning of ra Television. A story reported by Asrat on Tigray education bureau claiming that all October 11, 201915, for instance, claims that universities in Amhara regional state are the Amhara region is the best place for all: danger zones for the people who are coming from Tigray region but also want to show

13 hƩ ps://www.youtube.com/watch?v=ILIjzU1suu 14 hƩ ps://www.youtube.com/watch?v=zPXcqFCCHVc 15 hƩ ps://www.youtube.com/watch?v=OZXOY1mVU58 16 hƩ ps://www.youtube.com/watch?v=SjD5gSTOwNE 43 17 hƩ ps://www.youtube.com/watch?v=ONFNn4yAECk how the region is a place to live. tions the ethnic groups of the two students, Th e overall interest of Tigray Television, but simply refer to ‘our students’. Th ey des- DW, Asrat TV and Amhara Television is to cribe the students as . Th e story The overall interest give emphasis to their own ethnic group and tries to create a sense of nationality or one- of Tigray Televi- region. Pan, Isakhan and Nwokora (2020, p. ness. In fact, since EBC is a national medi- sion, DWET, Asrat 58) describe such kinds of discursive nar- um, it is not surprising that it frames stories TV and Amhara ration to construct self-positives as “charm in such a way by using words that can tone Television is to give off ence”. While there is a myriad of problems down the tensions among ethnic groups. emphasis to their happening in their regions that should get One most important thing that should be own ethnic group attention, the studied media channels tend appreciated here is that the station selected and region. to focus on selected positive aspects of rea- not only the right sources of the stories but lity to show how they are attractive. Th is is also inserted ideas in the story that are ne- on the other hand used in a way to undermi- utral and non-violent. ne others. In such a positive self, the media Unlike other media, EBC shows an inte- tend to refl ect only positive aspects of the af- rest in playing a centrist role. For instance, fi liated region and their ethnic group in the in stating the consequence of the problem, political narratives. the story does not refer to one ethnic group. Th e source, who is the head of the commu- E. Non-ethnic motive nication bureau of the Amhara regional Another technique applied in the annihila- state, says, “We lost two brothers, and the tion frame is the non-ethnic motive. Th is is regional state expressed its sadness at the a way that the media report the cases in a loss of the two students”. It is clear that the non-ethnic and non-contentious manner. two students who were killed in the incident In other terms, the media circumvent issues belong to the Oromo ethnic group. But the of ethnicity. For instance, EBC reported the source does not want to mention the dece- case of Woldia University which claims the ased students based on their identity. Rat- lives of two students (Oromo students) wit- her, the source as well as the story tend to hout siding to any group. Th e story reported embrace them as part of their ethnic group on November 10, 201918 attempts to defi ne and describe them as one. Th e source also the problem of the confl ict as not known. It denounces those who are try to give wrong does not seem to relate the case as ethnically perceptions of the case and incite further or politically motivated. Th e station inter- confl icts among students in the universities. views two offi cials in connection with the Th e other narrative that should be men- case: Th e Vice President of the university, tioned here is the solution suggested in the and the Head of Regional Communication story. In the stories OMN and Asrat, as ca- Aff airs Offi ce. Th e story was supported by ses in point, are expected to present soluti- up-sounds from the interviewed sources. ons on behalf of the ethnic group they serve. For instance, the vice academic president However, in EBC, the story indicates that says, the government and the police offi cers are working on investigating the case; the elders Students were in a peaceful mood the are also working in bringing reconciliation whole day. All the students were busy among the students who were in the con- fi ghting against the desert locust. Th ere fl ict. eTh story does not side with one ethnic was nothing unique until I received the group. report of the violence. Two of ‘our stu- In the case of the assassination of top dents’ have now passed away. Police are military offi cers and leaders of the Amha- investigating the cases. ra regional state, the story is reported in a non-ethnical approach by Addis Zemen and In another interview with the regional EBC. Except narrating the atrocity executed communication offi cer, the same idea is in- on those offi cials, the stories do not only link serted in the story. None of the sources men- up the case as ethnically motivated but also 44 18 hƩ ps://www.youtube.com/watch?v=R_oqeCfE5SE the deceased with their ethnic identities. II. OTHERING FRAMES Ethiopis in a report about the Sidama case Th is second type of frame is derived from in its edition on July 21, 2019 brings diverse the theory of othering. Unlike the previous ideas on the request of self-administration. frame, othering is more interested in giving Th e story describes the overall massive da- emphasis for ‘Us’ as a victim for an incident. mages in Sidama zone in connection with Th e media tend to purport that their own not getting a response by 11.11.11 (19 July group (‘Us’) has been vulnerable as well as 2019), which was the day when Sidama was victims of a certain incident. Th is frame is expected to be declared as a new regional also used to give emphasis to ‘Th em’ as the state. Th e story seems neutral and balanced cause of the problem. Th e frame uses vario- in describing the overall situation and dama- us techniques to expose intent such as expo- ge. Th e story does not connect the case with sition, externalization (blaming others), ethnic identity. It focuses on the problem by symbolic attachment, victimization, and un- describing what needs immediate attention dermining others. by all concerned bodies. All stakeholders’ ideas are refl ected in the stories. Sources qu- A. Exposition oted in the story are not only appropriate but As the term indicates, exposition is a techni- also stand out as neutral. que that a media channel uses to take certain Th e same newspaper is found to be centra- information from another outlet to expose list by reporting all round information about what the other ethnic group is doing wrong. the case of Jawar. On the story reported on In other terms, this technique is used to re- November 3, 2019, Ethiopis exposes detai- veal how the other ethnic group and their led and diverse sources of information that media are fermenting triggering messages support and are against the case. Th e news- that go against another ethnic group. So- paper thoroughly reports the case of Jawar metimes, the stories are not directly fetched and the subsequent violence in the Oromia from the original sources of information, region and the follow-up discussion carried but rather taken from the interviews or posts out by Prime Minister Abiy Ahmed and the in the other media that the sources favour. then Minister of Defense Lemma Megersa Th is can be explained with examples. in the region. Th e story seems to be fair in A story reported by DW on September 30, reporting the story and tries to counterba- 201919 states that, lance the tendency of reporting the case in a diff erent way by other media. Policy commissioner of the Amhara In the case of the establishment of PP, regional state, Abere Adamu, blames The Ethiopian me- some media, particularly the state-owned TPLF as a perpetrator of the recent cri- dia tend to hide, media, tend to refl ect centrist views. Reports sis and the loss of human beings and slant, and blur the on Amhara Television, OBN, EBC and Addis properties in connection with the Qi- most important Zemen show fair interest towards the esta- mant case in the Amhara region in the issues that could blishment of the PP. Th ey give much emp- Central Gondar area. Th e third-party hasis to the importance of the coming of PP (TPLF) is responsible for all the loss harm the image of for all the people and affi liated political par- and chaos in the Qimant area and the their ’own’ ethnic ties who have been isolated for long. people. Th e regional force did nothing group. In summing up the annihilation frame, the wrong except ensuring peace and stabi- Ethiopian media tend to hide, slant, and blur lity in the area. the most important issues that could harm the image of their ‘own’ ethnic group. For Th is is the lead of the story that was reported the most part, the media attempt to surpass by DW by extracting a quotation from an in- their weak sides or wrongdoings to show terview broadcasted on ESAT. One can say only the positive elements of the group they that DW systematically presented the story tend to serve. Also, the media have been un- so as to show how the other group (‘Th em’) dermining the pain of other ethnic groups. was thinking towards its own group (‘Us’, i.e. 45 19 hƩ ps://www.youtube.com/watch?v=LfFfSrVDdTg the Tigray from the perspective of DW). students to universities in the Amhara Th e story also tries to belittle the Amhara region as it is an attempt to dissociate regional state by showing how it is not wor- the two peoples [Tigray and Amhara], king to solve its internal problem by iden- and it is not a welcoming idea by all. tifying the actual cause of the confl ict. eTh Th e Tigray people should understand report quotes an extract from the ESAT20 that such a decision of the regional go- interview with the Head of the Police vernment is a strategy to pose confl ict Commission of the Amhara regional state, and distance its people with fellow brot- Abere Adamu, who says: her Amhara. It seems that Amhara Television Th is is the task of a terrorist group like It here seems that Amhara Television strate- strategically wants TPLF. Th is group provides training for gically wants to reveal how the Tigray regi- to reveal how the the group organized by the name of Qi- onal state does something unpleasant to the Tigray regional mant and armed them to dismantle the Amhara people. Th e story also extracts a qu- state is doing region. Th is is almost a proxy war of ote from an interview of a Tigray Regional something TPLF to destabilize the Amhara region. State Education Bureau Head: unpleasant to the Amhara people. Th e quoted idea is not new. It is a widely Th e regional government of Tigray will known claim in the political row betwe- not send freshman [fi rst year students] en the two political groups in the regions. students who are placed in universi- However, the question here is why DW ties in the Amhara regional state. We chooses to quote ESAT and Abere Adamu. also notifi ed the federal government to Also, in the story, the journalist explicitly give a new placement for all students of uses words that label the Amhara regional Tigray who are assigned in the Amhara force as ruthless by saying, “Recent atrocity region. Th is is our stand. happened in Gondar – the Qimant people is the worst in Ethiopia”. Furthermore, the Th e story pretends to expose misdeeds by story demands the Amhara Policy Force to the regional government of Tigray and their take responsibility for all loss of people and media, particularly towards the Amhara destruction of property of the Qimant pe- ethnic group. ople. Th e reporter articulates these strong A newspaper, Ethiopis, reports a story statements without quoting any source. Th e which has a similar tone. On July 11, 2019, overall impression is that one group is por- the newspaper ran a story about Irreecha (an trayed as evil. Oromo festival) by quoting a statement of a Th ere are also other examples of media top offi cial from Oromia in a press briefi ng channels which use quotations from non-af- to OBN. Th e lead of the story is, “Shemeles fi liated media to expose other ethnic groups. Abdissa, president of Oromia regional state, For example, on October 20, 2019, Amhara wants to give political meaning to Irreecha Television aired criticism of the Tigray Edu- celebration in Addis Ababa”. Th e news story cation Bureau Head who gave an interview articulates that the Oromia regional govern- to DW and Tigray Television. Like DW abo- ment is using the celebration of Irreecha as ve, which quoted a story from ESAT, Amha- a mechanism to ensure and strengthen the ra Television took this story from DW and political interests of his government in the Tigray Television, which have a diff erent quest of Addis Ababa to Oromia. Th e story agenda than Amhara Television. Th e lead of also emphasizes that such a celebration is the story goes as such21: like adding fuel to the fi re as the issue of Ad- dis Ababa is a hot political topic. Students of Bahir Dar University and Th e same newspaper is found to blame qe- dwellers of the city denounce the Tigray erroo for the killing of a large number of pe- government’s decision on not sending ople in Oromia based on their ethnic iden-

20 There is a common percepƟ on that ESAT favours the Amhara and EthiopianizaƟ on, up against the 46 interest of TPLF and DW. 21 hƩ ps://www.youtube.com/watch?v=ONFNn4yAECk tity and religious denomination (November 2019 can be a good example. In just one day, 3, 2019). Th e story attempts to show that the four consecutive long stories about the Qi- violence and inhuman actions taken against mant crisis were reported by DW in its news other ethnic groups living for long in the re- broadcast. Intensive reporting about this gion have been motivated by the Oromia re- case in Amhara Television may not be that gional state, qeerroo, and their leader Jawar surprising. Reports with an ardent interest, Mohammed. Unlike OBN, OMN, AMMA, highly exaggerated tone and overemphasi- EBC and Addis Zemen, Ethiopis is direct in zed themes that can send negative messa- identifying the causes and problems of the ges towards Amhara by DW and Tigray crisis in Oromia. Television, however, can be interpreted in a In recapping, the Ethiopian media show diff erent way. It shows an attachment of the a clear tendency to reveal antagonistic ide- media (DW and Tigray Television) with the as of their perceived enemies (‘Us’) through Qimant people who have demanded self-ad- monitoring the media favoring the other ministration in the Amhara region. As can (‘Th em’). Th e stories are presented and fra- be seen in the stories reported on September med in a way to expose a clear stand of other 30, 2019, the Tigray’s political group and the groups (‘Th em’), and to show how they are Qimant ethnic clan have a good attachment the real enemies of their own group (‘Us’). though they are neither geographically nor Th is technique shows plainly how ethnicity linguistically the same. From the 133 is played out in the media by means of em- In a news hour, four diff erent long stories stories, 109 were phasizing the negative intents and actions were reported about the case of Qimant. purely reported to of others. A similar result is observed in the One of the stories is about Qimant students expose how others quantitative data. Most of the stories tend to demanding immediate action by all concer- attempt to attack reveal bad experiences with other ethnic or ned bodies. Th e news goes as such: their own group political groups. From the 133 stories, 109 either physically or were purely reported to expose how others Mekelle University students belonging politically. attempt to attack their own group either phy- to the Qimant ethnic group call to all sically or politically. 101 of the stories expo- the people and organizations who have sed how their own ethnic or political group concern on humanities to protect the could be at risk. In other terms, the stories Qimant people from torture and ha- tend to narrate their groups as victims of rassment triggered by the Amhara go- the incidents, which will be presented in the vernment. next section. Several issues could be raised in relation to B. Symbolic attachment the fact that Tigray affi liated media showed (tactical alliance/solidarity) such concern in reporting the case of Qi- A second technique observed in the ot- mant. Firstly, it is a question why the lead hering frame (Us and Th em dichotomy) story is framed in such a way that it accuses is symbolic attachment. Th is technique is the Amhara regional state. Secondly, one is used to show how the selected media tend curious to know why the Qimant students to interlink with ethnic groups either in em- at Mekelle University approached only the phasizing or de-emphasizing their interest Tigray media. And fi nally, one could qu- in reporting cases of other ethnic groups. A estion why ethnically Qimant students who good example of this technique is the case of study in other universities of the country the Qimant in the Amhara regional state. As did not approach media organizations to noted, the quest for self-administration of complain about the case. Th e indication the Qimant in the Amhara region has been is that the coverage contains a secondary the cause for the death of many people in the agenda and that there exists an extended re- area and at the center of the political ten- lationship between the media in the Tigray sion not only in the Amhara regional state region and the Qimant people. Th ere are but also in the federal government. An inci- claims that DW is mostly serving the Tigray dent that happened at the end of September political elites, specifi cally TPLF. As can be 47 TPLF. Such kinds of strategic alliance with dif- ferent people who do not directly belong to their ethnic group can be observed else- where as well. For instance, OMN, DW and Tigray Television were found to be the main supporters of the Sidama self-regional admi- nistration demand. Also telling is DW and Tigray Television’s attitude towards Jawar’s call for action in Oromia and the subsequent violence in the region. Tigray Television re- ported the case of Jawar as a weak adminis- trative strategy of the federal government, which is both showing the station’s political stance against the federal Abiy government and an expression of symbolic solidarity with Jawar who is a political fi gure with a Amhara TV gave attention seen in the recent political narratives, TPLF large following in the Oromia region24 and a to the case of the abducted and the ADP are in diff erent political blocs22. strong opponent of the federal government. students at Dembi Dollo Th ey have recently been in a political row23. In short, symbolic attachment is employed University in the Oromia Some writers (e.g. Semir, 2019) note that the by diff erent channels to show some sense of region and referred to them clash in the Qimant area is a proxy war bet- belonging to a certain ethnic group to attack as ‘our students’. ween the Tigray (led by TPLF) and Amhara the common opponents (the third party). As states (led by ADP). Th is might be one of the the saying goes, ‘the enemy of my enemy is reasons that DW pays much attention to the my friend’. case of Qimant. In this particular story, one can arguably see that the media in Tigray are C. Victimization (exaggeration) seizing a kind of position that fi ts with the Victimization is a technique used by the me- interests of the political elites of its region, dia to give more sense and meaning to their particularly the ruling party, TPLF, to meet OMN, DWET and ethnic group. Th is approach is refl ected in the interests of its ethnic group. Hence, the Tigray Television the studied content. For example, compared station frames the story to portray the Qi- were found to be with other media outlets, a story reported mant people as the victims of the ‘atrocious on Amhara Television on February 6, 202025 the main suppor- attack’ of the Amhara police force. shows a great interest in the case of the 21 ters of the Sidama Th e attachment between the media (DW abducted students in Dembi Dollo Univer- self-regional and Tigray Television) and the ethnic group sity in Oromia regional state. Th e station administration (Qimant) is also refl ected in other ways. brought attention to demonstrations held quest. DW seemingly reports the case of Qimant in most parts of the region. In fact, as the to expose the malfeasance of the Amhara abducted students belonged to the Amhara regional government in particular, and its ethnic group, the people of this regional sta- people in general. Th e close collaboration te feel the pain more than the other regional between TPLF and Qimant comes out as a states. Th e story shows how the Amhara pe- tactical alliance which also can be used by ople are out crying due to its students being the affi liated media to send negative messa- abducted by unknown people in Oromia. To ges that can undermine the Amhara regio- quote from the story, “Amhara students are nal state and the ruling party ADP, as they still becoming victims of the conspiracy of are not having an amicable relationship with the ethnic politics in the country”.

22 ‘APD says ‘TPLF is trying to exploit the current situaƟ on to cover years of crimes’’, Borkena, 11 July 2019. Available at: hƩ ps://borkena.com/2019/07/11/adp-says-tplf-is-trying-to-exploit-the-current-situaƟ - on-to-cover-years-of-crime/ 23 ‘Regional power grab aƩ empt causes rare discord in Ethiopia coaliƟ on’, Dawit Endeshaw, 12 July 2019. Available at: hƩ ps://www..com/arƟ cle/us-ethiopia-poliƟ cs-idUSKCN1U725Z 48 24 hƩ ps://www.youtube.com/watch?v=xwBH8RXLqSo 25 hƩ ps://www.youtube.com/watch?v=J_1L8lSXPQs Th e story describes the Amhara people cularly Amhara people living in the region. as the prime victims of the pressure by the To quote, “so far many attacks targeting the unidentifi ed groups. eTh story quotes de- Amhara ethnic group have been happening monstrators who criticize the regional party, both inside and outside of the universities in which was “ADP continues to be obedient the region by Oromo gangs in the region”. and subservient to the governing system”. Th e other medium observed to give high In stating the solution for the problem, the emphasis to the ethnic group it serves is As- story underlines that unity and solidarity rat TV. Th e station focuses on the interests among Amhara people could help to bring of the Amhara ethnic group. Th e following immediate action for the release of ‘our stu- quote from a story reported on Asrat on No- dents’. Most of the sources for this story were vember 13, 201926 is illustrative: demonstrators who staged out to the street to demand the release of the abducted Amhara Soon aft er the death of two Oromo stu- students. One important thing that should dents in Woldia University while stu- be emphasized here is that unlike the other dents were watching a football match media aired in other regions, this medium on TV, Amhara students in several tends to use the phrase ‘our students’. Two Oromia universities are facing direct extremes are observed. On the one end, as attacks and the situation is not good for Some media discussed earlier, some media outlets, par- the Amhara people. While three Amha- outlets, particularly ticularly DW, Tigray TV, OBN and OMN, ra students in Meda Welabu University DWET, Tigray TV, which refrained from reporting about the were admitted to the hospital following OBN and OMN, abducted students in the Oromia region. On an attack on them, one student in Dem- refrained from the other end, channels such as Asrat and bi Dollo University was killed. Students reporting about the Amhara Television, which gave a thorough say that they stay there with fear and abducted students. report. Even some reports criticized both stress. the regional and federal governments for their weak measures to free the abducted From the extracted lead story, one can un- students. Th e story was also used as a sign of derstand that the story was reported in fa- all the intrigue happening on Amhara peo- vour of Amhara students in universities in ple by the “federal government as well as the the Oromia regional state. Th is does not Oromia regional state”. mean that the problems were not happening Th e ethnic tendencies were refl ected in in the mentioned universities. However, in three forms. Th e fi rst one as noted, the ab- looking at the narration of the problems in duction of Amhara students in Oromia was the universities, the story paid much atten- articulated in a way that was triggered to at- tion to the Amhara students. In defi ning the tack the Amhara ethnic group. Th e second causes for the continuous tension and con- angle is that there has been a clear diff erence fl icts in Oromia Universities and attack on between the Oromo and Amhara people in Amhara students, the medium gives names the current political discourse. According to for those who have caused the problem as the report, the Amhara–Oromo diff erences they appear to deliberately incite confl ict di- were refl ected not only by abducting Am- rected at the Amhara ethnic group. Th e story hara students but also the Oromia regional not only favours one ethnic group, but also state and its people had not yet denounced identifi es certain groups, particularly poli- the case which has been happening in its tical groups, which strive to create confl ict area. Th irdly, due to the crises in many uni- among students in line with ethnic identity versities in the country, Amhara students and religious denomination. have been targeted by means of killings and Among many other reasons that have harassment, and many of them are already been refl ected in the two media channels expelled from the Oromia region. Th e stories (AMMA and Asrat) in connection with the connect many problems happening in the abduction and displacement of the Amhara Oromia region that target students, parti- people in the Oromia regions, is the claimed 49 26 hƩ ps://www.youtube.com/watch?v=qmEX45Afer0 hate of the Oromo people towards the Am- University and the subsequent crisis in the hara society. Whatever the narrative in the universities in the Amhara regional state is stories, the two channels, particularly Asrat, provoked to target the Oromo community. seem to have a clear agenda set to reveal all Th e other important point that should atrocities targeting the Amhara people. Th is be discussed here is how the story of OMN is one form of the ideological square that suggests a solution to the problem. Th e gives much emphasis on the negativity of story does not point out possible solutions them by others so as to refer to the second that could solve this particular problem and OMN did not group as the cause of the problem. bring sustainable peace and reconciliation consult sources Aft er the death of two Oromo students in among the ethnic groups in the universities from the Amhara Woldia University in November 2019, hig- in the country. Instead, the story states that region or univer- her education institutions in the Amhara Oromo students should leave the universiti- and Oromia regions were not in a peaceful es and come back ‘home’. In fact, the chan- sity offi cials who situation. Death and displacement of uni- nel here suff ers from a source problem. Th e could balance and versity students were frequently reported. story does not consult sources from the Am- tell their version Quite diff erent frames were found in the hara region or university offi cials who could as to what went stories reported by OBN (to some extent), balance and tell their version as to what went wrong. OMN, Asrat and AMMA (to some extent). wrong. Beyond the interests of the sources, As giving ‘Us’ a high priority, OBN and the story applies the term ‘home’, referring OMN were the fi rst in describing the situati- to the Oromia region. Such kinds of descrip- on as desperate for the Oromo ethnic group. tions tend to delineate each regional state or OMN, for instance, reported a story by quo- ethnic group. Also, it is clearly depicted that ting the political leader Jawar Mohammed27. the Amhara region is not a home for them (the Oromo students). In addition, the story Jawar warns the government to protect quotes the students’ point of view that the ‘all Oromo students’ who are enrolled ‘Oromia government should take measu- in universities in the Amhara region. res to solve the problem while the problem He amplifi ed that the situation of Oro- is happening in the Amhara regional state’. mo students in the universities are fear- In the ‘Us’ and ‘Th em’ dichotomy, while all ful and leading their lives at risk. universities are under the mandate of the Ministry of Science and Higher Education, In this extracted lead of the story of OMN, stories assume much responsibility on the one can understand that there have been regional government to deal with the pro- some forms of isolation of one group from blem, which is not its mandate. Th e ethnic the others and describe it as a victim of the inclination of the media is underscored by case. In addition, the story uses phrases such the fact that the journalists did not intervi- as “our people” and “our students”, which ty- ew offi cials and other neutral sources, either pically refer to students of the Oromo ethnic the regional government of the Amhara or group. It is emphasized that students of this the federal government. Th e stories get stu- particular group need protection. Th e story ck with emphasizing the problem and ham- gives much attention to one ethnic group, mering the cause, and do not come up with namely Oromo students, by alienating oth- a solution involving diverse voices from all ers. Th e story does not deal with all students concerned people. As one of the manifesta- or students as Ethiopian citizens. tions of ‘Us’, OMN only interviewed Oromo In another story on the same day, Oromo students. students from Mekedela Amba, Debretabor One important trait that OMN, OBN (to and Woldia universities express their con- some extent), Asrat and Amhara TV (to cern about the situation, which they descri- some extent) share is that they do not report be as terrible. All the sources are students the cases that happen in their areas (Oro- who belong to the Oromo ethnic group. Th e mia and Amhara, respectively). While they story narrates that the problem in Woldia give much emphasis on the ‘worst situation’ 50 27 hƩ ps://www.youtube.com/watch?v=IzBgMyRbfxc in the universities, the media, for instance, typical: Asrat in the Amhara region and OBN and ‘THNeG’ (the Amharic version of OMN in the Oromia region, do not reveal TPLF) has been fi ercely working what went wrong in their respective regions, against the Amhara and other ethnic and they are hesitant to report the main cau- groups not to be peaceful and stable. It se of the problem and the exact happenings is still working on realizing its negative in the universities in each region. Th eir in- and false narratives towards Amhara. tent is to emphasize their ethnic group as a Currently, THNeG recruits the youth to victim of the crisis. train and support the illegal committee Asrat in the Am- Th e victimization technique can also be of Qimant so as to attack the Amhara hara region and found in newspapers. In its edition on No- people and aff ect its regional develop- OBN and OMN in vember 30, 2019, Berera published a de- ment and stability. the Oromia region tailed analysis of the Jawar case supported do not reveal what by pictures. Th e story describes Jawar as Th is extract can show not only the diff erence went wrong in their a perpetrator in the recent killings of Am- in political ideology and the ethnic tension respective regions. hara and other people living in the Oromia in the region (Amhara vs. Tigray and Am- region. Th e story bluntly claims that such hara vs. Qimant), but also how the cause of kinds of incidents should be likened to the the confl ict is defi ned. Firstly, the story des- Interahamwe of Rwanda (i.e. the Hutu para- cribes the armed groups as ‘musketeers’ that military organization). Th e allusion is that simply attack people or the regional govern- people are killed because of their ethnic and ment without good political grounds. Th e religious identity. Th e newspaper categorizes groups are directly linked with the neigh- one side as the perpetrator (the Oromo) and boring Tigray regional state which is descri- the other as the victim (mostly Amhara). bed as the main cause of the confl ict in the Amhara region. Th e news on Amhara Tele- D. Blaming the other vision claims that, (externalization) Th e blame technique is used in stories to It is an attempt to create instability in push away the cause of the incident, which the region, which ultimately aff ects the perhaps is caused by the group alligned to current development activities (sesame the media in question. Th is can also be used products will be harvested in this area) to label specifi c groups as the cause of a pro- of the people in the area. blem. It is a deliberate approach to accuse a group (ethnic or political) as a cause of the Th e cause of the confl icts and the overall crisis. A few stories from the selected media tension in the Central Gondar zone of Am- can serve as examples. hara regions are not explicitly narrated in A story on Amhara Television on Octo- the story. Instead of dealing with the politi- ber 6, 201928 reported the arrest of 53 well cal interests of the Qimant group, the story trained and armed infi ltrated soldiers who blames others (THNeG) in organizing and were sent to attack the Amhara region. In sending militants into the Amhara region. the story, political and ethnic issues are re- Th e TPLF group is labeled as traitors and the fl ected. For instance, by quoting a regional mastermind of all the chaos in the Amhara police offi cer, the story highlights that the regional state, particularly in the Qimant ad- current quest and armed struggle of certain ministrative area. groups in the Qimant area are not in the ge- Th e Berera newspaper refl ects a similar nuine interest of the Qimant people. All the trend of externalizing the causes of the con- confl icts and political tension in the Central fl icts in the Amhara region and its people to Gondar zone are described as deliberately the third party. A story reported on Novem- instigated by external groups with the aim ber 30, 2019 blames TPLF not only for the of violating peace and stability in the Am- case of Qimant but also for other incidents hara regional state. Th e following extract is in Oromia. For instance, the atrocities that 51 28 hƩ ps://www.youtube.com/watch?v=mXJXK2-fB8U) ic group (‘Us’) from blame by attributing all causes to the other ethnic group (‘Th em’).

E. Undermining others While blaming the other is a used in the me- dia (our media) to criticize the other ethn- ic group (them) for either the cause or the problem of the cases that have brought de- vastating impact on their own ethnic group (us), undermining the others is a technique applied to ignore others’ good values becau- se of their ethnic diff erences. As the name indicates, undermining others is intentio- nally disregarding someone’s positive activi- ties or initiatives. Th is can be, for instance, using some terms, phrases or ideas (themes) in the story. Th is can be illustrated by a few examples from the stories. For instance, in happened in the Oromia region on the Am- Facsimile of Ethiopis from the above-mentioned story, used to examine hara ethnic group is an intrigue plotted by October 2019, showing the the blaming technique, the name ‘THNeG’ the TPLF, according to the newspaper. Th e loyalty of Oromo young was applied to describe TPLF. In itself, there story goes: men (qeerroo) to Jawar is nothing wrong about the name. However, Mohammed. the Amharic abbreviation has a negative TPLF is an intrigue maker that instiga- connotation on the audiences. Th e story se- tes clashes between Amhara and Oro- ems to use the abbreviation ‘THNeG’ to un- mo people. TPLF was successful par- dermine all the values and power of TPLF. ticularly in sending negative and false In a similar view, Ethiopis in its edition on messages that not only demonize the October 26, 2019 wrote a headline that can Amhara people but also motivate the give a diff erent meaning of a person’s name. Oromo youth to be against the Amhara It reads ‘Ja-war’ to connect the person in people. question – Jawar , an activist and later po- litician of the Oromo Federalist Congress Another example of blaming the other oc- (OFC) – as a source of war. Th us, the story curred on DW Television on October 1, blames Jawar for the atrocity in the country, 201929 in a lengthy story about the case of particularly for the confl ict that claimed the Qimant and Gondar. As noted earlier, DW lives of 87 people in the Oromia region in presents the Amhara regional state in gene- October 2019. ral and the people of Gondar in particular as Th e return of the Irreecha celebrations to Generally spea- the main cause of the problem in the area. Addis Ababa in 2019 also triggers refl ection king, DW tends to Generally speaking, the channel tends to ac- on ethnic elements. Th e stories reported on accuse the Amhara cuse the Amhara region. OBN and OMN assert that the Oromo ethn- region. In short, externalization is one of the te- ic group has become a winner over the group chniques used in the Ethiopian media to re- whom they call neft egna, which literally me- fer to other ethnic groups perceived as ene- ans armed group. While it is a religious or mies as a cause for the problem occurring cultural celebration, the media use words in the country or some particular place. Th e as well as tone to undermine other political purpose of the media is not only to send ne- and ethnic groups. Similar reports were ob- gative messages to other ethnic groups but served in the media in connection with the also to alienate that group from the others call of Jawar. Th e stories reported on OMN who have a good relationship. Th e media on October 23, 2019 frequently used the tend to frame cases in a way to free its ethn- word neft egna and the phrase ‘Down, down, 52 29 hƩ ps://www.youtube.com/watch?v=LfFfSrVDdTg neft egna’ which has a negative connotation story tends to send a message that groups for the Amhara ruling group. established based on ethnicity will be highly All in all, most of the selected media in the aff ected by the coming of PP. Th e channel study were found to use labelling to show also criticizes the incumbent leaders who either hate towards the group or to under- have been using PP to maintain their power Most of the media mine the role of the group. According to Van and sustain their existence in the leadership. in the study were Dijk (1998), using terms can bring bias in Th e extreme negative tendency towards the found to use the interpretations of the word. PP is also refl ected in the story by includi- labelling to show Th e undermining technique has been ng a quoting from a source – “How a lemon either hate towards used in the selected media in the case of the and milk can go together” – just to indicate the group or to establishment of the Prosperity Party (PP). that the TPLF (as an ethnic entity) will never undermine the role In fact, unlike other issues, the case of PP merge with the PP (as a unifi ed group). of the group. is not clearly and directly used to describe Both OMN, DW and Tigray Television a certain ethnicity. Stories refl ect elements frame their stories in such a way that they of ethnicity through political diff erences undermine the initiatives of the PP as a coa- among political parties and regions. Th e dif- lition that will be against all the federal states ference among the political groups toward that focus on ethno-linguistics. Th is seems PP emanates from their perception and po- to be the idea of the political leaders of TPLF litical ideology toward ethnicity and federa- and Oromo political elites who are ‘against’ lism. the current leadership of Abiy. One should Looking at the selected stories reported by closely look at how these two groups are not Tigray TV, DW and OMN, they come out as in favour of the PP. Th e main reason can negative on the establishment of PP. For in- be the fear that PP will exclude their ethnic stance, a story reported on OMN on Febru- group from participating in political activiti- ary 8, 202030 explicitly rejects the coming of es32. Th e other unrevealed reason is that the PP. By quoting Jawar Mohammed, the story centralist point of view is always considered tries to undermine the initiatives of PP. His as the idea of Amhara people, where the argument is that PP has forgotten “those TPLF and Oromo political elites are most- who have been fi ghting for change” in the ly in opposition. Th e media channels from country. Th at is to emphasize the role of the Tigray and Oromo have been criticizing the Oromo youth (qeerroo), who were strugg- PP as a cause for the unilateral government, ling for change but may not benefi t from the which is governed by Abiy Ahmed, origi- new party. Also, the media criticize that the nally from former ODP. Th us, three media party promotes a centrist ideology while the channels frequently send negative messages Oromo people are against it. From this story to undermine the initiative of the establish- one can understand that the issue of PP has ment of the new party. a direct link to both ethnicity and political Unlike the above-mentioned media, the ideology, i.e. OMN rejects PP as it does not sampled state media show quite diff erent favor only the Oromo people. angles to the establishment of PP. Th e stories DW and Tigray Television raised similar reported on OBN, EBC, Amhara Television concerns. For instance, a story reported by and Addis Zemen describe the coming of the DW31 tries to undermine PP by emphasi- new party as a positive intervention in the zing the negative aspects of the coming of current political sphere. Amhara Television, the party. According to the story, the current for instance, reported on December 3, 2019 political power of the PP can put direct pres- with a lead story33 that “PP will resolve the sure on all other parties which have been problem of Amhara people who have been struggling to ensure federalism in the coun- living across the country”. Th e story emp- try. Th ough it does not explicitly say so, the hasizes that unlike other periods (the last 28

30 hƩ ps://www.youtube.com/watch?v=gQ_bLw6lhPM 31 hƩ ps://www.youtube.com/watch?v=1JjuX5OFurw&t=1420s 32 hƩ ps://foreignpolicy.com/2019/11/08/ethiopia-will-explode-if-abiy-ahmed-doesnt-move-beyond-eth- nic-based-poliƟ cs/ 53 33 hƩ ps://www.youtube.com/watch?v=5rZSdf7ELyw years), the coming of PP is a hope to bring ‘Us’ and ‘Th em’ based on their ethnic affi li- the best solutions for the Amharic speaking ation. Th e polarization of the media based people who have been living in many parts on political diff erences (private media for of the country and who have faced myriads opposition parties, state-owned media for of problems in the last 28 years. To quote, the ruling party), has been the nature of the “It will ensure freedom of expression, fre- Ethiopian media for a longer period. Howe- It is found a clear edom of movement and freedom of access ver, this polarization based on ethnicity is a polarization of the property, able to work”. A similar frame was new phenomenon in the country. media in describing reported on December 18, 201934, noting Th is study argues that such kinds of a cle- groups as ‘Us’ and the current merger of all parties which were ar and massive diff erence of and extended ‘Them’ based on excluded for the last 28 years in political, interests of the media towards their ethnic their ethnic economic, and other decisive decisions and groups can be reasoned out into fi ve major affi liation. powers in the country. Th e story articula- points. Th e fi rst one is related to the current tes that the establishment of PP is a timely open political system that gives the political and important action that will resolve the parties to organize and serve a particular complex problem of the country. ethnic group. As can be seen, the majority Amhara Television has reported frequent- of the Ethiopian political parties are organi- ly about the importance of the party (PP) zed to address issues of a particular ethnic not only including other groups but also en- group. Taking such nature of party establis- suring fundamental rights such as freedom hment and vested interest to address a cer- of movement and freedom of working by tain ethnic group has brought a direct im- moving from one region to the other. Ho- pact on the media. Th e study reasons that wever, the stories reported on OMN, Tigray the current political leaders and activists, as and DW are quite diff erent. As mentioned well as regional state offi cials, believe that above, they have a strong stance against the they could not get strong support from their establishment of PP. In this case, one can see ethnic group unless they amplify ethnic is- the ethnic elements on the diff erence and sues in their affi liated media. In most cases, the interest of the media either undermining politicians are seemingly interested in sen- or supporting the new party. Since the Am- ding messages to their supporters through hara affi liated media ardently support the the media, which can describe issues that establishment of the PP, opposing channels happened in their groups or their areas as such as OMN, DW and Tigray Television, a deliberate threat or an attempt directed to which favour the Oromo and Tigray elites their ethnic groups. Th is ultimately helps who have a clear opposition to the Amha- the leaders to show how committed they are ra people, have undermined the coming of to serve their ethnic groups. Th is is perhaps PP. So, in the establishment of PP, there has an easy way for politicians in motivating and been a clear contestation of ethnic identity mobilizing the ethnic group they belong to. and groups among the political parties and It is quite similar to the idea of Sotiropoulou the media they control. (2002) who did research on the former Yu- goslavia media. Th e author mentions that Conclusion “ethnicity played an important role in the A thorough textual analysis of the stories issue, as the political elites took advantage and a brief numerical analysis in the selected of the symbolic power that ethnicity has to media in this study show that the Ethiopian off er and used it as a tool for pursuing ter- media tend to take ethnicity at the center of ritorial, political, and economic objectives” their reporting. Th e overall investigation of (Sotiropoulou, 2002, p. 1). From the overall such reporting either intentionally or unin- analysis of the study, one can see a clear inte- tentionally favour their own ethnic group rest of political elites of each ethnic group in and undermine the other, which is percei- Ethiopia tend to use their ethnicity as an in- ved as an enemy. It is found a clear polari- strument to further strengthen their power zation of the media in describing groups as and meet their interest. 54 34 hƩ ps://www.youtube.com/watch?v=lUE1KOBLt2w Secondly, ethnicity and the media in the ly diminish their amicable relationship with current context can also be narrated in the or among the ethnic groups in the country. political positioning of the political parties, As argued by scholars, the media can infl u- regional states, and their ethnic groups. Th e ence the public not only by choosing the overall fi ndings of the study tend to lead us slant of a particular report but also merely to state that the media and their discour- by choosing what to report (Gerber, Karlan se on ethnicity emanate from three major and Bergan, 2009). Th at is what is typically pillars of the Ethiopian political narratives observed in the Ethiopian media. Unless the among the political elites and their ethnicity. story meets the interest of the ethnic group Oromo as a current winner and in power of the owners, there is less probability of get- want to exploit the situation for its ethnic ting covered. So the media system can also group. Th e struggles of political elites and be another factor that motivates the media the media seemingly stem from maintaining to focus on the ethnic issues in the country. the political power of the group, insisting Th e fourth argument focuses on the jour- that not enough is done for the Oromo by nalists’ level of professionalism. Th is argu- the government. Th e Tigray (TPLF) come ment stems not only from the framing ana- out as losers and struggling to get back to lysis itself, but takes the media context into power. Th e Amhara appear as victims and account as well. Unlike previous periods, targets for all threats. As can be seen from most of the media have the chance to hire the political narratives and the analyzed graduates from universities. However, the stories, these ethnic groups are portrayed journalists could not show their skills in re- as victims of the current political positio- porting those sensitive cases selected for this The media tend ning. In these particular pillars, the media, study. Most of the reports analysed here give to play a negative which are particularly ethnic centered, have priority to their own ethnic group. As a re- role in reinforcing brought the agenda that particularly favour sult, the professional integrity of journalism dissociation and their ethnic groups and undermine others. in the current ethnic sensitive media is al- hate in the As a result, they tend to play a negative role most ignored. Also, the values of journalism, Ethiopian in reinforcing dissociation and hate among for instance serving the public and reporting the Ethiopian community. Th is has been accurately, have been overlooked. As a re- community. refl ected in most of the studied media. eTh sult, fundamental ethics of the journalism media are inclined to behave in a way that profession, like minimizing harm and giving meets the interests of ethnic politics in the priority to humanity, appear to be in questi- country. Th is might be the reason that most on in Ethiopia. of the sampled media feature the selected Th e last argument of this framing analysis cases in line with the inter-ethnic political relies on the interests of the media in appro- diff erences in the country. aching sources who only favor their political Th e third argument of the study concerns agenda. A close look at the stories, except the media structure and media ownership. for a few, shows that they use sources that Since the current media, for instance the identify with their ethnic group. As presen- regional state media, are structured under ted earlier, most of the sources are political the regional governments which are demar- parties or activists who support their own cated by languages, the political narratives group. From this, one can argue that since of the regional states determine the media the media serve certain ethnic groups, they content and their frames. Also, the selected may confi ne to certain sources that belong commercial media are implicitly initiated to the same ethnic group. Representing pro- to meet the interests of some ethnic groups fessional media, journalists should go furth- so as not only to enhance cohesion among er in searching for information to make the the people in the same ethnic group but also story complete and balanced. In the analy- to sustain their media business. As a result zed stories, journalists did not succeed in of these, unlike another period, one can including diverse information from diff erent vividly see the political narratives of each sources. ethnic group in the media which ultimate- 55 References

Ethiopian authors are listed with their given name fi rst. Other authors are referenced accor- ding to surname.

Abbink, Jon (2011) Ethnic-based federalism and ethnicity in Ethiopia: reassessing the expe- riment aft er 20 years. Journal of Eastern African Studies 5(4), pp. 596–618. Abrham Fekede (2019) Th e role of social media in political reform in Africa: Th e case of Ethiopia. MA thesis, Addis Ababa University. Amanuel Gebru Woldearegay (2019) Fighting and broadcasting: A ’s Ra- dio Voice of Tigray People’s Liberation Front. Communicatio 45(4), pp. 64–96. Amanuel Yelkal (2019) Audience reception analysis on the documentary ‘Ye Fitih Seqoka’. BA thesis, Mekelle University. Ayele Anawetie Gessese (2020) Ethnic nationalists abuse of media: Lessons of Yogoslavia and Rwanda for Ethiopia. European Scientifi c Journal 16(16), pp. 98–122. Bekalu Atnafu Taye (2017) Ethnic federalism and confl ict in Ethiopia.African Journal on Confl ict Resolution 17(2), pp. 41–66. Busch, Tilman A. (2017) Ethiopian television: Report on the Ethiopian television landsca- pe’s recent developments. Report, European Union External Action Service, Addis Ababa. Danial Zemchal (2019) Assessment of media development in Tigray. MA thesis, Addis Aba- ba University. Daniel Birhanu (2020) Ethnically skewed media: Representation and stereotyping the audi- ence in the case of Asrat Television. MA thesis, Addis Ababa University. De Vreese, Claes H. (2005) News framing: Th eory and typology. Information Design Journal + Document Design 13(1), pp. 51-63. Dereje Teshome Birru (2018) Ethnic federalism implementation in Ethiopia: Th e paradox. Journal of Political Sciences & Public Aff airs 6(4). Dessalegn Yeshambel (2015) Media framing of the Nations, Nationalities and Peoples’ Day in Ethiopia: Th e case of some selected newspapers. MA thesis, Addis Ababa Univer- sity. Deuze, Mark (2006) Ethnic media, community media and participatory culture. Journalism 7(3), pp. 262–280. Dolan, Th eo and Will Ferroggiaro (2019) Social media and confl ict in South Sudan: A le- xicon of hat speech terms. In Allan Th ompson (ed.), Media and mass atrocity: Th e Rwanda genocide and beyond, pp. 319–339. Waterloo, ON: Centre for International Governance Innovation. Eden Niguse (2020) Assessing the practice of investigative journalism on broadcast media in Tigray. MA thesis, Addis Ababa University. Endalkachew Aklilu (2018) Th e polarization of state-owned and private media in Ethiopia: Th e case of EBC and OMN’s reporting on the current political crisis in the Oromia region. MA thesis, Addis Ababa University. Endalkachew Chala (2020, August 8) How the murder of musician Hachalu Hundessa in- cited violence in Ethiopia. Analysis, Global Voices. Available at: https://advox.glo- balvoices.org/2020/08/08/how-the-murder-of-musician-hachalu-hundessa-inci- ted-violence-in-ethiopia-part-i/ Entman, Robert M. (1993). Framing: Towards clarifi cation of a fractured paradigm.Journal of Communication 43(4), pp. 51-58. Fekede Menuta (2019) Language use in multilingual Ethiopia: Th e case of Southern Nation, Nationalities and Peoples’ Regional State. In Logan Cochrane (ed.), Ethiopia: Social 56 and political issues, pp. 27–52. New York: Nova Science Publishers. Fellman, Jack (1992) Language and national identity: Amharic in Ethiopia. Research in Afri- can Literatures 23(1), pp. 173–174. Frey, Lawrence R., Carl H. Botan and Gary L. Kreps (2000) Investigating communication: An introduction to research methods (2nd ed.). Boston: Allyn & Bacon. Fü rsich, Elfriede (2009) In defense of textual analysis. Journalism Studies 10(2), pp. 238-252. Gagliardone, Iginio (2016) Th e politics of technology in Africa: Communication, development, and nation-building in Ethiopia. Cambridge: Cambridge University Press. Gagliardone, Iginio and Matti Pohjonen et al. (2016) Mechachal: Online debate and elections in Ethiopia: From hate speech to engagement in social media. Report, University of Oxford. Available at: https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID2831369_ code2331437.pdf?abstractid=2831369&mirid=1 Gavriely-Nuri, Dalia and Tiki Balas (2010) Annihilating framing: How Israeli television fra- med wounded soldiers during the Second Lebanon War (2006). Journalism 11(4), pp. 409–423. Gerber, Alan S., Dean Karlan and Daniel Bergan (2009) Does the media matter? A fi eld experiment measuring the eff ect of newspapers on voting behavior and political opi- nions. American Economic Journal: Applied Economics 1(2), 35–52. Gezahegn Teji (2007) Music censorship in contemporary Ethiopia: Th e case of Ethiopian Radio and FM Addis 97.1. MA thesis, Addis Ababa University. Gow, Greg (2004) Translocations of affi rmation: mediascapes and cultural fl ows among the stateless Oromo. International Journal of Cultural Studies 7(3), pp. 301–319. Hailu Mideksa (2015) Th e discursive construction of Nations, Nationalities and Peoples’ Day by the Ethiopian Broadcasting Corporation. MA thesis, Addis Ababa University. Hallin, Daniel C. and Paolo Mancini (2004) Comparing media systems: Th ree models of me- dia and politics. New York: Cambridge University Press. Hanitzsch, Th omas, Folker Hanusch, Jyotika Ramaprasad and Arnold S. de Beer (eds.) (2019) Worlds of journalism: Journalistic cultures around the globe. New York: Co- lumbia University Press. Hewan Alemayehu (2019, June 18) Eskinder Nega, the Balderas Council, and the debate on Addis Ababa’s legal and political status: Implications to Addis Ababa residents. Analysis, Gobena Street. Available at: http://en.gobenastreet.com/2019/06/18/eskin- der-nega-the-balderas-council-and-the-debate-on-addis-ababas-legal-and-politi- cal-status-implications-to-addis-ababa-residents/ Horne, Felix (2018, December 13) ‘Ethiopia’s torture problem and the court of public opinion’. Commentary, Human Rights Watch. Available at: https://www.hrw.org/ news/2018/12/13/ethiopias-torture-problem-and-court-public-opinion Janas, Justyna (1991) History of the mass media in Ethiopia. Report, University of . Jawar Mohammed (2020, October 28) How Ethiopia’s transition to democracy derailed. Analysis, Addis Standard. Available at: https://addisstandard.com/in-depth-how-et- hiopias-transition-to-democracy-derailed-refl ections-by-jawar-mohammed/ Kebede Hordofa Janko (2003) Bariisaa. Entry in Encyclopaedia Aethiopica A–C, p. 479. Wiesbaden: Harrassowitz. Kejela Kena (2009) An exploratory study of journalists’ perceptions of their audience: Th e case of Oromia Television newsroom journalists. MA thesis proposal, Addis Ababa University. Kuwee Kumsa (1998) Oromo women and the Oromo national movement: Dilemmas, pro- blems and prospects for true liberation. In Asafa Jalata (ed.), Oromo nationalism and the Ethiopian discourse: Th e search for freedom and democracy, pp. 153–182. Asmara: Red Sea Press. Lyons, Terrence (2007) Confl ict-generated diasporas and transnational politics in Ethiopia. Confl ict, Security & Development 7(4): 529–549. Magsaganis, Matthew D., Vikki S. Katz and Sandra J. Ball-Rokeach (2011) Understanding ethnic media: Producers, consumers, and societies. London: Sage. 57 McCombs, Maxwell E. and Donald L. Shaw (1972) Th e agenda-setting function of mass media. Public Opinion Quarterly 36(2), pp. 176-187. McKee, Allan (2003) Textual analysis: A beginner’s guide. Th ousand Oaks, CA: Sage. Mekuria Bulcha (1997) Th e politics of linguistic homogenization in Ethiopia and the con- fl ict over the status of Afaan Oromoo.African Aff airs 96(384), pp. 325–352. Mekuria Mekasha, Th eodros Woudneh and Bent Nørby Bonde (2018) Overview of the Ethiopian media landscape 2018. Report, Media Progress. Available at: https:// www.media-progress.net/downloads/Overview%20of%20the%20Ethiopian%20 Media%20Landscape%202018.pdf Menychle Meseret (2017) Conceptualizing an African media system: An examination of media and politics in the Ethiopian context. MA thesis, NLA University College. Menychle Meseret (2019) Th e relations between the media and the political system in Et- hiopia's revolutionary democracy from May 1991 to April 2018. Ethiopian Renais- sance Journal of Social Sciences and the Humanities 6(2), pp. 22–38. Menychle Meseret (2020) Th e historial development of media in unconquered Africa: A historial review of Ethiopian media. Media History 26(4), pp. 373–390. Meseret Assefa Adamu (2020) Th e role of social media in Ethiopia’s recent political transiti- on. Journal of Media and Communication Studies 12(2), pp. 13–22. Meseret Chekol Reta (2013) Th e quest for press freedom: One hundred years of history of the media in Ethiopia. Lanham: University Press of America. Mesfi n Negash Bekele (2019) Political parallelism in diaspora-based transnational media: Th e case of Ethiopian Satellite Television and Radio (ESAT). MA thesis, Södertörn University. MoI (Ministry of Information) (1969) Th e handbook for Ethiopia. : University Press of Africa. Mulatu Alemayehu Moges (2017) Why silence? Reporting internal confl ict in Ethiopian newspapers. PhD dissertation, University of . Mulualem Daba Tola (2019) ‘Qeerroo fi qarree’: eTh engine of current transition in Ethio- pian politics. International Journal of Scientifi c and Research Publications 9(5), pp. 534–543. Mulugeta Abrha (2019) Mapping online hate speech among Ethiopians (2018): Th e case of Facebook, Twitter and YouTube. MA thesis, Addis Ababa University. Natinael Yohannes Habite (2019) Media representation of Ethiopian Prime Minister Abiy Ahmed: A comparative study of Al Jazeera Online, Ethiopian Herald and Reporter Amharic newspapers. MA thesis, Addis Ababa University. Pan, Chengxin, Benjamin Isakhan and Zim Nwokora (2020). Othering as soft -power dis- cursive practice: China Daily’s construction of Trump’s America in the 2016 presi- dential election. Politics 40(1), pp. 54–69. Østebø, Terje and Kjetil Tronvoll (2020) Interpreting contemporary Oromo politics in Et- hiopia: An ethnographic approach. Journal of Eastern African Studies. Published on- line ahead of print. Available at: https://doi.org/10.1080/17531055.2020.1796255 Said, Edward (1978) Orientalism. New York: Pantheon. Semir Yusuf (2019). What is driving Ethiopia’s ethnic confl icts? Report, Institute for Se- curity Studies. Senait Haile (2011) Th e framing of federalism: A case study of Ethiopian Television. MA thesis, Addis Ababa University. Sileshie Semahagne Kumlachew (2014) Challenges and opportunities of Facebook as a media platform in Ethiopia. Journal of Media and Communication Studies 6(7), pp. 99–110. Skjerdal, Terje (2010) Justifying self-censorship: a perspective from Ethiopia. Westminster Papers in Communication and Culture 7(2), pp. 98–121. Skjerdal, Terje (2011) Journalists or activists? Self-identity in the Ethiopian diaspora online 58 community. Journalism 12(6), pp. 727–744. Skjerdal, Terje (2013a) Competing loyalties: Journalism culture in the Ethiopian state me- dia. PhD dissertation, University of Oslo. Skjerdal, Terje S (2013b) Selective liberalization: Analysis of media reform in an emerging democracy. In Anthony A. Olorunnisola and Aziz Douai (eds.), New media infl uence on social and political change in Africa, pp. 32–50. Hershey, PA: IGI Global. Skjerdal, Terje (2017) Th e Ethiopian journalist: Loyalist and change agent. Research report, Worlds of Journalism Study. Available at: https://www.academia.edu/32459592/Th e_ Ethiopian_journalist_Loyalist_and_change_agent Skjerdal, Terje and Sintayehu Gebru (2020). Not quite an echo chamber: Ethnic debate on Ethiopian Facebook pages during times of unrest. Media, Culture & Society 42(3), pp. 365–379. Smith, Lahra (2008) Th e politics of contemporary language policy in Ethiopia. Journal of Developing Societies 24(2), pp. 207–243. Solomon Kebede Gebremeskel (2019) Journalistic practice and media production away from home (in exile): Th e case of Wazema Radio. MA thesis, NLA University College. Somerville, Keith (2011) Violence, hate speech and infl ammatory broadcasting in Kenya: Th e problems of defi nition and identifi cation.Ecquid Novi: African Journalism Studi- es 32(1), pp. 82–101. Sotiropoulou, Angeliki (2002) Th e role of ethnicity in ethnic confl icts: eTh case of Yugosla- via. MA thesis, University of Bath. Staszak, Jean-Francois (2008) Other/otherness. In Rob Kitchin and Nigel Th rift (eds.), In- ternational encyclopedia of human geography, pp. 1-6. Amsterdam: Elsevier. Stremlau, Nicole and Monroe Price (2009) Media, elections and political violence in Eas- tern Africa. Report. Available at: https://repository.upenn.edu/asc_papers/723 Stremlau, Nicole (2012) Th e press and the political restructuring of Ethiopia. Journal of Eas- tern African Studies 5(4), pp. 716–732. Stremlau, Nicole (2018) Media, confl ict and the state in Africa. Cambridge: Cambridge Uni- versity Press. Svensson, Mattias (2019) Ethnic federalism and political transition: A study of private me- dia opinions on ethnic politics, human rights and democracy in a changing Ethiopia. MA thesis, Uppsala University. Téwodros W. Workneh (2019) Ethiopia’s hate speech predicament: Seeking antidotes beyond a legislative response. African Journalism Studies 40(3), pp. 123–139. Téwodros W. Workneh (2020) Social media, protest, and outrage communication in Ethio- pia: Towards fractured publics or pluralistic polity? Information, Communication & Society. Published online ahead of print. Available at: https://doi.org/10.1080/13691 18X.2020.1811367 Tuchman, Gaye (1978) Introduction: Th e symbolic annihilation of women by the mass me- dia. In Gaye Tuchman, Arlene Kaplan Daniels and James Walker Benét (eds.), Hearth and homes: Images of women in the mass media, pp. 3–38. New York, NY: Oxford University Press. Van Dijk, Teun (1998) Ideology: A multidisciplinary approach. London: Sage. Ward, David (2011) Ethiopia media mapping 2011. Report, Electoral Reform Internatio- nal Services. Available at: http://www.eris.org.uk/images/userfi les/File/Ethiopia%20 Media%20Mapping%202011%20-%20Final%20Report.pdf Ward, David and Selam Ayalew (2011) Audience survey 2011. Report, Eris Electoral Re- form International Services. York, Geoff rey (2019) Social media in Africa: An emerging force for autocrats and activists. In Allan Th ompson (ed.), Media and mass atrocity: Th e Rwanda genocide and beyond, pp. 415–430. Waterloo, ON: Centre for International Governance Innovation. Yu, Sherry S. (2019) Ethnic journalism and the ethnic press. In In Tim P. Vos and Folker Hanusch (eds.), Th e international encyclopedia of journalism studies. Hoboken, NJ: John Wiley & Sons. 59 Master control room, Amhara Television, Bahir Dar. (Photo by Terje Skjerdal)

Terje Skjerdal (PhD) is Associate Professor in Journalism at NLA University College, Kris- tiansand, Norway. He coordinates a study programme in Global Journalism and has long experience in African media research. Terje Skjerdal has taught and researched journalism in Ethiopia since 2002 and was involved in setting up the master’s programme in Journalism at Addis Ababa University in 2004. He is currently heading the NORPART project ’Enhancing Norway-Ethiopia relations in Journalism and Communication education and research’, which is a bilateral 5-year collaboration between Addis Ababa University, Bahir Dar University and NLA University College.

Mulatu Alemayehu (PhD) is Assistant Professor of Journalism and Communication at Addis Ababa University. He has written and published a number of reports about journalism in Ethiopia. His doctoral studies dealt with the reporting of confl ict in Ethiopian newspapers (University of Oslo 2017). He is the local AAU coordinator of the NORPART project ’Enhancing Norway-Ethiopia relations in Journalism and Communication education and research’, in colla- boration with Bahir Dar University and NLA University College.