Abiy’s (First) 3 Years in Office

How Political Violence Has Changed

Outbound: تمريدة
6 min readMar 31, 2021


In late March, at a time when most of the international focus was understandably on Tigray, Benishangul-Gumuz, and other bordering areas with Sudan, news broke that a new round of violence in the Oromo Zone inside the Amhara region provoked a high number of casualties and extensive damage to properties. The trigger was reportedly the killing of an imam in front of the grand mosque in Ataye, Kemise woreda, with the Amhara Special Forces laying the blame on OLF-Shane and TPLF, a claim contested by others. In the days that followed, clashes between the Oromo residents and Amhara police spilled over other woredas and zones too. As the third year of Abiy’s premiership — inaugurated on April 2, three years ago — wraps up with yet another major episode of political violence, it seems a good time to take stock of the situation of political violence in Ethiopia and its recent transformations.

A data-driven look at the number of fatalities from episodes of political violence occurred in the three years under Abiy’s premiership, compared to the three-year period which preceded his inauguration, provides some important insights as to how instability has recently changed in the country. Overall, data collected by ACLED shows a 33% increase in the number of fatalities compared to the 2015–17 baseline. But a more nuanced picture emerges if one looks at disaggregated data related to each Ethiopian region. While Abiy’s 3-year premiership witnessed some success stories such as the decline in violence in the Somali region (thanks to the rehabilitation and reintegration of the ONLF in 2018 and a drop in the activities of the Liyu Police there), much bleaker pictures emerge from Tigray, Benishangul-Gumuz, and SNNPR. Halfway in between are, instead, two key regions for Abiy’s power base such as Amhara and Oromia, which have seen a relative decline in fatalities compared to 2015–17 but absolute numbers still remain high.

Others: Addis, Afar, Dire Dawa, Harari. SNNPR includes data for Sidama. Data as of end of March 2021.

Mixed Results

The explosion of violence in Tigray and Benishangul-Gumuz throughout 2020 radically changes the assessment of a situation that, until mid 2019, was actually not that tragic. In April 2018 Abiy had inherited a situation which was already extremely tense: 2015–2017 was a period of turmoil, especially in Oromia and Amhara regions, and Abiy’s appointment was itself the product of that growing instability inside and outside the EPRDF regime. In the first phase of his three-year premiership Abiy did succeed in reducing part of this instability. Yet, things changed thereafter.


In 2018–2019, with a series of reforms and decisions which received much praise internationally but also domestically, Abiy tried to rehabilitate politically some Oromo and Somali armed groups until then in exile (such as the OLF and ONLF), in order to stabilise their areas of operations. Unsurprisingly, then, Oromia and the Somali region saw a significant decrease in armed violence in 2018–2021 (-43% and -46% compared to 2015–2017 baseline). This does not mean that the situation has turned peaceful in these regions, but in relative terms fatalities did decrease significantly. In Oromia this was specifically the result of a new pact that Abiy, Ethiopia’s first Oromo Prime Minister, established with the local youth movement and parties, promising them a way out of marginalisation. However, these commitments were soon questioned after authorities intensified the crackdown on Oromo activists in 2019 (with each side accusing the other of breaking the pact) and returned to the politics of mass arrests, including of leading Oromo opposition figures such as Jawar Mohammed and Bekele Gerba in the aftermath of the wave of deadly unrest triggered by the assassination of the Oromo singer Hachalu in June 2020. The worsening health conditions of the imprisoned Oromo leaders have recently further increased the political tension in that critical region, where additionally OLA fighters (the breakaway faction of OLF) continue their insurgency (especially in the west and south of the region).

Very similar trends to Oromia and the Somali region have been recorded in the Amhara region, also affected by protests in 2015–2017 although on a smaller scale than Oromia. Fatalities almost halved (-47%) in the past three years despite recent upticks in violence (see the opening paragraph). But, overall, the region with the most dramatic decrease has been Gambela (-89%), as authorities have managed to put an end to the series of cross-border raids of mostly Murle militants from South Sudan which caused massacres and kidnappings in 2016 and early 2017.

Sharp deterioration

The top three regions witnessing an increase in political violence fatalities are instead Tigray (+1,690%), Benishangul-Gumuz (+974%), and SNNPR (+176%). The first two also have the highest per capita score among all the regions: in 2018–21, the yearly fatality rate every 100,000 inhabitants was 37 in BSG and 17 in Tigray — whereas SNNPR had ‘only’ 0,75, one of the lowest among all the regions (Somali had 1,75, Oromia 1,45, and Amhara 0,52. For replication purposes, I calculated per capita scores on the basis of 2017 population figures provided by the Central Statistic Agency).

Before November 2020, in Tigray only few fatal clashes had been recorded, largely between the Tigray regional police and the residents of disputed areas with the bordering Amhara region such as Welkait and Raya, or between Ethiopian troops and Eritrean forces/(then)Eritrea-based Ethiopian armed opposition groups. Similarly, in Benishangul-Gumuz (BSG), especially in Metekel zone, inter-ethnic violence carried out by ethnic militias (mostly Oromo, Amhara, Gumuz militias) escalated only in late-2018 because of a combination of historical grievances, land resources, and governance failures (when is this not the case? Here specifically linked to flawed ethno-federal arrangements and the fear to lose one own’s rights at a time of shifting visions of the future of Ethiopia as an ethno-federal or unitary state). Before that, in 2015–2017 fatalities were limited to clashes between the Ethiopian security forces and the Benishangul Gumuz People’s Liberation Movement (BPLM), back then still based in Eritrea and involved in operations against Ethiopia’s EPRDF and its projects, including the GERD dam located in BSG. Finally, in 2018–2021 violence has also been on the rise in the Southern Nations largely because of an increasing number of communal/ethnic-based attacks and of clashes between federal troops and protesters, including because of resettlement programmes.

Many of these recent episodes of political violence are interpreted as similar manifestations of ethnic groups’ dissatisfaction with the prospect that ethnic federalism, Ethiopia’s form of state system which since 1994 has been supposed to guarantee self-determination and self-rule to its ethnic groups, could be replaced by a more unitary state system, the one sponsored by Abiy and his Prosperity Party project.

As a final note, clashes in Fashaqa, the Sudanese farmlands disputed by Ethiopia’s Amhara nationalists, have also been on the rise. While a few major fatal events had already been recorded in the border area in 2015–2017, the number of clashes increased significantly after November 2020 (although related fatalities have so far remained relatively contained: around five dozen fatalities in total in 2018–2021). Things could well change if Sudan and Ethiopia are dragged into a full fledged war. For an overview on the Ethio-Sudan tensions, the Fletcher Forum of Tufts University recently published an in-depth analysis by Faisal Ali with some comments of Kholood Khair and me on the role played by Addis’ and Khartoum’s domestic considerations. An even deeper analysis on all the major border tensions in the whole Horn of Africa that I cannot recommend highly enough is this long read Dan Watson wrote for ACLED.

Who done it?

Additional insights can be obtained by looking at the actorness of such fatal episodes of political violence. The following interactive charts capture the evolution of political violence in every region since April 2015 and additionally categorize fatalities by actors, specifically checking (whenever possible, considering the difficulty in retrieving such information for all the observations) whether state forces were involved (according to ACLED’s codebook, state forces are “collective actors that are recognised to perform government functions, including military and police, over a given territory”). In April 2018–March 2021, Ethiopian state forces continued to be involved in at least 29% of fatal violence in the country (it was 55% in 2015–17). Browse by region: