Aid donors, democracy and the developmental state in Ethiopia

ABSTRACT The “developmental state” has become a prominent alternative development model defended by contemporary Western aid donors, particularly in Africa. Purported “developmental states,” such as Ethiopia and Rwanda, are argued to possess strong-willed, visionary leaderships whose commitment to delivering on ambitious development plans renders them attractive donor partners. These leaderships are also, however, often authoritarian and unapologetic when criticized for democratic backsliding or human rights abuses. For many Western donors this represents a tolerable trade-off. The purpose of this article is to interrogate, critique and explain the assumptions and ideas underlying this trade-off. Using the case study of Ethiopia, we argue that donor officials’ understandings of “developmental state” are varied, vague and superficial, the main commonality being a “strong” regime with “political will” and a non-negotiable approach to domestic governance. We suggest that donors have too readily and uncritically accepted, internalized and deployed these notions, using the “developmental state” concept to justify their withdrawal from serious engagement on democratic reform. This derives from a systemic donor preference for depoliticized development models, as well as from Ethiopian officials’ own savvy political manoeuvrings. It has also, however, weakened donors’ position of influence at a time when the Ethiopian regime is debating major political reform.


Introduction
First conceptualized during the 1980s in relation to Japan, the "developmental state" has become a prominent alternative development model defended by contemporary Western aid donors, particularlysince the mid-2000sin Africa. Purported "developmental states," most prominently Ethiopia and Rwanda, are argued by proponents of the concept to possess strong-willed, visionary leaderships whose commitment to delivering on ambitious development plans renders them particularly attractive partners for donors. 1 These leaderships are also, however, authoritarian, heavy-handed with domestic critics and unapologetic when criticized for democratic backsliding or human rights abuses. For many Western donors, supporting such regimes with foreign aidusually with, at best, muted criticism of their transgressionsrepresents a trade-off which is tolerable.
The purpose of this article is to interrogate, critique and explain the assumptions underlying this trade-off. We do so by moving beyond debates on whether states "fit" the developmental state model but, rather, focus on how donorand recipient stateofficials appear to comprehend the concept itself, and the implications of these understandings for their own engagement on democracy promotion. That is to say, we do not seek to define what the developmental state "is" but, rather, explore what marshalling of the concept enables international actors to "do" in the context of international engagement on democratization, or lack thereof.
We focus on the case study of Ethiopia, one of the most frequently cited examples of a "developmental state" in contemporary Western donor discourse. We argue that donor officials' understandings of the "developmental state" concept are varied, vague and superficial, the main commonality being a "strong" regime with "political will" and a non-negotiable approach to domestic governance. We suggest that Western donors have too readily internalized these broad notions, using the "developmental state" concept to narrate their voluntary withdrawal from serious engagement on democratic reform or human rights abuses. This derives, we contend, from a longstanding donor propensity to adopt depoliticized development models to frame and rationalize their activities, as well as from Ethiopian officials' own savvy political manoeuvrings. This tendency, we suggest, helps to produce an echo chamber within which discourses such as that on the "developmental state" become increasingly central to donor support rationales, crowding out alternative narratives around democratization.
To be clear, we do not argue that democratization in Ethiopia represents a catch-all solution to the country's complex political, socio-economic, governance and developmental challenges. Nor do we believe that Western democracy promotion would necessarily be effective, or lead to unqualified, positive transformation; there are limited success stories to be drawn upon in this regard, particularly in Africa. 2 We also do not suggest that commitment to the "developmental state" in Ethiopia is the onlyor even primaryexplanation for Western donors' often uncritical support for the country's authoritarian regime. 3 Our argument instead is twofold. First, that the "developmental state" provides an attractive and malleable imagery and language through which donors can justifyto themselves as well as to the outsidetheir lack of engagement on democratization and human rights in a profoundly authoritarian state. In service to broader developmental goals, this language represents a far more "acceptable" justificatory framework for donors than one based in bolstering authoritarianism in the name of national security.
Second, that Western donors' failure to engage substantively with the Ethiopian government on issues of democratization and human rights in recent years stands in contrast to their stated core values and priorities around foreign and international development policy. The language of the "developmental state" provides a vehicle for side-lining these priorities. This, after all, is a region where mass protests demanding an end to authoritarian rule and the opening of political space have led to major political transformation in both Ethiopia and Sudan since 2016. In this context, it is critical to unpack and interrogate the deployment of concepts such as the "developmental state" since they become the discursive and ideational basis for supporting authoritarian regimes on the one hand and for undermining domestic actors opposing them on the other.
In advancing these arguments, we challenge two key scholarly and practitioner rationales for aid policy choices based on supporting authoritarian "developmental" regimes. First, that the concept is coherent enough to base long-term planning around; our findings suggest that even within specific, country-focussed policy circles, donor and recipient state officials lack a shared understanding of the concept beyond vague notions of a strong, reformist government meriting support. Second, that donor calculations around promoting developmental change can assume that domestic forces in authoritarian "developmental" states will not resist the democracy/development trade-off upheld by international policy engagement. In the case of Ethiopia, Western donors largely refrained from pressuring the Ethiopian government on democratization throughout the 2000s and 2010s, instead bolstering its hold on power in the name of the "developmental state." By 2018, however, the same donors came to be wrong-footed when political dynamics around both democratic reform and the "developmental state" begun to shift in the country itself. The study therefore makes a broader contribution to scholarship on democracy promotion and the politics of foreign aid, placing "development state"-focussed policy decisions in the context of longer-term, systemic tendencies of Western donor agencies. Our study also introduces the notion of the "echo chamber" to characterize and account for the path dependency that depoliticized donor support rationales and narratives appear to follow in states such as Ethiopia.
We structure this analysis as follows. First, we present our methodological and conceptual approach. Second, we chart the evolution of the relationship between Western donors and the Ethiopian government since 1991. Third, we explore how the "developmental state" concept has been described in Ethiopiaby Ethiopian officials themselves and by donor personnel. Before concluding, the article's fourth section places these developments in a broader context, highlighting how donors have depoliticized their aid efforts by once again reproducing an overly simplified grand narrative of development in Ethiopia, and how the Ethiopian government manages donors.

Methodological and conceptual approach
We base our analysis on 64 semi-structured interviews undertaken with current and former Western donor officials based in Ethiopia, together with a number of Ethiopian civil servants and current and former senior figures within the ruling Ethiopian People's Revolutionary Democratic Front (EPRDF). These interviews were undertaken between 2009 and 2018 in Ethiopia (Addis Ababa and Mekelle) and, in one case, Uganda (Kampala). We draw particularly on data collected in Addis Ababa in March 2014 (Jonathan Fisher) and November 2016 (Stephen Brown). 4 Not all respondents are directly cited in this study, nor did they all explicitly invoke the "development state" in their comments. However, a significant number of them did raise the concept as part of their analysis of the relations between the Ethiopian government and international actors, a fact that inspired this article. All interviewees spoke in their personal capacity, rather than as representatives of their organizations. They are cited below with as much detail as they permitted, in most cases choosing to protect their identities due to the sensitive nature of the discussions.
We use the shorthand "Western donors," or "donors," to refer to employees of the members of the Organisation of Economic Cooperation and Development's Development Assistance Committee (OECD-DAC), plus the United Nations (UN) system, World Bank and International Monetary Fund (IMF). The DAC membership includes one multilateral institutionthe European Union (EU)and 29 states, of which the United States (US), United Kingdom (UK), Germany, Canada and the Netherlands are currentlyas of 2019the most significant in terms of Ethiopia's overall aid landscape. The study does not consider Chinese government engagement with Ethiopia; China remains the only non-OECD-DAC member state to provide significant support to Ethiopia. We acknowledge the danger of mischaracterizing as united the policies, positions and perspectives of as disparate and complex a group as the OECD-DAC by using the general term "Western donors" or "Western donor community" and underline differences in the relative position of this group's membership in the analysis itself where relevant and possible. Unfortunately, interviewees' anonymity requirements prevent us from discussing and comparing individual donors' positions more systematically. By "aid" and "foreign aid," we refer primarily to "official development assistance" (ODA), as defined by the OECD. 5 Some of Ethiopia's most significant donors notably the US and UKalso provide substantial amounts of military and security assistance to Addis Ababa which are not captured within the OECD definition, but which we also consider given their centrality to a number of key Ethiopia-donor relationships.
Finally, we adopt a broad understanding of democracy assistance, encompassing not onlyor even primarilydiscrete projects and programmes but, rather, overall dialogue between donors and a government around civil and political rights. This includes not only "behind-the-scenes" and public pressureincluding those linked to aid withdrawal, or threats thereofbut also the absence of this, particularly in the aftermath of "trigger" events such as election-rigging, opposition crackdowns or police firing on protestors. 6 Thus, when we refer to democracy or democratic governance, we mean it as shorthand for a range of civil and political rights.

Ethiopia and its donors
Early relations between Western donors and the EPRDF regimewhich came to power in May 1991 after a lengthy insurgency against the brutal dictatorship of Mengistu Haile Mariamwere uneasy. Although US policy-makers in particular were glad to see the end of the socialist Mengistu regime, they remained wary of the Marxist-Leninist EPRDF coalition and overall aid levels to Ethiopia declined by nearly 50% between 1992 and 1997. 7 This state of affairs was radically transformed, however, during the later 1990s and early 2000s as the Ethiopian regime, under the leadership of Meles Zenawi, aligned itself more explicitly around two key US and World Bank/IMF policy agendas. First, Ethiopia adopted neoliberal economic reforms prescribed by the Bank. Second, the EPRDF positioned Ethiopia as a decisive Western security ally in the Horn of Africa, aligning itself with the US against the Islamist regime of Omar al-Bashir in Sudan and securing a considerable increase in US military and economic aid as a result. This trajectory has continued apace to the present day. Between 1997 and 2008, ODA flows to Ethiopia increased by over 500%. 8 The latter stems in part from the major role Ethiopia has played since 2006 in counterinsurgency and statebuilding interventions in Somalia, which Western donors consider a "safe haven" for Islamist extremists.
The development and promotion by Meles of the "developmental state" model as a framework for Ethiopian economic growth and development since the early 2000s (see below) also further endeared Addis Ababa to Washington, with US and other Western officials variously interpreting it as an example of African "ownership" of aid and a long-term, sustainable plan for securing Ethiopia's economic future. 9 A corollary of this has been Western donors' growing reluctance to apply pressure on the EPRDF regarding its poor record around human rights and democratization. Indeed, Western support to Ethiopia has increased in parallel with a decline in both areas: Freedom House downgraded the country from "Partly Free" throughout the 2000s to "Not Free" since 2010, while aid levels have continued to increase. Restrictions on independent expression, media freedoms, opposition parties and civil society organization imposed in Ethiopia during the 2000s were largely ignored by Western officials. 10 The EPRDF's arrests of opposition figures and stacking of the electoral playing field in its favour in advance of general elections has also attracted, at best, muted international censure. Though Western donors diverted some aid following the killings of nearly 200 protestors by police in the aftermath of the 2005 election, some offered virtually no comment on the killing of hundreds more during crackdowns on anti-government protestors in the Amhara and Oromia regions during 2016 and 2017.
Indeed, the 2005 crisis represents a watershed in donor-EPRDF relations. Prior to this point, Western countries had consistently backed the regime, although they had sought to temper their support, at least to some degree, with criticisms and occasional aid diversions in response to particularly egregious acts of democratic backsliding or human rights abuse. The crisis nonetheless forced the issue of how far donors should maintain even this approach, and subsequently they effectively disengaged entirely from pressuring the regime on these issues. This can be explained in part by Ethiopia's growing role in Somalia during this period. This was also, though, the moment when Meles and his aides begun to develop and promote the "developmental state" as a critical change of direction, providing donors with a language and expansive concept through which to rationalize and defend their support for Addis Ababa.
Donor reluctance to engage Addis Ababa on issues of democratization or human rights nonetheless meant that they were blindsided by the resignation of Meles' successor Hailemariam Desalegn (Meles died in 2012) in February 2018 and his unexpected replacement by a reformer, Abiy Ahmed. Abiy's ascendancy came about against the backdrop of the 2016-2017 protests and major divisions within the EPRDF coalition, leading to the isolation and disaffection of Meles' long-dominant party, the Tigrayan People's Liberation Front (TPLF). 11 In the months following his accession to the premiership, Abiy strongly criticized his predecessors' records on governance and democracy and distanced his administration from a range of EPRDF "sacred cows," including the "developmental state." 12 He contended that his country had "no option except pursuing a multiparty democracy supported by strong institutions that respect the rule of law," released thousands of political prisoners and legalized a range of opposition parties and movements previously banned as "terrorist" organizations. 13 Though it remains too early to say whether Abiy's sentiments will translate into substantive change in Ethiopia's political system, what we emphasize here is that the democratic opening that his rise represents occurred not through the engagement or actions of Western donors but, rather, in spite of them. Moreover, the rapidity with which the "developmental state" framework has been jettisonedat least discursivelyfrom within the EPRDF leadership as a partisan legacy of a deposed faction underscores the naïveté of donors separating political reality from developmental models in approaching authoritarian regimes like that of Ethiopia. More generally, the authoritarian centrism around which the country's "developmental state" has been constructed is likely to sit incongruously with a reform agenda focussed on democratization and liberalization. Pursuing the latter may automatically lead the Abiy government away from the maintenance of the model. 14 In the following section we explore in greater depth the "developmental state" concept, examining how it is understood by scholars and, more significantly, by Western donor and Ethiopian officialsand with what implications for donor engagement on democratization.
The "developmental state" and democracy in Ethiopia Defining the developmental state The "developmental state" is a very elastic concept. Initially coined to describe and explain surprisingly rapid Japanese economic growth and industrialization after the Second World War, 15 the term was later applied to other East Asian economies that had similarly "miraculous" growth, attributable to judicious state intervention in the market. Early expositions of the concept associated it strongly with bureaucratic rule and the political dominance of "pilot agencies" such as Japan's Ministry of International Trade and Industry. However, such characterizations of East Asian political development and economic growth have been strongly criticized by scholars in more recent years as misleading and superficial, while attempts to broaden the concept to describe polities in Latin America, northern Europe and southern Africa have further decoupled the framework from empirical application. 16 The term has, for example, been used to describe a growing range of countries around the world characterized by vastly different political systems and development strategies, ranging from China and Brazil to Botswana and Norway. As Laura Routley argues, the "ephemeral, buzzword, nature of the concept of developmental states … highlight[s] how the concept could become utilized in ways that are unexpected and come to mean different things in different contexts." 17 Since the late 2000s, the "developmental state" has been perhaps most prominently used in Western donor and policy circles to characterize the developmental approaches pursued by the authoritarian governments of Ethiopia and Rwanda. Indeed, as demonstrated below, the term is often applied to Ethiopia by the Ethiopian government itself. The concept has also been popular in scholarly analyses of both countriessometimes being employed to criticize the two regimes' authoritarian tendencies, sometimes to laud their reported developmental successes. 18 In this article, we do not wish to debate the accuracy of applying the term to the Ethiopian case, which the malleability of the concept renders especially difficult. Rather, we want to interrogate the effects of uncritically describing Ethiopia as a developmental state, in particular as it relates to limits to democratic governance and the enjoyment of human rights, because that is what the concept of developmental state is often used to justify.
As discussed below, Western aid donors have tended to argueimplicitly or explicitlythat developmental states require a strong, authoritarian government to achieve economic results. Although many East Asian states characterized as developmental states were indeed authoritarian (at least during the period under study), there is by no means a consensus that democracy is incompatible with the developmental state; quite the contrary. 19 In other words, while Western donors have tended to link successful developmental states with authoritarianism, this is not the view of more recent studies of the concept.
The Ethiopian government's version of the "developmental state" Indeed, the need for authoritarianism has not even been the official view of developmental state proponents within the EPRDF itself. The Ethiopian government's promulgation of the "developmental state" concept has its origins in a leadership crisis which occurred within the EPRDF's dominant partythe TPLFin 2001. Meles Zenawi emerged from the crisis in a strong position, purging many of his most prominent internal critics and empowered to impose his own agenda. 20 Meles' victory in the 2001 showdown enabled him to articulate and implement a clear agenda for Ethiopia's future economic development within a broader debate on "revolutionary democracy" without fear of internal contradiction. The "developmental state" model that he promoted was fully articulated and mainstreamed across Ethiopia's expansive national bureaucracy in 2005, in the aftermath of the May election in which opposition parties made major gains, including all 23 parliamentary seats in the capital. 21 Critically, while the main contours of the "developmental state" model promulgated by Meles have been outlined in periodic national planning documentsnotably the Plan for Accelerated and Sustained Development to End Poverty (2005/06-2009/10) and the first (2010/11-2014/15) and second (2016-2020) Growth and Transformation Plansthe actual terminology of "developmental state" has largely been reserved for more ad hoc dialogues with mainly international audiences. Meles first articulated the idea in a policy paper presented at several international fora in 2006. He and his aides and successors have tended to do the samein interviews with foreign journalists or, in the case of Meles' former close aide Arkebe Oqubay, in a monograph. 22 This underscores two key arguments of this study: first, that the notion of the "developmental state" in Ethiopia is almost deliberately opaque; it is defined explicitly only in the scattered writings of senior officials. Second, that Ethiopian officials have consciously worked to "manage" the donor community and build the concept into donor rationales for supporting the EPRDF regime.
In that regard, Ethiopian expositions of the "developmental state" have eschewed the notion that the model is incompatible with democracy. Meles argued in a 2006 presentation in New York that: Even if a developmental state were to be solely concerned about accelerating growth, it would have to build high social capital that is vital for its endeavours [and] stamp out patronage and rent-seeking. These are the very same things that create the basis for democratic politics … A successful developmental state would thus be very well placed to be both developmental and democratic. 23 Meles reiterated this argument and his immediate successor, Hailemariam Desalegn, made similar points in encounters with Western journalists in 2015 and 2016. 24 For many analysts, however, these theoretical positions have been difficult to square with the Ethiopian government's domestic political agenda. Meles' early-to-mid-2000s reorientation of government policy away from what he referred to as "the neo-liberal paradigm of development" towards "the developmental state model" occurred during a period of political crisis for the ruling EPRDF. The comprehensive restructuring of national state and party institutions carried out in response to the crises of 2001 and 2005 was aimed not only at extending the state's developmental reach but also the ruling party's political dominance. Aided by crackdowns on opposition activists and domestic protestorsincluding the arrests of 60,000 people following the 2005 elections the EPRDF and its allies won 99.6% of parliamentary seats in 2010 and 100% in 2015. 25 Senior officials' ambiguous statements on democratization timelines 26 and heavy-handed reactions to more recent domestic protests have also strained observers' credulity regarding declared EPRDF commitments to pluralism and civil and political rights. Since 2015, hundreds of protestors calling for political reform have been killed by security forces in the Oromia and Amhara regions in particular and two national states of emergencies have been declared.
Moreover, notions of the "developmental state" and its centrality to government policy have co-existed alongside more longstanding and influential political traditions. The EPRDF emerged out of a Marxist-Leninist rebel movement whose senior members have understood "democracy" not in terms of multi-partyism and alternations of power but as a grand coalition of mass associations and cooperatives led by a vanguard revolutionary movement. The idea of "revolutionary (or abyotawi) democracy" emerged, in part, from statist notions of political authority embedded within successive Ethiopian polities but has been transformed and hybridized through its encounter with the post-2005 language of the "developmental state." 27 Indeed, official Ethiopian articulations of what the "developmental state" is, and necessitates, have been remarkably fluid and inchoate, givenor, perhaps, explaining the international resonance of the concept. Meles himself first argued for the importance of following the example of "Asian Tigers" such as post-1960s Taiwan and South Korea, employing statist interventions and management of the economy to promote industrialization and drive growth. This approach has been mixed, however, with an adoption of the "China model," whereby endogenous growth is propelled through defensive engagement with international capital and mass, modernist infrastructural projects facilitated by a vanguard party-state complex. However, EPRDF engagement with this model has been, at best, selective and at worst superficial. 28 Indeed, even Meles' more sympathetic critics have acknowledged that the late prime minister never "fully present[ed] his theory of the developmental state to an international audience." 29 Since Meles' death in 2012, the EPRDFand TPLFhave become increasingly divided and the contours of the "developmental state" concept in Ethiopia have become even less clear. 30 Two senior TPLF officials interviewed during 2016-2017 both explained the "developmental state" agenda in terms of "carrying on the vision of the late Meles" but gave quite different accounts of what this entailed in practice. 31 Following the unexpected elevation of Abiy Ahmed to the EPRDF chair and Ethiopian premiership in April 2018, the "developmental state" concept has increasingly been disavowed by senior government officials. Abiy, leader of another EPRDF coalition memberthe Oromo Democratic Partysecured the premiership against the wishes of the previously dominant TPLF and rapidly presented himself as a "new broom," critical of many policies favoured by Meles and the TPLF and committed to blazing a new, iconoclastic path across foreign policy, security and economic sectors. 32 The "developmental state," an idea particularly associated with Meles and the TPLF, has since come to be part of an acrimonious public debate between TPLF elders and Ethiopian government officials, with the former accusing the latter of abandoning the model in a wider effort to distance itself from the TPLF. 33 The "developmental state," therefore, not only lacks a clear, consistent meaning across Ethiopian officialdom, it is also tied to political dynamics and trajectories which appear increasingly unstable and unpredictable.

Donor understandings of the "developmental state"
Official articulations of the Ethiopian "developmental state" have thus remained considerably ambiguous on the core content of the concept, as well as on the relationship theorized between development and democratization. Given the significant diversity across the Western donor community in Addis Ababa it is perhaps unsurprising that there has also been a lack of clarity within this cohort regarding the model.
Even though the "development state" is frequently raised by donor officials in interviews and is in fact central to how they justify the nature of their engagement with Addis Ababa at a discursive level, as with the Ethiopian government itself the concept is conspicuously absent from official documents. This provides, we suggest, a separation between how donor support for Ethiopia is articulated de jure and de facto. It creates a sealed space for discourses on the "developmental state" to be promoted and developed within the donor community without the external scrutiny that would come with their inclusion in official, public documents. The term itself is not used once in any of the annual reports issued between 2004 and 2018 by the Development Assistance Group, which brings together 30 bilateral and multilateral donors. 34 Similarly, the US Agency for International Development's Country Development Cooperation Strategy 2011-2019 for Ethiopia makes no mention of the "developmental state." 35 The concept is, however, invoked in the World Bank's 2018-2022 Country Partnership Framework a few times in passing, albeit never defined. These references suggest that, for Bank officials, the "developmental state" involves an incongruous mix of visionary topdown economic policy-making, democratic participation and free markets. 36 In our interviews, donor officials nonetheless frequently referred to the "developmental state," but with significant variation in the ways in which they deployed the concept, at least in terms of its main features and how they connect beyond the general consensus that, in the words of one European donor official, the Ethiopian government possesses "a very clear vision … on development." 37 In explaining their understanding of the content of this vision, some, such as a Canadian aid official, focussed on "pro-poor development," 38 for others, however, the "developmental state" centres around promoting industrial growth, 39 while, for still others, including a World Bank official, it is grounded in the building of infrastructure. 40 While perhaps each of these could be argued to be a part of the whole, no two respondents seemed to understand the "developmental state" in Ethiopia in the same way. Very few officials included in their reflections consideration of the Ethiopian civil service or bureaucracy, despite the significance of autonomous bureaucracies to scholarly articulations of a developmental state. Indeed, the main observation by those respondents who did speak about the role of this constituency was that it has been a roadblock to innovation and progress, thwarting reform with crushing amounts of red tape.
Perhaps the most central divide in expositions of the "developmental state" model apparent in donor interviews relates to the credibility and durability of the Ethiopian model itself. Officials from the largest bilateral missions in particularthe US, Canada and the World Bankexpressed considerable confidence in the government's developmental plans and abilities to deliver. 41 Some portrayed this confidence as almost an act of faith: "Either you buy into the developmental state approach, or you don't.
[The Government of Ethiopia] have chosen a model, placed emphasis on the role of the state and political class. … the model has been highly successful … the model pays off." 42 Other donor officials, however, particularly those based in multilateral institutions such as the UN and the World Bank, expressed greater scepticism on the depth and sustainability of the "developmental state" model. 43 Two points of consensus on the "developmental state" in Ethiopia can nonetheless be identified across virtually all those interviewed. First, that whatever successes might be ascribed to the model stem from a "strong state" which cannot be meaningfully negotiated with. "The developmental state" model, one World Bank official noted, "is government-led, there is political will, a strong definite vision of where to go … [the government] only takes money for its own policies." 44 Likewise, as one UN official explained, "the developmental state implies a strong state [and this] shapes the government's outlook and relations with donors." 45 One Western donor official reflected that "there is not much policy dialogue" in donor-government discussions 46 , with another agreeing that "with the Government of Ethiopia there is no discussion." 47 "Strength," "will" and "vision" for donors in the Ethiopian context have tended to be linked implicitly or explicitly to the authoritarian character of the EPRDF regime, which allows it, respondents suggested, to implement developmental policies more efficiently and effectively than neighbouring polities. 48 In the words of one European donor official who transferred from Ethiopia to Uganda, for example, "Ethiopia is very top down, there's not much political freedom, it's quite autocratic … . but there is a very clear … government perspective on development and a very clear vision [compared to Uganda]." 49 According to a UN official, "you can trust Ethiopia more than Kenya." 50 Critically, this understanding of developmental progress and authoritarian rule as interdependent in Ethiopia has led to a second point of consensus across the donor community: that pressuring Addis Ababa on lack of democratization is undesirable or even, perhaps, counterproductive for securing longer-term developmental gains. "All development partners [donors] fall in line," as one aid worker observed, "there is very little pushback. They could push more but don't." 51 "There are trade-offs working with the Government of Ethiopia," suggested officials from one bilateral donor, "but there are benefits of doing it their way … because we deliver on the national [development] plan." 52 Many respondents presented the development/democracy trade-off as an acceptable means to help "craft portfolios like health and education, so you can be more strategic in the longer-term." 53 Indeed, several interviewees suggested that opening up Ethiopia's political system would be counterproductive: "human rights views are disadvantaging the people of Ethiopia," argued one European official, "opposition parties would have a similar approach to governing [to the EPRDF] but would do a worse job." 54 Donor approaches to the "developmental state" and democratization in Ethiopia have not, of course, been uncritical. Most of our respondents expressed scepticism at some of the claims surrounding Ethiopia's developmental success, and many portrayed wider donor approaches to Addis Ababa as naïve or simplistic. However, the Western donor community has collectively structured and rationalized its relationship with the Ethiopian regime since the mid-2000s around support for the amorphous "developmental state" projecta path dependency undergirded by the frequent staff turnover and relatively short postings commonplace to donor missions. Donor understandings of the specific content of this projectand the mechanisms linking this content to particular sectors and developmental outcomesdiffer across, and within, missions. What is common to all articulations, however, is the idea that the project's effectiveness stems from the authoritarian character of the EPRDF regime and that pressuring this regime on democratization runs the risk of undermining a broadly successful developmental model.

Understanding donors and developmentalism
The embedding of this development/democracy "trade-off" within donor-Ethiopia relations contradicts donor commitments to promote and uphold democratic values through development and foreign policy. The UK Department for International Development (DFID)'s stated objectives, for example, emphasize London's commitment to using aid "to champion British values around the globe: freedom, democracy, tolerance and the rule of law." 55 The stated vision of the US Department of Statewhich oversees or administers most of the country's aid programmesis to "promote and demonstrate democratic values and advance a free, peaceful, and prosperous world." 56 The European Union asserts that "commitment to democracy … [is] a principle underpinning its external action." 57 Why, then, would officials from these agencies and organizations so willingly overlook or deprioritize these agendas in the name of a developmental model whose shape and content they understand so vaguely and ambiguously?
Clearly, Ethiopia's perceived value to donors as a security ally is a central part of the explanation. As we and others have discussed elsewhere, Ethiopia's geostrategic significance and the EPRDF's willingness to align with, in particular, US, UK and EU security agendas in the region have persuaded many leading Western aid donors to backpedal on democracy promotion in the country and the region. 58 The Ethiopian government is adept at playing off donor countries' concerns with democratic governance and human rights against their other foreign policy objectives, especially when the latter are rooted in donor self-interest. One Western aid official we interviewed summed up the trade-off quite explicitly: "Why do donors put up with it? We are here because Ethiopia is a strategic country and we have mutual interests. It is not in order to save lives. Ethiopia holds a key strategic position and we want to support itmaybe 'at all costs'." 59 Similar geopolitical considerations nonetheless pervade a range of other donor relationships with African statesincluding Uganda, Kenya, Chad and Djiboutiwhere there is little pretence that this also forms part of an effort to bolster a "developmental state" project. Additional explanations are therefore required in the case of Ethiopia and donor narratives on the "developmental state." In the final section of the study, we provide these by contextualizing the donor-Ethiopia relationship within a broader dynamic, which includes two key processes: the depoliticization of development and the Ethiopian government's skill at preventing anything stronger than mild donor critiques.

Depoliticizing development
Donor agencies and bureaucracies have historically sought to understand development issues through a technical, depoliticized lens. 60 "The temptation of the technical" stems from the economistic, problem-solving focus of international development as an enterprise. As Thomas Carothers and Diane de Gramont argue, the donor community: defined their central mission as fostering economic growth … aid organizations held fast … to the belief that they could help economically transform poor countries by providing timely doses of capital and technical knowledge while maintaining a comfortably clinical distance from these countries' internal political life … these views … are still prevalent in the development aid community today. 61 For decades, successive Ethiopian governments have been adept at using "development speak" to obtain donor support, including under Emperor Haile Selassie (deposed in 1974), and thereby reinforcing their authoritarian power. 62 Most donor agencies have, since the late 1990s, explicitly recognized the fundamentally political nature of development and have sought to incorporate and streamline "thinking and working politically" into their organizations through a range of initiatives. 63 These efforts, however, have had limited impact beyond particular silos and cadresgenerally those populated by governance advisers. 64 The "developmental state" in this context represents an appealing, comforting and intuitive framework for donor officials since it presents development as a technical exercise of inputs and outputs overseen by a benevolent leadership existingseeminglyabove, or apart from, the political realm and commitments to democracy. Certainly, many of our respondents came to present issues and sites of political contestation or activity in Ethiopia in technical terms. One senior European official, for example, explained the basis of the 2016 protests as "about corruption … it is not about votes," 65 whereas in fact the protests were based in much broader, deep-seated grievances around injustice, political marginalization and state-sponsored human rights abuses. 66 Another European donor official defined governance as "things like the provision of basic services." 67 More generally, though, the appeal of the "developmental state" model for Western donors can be linked to the longstanding tradition within the international development community of searching for broad-spectrum, path-dependentand often ahistoricalparadigms to inform and guide policy. 68 The paradigms of development that underlay these shifts have tended to be at a high level of abstraction, conceptualized as a template that can be applied in a wide array of different contexts, but rarely reflecting the empirical experiences of economically developed states themselves. The appeal and deployment of the "developmental state" model by donor officials in Ethiopia, and elsewhere, should be partly understood in this context. It is a paradigm that helps donor officials to organize, justify and make sense of their engagement with Ethiopiaand provides a lens through which to understand and govern virtually every aspect of the donor-government relationship, while justifying their relative inattention to democracy and human rights.

Managing donors
It is important, however, not to overlook the critical role played by the Ethiopian government itself in persuading donors of the necessity of trading democracy for development. Ethiopia is one of the largest aid recipients in the world, with aid representing 15.6% of GDP in 2005 and5.7% in 2016. 69 Though this share is shrinking, it remains significant, but, paradoxically, donors have less sway in Ethiopia than in most Sub-Saharan African countries. 70 One donor official referred to the process of reviewing development policies with the Ethiopian government as "a joke." Discussions on development, for him/her, are actually an "empty, meaningless" performance rather than a "productive interaction." The official concluded: "The bottom line is that it is almost impossible to influence national officials." 71 How is the Ethiopian government able to resist so effectively potential pressure from aid donors, on whom they depend financially?
First, the government is effective in keeping donors isolated both from Ethiopian actors (and therefore alternate sources of information) and from coordinating among themselves. Donor officials are strongly discouraged, for example, from meeting with local actors. As a senior Western aid official admitted: "Donors don't visit prisons or talk to dissidents. The regime is extremely good at controlling information. Donors know nothing. They base their reports mainly on gossip and rumours, and read the tea leaves in presidential speeches." 72 When the Ethiopian government declared a state of emergency in October 2016, in response to nationwide anti-regime demonstrations, the measures included a prohibition on foreign diplomats travelling more than 40 kilometres outside the capital, allegedly for their own safety. These actions have helped to augment the "echo chamber" character of donor discussions on Ethiopia, and core support rationales.
In addition, the Ethiopian government discourages donors from speaking with one voice or coordinating their actions. It makes clear its preference for holding policy discussions bilaterally, rather than with the donors as a group, allowing it to control the agenda more closely, avoid political issues and potentially obtain more resources. A Western development agency official recognized that "When [donors] are alone [with the government], they are more shy about raising human rights and democracy." 73 In other words, preventing aid coordination mechanisms from being effective avoids situations in which donors can "gang up" on the Ethiopian government. The latter, moreover, does not hesitate to pit Western donors against each other, as does Addis Ababa's invoking of its separate development cooperation activities with non-Western countries, notably China.
More generally, Ethiopian government officials are particularly well-known within the Western donor community for being direct and explicit with development actors on what they will or will not accept in terms of aid and international censure. 74 The 2005 elections represented a turning point in this regard. As mentioned above, opposition parties did much better than expected, despite significant irregularities, and the government violently repressed the popular protests that followed, aborting the democratization process and returning the country unambiguously to authoritarian rule. 75 When donors raised objections quite prominently, the government "told them to go to hell," in the words of one UN official. 76 Donors did impose some post-election aid sanctions, but dropped them within six months, 77 and it appears that the Ethiopian government's robust response to donor concerns had a lasting negative impact on the latter's preparedness to criticize Addis Ababa on issues such as democratization.
Since that time, donor officials have been a lot more cautious in their interactions with the Ethiopian government, especially in their public statements. When donors and the government fundamentally disagree about democracy and human rights, "donors wilfully blink" and set aside their rhetorical commitment. 78 As summarized by one European aid official, "sometimes they [donors] talk a lot with the government, but they never clash." 79 Donor officials fear that if they press further, they could be "kicked out" of the country, which has occurred in the past. 80 Some donor officials argued that donor acceptance of non-negotiability in its relations with Addis Ababa is a convenient conceit on both sides; donors do not leverage the bargaining power they potentially possess because they prefer not to, and would rather hide behind a veneer of impotence, or support for a development model which they understand only vaguely and abstractly. 81 A few donor officials are concerned that donors' timidity and intellectually lazy contentment with developmental "success" in Ethiopia are likely to have negative long-term consequences. For instance, one argued that, at a minimum, "By taking the easy road, [donors] are having less of an impact." More fundamentally, the same official argued, "Donors need to act before the country collapses. Big donors want business as usual, but it does not help to hide problems. The country will lose the gains [it has made] if it falls apart," and called human rights "necessary for long-term development." 82 The 2016-2017 protests and emergence of Abiy Ahmed as prime minister in 2018both of which donors observed with surprise, from the sidelinesunderscores exactly this point, and the risks of a donor approach to Ethiopia that ignores popular demands for change.

Conclusion
Ethiopia is often touted as a model of how a so-called "development state" can produce rapid economic growth. Western donor officials have an amorphous understanding of what such a state is, in part due to a lack of consensus on what this malleable term means. Crucially, they have generally equated it with a government that has a strong vision of what it wants to achieve and the ability and "political will" to achieve it. They have also accepted rather uncritically that the curtailment of democracy and human rights is a necessary and acceptable trade-off.
However, as we have demonstrated, the academic literature on the development state does not actually support such claims, while Ethiopian leaders have also argued against the inherent need to postpone political liberalization in order to achieve economic developmentwhile nonetheless cracking down on dissent and limiting human rights. Nonetheless, aid donors have internalized this depoliticized vision of development and deployed it to justify a virtual withdrawal from the Ethiopian political sphere, unwilling to challenge a government that does not welcome their interference in any area. As recent political developments in Ethiopia have demonstrated, popular demands for democracy and human rights have been significant and growing, suggesting that the "developmental state" model has not been (or is, at least, no longer) tenable. By removing themselves from political debates, with not only the government but also other domestic actors, donors have been woefully unprepared to engage productively in policy discussions at a time when the rules of the political game are being rapidly rewritten.
Beyond the Ethiopian case, this article's findings highlight clear implications for both Western policy-makers and scholarsand, indeed, speak to the interface between the two. As discussed, Ethiopia is not the only African state where donor reticence to engage on democratization and human rights is often rationalized with reference to the notional "developmental state," and the concept is sufficiently inchoate and decontextualized to be repurposed for any polity combining authoritarianism with stated developmental aspirations. The Ethiopian government and Western donor community are not, however, the only actors in this process of knowledge production; a range of (often donor-funded) practice-focussed scholars have played an important role in intellectualizing and legitimizing the concept and its value. This article further underscores the different logicsbureaucratic, diplomatic and geostrategicthat lead analysts and practitioners to retreat behind simplistic development models, either because they truly believe in them or because it suits their other purposes. In both cases, our findings emphasize the importance for donors and analysts of challenging received wisdoms and discourses around "what works" in development. The Ethiopia case reveals how comprehensively path dependency and echo chambers can grow in this regardsqueezing out debates on democratization and human rights.