11 - Late-Late Development and Political Order

Summary

This final chapter reflects on the implications of the rise and fall of the EPRDF. The discussion considers, first, the EPRDF’s legacy for Ethiopian politics and development, and, second, the relevance of Ethiopia for debates about late-late development and authoritarian durability. While the beginning of the twenty-first century has seen resurgent interest in state-led development and industrial policy, there has been insufficient attention to the political economy drivers that might realise structural transformation in a context of late-late development. For many East Asian late developers, authoritarian leaders pursued development as a means of maintaining political order. While somewhat comparable processes underpinned rapid growth in Ethiopia, the EPRDF’s failings raise doubts regarding the potential for authoritarian state-led development in Africa. The realisation of hopes for structural transformation in Africa will require new political configurations that provide rulers with the incentives to pursue political survival through economic transformation within the constraints of late-late development.

This final chapter reflects on the broader implications of the rise and fall of the EPRDF’s project of state-led development. The discussion considers, first, the legacy of the EPRDF for Ethiopian politics and development, and, second, the relevance of the Ethiopian case for the debates raised in Chapters 1 and 2 about late-late development and authoritarian durability.

The preceding analysis traced the EPRDF’s project of state-led development to, on the one hand, the formation of the Ethiopian state and the 1974 revolution, which left the legacy of a relatively powerful and autonomous state, and, on the other, the threat perceptions of the EPRDF ruling elite once in office. In particular, elite concerns about the lack of a mass base and the threat of ethno-nationalism led the EPRDF to establish an ethnic federal system to allow for ethnic self-determination and to pursue broad-based distribution in ways that tied the masses to the regime and secured their acquiescence. While the government’s initial distributive focus was on land and agricultural inputs, the need for additional distributive resources was a major factor driving an evolving industrial strategy that was intended to maintain political order by generating mass employment opportunities. State-led development was therefore not just an economic strategy, but also a political strategy for maintaining power. Industrial policy failures and limited employment creation, in turn, constituted a political crisis, with large numbers of young adults escaping the government’s attempts to enmesh them in relations of dependence. It is precisely these groups that mobilised along ethnic lines from 2014 onwards, protesting the injustices of the ‘developmental state’, leading to the fragmentation of the ruling coalition, and the collapse of the EPRDF and its project of state-led development.

The TPLF/EPRDF, along with other groups formed in the revolutionary period of the 1970s, sought to tackle two foundational issues – the land question and the national question. Despite its developmental successes, the EPRDF failed to resolve either definitively and was ultimately undone as a result. Indeed, new manifestations of these same underlying issues continue to divide Ethiopian politics through to the post-EPRDF era. At the present time, the land question of the 1970s must be reframed in terms of the agrarian question of labour concerned with the creation of secure employment in sufficient numbers to meet the needs of a rapidly growing population. In the agrarian society of Imperial Ethiopia where a landed elite dominated land ownership and exploited tenants on their land, these distributive questions were understandably framed in terms of access to land. Reforms carried out by the Derg administration and the TPLF temporarily addressed this land question by providing broad-based access to the key factor of production. Yet, rapid population growth and a stalled process of structural transformation have meant growing pressure on limited land resources and a severe shortage of employment opportunities outside agriculture. The agrarian question of labour now facing Ethiopia – and many other late developing countries – is therefore how to provide decent and productive work and economic opportunities to significant sections of the population, particularly younger generations, when these are deemed surplus to the needs of capital.

The legacy of the EPRDF’s failed attempt to resolve the national question is equally problematic. The national question originates in the history of state-building, the incorporation of diverse ethno-linguistic groups into Ethiopian territory and the incomplete attempt at nation-building undertaken by past regimes. Inspired by Stalin’s approach to the national question, the EPRDF adopted federalism in an attempt to defuse ethno-nationalism through the provision of limited self-determination. Like in the Soviet Union, however, rather than defusing ethno-nationalist sentiment, federalism politicised ethnicity by linking access to resources and political power to ethnic affiliation. Moreover, federalism was based on a simplistic assumption that equates language, ethnicity and territory, resulting in conflict regarding, for example, the ethnic ownership of diverse urban centres, the borders between ethnic regions and the status of ethnic minorities outside their ‘home’ region. It is doubtful whether the EPRDF ever intended to devolve meaningful powers to regional administrations. However, the growing emphasis on state-led development after 2001 necessitated further restrictions on regional administrations, which were reduced to implementation of a federal development strategy, undermining any tenuous claims they had to representation of the ethnic group of their region. The result was that the distributive crisis and intra-elite divisions coalesced along ethnic lines. Ethno-regional political elites have mobilised their ethnic constituencies, drawing on historical grievances, inequalities and material deprivation, as a means of pursuing power nationally.

This combination of distributive crisis and ethnic politicisation was not only a major factor in the political transition in 2018, but also in the political and humanitarian crisis that has unfolded since. Far from resolving political problems, the EPRDF’s collapse led to an explosion of inter-communal violence due to the loss of central control that had previously placed a lid on growing tensions. Among the numerous conflicts across the country were those: between the Guji Oromo and Gedeo (SNNPR), between the Berta, Gumuz, Amhara and Oromo ethnic groups in Benishangul-Gumuz; along the contested Oromiya-Somali border; between the Afar and Issa Somali clan; and between the Sidama and Wolayita in Hawassa. Each of these conflicts had complex, historically rooted causes. However, a common theme was conflict over territory, resources and political power, with ethno-political elites seizing opportunities to mobilise their ethnic constituencies against ethnically defined others.

This political crisis is most dramatically and tragically illustrated by the civil war that broke out in northern Ethiopia in November 2020 and which has devastated Tigray, as well as parts of Amhara and Afar regions. With the selection of Abiy Ahmed as EPRDF chairman in 2018, the TPLF leadership retreated from the federal government in Addis Ababa to Tigray to consolidate control over that region. The result was that as inter-communal conflicts spread across the country after 2018, the TPLF continued to deploy its extensive party-state infrastructural power to maintain political order in Tigray. Relations between the TPLF and Abiy Ahmed were strained from the beginning, resulting in a power struggle as Abiy and his allies in the OPDO and ANDM attempted to consolidate power, while the TPLF sought to retain regional control and the party’s dominant economic position. Abiy removed the Tigrayan heads of the military and intelligence services in June 2018 (Ashine 2018), retired members of the old guard, including several Prime Ministerial advisors, and removed TPLF-allied leaders in Somali and Afar. Moreover, the leadership of MetEC – drawn from the Tigrayan-dominated military leadership – and Bereket Simon, a long time senior ANDM minister and advisor, were arrested and charged with corruption. The reform process was not just limited to political elites, however, with the new government removing Tigrayan civil servants – previously over-represented – from the federal bureaucracy (Lefort 2018). Abiy’s move to consolidate power by merging the EPRDF parties into the single Prosperity Party in late 2019 marked the complete breakdown of relations with the TPLF. The TPLF refused to join a party that it considered to be a means of centralising Abiy’s political power.

The following year saw the build-up of federal and regional security forces and escalating tensions between the two camps, not least over the delay to federal elections originally planned for May 2020.¹ Conflict eventually began when Tigrayan forces attacked the Ethiopian army’s Northern Command near the Tigrayan capital Mekele to pre-empt what it believed was an imminent federal attack on the region. The conflict quickly drew in other protagonists, beyond the Tigrayan regional and federal forces. Most notably, this included the Eritrean military which cooperated with the Ethiopian army to settle old scores with the TPLF and the Amhara regional forces and youth movement (Fano) that sought to take Tigrayan territory that had long been claimed by Amhara in Welkait and Raya. As the conflict spread, other actors were pulled into the conflict, including the Oromo Liberation Army, which had returned to its own armed insurrection in Oromiya and formed an alliance with historic rivals the TPLF, while Afar regional forces collaborated with federal military to resist TPLF incursions into Afar. The conflict also took on an international dimension with regional powers such as the United Arab Emirates, Turkey, Iran, Egypt and Sudan engaging through supply of arms, or military or diplomatic support. At the time of writing in late 2022, Ethiopian and Tigrayan forces have agreed a provisional peace deal which has limited the fighting but not yet resolved the key political and territorial grievances that drove the war. Meanwhile, the conflict has had a catastrophic human cost, likely resulting in hundreds of thousands of deaths, widespread human rights abuses, including sexual violence against civilians, the destruction of infrastructure including health facilities, schools and bridges, and the return of famine to northern Ethiopia for the first time in a generation.

The outcome of the war will have vitally important implications for the future of Ethiopian politics and development. The first relates to the national question. The conflict has increasingly been framed as a battle between federalists – represented by the TPLF and OLF – and those favouring a return to a centralised state under Abiy Ahmed (Young 2021). While there are ideological differences between the two camps, the differing views of federalism result as much from divergent elite interests as ideological commitments. While the TPLF was instrumental in establishing federalism, as the dominant player in the EPRDF, the party routinely undermined regional autonomy by enabling federal encroachment on regional responsibilities where this was in line with political or economic priorities. The TPLF’s renewed commitment to regional autonomy only came when the party was reduced to a regional administration resisting federal government intervention. Likewise, Abiy Ahmed rose up through the ranks of a subordinated OPDO into national office under the EPRDF and then rode the wave of Oromo nationalism to the premiership without any apparent objection to federalism. It is only once at the head of the federal government seeking to consolidate power and increasingly dependent on the support of Amhara nationalists that he appears to have developed a preference for a centralised state. While the war is, at root, a factional struggle between contending elites, its outcome will nonetheless have far-reaching implications for the structure of the Ethiopian state and the role of ethnicity within it.

The war likewise has important implications for national development and the distributive crisis highlighted in this book. Soon after Abiy came to office in June 2018, he announced the opening of key state enterprises, such as Ethio Telecom, Ethiopian Airlines, Ethiopian Electric Power, and the Maritime and Logistics Corporation to private investment and competition for the first time (Maasho 2018). Furthermore, large-scale financing from the World Bank, IMF and US government, required to address Ethiopia’s growing debt burden, were made conditional on a series of targets related to economic liberalisation (World Bank 2018b, IMF 2019). Subsequently, the government’s 2021 Ten Year Development Plan announced ‘a gradual transition from public to private sector-led growth’ with a focus on promoting macroeconomic stability (PDC 2021, p. 8).² The result is that the Prosperity Party has come to favour economic liberalisation and private sector development in sharp contrast to the EPRDF’s project of state-led development.

Economic liberalisation in agriculture and manufacturing has important implications for attempts to promote structural transformation and address the distributive crisis. With respect to agriculture, a longstanding political and academic debate concerns the land question and, principally, whether land should be privatised or remain under state ownership (Rahmato 1994, 2009, Nega et al. 2002, Githinji and Gebru Mersha 2007). There have been consistent suggestions that Abiy favours land privatisation as part of a future constitutional revision.³ While privatisation would have important implications for agricultural productivity and the distribution of landholdings, a single-minded focus on privatisation neglects the reality, discussed in Chapter 5, that while land access remains important, it provides a livelihood for an ever declining share of the population. Neither state ownership nor privatisation alone offers any realistic possibility of addressing the severe shortage of rural land and growing population of young adults in need of employment opportunities. Rather, the agrarian question of labour can only be addressed by situating land in the broader context of structural transformation, employment creation and social protection.

With respect to industrial policy, the new government reinforced the trend towards foreign investment promotion from the late EPRDF era, with Abiy announcing the construction of additional industrial and agro-industrial parks amid efforts to court foreign investors. In the short-to-medium term, however, industrialisation has been badly affected by, first, the Covid-19 pandemic and disruption to retail and global supply chains, and, subsequently and more significantly, the civil war. Many of the EFFORT-owned factories in Tigray were destroyed in the conflict, which shattered the perception of political stability that had been so crucial to attracting foreign investment in the first place. The main challenge, however, was the US government’s decision in late 2021 to remove Ethiopia from AGOA privileges in response to human rights abuses and the catastrophic humanitarian situation (Whitfield and Maile 2021). This withdrawal undercut the business model – based on tariff free exports to the US – of many foreign investors in Ethiopia’s industrial parks. The result has been the withdrawal of key firms such as PVH that was vital to the flagship Hawassa Industrial Park. At the very least, it will take years before investor confidence recovers and a foreign investment driven export oriented industrial strategy becomes feasible again.

As discussed in Chapter 6, however, while there is undoubtedly potential for a dynamic manufacturing sector to employ an increased share of the labour force, the EPRDF era has shown that manufacturing alone is unlikely to resolve the country’s distributive crisis. Even should investors return and fill up the existing industrial parks, it is implausible that the resulting manufacturing sector would be on a scale required to resolve the distributive crisis that grows with every passing year. In a best-case scenario an increasingly productive and dynamic agricultural sector would provide improved livelihoods for something like the 16 million households currently involved in agriculture and a revitalised manufacturing sector would create something like the two million jobs that were planned under the EPRDF. However, this would still be dwarfed by the growing population and the need for productive economic opportunities for young adults.

The EPRDF government resisted the idea that social protection could be more than a sticking plaster, providing modest levels of essential support to the poorest for a limited period of time, but ultimately destined to be eliminated as the economy grew and the poorest achieved self-reliance. The likes of South Africa, Brazil and many other Latin American countries – despite their very different histories – highlight a different potential future, however. In these countries – in the context of high land inequality and capital-intensive industry that provides relatively few employment opportunities – there has long been an implicit acceptance that labour availability exceeds the needs of capital. The result is that governments have expanded social transfers to broad sections of the population excluded from productive opportunities, using social protection as a means of maintaining political stability (Seekings and Nattrass 2005, Saad-Filho 2015). EPRDF ideological commitments and resource constraints have limited the expansion of social protection, but the distributive crisis may ultimately necessitate a growing and long-term role for social protection for those surplus to the needs of capital.⁴

Overall, then, the legacy of the EPRDF era is a complex mix of rapid economic growth and improvements in infrastructure and social indicators, as well as a growing distributive crisis and the politicisation of ethnicity. The result since 2018 has been widespread political instability that is conducive neither to structural transformation nor political stability, and which threatens many of the socio-economic gains of recent decades.

The EPRDF’s Ethiopia and the Politics of State-Led Development in Africa

The rise and fall of the EPRDF’s state-led development model is not just of vital importance for Ethiopian politics and development, but it also raises significant questions about the drivers of and prospects for structural transformation in Africa and other late-late developing countries. The beginning of the twenty-first century has seen resurgent interest in structural transformation in Africa and the role for industrial policy. As in Ethiopia, this growing concern with industrialisation is, in part, related to the challenge of youth unemployment (Garcia and Fares 2008, African Economic Outlook 2012, Meagher 2016). While Ethiopia’s experience with industrial parks has increasingly figured in these debates, there has been little attempt to analyse the political dynamics that shaped state-led development in Ethiopia and to set these in comparative context.

The first issue raised in Chapter 1 concerned the feasibility of African states pursuing authoritarian state-led development along similar lines to late developing East Asian countries. In this respect, the analysis in Chapters 3 and 4 showed that the EPRDF’s pursuit of rapid development as a means of consolidating political power shares certain similarities with the likes of South Korea and Taiwan, with political centralisation and elite threat perceptions playing an important role. Inevitably, the process of state formation differed in each case. In South Korea and Taiwan, Japanese colonial rule played an important role in enhancing state capacity and establishing basic manufacturing (Kohli 2004), while political crises after the Japanese withdrawal led to further strengthening of the state and land reform that eliminated the landed elite (Vu 2010). In Ethiopia, it was the 1974 revolution that served to bureaucratise the Imperial state, as well as eliminating the landed elite through extensive land reform. In Ethiopia’s case, however, the legacy of state formation was not uniformly positive: unlike in Korea and Taiwan, the post-revolutionary Ethiopian state had to wrestle with the challenge of linguistic diversity. Nonetheless, the result was that by 1991 Ethiopia had what was, in comparison with many of its African neighbours, a relatively powerful and autonomous state.

There are also certain similarities in the way in which the EPRDF, like the political elite in South Korea and Taiwan, came to see rapid and broad-based development as essential to consolidating power. South Korea and Taiwan faced both external threats from neighbouring communist regimes and severe internal threats from communist revolutionaries (Campos and Root 2001, Doner et al. 2005). Severe internal threats to the ruling elite were even more important in Malaysia, where class conflict overlapped with ethnic divisions (Slater 2010). The result in each case was that rulers pursued distributive strategies involving land redistribution and the provision of agricultural extension and social services aimed at binding the masses to the regime and denying opposition movements a political base (Albertus et al. 2018). Where distributive pressures exceeded resource availability, elites also had a strong incentive to pursue structural transformation as a means of ‘growing the pie’ and expanding resources available to distribute and consolidate their control (Campos and Root 2001, Doner et al. 2005). A comparable dynamic underpinned the EPRDF’s push for state-led development. The EPRDF came to power at a time when the nation-state system had greatly reduced the threat of foreign invasion.⁵ Nonetheless, the EPRDF leadership perceived ethnic and class divisions, and demographic pressures to pose a major threat to their rule. The result was a strategy that sought to enmesh large sections of the population through broad-based resource distribution, while pursuing rapid development to enable the maintenance of political order over time.

While there are certain similarities between the Ethiopian experience and that of several East Asian countries, therefore, there is a sharp contrast with the standard portrayal of African states. From this perspective, African regimes – far from centralising political power and building high capacity states – lack control over much of their territory and people, instead relying on chiefs and clan leaders to maintain political order and using state bureaucracies to distribute patronage opportunities, undermining state capacity and coherence (Evans 1995, Mamdani 1996, Herbst 2000, Kohli 2004). Explanations for this state of affairs largely focus on colonialism,⁶ which divided populations along ethnic lines and created arbitrary borders. In the absence of severe external threats or strong revolutionary movements, African rulers have often consolidated the political order left by colonial regimes, rather than pursuing political and economic transformation that might threaten their hold on power (Boone 1992, Mamdani 1996, Mkandawire 2002).

Ethiopian history is certainly unusual in comparison to other African countries, and, as argued in Chapter 1, this makes the country a leading contender for authoritarian state-led development. However, Ethiopia is not completely exceptional, as sometimes suggested. Africa is an enormously diverse continent, including in terms of the potential for states to play a role in development. While the legacy of colonialism in Africa has often been problematic for state-led development, there remains the possibility of relatively rapid state transformation in extreme circumstances of revolution and conflict, and more gradual and peaceful state-building over extended periods. Perhaps the most obvious comparison with Ethiopia is that of Rwanda, another authoritarian regime that has achieved relatively rapid developmental progress in recent decades.⁷

Rwanda traces its history to a fifteenth century pre-colonial kingdom, which undertook significant centralisation of power and territorial expansion in the nineteenth century (Newbury 1988, Vansina 2005). State-building reinforced the division between the Tutsi and Hutu, which was further racialised and institutionalised under Belgian colonial rule, framing the Tutsi ruling elite as foreign conquerors and the Hutu masses as the indigenous population (Mamdani 2001). This division was politicised in the lead up to independence resulting in the 1959 ‘revolution’ which overthrew Tutsi local chiefs and displaced the king, installing a Hutu-dominated regime and forcing many Tutsi into exile. The remaining Tutsi were politically marginalised in independent Rwanda, though it was not until the early 1990s that a multitude of factors combined to produce the 1994 genocide against the Tutsi. These included an economic crisis caused by the collapse of coffee prices in the late 1980s, an agrarian crisis manifest in land shortages and inequality (André and Platteau 1998, Verwimp 2013), multi-party political competition that politicised ethnic divisions and the invasion of the Rwanda Patriotic Front (RPF) dominated by Ugandan Tutsi exiles that was framed by Hutu extremists as an attempt to restore Tutsi political domination.

The RPF defeated the Rwandan army and ended the genocide, coming to power in 1994 in a context of extreme insecurity. The Rwandan state and economy were devastated, while the RPF – dominated by Anglophone Tutsis who had grown up in neighbouring Uganda – sought to consolidate power over a predominately Hutu population that had been indoctrinated to despise and fear the Tutsi and, in particular, the RPF (Prunier 1998). Moreover, there remained a significant military threat posed by remnants of the Rwandan Army in the Democratic Republic of Congo. The RPF initially focused on restoring political order, and then formulated its longer-term strategy for ruling Rwanda from the late 1990s. Like the EPRDF, the RPF faced a situation in which ethnicity was politicised and the government sought to rule over a population within which it enjoyed little support. Rather than allowing for ethnic self-determination as in Ethiopia, the RPF argued that ethnic divisions were artificially imposed by colonial rule and sought to build a post-ethnic Rwandan nation. The result is that ethnic self-identification has been restricted, while the government has sought to inculcate a Rwandan national identity (Purdeková 2012).

However, like the EPRDF, the RPF placed also mass distribution at the centre of its political narrative in order to ensure that all Rwandans had a material stake in the new government and its vision for the country. Protais Musoni, former minister and leading RPF member, articulated this thinking,

we believe we need a unified nation to achieve development and provide services to people. Singing about unity is not enough, we still need the cement that is an equitable distribution for everybody in order to cement the sense and pride of a nation … With poverty, people can say ‘we have a nation that is unified, but what is in it for me?’⁸

Broad-based distribution became a central feature of the RPF’s political narrative and the construction of a post-ethnic society, while, in the context of extreme resource shortages and intense demographic pressures, rapid development was considered essential to ‘grow the pie’. Not unlike the EPRDF then, for the RPF rapid development became synonymous with consolidating political power (Booth and Golooba-Mutebi 2014, Mann and Berry 2016, Harrison 2017). Unlike Ethiopia, however, this impulse did not initially lead to a clear focus on agrarian transformation or industrialisation. The government sought to leapfrog directly from low productivity agriculture into high productivity services such as tourism, communications, finance and real estate in recognition that Rwanda’s tiny domestic market and landlocked geography provided limited potential for manufacturing. Nonetheless, these emerging high productivity service sectors provided little in the way of employment creation, with most new jobs instead emerging in low wage, low productivity service activities (Behuria and Goodfellow 2019).

By the mid-2000s, the RPF was forced into a re-think, with high inequality and stagnant poverty reduction threatening the RPF’s political narrative. The result was a renewed focus on agricultural extension and subsidised agricultural inputs to raise productivity through regional crop specialisation (Booth and Golooba-Mutebi 2014, Harrison 2016), as well as attempts to expand horticulture and promote value addition and upgrading in the coffee sector (Behuria 2020). From 2015 onwards, the government has also sought to promote industrial expansion by establishing a special economic zone with the aim of attracting foreign investors to take advantage of preferential trade agreements with US and Europe. Thus far, however, industrial development has remained at an incipient stage, with government policy inconsistent and lacking in effective implementation (Behuria 2019).

This developmental push has, like in Ethiopia, been accompanied by a massive expansion of state infrastructural power, as a means of implementing government policies and consolidating control over society. Of particular importance is the decentralised state structures that reach down to village level, with all state officials subject to performance contracts (imihigo) that aim at alignment between national priorities and local development activities (Chemouni 2014). These state structures, party- and military-owned business conglomerates and related organisations such as agricultural cooperatives (Harrison 2017) are used to promote economic transformation and to establish political control down to the individual level, with Rwandans given clear indications of expected behaviour (Ansoms and Cioffo 2016).

The results of the RPF’s development project have been hotly contested. Official figures show rapid economic growth of 7.2 per cent per annum in the decade to 2019, while crop yields massively increased after 2007 (Booth and Golooba-Mutebi 2014, Harrison 2016). Growth in smallholder productivity, in turn, was believed to have had a major impact on poverty (World Bank 2015c, p. 20). Subsequent assessments have questioned these trends, however, suggesting that poverty actually rose, rather than falling rapidly between 2011 and 2014 (Desiere 2017) and that agricultural productivity was stagnant after 2007, with claimed growth actually the result of overestimation and inaccurate yield assessments (Desiere 2016, Heinen 2021). Whatever the extent of socio-economic progress, both critics and those more positively disposed to the RPF agree that the massive expansion of state capacity has been channelled effectively to consolidate the position of the ruling coalition (Golooba-Mutebi 2008, Reyntjens 2013). In a country that is much smaller and more densely populated than Ethiopia, the government has thus far been able to use the party-state infrastructure to stifle any hint of domestic political opposition, whatever the distributive challenges presented by its development model. Time will tell whether the attempt to forge a united Rwandan nation will succeed and whether the modest progress made in the pursuit of structural transformation will contribute to political instability in Rwanda as it did in Ethiopia.

Overall, the Ethiopian and Rwandan cases therefore show that, while comparatively rare, under particular circumstances, elite perceptions of internal threats can provide the incentive for ruling elites in Africa to pursue state-led development as a means of consolidating power. Moreover, state capacity – rooted in the legacies of state formation, but expanded where rulers prioritise developmental progress – is vital to deliver on these developmental plans. Despite the progress that has been achieved, however, elite commitment to development in Ethiopia and Rwanda has, as yet, been insufficient to realise structural transformation.

The Challenge of Late-Late Development

The EPRDF’s commitment to state-led development and its considerable socio-economic achievements can be understood through comparison with examples of state-led development in East Asia and the theories advanced to understand these processes. However, an assessment of the broader significance of the Ethiopian experience for late-late developing countries requires a comprehensive analysis of the reasons for the ultimate failure to achieve structural transformation and the political crisis that followed. Chapters on agrarian transformation (Chapter 5) and industrial policy (Chapter 6) highlighted multiple factors that contributed to the failure or inefficiency of particular projects, constraining economic performance and, ultimately, limiting the distributive resources available to secure mass acquiescence.

Many problems relate to domestic factors and, in particular, the limitations of both domestic capitalists and the state. While past regimes left a somewhat positive legacy in terms of state capacity and autonomy, the EPRDF was at a disadvantage with respect to domestic capitalists. Haile Selassie’s modernisation drive prioritised foreign investors, providing little incentive for the landed elite to turn to manufacturing, while the Derg nationalised private firms, providing protection for inefficient state enterprises. The result was that in 1991 there was no existing domestic capitalist class with manufacturing experience on which to build. Despite the EPRDF’s pronouncements regarding the embrace of capitalism post-2001, the reality was more ambiguous with the EPRDF maintaining its suspicion of the ‘rent seeking’ tendencies of the private sector. The result was that party-state enterprises and politically affiliated firms were often prioritised as part of the industrialisation drive. This lack of sectoral experience among domestic firms proved a major stumbling block when the government sought to establish basic manufacturing in the early 2000s (Whitfield and Staritz 2021).

Furthermore, while the EPRDF inherited a relatively strong state compared with neighbouring countries in Africa, Ethiopia in the 1990s was far from the technocratic Leviathan portrayed – perhaps to the point of caricature – in some of the literature on East Asian development. On taking power, the Ethiopian state was inefficient and lacked technical expertise in priority economic sectors. The EPRDF invested heavily in education, training and bureaucratic reforms to improve technical expertise and implementation capacity. However, the EPRDF’s state-led development project remained an elite-led, politically driven effort, in which senior party officials – and Meles in particular – invested considerable time and effort studying East Asian development strategies and re-interpreting them for an Ethiopian context. The result was that the bureaucracy was dominated by party hierarchies and largely reduced to implementing centrally defined priorities (see also Weis 2015). As the distributive crisis grew during the period covered by the two Growth and Transformation Plans (2010–2020), political leaders set ever higher and more unattainable targets. This over-ambition is surely part of the explanation for the growing focus on high-risk mega-projects, which offered the faint possibility of outpacing the growing distributive crisis and digging the coalition out of a deepening political hole. The reality, in contrast, was that developmental ambitions outpaced or entirely bypassed moderate state capacity, resulting in a series of enormously expensive projects that either did not deliver the expected benefits or failed catastrophically, ranging from the Ethiopia-Djibouti railway, the Addis Ababa light railway, the sugar plantations and fertiliser factory contracted to MetEC, and a series of ever larger hydroelectric dams (Gebresenbet and Kamski 2019, Rode et al. 2020, Lavers et al. 2021, Terrefe 2022).

While the state and affiliated party-state enterprises undoubtedly made several expensive mistakes as part of the EPRDF’s drive for structural transformation, the failings of this strategy cannot be solely attributed to limited state capacity, however. Research repeatedly underscores that industrial policy is an inherently high-risk activity, requiring pragmatism and adaptation in response to initial failures. South Korea and Taiwan succeeded not through an enlightened technocracy laying out a blueprint that was diligently implemented over the following decades, but through a process of trial and error, designing industrial policies that responded to previous failures and an evolving context (Amsden 1992, 2001). Though there were many limitations to the EPRDF’s development strategy, analysis of the government’s approach to agricultural, industrial, urban and social protection policy clearly shows some willingness to adapt and innovate in response to policy failures and new challenges, often with considerable success.

An important argument put forward in this book, in contrast, is that the EPRDF’s hope that structural transformation could provide the mass employment that would address the distributive crisis and consolidate political order was unrealistic, whatever the state’s capacity or the leadership’s commitment to state-led development. The infeasibility of this strategy, moreover, was the result of, on the one hand, the constraints imposed by the contemporary global economy and, on the other, rapid population growth resulting from a delayed demographic transition. The contemporary global economy presents a far more challenging context in which to undertake structural transformation than that which faced earlier industrialisers. The technological frontier, even for relatively low productivity activities such as assembly roles in apparel, has advanced so far at this point that it is extremely hard for new domestic firms which lack required technological capabilities to bridge the divide or to secure market access in global value chains. The absence of an experienced manufacturing class is a condition shared to some degree by all late-late developing countries (Amsden 2001). However, as noted, Ethiopia was particularly disadvantaged in this respect given the severe lack of manufacturing experience on which to build (Whitfield and Staritz 2021). The government’s attempt to buck this trend was ultimately unsuccessful. The subsequent promotion of foreign investment in industrial parks was, arguably, better aligned with the opportunities available to late-late developing countries in the contemporary global economy. However, the pursuit of broad-based development through foreign investment has its own challenges. Resolving the agrarian question of capital through foreign investment breaks many of the potential links between industry and agriculture, leaving the politically sensitive agrarian question of labour (Bernstein 2004). Moreover, for foreign investment to translate into significant value added, mass employment creation and industrial upgrading requires strong state support to promote linkages between foreign investors and domestic suppliers, a challenge that has proved beyond not just the Ethiopian state, but many others also.

The EPRDF’s evolving industrial strategy did deliver a number of successes, notably in floriculture and cement, and, as investment flowed into new industrial parks, there was considerable promise that this could generate significant employment. However, even under the most optimistic assessment – one that often prevailed within the ruling elite – this industrial employment would have paled in comparison with population growth and the demand for jobs. By the 2010s, agriculture had long since exhausted any capacity to absorb this growing labour surplus, while industrial expansion was far too modest and slow to address the problem. In Ethiopia, as elsewhere, the result was rapid growth in low wage, low productivity services (Benanav 2014). While the EPRDF’s concerted pursuit of state-led development was relatively unusual, the ultimate result is a context-specific manifestation of a global problem in which population growth has expanded beyond the needs of capital (Benanav 2014, 2019).⁹ Rapid population growth and growing numbers of young adults without access either to land or urban employment is ‘not a Malthusian problem of resource constraints but a capitalist one of distended and broken labor markets’ (Benanav 2014, p. 116).¹⁰

The introductory chapter to this book made the case that in many respects Ethiopia constitutes a most likely case for authoritarian state-led development whereby a ruling coalition pursues structural transformation as part of its strategy for political survival. Detailed analysis of the EPRDF era partially supports this view, in that the EPRDF did pursue rapid development as a means of maintaining political control with considerable success. However, the distributive crisis that resulted from the limitations of this strategy and the ultimate collapse of the ruling party suggests that the prospects for late-late developing authoritarian states to maintain political order through the structural transformation of their economies may be limited. If African countries are to tackle the challenge of structural transformation, they will need to identify strategies and supportive configurations of political forces suited to the specific challenges of late-late development.

Mass Politics and Authoritarian Durability

The arguments in this book also have important implications for resurgent debates about authoritarian durability. Ethiopia under the EPRDF was a noteworthy example in which the ruling coalition came to associate its own political power with rapid development. For the EPRDF, agricultural productivity growth and industrial employment creation were not just desirable, but an essential means of creating the distributive resources that would bind the masses to the regime and maintain political order. The failings of the EPRDF’s development strategy were, in turn, a major contributing factor in the coalition’s ultimate collapse.

The dominant trend in recent work on authoritarian durability has been to focus on intra-elite cohesion, for which strong political parties are of particular importance (Brownlee 2007, Levitsky and Way 2010, Slater 2010, Geddes et al. 2018, Lachapelle et al. 2020). Well institutionalised political parties can regulate access to power between contending factions, providing opportunities for advancement and compensation for those not selected for senior positions (Brownlee 2007). A common analytical focus in this literature concerns critical junctures such as political crises, civil wars or national independence that lead to party and state building, which underpins authoritarian durability in the long run (Slater 2010). This elite focus is important and has formed a central thread of the analysis in this book. Intra-elite wrangling, ultimately leading to cohesion was an important factor in the TPLF’s ascent to power, the focus on state and party building that sustained the EPRDF in power over nearly three decades and the pursuit of state-led development.

Nonetheless, the dynamics of authoritarian rule under the EPRDF can only be partially understood through the lens of elite politics. A more complete understanding of the EPRDF’s initial consolidation of power and its eventual collapse necessitates analysis of mass politics. As argued in Chapters 3 and 4, a central concern of the TPLF/EPRDF from very early on was how to gain and retain the support, or at least the acquiescence, of the peasantry. This focus on building a mass base drew on Maoist ideology and subsequently informed the EPRDF’s focus on mass distribution and development on taking power in the 1990s. However, the Ethiopian experience highlights another important insight that is often downplayed in recent work on elite politics and authoritarianism – namely, that authoritarian durability is a dynamic challenge. While critical junctures can shape patterns of state and party building over extended periods, demographic changes, evolving class structures and ethnic divisions present a continually changing political challenge to incumbents. As such, ruling coalitions must continually refine and adapt their political strategies, even where they have built powerful party and state institutions to maintain political order. In Ethiopia, the EPRDF initially consolidated political power based on control over the distribution of land and agricultural inputs. However, rapid population growth in the absence of structural transformation led to a generational divide in access to land and employment, undermining the government’s strategy of coercive distribution. Moreover, federalism, which had initially stabilised the political system by allowing for some degree of ethnic self-determination, led to growing ethnic politicisation, with the result that mass distributive pressures were framed primarily as ethnic conflicts. Ultimately the EPRDF’s inability to respond to the dynamic challenge of population growth, urbanisation and the distributive crisis that this represented proved to be its downfall.

Consequently, the literature on elite politics and authoritarian durability can usefully be complemented with a focus on the dynamics of mass politics and elite-mass relations. To do so, research can build on a long line of work that highlights the importance of long-run factors such as the changing balance of class forces and their relations with the state in patterns of authoritarianism and democratisation (Moore, Jr. 1967, Rueschemeyer et al. 1992). Moreover, Albertus’ (2015, 2021, Albertus et al. 2018) recent work has highlighted the importance of mass politics and distribution as a strategy for mass control. Such an analytical approach must focus both on intra-elite politics and elite-mass relations, paying attention not only to critical junctures, but also the dynamic nature and evolving challenge of authoritarian durability. Even the strongest authoritarian regimes cannot afford to rest on their laurels, but must continually adapt to new threats resulting from demographic change and structural transformation, and the particular political manifestation of these in class and ethnic cleavages.

Analysis of the Ethiopian case – as a leading contender for authoritarian state-led development in Africa – highlights the challenges for late-late developing authoritarian regimes pursuing the structural transformation of their economies as a strategy for maintaining power. However, this certainly does not imply the demise of authoritarianism in general. Indeed there have been growing claims of an authoritarian wave that is in the process of rolling back the gains of the third wave of democratisation in the 1990s (Lührmann and Lindberg 2019). Rather than developmental forms of authoritarianism, a more common outcome in recent decades has instead been a transition to electoral or competitive authoritarianism. While past authoritarian regimes would routinely outlaw all political opposition, by the early 1990s few countries – and none in receipt of significant western aid – failed to hold elections at all. A combination of the breakdown of Cold War allegiances, diplomatic pressure and political conditionality from western donors, and domestic pressures have led most authoritarian regimes to introduce some degree of electoral competition (Levitsky and Way 2010). However, a common result has been that incumbents have resorted to ‘feigning liberalism while monopolizing power’ (Brownlee 2007, p. 17) rather than undertaking meaningful democratic transitions. Indeed, competitive authoritarianism has become the dominant form of authoritarian rule in the post-Cold War era (Magaloni and Kricheli 2010).

Under competitive authoritarianism there are constitutional mechanisms for competing for office and meaningful electoral competition takes place, but competition is heavily skewed in favour of the incumbent through its influence on the electoral system, the media, distribution of state resources and the state’s coercive apparatus (Levitsky and Way 2010). Elections have failed to transform politics in many competitive authoritarian regimes. Yet, elections do present leaders with additional challenges. Notably, these include how to limit the defection of political elites seeking to advance their careers by establishing opposition parties and how to secure a majority of votes as the ballot box. A common tendency has been to build broad coalitions, often through allocation of patronage, that co-opt would be political challengers (Arriola 2009), while also extending patronage to politically important constituencies in an attempt to secure votes (Van de Walle 2014). While such coalition building can be a means of maintaining political stability, it is likely to erode the autonomy, capacity and elite cohesion required for state capacity and state-led development (Slater 2010, Vu 2010).

A useful comparison to Ethiopia and Rwanda in this respect is Uganda, where Yoweri Museveni and his National Resistance Movement (NRM) have consolidated power through competitive authoritarianism at the cost of undermining economic transformation. In many ways, the Ugandan experience is more typical of contemporary authoritarianism than developmentally focused regimes in Ethiopia and, to a degree, Rwanda. British colonial rule in Uganda produced a postcolonial political and economic elite that was fragmented and polarised along ethnic, religious and class lines. This fragmentation presented a challenge to post-independence stability that ultimately led to a military coup and a civil war that ended with military victory for the NRM in 1986 (Mamdani 1996, Golooba-Mutebi and Hickey 2013, Whitfield et al. 2015). While conflict continued in northern Uganda for much of the following 20 years, the NRM consolidated power and brought about political stability in the rest of the country by expanding its broad but fractious coalition. Having initially curtailed political competition with a view to prioritising political stability, the NRM was forced to accept multi-party elections from 2006 as a result of a series of defections from the Movement from those that saw little potential for advancement (Tripp 2010, Golooba-Mutebi and Hickey 2013, Whitfield et al. 2015). Since that time, Uganda has closely resembled competitive authoritarianism, with the government retaining the formal trappings of democratic institutions, but routinely manipulating these to strengthen the regime’s hold on power (Tripp 2010). In particular, the NRM has used the distribution of patronage to buy the support of key elite groups and electorally important political constituencies (Tripp 2010, Hickey 2013, Whitfield et al. 2015), while also resorting to arbitrary violence which ‘makes civic spaces fragile, undermines collective action, and destabilizes those who would seek to challenge the state’ (Tapscott 2021, p. 5). As such, the NRM governs ‘not by crushing opposition outright, but instead by destabilizing, fragmenting, and diluting it’ (Tapscott 2021, p. 19).¹¹

While the NRM has solidified its hold on power, the unpredictable nature of competitive authoritarianism has undermined the bureaucratic capacity, autonomy and elite cohesion required to bring about state-led development. Pockets of effectiveness within the state are largely limited to those sectors required for the maintenance of political order (Hickey 2023). In its early years, the government relied heavily on foreign aid to finance the state, resulting in a strong focus on macroeconomic stability and expansion of social services (Whitfield et al. 2015). While this resulted in relatively high economic growth rates, this did not translate into a strong focus on the reforms required to bring about agricultural productivity growth or expansion of manufacturing. Indeed, the attempt to build a broad and diverse political coalition undermined attempts at land reform and the expansion of agricultural extension (Joughin and Kjær 2010, Kjær 2017). Moreover, despite occasional government claims to promote structural transformation along East Asian lines (Hickey 2013), industrial policy has been inconsistent and poorly implemented due to divergent political and economic interests, and ideological differences within the ruling coalition (Whitfield et al. 2015, Behuria 2021). As is the case in many other African countries (Whitfield et al. 2015), the NRM’s efforts to retain political support in a context of competitive authoritarianism have undermined the pursuit of structural transformation.

In sum then, while the beginning of the twenty-first century has seen resurgent interest in state-led development and industrial policy in Africa, there has been insufficient attention to date to the political economy drivers that might realise structural transformation in a context of late-late development. For many East Asian late developers, authoritarian leaders pursued rapid development as a means of maintaining political order. While somewhat comparable processes underpinned rapid economic growth in Ethiopia, the ultimate failings of the EPRDF’s project of state-led development raise doubts regarding the potential for authoritarian state-led development in Africa. Rather than developmental forms of authoritarianism, a more common pattern to date has been for authoritarian regimes to introduce constrained forms of political competition that have enabled rulers to retain political office while undermining the potential for state-led development. The realisation of hopes for structural transformation in Africa will require new political configurations – democratic or authoritarian – that provide rulers with the incentives to pursue development as part of their political survival strategies within the constraints of late-late development. Our understanding of these processes and possibilities, meanwhile, will require a frame of analysis that focuses as much on the strategies for distribution as well as accumulation, of mass politics and elite-mass relations as much as intra-elite politics, and of the structural constraints imposed by class, demography and the global political economy as well as the agency of leaders.