The governance of Addis Ababa Light Rail Transit

Chapter 11 in the book 
 Refractions of the National, the Popular and the Global in African Cities.


the national vision of becoming a middle income country, Africa's diplomatic capital and international competitive city. (Addis Ababa and Special Zone Development Planning Project Office 2013: 4) Furthermore, evidence from recent studies shows that the implementation of mega city projects was part of the government's conviction in making Ethiopian cities global cities (Terrefe 2020;Kloosterboer 2019). In arguing how Ethiopian cities would be a hub for industrialisation, diversification and structural transformation, a chapter prepared to guide Ethiopia's structural transformation shows 'building both "hard" and "soft" infrastructure' would be instrumental to improve Ethiopian cities' competitiveness in the global economies (UN-Habitat 2014). Thus Ethiopia's commitment to urban governance is closely related to its perceived and tangible benefits in facilitating rapid growth and transformation. Within this background, this chapter aims to explore how one of the mega urban projects implemented in the city of Addis Ababa, that is, the Addis Ababa Light Rail transit (AA-LRT) has evolved; the government's discursive rationale in choosing AA-LRT as an urban transport option through elucidating whether or not the AA-LRT was developed out of a felt need or the pressure to be global. In addition, the study discusses the politics around the investment in AA-LRT and how it is governed through highlighting the governance structures and institutional setups.

Urban governance under a revolutionary developmental state
Urban governance refers to processes and mechanisms through which diverse actors in a given urban centre/city organise their actions and resources for the common urban good through utilisation of economic, social, political and environmental resources (Harpham & Boateng 1997). Multiple actors include formal and informal, governmental and non-governmental, public and private. For these actors to function effectively and achieve collective goals, institutionalised arrangements establishing structures and systems have to be put in place, and these will emerge out of the stakeholders' diverse interests through consensus building and/or contestation (Hendriks 2013;Lindell 2008;Melo & Baiocchi 2006). Under normal circumstances urban governance is robust enough to accommodate contestation, as relevant actors promote their own vested interests on what they want to be done and how. Ideally, urban governance as a process facilitates persuasion, contestation, bargaining and, finally, consensus building among actors during urban policy development. Lindell (2008Lindell ( : 1880 further suggests that urban governance consists of 'multiple sites where practices of governance are exercised and contested, a variety of actors, various layers of relations and a broad range of practices of governance that may involve various modes of power, as well as different scales'. Lindell's characterisation of urban governance as made up of multiple, contested sites highlights how any given urban policy is supported or opposed at different levels, and its fate will depend on the efficiency and effectiveness of formal institutions and the relative power between policy actors. Hence, the effectiveness of urban governance is dependent upon how these multiple actors engage with different levels of governance structures and with each other, shaping the exercise of governance. The relationship of relevant actors as well as their active participation in the process of planning and implementation of urban policy and strategy vary by context. Ethiopia is one of the countries that endorsed the global campaign on 'urban governance,' which openly advocates for a participatory democratic process between citizens, public and private actors in urban agenda setting and decision-making processes (see UN-Habitat 2002). The globally promoted notion of urban governance includes planning and governing a city's common affairs through an ongoing process that accommodates conflicting or diverse interests, leading to cooperative action. In reality, however, Ethiopia's political landscape limits the applicability of many of the attributes of commonly accepted definitions of governance. This relates mainly to Ethiopia's ruling party's adherence to 'revolutionary democracy' as its guiding philosophy and political economic strategy. Ethiopia's 'revolutionary democracy' is an ideology that draws its inspiration from Marxism, Leninism, Maoism, and liberal ideologies in defining the socio-economic and political structures of the country (Bach 2011). The ideology is promoted as the only doctrine that would help defeat Ethiopia's major social, economic, and political challenges based on one vanguard political party system. It thus grants discretionary power to party leaders and bureaucrats at all levels, effectively eliminating other voices. In other words, a decision taken by the party can override a decision arrived at within state institutions following the normal conventions. Vaughan (2011) analysed the state party and the Ethiopian People's Revolutionary Democratic Front's (EPRDF) revolutionary democratic nation building process and highlights 'democratic centralism' as a fundamental tenet of EPRDF's ideology. Hence, Ethiopia's revolutionary democracy is an ideological tool legitimatising the way central committee members of the EPRDF leadership interact with local level structures to guide and lead the socio-economic development processes. According to Fantu Cheru (2016: 606), Ethiopia blends different ideologies together in 'a cleaver meshing of selected development experiences from East Asia (China, Taiwan, Vietnam, South Korea, Singapore and Japan) with 'soft neoliberalism' from the West under the guidance of a strong developmental state.' Thus, understanding Ethiopia's urban governance requires situating it within its historical, social, economic, and political history. Against this backdrop, this study looks at the governance of the Addis Ababa Light Rail Transit (AA-LRT) project as the country's major attempt to provide mass urban transport for a rapidly expanding city using a political economy approach.

Analytical approach and research questions
As discussed above, Ethiopia's political and economic strategy based on 'democratic centralism' empowers party leaders to make policy decisions and inhibits the application of globally promoted urban governance models that assume contestation and bargaining between interest groups with competing interests. This chapter applies a political economy lens in analysing the governance of AA-LRT. Mcloughlin (2014) describes the aim of political economy analysis as locating development interventions within an understanding of the prevailing political and economic processes in a given society. Applying a political economy approach offers a means to situate the structure and processes of Ethiopian urban affairs within the ruling party's overarching political ideology and how it dictates policy agenda setting and the interaction between the state and its citizens. According to Drake (2001), political ideologies are clusters of values espousing the underlying philosophies and principles of the government through which socio-economic and political systems and structures are operated. Drake further argues that political ideologies generally are determined by state views of how a society should be organised/governed, and the extent to which a state may interact with its citizens and intervene in their daily lives.
Specifically, this study draws on a 'state centred approach' to political economy in studying the governance of AA-LRT. In their seminal work on diverse strands within political economy, Caporaso and Levine (1992) show that different approaches vary by their assumptions about the primacy of economics or politics in decision-making processes. A state centered approach to political economy focuses on situations where states systematically exclude non-state actors from the policy process and generate policy initiatives on their own. This accurately describes how the political regime in Ethiopia has instituted a developmental state governance structure and allocates major decision-making power to party leaders in guiding urban development priorities. Thus, this chapter critically analyses AA-LRT's institutional and governance arrangements including its structures and capacities. Specifically, this chapter focuses on the following interrelated research questions: 1. Where did the idea for the AA-LRT originate? 2. What was the discursive rationale and strategy for the government of Ethiopia choosing AA-LRT as an urban transport option? 3. What are some of the politics surrounding the investments in AA-LRT and were there external influences? 4. How is AA-LRT governed? What are the structures and institutional setups?

Methodology: Design, data collection methods and data analysis
In order to address the aforementioned research questions, this study used a qualitative research design. Creswell (2007: 17) defines qualitative research as 'an inquiry process of understanding based on distinct methodological traditions of inquiry that explore a social or human problem. The researcher builds a complex, holistic picture, analyses words, reports the detailed views of informants, and conducts the study in a natural setting.' There were two reasons for choosing a qualitative study design. First, qualitative research is appropriate when not much is known about the topic under inquiry (Creswell 2007;Padgett 1998). As the AA-LRT was launched in 2015, there is no literature that studied the AA-LRT from a policy and urban governance perspective. Second, a qualitative design provides the most appropriate methodology to obtain indepth information on the process through which AA-LRT received prominence and emerged as a major urban transport option in Ethiopia by generating explanations embedded within empirical observations. Hence, qualitative data were collected using in-depth interviews and focus group discussions 1 conducted with AA-LRT users in two stations (i.e. CMC station and Meskel Square). A total of 17 in-depth interviews with key government officials relevant to the AA-LRT project, research institutes, and civil society members were conducted (see Annex I: Summary of study participants). Specific attention was given to government officials who were involved at the initiation phase of the AA-LRT and were on the board of directors. Data were analysed using Padgett's (1998) multiple interrelated steps. Transcripts of interviews and discussions were reviewed and frequently emerging themes relevant to the study questions were identified. Once relevant themes were organised based on the research questions, interpretation of themes was conducted. To ensure the credibility, transferability and dependability of the data (Lincoln & Guba 1985), triangulation of emerged data by data source was used.
In presenting findings, this chapter is structured into three parts. First, a description of the overall status of public transport in the city of Addis Ababa is presented to contextualise the LRT's emergence as an important mass transit service in Addis Ababa City. In this section, the process that led to prioritisation of the AA-LRT in the public policy arena is discussed vis-à-vis the politics around financing the infrastructure, operation and maintenance costs of AA-LRT. The question of finance and delivery of the infrastructure goes to the core of external influences in the process. Second, the governance of AA-LRT is described, including the governance structure, institutional capacity, and relationship between national and city governments. Finally, the chapter discusses implications of the findings for urban governance policy and practice in the Ethiopian context.

Public transport in Addis Ababa: Current status and the quest for mass transit
Literature that provides a comprehensive understanding about the status of public transport in Ethiopia, in general, and the city of Addis Ababa, in particular, is sparse. The few available materials on public transportation do not have up-to-date data. Existing literature and government documents show that public transport in Addis Ababa has been dominated by city bus services provided by the public enterprise Anbessa City Bus Enterprise and taxis operated by private sector actors (Gebeyehu & Tekano 2007). Fenta (2014) conducted a study on public transport demand in Addis Ababa and estimated that there were nearly 10 000 taxis with a capacity of 12 persons, 460 buses that seat 22 to 27 persons, and 6 500 sedan taxis that seat four persons. Three-wheeler taxis (bajaj) and animal carts are also used in the peripheral areas of the city.
Qualitative findings for this study also indicate that public buses make up the bulk of public transport in the city, in addition to privately owned taxis (mini-buses and sedan taxis). According to key informants, different bus systems exist. For instance, the oldest bus service is Anbessa City Bus Enterprise, which is publicly owned and has 814 buses and currently transports nearly 500 000 passengers daily. After the city administration issued its transport policy in 2011, the city administration has been making efforts to improve the mass transport service in the city. For instance, the city administration introduced a Public Service Employees Transport Service Enterprise, which has provided free transport to civil servants, running during rush hours in the morning and after office hours since 2015. There are 410 of these buses transporting 98 000 persons per day. Furthermore, in May 2016, the Addis Ababa City Administration Transport Programs Coordination Office introduced a 'Sheger Express' bus system and deployed 150 buses. Sheger Buses are equipped with modern features such as automated ticketing and are accessible for the elderly and people with physical disabilities, which enhances the city's ambition to provide inclusive transport service for nearly 25 000 individuals. In addition to the newly introduced buses, the city administration planned to bring in 1000 metered taxis, of which 150 are already functioning. After critically examining the inadequacy of existing public transport systems, the city administration granted permission for 300 privately owned cross-country buses to serve citizens before and after office hours; these serve 65 000 individuals. Private cars only transport 20% of city residents 2 (see Table 1).  (2014) estimates that pedestrians make up 55% of daily journeys but do not have appropriate walking facilities and confront a transport network characterised as 'non-pedestrian friendly'. The inadequacy of public transport coupled with a lack of planning and integration of motorised transport systems makes mobility in Addis Ababa chaotic, unreliable, unsafe, unaffordable and inefficient for an expanding city (Fenta 2014). A recent study by UN-Habitat (2017) confirms that public transport in Addis Ababa is inaccessible and unaffordable for the majority of city residents. As a result, the Addis Ababa City Government has been searching for an alternative mass transit system. The federal government eventually embarked on a massive project and introduced the AA-LRT service as the first modern mass transit system in the country (Ministry of Transport 2011a, 2011b). It has been promoted as a major breakthrough in the transport industry, not only for Ethiopia, but also in the history of sub-Saharan African countries (Jemere 2012;CNN 2015;The Economist 2015). The AA-LRT development took three years with a total infrastructure investment cost of USD 475 million and the project was inaugurated on 21 September 2015. The AA-LRT consists of two lines: the East-West axis from Ayat to Tor-Hailoch and the North-South axis from Minilik II square to Akaki/Kality. The LRT covers a total length of 34.25km (North-South line 16.9km and East-West line 17.35km) (see Figure 1).

Source: Maximilian Dörrbecker
Light Rail Transit: Its origin, discursive rationale and the politics around financing the project Over the past 15 years, several initiatives were considered and tested as the Addis Ababa city government searched for an afforfable and efficient mass transit system to reduce traffic congestion and urban pollution. A Bus Rapid Transit (BRT) system similar to successful bus transit systems in several Brazilian and Colombian cities was the first seriously considered initiative. The city government in collaboration with 1 East-West route 2 North-South route the French Development Aid agency commissioned a French consulting company to undertake a feasibility study for a Bus Rapid Transit (BRT) system for Addis Ababa in 2004. 3 The study recommended rapid bus corridors along the North-South and East-West direction of the city. In 2005, the city contracted and commissioned an Indian consulting firm to draw up a transport master plan (Lyon Town Planning Agency 2011). 4 Based on its suggestions, several initiatives were taken to translate the BRT idea into practice on the ground. For instance, the current two LRT route (i.e. north-south, and east-west) corridors were initially planned for the implementation of BRT. 5 Long before the Light Rail came to the fore, a BRT demonstration corridor was built in the north-south direction, starting from the centre of the city (i.e La Gare) to northern Kality. The current La Gare to Kality LRT route was a full-fledged asphalt road, which was designated as a BRT route. Buses with a carrying capacity of 23 and above started providing a rapid bus service along the proposed corridor. While the city of Addis Ababa was undertaking the BRT pilot project as a preferred mass transport option, a swift discussion on the idea of Light Rail Transit (LRT) emerged, seemingly out of the blue and without warning. As one key informant succinctly put it: BRT was initially considered as a solution to the city's mass/public transport need and the two lines; one from La Gare to Kality and Megenagna to Ayat round about was left as BRT corridors. Without any further consultation to relevant stakeholders. No justification was given on abandoning the BRT and embarking on LRT. The current LRT routes (North-South and East West lines) were originally planned to be BRT routes. 6 Moreover, a key informant reported that the rise of LRT, while abandoning BRT, had been a public transaport idea imposed on the Addis Ababa City administration by the federal government: The city undertook various studies in search of mass transportation means and Bus Rapid Transit (BRT) was initially considered as solution. The city administration constructed one BRT corridor from La-Gahar to Kality, which now serves as LRT south-north route in due time though, LRT got momentum, and the focus shifted from BRT. 7 Another key informant noted: BRT was the first choice of the city Administration and much was done from allocating a BRT corridor and running pilot rapid bus service in the city. The 3 Interview with Bureau of Addis Ababa City Administration 4 A report by Lyon Town Planning Agency reported that the pre-feasibility study and the development of Addis Ababa city transport master plan were part of the Ethio-French project from 2002-2010. 5 Interview with Addis Ababa City Road Authority 6 Interview with AACRA 7 Interview with Ethiopian Railway Corporation BRT corridors were now serving as LRT routes. The LRT idea won and the city dropped the BRT idea. However, after the completion of the LRT, BRT is rising. The city has completed a study to introduce BRT so that it can be integrated with the LRT service. (Interview with Addis Ababa City Transport Authority) The qualitative findings strongly suggest that Addis Ababa City's transport planners and the mandated city administrators developed the BRT strategy to curb the ever growing city residents's transport need. However, the swift decision to focus on the more expensive AA-LRT project poses a practical question: What was the rationale behind choosing AA-LRT over BRT?
The rise of LRT: Issue emergence and discursive rationale 'The late prime minister, Meles Zenawi, first suggested LRT as a quick solution to the city's growing transport problem.' 8 Almost all key informants who participated in this study unequivocally confirmed that the late Prime Minister Meles Zenawi was the person responsible for singlehandedly promoting the LRT as the best mass transit option. A key informant from the Addis Ababa City Road Authority stated: Once LRT was introduced as his 'Ethiopia renaissance' showcase, all cabinet members backed the idea without much confrontation. 'Once the late PM introduced the idea, no one dared to challenge,' as one key informant pointed out. 9 The LRT was thus a priority project for the government as it strived to demonstrate to the visiting African heads of state and to the people of Ethiopia the regime's commitment towards enhancing the image of Addis Ababa as the capital of Africa, as well as to reinforce the 'Ethiopian Renaissance' narrative that was getting political traction. In short, LRT had strong political backing from the highest levels of government. The urgency of completing the LRT project in a very short time-frame was promoted by the preparation surrounding the 2015 national elections. As one key informant succinctly put it: In addition to the social and economic significance of LRT for the residents of Addis, it is a political project that had symbolic meaning enhancing the renaissance of the country as well as helping the ruling party gain popular support.
Despite the efforts made in seeking an in-depth explanation, the swift government decision from implementing BRT to initiating the LRT remained a mystery. The late prime minister singlehandedly advocated for the emergence of LRT as a mega city project.

Discursive rationale
This study found that the interaction of two major factors contributed to the sudden abandonment of the BRT initiative in favour of the LRT in 2007. The first factor has to do with Addis Ababa's rapid population growth due to the dramatic growth of the Ethiopian economy, particularly in the capital and its surroundings. Addis Ababa's population is thought to be growing at an annual rate of 2.1% (Central Statistics Agency 2013). With the spatial expansion of the city, fuelled by expanding industries and other economic enterprises, more and more people were migrating to the capital in search of jobs and other economic opportunities. This expansion was further exacerbated by massive urban renewal projects initiated by both the federal and city governments. Since the 2005 urban policy, the city was busy clearing major slum areas in the inner city. As a result, 80% of inner-city dwellers who used to live in public housing were relocated into condominium apartments that were built in satellite cities, which have appeared in the outskirts of Addis since 2005. 10 Relocated residents have been forced to commute long distances to and from the city centre where their work and social life are concentrated. As the city expands and creates satellite cities, a huge demand for transport is emerging that can no longer be satisfied with existing systems. Consequently, the government concluded that the BRT option was unlikely to meet the growing mobility needs of the fast expanding urban population. Instead, the federal government opted for the Light Rail Transit (LRT) option as the best solution given the changing dynamics of Addis Ababa and the surrounding regions.
Second, the government wished to modernise the city's image by introducing a modern mass transportation system, and that the LRT initiative fits very well with the evolving 'Ethiopia's Renaissance' narrative. Similar to the 'Africa Rising' narrative, the notion of 'Ethiopia's Renaissance' evolved internally while the country was celebrating the Ethiopian millennium nine years ago in 2007. Since then, the notion of 'Ethiopia Rising' has dominated the national development discourse, and is increasingly reflected in key strategic plans and documents, such as the Growth and Transformation Plan (I & II). For instance, the Growth and Transformation Plans (I & II) emphasise government financed mega infrastructure projects such as the 'renaissance dam' and industrial zones. The LRT represents one of the key mega projects that are designed to showcase the country's economic renaissance and the government's success in reducing poverty through accelerated economic growth and diversification.
Besides attractiveness, environmental friendliness was used as an important indicator for prioritising LRT as the most suitable form of public transport for Addis Ababa. Growing acceptance of the LRT among policymakers contributed to the 10 The city of Addis Ababa has been building condominium houses on the outskirts of Addis and established satellite neighbourhoods that can accommodate an average 15 000 households in areas such as Tulu Dimtu, Ayat, Bole Arabsa and Abdowolo. These new sites do not have transport mobility networks. And the city has been working hard to establish a mass transport service to and from the city centre to these new satellite sites.
fading away of BRT, although BRT would have been significantly cheaper than LRT. 11 This study found that arguments for LRT's capacity, attractiveness and environmental friendliness received more support than the initial investment cost that the LRT would incur. In other words, cost considerations were not given serious attention in the decision to go with the LRT. The pragmatic stance that the government had taken in terms of LRT's capacity and environmental friendliness, the government of Ethiopia's ambitious plan to make Addis Ababa a global city have been major drivers for choosing AA-LRT over BRT. Perhaps this fits into the 'urban fantasies' that decision-makers are made to believe, such as 'Ethiopia is rising' (Terrefe 2020;Watson 2009)

The politics around LRT infrastructure investment
Once the LRT was recognised as a viable mass transit option with significant political backing and potential financing, the next stage of the process was the selection of its mode of delivery and construction. A conventional approach of 'design and build' was taken to choose potential contractors with prior experience in LRT infrastructure development. The Ethiopian government issued a Request for Expression of Interest, and companies from Russia, Italy, Turkey, India and China responded. It is striking what the geographic footprint of international interest is when such calls are issued. The preliminary selection process identified an Italian company based on its technical competence and prior experience in designing and building railways following EU standards. However, Ethiopia was challenged at that time by its capacity to finance the project in its entirety. According to one informant: 'For Ethiopia, it was not possible to secure financial resources from European countries, as the timing was when the European financial crisis emerged.' 12 The rationale for the 'design and build' approach was twofold. First, it was clear that the LRT would require large financial investment that the Ethiopian government would not be able to generate on its own. Second, the country neither had prior experience in designing and building an LRT system nor adequately trained human resources for planning and executing such a project. Hence, the government deliberately sought a company that would bring both financial and technical resources. In the end, the government of Ethiopia decided to award the AA-LRT construction to the Chinese LRT turnkey developing contractor (China Railway Engineering Corporation [CREC]) after considering the company's experience in rail construction and its willingness to mobilise 85% of the financial cost of the LRT.
There was no debate or opposition to this decision due to the political backing of the LRT. The fast-track approach had several limitations. According to key informants, the 'design and build' approach by its very nature does not require precise planning in advance of construction, making it difficult for decision-makers to determine the actual project cost prior to its completion. For example, several amendments to the 11 ITDP estimates that BRTs are between four and twenty times cheaper than tram or light rail transit (LRT) and ten to a hundred times cheaper than metro. 12 Interview with representative of Addis Ababa City Road Authority (one of the board members for AA-LRT).
AA-LRT north-south and east-west routes were made during the construction phase, such as building additional bypasses, elevator shafts, which considerably inflated the investment costs. For instance, the construction of the LRT line is for a total of 31km, of which 23km is ground level rail; 7.33km is bridge construction, 0.9km underground tunnel. The 23km ground level line is fenced on both sides of the tracks. Decisions regarding the bridge construction, creating an underground tunnel, and fence to separate the rail tracks were made after the start of LRT construction, again leading to an increase in construction costs. Furthermore, lack of proper planning introduced unintended social and economic costs. For example, the decision to fence off the rail tracks from the main road disrupted the urban fabric by blocking free pedestrian movement and cutting cars off from existing intersections and turning points. The LRT project also caused traffic congestion near train stations and at the new roundabouts where the LRT line is at ground level. Placing the LRT line between opposite traffic lanes has led to poor pedestrian safety, as pedestrians must cross through moving traffic while walking to and from the train stations. Furthermore, the average distance between LRT stations is 800 metres and pedestrians must watch out for both the fast driving cars on the road and the LRT when crossing. This makes crossing the LRT lines from one side to the other side of the road difficult for older people and people with disabilities. According to one informant, 'Since we did not have the design before the construction, we were not able to make relevant decisions in a timely manner.' 13 On 12 March 2017, MailOnline reported that the half a billion-dollar investment in public transport failed to curb Addis Ababa's traffic problems, featuring the headline, 'light rail fails to fix Ethiopia's traffic troubles'. The report went on to say: Electric light railway tracks soar over Ethiopia's capital Addis Ababa, a rare example of mass transit infrastructure on the continent … But despite government promises, the roads below are still clogged with traffic 14 months after the light rail system's opening, and for many residents the city's network of overcrowded minibus taxis remain the only option. (MailOnline 2017) During data collection, the current author observed how the lack of prior planning on the design of AA-LRT routes created enormous traffic in all roundabouts where the LRT line is on the ground level and near all train stations (see Figures 2 & 3).
Economically, LRT led to financial loss to individual business operators whose businesses face the fenced roads. The fence disrupts the normal conduct of business as communities are split apart. Furthermore, construction of LRT lines resulted in economic waste by the government. A report by the Ethiopian Institute of Architecture, Building Construction and City Development (EiABC 2015) identified major waste and costs incurred during the AA-LRT development, including the following: 13 Interview with key informant from the Addis Ababa City Road Authority.
• A 40m wide road that was part of a BRT corridor and completed only in 2010 with a total cost exceeding 283.6 million birr (~USD 13.5 million) was demolished. • The construction of the 8km LRT line from the CMC area to Mexico roundabouts involved the relocation of an 8km main water pipeline that provides water to over a million people from the Legedadi Dam. The total cost of the relocation of the pipeline was about 421 million Ethiopian birr (~USD 21 million). • The AA-LRT line construction required relocation of electricity lines costing the Ethiopian Electric Power Authority 130 million birr (~USD 6 million).
Hence, a total of nearly USD 42 million was wasted that could have been used for other infrastructure or urban service provision. This could have been avoided if decision-makers had taken the time to assess the potential social, economic and financial consequences of the fast-tracked approach as well as if different actors had had a greater role in decision-making. Thorough planning using benchmarking could also have streamlined the process. For instance, countries such as Morocco have done well in constructing an efficient tram system instead of expensive LRT lines. As a result, the Casablanca tramway has been custom-built to facilitate free movement of passengers and significantly lessened traffic jams. 14 If the tram had been chosen, 'better traffic flow could have been achieved since tram lines/routes can be shared by other transport/mobility systems.' 15 For instance, in a guideline developed for the planning of BRT, Institute for Transportation and Development Policy (2007) asserts that BRT is a cost effective mechanism compared to LRT and can help cities to rapidly develop a public transport system while delivering quality service.

Accessibility, affordability and reliability of AA-LRT service
Although impossible to get an official report, attempts have been made to assess the experience of AA-LRT service users with the specifc emphasis on accessibility, affordability, and its reliability. Emerged qualitative data shows an average 135 000 residents use the LRT service per day. 16 A key informant reported: Although operated by the federal government, LRT daily ticket sells indicates that the 28 LRT cars transport 120 000-150 000 passengers daily. 17 The LRT also provides multiple benefits such as time-saving to passengers as the LRT lines are fenced, and reduction in pollution as the LRT consumes electricity. In addition to the effectiveness of the LRT in terms of its benefit, assessing the effectiveness of the AA-LRT as a mass transit service needs to take transport users' preferences, the reliability of services, convenience, safety, comfort, accessibility, and affordability issues into consideration (Imam 2014

Accessibility and affordability
FGD participants' experience in using the LRT and their overall assessment of the service is positive. In relation to its accessibility, study participants expressed their dissatisfaction that the LRT lines cover only certain areas. As shown above, in the AA-LRT map, the AA-LRT has only two corridors, these being, the East-West corridor that runs from Ayat to Tor-Hailoch and the North-South corridor that runs from Minilik. As a result of its limited connectivity to the different parts of the city, AA-LRT service users expressed their strong desire to see AA-LRT service in every direction of the city with better horizontal connectivity. Hence, its accessibility is only for people who are currently living near the LRT lines. This limits its accessibility for residents outside of the LRT lines. When it comes to affordability, the most recurrent themes emerged from all FGDs were that the LRT is affordable, fast and covers long distance. An observation on the transport fee shows that the train tariff fee is significantly lower compared to other transport services such as city bus and mini-bus taxi fees for similar distance. The tickets cost two to six birrs (USD 0.08 to USD 0.20) depending on the distance of the trip. A discussion participant noted: The LRT is cheap. I could have payed more than 10 Bir for the trip via taxi from where I live (Hayat area) to the centre of the city. Now, I only pay 4 Birr.

Reliability of LRT service
Service users' perception of service frequency in relation to travel and waiting times, as well as frequency of departure was assessed and findings from FGDs consistently confirm that the LRT is a reliable and efficient service.
In the city where you wait in line for an hour or two for taxis, the waiting time for the train is 15 minutes which impacts other parts of the city's economy positively.
Most FGD participants agree that the waiting time is on average 15 minutes during rush hours. However, there are times where service users experienced 40-minute delays (two FGD participants) and 50 minutes (one FGD participant). Furthermore, service users reported that they also witnessed a train arriving on time and not boarding passengers. This might be explained in terms of a dedicated train as 'Express Line' that might avoid some stations or a train that might have no space to board additional passengers. Furthermore, despite the dedicated power supply for the LRT, there were some incidents where the train stopped mid-trip and later announced to passengers that electric power was the cause. The challenge in relation to power disruption is also confirmed during key informant interviews. However, the integration of the city's major infrastructure and utility providers was instrumental to reducing the impact of electric power.

AA-LRT governance structure
The city of Addis Ababa is a federal government's capital and a city administration by itself. This offers opportunities for maximising the city's unique position in the country's urban development. Ethiopia's Urban Local Government framework guides the overall city development process and the city's legitimacy in collecting taxes and municipal revenues for its socio-economic development purposes. The city is mandated to provide major urban infrastructure and services. 18 The budget for these administrative functions is approved by the city council, and thus the city retains some decision-making power. This puts Addis Ababa in a better position than other Ethiopian cities.
Despite this presumed advantageous position, Addis Ababa is not fully exercising its administrative autonomy. The lines between its State and Federal functions are blurred due to excessive federal involvement. According to the World Bank and Cities Alliance ( According to key informants, the rationale for the establishment of an ERC independent of the old bureaucracy was threefold. First, the train network needs an independent regulatory body, which raises a couple of issues that need to be addressed by higher officials. Issues like whether or not to have one regulatory body for LRT and another for the ongoing and emerging national train networks that the country has embarked on. Second, the country did not have the technical capacity and experience in managing the day-to-day functioning of the AA-LRT. Third, the country had to find a way to cover the upfront infrastructure and operation costs. Hence, the ERC was established.
The political backing from the prime minister's office and the personal interest of the late PM in the LRT project warranted the creation of a management system independent of the normal government bureaucracy. The ERC, as an independent institution with strong political backing from the highest office in the land, was able to freely formulate and implement rules and regulations for the operation of railways in the country. The decision to have a single management and governance strategy in order to avoid red tape and unnecessary gridlock from the existing city and federal bureaucracy was instrumental in increasing the efficiency of the AA-LRT project. However, the establishment of ERC in 2008 introduced a gap concerning the role of the city government of Addis Ababa vis-à-vis AA-LRT management. A key informant from the city government of Addis Ababa noted: Our role was visible during the infrastructure phase of the project. It was the city administration that was responsible for providing land for the LRT lines. As a result, the mayor's office was active. We used to be involved and informed. Currently, in the operation phase, the relationship between the city 18 Basic urban infrastructure and services include: housing, land, water, electricity, telephone, road construction; road lights; drainage and sewerages; solid waste disposal systems; poverty reduction; maintaining vital statistics; certificates of marriage, birth and death; abattoirs; bus terminals and market places; combatting soil erosion, landslide disasters and environmental pollution.
administration and the AA-LRT project office as well as with the Board needs to be reviewed. Our present involvement is extremely minimal. 19 As a federal project, the ERC is currently working on AA-LRT phase 2, that is, expansion of existing lines. In the current operational phase, the role of representatives of the board, including the city of Addis Ababa, is minimised: 'ERC is managing both the AA-LRT as well as other railway projects of the country. Structurally, ERC is accountable to the federal government, not to the city Administration.' 20 In operating AA-LRT, the ERC is dependent on international external institutions for infrastructure design, operation and maintenance. The ERC is currently functioning with the support of two Chinese companies and a Swedish Road consultancy firm. The China Railway Engineering Corp (CREC) built the AA-LRT infrastructure. CREC is also partially involved in the current operation phase of AA-LRT. Another Chinese company called Chenzen Metro, which is experienced in the operation and maintenance of LRT, is contracted to manage the operation and maintenance of AA-LRT for the next five years. Significantly, the Swedish Road consultancy was hired to evaluate the work of the Chinese companies in all phases. Whether the AA-LRT will remain as a federally administered project for the foreseeable future is not clear. Questions remain whether the federal government will at some point hand over the day-to-day management and operation of the AA-LRT to the city administration. If the latter were to happen, city authorities will need time to prepare the appropriate legal, regulatory and management systems.

AA-LRT institutional capacity building
There were no railway engineers in the country prior to 2012. 21 A critical look into the AA-LRT institutional setup shows that the nature of the project contributed to the government's decision about how the AA-LRT should be managed. As a new industry, LRT requires specific competences in order to successfully manage its day-to-day operations. The LRT idea was embraced in the absence of relevant professionals who could plan, implement and manage the project. As one informer put it, 'At that time [during the inception of LRT], the country did not have professionals on railway engineering. We also had no experience in managing LRT.' 22 The establishment of ERC as an independent institution is not adequate to manage the AA-LRT in a sustainable way. Considering the complete lack of relevant skills in railway engineering and management expertise, the ERC relies on Chinese companies in managing the entire LRT design and operation process. Currently a Chinese company, Chenzen Metro, works under the AA-LRT project office to manage the day-to-day operation and maintenance of the AA-LRT. At the same time, the ERC decided to establish a Railway Engineering Institute in the Technology faculty of Addis Ababa University in order to train engineers to be responsible for the management and maintenance of the LRT in future. According to one informant, 'There were no railway engineers in the country prior to 2012.' 23 This is a commendable move, as it would reduce the ERC's dependence on Chinese management and maintenance services in the long run. Hence, a two-year Master's of Science programme focusing on three fields of study at Addis Ababa University's technology faculty was opened. These fields include: civil railway engineering, mechanical or rolling stock engineering, and electrical engineering for railway systems. After signing a memorandum of understanding with the ERC, the institute enrolled 540 students in four cohorts. Since its establishment, a total of 58 students graduated in the railway civil, 55 in the mechanical or rolling stock, and 27 in the electrical engineering fields. 'All students are employees of ERC and will continue to work for ERC after graduation.' 24 This initiative demonstrates that Chinese experts are not meant to remain in the agency indedifinitely. Instead, local experts will progressively occupy operational and management roles in all aspects of the service.

Discussion
The political economy of Ethiopia's developmental state approach to governance clearly limited involvement of diverse actors during AA-LRT inception and implementation. The former Prime Minister Meles Zenawi's sole decision to impose the AA-LRT system on the city, despite the city administration's choice and perhaps the expert's well designed concrete plan to implement BRT, was a political decision. The government's grip on decision-making leaves no room for other visible and invisible policy actors to contribute to policy decisions. This reflects the uncontested 'democratic centralism,' a driving principle of the the ruling party that guides its strategic day-to-day activities (Bach 2012; Vaughan & Gebremichael 2011). There is a notable absence of a 'level playing field' that would offer favourable conditions for organisations and civil society interest groups to negotiate and debate the merits of an urban project on the scale of the AA-LRT to maximise its benefit and reduce the negative social and economic impacts on Addis Ababa residents. The AA-LRT's emergence as the only viable transit system and the decisions taken during its implementation illustrate the idea of 'developmental patrimonialism ' (Kelsal et al. 2010 AA-LRT's governance structure and institutional system also demonstrate a topdown institutional arrangement and the decision-making based on embedded consensus within EPRDF's elite circle. As a result, open contestations and entertaining alternative policy ideas on how to provide an environmentally sound and economically beneficial mass urban transit system that would have averted the socio-economic chaos the AA-LRT project engendered were not allowed.

Top-down institutional setup
The AA-LRT demonstrates how a 'developmental state' can usurp the functions of local governments in the interest of optimising the delivery of urban services from its perspective. While the government has committed itself to decentralising decision-making to lower tiers of government, in reality, it bypasses these same local institutions or fails to consult them fully. Hence, the devolution of autonomy at local level government institutions remains unrealised. It would be fair to suggest that the requirements of strategic external actors, such as the Chinese investors, reinforce this pre-existing tendency. The federal government's decision to institute an independent railway corporation with a separate structure to be managed by the board rather than the city government of Addis Ababa, appeared as a pragmatic decision rather than adhering to and translating the country's established urban governance principles into practice. The narrative of good urban governance in key urban development initiatives is merely an alibi to cascade top-down decisions through embedded consensus at all levels. In this regard, the decision that was made in 2007 by the EPRDF elites to elevate the LRT service as the sole option to meet the city residents' transportation needs and showcase Ethiopia's 'rise' blurred existing urban governance structure and systems. Despite the city of Addis Ababa's autonomy and mandate to provide urban services such as a reliable, fast and affordable transport system, the top-down decision from the central government to introduce LRT has not to date created any controversy. The swift decision to introduce the LRT over the city's own preference to introduce a pilot BRT service perfectly shows the hierarchical power relationship between the federal and city governments. Furthermore, the federal government's sole decision on loan agreements with Exim Bank of China shows the limited autonomy of the city administration in terms of initiating and implementing socio-economic related development plans.

Urban governance based on 'embedded consensus'
The absence of any complex or dynamic interactions between policy actors is obvious within the 'developmental state' or 'developmental patrimonialism' approach to governance. What is interesting in the AA-LRT case study is that the AA-LRT's emergence faced no contestation, and its implementation did not raise debate between the national/federal and local level governance structures. There is no public document showing any conflict between the federal and city government bureaucrats throughout the entire political process. The consensus surrounding the introduction of LRT was seemingly absolute. The city administration expressed no concerns over the amount of investment neither for the initial LRT infrastructure development nor the ongoing operation and maintenance costs. In addition, none of the authority figures/ elites questioned the relevance of the project as it led to financial waste of nearly USD 42 million. Although the loan agreement has its own grace period and maturity time, the city government of Addis Ababa was not involved in any of the negotiations and agreements pertaining to repayment of loans. Furthermore, the city administration did not undertake research into the feasibility of the project in terms of sustaining the service through revenue generation. The level of consensus and cohesiveness between the federal government and the city administration resonates' with Kelsall's (2011) neo-patrimonial 'embedded bureaucratic autonomy' of the federal government excluding other non-state actors including the private sector. The cohesiveness of the federal and the city government of Addis Ababa show the effectiveness of the regime's commitment to govern based on the principle of democratic centralism where the State and ruling party are one and the same. Top party leaders are the leaders of the country and focus only on State effectiveness in both policy decision-making and service delivery process. Hence, the structures that are instituted around the three tiers of governance (federal, regional and local level) are more symbolic than substantive.

Conclusion
The political economy of the state-society relationship within 'developmental state democracy' operates within the deeply entrenched bureaucratic centralism through which patrimonialism is harnessed. Hence, the system of governance does not promote dynamic interactions between various actors in line with the principles of good urban governance. Instead, the overriding concern is State effectiveness, which operates through embedded consensus and this seems the most likely way forward. AA-LRT's contribution in providing transportation for a significant number of Addis Ababa residents is commendable. Despite the effectiveness of the Ethiopian state in achieving the completion of AA-LRT construction in a few years to fit into the 'Ethiopian Renaissance' narrative, I argue that the absence of citizens, civil society organisations, and Ethiopian business involvement in the decision-making process is responsible for a huge amount of wastage. Had an authentic and genuine process of decision-making involving the federal government and the city authorities existed, a more cost-effective alternative form of transport services would have been proposed and debated without the federal government having to impose its will on the city government. The governance systems in place narrowed the space that could provide democratic checks and balances as well as feedbacks, which are fundamental in facilitating the social and financial efficiency of the LRT.
Furthermore, the government's ambition to reposition the city of Addis Ababa as a global city perhaps fits into the completely imported urban fantasies of which decision-makers were convinced in pursuing the 'Ethiopia is rising' and 'eco-friendly' project that also provided mass transit service (Terrefe 2020;Kloosterboer 2019;Watson 2014). As a result, local realities were ignored, and emphasis was more on the symbolic significance of AA-LRT as a mega city project that had helped the country get international attention. Moreover, the project has helped to provide affordable and reliable transport service although further research is needed to explore how the expensive modern mass transit has helped ease the city residents' transport need. Obviously, the desire to have modern transportation prevented decision-makers from averting LRT-induced financial wastage that could have been spared for other infrastructure developments. In fact, further research is needed to actually determine the cost and benefit of AA-LRT. The persistence of the government's decision to expand the LRT lines 25 without critically examining the unintended negative socioeconomic impacts sufficiently explains the lack of transparent and accountable urban governance. Hence, the overall governance systems and structures in Ethiopia do not provide democratic checks and balances as well as feedbacks, which are vital to increasing the social and financial efficiency of LRT. External actors and their agendas are folded into these overriding dynamics, they are not determined by external actors.

Institutions
Number of participants

Government institutions* 12
Civil society groups 3 Research institutes 2 Focus group discussants 8