The policy of job creation in Kolfe Keraniyo

In much of the global South, recent years have witnessed a renewed interest in entrepreneurship and microfinance to support the formalization and legalization of the economy, and hence to combat poverty and sustain growth and development (Amoros Reference Amoros2009; UN 2013; World Bank 2013). While the impacts of these initiatives cannot be generalized, critical studies have highlighted that marginality and destitution should be regarded beyond the framework of economic empowerment, as they are entrenched in broader power relations reproducing dynamics of social, economic and political exclusion and inclusion (Roy Reference Roy2010; Bateman Reference Bateman2012). Other studies have noted that the ideas underpinning the support given to microfinance and small businesses should rather be regarded as part of wider processes restructuring global capitalism that reflect new strategies of market liberalization (Benerjee and Duflo Reference Benerjee and Duflo2011; Weber Reference Weber and Fernando2006).

Recent studies conducted in Ethiopia found that MSEs that were designed to extend the ruling party's structures of political mobilization in the urban society (Di Nunzio Reference Di Nunzio2015) have played an important yet contradictory role in shaping the future for unemployed youth (Mains Reference Mains2012), that their success has largely depended on the closeness of beneficiaries to local administrative structures (Hundera Reference Hundera2014), and that they have had a limited socio-economic impact on the majority of beneficiaries (Gebre-Egziabher and Aynew Reference Gebre-Egziabher and Aynew2010), especially when businesses were operated by women (Bekele and Worku Reference Bekele and Worku2013).

In rapidly transforming Ethiopian peri-urban settings, the promotion of entrepreneurship initiatives underpin important dynamics of negotiating public authority. Driven by expanding services and the agricultural sector, Ethiopia experienced a yearly average growth of 10.8 per cent between 2003–04 and 2012–13, compared with a regional average of 5.3 per cent (World Bank 2015). In the context of the country's rapid growth, in recent years peri-urban Addis Ababa has experienced profound socio-economic transformation mainly driven by large infrastructural projects, including low-income housing known as condominium, as well as the rapid expansion of the manufacturing industry.

Kolfe Keraniyo is located on the western edge of Addis Ababa, covering a large tract of the border with Oromia Regional State. From an administrative perspective, in 2003 Addis Ababa was restructured into ten sub-cities (Kifle Ketema), with Kolfe Keraniyo currently being the most densely populated.Footnote ⁴ Since at least the 1970s, the area has served as an important business hub, attracting a significant number of people from different parts of the country, especially from southern Ethiopia. This reflects the sub-city's current heterogeneous ethnic and religious composition.Footnote ⁵

Immigration accelerated from the mid-2000s, mirroring broader dynamics of economic restructuring as well as generating growth in informal business activities.Footnote ⁶ Kolfe Keraniyo is currently characterized by migration to and from both the inner city and rural areas.Footnote ⁷ The former is mainly driven by the rising cost of living in more central areas, as well as government policies of relocating low-income people from the inner city (see also Yntiso Reference Yntiso2008). The latter is marked by an increasing trend of rural out-migration of people in search of job opportunities and more reliable public services in the expanding areas of the city (Tadele Reference Tadele, Pankhurst and Piguet2009). This migratory pattern is not always permanent – people often commute from neighbouring ‘rural’ localities, or take on seasonal and temporary jobs – and it does not necessarily involve the relocation of the entire household: often only one member moves in search of employment. A rapidly increasing informal economy in the sub-city formed the specific backdrop to the policy of promoting MSEs against a broader context in which the formalization of informal businesses became a central priority of the political elite's developmental agenda.

Physical, social and economic boundaries in Kolfe Keraniyo are by definition ‘unstable’, as they are constantly redefined by increasing rural–urban interactions. Arguably, in these fluid contexts public authority is constantly redefined along the lines of access to resources and the construction of multiple identities, as well as being shaped by policies and development interventions targeting specific individuals or groups. The political processes underpinning the promotion of MSEs in Kolfe Keraniyo constitute a privileged entry point – or, to use Hagmann and Péclard's (Reference Hagmann and Péclard2010) term, ‘negotiation arena’ – where some of the defining features of the developmental state project are more pronounced, as well as reflecting its inherent tensions and contradictions.

MSEs and the construction of the developmental state in Kolfe Keraniyo

The emergence of MSEs as an important policy tool to combat unemployment has to be contextualized within the country's post-socialist transition since the 1990s, and more specifically in the reformulation of the long-term development strategy after the mid-2000s (Aalen and Tronvoll Reference Aalen and Tronvoll2009; Lefort Reference Lefort2012; Di Nunzio Reference Di Nunzio2014b; Bach Reference Bach2011). After the war with Eritrea (1998–2000), the purge of the so-called ‘left wing’ of the inner circle of the EPRDF after 2001, and the outcomes of the controversial 2005 general elections (Lefort Reference Lefort2007), the long-term strategy for economic development was reformulated towards what the political elite termed ‘developmental capitalism’ (Lefort Reference Lefort2012; De Waal Reference De Waal2013; Clapham Reference Clapham2006).

The former leader Meles Zenawi felt that promoting a strategy of rapid economic growth through strong state interventionism was the only way for Ethiopia to achieve broad-based socio-economic development and at the same time to keep the plurality of ‘people, nations, and nationalities’ together in the Ethiopian state (Turton Reference Turton and Turton2006: 1). This strategy was conceptually juxtaposed with experiences of neoliberal market liberalization that were considered to have performed poorly from an economic and social perspective in much of Africa and the developing world (De Waal Reference De Waal2013). Among the most tangible transformations, the political reform entailed the expansion of local bureaucratic structures, as well as a massive campaign of recruitment of EPRDF party members, which, as noted by Di Nunzio, went from ‘760,000 in 2005 to more than 4 million in 2008’ (Reference Di Nunzio2014b: 420).

In this context of profound transformation, the developmental state had both a material and an ideological value. From a material perspective, it embodied the political elite's attempt to promote a strategy of fast economic growth by means of the collective mobilization of the plurality of people of Ethiopia. As its ideologue, Meles Zenawi admitted: ‘I am convinced that we will cease to exist as a nation unless we grow fast and share our growth.’Footnote ⁸ The ideological dimension of the developmental state reflected the performativity of the political project behind it – ‘changing the mind of the people’ – which advocated for the creation of positive experiences of lasting change. The ideological and material logics of the developmental state underpinned three important aspects of the promotion of MSEs in Kolfe Keraniyo.

A first element was a policy of regularization aimed at formalizing existing economic activities, creating new job opportunities, and enforcing the local structures of social and political control (see also Di Nunzio Reference Di Nunzio2014a). The job creation strategy was based on a regularization policy aimed at combating behaviours labelled ‘social evils’, for instance ‘prostitution, illegal street vending, smuggling, and illegal tree cutting’.Footnote ⁹ As argued by one local official, the policy of regularization was motivated by the current trend of rapid in- and out-migration in the sub-city and entailed the promotion of MSEs to formalize and regularize a growing peri-urban population and to complement other government programmes, such as the resettlement and housing policies.Footnote ¹⁰ The policy can be regarded as both a strategy for economic development aiming to ‘turn unemployed people into businessmen’Footnote ¹¹ and a tool for the social and political control of a rapidly increasing urban population. An important aspect of the MSE policy in Kolfe Keraniyo was to provide for the registration of the unemployed in the kebele employment lists, and hence ‘to keep track of who is moving into the sub-city’.Footnote ¹²

A second element was the policy emphasis on ‘group first’. In Kolfe Keraniyo, paying back a loan in a newly government-sponsored MSE was primarily the responsibility of the group, which provided a guarantee against individual members by establishing a kind of moral and political – rather than strictly legal or financial – sense of accountability.Footnote ¹³ The centrality of the group to drive socio-economic transformation reflected an emphasis on the population's collective, rather than individual, participation in the development process. This emphasis was regarded as a feature of developmental capitalism that distinguished it from neoliberal strategies.Footnote ¹⁴ The notion of collective participation must be contextualized within the historical trajectory of the political elite, who originated from two distinct revolutionary moments in the 1970s and the 1990s; this trajectory informed an approach to the national question based on the self-determination of the plurality of groups and people of Ethiopia (Vaughan and Tronvoll Reference Vaughan and Tronvoll2003: 15; Markakis Reference Markakis2011). MSEs embodied the notion of collective participation to the extent that the business model underpinning this government strategy was organized around the work of small groups rather than individuals. This aimed to pull together the joint efforts of poor people for common business projects and, at the same time, to facilitate access to financial services, primarily credit, by means of group collateral (Gebre-Egziabher and Aynew Reference Gebre-Egziabher and Aynew2010).

A third dimension related to the project for the socio-economic transformation of individuals’ attitudes towards saving. ‘Saving first’ – which is about instilling an attitude to saving through training and other activities – was a central element in the promotion and consolidation of MSEs in Kolfe Keraniyo. As explained by local officials, the socio-economic transformation of the sub-city is impossible without ‘a structural change in the way people think about saving and approach business’.Footnote ¹⁵ A similar concept informing the promotion of MSEs was the notion of kaisen. Kaisen, derived from Japanese, means quality production in the most efficient way and at the lowest cost possible.Footnote ¹⁶ In Kolfe Keraniyo, the idea of kaisen was taught during vocational training before the formation of MSEs, and particularly at the stage when business plans were being formulated.

The promotion of MSEs in Kolfe Keraniyo reflected key mechanisms in the construction of the developmental state at the local level. Individual and societal transformation cannot be separated conceptually from rapid economic growth and broad-based development (Zenawi Reference Zenawi, Noman, Botchwey, Stein and Stiglitz2012). This project underpinned a political plan for transformation that emanated from a policy of regularization based on collective participation in the country's socio-economic development, and, at the same time, on a change in attitude towards saving, business and development more broadly among individuals.

Combatting unemployment through small businesses

At the time when my fieldwork was conducted, MSEs were an important pillar of the Ethiopian government's poverty reduction strategy, and they complemented other initiatives, including the Productive Safety Net Programme, the Food Security Programme, the Household Asset Building Programme and the Voluntary Resettlement Programme. MSEs also embodied the political priority of formalization and regularization of both urban and rural economic sectors, with an emphasis on business development to lay the foundation for medium and large export-oriented enterprises in the long run.

MSEs were first regulated with the publication of the 1997 National Micro and Small Enterprises Strategy and the establishment of the Federal Micro and Small Enterprises Development Agency. However, it was only after the political shift of the mid-2000s discussed above that a strategy of employment creation that drew heavily on the promotion of MSEs was put into practice. Informants noted that, in Kolfe Keraniyo, the aftermath of the 2005 elections was a period of profound transformation in which the presence of party/government structures intensified significantly. Job creation through MSEs and the political mobilization of the peri-urban population – particularly unemployed youth – became overarching government priorities to prevent and mitigate growing political instability and social unrest. This reflected the fact that unemployed youth – along with women – were cast as the main priority group to be targeted by government-sponsored MSEs from the very beginning.

In Kolfe Keraniyo, MSEs were organized by the sub-city MSE office, as well as by local promotion offices at wereda (district) and kebele levels. Local promotion offices functioned as employment agencies, and, coordinating with other offices, they compiled employment lists and recommended priority areas where potential new businesses were needed. Another important responsibility of these offices was to provide technical and vocational training on a wide range of topics. While technical training sessions essentially comprised workshops teaching specific technical skills, vocational training reflected the transformational dimension of the political project behind MSEs. Vocational training was primarily aimed at promoting changes in individual attitudes towards saving (‘saving first’), as well as inculcating the primacy of group work over individual work (‘group first’). At the time when my fieldwork was conducted, a wide range of local institutions were constantly busy organizing vocational training sessions every few days. One vocational session I attended went on for several hours, with the convener repeatedly espousing the benefits of working in groups rather than individually – the ‘only way possible to make it successfully is through mutual help and by pulling together resources’ – as well as the ethical importance of saving money as ‘the only way possible to build a future without poverty’.Footnote ¹⁷

Depending on the number of members, MSEs were categorized as individual, as shared companies, or as small enterprises or cooperatives (more than ten members). Although in practice it was quite difficult to distinguish between cooperatives and small enterprises, a main difference in Kolfe Keraniyo was the requirement that enterprise members had to provide 20 per cent of the initial capital themselves in order to be eligible for an 80 per cent loan from an MFI. Another difference was the interest rate applied to the loan, which was sometimes lower for cooperatives.

Equally important is the concept of ‘graduation’. Promotion offices monitored existing businesses, and, depending on the enterprise's capital, ‘graduated’ successful enterprises. A business should achieve graduation within five years of its inception, meaning that it was then expected to be able to handle licensed activities autonomously. Enterprises are divided into four categories depending on their capital: micro or start-ups (up to 50,000 ETB); small (50,000–200,000 ETB); medium (200,000–500,000 ETB); and industrial (above 500,000 ETB). MSEs operate in priority sectors dictated by the Kolfe Keraniyo sub-city promotion office; these include manufacturing, construction, trade, services and urban agriculture.

Beneficiaries, leaders and groups

As explained by a local official, vocational training was also the main tool through which promotion offices registered unemployed people and targeted new beneficiaries.Footnote ¹⁸ Registration was carried out through a ‘door-to-door’ model and implemented by both the MSE office and other kebele offices.Footnote ¹⁹ The ‘door-to-door’ model highlighted the extent to which the vocational project of the developmental state was meant to have universal reach and target the entire peri-urban population, and the way in which it constituted a main arena within which state–society relations were negotiated.

Awareness creation was based on what local officials in Kolfe Keraniyo referred to as the three ‘development alarms’:Footnote ²⁰ that is, how different institutional levels engaged the unemployed and involved them in development activities. The ‘people's community wing’ referred to MSE beneficiaries who performed advocacy activities to raise awareness about the benefits of establishing an MSE. The ‘government wing’ included all the offices providing training and advice on how to establish an MSE, including MFIs, technical and vocational schools, and promotion offices. The ‘political party wing’ covered members of the local EPRDF branch who recommended people who could establish an MSE, as well as sectors and specific activities. In practice, the boundaries between the three wings were much more blurred. As expressed by one local businessman: ‘The very same people come on behalf of the EPRDF, of the MSE office, or as your neighbour … depending on what they have to ask you.’Footnote ²¹

As noted by one informant, the ‘door-to-door’ model, which entailed the deployment of all three wings to advocate for the establishment of MSEs, reflected both continuities and discontinuities with previous job creation strategies. ‘What has not changed is the pressure of the government to control and promote business … The difference is that now the poor are expected to become businessmen in a short time.’Footnote ²² The three wings acted on several institutional levels, starting below the wereda. The organization of these sub-units varied significantly from context to context, and no official document provided comprehensive information about how they interacted with the formal administrative and party structures. Starting from the lowest level, in Kolfe Keraniyo these included the block (no more than fifty people), the mender (composed of two or three blocks), and the katana or sub-wereda. A central characteristic of the Ethiopian model is that the local institutional structure was conceived to include and mobilize the entire population, although it was not without its contradictions, and thus to extend the reach of the central state/party structure (Di Nunzio Reference Di Nunzio2015; Chinigò Reference Chinigò2015).

Other institutional constellations mediating access to MSEs included the micro-institutional structures of the ‘one to five’ networks (and le hammest) and development groups (ye limat budin). These networks were the basis on which groups were formed for various purposes and had both political and technical value (Lefort Reference Lefort2012). The ‘one to five’ was a group of six people where one was the leader or model and was responsible for passing relevant information from government offices to the other five, coordinating development activities, and collecting taxes. The development groups comprised twenty-five to thirty people and kept a record of individuals seeking new jobs, drew up unemployment lists, and identified potential jobs or professions where there was a shortage in a specific area. Development groups also performed other activities relating to neighbourhood security, political advocacy for the ruling party and public works. Both of these networks had the aim of ‘bringing and sustaining change through good governance activities’Footnote ²³ and explained ‘why today Addis is peaceful’.Footnote ²⁴ More broadly, these state-controlled networks played an essential role in terms of surveillance of the peri-urban population. While this reflected an effective ‘outsourcing’ of state control to party members and affiliated individuals, they also played a key role in mediating access to MSEs and service delivery.

In Kolfe Keraniyo, new MSEs were sometimes formed in conjunction with ‘one to five’ networks. In four cases analysed in wereda 8 and 14, recently graduated students were first asked to form ‘one to five’ networks, and only then to submit their application to start an MSE to the promotion office. When an MSE had fewer than five members, the ‘one to five’ network was formed in conjunction with a neighbouring business. When an MSE had more than six members, more networks were formed as a consequence. In some cases, ‘one to five’ networks formed for MSE purposes did not necessarily require members to be neighbours. More broadly, these networks operated performance checks on individual businesses and were meant to be a forum for discussion on saving and entrepreneurship.

The ‘one to five’ networks and the development groups also performed the very important role of keeping track of jobless people in each neighbourhood and recommending new potential beneficiaries to the MSE Promotion Office. As noted by an informant, ‘Members of the “one to five” networks and development groups are your witnesses and can recommend you to the promotion office to get loans and establish MSEs.’Footnote ²⁵ According to another informant, the role of these networks was to provide an opportunity for people to support each other, including with regard to finances, sharing knowledge and experience about their businesses, instilling the habit of living together, sharing work responsibilities, providing security in the neighbourhood and maximizing business profits.Footnote ²⁶

Central to all these institutional configurations was the role of leaders. Each network had one leader who bore the moral, practical and political responsibility for the group members on behalf of higher institutional levels. The leaders of blocks and menders were individuals with a good reputation in the community and were elected through internal mechanisms. At the katana level, leaders had a much more prominent political orientation and were involved in the party organization. Group leaders within the ‘one to five’ networks and development groups were usually party members and were socially, economically and politically responsible for the other members. In the case of MSEs, the key duty of leaders was to collect money to pay back loans. Leaders were meant to facilitate and coordinate knowledge sharing and information flows from and to local government institutions. And although leaders were normally chosen directly by other members, the kebele retained the power to dismiss the proposed leaders.

The constitution of groups, networks and associations of various kinds was a defining feature of the developmental state project for the political mobilization of the peri-urban population. The ‘three wings’, together with other local institutional configurations, constituted important arenas in which state–society relationships in Kolfe Keraniyo were negotiated.

MSEs and microfinance

The institutional arrangements providing access to credit for MSEs illustrated further political processes underpinning the negotiation of state–society relationships in Kolfe Keraniyo. The primary channel through which MSEs had access to credit was through local MFIs (see also Amha and Alemu Reference Amha and Alemu2014). MFIs, whose growth in Ethiopia by the time when my fieldwork was conducted was remarkable, were de facto public institutions with capital almost exclusively controlled by central government (Deribie et al. Reference Deribie, Nigussie and Mitiku2013). In Kolfe Keraniyo, the main MFI was the Addis Saving and Credit Institution (ASCI), with the government holding 98 per cent of the shares. Interest rates offered by ASCI were not significantly different from those offered by banks, and normally ranged between 12 and 15 per cent. However, banks seldom granted loans to MSEs unless proper guarantees were in place.

As explained by one branch manager, MFIs evaluated and approved the business plans of MSEs in conjunction with the promotion office, normally based on the criteria of feasibility, skills required and guarantees.Footnote ²⁷ There were different types of guarantee an MSE could provide, including private collateral, group guarantees, a letter of guarantee from a public institution and the provision of seed money to make up part of the initial capital. The vast majority of newly established MSEs obtained credit through the group guarantee system (or group collateral).Footnote ²⁸ In such a system, the group was liable for the collective amount of credit provided on behalf of individuals. Group collateral illustrated a key mechanism operationalizing the notion of ‘group first’. It reflected the ways in which a wider sense of political and ethical accountability could not be conceptually separated from the financial accountability underpinning the individual repayment of a loan. As noted by one informant: ‘MFIs are state-controlled institutions through which the government channels credit … to promote a business model based on political groups … that in one way or another are connected to the EPRDF.’Footnote ²⁹

Similarly, other aspects of the relationship between MSEs and MFIs reflected the notion of ‘saving first’. Newly established MSEs were required to contribute an initial 20 per cent (from savings or seed money) of the capital they intended to borrow (see also Hundera Reference Hundera2014). In addition, individuals were required to save a variable amount of money every month – known as compulsory saving – before and after their MSE was granted a loan. Individual saving was meant to demonstrate the willingness and ability to save a certain amount of money for several months consecutively, as well as raising cash to be used as collateral. In addition, as noted by a local official, saving was meant ‘to change the habits of people towards their future … Saving is transformative because it changes the mind of people.’Footnote ³⁰

The notions of ‘saving first’ and ‘group first’ embodied in the mechanisms of credit provision in Kolfe Keraniyo played both an economic and a disciplinary role. The policy emphasis on group saving reflected the political elite's attempt to encourage virtuous cycles of micro-capital accumulation in order to achieve the developmental state project for the socio-economic transformation of Ethiopia. The disciplinary role reflected the fact that such a project was expected to promote a change in unemployed people's attitudes through their participation in a number of state and party political networks. Ultimately, individual action was subordinated to the collective effort – directed by the state – at transforming the Ethiopian nation and sustaining economic growth.

‘Keeping good relations with the government’: individual experiences of group work in Kolfe Keraniyo

The promotion of MSEs embodied important aspects of the territorialization of the power of the developmental state in a peri-urban milieu. However, this does not mean that such a process was always top-down. Experiences of group work in MSEs highlighted that people and social groups actively contributed to the making and unmaking of the developmental state by negotiating, refashioning or challenging the norms and rules with which they were confronted. Far from constituting a coherent process of structural refashioning, these dynamics took the form of diverse strategies of coping and subverting the normative framework of the developmental state. The ultimate effect was that they contributed to destabilizing the peri-urban space, for instance by promoting the informalization of business activities. This revealed that state authority in peri-urban settings was produced and challenged at the same time.

In line with recent studies, many of the MSEs observed in Kolfe Keraniyo were not successful from an economic perspective (Di Nunzio Reference Di Nunzio2015; Gebre-Egziabher and Aynew Reference Gebre-Egziabher and Aynew2010). Through MSEs, members were at best able to obtain basic subsistence, but usually the businesses were not financially sustainable. One of the main reasons why MSEs often failed to become successful businesses was that they felt like impositions from above rather than being genuinely demanded by the people involved. For instance, some respondents contended that they did not want to get loans because they were afraid of being in debt. In an interview, one woman articulated the following paradox:

In a focus group discussion about the challenges of group saving, participants lamented that paying back the money in time was often very difficult. ‘Saving individually is much less risky … and what does saving mean in the end?’Footnote ³² In general, a large majority of respondents were very sceptical about establishing groups and expressed their preference for individual businesses. ‘We work in [a] group because the government wants us to work in groups … but you don't make money, you keep good relations with the government.’Footnote ³³ This is further corroborated by cases in which MSEs were established with initial capital provided by local or international NGOs, where the formation of groups was a prerequisite to receive the initial loan. As one woman put it: ‘I did not know the other women before and I don't know what kind of business we will do together … the reason why I joined the micro [i.e. MSE] is that they give us some money and you can get other loans.’Footnote ³⁴

From the perspective of local officials, the main challenges of the MSE policy were ‘changing the mind of people when it comes to saving’ and ‘creating a saving culture’.Footnote ³⁵ One official from the Women and Youth Affairs Office contended that loans are often spent first on household consumption, and much less frequently as investments for productive activities. From the perspective of the poverty reduction strategy, loans and group businesses in Kolfe Keraniyo often performed the function of a social protection system – something that was largely absent in the peri-urban context – rather than creating real opportunities for economic growth.

However, a narrow focus on financial and economic viability obscures the social and political significance of the emphasis on MSEs. MSEs played a very important role as they opened up new opportunities for access to credit for individual members, for the redistribution of resources from participants to party and government networks, and for assistance from government institutions and development organizations. Arguably, the shift in the intended objective of the job creation policy reflected the fact that peri-urban dwellers actively contributed to the making and unmaking of the developmental state. Many of these issues are illustrated in the life histories of peri-urban dwellers who have been involved in the constitution of groups, such as the following.

Z. M. is a member of an MSE for which the initial capital was provided by the government. The MSE was established some years ago to gather together women selling tree seedlings and vegetables who had previously been engaged in illegal tree cutting. Despite being in place for a long time, the association was not very successful from an economic perspective: with the profits, members were barely able to cover their current expenses. Nonetheless, Z. M. worked three days a week from 6 a.m. to 3 p.m. for the association. Ironically, the motto of the association was ‘If we work we will have hope’. But why was she a member of the MSE? She explained that the association was not for making a living; rather, through the association she was able to get individual loans for her own ‘real’ business. The association provided the group collateral that each individual member needed in order to apply for an individual loan at the local MFI. The MSE was organized in groups of eight women for credit and loans purposes. Each group had a leader, and she was the leader of her group. To make her living, Z. M. bought vegetables from the main market in Addis Ababa (Merkato) and sold them in her local neighbourhood market. She went to Merkato whenever she was not busy with the association. She worked four days a week in her own business and said she was able to make a profit of about 25–30 ETB per day.Footnote ³⁶ An interesting aspect, however, was that she had been taught to draw up a daily business plan in the training about saving she received from the kebele. Her profits did not include her daily salary, which she claimed was 15 ETB; as she explained, ‘This is the cost for the labour as if someone else would do the business on my behalf.’ She therefore was taught to consider herself an employee in her own business. The other peculiarity was that she claimed she could not use the current day's profits, but only the profits she had made the day before. The reason for this was that she had been taught that the day's profits had to be considered as insurance against any inconvenience or problem happening during the day. ‘I can't use the profit I made today. I have to use the profit I made yesterday. The profit of today is the insurance in case I'll get sick.’ She was very proud of herself and extremely happy about the training on saving that she received every few months at the kebele. ‘Saving transformed me, and it is important because [it] helps you when you are old.’Footnote ³⁷

The story of Z. M. – which can be seen as a success story – elucidated both the opportunities and the tensions of the project for economic development based on the promotion of small businesses and groups in Ethiopia. The main contradiction of many group initiatives was that they were not productive, with people often engaging in free labour. Group work, especially in MSEs, was a driver of successful businesses only in a minority of cases. People often complained that they would prefer to work individually but the only way to get support was by establishing groups. Complaints included that it was often difficult to ‘put different heads together’, the free-riding of some group members, the inflexibility of promotion offices when deciding what businesses to undertake, and the time spent in training vis-à-vis actual work. However, as there were limited opportunities to succeed in other ways, participation in MSEs was an entry point to other opportunities provided by the local state/party structure. Participation in government initiatives played an important political and social role and was often the only possibility many peri-urban dwellers had to conduct other economic activities and to make a living. Even if this came at the cost of working for free for several days every week, as the case of Z. M. shows, this opened up opportunities for access to credit, licences and other redistributive mechanisms. Ultimately, efforts at formalizing and legalizing the peri-urban economy through the promotion of MSEs had the paradoxical effect of engendering a tendency towards the informalization of other productive business activities.

More broadly, the negotiation of the local developmental state operated through the construction and segmentation of time and space, and hence constantly redefined what was legal and illegal, as well as formal and informal. The following example of street vending helps elucidate this concept.

F. S. is a street vendor in a very crowded street in Kolfe Keraniyo, selling petty merchandise that he keeps in a wooden tray with a neck strap. Despite the fact that he had no licence for his business, he explained that his activity was neither legal nor illegal; this depended very much on how and where he sold his goods. Selling to the cars queuing on the street was strictly illegal; he explained that this was often the most remunerative activity, but if he were caught by the police, they would normally take away his products and ‘you'll have a very bad day’. Conversely, it was not really legal to conduct business on the pavement but it was tolerated. Sometimes the police or other people controlling street businesses would say something and he would have to move somewhere else. His activity was legal just outside the house where he lived, located in a secondary street. Here, he was allowed to conduct his business through the MSE that he had established together with his neighbours. However, spending his entire day there meant ‘gaining nothing at the end of the day’. For this reason, ‘you have to be smart and find your own way on the street’. By this, F. S. meant that he had to negotiate the space where he could conduct his business on a daily basis, as being ‘tough’ was the only way he could make a living and be independent.Footnote ³⁸

In a similar vein, G. S., another illegal street vendor, contended that he was allowed to perform his activity only at specific times. He explained that, until a few years earlier, there had been much more flexibility with regard to street vendors and, within certain limits, he was allowed to conduct his business at any time of the day. ‘Today things are different … in very crowded areas, where vending is more remunerative, you are allowed to sell only at certain times of the day … After 5 p.m. is always good as the traffic police officers are no longer around.’ He also mentioned that being ‘tough’ was a central requisite to navigate street business, because ‘what you can do and what you can't do depends very much on who you are and who you know’. He said that he had enrolled in the kebele employment lists, but that so far he had not been shortlisted for training or selected to establish a savings group. G. S. also mentioned that ‘MSEs are a good entry point to the system’ and that ‘once you are in it's easier to do your own business’.Footnote ³⁹

In rapidly transforming peri-urban settings, the boundary between formal and informal, legal and illegal, is constantly shifting, and its redefinition can be analysed through the lenses of time and space. To get by, people constantly challenged and negotiated dominant configurations of time and space. From the government perspective, MSEs were allowed to perform their activities within certain spaces – in terms of physical location and specific business activities. In an interview, members of an MSE comprising five recently graduated students in Kolfe Keraniyo complained that the promotion office suspended their licence for selling oil and sugar – a business that enabled them to make a good living – and replaced it with a metalwork licence. While their complaint had little effect, the outcome was that they continued to conduct the oil and sugar business illegally, and, at the same time, they complied with the request of the promotion office ‘because you can't break relationships with them’.Footnote ⁴⁰

Social rules and boundaries are constantly negotiated and redefined, with political priorities sometimes changing to reflect a changing reality. Many of the peri-urban poor tended to have no licence for their business, did not pay taxes, and might or might not be enrolled in the kebele employment lists. To some extent, the employment list embodied the expectations and aspirations of peri-urban dwellers in their relationship to the developmental state (Mains Reference Mains2012). The high-modernist ethos (Scott Reference Scott1998) of the developmental state provided a very powerful – and normative – vision of the future, which was embodied in the collective mobilization of the population to achieve broad-based economic growth and development. In such a context, ‘waiting’ to be selected for an MSE, for training and more generally for developmental initiatives reflected everyday encounters with the developmental state, and, more importantly, established its dominant temporality. Whether or not people were included in the employment lists often reflected their relative position vis-à-vis local power, as did the length of time they had to wait before being involved in developmental activities that allowed them access to socio-economic mechanisms of redistribution. Once an MSE was established, the dominant temporality became the notion of ‘graduation’ – the expectation that the business would become self-sustainable within a certain temporal span, normally five years. However, being enrolled in the employment lists or being part of government initiatives was very often time-consuming, and people were expected to perform a number of activities, including working for free and attending meetings and training sessions, and were also expected to respect some further rules. It was not uncommon to hear people enrolled on the waiting lists complaining that they did not benefit from government initiatives because ‘they were still waiting’. ‘Waiting to be employed’ and ‘waiting to graduate’ were only two examples of how the negotiation of state authority operated in the context of the policy of job creation in contemporary Ethiopia.

In the context of all these people living in the limbo of peri-urban Ethiopia, the MSE policy constituted the normative framework confronting peri-urban dwellers, one that was actively negotiated, challenged and therefore reproduced. The border between legal and illegal, formal and informal, was constantly redefined – and reflected the production of state authority – through everyday negotiations of spatial and temporal boundaries in the context of the job creation policy. The everyday making and unmaking of the local developmental state project reflected its central contradiction: that the dynamics of social, economic and political exclusion operated simultaneously, despite the claim of the developmental state to be more inclusive than its neoliberal alternatives.