Good Intentions Gone Awry: Government Intervention and Multistakeholder Engagement in a Frontier Market

How to achieve sustainable communities with decent work and economic growth without negative environmental impact, is at the heart of the United Nations Sustainable Development Goals and a top priority of many governments around the world. This article critically explores the role of government intervention for achieving sustainable local prosperity in frontier markets of developing countries, where such advancement is especially crucial. More specifically, we explore by an in-depth case study how multiple stakeholders cooperate to enhance local development and export from firms in the leather and leather products industry in Ethiopia. From a multistakeholder engagement perspective, including representatives of local businesses, United Nations, Ministry of Trade and Industry, and other development partners, we analyze how government interventions have resulted in unintended outcomes despite their good intention. We contribute with a new understanding of why development initiatives in frontier markets struggle with stakeholder integration, caused by power asymmetry and lack of institutional trust which prevents the achievements of sustainable development goals. Contextualized implications for firms, government, and non-governmental actors on how to improve collaboration are provided, and policy implications are proposed.


Introduction
Agenda 2030 and its 17 Sustainable Development Goals (SDGs) are presenting a highly complex, albeit necessary quest for our global society. How to collaborate for sustainable communities with decent work opportunities and economic growth without negative environmental impact, is at the heart of the UN SDGs 7,8,9,11,and 17 (www. un. org/ susta inabl edeve lopme nt), and a top priority of many governments around the world. Currently, the speed of progress has not been satisfactory, which calls for systemic change in need of interventions (van Tulder, 2017;van Tulder & Keen, 2018). Networks of stakeholders from civil society, businesses, and governments are formed to tackle this increasing complexity (Heikkinen, 2017).
For multistakeholder cooperation to achieve intended outcomes, it becomes incumbent to consider the institutional context (Scott, 2005). Africa-based research is highlighted as a particularly relevant context for building new knowledge of stakeholder relations (Kolk & Rivera-Santos, 2018). Musacchio and Werker (2016) point out that MNEs are increasingly interested in low-income, high-risk countries as manufacturing destinations, as hubs from which to export goods and services elsewhere. They call such economies "frontier economies", defined by weak legal systems, politically manipulated markets with unsteady GDP, but offering better returns on FDI because of their high growth rates.
Africa, together with Eastern Europe, the Middle East, South America, and Asia, has a concentration of frontier markets with rapid progress in need of stability and sustainability (Musacchio & Werker, 2016;Opute & Madichie, 2017). These markets are sensitive to exploitation and are in dire need of finding ways for sustainable growth with 1 3 lasting prosperity for the local businesses and their community. Research in this context is scarce (Birru et al., 2019;Ibeh et al., 2012), yet its impact on progress toward sustainability goals holds great potential. We propose that studying the extreme context (Hällgren et al., 2018) of frontier markets can expand our understanding of multistakeholder engagement and develop new knowledge toward reaching the SDGs.
One of the countries that has moved from developing to frontier market status is Ethiopia that is undergoing institutional changes to manage its high growth rate (Culverhouse & Dielmann, 2019;Musacchio & Werker, 2016). Ethiopia hence provides a fertile ground to study multistakeholder engagement in a frontier market, where sustainable development is particularly important. To that end, this study explores the development of the leather and leather products industry in Ethiopia with a qualitative case study approach. Our main research question is: How can government intervention and multistakeholder engagement for SDGs in a frontier market achieve favorable outcomes?
By taking a multistakeholder engagement perspective, government intervention, intention, process, and the ensuing outcomes, are explored. This study contributes in two ways. First, we contribute by exploring government intervention and multistakeholder engagement in a frontier market to shed light on this hitherto under-researched context. We demonstrate how interventions go awry despite their good intentions and not only fail to produce the desired improvements but result in unintended negative outcomes. Second, our findings contribute to a new understanding of how multistakeholder engagements can produce desired outcomes for SDGs by identifying hurdles and discussing ways to overcome them through seven aspects for improved multistakeholder collaboration.
The remainder of the paper is structured as follows: First, we describe our theoretical framework of multistakeholder engagement, institutional and international entrepreneurship, and the role of family firms. Second, we describe the research context. Subsequently, we describe our qualitative case study approach. Then, we present our findings and discuss our conclusions and their implications for theory and practice.

Multistakeholder Engagement and Institutional Entrepreneurship
A stakeholder approach considers all of those groups and individuals that affect, or are affected by, the accomplishment of an organization's purpose (Freeman, 1984, p. 25). Stakeholder theory posits that managers should consider the interests of stakeholders in a firm in their decision-making (Jensen, 2001, p. 8). In discussing the root of stakeholder theory, Jensen (2001, p. 14) further suggests that it taps into "the deep emotional commitment of most individuals to the family and tribe". As such, stakeholder theory is a theory that moves people by its emotional resonance (Laplume et al., 2008;Weick, 1999). Thus, making the necessary evaluation about multiple good and bad options will determine the success of an organization (Jensen, 2001). In a frontier market characterized by weak institutions (Khanna & Palepu, 1997), achieving organizational success requires multistakeholder engagement with deep emotional commitment (Jensen, 2001). For inter-organizational collaborations to create value, partners need complementary resources and to make relation-specific investments to build trust (Weber et al., 2017). Successful multistakeholder cooperation can leverage resources to create new institutions or transform existing ones (Garud et al., 2007). According to Pacheco et al. (2010), institutional entrepreneurs can be characterized as institutional innovators or agents of institutional change.
Firms from emerging markets are situated in an increasingly dynamic and competitive environment characterized by disruptive changes in technology, markets, and regulatory systems (Eddleston et al., 2020;Khanna & Palepu, 1997). DiMaggio and Powell (1983) define three institutional pressures on organizations leading to isomorphic change: (1) coercive pressure exerted from state actors, (2) mimetic pressure that drives firms to adopt the characteristics of other organizations, and (3) normative pressure by and within firms (e.g., via associations and professionalization). An organization striving for legitimacy might encounter conflicts of interest (Meyer & Rowan, 1977). Firms from emerging and frontier markets may face pressure to legitimize themselves by conforming to unrealistic requirements (Khanna & Palepu, 1997). Moreover, regulatory institutions may mimic each other regarding recent policy fads, paying little attention to the intricacies of managing, for instance, global value chains (GVCs) and how they can be configured (e.g., Khanna & Palepu, 1997;Melin & Nordqvist, 2007;Taglioni & Winkler, 2014).
The institutional context plays a strong role in the ability to manage change. This requires successful frontier market firms to join conglomerates with diversified portfolios to cope with the institutional voids in their domestic market. An institutional void can be defined as the lack of a welldeveloped institutional structure that is a regulatory system and subsequent mechanism for enforcing contracts that in turn facilitate the smooth functioning of the product, capital, and labor markets (Palepu & Khanna, 1997). The question is at what cost frontier market firms are invited to collaborate with the international conglomerates. High-impact entrepreneurship necessitates an institutional environment characterized by knowledge spillovers and access to capital (Stenholm et al., 2013). Oparaocha (2015) suggests that international entrepreneurship depends on how firms are aware of, have access to, and use resources available through institutional networks. This means that the institutional environmental setup can enhance or hinder the entrepreneurial activities of firms, in general, and their international entrepreneurship activities in particular.

International Entrepreneurship in Frontier Markets
Governments in most frontier and emerging markets seek to support international entrepreneurship (Michael & Pearce, 2009) as it is considered the engine of a nation's economic growth and employment (Audretsch, 2007;Michael & Pearce, 2009). International entrepreneurship enables value creation through access to foreign stakeholders, resources, and institutions. Moreover, it can allow firms to enhance their core competencies, gain unique knowledge, access new technology and growth opportunities abroad (Hitt et al., 1997(Hitt et al., , 2006. Successfully accomplishing change and innovation requires a capacity to act on new opportunities. Thus, absorptive capacity is the "ability of a firm to recognize the value of new, external information, assimilate it, apply it to commercial ends and is critical to its innovative capabilities" (Cohen & Levinthal, 1990, p. 128).
Governments are motivated to support internationalization to access foreign currency, create jobs, increase the competitiveness of an industry, and encourage innovation. However, we know little about the success of government interventions to increase export and the realization of SDGs in frontier markets.
In our context of a frontier market in a developing country, international collaborations run particular risks of exploitation (Prashantham & Birkinshaw, 2020). The uneven distribution of power makes local firms vulnerable due to their position of dependency on their foreign collaborators. In inter-organizational relationship literature (e.g., Galaskiewicz, 1985), and more specifically in strategic alliance (e.g., Wang, 2011) and supplier-customer relationships (e.g., Caniëls & Roeleyeld, 2009;Cox, 2001;Johnsen & Ford, 2008), power asymmetry has been used to discuss differences among firms in their performance. Power asymmetry can be defined as "the degree to which one firm holds substantially more or substantially less power than another in an alliance relationship" (Wang, 2011(Wang, , p. 1269. The issue of power and dependence and how to harness them for better firm performance has also been examined in global value chains (GVCs) Grereffi, 1999;Johnsen & Ford, 2008;Kano et al., 2020;. Prior studies point out the need for economic, social, and environmental upgrading and the crucial role that governance plays in GVCs including host country firms Grereffi, 1999;. "Upgrading" refers to a process of improving the ability of a firm or an economy to move to a more profitable and/or technologically sophisticated capital-and skill-intensive economic niche (Gereffi, 1999, p 52;Lee & Gereffi, 2015, p. 323). Yet, we know little about how multistakeholder engagement for SDGs successfully can achieve upgrading.

The Role of Family Firms in a Frontier Market Context
Family firms are a prominent part of most economies but play a particularly important role in countries with less stable environments, where relying on family structures is found to make up for a lack of institutional structure (Hatum & Pettigrew, 2004;Luo & Chung, 2013a, 2013b. Chua et al., (1999, p. 25) define family business as "a business governed and/or managed to shape and pursue the vision of the business held by a dominant coalition controlled by members of the same family or a small number of families in a manner that is potentially sustainable across generations of the family or families". Sirmon and Hitt (2003) point out that family firms have unique resources stemming from the influence of family in the business such as survivability, patient, social, and human capital. Several studies have determined that family firms show outstanding longevity and outperform their non-family counterparts (Anderson & Reeb, 2003;Zellweger et al., 2012). Although the results are somewhat inconclusive, family firms generally show strong local embeddedness (Bingham et al., 2011;Dekker & Hasso, 2016), CSR engagement (Faller & zu Knyphausen-Aufseß, 2018) and a pursuit of both economic and non-economic goals (Chrisman & Patel, 2012), making them care more about stakeholder relations than non-family firms (Cennamo et al., 2012).
Family firms are inclined toward a stewardship orientation, which can be conceptualized as employee, customer, and long-term orientations (Miller et al., 2008;Segaro et al., 2014). Furthermore, Le Breton Miller and Miller (2018) argue that the ways family firms demonstrate stewardship or agency behavior have important links with the institutional context. A small growing body of literature has examined the role of a family business in emerging markets (Luo & Chung, 2013a, 2013b. While research on the role of family firms in frontier markets is scarce, their characteristics bode well for engagement in achieving SDGs. Following an inductive logic, this review of prior literature aims to describe a theoretical background of concepts that are helpful to the interpretation of our explorative case study (Stake, 1995). The assumption that governments implement institutional changes that achieves improvements toward strategic development goals facilitated by multistakeholder collaboration does not explain why, in the frontier market context, stakeholders with aligned aspirations not just fail to reach these intended results but experience negative consequences. We aim to contribute to this theoretical puzzle (Alvesson & Kärreman, 2007) by exploring the research question of how government intervention and multistakeholder engagement for SDGs in a frontier market can achieve favorable outcomes. Next, we describe our particular frontier market context.

The Frontier Market of Ethiopia
Frontier markets combine characteristics of low income and high growth with volatile institutional frameworks (Graham & Emid, 2013;Opute & Madichie, 2017). This research is conducted in the frontier market of Ethiopia (Gordon & Eghabal, 2013;Mataen, 2012;Musacchio & Werker, 2016), a sub-Saharan African country, with 110 million inhabitants. Ethiopia is growing rapidly, with GDP growth averaging 10% throughout the last decade, compared with the regional average of 5.4% (Afrieximbank, 2018;MPO, 2020;Worldbank, 2019). Being one of the poorest countries in the world with a GDP of US$86.6 billion and GDP per capita of US$786 in 2018, Ethiopia aspires to reach lower-middle-income status by 2025. The manufacturing sector is underperforming because of several challenges, such as the structural-and logistics-related issues, acute foreign exchange shortages, and an unstable political environment (MPO, 2020;World Bank, 2019). Concerning the business environment, Ethiopia ranks as 126th out of 140 countries, pertaining to macroeconomic instability, workings of institutions, and business dynamism (Schwab, 2019).

Ethiopian Industrial Policy
The Ethiopian government has implemented several policies and programs aimed at transforming and growing the economy to improve the wellbeing and sustainable livelihood of its people in line with the UN SDGs. The first growth and transformation plan (GTP I) was a five-year plan (2010)(2011)(2012)(2013)(2014)(2015), intending to enhance export competitiveness. However, this goal was not realized as planned (National Planning Commission, 2016). GTP II has prioritized accelerating the growth of the manufacturing sector, and partnerships (SDG 17) are identified as key to mobilize resources for climate change and the green economy (GTP II, p. 48). To this end, the leather and leather products subsector is selected as one of the priority sectors (National Planning Commission, 2016).

The Leather and Leather Products Sector in Ethiopia
The leather sector is an old manufacturing subsector, primarily consisting of family firms. It has existed for more than eight decades in Ethiopia thanks to the country's huge livestock resources (Oqubay, 2015). The leather sector has the following three major industrial components: tanneries (where the raw skin and hides are processed into leather), footwear manufacturers, and leather product manufacturers. Moreover, the industry has formal large and medium-sized companies as well as many micro-enterprises operating in the informal sector of the economy. The leather and leather products sector is a subsector contributing immensely to the industrialization and growth of the economy (Oqubay, 2015). Yet, each stage of its value chain has been characterized by challenges and bottlenecks; quality problems with raw skin and hide, lack of access to finance, competence deficiencies, technology gaps, weak marketing, and electrical power interruptions (Addis Ababa Chamber of Commerce, 2015; Grumiller & Raza, 2019;UNIDO, 2012). In particular, SMEs struggle to be successfully integrated into larger value systems and to create linkages with support providers (UNIDO, 2016). The sector has exhibited liquidity constraints, inefficient regulations and law enforcement capacity, and coordination problems (Abebe & Schaefer, 2015;Dalberg, 2013;Gebreeyesus, 2013). Collaboration and multistakeholder engagement appear as a missing link to unlock the potential of this industry.
In sum, the leather and leather products industry is important to Ethiopia but is struggling to achieve sustainable prosperity. To this end, several stakeholders, namely, firms, government (Ministry of Trade and Industry and Leather Industry Development Institute), United Nations Industrial Development Organization (UNIDO), the Ethiopian Leather Industry Association (ELIA), Japan International Cooperation Agency (JICA), and relevant others are collaborating toward this common goal. To achieve development, the government has undertaken several actions that, so far, unfortunately, have fallen short of producing the intended outcomes.

Research Design
Given the scarcity of research on multistakeholder engagement for SDGs in frontier markets and the context-bound nature of the phenomenon, this study uses an exploratory research design (Creswell, 2007) with a qualitative case study approach (Eisenhardt, 1989;Yin, 2014). As such, we follow Eisenhardt's (1989) strategy for theoretical sampling by selecting a theoretically useful case with the potential to extend emergent theory. Our case is the phenomenon of multistakeholder engagement for SDGs in the Ethiopian leather and leather products industry, which includes several sub-cases of different stakeholder groups. This embedded design is chosen to incorporate several levels of analysis for a rich understanding of the phenomenon from both the sector and individual stakeholder levels (Yin, 2014). Given the aim of producing contextualized knowledge of the phenomenon, our emphasis is not on the generalizability of findings in a statistical sense, but on their case-to-case transferability (Lincoln & Guba, 1985). We reflect on this transferability and its boundary conditions in the final section of the paper (Busse et al., 2017).

Sample
Sub-cases were identified through initial interviews with representatives of the Ministry of Trade and Industry's agency called the Leather Institute Development Agency (LIDI) and the Ethiopian Leather Industry Association (ELIA). Regarding the criteria, we aimed for a diversity of firms along several dimensions. We primarily included family businesses because those constitute the vast majority in the sector and the country. For family firms, we made sure to include both those that have been doing well in terms of internationalization and those that are struggling. One nonfamily firm and one FDI firm were also included, both part of large conglomerates. The final sample constitutes nine case-firms representing the stakeholder group of local businesses, which is presented in Table 1.
From the government side, the Ministry of Trade and Industry is represented by two respondents, whereas the Leather Industry Development Institute is represented by four respondents, including the former Director General. Thus, seven respondents constitute the government side, as shown in Table 2.
Other interest groups included are the Leather Manufacturers Association with two respondents and two Development Partners (see Table 3). Some respondents were identified with a snowballing technique from previous interviews. In total, data were collected from five stakeholder groups (one group of firms, two governmental, and three non-governmental groups; Tables 1, 2, 3).

Data Collection and Analysis
Data collection and analysis are highly intertwined in explorative case study research (Stake, 1995). The main data source is 23 in-depth interviews with different stakeholders (see Tables 1, 2, 3). All interviews followed an interview protocol with open questions and were conducted in person by the first author in the respondent's preferred language (English or Amharic). All interviews were recorded and transcribed. The transcripts were translated to English when necessary to allow the involvement of the second author. The first author is native to the culture with deep knowledge of the business climate and industry characteristics. Throughout the analysis, we were inspired by Salvato and Corbetta (2013) to use the insider/outsider composition of the research team. Where the familiarity of the first author was necessary for a deep "insider" understanding, the "outsider" perspective provided by the second author was used for a critical view. The sub-case descriptions emerging from the interviews were complemented with secondary materials, comprising media coverage, annual reports, and company webpages (86 documents in total) to triangulate the material. To capture multistakeholder engagement; government actors and non-governmental support agencies were interviewed from their viewpoint. A round of follow-up interviews was also made with the Ministry of Trade and Industry (MOTI) and with the United Nations Industrial Development Agency (UNIDO) to further triangulate and deepen the SDG aspects (see Tables 2 and 3). Following this procedure, we reached theoretical saturation by continuing data collection until the point at which incremental learning was minimal because we encountered phenomena seen before (Eisenhardt, 1989). To manage large volumes of data, all empirical materials were collected into one case-database for systematic processing.
Data were analyzed through a gradual move from casespecific to conceptual insights following three broad steps: (i) building thick descriptions of the sub-cases; (ii) crosssub-case analysis; and (iii) sensemaking process. The thick descriptions were staying close to the language of the informants resulting in an overall case-description of 245 pages, aimed at reaching a deep understanding of various stakeholder perspectives. In the second step, we search for patterns of similarities and differences and constructed a 60-page case analysis describing the whole process. For each emerging theme, we organized quotes in a data table with supporting quotes from each stakeholder group which was finally reduced to a 14-page table serving the purpose of showing our data (Pratt, 2009) and made available to the reader in an electronic appendix (see Online Appendix 1). Finally, through the third step, we read and re-read the empirical material, combined with iteration between data and previous literature, until our analysis resulted in the conceptual insights that form our theoretical contributions (Eisenhardt, 1989).

Findings
The different stakeholder inputs are merged to present the story around the intention, process, actual outcome, and learnings for each of the three Government Interventions. Important stakeholder perspectives are illustrated with quotes in the text and supported by more quotes in Online Appendix 1. The three interventions identified and described are chronological with some overlapping, and the third intervention is recently initiated and ongoing.

Government Intervention I: Value Addition
Intention The value addition intervention by the government had the main objective to move the local leather industry up the value chain, by stimulating the leather producers to upgrade their products instead of selling raw material and products with low levels of refinement. The leather production process has four common stages: pickling, tanning, crusting, and finishing. 1 First, the raw skin and hides are pickled in a salt treatment to make the hide acidic and receptive to the following tanning. Tanning is a chemical process often done with Chromium giving the leather a pale blue color, which is why the product at this stage is called "wet blue". The third step, crusting, is a series of chemical and mechanical procedures to make the leather thin and soft and most often also colored. The last stage is finishing, to make the hide softer and to mask small mistakes. Leather can be traded at any of these stages of processing and the value increases at each step. Government representatives were convinced of the need for radical development to reach great benefits. Several stakeholders, including governmental, nongovernmental, suppliers, research institutes, leather tanneries, and leather goods manufacturers have been involved to bring the intervention to fruition. The government is the main driving force, with considerable support from UNIDO. The intended outcome was job creation, increased income, increased revenue from taxation, and increased export. As described by a representative of the government: Actually, a lot has been done. We first asked, how can we work on the value addition of the leather sector? A big package was developed to expand the tanneries. After expanding the tanneries, we provide incentives for the finished product manufacturers such as tax relieved duty-free import of equipment and machinery needed for their manufacturing. The direction highlighted is clear: instead of exporting raw hide and skin, export semi-processed, processed, and finished leather products. (State Minister, MOTI1) LIDI worked to execute the new strategy with ambitious earning goals: To have manufacturing that leads to value-addition, we must use both carrot and stick. When going to valueaddition, the carrot is the incentive, and the tax is one instrument we started to use to discourage the export of pickle and wet blue.
[…] The world footwear market right now is estimated to be around US$141 billion. (Deputy General Direct, footwear and Leather Products, LIDI) According to UNIDO, the value addition intervention was trying to attain SDG 7 (Affordable and Clean Energy), SDG 8 (Decent Work and Economic Growth), SDG 9 (Industry, Innovation, and Infrastructure), and SDG 17 (Partnerships for the Goals). The industry level focus on value addition naturally puts SDG 9 central stage: SDG 9 is the main one promoted by UNIDO, and it is a mandate of an inclusive and sustainable industrial development. So, our main bulk of the activities is there. Then, we have another important SDG related to the environment and energy, SDG 7. (Representative and Director of the Regional Hub, UNIDO) Most stakeholders agree that, in theory, value addition positively contributes to economic development and increases the rate of employment for the country.

Process
The value addition intervention has been called "Fast and Furious" for the forcefulness and speed of its implementation. The Ethiopian government implemented policies through export bans and new export taxes to push tanneries up the value chain (Brautigam et al., 2016;Oqubay, 2015). Specifically, in 2008, Intervention I (Stage A) included a 150% export tax on raw skins and hides, 20% on wet blue and 10% on pickle. In 2012, intervention I (Stage B), imposed a 150% export tax also on pickle, wet blue, and crust (Raw and Semi-Processed Hides and Skins Export Tax Proclamation No. 567/2008; Abebe & Shaefer, 2013). The final step in this development was to require firms to only produce finished leather goods through export taxation, sealed with a value addition legislation in 2012. The businesses were struggling to adapt to the interventions: For a short time, we could continue to export wet blue until the government made a policy change, demanding the export of semi-finished leather. Then, we started to export crust. We were also only able to export crust for a short time. Then, the government changed the policy again and demanded that we should only sell finished leather. (Commercial Manager, Joady) The intervention forced all tanneries to try rebuilding the factories to produce finished leather and finding new customers for the new product categories this resulted in. All firms in the industry were affected by the value addition regulations but to a varying degree depending on their prior level of refinement. Interviewed firms point out that those who are responsible for the regulation and legislation did not seem to sequence the regulation with the businesses and the level of capital needed for compliance. To try to support the transformation, several stakeholders have been engaged: We have much support from our development partners such as EU, DFID, and UNIDO; they are supporting us to minimize gaps. JICA, for instance, works with us on this project called the 'Ethiopian Highland Leather'. They do promotion work and even have video production. (State Minister, MOTI1) With UNIDO's standing as the industrial development organization of the UN, it has been instrumental in opening doors: We must create a partnership with the Association from other parts of the world. In particular, to benchmark the Ethiopian leather sector with the major Associations, for example, we must work with ASSOMAC and PICIE. ASSOMAC is an association for manufacturers of equipment for the leather and leather products located in Italy close to Milano. (Representative and Director of the Regional Hub, UNIDO).
Concerning capacity-building activities, UNIDO also facilitated trade fair participation in important foreign markets and connected the Ethiopian industry to leading research institutes in Shanghai and India. Despite all engagements and efforts, the intervention still resulted in unforeseen export problems: When the policy changed to exporting value-added crust, the Italians and most leather factories still preferred to only buy the wet blue because they do not accept the chemicals we use for further processing. (General Manager, Genket) When the value addition policy was introduced at different intervals, the financial requirements of each phase, and the time provided for firms to fulfill these requirements, did not go hand in hand. In addition, when studies were conducted, learning from them was difficult because of the high turnover of government officials: When the direction is given to UNIDO to conduct studies and when the officials giving those directions are replaced with other officials, it will take time to explain and for them to understand the project goal. Moreover, the whole system will be disturbed. This problem could apply to all and is a problem from the government side. (UNIDO, Representative and Director of the Regional Hub) The development was disappointing.

Actual Outcome
The good intention to increase the earnings of the domestic leather sector through the export of refined goods resulted in an outcome with worsened circumstances. Instead of the improved local job market and increased earnings, all sorts of unintended effects arose. Perhaps the most striking awry result is the decreased export, which was the total opposite of the intention: The government's perspective was that the more value-added to raw skin and hide, the more inflow of foreign currency will be generated. However, the market for finished leather for all leather firms in the sector is weak. Especially after the regulation demanding us to export finished leather, we are not exporting. (Commercial Manager, Joady) All businesses witness plummeting exports over the years, which are directly related to the different steps of prohibiting the export of wet blue and crust: When we were exporting Crust, we had reached the export of US$9 million. When we started to produce finished leather, the maximum we have been exporting has been US$6 million. Since I have become commercial manager, we have been exporting US$5 million, then US$2.5 million, and now US$1.5 million last year. (Commercial Manager, Pedie) The intended demand for finished Ethiopian Highland Leather did not materialize. ELIA worked with JICA to analyze the decline in international markets. They confirm that the problem is due to the forced shift from raw skin and hide to finished leather: In the past, Japan and Italy used to take the Ethiopian Highland raw skin and hide. So, when we switched our policy to add value to raw skin and hide, Japan did not come here to source and build a manufacturing plant. Thus, demand for Ethiopian leather decreased because of the value addition policy. (Director of Leather and Leather Products, MOTI2) Not only did export decline and the cost of production increase, but the balance between local producers was offset. An unfortunate competition was created when the domestic market was forced to become less differentiated.
Our business was tannery to tannery. When it was decided that we all should produce finished products, the tanneries became competitors, without any research. In this process, they drove our customers to do our job, and this left us with nothing. (General Manager, Kinde) Many businesses express frustration about trying to comply but struggle to handle the huge increase in requirements. Although institutional level efforts were made, timing and coordination of support halted.
That is because the other governmental and banking services support providers were not ready to fully engage, and second, the international market was not ready to accept our finished leather products. The policy was implemented, but there was no promotion or support. (General Manager, Genket) However, governmental stakeholders do not have the same picture of the outcome and see several positive effects: Thanks to the policy intervention, the manufacturing sector was able to transition from sending raw skin to sending semi-finished, or processed, and finished leather products. Developing policies and implementing them was the government's task. Getting the task done has enabled the export of finished leather products. (Director of Leather and Leather Products, MOTI2) Development partners indicate many support systems, but they are for a limited time period and can cease abruptly if project time runs out, pointing to the challenge of achieving lasting results also after aid is terminated.
The government has received criticism for failing to provide appropriate support to their intervention. This is a daunting task under the conditions of a frontier market. Some business managers have experienced access to useful support, but many aspects must come together at the same time to manage the challenge of exporting highquality products from a developing country.
Stakeholder groups diverge in their view on the sufficiency of support provided and the local businesses are receiving criticism: Finally, the private sector comes to a meeting and complains. The private sector does not look at what is lacking from their side. This is the problem. These three, namely, the private sector, the government, and the donor, must play their roles. In addition, the private sector must grasp all the knowledge and institutionalize it. (UNIDO, Project Manager) Some new entrepreneurs are flourishing in the leather products sector that entered through backward integration and were therefore not adversely impacted by the value addition intervention. The question is if they have resulted in more jobs or if they have replaced old local businesses.

Learnings
Government intervention I on value addition had good intentions but resulted in problems that can be viewed as worse than before the intervention. Several problems that arose in the implementation stem from the forceful legislative approach chosen by the government. The primary means chosen were prohibitions, tax barriers, and tougher requirements. Stakeholders point to a culture of relying on restrictions rather than positive reinforcements: The biggest challenge for us is a 'preventive mindset' in Ethiopia. A ranking government official had said that almost all the government authority is regulatory, and not supportive. There is no interest in 'how can I help you?' (Key respondent, ELIA) For the top-down regulation to achieve the intended outcomes, they must be based on well-informed decisions which are impossible without input from the primary stakeholders. Several different stakeholders indicate a lack of industry knowledge from the highest decision-makers.
What should have been done was to establish and develop the manufacturing industries despite the type of sale they make. Moreover, the industries should be allowed to be guided by their own entrepreneurs and choose what's best for them. (Former Director General, LIDI) Even though support schemes were available, coordination and sequencing were lacking to sufficiently achieve the right support at the right time. Moreover, enduring support is needed for local producers to manage through the restructuring. Good efforts were made, but the benefit for local communities was not reached.

Intention
When the awry results of the value addition intervention were realized, the government focused a new intervention on inviting FDI firms. The reason for developing countries to seek FDIs are generally as a catalyst for development, increased employment, forward and backward linkages, and technological spillovers (Waldkirch, 2011). Ambitious strategies were formulated by the Ethiopian government to attract FDI to the leather and leather products industry to amend the lack of local development and export. UNIDO representatives saw many possibilities: SDG 9 is definitely the major one, in cooperation with SDG 7. Then, we have SDG 17 that is also critical with partnerships […] Moreover, the socio-economic ones like SDG 5 and SDG 8 are critical as they refer to gender, women empowerment, and decent work. (Representative and Director of the Regional Hub, UNIDO) With capital strong FDI firms entering the local market, the export would be boosted, creating an increased size and activity of the entire industry.
Concerning those who are from abroad with experience in the sector, we try promoting transfer of experience and technology to the local manufacturer. Such activities occur, starting with developing industry strategies. (Director of Leather and Leather Products, MOTI2) Preferably, the foreign firms would focus on the later stages of the value chain providing demand for the local firms to produce the different stages of processed leather. Through their knowledge and access to technology and supplies, the FDI firms would educate the local firms on international environmental demands and provide them with the right chemicals to overcome previous export obstacles and improve local environmental conditions. The benefits of FDI firms engaging to fill these gaps received unambiguous acceptance.

Process
The government intervened by inviting FDI firms to boost the sector. Concerning non-governmental stakeholders, UNIDO was extensively involved in the FDI entry strategy: By promoting Ethiopia and by preparing the baseline data to attract investors and show the potential of Ethiopia regarding the leather sector, the leather products, the labor force, and the investment codes are prepared by UNIDO with the Ethiopian Investment Commission.
[…] Sometimes headhunting is involved. Who is the best shoe producer in Italy? (Project Manager, UNIDO) The intended positive outcomes of FDI were many but once again, problems arose in implementation. If Government Intervention I was relying too heavily on the legal regulation, the second intervention can instead be viewed as too unregulated. The government invited firms to produce finished leather products such as shoes, bags, and garments to create demand for the local leather tanneries. However, the foreign firms were invited without any legal requirement to work with the local tanneries. The FDI firms, therefore, were not prohibited to source skin and hide directly and performing their own pickling, tanning, crusting, and finishing with great savings. The prevailing law does not allow foreign firms to directly buy and sell products from raw material suppliers because of the export bans. However, upon entry to Ethiopia, the FDI firms engaged in procuring raw materials on their own. This has intensified the level of competition and created an ongoing price war between local and foreign leather investors instead of the intended collaboration: As I understand it, foreign investors are using us to get raw material and resources, not to interact and work with us both in terms of experience and technology. A policy that can fill the gap between foreign and local leather investors is necessary so that they can work together. (Marketing and Business Development Manager, Goshen) The backward integration goes back all the way to compete with local tanneries for the sourcing of raw skin and hides from small local animal keepers. In this way, the conditions for the domestic producers are severely worsened.
When the FDI firms came, it was planned that they come with the technology, chemicals, and machinery setup. It was not intended for them to gather our raw hide and skin, but rather, for our local factories to process at a certain level and sell it to them.
[…] Our policy does not prohibit direct buying from local sellers.

(Director of Leather and Leather Products, MOTI2)
The government anticipated a need for supervising the development. Inspectors were to monitor the expected integration of the FDI firms. However, this monitoring did not function properly: We have a leather department and goals set. We have experts, who will visit the FDI firms. They must determine whether they are operating according to the goals set and how integrated they are with the local firms. Even though we have this system, a capacity problem exists on our side; the experts who conduct these visits are not the right experts for the job. In addition, they are not paid well. The salary is discouraging and therefore not attracting the right people for the job. The platform is in place, but the actual performance is not as expected. (State Minister, MOTI1) Development partners also indicate that the lack of knowledge about the sector by the decision-makers has hindered the progress. As the government tries to create a conducive environment for FDI firms to invest locally to achieve economic growth of SDG 8, UNIDO and other support-organizations question the achievement of decent work conditions, which is the other part of SDG 8. Providing decent work for their workers does not seem to be the priority of FDI firms.
Of course, it is the role of ILO as UN organization, we ensure that jobs should have dignity: wages are paid to the workers and operation has been safe. You can see there are some FDIs that are exploiting, I would say. (Representative and Director of the Regional Hub, UNIDO) There are also a few successful cases. For Sobrel, FDI entry worked as intended but, in this case, the process developed from an established business relationship, which was continued thanks to the invitation of FDI firms: We used to export a lot to China. Then the tannery shoe factory in China started opening shoe factories in Ethiopia, such as Huajian, George shoe, and New-Wing. This was easier for us because we produce and do not have to wait to export. They take leather from us and make the shoes here and export to the US. (General Manager from Sobrel) Unfortunately, Sobrel seems to be one of the rare exceptions, and overall, the process of implementing the FDI strategy suffers from a lack of integration and regulation of the industry actors.

Actual Outcome
The great expectations on inviting FDI firms into the domestic leather and leather products industry were not met. Although export was increased by the FDI firms, it had little positive effect on the total export from the industry. Actually, some point to a negative trend where local firms are replaced by foreign-owned ones that crowd out local firms through backward integration.
The tanneries that were already on their knees from trying to comply with the previous value addition policies are now outcompeted by capital strong FDI firms and are very disappointed with the outcome. Corruption is implied as an explanation for why FDI firms were allowed to proceed with such freedom when they were not honoring the expected behavior.
LIDI is responsible for all this mess, and the people in power should be accountable for destroying such a big industry. You can come and see; my factory alone has machinery worth 600 or 700 million. How can such massive industries like ours be allowed to be mistreated by an institute like this? What did we get in the past decade? Nothing. So why are some people still defending the process? I say, 'follow the money!' (Owner-Manager, Kinde) There seems to be a consensus among all stakeholder groups that FDI firms should have been integrated into the local economy as a catalyst, but not as a force for replacement. How to achieve this remains unsolved. Government stakeholders are acknowledging several problems: There are diverse problems: one is of the raw material, then that of the shortage of technology, especially for the locals, shortage of technology and machinery, and then the government, the things we supply, such as power, the dollar, etc. based on the countries' limited financial condition. We have not been able to provide in the way their need entails. (Director of Leather and Leather Products, MOTI2) The proportion of the industry export from local firms is declining, and the positive spillovers are not achieved. According to the business managers interviewed, a problematic outcome is a replacement occurring when FDI firms are engaged in the same business as they are. This view is supported by non-governmental stakeholders: In fact, what FDI does is use only your cheap labor and your cheap energy, you know, the water and those facilities, but it does not create knowledge to develop the Ethiopian leather. When they find another place where these are cheaper, they take everything back with them. (Representative and Director of the Regional Hub, UNIDO) The governmental stakeholders acknowledge the awry results: When the FDI firm is permitted to operate, it is with the assumption that there will be technology transfer. Especially, they will integrate and work with the local firms. A huge gap definitely exists between the intended and actual outcome. […] We had expected big things to happen, but so far, the outcome is not what we have expected. (State Minister, MOTI1)

Learning
Government Intervention II had good intentions but resulted in unforeseen problems. Because of the lack of integration, the technology transfer and knowledge spillover did not materialize. What makes matters even worse is that the FDI firms have not only refrained from the expected collaboration (apart from a few exceptions) but have also outcompeted the local firms on their place in the value chain.
There should have been a supply chain integration, where we provide wet blue and pickle, and then they buy it and processes it further. (Owner-Manager, Kinde) Some lessons we took from this include coordinating the relationships between the FDI and local firms by making the FDI firms source semi-processed skin from the local firms except for special permission sourcing. The FDI firms should only be able to maneuver around the top part of the supply chain. (Deputy Director General, Leather Technology, LIDI) A major problem was the lack of control. The government should control that treaties are honored. Launching the platform was not enough without monitoring the follow-through of the control function. FDI firms were invited to enter with little regulation over their advancements in the market. When the FDI firms could roam free, they soon spread out of control.
The intended learning from FDI firms was also hindered by the huge difference in knowledge. For knowledge transfer to occur, the gap cannot be too wide which is why efforts would have been needed to support the narrowing of the competence gap.
Our challenge has been on how to minimize this gap and how to create industry-university linkage. When the graduates have the necessary knowledge, it is much easier to support them. (State Minister,MOTI1) There is a worry regarding the long-term commitment of FDI compared with the vision of local family firms: Their [family firm's] vision is not limited to profit today -it is for the generations to come. Instead of depending on FDI, it is better to have more firms, five/six, like Earls. The export will come late, but it will be sustainable. An FDI, whatever support it gets, leaves at any time they sense danger. (Deputy General Manager, Earls) UNIDO is trying to promote SDG 8 of decent work conditions in the FDI firms: You see many of these workers kept with the misery of 800-birr/ month. How can you eat? So, this is something that UNIDO tries to address, ensuring that the companies, especially FDIs, that are working in the country are also supporting their workers. It's not only value addition and the technical one. We must always think that the industry should have a human face, so decent work is extremely important. (Representative and Director of the Regional Hub, UNIDO) Increasing demands on FDIs can, however, decrease their willingness to stay. From the government perspective, FDI is still considered important for the economy of the country even if they recognize that it could have been better managed. The eager approach to attract FDI led to an unregulated experiment that generated unintended negative outcomes for the local firms. Jeopardizing the domestic business climate is a high gamble because what remains if FDI firms leave?

Intention
Neither the value addition requirements nor the FDI entry has so far resulted in the intended outcomes. Exports to Europe and the US are held back by capacity, quality, and environmental compliance problems. An unwanted sideeffect of banning the export of semi-processed leather is increased pollution from more steps of the production chain occurring in the country. For these reasons, a third Government Intervention of environmental compliance has been initiated. It is highly influenced by the UN development goals: Irrespective of the fact that these tanneries provide job opportunities, or provide an opportunity of exporting their product abroad, the process of manufacturing should not disturb inhabited areas. (Director of Leather and Leather Products, MOTI2) Government Intervention I and II were primarily focused on SDGs 8 and 9, creating decent work and economic growth in the industry. The third intervention is putting SDGs 7 and 11 center stage, to achieve sustainable communities and clean energy. The government of Ethiopia does in its GTP II (2015/16-2019/20) make explicit reference to incorporating SDGs. Regarding the environment, GTP II (1.3.9), states that it aims to build a climate-resilient green economy, linking to SDG 11 and target 6. International level environmental conformity is expected to provide healthy living conditions for the citizens and to overcome the export problems encountered during previous interventions. To that end, new requirements for treatment plants are planned.
As an addition to the export challenges, environmental compliance is a major issue. There are two aspects of this: solid waste and effluent waste. (Deputy Director General, Leather Technology, LIDI) Tanneries use many chemicals and discharge waste that contaminates nature. Primary, secondary, and tertiary treatment plants are needed to protect the environment. To manage the industry transformation, the government plans to cluster the leather sector in Mojo city, with a common tertiary treatment plant. "Mojo Leather City" is an eco-industrial park in the making. This is supported by UNIDO and other stakeholders for its expected contribution to a clean environment and other positive outcomes.
We are much involved in the creation of a leather district in the town of Mojo. Moreover, we are right now finalizing the plan to establish a satellite of the leather Institute and a common facility for footwear for young people and for job creation there. Therefore SDG 9 […] then, SDG 7, which are related to infrastructure and clean technologies and energy, are important (Representative and Director of the Regional Hub, UNIDO) All stakeholders agree that environmentally friendly production is good for the whole society. For local companies, the quick transformation demanded is, however, perceived as impossible to comply with.

Process
Several stakeholders are engaged to protect the environment for an inclusive and sustainable industrial development in Ethiopia. From the government side, MOTI, LIDI, Ministry of Environment, Forest, and Climate Change, and the Environmental Commission are involved. While from the non-governmental side, UNIDO is the primary stakeholder for assisting firms in meeting the environmental requirements. All firms are now required to build treatment plants that meet international environmental standards and only use internationally approved chemicals. The remaining firms that survived the previous interventions are now required to finance this major transformation in competition with FDI firms. This has caused some long-lasting family firms to crumble.
We are stuck between a rock and a hard place. We had customers for crust but not for finished leather. The loans that we have taken to update our machinery are piled up. Consequently, we entered the Non-Performing Loan status. While we were undergoing all these challenges, the environmental compliance law started requiring us to build primary and secondary treatment plants. (Commercial Manager, Pedie) The firms have no choice but to try to transform, resulting in a shake-out of the weakest and huge risks taken by the remaining ones. In this way, at least some steps toward environmentally friendly production are taken.
Almost all tanneries have established waste treatment plants for effluent wastes. Even though we cannot say they have addressed the issue entirely, they have somewhat managed it. (Deputy Director General, Leather Technology, LIDI) The increasing environmental awareness in society is causing anxiety among residents close to production facilities. Sometimes this causes unintended consequences for the tanneries that are complying.
We have built both primary and secondary effluent treatment plants. […] People will start talking that we have released hazardous chemicals. However, we have not done so! Then, the environmental monitoring agency starts to threaten us to close the factory. (Commercial Manager, Pedie) Being forced to move production would again add tremendous costs. It would also take away jobs in the towns. For the surviving firms, it is constant stress if their factory will be shut down from insufficient compliance or from failing to pay back the loans, or what the next change will hold.
In addition, the chemicals and the environmental requirements are changing all the time. (General Manager, Loewe) In the process of protecting the environment, UNIDO has been supporting firms to implement cleaner production processes. The efficient consumption of energy by using improved production systems for tanneries has been a key aspect to advance SDG 7. However, existing tanneries are at best reaching secondary wastewater treatment levels. For the realization of Mojo Leather City, with tertiary capacity, the government of Ethiopia is working together with UNIDO, European Investment Bank, and other agencies.
In the Mojo Leather City, we are trying to do the whole in bulk. We have conducted a study, but the government of Ethiopia should work with the European Investment Bank and other donors to finalize. The Italian government has also expressed its wish to participate in the endeavor. So, we hope that this will be done soon. (Representative and Director of the Regional Hub, UNIDO)

Actual Outcome
The good intention to achieve international level environmental compliance in the leather and leather products industry is yet to meet any intended outcomes and is, unfortunately, heading toward disaster for local firms that had started to recover and adapt to Intervention I and II. When the first thrust was implemented in 2017, the intervention made headlines as six tanneries were closed by Addis Ababa City Administration's Environmental Protection Authority, involving negotiations with the Prime Minister of Ethiopia (The Reporter Ethiopia, 2017). More shutdowns have followed: The Ethiopian Commission for Forest and Environment has lately closed tanneries in town because of the environmental problem. (Representative and Director of the Regional Hub, UNIDO) Some tanneries were hoping to move to Mojo Leather City, where also the tertiary common treatment plant was supposed to be constructed. However, the land for the Mojo Leather City has not yet been secured. Although still at an 1 3 early stage, the good intention to achieve a better environment, and thereby comply with international standards that enable export, is yet to be realized. Instead, the situation is critical: Due to this new policy, the whole leather industry is at risk. Almost all local firms are facing bankruptcy, and they are exiting the sector. Some factories are closed and are under bank foreclosure procedures. Some people lost their lives due to the crushing financial burden. (Commercial Manager, Pedie) The capital strong FDI firms have had a better position to comply. However, the tougher the demand becomes, the more their gain will decrease, and their exit could follow. In the meantime, foreign-owned firms have less motivation to care for the local environment. The actual outcome on firm level of our case companies is gloomy. All but two are heading toward disaster. A summary of their export performance over the time of the interventions and a firm-by-firm comment is presented in Table 4 and illustrated in Fig. 1.

Learning
The intention is good; to have all tanneries updated with the latest technology for environmentally friendly production. However, that transformation cannot be forced overnight.
Firm representatives feel that it is impossible to comply, and that decision-makers do not understand their reality: In the case of environmental issues, the leather businesses seek support from the government since it is unmanageable. We have tried our best to complain to the government officials, some of them have leather sector know-how, but some have no idea about our business. (Commercial Manager, Pedie) The government actors stress the importance of environmental compliance for the firms and the society at large. Most of them do not see how the government could have alleviated any burden on the firms on the speed, sequencing, or provision of grants. However, some point out that best practice of other countries such as India could have been better used. They are, however, acknowledging the problems of implementation and are trying to tackle it: In Mojo, gathering all the leather firms together, with one treatment plant up to tertiary level, can clean the wastewater and discharge cleaned wastewater. […] this is a so-called ionization plant. It is being prepared. (Director of Leather and Leather Products, MOTI2) This is a big scale intervention with high potential. However, it is not yet put into action, and many uncertainties regarding the existing businesses remain. The aim

Discussion
Thanks to the large prevalence of livestock and long tradition of leather processing in Ethiopia, the leather and leather products industry has unique resources, with the potential to become a source of decent work and sustainable economic growth for the country in line with SDGs (National Planning Commission, 2016). The interventions attempted have not succeeded to allow for different types of leather and leather products firms to thrive in the sector. Although collaboration is understood as important, the much-needed partnerships for the SDGs (SDG 17) are malfunctioning. Although different regulations were enacted, the knowledge spillovers and the capital necessary for international entrepreneurship were not realized (Stenholm et al., 2013). Coercive pressure exerted on the leather firms, specifically on tanneries, did not render the intended outcome of achieving high export earnings and social welfare for the country (DiMaggio & Powel, 1983;Kelling et al., 2020).

Attempting Value Addition by "Moving Fast While Staying Static"
Concerning Intervention I, the main goal was to achieve decent work and economic growth (SDG8) through the development of the industry (SDG9). The government's approach can be described as "moving fast while staying static": The paradox revealed in the value addition policy, is in how the policy has been decided and implemented. It is "fast and furious" in forcing rapid value addition by legislative action, on the one hand, and "remaining static" after implementation, on the other hand. Rapid policy changes were forcefully imposed with inadequate support mechanisms compounded by a lack of coordination of multistakeholder engagement. The "fast and furious" top-down decision-making agrees with the findings of Abebe and Shaefer (2013) and Brautigam et al. (2016). The staying static aspect, as revealed in this study, can be attributed to the frontier markets lack of institutional trust and collaboration (Khanna & Palepu, 1997), thereby reverting to a regulatory approach. According to institutional theory, mimicking and legitimacy seeking happens in a local context in an institutional setting (DiMaggio & Powel, 1983;Kelling et al., 2020). In our frontier market, we observed the "rationalized myth" that "value addition will lead to more export earnings". Due to this myth, the successful "East Asian model" was adopted, but critical studies on how to adapt foreign models were not conducted. Actors expressed a perceived lack of "an owner" of the multistakeholder engagement for stakeholder coordination. Fruitful collaborations are to be achieved by actively searching for complementary interests, aims, and competencies (Van Tulder, 2010). Therefore, "an owner" should rather be sought in shared ownership with members acting as stewards. Such stewardship was found in family businesses in their care for their family and the long-term prosperity of the firm and its community (e.g., Miller et al., 2008), which is at risk if the value addition intervention is harming the local businesses. Thus, the development agenda must become a common agenda of all stakeholders engaged, which is not yet achieved in our case, even if the intended outcome was widely desired.
In the eagerness to achieve the overall goal of SDG 8, including particularly the economic growth (target 8.1), and the upgrading of the sector (target 8.2 and SDG 9), other vital sub-targets of promoting development-oriented policies that support entrepreneurship with access to financial service (target 8.3, 8.10, and 9.3) were neglected (www. un. org/ susta inabl edeve lopme nt). It stayed static until January 2020 when reverting some of the most damaging requirements were initiated, allowing export of semi-processed leather without 150% taxation.
For multistakeholder partnerships to work effectively, deliberation, decision-making, and enforcement are necessary (Dentoni et al. 2018). According to firm and leather association respondents, the leather sector instead experienced a fragmented bureaucracy. Shared trust and localized learning can be sources of competitive advantage (Maskel et al., 1998, p. 53). In contrast, our stakeholder groups seem to lack both the shared trust and localized learning. The voiced "preventive mindset" of approaching change through restriction rather than nudging with positive reinforcement (Jolls et al., 1998) impedes multistakeholder engagement. Respondents indicated that the government usually blames firms for malpractice, whereas firms complain that the government is regulatory and nonsupportive. This "blame game" is a hotbed for mistrust that hurts the prerequisite for a fruitful multistakeholder engagement, even if each stakeholder group wants the same result. Development partners such as UNIDO suggest that firms and government actors both should stop externalizing problems and devise ways to work together.
Despite all these adversities, a few firms still managed to shine. Accidentally, these firms' goal to produce finished leather and the government's goal of value addition was aligned. According to Garud et al., (2007, p. 957) definition of institutional entrepreneurship, these firms can be considered accidental institutional entrepreneurs from this fortuitous alignment. Two of the case companies in our sample started to supply to an FDI firm, and they have been able to cope with the series of regulations and prosper. Hence, the few that thrive did not practice value addition as intended by the government intervention, whereas those forced to climb up the value chain suffer. In this way, the weak were weakened (or demised). Although firms want to include a broad stakeholder perspective in their organizational purpose (Carroll & Buchholtz, 2009), a prerequisite to any societal obligation is to first generate a profit that enables survival.
In sum, regarding intervention I, the coercive approach had good intentions, but lack of enduring measures to support implementation caused unintended outcomes. Important lessons were not absorbed systematically and fed into the implementation system, and the multistakeholder cooperation was not well managed.

Attracting FDIs to "Follow Fads and Fashion"?
Many countries benefit from FDIs (Buckley et al., 2002). The allure of FDI can be viewed as mimetic government behavior to follow "fads and fashion". After the awry result of Intervention I, the Ethiopian government aimed to achieve improved industry, innovation, and infrastructure (SDG 9) by letting FDI firms lead the way. In China, non-Chinese MNEs have provided both technological and international market access spillovers (Buckley et al., 2002), whereas, for our Ethiopian firms, those benefits have not occurred. Attracting big firms to invest in Ethiopia was intended to boost the industry (Oqubay, 2015). This eagerness led to an unregulated entry with unfortunate consequences for the local firms, seeing themselves replaced instead of supported. Furthermore, the government failed in its role to "control the controller" (Accinelli et al., 2017). Due to the lack of deep knowledge of the leather and leather products sector by top policymakers and a rapid turnover of key government actors, the much-needed regulation of the FDI entry and orchestration of multistakeholder engagement remained largely absent. Unintentionally, many local firms have perished. Buckley et al. (2002) suggest that whether firms manage to benefit from FDI depend on their absorptive capacity (Cohen & Levinthal, 1990). In this study, local firms rarely came close enough to absorb new knowledge from FDI firms, that instead of collaborating were competing. Because of their power advantage (Wang, 2011), neither the Ethiopian government nor local industry actors could control the FDI rampaging. The lack of knowledge spillover to local firms can be attributed to both the expressed knowledge gap being too wide to overcome and a power asymmetry permitting FDIs to neglect local needs.
Multistakeholder engagement aspires to create a fair playing field among stakeholders to solve complex problems where all stakeholders are engaged in the decision-making (Dentoni et al., 2017). In our case, these steps have not been taken, and the desired upgrading of local firms is still absent . Undue attention on FDI firms as a source of foreign currency and employment has skewed the playing field in favor of these firms and crowded out local firms (Buckley et al., 2002;Fikade, 2019). As a result, the locally embedded family firms are at risk. A few firms have started to supply their finished leather to Chinese shoe firms Both the value addition and the FDI entry interventions gave awry results with worsened conditions for the frontier market of Ethiopia. This, coupled with mimetic pressures for international environmental standards, led to enforcement of environmental compliance.

Attempting Environmental Compliance by Building Rome in One Day?
The recent thrust to achieve environmental compliance of international standards can be perceived as "attempting to build Rome in one day." The environmental compliance intervention is expected to provide healthy living conditions for the citizens and to overcome the export problems encountered during previous interventions. To that end, new enforcement of requirements on treatment plants and chemicals were introduced. The third intervention puts SDGs 7 and 11 at the center stage to achieve sustainable communities. Even if all actors agree that environmentally friendly production is good, the challenge and disagreements stem from how to conduct this conversion of an industry in a frontier market where most local firms are struggling to survive the previous interventions.
Several local firms had gone bankrupt or are under Non-Performing Loan status due to the value addition law. The surviving rest are struggling to, once more, comply with government interventions. Government Intervention III on environmental compliance is at the risk of aiming too high with negative consequences. Working together toward systematic improvements with stepwise development in the right direction would more likely achieve better outcomes (Dentoni et al. 2018), especially when the local family firms with a long tradition in the community are hit the hardest. Now the leather and leather products industry in Ethiopia is pushed to a breaking point. The pacing of interventions needs to be considered not to risk ending up with nothing.

Lessons for SDG Realization Through Multistakeholder Engagement
All stakeholders share the goals but are misaligned in how to achieve them. Except for the government actors, most of the firms are family-owned; thus, multistakeholder engagement could benefit from acknowledging a stewardship orientation (Miller et al., 2008) if those actors were included in decision-making throughout the development process. Furthermore, in a weak institutional context, we observe the need for incorporating emotional commitment (Jensen, 2001). This study extends our understanding that for multistakeholder engagement to work effectively in tackling complex problems (Dentoni et al. 2018); stewardship orientation and emotional commitment of all stakeholders need to be fostered. Learning from previous interventions means acknowledging stakeholder engagement and seriously enable multistakeholder collaboration. Doing so requires a change in conditions for engagement along the needs identified and elaborated in this discussion. In Table 5, we summarize our contribution in seven aspects for improved multistakeholder engagement needed to achieve forthcoming progress for the SDGs: (1) perceptual shift from top-down imposed intervention toward multistakeholder collaboration with a stewardship orientation, currently hampered by a lack of institutional trust characteristic for frontier markets; (2) a deep understanding of the practice to ensure feasible requirements in policy formulation through multistakeholder input; (3) a well-consorted and enduring implementation scheme with an effective feedback system integrated, which is extra challenging due to high turnover rates of government officials in this context; (4) orchestration of collaboration to ensure global supply chain integration including economic, social, and environmental upgrading even when challenged by the dependence on powerful external stakeholders; (5) a value-based system to control the controller that prevents breaches from being ignored by corrupted individuals; (6) educational efforts that narrow the competence gap to enable the development of absorptive capacity; and (7) pacing the interventions to get somewhere with incremental steps instead of risking ending up nowhere with failed radical leaps. These seven aspects are prerequisites for fruitful multistakeholder engagement for useful interventions resulting in sustainable societal prosperity contextualized for frontier markets (Table 5).
Studying this extreme frontier market context (Hällgren et al., 2018) leads us to expand our understanding of multistakeholder engagement for SDGs. Our study explored how interventions with good intentions could render negative outcomes. Learning from the factors identified as preventing multistakeholder engagements for SDGs in this frontier market leads us to propose the following process model for SDG realization through multistakeholder engagement as illustrated in Fig. 2.

Conclusion
This study provides empirical exploration and a refined theoretical understanding of how multistakeholder engagements can achieve SDGs in a frontier market through international entrepreneurship activities. By using a qualitative case study approach adhering to different stakeholder perspectives and conducting in-depth analyses, our study reveals how different actors cope with challenging conditions toward desirable outcomes. This study contributes to institutional theory and more specifically extends our knowledge of frontier Environmental compliance International environment standards for increased export and less pollution Push to breaking point and more pollution from more steps performed locally High-risk strategies to show accomplishments Pacing the interventions with incremental steps to get somewhere instead of nowhere markets. It considers how different actors exhibit institutional entrepreneurship and why others do not. Similarly, the study considers how, in a weak institutional environment, multiple stakeholder groups engage to achieve intended outcomes. The trial and error, speed of intervention, and lack of multistakeholder collaboration at different phases of intervention also extend our understanding regarding stakeholder theory in a frontier market. The identified lack of the overall "owner" of the multistakeholder engagement suggests the need to foster a fruitful perception of shared ownership of the engagement by all relevant stakeholders for which a stewardship orientation is preeminent. To achieve this, preserving the local family businesses where the stewardship inclination is widespread appears beneficial. Thus, government policies, strategies, and programs, which are built on SDGs coupled with flexibility in implementation is expected to attain the intended outcomes. In the long run, well implemented SDGs will strengthen the desired firmlevel outcomes for both local and foreign firms operating in the Ethiopian context.

Boundary Conditions, Limitations, and Future Research
Our findings are based on the leather and leather products industry in Ethiopia; therefore, we primarily consider our findings transferable to other traditional industries in other frontier markets. Limitations stem from the challenge of accessing different stakeholder groups under the conditions of developing governmental institutions. We are aware of the sensitive nature of providing a balanced view of diverging stakeholder perceptions. To mitigate this, we have strived for transparency and trustworthiness by providing many quotes and incorporating different voices in the narrative. Our study is an initial contribution in an under-researched context. We see many avenues for future research on sustainable development in developing countries. Frontier markets in particular, are at the crossroads; pushed to the breaking point by international competition and well-intended government interventions with risk of going awry. So far, the efforts to reach SDGs have led to negative outcomes, especially for local businesses and their communities in this context. Yet, by taking multistakeholder collaboration seriously, progress is attainable. We suggest future research to further unpack the role of collaboration and learning to achieve progress and the long-term sustainability of local businesses. In addition, future research can focus on the role of emotions and stewardship orientation in multistakeholder engagement, to explore under which conditions it can be gained or lost.
Funding Open access funding provided by Jönköping University.