Jump to ContentJump to Main Navigation
The Practice of Industrial PolicyGovernment—Business Coordination in Africa and East Asia$

John Page and Finn Tarp

Print publication date: 2017

Print ISBN-13: 9780198796954

Published to Oxford Scholarship Online: April 2017

DOI: 10.1093/acprof:oso/9780198796954.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 18 January 2022

A Natural Experiment of Industrial Policy

A Natural Experiment of Industrial Policy

Floriculture and the Metal and Engineering Industries in Ethiopia

(p.191) 10 A Natural Experiment of Industrial Policy
The Practice of Industrial Policy

Mulu Gebreeyesus

Oxford University Press

Abstract and Keywords

Ethiopia represents an excellent case study of recent industrial policy experimentation in Africa. The country is well known for its successful promotion of the cut-flower industry through business–government coordination. What is less known is that, at almost the same time, it was also using coordination to promote the metal and engineering industry, albeit with little success. This study provides comparative analysis of the policy process and outcomes of the interventions in these two industries. Examining why one intervention worked and the other failed in the same political context and institutional setting provides a natural experiment from which to draw valuable lessons.

Keywords:   Africa, business–government coordination, Ethiopia, industrial policy, policy experimentation

10.1 Introduction

Industrial policy is back on the development agenda (for example, Hausmann and Rodrik 2003, 2006; Khan and Blankenburg 2009; Cimoli, Dosi, and Stiglitz 2010) and several countries in Africa and the rest of the developing world have already started to re-introduce industrial policy. African governments have also collectively taken initiatives to promote industrialization in Africa under the theme ‘Industrialization of Africa’ at their January 2008 summit (UNECA and AU 2013).

Learning from past experiences is an inevitable process of policy-making. And yet the focus of learning has been on the successful industrialization of East Asian countries. Industrial policy-making in Africa needs to take account of the specific character of the continent. In this regard, the continent should critically evaluate its own experience and learn from its success and failure.

This chapter aims to contribute towards this by providing evidence of success and failure of industry cases from Ethiopia. The country represents an excellent case study as it is among the leading countries in Africa of the recent wave of reintroducing industrial policy. The focus of this study is on two sectors, floriculture and metal and engineering, both of which are among the priority sectors for promotion by the government but performed differently despite similar policy environment. Understanding why one intervention worked and the other failed in the same political context provides valuable lessons.

(p.192) The main critic against industrial policy has been the inability of state bureaucrats to pick winners and the high probability that they promote rent-seekers instead (for example, Krueger 1974; Pack and Saggi 2006). As a result, the state–business relation (SBR) and particularly the rent management has taken centre stage in industrial policy debate (for example, Hausmann and Rodrik 2003, 2006; Khan 2008; Khan and Blankenburg 2009; Schmitz, Johnson, and Altenburg 2013). Schmitz, Johnson, and Altenburg (2013) defines rent management as government intervention for raising (lowering) profitability in selected sectors and thus making private investment in these selected sectors more (less) attractive. They argue that success of industrial policy depends on how we deal with the following four categories of risks of rent management.

  1. (i) political capture by private investors and allied policy makers;

  2. (ii) choice of wrong instruments;

  3. (iii) targeting the wrong sector/technologies; and

  4. (iv) doing too little.

This chapter adopts these critical success factors for rent management as analytical framework to understand the success or failure of industrial policy in the two sectors under consideration. Some more specific questions we would like to address in our quest for lessons from the rent management practice in the two sectors in Ethiopia are:

  • How were the sectors selected and what are the instruments used by the government to create incentives (rents) in each sectors?

  • How are results defined and monitored and what are the sanctions instituted?

  • What are the institutional arrangements and particularly how does the government manage the tension between coordination and capture?

  • How is the private sector organized in its dealing with the government and each other?

  • How did these sectors perform and what are the key drivers of success or failure?

Many of these questions are about institutional linkages and coordination, thus, cannot be answered quantitatively. The analysis is, therefore, mainly qualitative and descriptive. A number of available databases (for example, Central Statistics Agency (CSA), Ethiopian Revenue and Customs Authority (ERCA), and surveys conducted by Ethiopian Development Research Institute (EDRI) on floriculture and metal industries) have been used to highlight each sector’s structure and relative performance. Moreover, secondary sources (published and unpublished documents from academia and policy circles) (p.193) and additional interviews with key players in each sector (associations and government agencies) have also been used in the analysis.

The rest of the chapter is organized as follows. Section 10.2 highlights the current debate concerning the implantation of industrial policy. Section 10.3 provides a general background of industrial policy in Ethiopia. Sections 10.4 and 10.5 respectively discuss the implementation and performance of industrial policy in the floriculture and metal and engineering industries. Section 10.6 tries to draw lessons from the two cases.

10.2 Principles of Successful Industrial Policy: Brief Review of the Literature

There seems little dispute on the theoretical arguments justifying industrial policy. The objection towards industrial policy often revolves around implementation and management. For example, the rent-seeking literature (for example, Krueger 1974; Tullock 2005; Pack and Saggi 2006) provides at least two main reasons as to why industrial policy is doomed to fail. The first is related to the fact that governments have no sufficient information to make the right choice, that is, to pick the winners. The second is that industrial policy is an invitation to corruption and rent-seeking, referred to as political capture. They, therefore, argue that rents should be avoided or minimized.

Khan (2008) and Khan and Blankenburg (2009), on the other hand, emphasize the advantage of looking at industrial policy through the lens of rent management. They argue that catching up with advanced countries requires rapid and sustained productivity growth which in turn depends on the creation (or learning) of new technologies. Building up technological capabilities can create a very high return in the future but the private sector will not play a big role particularly at the early stage because of the ‘risk’ involved. Governments should, therefore, create rents to attract investment in the selected sectors through different instruments. According to them, the challenge is then not to avoid rents but to manage them.

There are different perspectives on how to ensure good governance and avoid political capture in implementing industrial policy. One view is to create arm’s-length relationships between the government and private sector implying the need for autonomy of bureaucracy from the pressure and lobby group. The criticism against this view is that it implicitly assumes an omniscient government (i.e., the principal has already needed information), thus, limiting the importance of flow of information from below.

An alternative and widely held view is the concept of ‘embedded autonomy’, which highlights the need for dense links between government and the private sector. According to Evans (1995: 12), the government should be ‘embedded in (p.194) a concrete set of social ties that binds the state to society and provides institutionalized channels for the continual negotiation and re-negotiation of goals and policies’. Rodrik (2013) emphasizes on three key principles regarding the state business relation: embeddedness, discipline, and accountability.

The presence of strong private sector association is regarded as helpful for effective SBRs. According to Schmitz, Johnson, and Altenburg (2013) high organizational capacity of the private sector and broad-based membership of its association increases the likelihood of an alliance becoming effective and legitimate. Hausmann and Rodrik (2006) also argue that the trade associations might be a cooperative solution not only to the free-rider problem among private participants but also to improve the transparency and legitimacy of the relation between the private and public sectors. According to them the association should be left free to organize by itself at the level of what it deems is necessary. Forcing groups to organize according to some predetermined criteria, for example, by sector classification, may create groups that have few specific needs in common.

The rent management literature seems to implicitly assume that the potential sector is correctly identified and instruments are readily available. But the debate on how to choose potential sectors is far from settled. One dimension of this debate is whether the choice of the sectors should follow or defy existing comparative advantage (Lin and Chang 2009). Another strand of the debate is whether the factor endowment model that predicts broad structure of comparative advantage (for example labour-intensive, natural resource-based products, etc.) can give any practical guide to the choice of potential activities (Hausmann and Rodrik 2003). Hausmann and Rodrik (2006) argue that the new activities that become successful cannot be known with certainty ex ante. It is an ‘ongoing learning’ process, which requires entrepreneurial experimentation of the private sector. According to them it is, therefore, important whenever possible for the government not to predetermine whom it will deal with in terms of sectors or activities but making choices endogenous to an open process. However, this should not preclude setting a priority sector by the government ex ante. Rather, policy makers should be open to adjust their list to include new ones and/or exclude from an existing list upon information.

Similar arguments arise in relation to the required instruments and institutions for successful industrial policy (for example, Khan 2008). There are no bullet-proof instruments and institutional configurations that can work in every circumstance. The choice of instruments is also a matter of experimentation and compatibility with the organization and structure of a political poser in that society. Choosing wrong sectors and instruments is an inevitable but the most important thing is to put in place a mechanism to learn from mistakes. Industrial policy should, therefore, be seen as experimental process of trial and error. And again, an effective SBR can help reduce the risk of uncertainties in the choice of sectors and instruments.

(p.195) 10.3 The Evolution of Industrial Policy in Ethiopia: Some Background

The Ethiopian People’s Revolutionary Democratic Front (EPRDF)-led incumbent government adopted a Structural Adjustment Programme (SAP) in 1992/3, a year after it seized power. A series of reforms were then introduced to replace the command economic system by a market-oriented one. An export promotion strategy was adopted in 1998 as a reaction to the lack of progress in export diversification. A comprehensive industrial development strategy (IDS), that gave impetus to the export promotion, was then formulated in 2002/3. The IDS which is based on the government’s broader development vision—Agricultural Development-Led Industrialization (ADLI)—states some key principles such as: (i) strengthening the linkage between industry and agriculture, (ii) export-oriented sectors as a leader of the industrialization process, and (iii) due emphasis on labour-intensive industries.1

The strategy also recognizes the private sector as an engine in the industrialization process and at the same time explicitly states the need for a strong state role not only as facilitator but also as a leader. It made a distinction between ‘rent-seeking’ and ‘developmental’ capitalists and the need to curtail the former and promote the latter. The strategy cited two important mechanisms in which the government could engage and promote the private sector; creating conducive environment for investment and direct support for priority sectors. The sectors that were declared as priorities include textile and garment; meat, leather, and leather products; other agro-processing industries; construction; and micro and small enterprises (MSEs). The priority list has been adjusted through time. For example, the flower industry starting from the Plan of Action for Sustainable Development and Eradication of Poverty (PASDEP) covering 2005/6–2009/10 and import-substituting industries (ISI) such as metal and engineering and chemical industries with the launch of Growth and Transformation Plan (GTP) I (for the period 2010/11–2014/15) were officially rendered as priority sectors.

10.4 The Ethiopian Flower Industry

10.4.1 The Emergence of the Flower Industry in Ethiopia at a Glance

Table 10.1. The entry process and export patterns of the Ethiopian flower industry








Ownership type of the farm








Fully foreign







Joint venture





Total number







Total exports (million US$)







Source: Author’s compilations. The export figures are taken from the UN-COM TRADE database, while the number of firms from the author’s own surveys (for more on this see Gebreeyesus 2013).

Ethiopia’s attempt to export summer flowers, produced under the state farms, to Europe in the early 1980s did not last long. The private sector-led flower industry began to emerge in the mid-1990s in the aftermath of the extensive (p.196) reform programmes to transform the command economy to a market-oriented one. Table 10.1 gives the number of firms and the pattern of export values of the Ethiopian flower industry. The entry process at the early stage was slow and there were only five flower firms exporting no more than US$5 million worth of cut flowers until 2003. But things started to change drastically after 2003 with the beginning of active government support to the sector. The number of flower firms doubled in only one year (2003–4) and reached sixty-seven by 2008. The sector was estimated to generate direct employment for about 50,000 people out of which above 70 per cent are women. In 2008, the industry became one of the five top export commodities for the country with more than US$133 million foreign exchange earnings. With an export value of nearly a quarter billion US$ in 2012, Ethiopia ranked as the second-largest floriculture products exporter from Africa (Gebreeyesus 2014).

10.4.2 Some Factors behind the Success of Ethiopia’s Flower Industry

How did the Ethiopian flower sector become visible and successful in such a short period of time? The following subsections try to examine the main factors behind this extraordinary performance with a focus on the policy management. Private Entrepreneurial Experimentation that Revealed Ethiopia’s Comparative Advantage in Floriculture

Ethiopia is endowed with the conditions required for a successful flower industry especially suited to the production of high-quality roses. The country also has comparative advantage due to its geographical proximity to the world’s largest flower market, the European Union (EU), which is the primary destination of flower exports from sub-Saharan Africa (SSA). However, (p.197) this latent comparative advantage was not realized until the mid-1990s when two domestic entrepreneurs (namely Meskel Flowers Plc. and Ethio-Flora) started production of flowers with the aim of exporting them directly to Europe.

In 1999, Golden Rose Agrofarms Ltd, a foreign-owned UK-based business started its production after thoroughly examining the performance of the two early entrants as part of its feasibility study. Unlike the two early entrants, Golden Rose located its farm in a highland area (about 2,060 metres above sea level) to enable the production of high-value roses, and introduced modern steel-structure greenhouses. While the first two movers ceased their flower production in the course of the time, between 2001 and 2003 four other rose farms joined the industry following the success of Golden Rose.

The early entrants had faced various uncertainties and incurred substantial searching costs. The first source of uncertainty was the choice of appropriate technology. Despite Ethiopia’s comparative advantage lying in the production of large budded highland rose varieties, the two early entrants started production in lowlands which is only suitable for small budded roses and summer flowers. It was Golden Rose that pioneered the production of highland roses in modern steel structure greenhouses in 1999. The early entrants had also encountered marketing problems. All had started exporting through the Dutch auction but later on at least two of them forced to change their route through direct sales in Germany as a result of low price and unexpected service charges at the Dutch auction. Shortage of skilled personnel was also another acute problem, whereby the firms opted for recruiting expatriates and initiating on the job training. Lately Aware but ‘Pro-Active’ Government

Besides technology and marketing knowledge the early entrants had also other challenges such as transport, land, and finance. In order to seek support from the government the few existing entrepreneurs (only five) at the time organized themselves and formed an association called Ethiopian Horticulture Producers and Exporters Association (EHPEA).3 The government was not aware of the potential of the flower industry in Ethiopia until the end of 2002 when it was approached by private entrepreneurs. Upon realization of the big opportunity the government decided to actively support the sector development not only as a facilitator but also assuming its coordination role. The government prepared a five-year action plan for the sector development with the participation of representatives from the association.

Accordingly, targets were set up to put 1,000 hectares under flower production by the end of five years (2003–8). To scale-up from the low base (covering no more than thirty hectares in 2002), government came in with a (p.198) multifaceted support focusing on: access to land, access to long-term credit, infrastructure, and air transport coordination. Starting from 2003, land held by the government was made available at a very cheap price within the vicinity of the airport in Addis Ababa, leading to the creation of flower enterprise clusters. Long-term credit was also made available through the state bank (Development Bank of Ethiopia) on very generous terms. Lending was undertaken against a project plan with no collateral requirements. The government has also played an important role to solve the transport problem by initiating close cooperation between the exporters and the state-owned Ethiopian Airlines. In an effort to support the booming flower industry Ethiopian Airlines purchased extra cargo planes as well as relocating its flight destinations. FDI as a Catalyst in the Sector Development

The government’s announcement of its engagement with the flower industry sent a positive signal on the prospect of the flower industry. As a result the sector witnessed speedy flow of domestic and foreign investors leading to the industry takes-off beginning 2004. Since then the foreign-owned firms started to dominate the flower industry (see Table 10.1), a significant number of them coming from neighbouring countries including Kenya, Uganda, and Zimbabwe. The large flow of foreign investors had helped not only scaling up the industry but also in the diversification of activities into summer flowers and production of cuttings as well as other types of roses. Managing the Political Capture and Coordination

How did the government organize itself and manage the rents it provided to the private sector in the flower industry? What were the monitoring mechanisms of performance and sanctions instituted in case of failing to meet expectations? The remaining part of this section provides some highlights on this focusing on the SBR.

The relation between the government and the private sector was exemplary from the commencement. Meles Zenawi, the late prime minister, was personally involved in the promotion of the flower sector, holding regular meetings with investors to get up-to-date information and seeking immediate solutions to the business constraints faced. See Box 10.1 for further elaboration on the involvement of Meles Zenawi. The minister of Ministry of Trade and Industry (MOTI) was also personally involved making regular visits to the farms to get first-hand information. The Horticulture Development team that was formed within the MOTI around 2002 as a window of interaction with the private sector was upgraded to agency in 2008 and named the Ethiopian Horticulture Development Agency (EHDA). The agency was set up to act as a one-stop shop (p.199) for services required by investors including capacity building, investment support, and market promotion.

The industry association, EHPEA, was the conduit for the successful collaboration between the private sector and the state in the sector building. The potential of the flower industry was revealed by the private sector lobbying process and not by individual firms but through association. The association was self-organized and focused on its activities. More importantly, the association was not merely involved in extracting rents from the government but also active in mobilizing its members and other stakeholders to resolve the coordination problems. It acted as a developmental partner involved in a (p.200) range of activities aiming to promote the sector. For example, in 2004 the association established a subsidiary named Ethio-Horti Share Company, which handles both input supply and export handling and forwarding services. In 2006, responding to the mounting pressure for standards compliance the association in collaboration of other actors initiated the national scheme of good agricultural practice (GAP) known as EHPEA Code of Practice for Sustainable Flower Production which is in alignment of international standards. To implement the GAP it provided a series of training programmes to its members towards certification.

The association has also developed a very strong connection with the international community which enabled to mobilize additional resources. It is also engaged in market diversification efforts through visits to potential market countries and invitation of potential buyers. It continues to organize international trade fairs in Addis Ababa on a two-year basis attracting hundreds of floriculture companies from abroad. These all give evidence that the emergence of the self-organized industry association was instrumental in managing the rents created to the sector.

The relationship between the government and the private sector was not only rosy. Some tensions have appeared which are worth stressing here to in the context of the rent management practice in the sector development. One such issue was the repatriation of foreign exchange of flower exports. As a reaction to the under reporting of export sales by many exporters the National Bank issued a directive in 2006 to monitor the export of flowers, foreign exchange repatriation, and trading prices.

The second tension arises in relation to bank loans. As shown in Section, the government, through the Development Bank, was the main source of long-term investment loan for the sector. Due to mismanagement or misconduct, some firms were not able to profitably run their flower business and put the Development Bank of Ethiopia (DBE) under dilemma whether to foreclose or reschedule the loan. The financial crisis in 2008–9 had aggravated the problem of servicing debt of many exporters and led to an industry level request of debt rescheduling. According to officials at the EHDA, thirty-nine out of the forty-two investors who had borrowed from DBE tabled such a request. Based on the request of the National Export Promotion Committee, a team comprised of staff members from different government agencies, was established to evaluate each farm individually and produce recommendations with the aim of preventing foreclosure. Based on the recommendations of the team DBE implemented a number of measures including rescheduling of loan repayments, additional credit for some farms and even direct management intervention in the most heavily affected farms. These measures have not only saved the sector from collapse but also the bank itself.

(p.201) Another emerging area of rent management concerns profit tax. Investors in this sector have been granted up to five year profit tax holiday. This privilege has, however, started to expire for most of the firms as the time passes. We have learned from the EHDA officials that some firms have already requested for an extension of the years of tax holiday on different grounds. This shows that tax have become an area of rent management. Profit tax is practically handled by the revenue authority. Yet we have no sufficient information whether these requests are genuine and on the response of the authorities.

Land provision has been one component of the incentive packages for investment in the flower sector. The government through the regional governments has made available thousands of hectares of land at a very cheap price primarily in the vicinity of Addis Ababa for interested investors. But not all investors have actually developed the land they acquired. We learned that the government is in the mood to reclaim the land from investors that are not able to develop for an extended period of time following its repeated calls to the private sector to rectify that.

10.5 The Basic Metal and Engineering Industry

10.5.1 The Five-Year Metal and Engineering Industry Development Plan and Actions

The focus of the industry promotion of the incumbent government in Ethiopia has been the export sectors. However, beginning 2006 the authorities entertained the idea of adding some import-substituting sectors in to the priority list as a reaction to the growing import dependence of the country. Consequently, the five-year GTP covering the period 2009/10 to 2014/15 identified the basic metal and engineering industry (hereafter, BMEI) as primary industries for import-substitution based industrial development. The BMEI was then declared to be included in the list of priority industries for promotion.4

In 2010, the BMEI sector strategy was designed with the objective and targets taken from the five-year GTP. The objective of the BMEI development strategy is not only to substitute the BMEI imports but also to strengthen other sectors including the export industries by developing local capacity of design and manufacturing and facilitating technology transfer. Accordingly, targets were set for the sector to be achieved by the end of the GTP period (2014/15).5

(p.202) The government has taken various measures to implement the sector strategy and meet the targets. First, similar to the other priority sectors, the BMEI has been granted economic incentives such as provision of land at a reasonable price, generous credit schemes (for example, 70:30 credit/equity), free duty on imported investment capital goods and raw materials, and up to five years’ tax holidays on profits.

Second, the responsible government entity coordinating the sector development was upgraded to institute level. Accordingly, the Metal Industry Development Institute (MIDI) was established in 2010 under the Ministry of Industry by ministries regulation number 182/2010 (Council of Ministers Regulation 2010) to lead the development of the sector.

MIDI has been actively engaged in the support of the sector development since its inception. It has produced a number of potential project profiles to attract domestic and foreign investment and provides continuous support to ongoing projects. It also supports the private sector capacity development such as investment capability (e.g., project development and acquisition of capital) and monitoring performance including quality and productivity improvement through the introduction of Kaizen. Towards this it has created a database for fifty-two medium and large (M&L) firms and eight ongoing projects, and allocated one engineer for fifteen companies to provide support and monitor performance of the private sector.

The Institute has also established close relation with the private sector through the industry association. It provides office for the association free of rent in its premises and also holds monthly meetings of the joint committee of the industry association and MIDI often identify the challenges and discuss on solutions. MIDI also facilitates dialogues between the private sector and the responsible government organs such as banks, revenue and customs authorities, and regional administrations in an effort of smooth implementation of the incentive schemes provided to the sector.

Third, in addition to the efforts to strengthen the private sector the government has also sought state direct investment to play a catalytic role in the sector development. Accordingly, in 2010 the government reorganized fifteen state-owned enterprises operating in the metal and engineering sector, most of which were part of the Defence Industry aimed to satisfy the military needs, and form a large corporation, namely the Metal and Engineering Corporation (METEC). The objective of the corporation is to help realize the GTP by accelerating technological transfer and capacity in the sector.

10.5.2 The Structure and Performance of BMEI in Ethiopia

Table 10.2. The structure of Ethiopia’s BMEI, 2012–13

Number of Firms

Share (%)

Gross Production Value (Billion Birr)

Share (%)

Manufacture of basic iron and steel





Manufacture of fabricated metal products





Manufacture of machinery and equipment (not elsewhere classified)





Manufacture of motor vehicles, trailers, and semi-trailers





Total BMEI





Source: Author’s calculations based on CSA survey report (2012/13).

In this part we examine the structure and performance of the BMEI in relation to other industries and against the targets set in the five-year development (p.203) plan. Table 10.2 presents the current structure of the BMEI in Ethiopia based on recent survey report of the CSA on M&L manufacturing sector consisting firms with ten and above employees. According to the survey report, in 2012/13 there were 243 M&L firms operating in the BMEI producing in sum about Birr 19.7 billion value of production. This is below 10 per cent of the total number of firms and 18 per cent of the value of production of the whole manufacturing sector.

The BMEI is customarily classified into two broad categories the basic metal and the engineering sub-sectors. The engineering sub-sector is larger in size but dominated by the manufacture of fabricated metal producing hand tools, spare parts, and cutleries, which are characterized by low technology activity. The manufacture of fabricated metal alone account respectively for 75 per cent and 53 per cent of total BMEI number of firms and gross value of production. The engineering industry also consists of about eleven vehicle assembly plants and twelve machinery and equipment manufacture plants, both of which are characterized by low value added activity.

The manufacture of basic metal industry in Ethiopia is small in size accounting for only 15.6 per cent and 18.8 per cent of the BMEI’s total number of firms and production, respectively. It is concentrated in producing basic construction materials such as reinforced bars, hollow sections and corrugated sheets as well as billets. Moreover, this sub-sector is heavily dependent on imported raw materials and locally available scraps rather than domestic iron making. (p.204)

Table 10.3. Performance of the BMEI against the target set under GTP

Domestic Production

Capacity Utilization

Per Capita Consumption in Kg

Target (billion Birr)

Actual (billion Birr)

Achievement against target (%)

Target (%)

Actual (%)

Achievement against target (%)

Target (%)

Actual (%)

Achievement against target (%)













































(*) Note: The performance of 2013/14 is based on assessment of MIDI in its communication with the private sector.

Source: Author’s calculation based on MIDI (2014) internal report.

Next we assess the performance of the BMEI sector against the targets set under GTP. Table 10.3 gives performance of the sector against the targets for the first four years of GTP-period (2010/11 to 2013/14). The figures are taken from MIDI recent reports, which in turn are based on information from CSA for the domestic production and ERCA for imports. The domestic production covers only the production of M&L manufacturing firms with ten and above employment size. Based on the M&L sector statistics we can show that the (p.205) domestic production of BMEI products grew more than four times in the first four years of the GTP period. This achievement is, however, only 59 per cent when compared with the target set. The capacity utilization also shows very weak performance. Despite the targets to reach above 90 per cent, the BMEI average capacity utilization remained below 62 per cent of the installed capacity over the period under consideration.

According to the MIDI report annual per capita metal consumption of the country has doubled in the first four years registering better performance (73.4 per cent) even against the target. The rise in demand for metal and engineering products is driven by the fast economic growth in the country and particularly booming construction sector. Unfortunately, this demand is still largely satisfied by the growing imports. For example, in 2013 (the recent available data) the total demand for metal and engineering products was about 116.3 billion Birr, out of which 83 per cent were met by imports.6 The imports share in this sector shows no sign of decline in the last fifteen years, suggesting failure in terms of imports substitution.

10.5.3 Main Challenges Facing the Ethiopian BMEI Perceptions of the Stakeholders

There exists some common understanding among the stakeholders (the government and private sector) on the BMEI underperformance. For example, a study committee was established constituting experts from different government organs such as the parliament, Ministry of Industry, and MIDI in an effort to understand and resolve the challenges facing the sector. The study report which was placed for discussion among the stakeholders on the 1 March 2014 identified a number of bottlenecks deterring the sector development some of which are listed:

  • Unfavourable tariff/tax structure: the existing tariff/tax structure encourages imports rather than domestic production and does not identify low and high value-added activities.

  • Lack of market and unfavourable government procurement system: the government is the main buyer of BMEI products but most government organizations prefer imports and have less interest in domestic products. And at the same time the government procurement system favours bulk purchases, which can only be met by imports.

  • Shortage of finance: the financial system favours importers over producers as well as short-term over long-term loans.

  • Poor-quality infrastructure: This includes frequent power cuts, low-quality telecom services, and shortage of land. (p.206) Absence of Distinct Instruments

The author’s recent discussions with the association and MIDI representatives confirmed that there is a shared view among the private sector and government regarding the challenges identified, most of which imply the absence of sufficient protection of the selected industries from imports. As shown in Section 10.3, the industries selected for import substituting including the BMEI and the chemical industries were granted the same package of incentives that have been given to the export sectors such as credit, land, free duty on imported capital and inputs, and tax holidays. The programmes that aim to enhance the capacity of the private sector have also equally applied in the export-oriented as well as the import-substituting selected industries. As far as our review, there was no revision of the tariff structure of competing imported products in favour of the domestic metal and other ISIs following the decision to promote them. The package of incentives that works for the export industry might be irrelevant or less sufficient for the promotion of ISIs. In sum, the promotion of the ISI needs to have distinct instruments particularly designed to address the bottlenecks facing the sectors and trade policy is one of them. Choice of the Sector and the Absence of Champion Products

Even if the development of the import-substitution industries requires some form of protection, in the present globalized world it is difficult to completely shield the industry from foreign competition. It is, therefore, crucial to identify the sectors whereby the country has some comparative advantage to kick-off the industrialization whether through the promotion of export-oriented or import-subtitling industries. In theory, there is no inherent conflict between the export-oriented industrialization (EOI) and ISI strategies. In fact the development strategies in the successful Asian countries (e.g., South Korea, Taiwan, and China) have always been a combination of ISI and EOI strategies. The most important issue is then in which sectors do you start the promotion. Taiwan and South Korea start implementing not only the EOI but also the ISI strategy with consumer goods such as textile, food, and other labour-intensive industries. In both countries, the heavy and chemical industry drive for import substitution only came lately in the 1970s following the success in the labour-intensive industries through EOI as well as ISI strategies.

Unfortunately, despite the comparative advantage of the country lies on labour-intensive industries, the sectors that have been selected to kick-off the ISI-based promotion in Ethiopia, the basic metal and engineering and chemical industries, are by nature capital- and technology-intensive sectors. The ISI strategy lacks the analysis of existing and future comparative (dis)advantage rather it relies on the need to tackle the growing import dependence and the possible role of BMEI and the chemical industries in import-substitution.

(p.207) The absence of careful analysis of opportunities and gaps has also resulted in the identification of a sector that is too broad to make effective action plans as a single industry. The BMEI is a highly diversified sector constituting resource-intensive industries (e.g., metal making), technology intensive industries (e.g., office, accounting and computing machinery, radio, TV and communication materials, machinery and equipment, vehicles and transport equipment), and low-tech or labour-intensive activities (e.g., fabricated metal, consumer electronics). As a result, no champion product has emerged yet in this process. Weak and Less-Motivated Private Sector

In subsection 10.5.1 we have shown the government’s extensive efforts to promote the BMEI and engage the private sector in the development course. However, there lies some weaknesses in the organization and motivation of the private sector. First, the industry association was formed by the recommendation of the government in a consultation assembly held in April 2007. Four months later (i.e., September 2007) the association was established with ten members and named as the Ethiopian Association of Basic Metal and Engineering Industries (EABMEI). Immediately, the EABMEI has successfully gained affiliation with MOTI. In the beginning the private sector was largely reluctant. Thus, the association continued to have few members, for example no more than thirty-two members until 2010. In recent years, the number has grown relatively faster and currently reached about eighty following the realization of some benefits from the close consultation with the government through their association. But still this number is only about half the number of operating companies suggesting the continued lack of interest among many.

A second weakness is that the organization of the association is too broad amalgamating various industries, which might be one cause of the lack of interest in the association. This is also a reflection of the fact that the initiative of forming private associations was driven from the government side. This is not consistent with the emerging literature that suggests that associations need to be organized at sub-sector level to address specific industry constraints and motivate members (for example, Hausmann and Rodrik 2006).

Third, the activities of the association have been limited to facilitating government and private sector consultations and particularly lobbying the government for the provision of improved services and other supports. So far, it has made very limited efforts to mobilize the private sector itself and other stakeholders (except the government) in the sector building. The contributions for members of the association are very small (about US$150 per annum) and unable to undertake extra activities. The association needs to be a developmental partner in the sense that it is not only a forum of government lobbying but also discharges its own fare share.

(p.208) 10.6 Lessons to be Learned

The aim of this chapter was to provide some insights from the success and failure cases of present industrial policy experiment in Ethiopia. We selected two sectors, the floriculture and the metal and engineering industry which are considered to represent respectively the success and failure cases. The two sectors are different in the context that the former is export oriented while the latter is importsubstituting. Moreover, unlikely to the floriculture the BMEI industry promotion has started relatively lately, thus, it might be too early to evaluate the full outcome in this sector. One can, however, draw some useful lessons from the undergoing experiments, which is what we have tried to do in this study. The main lessons to be learned from the case studies can be summarized as follows.

First, the selection of a potential industry for promotion has long been a source of controversy in industrial policy debate. The analysis of comparative advantage (actual or latent) can be an instant aid in this regard. Our case study shows that Ethiopia has a clear comparative advantage of high-value cut flower production over other competitors. This opportunity was identified by the private entrepreneurs through their costly experimentation. In contrast, the import-substituting strategy declares the capital- and technology-intensive sectors such as the BMEI as a priority despite the country’s advantage lies over labour-intensive industries such as textile, leather, or food. The import-substituting experience of successful Asian countries, however, suggest otherwise. The heavy and chemical industries were promoted at a later stage following the success in the labour-intensive industries through EOI as well as ISI strategies. The current performance of the BMEI as well as the experience elsewhere leads to the question that whether the selection of BMEI (but not other labour-intensive industries) for import substitution in Ethiopia at this early stage of development was proper.

Second, the emerging literature shows that a sector promotion had to focus on a specific activity with clear analysis of opportunities and challenges. The two sectors are very different in this regard which might have an impact on their relative success. The floriculture is a specific activity, while the BMEI which consists of highly diversified industries was broadly defined to make any effective industry action plan. The implication is that the future development plan of the BMEI sector ought to be based on a specific activity level.

Third, the specific activity-level promotion has also great implications to the choice of policy instruments and the nature of the engagement with the private sector. The successful discovery of the flower industry in Ethiopia was a result of private entrepreneurs’ experimentation. These enlightened entrepreneurs demonstrated that Ethiopia has great potential for the flower industry. In their effort to seek support they organized themselves and convinced the government to take this opportunity seriously. A warm relationship (p.209) was established between the private sector and the state since then. This relationship not only enabled the government to pick the floriculture as a priority sector but also facilitated the design of appropriate policy instruments addressing emerging bottlenecks. Identifying the binding bottlenecks and designing appropriate policy tools requires careful understanding of the dynamics of the system at local and global levels. This was not possible without the emergence of a vibrant industry association, the EHPEA, which facilitated among others sharing emerging concerns, disseminating the available knowledge and consensus building among relevant stakeholders. This is consistent with the view (Rodrik 2007) that a strategic collaboration between the private sector and the government is needed to uncover where the most significant obstacles to restructuring lie and what type of interventions are more likely to remove them.

Unlike the floriculture, the establishment of the private sector association in the BMEI was initiated by the government on a predetermined criterion. As a result, the EABMEI consists of diversified sectors that have few specific needs in common and the less motivated private sector. The association activities mainly focus in lobbying the government support but lack any significant initiative to mobilize private sector and other stakeholders towards the sector development. The presence of diverse sectors under this organization might have also hindered the design of appropriate instruments addressing sector-specific binding constraints. More importantly, there is a lack of distinct instruments addressing the special needs of the ISIs such as the BMEI, rather the authorities apply a more or less similar set of incentives and support programmes designed for the export sectors.

Lastly, the analysis shows that the management of rents that are provided to the private sector was crucial in the success of the flower industry. The SBR in this sector was exemplary but not necessarily always smooth. Over the course we have seen some sources of tensions between the public and private sector, for example, repatriation of foreign exchange, debt rescheduling and termination of tax holidays. The responsible government institutions have learned from their experiences. They began to strengthen their internal capacity and have developed a mechanism to monitor the performance and properly manage the rents. In contrast, no clear pattern has emerged yet in the BMEI sector in this regard.


Bibliography references:

Cimoli, M., G. Dosi, and J. Stiglitz (2010). Industrial Policy and Development: Political Economy of Capabilities, Accumulation. Oxford and New York: Oxford University Press.

Council of Ministers Regulation (2010). ‘Establishment of the Metals Industry Development Institute’, Regulation Number 182/2010, Federal Democratic Republic of Ethiopia, Addis Ababa.

(p.210) CSA (Central Statistics Agency) (2012/13). ‘Report on Small Scale Manufacturing Industries Survey’. Addis Ababa: CSA.

Evans, P. (1995). Embedded Autonomy: States and Industrial Transformation. Princeton, NJ: Princeton University Press.

Gebreeyesus, M. (2013). Industrial Policy and Development in Ethiopia: Evolution and Present Experimentation, WIDER Working Paper 2013/125. Helsinki: UNU-WIDER.

Gebreeyesus, M. (2014). A Natural Experiment of Industrial Policy: Floriculture and the Metal and Engineering Industries in Ethiopia, WIDER Working Paper 2014/163. Helsinki: UNU-WIDER.

Hausmann, R., and D. Rodrik (2003). ‘Economic Development as Self-Discovery’. Journal of Development Economics, 72: 603–33.

Hausmann, R., and D. Rodrik (2006). Doomed to Choose: Industrial Policy as Predicament. Cambridge, MA: Harvard University Press.

Khan, M. (2008). ‘Governance and Development: The Perspective of Growth-Enhancing Governance’. In GRIPS Development Forum (ed.), Diversity and Complementarity in Development Aid: East Asian Lessons for African Growth. Tokyo: National Graduate Institute for Policy Studies.

Khan, M., and S. Blankenburg (2009). ‘The Political Economy of Industrial Policy in Asia and Latin America’. In M. Cimoli, G. Dosi, and J. E. Stiglitz (eds), Industrial Policy and Development: The Political Economy of Capabilities Accumulation. Oxford: Oxford University Press.

Krueger, A. O. (1974). ‘The Political Economy of the Rent-Seeking Society’. American Economic Review, 64: 291–303.

Lin, J. Y., and H. Chang (2009). ‘DPR Debate: Should Industrial Policy in Developing Countries Conform to Comparative Advantage or Defy It?’ Development Policy Review, 27(5): 483–502.

MIDI (Metal Industry Development Institute) (2014) ‘Metal and Engineering Industry Plan for the Second Growth and Transformation Plan Period (2015–20)’, draft mimeo.

Pack, H., and S. Saggi (2006). ‘Is there a Case for Industrial Policy? A Critical Survey’. World Bank Research Observer, 21(2): 267–97.

Rodrik, D. (2007). ‘Industrial Policy for the Twenty-First Century’. In D. Rodrik (ed.), One Economics—Many Recipes: Globalization, Institutions and Economic Growth. Princeton, NJ: Princeton University Press.

Rodrik, D. (2013). ‘Unconditional Convergence in Manufacturing’. Quarterly Journal of Economics, 128(1): 165–204.

Schmitz, H., O. Johnson, and T. Altenburg (2013). Rent Management: The Heart of Green Industrial Policy, IDS Working Paper, 418. Brighton: Institute of Development Studies.

Tullock, G. (2005). The Rent Seeking Society. Indianapolis, IN: Liberty Fund.

UNECA and AU (United Nations Economic Commission for Africa and African Union) (2013). ‘Industrialization for an Emerging Africa’. Paper presented at 6th Joint Annual Meetings of the ECA and AU Conference of African Ministers of Finance, Planning and Economic Development, Abidjan, Côte d’Ivoire 25–26 March. Available at: <http://www.uneca.org/sites/default/files/uploaded-documents/CoM/cfm2013/industrialization-for-an-emerging-africa-issuespaper.pdf> (accessed 16 September 2016).


(1) Gebreeyesus (2013), on which this section is heavily dependent, gives extensive review of the history of Ethiopian industrial policy.

(2) Note that the total number of firms at this time was about eighty-one, which means four firms were missing from the 2010 survey.

(3) The association now consists of about eighty-five members.

(4) The chemical industry was also similarly declared among the priority sectors for import-substitution-based industrial development.

(5) See Gebreeyesus (2014) for the detailed targets.

(6) See figure 2 in Gebreeyesus (2014).