Industrial policy, local firm growth paths, and capability building in low-income countries: lessons from Ethiopia’s floriculture export sector

Abstract

The industrial policy literature has made important contributions to understanding how to catalyze industrialization, but it tends to focus more on the industrial policies required to incentivize learning than on studying firm-level learning dynamics and investments in technological capabilities. Furthermore, global value chain studies on export sectors in low-income countries rarely consider whether local firms are part of broader business groups and if so, how that position within business groups affects local firms’ business strategies regarding the global supply chain in question as well as investments outside it. This article examines local firms’ investments in a new export industry, and they are shaped by dynamics linked to their family business groups, industrial policies, and the wider national economic and political contexts of low-income countries. It does so through a case study of Ethiopia and the emergence of its floriculture export industry. The article explains firm-level motivations in acquiring knowledge and building capabilities as well as their export trajectories and firm growth paths and how they are shaped by owners’ decisions related to the overall business group. It argues that industrial policies need to consider not only the existing capabilities, organizations, and networks of local firms but also tailor industrial policies to the potential of the family business groups to drive capability building in new industries.

1. Introduction

The process of capability building among local firms historically has been guided by government industrial policies in most, if not all, industrializing countries. Horizontal industrial policies removed or reduced constraints with economy-wide implications, while vertical industrial policies reallocated resources from relatively low- to high-productivity sectors and steered investment in desired directions (Chang, 2002). The role of industrial policy remains crucial in the 21st century, as it can still shape local firms’ resources, their utilization, and economic performance. Industrial policy is expected not only to provide the necessary support to local firms to increase their resources and facilitate national or sectoral innovation systems but also to leverage the potential role of foreign direct investment and global production networks in accelerating knowledge transfer (Chang and Andreoni, 2020; Hauge, 2020). However, increasing firms’ resources and facilitating access to foreign knowledge do not automatically lead to local firms’ investment in learning. Therefore, industrial policy needs to ensure that firms’ resources are exploited to build capabilities in the desired sectors through reciprocal control mechanisms (Amsden, 2001).

The industrial policy literature has made important contributions on how to catalyze industrialization in sub-Saharan African countries through discussions of policy design, bureaucratic effectiveness, capability building, and state–business relations (Cimoli et al., 2009; Altenburg and Melia, 2014; Page, 2015; Whitfield et al., 2015). However, this literature tends to focus on governments and external factors and to neglect firms’ agency in guiding resource reallocation as well as how firms respond to targeted industrial policies. As Chang and Andreoni (2021) note, structural development economic scholars tend to focus more on the industrial policies required to incentivize learning than on studying firm-level learning dynamics and investments in productive capabilities.

There is a nascent literature that examines African exporting firms’ learning and capability-building processes considering how global value chain (GVC) dynamics, foreign direct investment, and government industrial policies shape firm-level efforts (Whitfield et al., 2020; Whitfield and Staritz, 2021a,b). These studies affirm the importance of local firms’ initial resources, specifically finance, knowledge, and social networks, and show how those initial resources determine firms’ ability to assess the capabilities required; the capability gap they face; and the costs, risks, and rewards of their effort to bridge the gap. In revealing those firm-level dynamics, these studies emphasized the broader role of local firms’ agency. However, the GVC approach adopted in these studies focuses on firm-level economic upgrading within a particular global supply chain, which is implicitly assumed as the main trajectory for economic development. Thus, these studies do not consider whether the local exporting firms are part of broader business groups and if so how that position within business groups affects local firm business strategy regarding the global supply chain in question as well as investments outside it.

We build on this work and further extend the emphasis on firms’ agency by examining the role of family-owned diversified business groups in influencing local investments in the floriculture export industry in Ethiopia. The article does this by combining the GVC approach with insights from evolutionary economics and the diversified business group literature, which stresses that firms exercise their agency in much broader ways and decisions can be significantly affected by the other constituents of the business group.

The floriculture industry case in Ethiopia, a low-income country that is still in the process of industrializing, is significant as it is one of only a few new agribusiness or manufacturing export sectors that emerged in Ethiopia since the 1990s when the country began economic reforms to shift from a command economy to a mixed economy. Ethiopia became a top exporter of cut flowers, placing fifth behind the Netherlands, Kenya, Colombia, and Ecuador. The cut-flower industry emerged in Ethiopia through a combination of local entrepreneurial investments, government industrial policy targeting the sector, and Dutch development programs that provided direct support to the industry. The cut-flower industry created direct employment for over 40,000 people, generated around US$200 million annually, and induced investment and employment in other productive activities through backward linkages to the packaging industry and forward linkages to transport and logistics businesses.

However, outside of the nascent floriculture and apparel export sectors, Ethiopia’s level of industrialization remained low. Manufactured exports in total exports were less than 13% (Oqubay, 2019). Domestic private investment was largely focused on the service sector rather than productive sectors, due to high and quick profits in the real estate and rental, construction, trading, and banking sectors compared to the profit margin in manufacturing, particularly for exports, which operate in more competitive markets (Gebreyesus, 2019). Most local manufacturing firms were small, had low levels of productivity, and focused on the domestic market.

In this context where Ethiopia’s domestic private sector had not yet actively entered manufacturing, especially exports, the floriculture export industry offers a strategic case to assess the conditions under which local firms do invest in new industries and how they build capabilities to enter and remain competitive in labor-intensive global supply chains such as floriculture. The locally owned flower export firms were all part of family business groups that owned firms in several domestic market-oriented industries. Most of these business groups emerged after 1991, and this was their first investment in an export industry. This article examines local firms’ investments in capability building in the floriculture export sector and how they were shaped by dynamics within their family business groups as well as the external environment in which they operated, including the features of specific government industrial policies. We first assess the technological capabilities of the local floriculture firms and then analyze the technological effort they demonstrated and their motivations to invest in learning, based on data generated through an original firm survey and in-depth interviews with firms and other actors in the industry carried out between 2016 and 2018.

The analysis shows that owners often sought only to develop the minimum capabilities needed to sustain their cut-flower exports. Government industrial policies targeting the floriculture export sector included loans at below-market rates from the state-owned Development Bank of Ethiopia, which incentivized local entrepreneurs to invest in the sector, but they did not entail performance standards that tied the financing to export performance and thus firms had little incentive to continue to invest in learning. In this context, firms’ decisions and growth paths were influenced by the limited resources of their family business groups and showed significant variations. Firm owners sometimes chose to invest less in building the capabilities of their flower export firms in order to release resources and managerial talents to other affiliate businesses or make investments in new domestic market industries. On the other hand, some family business groups transferred resources from their domestic market-focused firms to subsidize the costs of building capabilities in their flower export firms. The analysis also shows the development of this first generation of family business groups in Ethiopia, as the higher organizational and managerial demands of operating a firm in a competitive export industry precipitated a move toward modern management practices. These findings emphasize the need to consider local firms’ capabilities and position with family business groups when designing industrial policy (Puente and Schneider, 2020).

2. Conceptualizing how local firms build capabilities in a mature technology export sector

The costs for local entrepreneurs in low-income countries to enter a mature technology sector like floriculture vary depending on the characteristics of the “imitator firm” acquiring the technology and the environment in which it operates (Perez and Soete, 1988). Perez and Soete (1988) outline four elements determining the actual cost of entry, of which fixed investment costs of the plant and equipment are only one. The others include the cost of scientific and technical knowledge as well as the costs of acquiring the experience required to use the equipment effectively to make products, achieve a viable market share, and make a profit. These intangible assets—knowledge and experience—are not free, and their costs increase the larger the gap between the knowledge and experience required and the relevant knowledge and experience already possessed by the local “imitator” firms. The process of acquiring knowledge and experience is not easy, even for the most basic capabilities. Floriculture export production is a low technology sector, but given local firms’ limited exposure to international competition, can require complex organizational capabilities relative to the capabilities that local firms have.

The fourth element in the actual cost of entering a new industry for a local firm includes costs of overcoming “locational” disadvantages related to the general infrastructure and other economic and institutional conditions surrounding local firms (Perez and Soete, 1988). In most low-income countries where there are few international competitive manufacturing and agro-industry sectors, local firms do not benefit from the economic externalities found in more advanced economies, including closeness to innovating firms and input suppliers; adequate transport infrastructure; skilled workers and experienced managers; and advanced financial services, transport facilities, and utilities.

Thus, the cost of entry is high for local firms where knowledge and experience gaps are large and there are significant locational disadvantages that have to be overcome (Helfat, 2018). These costs of entry can be reduced through direct government support. A key aspect of industrial policy is to absorb directly or indirectly what would otherwise have been a cost to the firm. Such rebates are necessary for new industries to emerge in low-income countries because the initial costs of entry are very high for local firms, and it is uncertain how long it will take local firms to build the capabilities required to become competitive and thus profitable. The larger the capability gap, the higher the cost and uncertainty and thus the risk. At the same time, local firms are entering the lowest value functions in global supply chains which are characterized by relatively high competition and thus low prices. Local firms consider this risk/reward ratio when considering in which sectors to invest (Khan, 2009, 2018; Whitfield et al., 2020). Industrial policy works to reduce the risk by subsidizing the costs and incentivizing firms to invest in learning (Amsden, 2001).

In many developing country regions, family business groups are predominant among local firms that move into new manufacturing and agribusiness sectors because they perform a similar role of reducing or hedging risks, as the costs of learning in new industries are financed from the profits of firms in existing industries. However, the economic performance of family business groups has varied across developing country regions. Northeast Asian conglomerates developed in-house, largely fungible capabilities that they centrally managed, coordinated, and deployed to efficiently diversify into various businesses, where they engaged in a gradual process of learning to compete at the bottom end or at the middle layers and occasionally at the top of each industry (Amsden and Hikino, 1994). The economic performance of Korean family–owned diversified business groups, especially in becoming internationally competitive in important manufacturing sectors, is attributed to government industrial policies that incentivized firms to invest in targeted sectors and to become internationally competitive using performance standards tied to exporting (Amsden, 1989, 2001).

In the Latin American context, family business groups reduced their investment in proprietary technology and manufacturing following the investment of multinational corporations (MNCs) in those technology sectors to produce for the domestic market. The presence of MNCs with lower production costs in the domestic market reduced prices and thus lowered profit margins for the family business groups in manufactured goods compared to their businesses in relatively lower technology sectors such as commodities and services. In the absence of government industrial policies in most manufacturing sectors, the family business groups allocated more resources to sectors with higher profit margins without necessarily building high capabilities to compete at the top layers of industries (Schneider, 2009).

In Southeast Asia, family business groups were prominent among ethnic Chinese, whose strategic goals were family wealth creation and preservation rather than building the managerial and organizational capabilities of their conglomerate (Carney and Gedajlovic, 2002). These goals were their defense mechanism against the hostile environment and potentially predatory state that the Chinese entrepreneurs faced as a minority ethnic group. Such family business groups could decide to diversify to related activities or exploit fungible capabilities as far as it served their strategic goals.

Theorizing based on empirical cases across different regions, we see that family business groups chose different growth paths depending on the national economic and political contexts in which they operated, their existing capabilities, and government industrial policies. The costs of entry are a moving target, as they vary with firms’ capabilities and external environment. If we add GVC dynamics to the original arguments of Perez and Soete (1988), we can also say that family business groups may have chosen unrelated diversification growth paths initially due to their low capabilities and high costs of upgrading in each industry, given that they did not have proprietary knowledge and had to compete with lead firms in advanced capitalist countries (Amsden and Hikino, 1994). However, as the large Korean conglomerates built capabilities in new leading sectors such as electronics in the 1990s and 2000s, they moved to coherent diversification and then specialization growth paths where they engaged in proprietary technology innovation (Hobday, 1995; Mathews and Cho, 2000; Lee, 2013).

The cut-flower export industry is a mature technology sector. It is part of the floriculture GVC, where the economic activities of flower production are labor-intensive and thus over time shifted to developing countries using technologies and capital equipment produced in developed countries. Thus, the lowest value-capturing activities are in agricultural production, but they require sophisticated technical, organizational, and managerial capabilities typically found in industrial production in order to deliver a consistent, high-quality, perishable product on demand. The floriculture GVC is governed by “lead firms” such as the Dutch auction (where a majority of cut flowers are traded and sold to wholesalers or supermarkets), retailers, and other big buyers. These lead firms set the requirements and standards that dictate to varying degrees what, how, and under what conditions to produce, sell, and deliver. We argue that the floriculture GVC can be described in terms of a set of functions that suppliers pursue and that economic upgrading then means moving into higher functions.¹

The initial function that suppliers in low-income countries enter is producing cut flowers, packing them, and arranging the transport logistics. To enter this Grow, Cut, and Pack (GCP) function, suppliers must deliver flowers with buyers’ specifications in terms of varieties, quality, and quantity at a specified time and place so that buyers/retailers can cater to consumers with fresh flowers without having a need to store them. This implies that supplier firms need to have a strong capacity in process planning from production to delivery and logistics capabilities (Labaste, 2005). The Dutch auction’s rule dictates that supply and demand determines auction prices. In practice, however, prices can be influenced by the capabilities of a supplier to achieve consistency in terms of availability (at the auction clock), quality, and volume. Suppliers with a good track record in consistency and reliability have a better chance of negotiating the prices. The specifications of buyers in the direct sales channel depend on their end markets and market segments. Buyers’ specifications in the Middle East generally are not as extensive as in Western end markets, and there is variation in consumer taste between the two regions that shapes buyers’ requirements.

Economic upgrading in the floriculture GVC, or a specialization growth path, can take several directions that entail moving from just producing flowers to taking on new functions related to design, marketing, and technological innovation (Melese, 2018). One type of upgrading is to move into designing and bouquet-making, which requires producing a greater variety of flowers or arranging the subcontracting of certain varieties to other firms, expanding capabilities in designing and packaging, and closer relations with buyers. Firms can also invest upstream in commercial propagations and cuttings, for both direct export or to supply to export producers. This function requires capital investment in advanced greenhouse technology and higher skills among staff as well as securing licenses from plant breeders. Breeders exercise the right to control their varieties and “proper” usage of it and thus shape supply and demand, and hence price, through managing access to varieties. Breeders can enter exclusive agreements with producers, which allow them to introduce unique varieties that few other firms have opportunities to offer globally (Azizi, 2020).

The move from exclusive rights to produce certain varieties to breeding new flower varieties in-house is a major technological jump, which requires capital investment in laboratory equipment, higher and innovative staff skills, and advanced research and development that involves years of experimenting. It also requires strong marketing and branding capabilities and the ability to enter socially embedded networks, as the floriculture GVC is governed by a handful of Western firms (Levelt, 2010).² Furthermore, the breeding function requires capabilities to protect and control property rights on which their royalties depend and to engage in deals that shape the demand and supply of their new varieties. Plant breeders are protected by the International Union for the Protection of New Varieties of Plants; however, the enforcement of this regulation differs, as Western countries tend to observe it more rigorously than countries in the Middle East.

The conceptual framework presented here combines insights from the evolutionary economics theory of the firm, especially the costs of becoming competitive, the role of diversified business groups in dealing with those costs, and the importance of industrial policy in lowering them, with the GVC approach to understanding the exact entry barriers for specific GVCs, what a specialization growth path means in terms of upgrading, and the challenges to specialization. Using this framework, we show in the following sections that government industrial policies were central to incentivizing the owners of Ethiopian diversified business groups to invest in the floriculture export industry because they reduced the costs of entry. The diversified business groups had no previous experience in agro-industry or manufacturing export sectors and thus were far away from the minimum required capabilities to achieve the productivity level, quality standards, and delivery speed of existing suppliers. They faced a large capability gap between the capabilities in their existing firms and thus higher costs: investments in building technical, organizational, and managerial capabilities to just enter the floriculture GVC and sell flowers at the Dutch auction.

Due to their low capabilities, the local flower firms generally sought to diversify their end markets for selling cut flowers to reduce their risks, but their growth paths did not involve upgrading. Upgrading within the floriculture GVC required significantly higher capabilities and building social networks linking them to buyers and breeders. Given the costs associated with building such capabilities and networks and the absence of government industrial policies providing support in these areas, local firms did not pursue upgrading. Instead, the Ethiopian family business groups sought to grow through related investments in horticulture for domestic and export markets, where they could use capabilities acquired through their flower farms and through unrelated investments in the production of goods for the domestic market where risks were lower and rewards were higher.

3. Development of Ethiopia’s floriculture export industry

The floriculture export industry in Ethiopia began in the 1990s because of the initial experiments of two local entrepreneurs to produce summer flowers and then roses, using the support of international donors to access the necessary resources. New macroeconomic and trade policies by the government attracted the first foreign firm to the sector, which started growing roses using more advanced greenhouse technologies, despite widespread market and institutional failures in the country. The rapid growth of the floriculture export industry was due largely to sector-specific government industrial policies launched in 2002, which were the result of lobbying by the floriculture industry association in Ethiopia. The industrial policies provided cheap land in water-abundant areas near the international airport; soft loans from the state-owned Development Bank of Ethiopia; income tax holidays; fiscal incentives that effectively reduced corporate tax; and removal of tariffs and duties on capital goods, spare parts, and inputs. The Ethiopian export promotion agency worked closely with the floriculture industry association to address infrastructural and bureaucratic problems in implementing the industrial policies, and in 2008, the Ethiopian Horticultural Development Agency was set up to continue supporting the development of the sector.

These incentives attracted further local and foreign investment in the sector, including experienced Dutch and Israeli firms. The entry of foreign firms, particularly the Dutch, significantly contributed to the rapid development of sector-specific services and the creation of critical mass which was essential for collective actions. The Dutch firms were mostly family-owned businesses that came with important experience and knowledge and included Sher Ethiopia, a subsidiary of Afriflora, one of the biggest rose producers in the world. Local firms benefited from the presence of these foreign firms through demonstration effects, information shared among production managers through their informal managers’ network, and sometimes via subcontracting and joint ventures between local and foreign firms.

However, growth in the sector slowed after 2008, as Figure 1 shows, partly because of the global financial crisis. The number of foreign and local firms declined, and export volumes largely stagnated until the mid-2010s. The Ethiopian government tried to shield flower firms from turbulence in international markets through loan rescheduling and additional financing, but its efforts did not stop the exit of many local firms from the sector. The number of local firms had been declining already from 2007 as the sector matured. By 2015, local firms accounted for only 15% of export volumes.

Local firms exited the sector for various reasons but crowding out by foreign firms was less likely as they were not competing in the domestic market, but rather exporting to both high- and low-value export markets. The collapse of local firms was due to political reasons such as politically motivated arrest of owners and natural disasters such as a flood, as well as the low capabilities of many local firms. The low capabilities resulted in poor feasibility studies structuring initial investments, diverting the soft loan for unintended purposes, and poor management of the flower firm by family business groups (e.g., assigning family members with no appropriate expertise or skills). Additionally, the Development Bank of Ethiopia and other government agencies lacked the required capabilities and political will to select appropriate investors (firms), set performance criteria, effectively monitor their loans, and take credible actions to enforce contracts.

The remaining local firms operating in the Ethiopian flower export industry in 2016 were only engaged in the GCP function of the floriculture GVC. Within the GCP function, there are different end markets and sales channels, as indicated in Figures 2 and 3. These end markets carry different risks and rewards and involve different relations between buyers and suppliers. Furthermore, within the Dutch auction sales channel, there are layers of governance such as the institutional rules of the auction “clock system” which largely serve as a minimum standard for the entire floriculture GVC; governance exercised between “auction-direct” buyers and supplier firms; and governance between unpacking agents and supplier firms. These governance layers are largely ignored in existing floriculture GVC literature, but they played a crucial role in catalyzing Ethiopia’s local floriculture exporting firms to invest in learning and building their capabilities, as shown in the next sections. The Dutch auction remained the most important sales channel for Ethiopia, as shown in Figure 3, but the direct sales channel increased in both traditional and emerging markets, particularly in the Middle East.

4. Characteristics of the Ethiopian-owned floriculture firms

Of the 15 surviving local flower firms, 10 were connected to family-owned business groups. Five of the firms were part of diversified business groups that had other firms in trading, agriculture, and manufacturing for the domestic market. Another five firms were owned by the Ethiopian diaspora who had foreign citizenship and lived abroad for a long period, but almost all these diaspora owners were connected to family business groups in Ethiopia. Once they returned to Ethiopia, they invested in one or more businesses alongside floriculture, and some of them set up their floriculture firm under an existing umbrella of their family business.

Two firms were owned by endowment funds affiliated with political parties in the Ethiopian People’s Revolutionary Democratic Front (EPRDF) coalition government, one of which was affiliated with the ruling party in the coalition at the time of the research. The last three firms were owned by the MIDROC Investment Group of Sheikh Mohammed Al-Amoudi (born in Ethiopia to Ethiopian and Saudi Arabian parents).

Table 1 summarizes the key characteristics of the Ethiopian-owned floriculture firms, with anonymized names. These firms had been operating in the industry for 5–13 years, and on average they employ 464 workers and have a landholding size of 45 ha, out of which 19 ha was cultivated with 12 rose varieties. Most of the firms (10 out of 13) financed their investment using loans from the Development Bank of Ethiopia, two firms used a private bank loan, and one firm that was a joint venture with a Dutch firm used a development bank loan and a Dutch subsidy, but during the research period, the ownership was fully transferred to the Ethiopian partner.

Many owners decided to invest in a flower exporting firm in response to the industrial policy incentives, which were being promoted by “investment brokers,” or observation of the growth of the floriculture industry in areas near them. The exception was the owner of D-Rose, whose entry into the flower sector was a deliberate plan to diversify to a related sector by tapping into its existing resources—its accumulated experience in growing vegetables for European markets in a joint venture with a Dutch company—and combining it with external resources made available through the sector industrial policy. Notably, the owner of D-Rose was also one of the pioneers of the sector that played a central role in its subsequent development and in designing the industrial policy as well. The rationale was different for the party-owned flower firms, which joined the sector to support the government’s industrialization effort, and MIDROC’s flower firms, where Al-Amoudi wanted to keep its good relations with the government.

We assessed the capabilities of local firms using a questionnaire designed based on a floriculture GVC-specific matrix of technological capabilities that we developed, inspired by the general matrix of technological capabilities introduced by Lall (1992) and incorporating the types of upgrading in the floriculture GVC described earlier. Based on the floriculture GVC literature and our experience in the field, we described the capabilities that were required for firms to successfully compete in the floriculture GVC at different functions or levels of complexity. The capabilities at each function were divided into categories that included investment, production, linkages, and end-market capabilities. We identified indicators for each category of capabilities, and then designed questions that could capture information on these indicators. The questionnaire was administered through face-to-face interviews with firm owners and managers of 13 local flower firms. We assessed the capabilities of the firms at each function in which they operated and then scored firms on each category of capabilities based on data from the questionnaire.

The analysis of these data presented in the next section focuses on the flower firms that were part of family-owned business groups. The owners of these flower firms, who were simultaneously the general manager of their flower firm and leader of a business group, had control over resource allocation and building capabilities across the firms in their business group. We carried out further interviews with the owners and the production managers of these floriculture firms, through which we aimed to understand firm-level decisions regarding export strategy and investments related to building capabilities. Based on these interviews, we explain how local firms built the capabilities required to enter the floriculture GVC and their subsequent export strategies, which dictated which capabilities they needed to build.

5. Capability building and export strategies of local flower exporting firms

In the absence of relevant knowledge and experience inside their existing businesses, local flower firms needed to acquire foreign technology and knowledge related to floriculture production and export and to build capabilities from scratch. In doing so, owners drew on their social networks and existing managerial talents to gain insights about the business and available resources for it. They gathered general information about the market and where to find experts and input providers by interacting with other potential local investors in their respective networks, the industry association, and talking to foreign and local people who were working in the floriculture export sector. Most owners hired foreign consultants with sector-specific knowledge to conduct a feasibility study, but they also reached out to local consultants who were familiar with the requirements of the Development Bank of Ethiopia and its bureaucracy so that they could secure the necessary finance.

To execute the project and unpack the foreign technology, local firms hired foreign consultants who had experience in Kenya as well as in Ethiopia. Prior to the targeted industrial policy that provided subsidized investment financing, local investors such as the owner of B-Rose had relied on their ability to mobilize finance from the domestic financial sector to pay for foreign expertise, which was possible only for a limited number of investors. The owner of D-Rose was able to secure finance and knowledge through a joint venture with a Dutch firm, which was made possible through the owner’s strong political and business networks. Loans from the Development Bank of Ethiopia made it possible for more investors to buy foreign expertise and conduct tours in Kenya and the Netherlands to gain knowledge from the flower sectors in those countries.

The foreign consultants hired by local firms played a fundamental role in enabling access to tacit knowledge and slowly building technical, organizational, and managerial skills. The industry association and Dutch development programs were also important sources of knowledge and catalysts of learning for local firms, especially in the early phase of local firms’ greenhouse investments and in mastering production capabilities. Later, relations with unpackers at the Dutch auction were crucial for firms to deepen their capabilities and thus capture higher prices.

Table 2 presents an overview of the technological capabilities and export strategies of the local flower firms. All local flower firms operated at the GCP function of the floriculture GVC and built their capabilities mostly to a medium level. Once the local firms entered the flower export sector, they were compelled to build their capabilities at least to the basic export production level due to the absence of a significant domestic market for cut flowers as well as because of the inherent narrow margin for failure that exists in the sector linked to the products’ high delicacy and perishability. None of the local flower firms invested in building capabilities to move into higher functions.

Rather, their growth paths involved deepening their capabilities within the GCP function to various levels depending on their export strategies and investing in largely unrelated businesses. We classified local firms’ export strategies within the GCP function into three types: deepening in the auction channel; concentrating in low markets using one or both sales channels; and diversifying into high- or both high- and low-end markets via auction, direct sales, or both channels. Even firms that followed the same export strategy demonstrated different degrees of success with that strategy due to idiosyncratic learning curves, as firms differed in their responses to internal and external stimuli as well as constraints.

Local investors had entered the flower export sector based on their perception of the risks and rewards of the sector compared to other business opportunities available to them as well as in relation to their family business groups. Moreover, the EPRDF government viewed businesses in a binary way: either developmental or anti-developmental (rent-seeking). Local investors in flower exports used their investments as a way to be seen as “developmental,” which allowed them easier access to foreign currency and land (which is largely state-owned in Ethiopia). It also allowed them to exploit the associated privileges beyond their flower firm to protect and expand their family business group. Operating in such a political environment, however, caused insecurity and precariousness among local flower firms, which might explain some of their decisions and reluctance to invest in upgrading.

Furthermore, as the domestic market was partly liberalized in the 1990s and 2000s, it led to economic growth and the emergence of a consuming class, which created wider business opportunities with undefined profit rates.³ Most owners of local flower firms had domestic market-oriented businesses such as importing, real estate, construction, and retail that required access to foreign currency and storage places. Generally, the other firms in their family business groups were less demanding in terms of the owners’ time but offered higher and quicker returns compared to the flower export business.

These factors influenced owners’ decisions about the allocation of resources and their management time across firms in the family business groups, which in turn affected the capability-building processes within their flower firms. Owners’ perceptions about the risks and rewards of the floriculture export business were subject to change as they gained more knowledge about the sector and as new attractive business opportunities emerged. We illustrate how these factors shaped the capability-building process and growth paths of the local flower firms through firm case studies.

A-Rose developed medium capabilities in the GCP function and exhibited growth for some time, but then its performance in the Dutch auction fell and its products garnered lower prices. The owner had decided to invest more in other family businesses such as construction of a hotel, real estate, and importing pharmaceutical products instead of reinvesting in the flower farm to replace old plants, test new varieties, or periodical expansion, and participating in marketing expos. The owner emphasized that he needed to invest in his other businesses, as they were more profitable and less demanding of his time compared to the flower business. The family kept the flower business due to the significant employment it created and to protect the family’s reputation and good relations with the government. However, instead of upgrading in the floriculture GVC, the family business group decided to diversify into producing vegetables and herbs using the capabilities accumulated through their flower export (greenhouse technology, organization, process and labor management, and export marketing skills). Horticulture products could be sold in domestic and export markets, lowering the risks related to currency fluctuations.⁴

G-Rose, one of the local firms with a lower capability score, exported via the Dutch auction channel until 2009 when it was hard hit by low prices due to planting the wrong varieties for the location of his flower farm. In response to this shock, the owner decided to channel money from his other businesses into the flower business to save it from collapsing. The owner identified marketing as the major weakness of this flower firm and allocated more of his management time to acquiring market information, searching for new buyers, and building relations with buyers while delegating more production-related responsibilities to salaried managers. Consequently, the owner of G-Rose changed its export strategy from Dutch auction to direct sales to the Middle East and invested in building the firm’s capabilities to the level required to succeed in this new export strategy. At the same time, the owner focused on his paper factory and continued diversifying into unrelated industries such as real estate, consultancy, and importing. The owner’s future plans regarding the flower export business were to diversify its end markets to southern Europe and Russia.

In contrast, the owner of B-Rose reallocated money and time from his flower firm to other businesses. As an early entrant to the sector, B-Rose set up the first local rose propagation facility and enjoyed first-mover advantages, but the owner’s commitment to the flower business declined over time, partly due to the global financial crises which forced some of its market agents in Europe out of business. B-Rose stagnated, which explains why it had the lowest capability scores. Apart from the market problem, one of the major reasons for its low performance was the owner’s decision to reallocate revenue generated by flower exports and his management time to his construction firm, which tied up a great deal of working capital and had a slow cash flow. At the same time, the owner continued investing in unrelated businesses such as a cement factory, beer factory, real estate, and rental properties. The owner explained: “we can’t let this farm tarnish the reputation of our family businesses so we did what we had to do.” Therefore, during the period in which this research was conducted, the owner decided to rebuild the production capabilities of his flower firm and was re-establishing his marketing networks at the Dutch auction while trying to establish new relations in the Middle East. In the future, alongside repositioning its flower business in higher-end markets, the owner intended to diversify into related businesses such as summer flowers, vegetables, and herbs, exploiting accumulated capabilities while reducing market risks, similar to A-Rose.

K-Rose was successful in the Dutch auction market and drew on existing capabilities to start production of fruit and vegetables for export, anticipating the arrival of sea freight cargo following the completion of the Ethio-Djibouti railroad. However, these plans were undermined temporarily by a fall in prices for its flowers on the Dutch market, partly due to the owner’s decision of assigning a family member as the marketing officer to operate in the Netherlands in proximity to the Dutch auction and Dutch unpackers. The unpackers did not see this decision as a positive move, and poor relations between the unpackers and the marketing officer negatively affected the performance of the firm’s products at the Dutch auction. The owner’s efforts to resolve the marketing crisis in the Dutch auction demanded more of his time, which forced him to hire a salaried deputy general manager to oversee the flower production. In the meantime, he sought new flower markets in the Middle East and invested in a new domestic market business producing bottled water.

Even the owner of D-Rose who had prior experience in the flower sector did not choose specialization as its growth path. D-Rose was established initially as a joint venture with a Dutch firm but became fully locally owned. The owner developed political and business networks at national and international levels, respectively, that he exploited to expand his other businesses. The owner neglected his 20-ha flower firm to expand his retail and food service businesses as well as build its network in Middle East markets. European buyers with whom the owner already built a strong network and trust perceived the drop in the flower firms’ product consistency as a temporary challenge and kept their contract with the firm. Despite his success in the Dutch market, the owner thought that “the auction is risky gambling business,” so he allocated more resources to expanding his maize seed production and vegetable exports to Europe and the Middle East, which reached over 280 ha and became the largest exporter of vegetables in the country.

However, some owners consistently built the capabilities of their flower businesses while also investing in unrelated businesses. For instance, E-Rose stood out with its higher competency and speed in accumulating the necessary tacit knowledge, developing relatively higher and dynamic capabilities that enabled it to quickly identify and respond to bottlenecks and redesign production processes. The owner built a strong network around the Dutch market and served on the Advisory Councils of the Dutch auction, representing member firms from Ethiopia (both local and foreign),⁵ and he was able to purchase and make profitable a collapsed foreign-owned flower firm. However, the owner emphasized that he has been cautious not to expand the flower firm beyond what he can properly manage, as he cannot assign other family members for the role because they do not have adequate experience in the flower export sector. The owner highlighted that his flower business, unlike his other businesses, helped him to build managerial and marketing skills in a dynamic environment, which he intends to exploit in running existing and new family businesses. Alongside deepening its capabilities to supply the Dutch auction, the owner plans to utilize the firm’s marketing capabilities to engage in market diversification and to expand his investment in manufacturing of packaging materials for the flower industry in Ethiopia as well as to strengthen his other existing businesses in furniture and paint manufacturing, real estate, and rentals.

For many firm owners, there was an internal competition between affiliate firms in their diversified business groups not only for financial resources but also for owners’ managerial talents and management time. Owners began to use salaried managers in the cases where trust was built between owners and salaried managers and where managers learned to successfully run the flower export operation. The owners of E-Rose and D-Rose even began to centrally control and coordinate other resources across their family business groups, such as sharing labor management and organizational skills among firms through joint training. However, most of the local firm owners struggled to find competent salaried managers and/or build trust with them, which affected their capability-building process. In such situations, the owners tended to neglect their flower firms, which caused significant damage at production and marketing levels. Nevertheless, all owners of local flower firms indicated their interest to delegate more responsibility as well as authority to salaried managers, and they were taking steps toward that objective.

Regardless of their level of capabilities in the flower export sector, the overall business groups of flower firm owners expanded. Table 3 summarizes local firms’ business groups before and after entry into the flower export sector. Given the national economic and political environment and governance in the floriculture GVC as well as the limited available resources and capabilities of firms, firm owners expected to continue allocating and reallocating resources across their diversified businesses based on internal competition or owner-managers’ strategy on the best way to increase their overall profitability as a conglomerate.

It is not clear from this research if the flower business served as a springboard for the growth, and to what extent business group owners could draw on the capabilities of the flower firms. However, some business groups that engaged in coherent diversification into summer flowers, vegetables, fruits, and herbs indicated that they planned to utilize their experience from the cut-flower sector in their new line of businesses, which are not technologically more sophisticated than their existing flower export business. None of the business groups entered other export sectors, although some groups entered new manufacturing industries focused on the domestic market.

6. Conclusion

The Ethiopian EPRDF government allocated resources to promote the flower export sector that incentivized not just foreign but also local investment in the sector. Local flower exporting firms, most of which were part of family business groups, built the capabilities required to be competitive at the lower end of the floriculture GVC. The risks of entering a new export sector like floriculture were high because there was not yet a considerable pool of local managers and workers that could be drawn upon and because the domestic market for cut flowers was limited and thus the costs of failure were higher than if the product could be sold in the domestic market should the firm fail to become internationally competitive.

Being part of a family business group was important as it allowed the owners of local flower firms to finance the losses of their flower export business during the learning period and when external crises occurred. Without such a financial buffer, single-product firms would have been forced to close if they had not achieved profitability by the time the investment loans from the state-owned development bank came due, or because of external crises that affected profitability. Notably, this is what occurred in the apparel export sector in Ethiopia, where the local apparel export firms that were not part of family business groups struggled to learn and compete in the lower functions of the apparel GVC and were forced to the domestic market or declare bankruptcy (Whitfield and Staritz, 2021b). Further research is required on how family business groups spread the risks and costs of learning in new agri-business and manufacturing export sectors.

Owners of the family business groups that invested in the floriculture export industry exercised agency in favoring one affiliate business over the other as they sought to increase the overall profitability of their business groups. The Ethiopian government’s industrial policy created incentives for local investment, but it did not include performance standards related to exporting in return for those policy rents. As a result, the local flower exporting firms built the minimal capabilities needed to sell their flowers in various end markets, shifting resources to other firms in the business groups. When crises occurred, they sought to keep the flower firm competitive, returning attention to it and investing in building further capabilities. Without incentives or compulsions to move into higher functions in the floriculture GVC, owners of the family business groups largely chose unrelated diversification as their growth path. Moving into the flower design and bouquet or the propagation and cutting functions of the floriculture GVC required significant investments by local firms in higher packaging and greenhouse technologies, higher staff skills, marketing capabilities, expanding assortments, and acquiring licenses from breeders. The business groups perceived such investments for upgrading as less attractive relative to other economic opportunities in the domestic market, given the high cost and uncertainty of becoming internationally competitive and thus the high risks associated with upgrading.

We have argued in line with Chang and Andreoni (2021) that the key aspect of industrial policy is incentivizing learning but that research on industrial policy in low-income countries needs to focus more on studying firm-level investments in productive sectors and learning dynamics. Understanding firm-level motivations for investing in new export industries, especially the process of acquiring knowledge and building capabilities, and their decisions related to export trajectories and firm growth paths are essential to designing industrial policies. Family business groups are a dominant form of business organization in developing countries because they reduce the costs and risks of investing in new sectors, and they provide finance and other resources in a context of underdeveloped financial markets.

Thus, family business groups are best placed to invest in new export industries but are unlikely to do so without government industrial policies because the costs and risks associated with becoming profitable are higher in export industries than in domestic market-oriented sectors. At the same time, it is export industries that demand higher capabilities and drive catch-up industrialization. Industrial policies need to consider the existing capabilities, organization, and networks of local firms, as Puente and Schneider (2020) argue, but also tailor policies to the potential of the family business groups to drive capability building in new industries.

Footnotes

References