2.1. The rationale for industrial policy in fostering structural change

The state has a central and pivotal role in fostering structural transformation in an economy. According to the ‘New Structural Economics’, although markets are the basic and first-best mechanism for the effective allocation of resources, the state should provide an enabling environment for firms to thrive in (Stiglitz et al., 2013; Lin, 2012; Lin and Monga, 2011). By accelerating the transfer of resources (capital, labour and knowledge) from low to high productive sectors, industrial policy has the potential to promote and sustain inclusive economic growth. Theoretical and empirical literature points out the importance of industrial policy in addressing distortions that constrain structural change. The first distortion relates to the presence of market failures; the second to coordination failures; and the third to technological accumulation and the acquisition of knowledge.

The traditional view in economics was that markets are efficient and state interventions should not influence the allocation of resources across sectors. However, there is a growing consensus that markets do not necessarily lead to efficient or desirable outcomes and the state has a role to play in this regard. One of the well-known market failures that industrial policy can address is information and cost discovery externalities (UNCTAD and UNIDO, 2011). According to Hausmann and Rodrik (2003), information externalities deter firms from exploring new economic activities, especially in developing countries where property rights are not enforced. This arises because the first firm to invest in cost discovery bears all the costs, while rival firms learn from the outcome of the first entrant. Due to this free riding problem, investment is minimal as no firm is willing to make any effort in the discovery of new products. Industrial policy can thus be used to promote entrepreneurial entry, survival and compensation for innovation through patent rights and copyright laws (Lin and Chang, 2009). In a cross-country study, Yasar et al. (2011) provide empirical evidence that supports a positive and robust relationship between firm performance and property rights. The authors show that institutions that secure property rights and promote sound legal systems are a pre-requisite to increasing firms’ performance and productivity in developing countries.

Another type of market failure relates to environmental externalities. These arise because firms, motivated by profits, do not incorporate pollution and environmental degradation costs in their investment decisions. Industrial policy can be relied upon to correct this, by supporting the development of green technologies, as well as production processes that are environmentally friendly, resource efficient and low carbon intensive (Hallegatte et al., 2013). In a study of industrial waste management in Taiwan, Tsai and Chou (2004) empirically show the effectiveness of industrial policy in addressing air pollution. They show that government policies such as providing financial support to firms to engage in waste-reducing initiatives such as recycling have been effective in promoting clean production techniques in the industrial sector.

The second need for industrial policy arises due to the presence of coordination failures (Pack and Saggi, 2006). Coordination failures occur because the feasibility and profitability of most economic activities is contingent on the existence of complementary investments. This implies that a firm is willing to invest in a particular sector if there are other firms that support its production process. In the absence of such an environment, entrepreneurial and domestic production may be adversely affected. Therefore, the state has a responsibility to promote and coordinate collective investment decisions from independent actors and firms (Altenburg, 2011). In an analysis of manufacturing firms in Ethiopia, Gebreeyesus and Mohnen (2013) provide evidence that supports the importance of firm coordination and networks in promoting technological innovation. The authors show that local business relations constitute the key channel through which firms acquire knowledge on market opportunities, new products, competitors and production techniques. In a case study of the flower sector in Ecuador, Hernandez et al. (2007) highlight how industrial policy fostered coordination between production on the one hand and the transportation of flowers to foreign markets on the other. The authors highlight the role of the association of flower exporters in convincing the government to increase the number of cargo flights by its national airline in order to promote the production and export of flowers.

Besides the need to correct market and coordination failures, industrial policy can address deficits in technological accumulation and learning among firms. In developing countries, domestic firms rely on existing technologies to boost their technological capabilities. However, adopting technology is usually costly and time consuming, as most firms either lack the necessary information or access to existing global technologies. In addition, inter-firm spillovers may hinder firms from investing in technology or knowledge accumulation as labour is mobile across firms. Given that knowledge is a public good whose usage is non-rivalrous, the market becomes an inefficient provider, necessitating the government to encourage the private sector to conduct research and innovation by ensuring strong patent and copyright laws that secure property rights (Stiglitz et al., 2013). Empirical evidence tends to confirm that the income convergence of East Asian countries towards that of developed countries was accelerated by industrial policies that promoted constant learning and knowledge accumulation among firms (Rodrik, 2009). This is in line with firm-level evidence that shows that patent rights have a positive and significant impact on the ability of firms to allocate their investment resources to research geared at developing new production techniques (Allred and Park, 2007).

2.2. The evolution of industrial policy in Africa

The evolution of industrial policy in Africa can be categorized into three distinct phases. These include the import substitution policies (ISIs), which began after the independence era in the 1960s through to the 1970s, the structural adjustment programmes (SAPs) of the 1980s, and the poverty reduction strategy papers (PRSP) in the 1990s. During these periods, most governments implemented various industrial policy strategies and interventions to promote industrial development. However, analysis of these strategies shows that industrial production did not take off, and de-industrialization occurred in most countries. Manufacturing as a share of output and employment decreased or remained low over most of the period (ECA and AUC, 2013).

During the ISI phase, industrial policy emphasized the production of consumer goods, which were initially imported. These policies were implemented in the belief that there already existed a pool of domestic consumers who would provide demand for these goods. It was further contended that existing local firms would eventually upgrade their production structures from consumer goods to capital intensive and high value added products, promoting structural change in the country (UNCTAD and UNIDO, 2011). However, ISIs did not spur industrial development as they required substantive government support, which proved costly and unsustainable. ISIs emphasized supporting and protecting firms rather than nurturing their competitiveness and as a result, most firms were not competitive or efficient.

The failure of ISIs led to the implementation of the SAPs. The SAPs were founded on the belief that industrial development was hindered by poor economic policies such as real exchange rate misalignments, trade protection, the neglect of the agriculture sector and interest rate controls (UNCTAD and UNIDO, 2011). Countries were required to liberalize their trade markets, de-regulate interest rates, devalue their currency and privatize state institutions. A key component of the SAPs was the combination of trade liberalization and withdrawal of government support to domestic firms. This limiting of the role of the state led to the collapse of most domestic firms as they were not able to compete with established international firms. Mudenda (2009) analyses the impact of SAPs on domestic garment firms in Zambia. The analysis shows that trade liberalization led to an influx of cheaply manufactured textile products, and as a result, around 35% of existing domestic firms closed down. This evidence is also consistent with other studies such as Tekere (2001), who shows that trade liberalization resulted in the collapse of clothing firms in Zimbabwe; Mohammed et al. (2011) finds the same in the case of Sudan, whilst Grosen and Coflkun (2010) argue that a decade of SAPs implementation in Tanzania did not address structural factors constraining industrial development.

Following the failure of both ISIs and SAPs, a third wave of industrial policy was implemented under the PRSP. During this phase, Africa’s lack of industrialization was attributed to the prevailing low levels of human development in the continent. Countries were required to channel their resources to social sectors, especially in education and health. However, PRSP did not translate to industrial development as emphasis was laid on providing basic education and health with no clear linkage to developing industrial skills for high productive and value added sectors such as manufacturing (ECA and AUC, 2013).

2.3. Industrial policy challenges in Africa

Industrial policy in Africa is constrained by several structural and policy factors. First, industrial policy in most countries has not conformed to comparative advantage. Most countries have ignored their relative abundance factors and instead promoted capital intensive industries, which are not able to take off. This has instead resulted in high production costs, over-reliance on government support, inefficient production structures and lack of competitiveness (Lin and Chang, 2009). Empirical evidence from Teixeira et al. (2014) shows that sub-Saharan Africa fails to ignite industrialization due to the support for production processes that are beyond its human and financial capacities. The authors argue that most countries failed because they promoted capital-intensive sectors rather than capitalizing on their comparative advantage in labour and natural resources. Hornsby (2012) critically examines Kenya’s industrialization history in manufacturing private vehicles in the mid-1980s. The analysis shows that despite significant subsidies, the eventual collapse of the industry in 2001 was attributed to a lack of sufficient financial resources and technology, factors in which the country did not have a comparative advantage. These examples are reminiscent of several case studies of industrial policy failure highlighted by ECA and AUC (2013).

Both theory and empirics underscore the need for industrial policy to support firms that effectively exploit the economy’s comparative advantage. The Ricardian theory of international trade highlights the benefits emanating from specializing and trading in sectors of comparative advantage. However, due to market and coordination failures, the new structural theory as well as the new trade theory lays the foundation for state interventions to address these failures. Murphy (2007) examines the ability of firms to upgrade their industrial production capacity in rural Tanzania. Based on empirical analysis, the author finds that government support in building local firm capabilities through skills formation can foster industrial development. Mccormick (1999) presents an analysis of industrial clusters in Ghana, Kenya and South Africa. The author shows that the differences in effective clustering and industrialization can be attributed to state interventions. Successful cases portray the role of the government’s support in providing infrastructure and a policy environment that favours business activities and enterprises. In the same vein, Rijkers et al. (2010) shows that supporting firms through the expansion of access to capital is a vital factor that can increase the productivity of manufacturing firms located in rural Ethiopia.

Proponents of this comparative-advantage-following development strategy argue that industrial policy should target a country’s abundant resources, such as commodity endowment and labour. By focusing on the availability of these resources, domestic firms will capitalize on existing resources and become profitable in domestic and international markets. Subsequently, this will lead to an increase in investment, market share, and competitiveness in domestic, regional and international markets (Stiglitz et al., 2013). Muchie (2000) examines the drivers of industrial development in Ethiopia. The author shows that the success of the leather sector hinges on the country’s comparative advantage in livestock resources as well as government initiatives to boost skills formation and technology adoption.

Second, the absence of sound institutions, bureaucratic capacity and governance frameworks constrains industrial policy across Africa. Industrial policy is designed with minimal consultation with key stakeholders, especially the private sector. Institutions in charge of industrial policy are characterized by competing and conflicting objectives, and industrial policy decisions are captured by interest groups and the political elite (ECA and AUC, 2014). Duru (2012) presents the challenges of the Nigerian experience of industrial policy, arguing that political interference and rent seeking practices have constrained its effectiveness. On the contrary, experiences of successful industrial- ization in both developed and developing countries underscore the importance of a visionary and highly skilled bureaucracy as well as government institutions that incorporate the private sector in the design and implementation of industrial policy (Hosono, 2015; Wade, 2009).

Third, industrial policy in Africa has disregarded country and sector specific conditions. Industrial policies are normally adopted from other countries without being tailored to local economic and political conditions. As a result, most policies turn out to be ineffective as they are not context specific and do not take into account a country’s endowments and initial conditions. Loewe (2013) presents a case study of industrial policy in Egypt. The author shows that its ineffectiveness could partly be attributed to minimal consultation with the private sector and state interventions that were only targeted to large and established firms. The analysis shows that industrial policies were not formulated in a clear and transparent manner, and they were not based on creating incentives for firms to engage in the production of high value-added products.

Chang (2013) examines four common industrial policy challenges argued to be the most binding in the African context. These include structural impediments such as climate, geography, culture and history; the abundance of natural resources; political economy issues; and bureaucratic capabilities. However, despite their validity, Chang argues that these arguments are largely theoretical and lack any empirical support. Chang argues that what African countries need is a deeper understanding of how to get industrial policy right by implementing sound and timely government interventions. Therefore, these arguments should shape the design and the form of industrial policy but not negate its importance in accelerating a country’s structural transformation process.

2.4. Industrial policy and state interventions

Despite the crucial role of industrial policy in spurring structural change, a contrasting body of literature argues that state interventions should not be used as they induce economic distortion through corruption and rent seeking (Rodrik, 2009). Corruption sets in as entrepreneurial firms are incentivised to demand favours rather than engaging in productive investment. This in turn distorts the proper functioning of the market and shifts economic rents and benefits to politically connected firms and the elite (Rodrik, 2009). It is also argued that the state lacks adequate information and capacity to identify sectors that industrial policy should target. Therefore, any failure in government intervention is more costly and harmful than the targeted failure to be addressed, and industrial policy is likely to result in supporting non-viable firms (UNCTAD and UNIDO, 2011).

Despite the validity of these claims, it has been argued that devising sound and effective institutions is vital in mitigating these failures (ECA and AUC, 2014). Bjorvatn and Coniglio (2012) develop a theoretical model, which shows that even in the presence of inefficient governments, industrial policy can provide a big push and enable developing countries to boost their industrial capacity. Successful experiences from the East Asian countries show that government failure was avoided through transparent and accountable institutions. Wade (2009) presents a case study of Taiwan and argues that the Industrial Development Bureau addressed these failures by designing incentives that had a sunset clause, promoting an educated bureaucratic workforce, and recognizing the private sector as a key player in the industrial process. Holcombe (2013) argues that industrial policy was successful in Japan and South Korea as it was targeted to turning small firms that had demonstrated their entrepreneurial and productivity capabilities into globally successful firms.

Given these successful experiences with industrial policy, most studies and policy documents (Hosono, 2015; Stiglitz et al., 2013) have focused on drawing industrial policy lessons for African countries from the experiences of East Asian countries. While this is insightful in drawing best practices, blueprint approaches have often resulted in policy failure because they do not take into account country and context specific circumstances (Rodrik, 2009). Industrial policy tools such as directing FDI into high technology sectors implemented in Singapore and Taiwan, export processing zones and under-valued exchange rates in China, and trade liberalization in Korea are less likely to succeed in most African countries, given differences in initial and present conditions as well as heterogeneity in factor endowments and economic institutions. The next section presents a case study of industrial development in Ethiopia, with the aim of distilling policy lessons that are likely to be suitable in fostering structural change in other African countries.

3.1. Overview of Ethiopia’s industrial performance

The Ethiopian economy is among the fastest growing on the continent. Table 1 highlights the country’s key macroeconomic indicators since 2000. The table shows that the economy has recorded a double digit real GDP growth rate since 2004 (with the exception of 2009 as a result of the global financial crisis). Real GDP per capita growth has steadily increased from 3.3% in 2000 to 10.8% in 2004, reflecting improvements in purchasing power and standards of living, before slightly declining to 6.5% at the end of the period. Sectoral composition of aggregate output in the country shows that the industrial sector has gained momentum in recent years.

An analysis of the industrial sector shows that value addition has consistently remained above 10% of GDP during the period from 2000 to 2009, as a result of increased industrial activity and coherent industrial policies. Based on the country’s projections, the industrial sector is envisioned to be a major source of economic growth and job creation in the long run (MoFED, 2010). However, the manufacturing sector has oscillated around 5% of GDP, as performance in the cut flower, textiles and apparel, as well as wood, paper and pulp sectors has slowed down (AACCSA, 2015). Nevertheless, as discussed below, the leather and leather product sector continues to play a significant role in the country’s manufacturing sector. In fact, according to Oqubay (2015: 5), industrial policy in the leather sector has mirrored the experience of the East Asian countries, although it has been tailored to take into account industry specific factors such as its low industrial base, path dependency in the form of low value trap, and integration into global value chains. In addition, ‘the depth of political commitment to consistency of policy instruments and compatibility of policy institutions has been significant to the success of the leather sector’.

3.2. Background of the LLP industry

The LLP industry was established in order to capitalize on the country’s comparative advantage in livestock resources. The country is endowed with a cattle population of around 52 million, ranking it first in Africa and sixth in the world. In addition, the sheep population is estimated at 27 million, placing it third in Africa and tenth in the world, while its goat population of 23 million ranks third in the continent and eighth in the world. The off-take rate (annual proportion of animals sold or consumed) for cow hides is estimated at 14%, 27% for goat skins and 40% for sheep (LIDI, 2013).

Given these vast resources, the country’s industrial policy prioritizes the LLP industry as one of the sectors that can foster structural change in the economy when aided by effective industrial policies. The LLP sector has been dynamic in nature, growing over time to fit into the country’s industrial development framework. In 2013, the sector comprised 29 tanning companies that added value to hides and skins, 16 medium and large scale footwear manufacturers with an annual production capacity of 15 million pairs, and 15 leather goods and garment companies that produced gloves, garments and bags (LIDI, 2013). Export performance has also been remarkable, as the sector has experienced rapid growth in export revenues from 880, 000, 000 birr (US$ 44 million) in 2004 to 2, 200, 000, 000 birr (US$ 110 million) in 2012 (UNIDO, 2012). In addition, foreign direct investment has increased from 920,000 birr (US$ 46,000) in 2004 to 304,725,000 birr (US$ 15,236,250) in 2009, primarily due to inflows from China and India (UNIDO, 2012). As elaborated in the country’s industrial development plan, the government has devised several policies to boost investments and expand the sector.

The government has engaged several donor agencies in the development of its industrial policy. For instance, the United Nations Industrial Development Organization (UNIDO), in conjunction with the MoI, has implemented a technical assistance project aimed at upgrading tanneries and leather laboratories. It has also organized international trade fair collections to boost market access. The government benefits from advisory services from the Japanese government and the Japan International Cooperation Agency (JICA) has engaged the government in policy dialogue and sharing of best practices and industry experiences. It has also been involved in productivity enhancement of domestic firms, through technology transfer and boosting managerial skills.

The leather sector provides significant employment opportunities. Fig. 1 shows that employment has increased over time, from 6989 workers in 2000 to 10,707 workers in 2010, representing an annual growth rate of almost 5%. This has been vital in providing job opportunities and income to the country’s abundant labour force, consistent with theoretical assertions that the industrial sector can be labour intensive and instrumental in absorbing excess labour, especially that of unskilled workers.

Fig. 1.

Employment trends.

Source: UNIDO, 2012.

In addition to creating job opportunities, the LLP industry acts as a source of earning and livelihood, contributing positively to poverty reduction. Fig. 2 depicts the trend of average monthly earnings, expressed in real birr (the local currency) for workers in major leather processing firms.

Fig. 2.

Average monthly earning. Note: I US $ = approximately 20 birr.

Source: UNIDO, 2012.

Over the period 2000 to 2009, the figure shows that monthly earnings in the leather, luggage and handbags sub-sector were consistently higher than those in the footwear sub-sector, given the high level of skilled workforce required in designing, sewing, coating and polishing of luggage and handbags. Although the figure shows a general downward trend in total monthly earnings of workers engaging in the leather industry – from almost 2000 birr (US$ 100) in 2000 to around 1000 birr (US$ 50) in 2009, these earnings are considerably higher than those of workers engaged in the exportation of raw and unprocessed leather products.¹ For instance, in 2009, the average monthly earning for workers exporting unprocessed products such as raw hides and skins of sheep – which is the most profitable - was around 625 birr (US$ 31), an amount that is 375 birr (US$ 19) lower than the earnings of workers in the major processing firms (USAID, 2013). Therefore, workers engaging in value addition in the leather sector seem to receive on average higher earnings compared to those exporting unprocessed products such as hides and skins.

Finally, Fig. 3 shows the average monthly distribution of workers’ earnings in 2009 and reveals that the sector’s compensation is significantly skewed towards low and medium wage employees who normally constitute a large portion of the workforce. The figure shows that, per month, around 20% of workers in the leather and leather product industry (LLPI) earned between 200 birr (US$ 10) and 400 birr (US$ 20), while the national absolute poverty line in 2009 was around 315 birr (US$ 16) per adult equivalent per month, implying that some of the workers’ earnings were below the estimated amount needed to secure the basic necessities of life (World Bank, 2014). Disaggregating the average monthly distribution of earnings according to the different leather products, the analysis shows that this skewness is more pronounced in the footwear sub-sector, where around 30% of the workers earned between 200 birr (US$ 10) and 400 birr (US$ 20). That workers remunerations are on average low in the leather industry suggests that while industrialization is desirable, especially in contexts where labour is abundant and considered to be cheap and attractive to foreign firms, government interventions are crucial in ensuring that the process of industrialization is pro-poor. This might involve policies that enhance income growth but also income redistribution, ensure that industrialization does not lead to vulnerable employment, and is ultimately associated with poverty reduction through higher earnings (Fosu, 2016; UNIDO, 2006).

Fig. 3.

Average monthly distribution of earnings - in birr (2009).

Source: UNIDO, 2012.

4.1. Industrial policies implemented by LIDI