Tension between state-level industrial policy and regional integration in Africa

Abstract

How does industrial policy currently interact with Regional Economic Communities in Africa? This article examines the tension between Regional Economic Communities and state-level industrial policy in Africa through two main problems: (1) how state industrial policies undermine Regional Economic Communities through a crisis of implementation and (2) how Regional Economic Communities undermine industrial policies through crisis of coordination. In discussing these issues, using indicative examples from the regions, especially the Economic Community of West African States (ECOWAS) and the East African Community (EAC), the article brings together industrial policy and regional economic integration in Africa.

Keywords:

• Industrial policy
• regional integration
• Africa
• implementation
• coordination

Introduction

Recently, there has been renewed academic interest in industrial policy in Africa (Chang, Hauge, and Irfan 2016; Chang 2012; Oqubay 2015). Meanwhile, African countries are in the process of forming Regional Economic Communities (RECs) and pursuing continental integration (African Union 2022). How does the industrial policy practiced in Africa interact with RECs and their free trade areas/customs unions? This article defines industrial policy as the government-led promotion of specific sectors and reveals a tension between state-level industrial policy and the form of RECs established in Africa. This tension is an expression of the globalisation paradox, namely the possible contradiction between states’ pursuit of development through industrial policy on the one hand and globalisation on the other (Rodrik 2011). At the regional level, the tension is a manifestation of the lack of institutional and political underpinnings of national markets in RECs. This article shows how the tension between industrial policy and RECs plays out in Africa through two core challenges: (1) a crisis of implementation, whereby industrial policy undermines RECs; and (2) a crisis of coordination, whereby the RECs undermine industrial policies.

On the crisis of implementation, Africa’s RECs have followed the sequence of market integration first outlined by Béla Balassa (1961): they started as free trade areas and later became customs unions. However, this standard model neglects the reality of domestic politics in African countries (Gibb 2009, 3), which is crucial for understanding the form and features of state industrial policy. The tension between RECs and domestic politics produces a crisis of implementation in the form of delays in the state implementation of their REC agreements, partial implementation, administrative difficulties, and policy reversals. This article will discuss the role of industrial policy in the crisis of implementation. Furthermore, implementation problems have negatively affected both industrial policies and firms’ regional strategies in Africa. As industrial policy is deeply political, given the link between domestic political interests and the distribution of resources in developing countries, it follows national politics. Regions, in contrast, lack the political structures required to support or generate a sector-based policy at the regional level. The structures of regions are heterogeneous, leading to variation in the form and origins of industrial policies and therefore potential coordination problems. Through various coordination problems the article will discuss how RECs are obstructing state-level industrial policies in Africa.

In discussing these issues, using indicative examples from the regions, especially the Economic Community of West African States (ECOWAS) and East African Community (EAC), the article brings together industrial policy and REC in Africa. The growing literature on the return of industrial policy has taken some recent developments into account, such as changes in the organisation of production in the form of the value chain (Oqubay 2015; Chang, Hauge, and Irfan 2016). The fact of regional economic integration and how it affects industrial policies in Africa have not been properly considered. Several RECs in Africa have produced documents on regional industrial strategies, such as the West Africa Common Industrial Policy 2015–2020 and the East African Community Industrialisation Policy 2012–2032; however, none of these could be meaningfully called an industrial or production strategy. Rather, they are statements of intention unaccompanied by a clear policy framework for specific sectors that countries are required to implement. At the same time, different states have used selected industrial policies. By demonstrating the tension between actual state-level industrial policies and regional integration, the article may inspire and guide future empirical research on the intersection and interaction between industrial policy and RECs. The article may also contribute to policymaking, because an understanding of how regional integration and industrial policy interact may help policymakers to design better production policies.

The article is organised as follows. First, it conceptualises industrial policy and situates it within the domestic politics of African countries. Second, it discusses the nature of market integration in terms of regional economic integration in Africa. Third, it deals with how industrial policy undermines RECs in Africa. Finally, it discusses how RECs undermine industrial policy in Africa. The third and fourth sections provide a discussion of indicative examples.

Politics and industrial policies in Africa

Several scholars have conceptualised industrial policies as government policies designed to catalyse the development and growth of specific industries or to alter the structure of production to emphasise sectors that are expected to generate structural transformation (Aiginger and Rodrik 2020, 16–17). Although the term ‘industrial policy’ traditionally refers to policies in the manufacturing sector, some scholars have used it more broadly. For example, Dani Rodrik used it to cover policies in agriculture and services, arguing that ‘there is no evidence that the types of market failures that call for industrial policy are located predominantly in industry’ (Rodrik 2004, 3). The use of the term in the current article is not restricted to manufacturing. For the purposes of this article, industrial policy is defined as the government-led promotion of specific sectors. This section reviews some practical political limitations on the implementation of industrial policy that makes it vulnerable to national politics, and hence creates problems for REC in Africa.

First, as industrial policies select among sectors, technologies, and firms (along with other implicit categories, such as ethnicity), they involve the distribution of state resources, which makes them contentious at least in a concealed form (Chang and Andreoni 2020, 335). By selecting between categories, industrial policies create winners and losers, at least in the short term, and therefore engender several lines of distributional conflict. For example, by creating policies that favour industrial capitalists in a specific sector, industrial policies generate distributional conflicts between industrial capitalists on the one hand and workers, subsistence communities, rentiers and competing capitalists on the other (Itaman and Wolf 2021). Some of these distributional conflicts play out at the level of the economy while others play out at the firm level. Other distributional struggles play out at the sector level, eg among capitalists operating in the same sector and capitalists in different sectors. The example below helps to explicate lines of conflict. The Nigerian Sugar Master Plan, an industrial policy with the goal of increasing domestic production from 40,000 tons to 1.7 million tons through a host of policies (such as banning imported processed sugar to give domestic producers the impetus to produce; restricting the domestic market to three producers/suppliers, ie the winners of the policy; providing tax credits, subsidies, training grants, etc.) created conflict between the selected capitalists and several groups: downstream beverage producers, who had formerly imported processed sugar but now have to depend on domestic producers; subsistence communities (eg small-scale land holders dispossessed to aid setting up sugar estates and petty sugar traders); importers of processed sugar; workers employed in the existing sugar supply chain; and rentiers (Majeed 2019; Premium Times 2019). Some of these conflicts made implementing the policy difficult. Within the underlying power relationships in society, individuals and groups with holding power are likely to impose a huge political cost on any resource redistribution that negatively affects them.

Second, appropriate policy design will depend on initial institutional and political conditions, as not all instruments are equally effective in different contexts, and success will require instruments that are most likely to be effective in specific country contexts (Khan 2015). As industrial policies involve the distribution of state resources, distributional policies that run against the survival interest of ruling elites (for example, by emboldening oppositional groups) will face direct challenges from ruling elites. At a minimum, therefore, the actual motivations behind the practice of industrial policies in Africa are better understood within a broader framework of patron–client politics. Accordingly, the origin of industrial policies in Africa has been studied in terms of ruling elites’ strategies to remain in power (Whitfield and Buur 2014) or, in the case of state–business relations, capitalists who are part of a ruling coalition (with ties to ruling elites; eg party funders) using their closeness to ruling elites to influence the creation of particularly advantageous policies in their area of business (Akinyoade and Uche 2018). A political problem that follows from the above is policy continuity during changes in political conditions (Odijie and Onofua 2020). The classical example is Ghana’s industrialisation attempt in the 1960s under the leadership of Kwame Nkrumah, which had to be discontinued by the succeeding administration partly because the policy was too closely connected with the ruling party (Killick 1976). An industrial policy that arises in a given political setting and is tied to a set of ruling elites may be threatened by significant changes in those conditions. Under competitive clientelism, where industrial policies are deeply political and designed to satisfy specific capitalists tied to specific ruling elites (such as policies that support party donors), political changeover will create problems of policy continuity. After taking over the affairs of the state, new ruling elites tend to discontinue or redirect policies that pay off the clients of their opposition.

Third, as the implementation of industrial policies involves input from different agencies, the fact of dispersed governance generates the challenge of consistency at the administrative and ministerial levels, especially where government agencies have some degree of autonomy. As government agencies/ministries may also be winners and losers in a given industrial policy, losing agencies could create problems. Ideological hegemony (which is a component of the developmental state literature) could help solve this problem of dispersed governance. The centralisation of policies in competent bureaucrats (either sector bureaucrats or general industrial bureaucrats) could also help solve the problem of dispersed governance. Even so, there is no guarantee of coordination, as some government agencies have an internal logic for revenue generation that may be threatened by industrial policies. For example, in the Nigerian Sugar Master Plan mentioned earlier, which was prepared and administered by sector bureaucrats on the National Sugar Development Council (a parastatal in charge of promoting sugar production), some government agencies voiced reservations due to the loss of revenue from the concessionary tariff and the quota allocation regime that affected their departments specifically, notably the Ministry of Finance and the Central Bank of Nigeria (National Sugar Development Council 2017). At the same time, these agencies were part of the implementation process, and their lack of actual cooperation affected the policy execution. There are several reports of the National Sugar Development Council writing to the Central Bank of Nigeria (which is in charge of monitoring the concession agreement) to call attention to the open abuse of the policy due to a breach of the concession agreement (NSDC Newsletter 2015). The NSDC urged the Central Bank to fulfil its responsibility by clamping down on such companies, but there was no evidence of such enforcement (NSDC Newsletter 2017). In a direct interview with the NSDC concerning the policy, ‘lack of inter-agency cooperation’ was mentioned as a main challenge affecting the policy (NSDC Interview 2019).

Fourth, the successful implementation of industrial policy in a given sector might be linked to complementary development in supporting sectors and areas (or structures), whose absence could be constraining (Khan 2015, 10). Even in selective industrial policies based on a single sector, the linkages between that sector and affiliated areas, such as production factors and other input markets, mean that growth and production transformation in the selected sector could be based on growth and transformation in other areas. The correlating demand for the output of associated sectors/areas (by virtue of development in a given sector) is not always sufficient to coordinate investment. The efficiency attribute of demand and market prices breaks down in several circumstances – for example, if affiliating sectors require first-mover investment without being able to capture the full benefits of the cost of learning or if a domestic power interest is preventing changes. To achieve successful sector coordination, the government agencies in charge of a given policy must be able to identify the affiliated sectors for inclusion in sector industrial policies.

The four points above consider the practical political limitations in the implementation of industrial policies in Africa. These limitations are made worse by REC, as discussed below. The importance of national politics means that industrial policy could be understood as limited by the contours of power and shaped by the playing field of national politics. Considering this, how would industrial policies be affected by a regional economic project that does not have an underlying state-like power structure or sectoral coordination? Theoretically, there is an incompatibility between the politics of industrial policies at the state level in Africa, which must take political reality into account, and the fully fledged market-based model of regional integration, which is the dominant model of integration in Africa (see Gibb 2009). To the degree that several patron–client neighbouring countries have different political processes and demarcated political spaces, it is difficult to imagine a politically attuned fully fledged integration. Accordingly, the politics of state industrial policy and regional market integration would seem to undermine each other, as discussed below and shown in the last section.

The nature of regional integration in Africa

Before discussing the tension between state industrial policy and regional market integration, this section critically presents the nature of RECs in Africa. The main families of theoretical approaches to regional integration lie in both political science and economics. The two sets of theories explain different things; while political science theories explain how regional integration comes about, economic theories explain why regional integration is desirable. In political science, the dominant approaches are neofunctionalism and intergovernmentalism (Schimmelfennig 2018). Neofunctionalism refines many of the ideas developed by functionalism, where integration occurs gradually through areas of mutuality, creating a web of international governance (Mitrany 1966; Mattli 1999a, 22). Neofunctionalism begins with the assumption that integration is optimal and provides an account of how integration evolves using concepts such as the spillover, collective governance of material interdependence, updating of common interests, and sub- and supranational group dynamics (Schimmelfennig 2018; Mattli 1999a). In contrast with functional theories, intergovernmentalism places state political leaders at the core of regional integration; it assumes that governments delegate authority to regional organisations to secure their bargaining outcomes but remain in control of regional organisations and the integration process (Moravcsik 1991). This realist framework assumes that state political leaders are always in charge of the process of regional integration and a hegemonic state can exercise control by imposing its interests (Mattli 1999a, 5). While these theories are often used to explain the development of European integration (Schimmelfennig 2018), they fail to explain RECs in Africa.

In relation to Africa, where different types of integration have been ongoing since independence, John Ravenhill distinguished between political regionalism, functional regionalism and economic regionalism (Ravenhill 2016). He defined political regionalism as involving institutions whose primary goal is to promote a sense of collective identity to enhance the voice of a group of states in global affairs. The debate that preceded the Organisation of African Unity in the 1960s could be viewed as political regionalism (Elias 1965). Functional regionalism, on the other hand, is basically a form of interstate collaboration for governance on specific overlapping issues (Ravenhill 2016, 3). Ravenhill identified a third form of regionalism in Africa, economic regionalism, which he conceptualised as ‘collaborative action by states to remove barriers to the flow of goods and services, migration and capital’ (Ravenhill 2016, 3). RECs are economic forms of regionalism; in RECs, facilitating trade is at the core of integration discourses. As the United Nations Conference on Trade and Development observed in their study of RECs in Africa, the regions ‘have all tried to follow a sequence mostly based on standard customs union theory’ (UNCTAD 2017, 5). Customs union theory is the dominant economic theory explaining integration; it emerged from the work of Jacob Viner and Bela Balassa (Viner 2014; Balassa 1961). In brief, customs union theory explains the welfare implications of economic integration in connection with trade creation, trade diversion and terms of trade (Mattli 1999b, 32).

However, some researchers have criticised this mainstream theory when applied to Africa (Rekiso 2017, 88–89). Africa’s main export products are oriented towards supplying developed countries with raw materials and therefore primarily face questions of how to diversify and industrialise. Some traditional proponents of industrial policy for diversification and building domestic industries, such as Friedrich List (1856) and his modern explicators, most notably Reinert (1999, 2019) and Chang (2002, 2007), have situated their historically attuned arguments for industrial policy in opposition to premature free trade. However, the rise of value chain analysis has altered the way free trade and industrialisation interact with each other (Gereffi 2013; Hauge 2020). Given the geographical distribution of modern production processes, free trade in a region could facilitate cheaper inputs as well as functional cooperation to aid both regional and global production networks. The goal of harnessing regional value chains is becoming the primary argument for both regional and continental integration in Africa (Chivunga and Tempest 2021; World Economic Forum 2021; African Union Commission 2022).

Both development institutions and scholars have created a new research agenda around building regional value chains as the route to industrialisation and structural transformation in Africa (Avom, Tiako, and Nguekeng 2021; Morris and Staritz 2019; Andreoni 2019). While some researchers focus on case studies of existing regional value chains (Banga, Kumar, and Cobbina 2015; Keijser, Belderbos, and Goedhuys 2021) others show the prospects of new value chains and argue for more integration to bring this about (UNDP 2021; African Union Commission 2022). The problem with this discourse is that it is superimposed on the form of RECs that already exist in Africa, a fully fledged market integration (Gibb 2009). The model of market integration to REC is based on Balassa’s analysis of economic integration, which involves a planned sequence, based primarily on standard customs union theory (Balassa 1991). It involves five stages of integration, as shown in Table 1 (Balassa 1976; Crowley 2006; Gibb 2009, 707).

Table 1. Balassa’s five stages of market integration.

Stage	Features
Free trade area	The abolition of tariffs and quotas within member states.
Customs union	Free trade area with a common tariff fixed for trade with non-members.
Common market	Free movement of goods and services as well as factors of production.
Economic union	Both supranational authorities and supranational laws need to be in place for harmonisation of some national policies and transfer of some policies to the supranational level.
Total integration	Single currency; harmonisation of taxes and democratic body effective at supranational level.

Source: Balassa (1961), updated by author.

Richard Gibb argued that when applied to Africa, a fully fledged market integration ‘neglects the influence of domestic politics’ (Gibb 2009, 3). Imagining an economic integration for encouraging regional value chains in specific sectors in Africa may be different from a fully fledged market integration. This is because while a fully fledged market integration is likely to be blocked by national politics, regional value chains could be negotiated at the sector level. For example, while it is difficult to trade cement in West Africa under the ECOWAS Trade Liberalisation scheme, the Nigeria cement giant Dangote Cement has successfully negotiated a regional production chain with several governments in West Africa according to their political realities. Dangote Cement currently sources limestone from Senegal where it operates a 1.5 Mta integrated plant and operates a bagging terminal in both Ghana and Sierra Leone; in Cameroon, it operates a grinding facility (Dangote Cement 2022). In other countries where there was no direct sector negotiation, for example Benin and Burkina Faso, local politics blocked Dangote Cement from selling its product.

At their core, regional value chains are a form of sector coordination that could be negotiated at the sectoral level. While arguing against market integration in Africa, Jonathan Bashi Rudahindwa and Sophie van Huellen proposed a commodity-based industrialisation through regional cooperation in sectors like cocoa–chocolate, where there is an overlapping need for sectoral transformation in West Africa (Rudahindwa and van Huellen 2020). Such a sector integration will be negotiated by the sector players and can easily align with the national politics of industrial policies in all the countries involved. Indeed, a bilateral agreement has been signed between Ghana’s Cocoa Board and Côte d’Ivoire’s Conseil Cafe Cacao (Ghanaweb 2021). Although this is a bilateral agreement and it arises from the need to create a living income differential, other cocoa producers in the region (such as Nigeria and Cameroon) are negotiating to join the agreement. The agreement could equally include issues of regional production chain as well as structural transformation in the cocoa–chocolate chain.

Such a sector coordination, which is needed for promoting regional production chains, is absent from market integration. Therefore, even though stakeholders are using the argument of building regional value chains to promote market integration in Africa, research has shown that RECs have not significantly increased participation in regional value chains (Obasaju et al. 2021; Slany 2019). Instead, state sector policies are undermining RECs and vice versa due to tension between sector policies pursued by different states. Using empirical examples from several regions, the rest of the article will show how this tension plays out through two core challenges. The first challenge is a crisis of implementation of the fully fledged RECs whereby state-level sector policies undermine RECs. The second challenge is a crisis of coordination, whereby RECs undermine industrial policies.

Implementation challenges: how industrial policy undermines RECs in Africa

RECs in Africa are faced with the political challenge of translating the goals and objectives of market integration into action at the state level. Theoretically, this problem of implementation is one of aligning regional market integration with the domestic political realities of the different countries involved with their different domestic groups struggling to retain their market advantages. Industrial policy tools such as import bans, import tariffs, import quotas, restrictive import licences, waivers and subsidies are used by political elites to allocate benefits to some of their supporters. Domestic actors that enjoy such advantages, which have to be forgone for REC to be fully implemented, will attempt to retain them. Within this dynamic, of potential blockage of market integration by domestic actors, is a range of problems, such as delayed implementation, the total rejection of REC, partial implementation, future unilateral reversals of implementation, and implementation without full administration. All of these problems have already arisen in African economic regional market integration.

Long delays in implementation are perhaps the most common problem facing market integration in Africa. Crises of implementation have affected all dimensions of market integration (UNECA 2019a). ECOWAS, for example, originally agreed to implement the CET in 2015, after lengthy negotiation (De Melo and Laski 2014). However, its implementation is still delayed in several countries (ECOWAS 2019). The recent Africa Regional Integration Index Report (UNECA 2019a) (conducted by the African Union, African Development Bank and United Nations Economic Commission for Africa) studied the level of integration in the eight RECs in different dimensions (trade, production, macroeconomics, infrastructure, and the free movement of people), using a composite index made up of several indicators reflecting national efforts to achieve regional integration in Africa (UNECA 2019b). For the Southern African Development Community (SADC), the average score for overall regional integration was 0.3 out of 1, but the SADC performed best in the dimension of free movement of people, mainly for historical reasons. For ECOWAS, the score for trade integration was only 0.4, and that for productive integration was 0.2. While ECOWAS also scored high for movement of people (0.73), only three countries – Burkina Faso, Mali and Togo – adhered to the Free Movement of Persons Protocol. The rest adhered to or withdrew from the protocol according to what domestic politics would accept (for the full report see UNECA 2019c). The EAC scored 0.44 for trade integration and 0.66 for free movement of people, similar to ECOWAS. Some of the political problems listed earlier (especially partial implementation, future unilateral reversals of implementation and implementation without administration) are responsible for the low level of trade integration.

As a practical example of this, under ECOWAS trade rules, cement produced in the region should not attract import custom duties and taxes when traded across borders within ECOWAS. Although Nigeria agreed to liberalise its cement sector within ECOWAS, a perpetual effective ban has been engineered by domestic producers (Byiers and Karaki 2017). Domestic cement producers have a strong political influence, leading to restrictions on importing licences as well as a ban on the importation of bagged cement (ie an effective industrial policy) since the early 2000s (Akinyoade and Uche 2018). Likewise, the attempt to develop rice production in Nigeria recently led to a raft of policies that that undermined ECOWAS trade; this included a total border closure with neighbouring countries to curb smuggling (Orjinmo 2019). Journalists’ interviews with members of the Rice Farmers Association of Nigeria (RIFAN) showed not only that the association supported the border closure, but also that it had long been lobbying the government to keep the border shut (Adeyeye 2019; Pulse 2020; Tribune 2019; Vanguard 2020). As the RIFAN chairman said in an interview, ‘those criticising the closure of the borders are not true Nigerians because it is one of the best decisions of this administration’ (quoted in Premium Times B 2019). Likewise, the rice producers of Ghana have been lobbying the government to ban importation of rice (Ghanaweb 2019) – as is the case with other West African countries. Meanwhile, the border closure resulted in a large shortfall in customs tax in Benin which is a significant part of the government revenue (International Monetary Fund 2021, 10). The Benin government responded by imposing a duty to make up for the shortfall in government revenue. It imposed a new transit duty of CFA9 million (over US$15,000) per truck on Nigeria-bound cargo transiting through the country (Adekoya 2021; Tribune 2021). Both the border closure and the transit duty undermined the West African REC.

A sectoral negotiation in rice production (ie regional integration in rice) could have prevented the above problems in West Africa and helped to develop the regional value chain. Researchers have shown that there was already a nascent rice value chain between Nigeria and some countries in the region (Karkare et al. 2022). The pattern of different groups calling on their governments to implement policies that potentially contravene regional trade rules is common in West Africa. The politically influential among them (with the ability to impose costs on ruling elites) seem always to have their way at the expense of the rules of market integration. There are copious examples of this in the EAC as well; and in the case of rice, poultry, sugar and milk it has led to trade tensions and problems of implementing the EAC free trade (Njeru 2021; Nakaweesi 2021; The Citizen 2018; Blanshe 2021; The East African 2021; EAC 2015).

Furthermore, there are times when ruling elites have to promote diversification in pursuit of their own political survival. When ruling elites face problems of reduction in the prices of their main export products or reduction in the value of domestic currency, they usually respond in ways that could hinder free trade commitments. For example, the need to diversify led the post-2015 Nigeria administration (due to the reduction in the prices of crude oil) to adopt a new approach to existing trade agreements, such as flagrantly refusing to obey regional trade commitments and imposing trade restrictions in many sectors. Besides the increase in products placed under import restrictions, the Central Bank of Nigeria (CBN), also in 2015, restricted access to foreign exchange at the official window for importers of several items that were traded in the region, such as dairy, textiles, poultry and rice (see International Trade Administration 2021). When a product receives government support, there is usually an accompanying change in policy that undermines trade with reference to the product. This could take the form of a unilateral import ban, the imposition of import levies and more creative policies like restriction of foreign exchange. For example, although rice is supposed to be traded freely within the EAC regional market, when Uganda invested in rice production in 2015, it introduced an extra 18% Value-added tax on rice originating from neighbouring Tanzania and a 75% Common External Tariff (CET) duty or $200 per metric ton on rice produced in Kenya (EAC 2015). Similar problems have occurred with reference to beef, cigarettes, etc, with negative effects for the REC (The East African 2016).

Coordination challenges: how RECs undermine industrial policy in Africa

Even though RECs are not properly implemented, there is still some evidence that they are undermining industrial policy in Africa. All of the political constraints on the implementation of state industrial policies are complicated by REC. The problem of winners and losers, for example, and how losers may jeopardise industrial policy, extends to the region as well. If it is assumed that industrial policies create national winners and RECs are used to promote the interests of domestic capitalists (by expanding their market share, for example), then this very process is creating losers in regional countries though winning over market share from less competitive firms. The national winners, however, do not have political power in other countries of the region and will therefore encounter problems of market access. However, as some sectoral policies in Africa now accommodate regional integration, such challenges reduce the predictability needed to develop a regional strategy. For example, Nigeria’s policy in cement production involved two stages. The first was to raise domestic production to the level of self-sufficiency through different production incentives, and the second stage was regional participation (Odijie 2020). However, with a lack of clarity in implementation of REC, the cement industry has found it difficult to develop a coherent long-term regional strategy for exportation. There is an effective ban on Nigeria’s cement in the neighbouring Benin (Global Cement 2019) and Burkina Faso (Nairametrics 2019), because national capitalists in these countries object to losing their market to Nigeria cement. Another example is that the Cement Manufacturers Association of Ghana, representing local cement producers such as Ghacem and Diamond Cement, has been lobbying the Ghanaian government to ban the importation of cement from Nigeria (Global Cement 2016).

While Nigeria’s cement sector has enough power to drive policies in Nigeria, this does not extend to other countries. Nigerian cement producers have had to reverse some regional strategies due to problems with implementation in other countries. For example, while Dangote Cement is interested in building roads linking Nigeria to neighbouring West African countries, which are target markets for the firm’s cement (Ezeoha, Uche, and Ujunwa 2020, 6), in practice the company faces political problems with exporting cement to neighbouring countries (Nairametrics 2019). Although Dangote’s plant in Ibese is less than 30 km from Nigeria’s border with Benin (a target market), the country imports from China and initially refused to allow Nigeria cement exporters to pass through its country to sell in other neighbouring countries (Global Cement 2019). Furthermore, while Dangote Cement’s company’s 2018 report disclosed its plans for a regional policy, the shutdown of the Nigerian borders affected the firm’s strategy for exporting cement through road networks. The executive director Devakumar Edwin told The Africa Report that ‘we were taken by surprise and had cement in the border for 30 days, then 45 days. We then had to start pulling them back because the cement will get caked’ (quoted in Adeshokan 2020). Without REC and the impression of a common market, Nigeria cement producers would have had to work bilaterally by negotiating directly with target markets. It is possible to imagine a situation of ‘shared winning’, where the winners in a given country cooperate with the potential losers of another country instead of taking over their market share. Such shared winning would, however, require a sectoral integration. In areas where Nigeria’s cement producers have succeeded more in the region, such as Senegal, there has been a sectoral negotiation with domestic groups that amounts to market sharing outside of the regional trade rules. Dangote Cement is beginning to employ such bilateral strategies to coordinate with domestic groups: for example, it negotiated with the government of Togo to pay an informal fee for using the country’s road network (The African Report 2020) and partnered with domestic groups to access the market (TogoFirst 2019).

Furthermore, because regional regulations are based on domestic implementation where domestic political logic applies, coordination problems are likely to arise from dispersed governance. The policy literature dealing with policy conflict within an overall government, whose different ministries are autonomous and therefore operate by an internal logic, could help theorise the various forms of coordination/competition problems that may arise within a region where different countries use national logic to approach their REC (Bouckaert, Peters, and Verhoest 2016; Peters 2018; Candel and Biesbroek 2016). Guy Peters listed seven such problems (Peters 2018), all of which could be adapted here. These are duplications, contradictions, displacement, a lack of vertical management, changing demands, cross-cutting problems, and simple tidiness (Peters 2018). As space does not allow for a rigorous elaboration of all, I discuss a few below.

In terms of how REC undermines industrial policies in Africa, duplication occurs when most of the countries in a given region have the same form of industrial policy in the same sector, without sector coordination, and in the process obstruct each other at the regional level. Traditionally, successful policies in Africa have involved the development of domestic sectors to the point of maturity before exporting to the region or competing in the international market. In this sense, market integration is premised on the idea of providing opportunities for African firms and industries to benefit from the economies of scale inherent in larger regional markets (Byiers, Karim, and Woolfrey 2018, 4). However, there is a sense that the success of a given sector policy in a regional market depends partly on other countries’ not pursuing production in the same sector (especially in perfect competition where product differentiation is difficult). If most of the countries in a given region have an industrial policy in a particular sector, the sector will be added to their national list of sensitive products, therefore blocking neighbouring countries from importing to them. There are theoretical grounds to expect that the process of market integration in Africa is leading to duplication, whereby regional free trade creates a signal for domestic interest groups to replicate the policies of neighbouring countries.

For example, the process of trade liberalisation usually includes a list of sensitive products, which are products that individual countries can either exclude from trade liberalisation or on which they can impose higher tariffs to protect domestic firms. In terms of regional protection, regionally sensitive products (which are negotiated on the basis of production interests, existing industrial policy, or revenue concerns) will be replicated at the level of several states. That is, the exclusion list presented by a single country that reaches the regional exclusion list (during CET negotiation or trade negotiations with other regions) will become the political basis for industrial policy in neighbouring countries. For example, if a given country in a region has an ongoing industrial policy in a given sector and therefore negotiates for the sector to be added to the regional list of sensitive products (which would make the sector’s products more competitive in the regional market), the sector’s inclusion on this list could lead domestic interest groups in neighbouring countries to request a similar policy in the sector, as it is now protected in the region. This type of coordination problem has occurred in all African regions. For example, in ECOWAS, within a week of agreeing on the regional exclusion list in 2014, the Ghanaian government yielded to pressure from interest groups to create a production policy in poultry using the advantage of a regionally sensitive product (Government of Ghana 2014). The regional protection of poultry was partly instigated by Côte d’Ivoire, which had had a production policy in poultry since 2004 and an active industrial policy developed on the basis of the regional market since 2011. Far from creating a regional market for potential exportation (which Côte d’Ivoire wanted), regional integration closed the market for Côte d’Ivoire through duplication, as the target markets of Ghana, Mali, Benin, Burkina Faso, Senegal, Gambia and Togo all instituted a similar policy within a few months of concluding the exclusion list in 2014. Theoretically, this is unsurprising in developing regions, as organised interests use the fact of regional protection (negotiation of a CET or trade agreement with other regions, such as the Economic Partnership Agreement) to activate a national industrial policy.

Duplication explains why most EAC countries are pursuing production in the same sectors (sugar, rice, milk and wheat) through industrial policies identical to those for products protected under the regional CET (The East African 2020; Shinyekwa and Katunze 2016, 10). Product differentiation is difficult in most of the following products; therefore, there is a constant regional obstruction in form of trade dispute in the EAC (The East African 2020). Duplication is not against regional trade rules, so there is no platform for its detection and correction at the regional level. Nevertheless, it obstructs regional markets in precisely the areas in which countries have industrial policies, which is where they needed regional markets in the first place. Furthermore, as states evaluate their performance internally, rather than as part of a larger unit, they are likely to see each other as threats. Indeed, the comparative advantage of a given country could be seen as a threat by another country (even when the other country does not have any production interest in the area or intending to duplicate it). For example, in the EAC, Tanzania and Uganda once imposed a 25% tax on Kenya confectionery specifically due to Kenya’s cheap source of industrial sugar, which makes its confectionary more competitive (Trademark 2018). Uganda imposed a similar tax on Savannah cement produced in Kenya for the same reason – not necessarily to protect domestic cement (in which Kenya had an advantage) (EAC 2015).

Lastly, contradictions could occur when two or more countries pursue policies that are directly contradictory within a REC. Contradiction can occur within sectors, between sectors, between production processes, between policy and structure, and so on. To expand on the contradiction between policy and structure (as the others seem straightforward), the learning dynamics that make up the process of production upgrading do not relate solely to the internal organisation of a sector; they also relate directly to the external and broader organisational and institutional structures within which firms and sectors are located. Therefore, industrial policies are not concerned merely with the physical processes of production or the organisation of productive capabilities with sectors. They are also concerned with (and based on) structural processes centred on a plurality of networked organisations and institutions (Khan 2015, 10; Andreoni and Chang 2019; Bianchi and Labory 2019). For example, the Ivorian government’s policy on cocoa production in the 1960s and 1970s led to changes in structures of land tenure and labour arrangements (to free immigration), both of which fuelled the succeeding increase in cocoa production (Odijie 2019). Furthermore, policy alignment must occur between complementary sectors; for example, education systems may have to adapt to provide necessary training and skills for newly developed sectors. REC creates a problem of contradiction between industrial policy and regional structure if policies take regional economic integration into account. The first form of this problem is that it would be difficult to induce a losing country to adjust its institutional arrangements to assist a winning country. Furthermore, slight changes in institutions, standards, preferences, organisations, etc. in a given country may create problems for policies in another. For example, land reform in one country (especially within a region like the SADC, where land issues are central to production policies) could create problems for neighbouring countries and their policies.

Conclusion

The article argues that there is a tension between state-level industrial policy and the nature of regional economic integration in Africa. The tension is discussed through two interrelated problems: how industrial policies undermine RECs and vice versa. The vision of REC conflicts with the politics of industrial policy, leading to problems of implementation. Problems of implementation negatively affect industrial policies that take regional integration into account. Furthermore, because regional regulations are based on domestic implementation where national political logic applies, other problems of coordination are likely to arise. Although promoting regional value chains is a core rationale for REC, value chain specialisation usually involves a highly disciplined process of production governance that is intentionally provided by lead firms and other groups without which coordination would not be possible. A direct sector integration/negotiation is the certain way to avoid coordination problems in Africa. In conceiving of a system of coordination at the regional level there are two broad categories under which different types of regional production can be discussed: positive and negative coordination/integration (Scharpf 1994).

Positive coordination goes beyond simply avoiding conflict to seek integrative ways of dealing with production issues. Positive coordination can take the form of a single or regional industrial policy. However, it is impossible to conceive of a regional industrial policy that states have to follow unless it is prepared to take different political realities into account. Several regions in Africa have produced documents on regional industrial strategies (which are not sector focused), such as the West Africa Common Industrial Policy 2015–2020; the East African Community Industrialisation Policy 2012–2032; the SADC Industrialization Strategy and Roadmap 2015–2063; and the The Common Market for Eastern and Southern Africa Industrialisation Policy 2015–2030. However, none of these could be meaningfully called an industrial or production strategy. Rather, they are statements of intention unaccompanied by a clear production policy framework for specific sectors that countries are required to implement. These regional industrial strategies were not necessarily created based on foreseeing and solving problems of coordination, and none have amounted to any practical action in the production sphere. Negative coordination, on the other hand, is broadly a negotiation process to avoid conflict in production; this could involve sectoral coordination at the regional level. A sector negotiation would necessarily take the form of a division of labour that could help promote a regional value chain. As shown in this article, Dangote Cement has adopted a system of regional coordination in the cement sector of some West African countries instead of trying to win against local producers. However, this process of negotiation cannot be left to private capitalists, since only a handful of West African producers have the influence of Dangote.

Acknowledgements

I thank all the participants in the special issue workshop for their valuable comments as well as the anonymous referees and the guest editor for constructive suggestions on this article.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Michael E. Odijie

Michael Ehis Odijie is currently a research fellow in the History Department of University College London. He has various research interests, ranging from slavery and abolition in Anglophone West Africa (current focus), to cocoa farming as well as industrial policy and the role of state in economic change. His recent relevant publications include ‘The Need for Industrial Policy Coordination in the African Continental Free Trade Area’ (in African Affairs: https://doi.org/10.1093/afraf/ady054), ‘Is Traditional Industrial Policy Defunct? Evidence from the Nigerian Cement Industry’ (in Review of International Political Economy: https://doi.org/10.1080/09692290.2019.1657480) and ‘Political Origin and Persistence of Industrial Policy in Africa’ (co-authored with Anthony Onofua, in Globalizations: https://doi.org/10.1080/14747731.2020.1714851).