The ownership aspect

Whether lead firms are foreign or nationally owned may have important consequences for linkage development. In general, nationally owned firms are seen as being more embedded in the local economy, in turn stimulating more extensive local economic linkages (Dicken Reference Dicken2011; Morris et al. Reference Morris, Kaplinsky and Kaplan2012). However, in the Mozambican construction sector, foreign companies – Chinese in particular – tend to win national and international tenders. A representative of a US consultant company (Int. 19.6.13) cited the example of a project tender in Maputo: of the 19 contractors participating, 11 were Chinese, two were Korean and the remainder European. Domestic companies are rarely involved in the bids, so Chinese companies compete with traditionally strong South African and Portuguese companies (Alden et al. Reference Alden, Chichava, Roque, Alden and Chichava2014). To understand challenges to linkages between Chinese and domestic companies in Mozambique, we need to explore how Chinese companies enter the Mozambican market.

They do this in two specific ways. The first involves bilateral relations, where China offers a complete project, from financing to the actual construction, on terms favourable to the Mozambican government. From the Chinese perspective, this modality is preferable, given Africa's history with cash expenditures: the method of disbursing funds through specific projects neutralises the possibility of graft and corruption. Moreover, the money will stay within the Chinese system and ensure that aid-financed projects remain under the purview of the Chinese government, especially regarding the contract-tendering process (Davies et al. Reference Davies, Edinger, Tay and Naidu2008). Many of these loan packages in construction come through China ExIm Bank, which is fully owned by the Chinese government. Its loans are designed to: (1) fund manufacturing projects, infrastructure construction projects and social welfare projects in the borrowing country, which can generate promising economic returns or good social benefits, and (2) finance the procurement of Chinese mechanical and electronic products, complete sets of equipment, technology and service and other goods by the borrowing country (Corkin Reference Corkin2013: 64–66). Such funding should be understood as tied, which in turn does not stimulate local supply-chain development or domestic backward linkages.

A second mode of entry relates mainly to competitive pressure in the Chinese construction sector, and support provided to Chinese businesses for globalising their activities. Chinese foreign investments gained momentum after the Chinese government formalised the ‘Going Out’ policy in 2001, encouraging Chinese companies to establish themselves abroad and compete globally. This policy package included a reduction in bureaucratic procedures, to make Chinese producers more foreign investment friendly, with tax incentives, subsidies and access to loans credit and support from state institutions (Kragelund Reference Kragelund2009). Mozambique is among the developing markets where Chinese companies have found it lucrative to establish themselves. Morris et al. (Reference Morris, Kaplinsky and Kaplan2012) found that Chinese companies entering African markets without attachment to specific Chinese-funded projects would tend to be more embedded in the local economy, but only very subtle differences emerge in the cases examined here. The lack of local integration was generally understood as ‘business as usual’ for Chinese contractors, and was seldom contested:

Morris et al. (Reference Morris, Kaplinsky and Kaplan2012) found that Chinese (and Indian) lead firms were more reluctant to outsource activities outside their core competences, in comparison with South African and other ‘northern’ lead firms. Chinese interviewees claimed that the nature of the work did not require extensive outsourcing, as the company had the capacity to execute the entire project. Moreover, this made it easier to control costs. Cost is relevant here: price remains a major competitive factor for many Chinese firms, and Chinese control of the main service and supply chains is seen as an important element in their ability to keep prices competitive:

In light of the price element, Chinese firms could become ‘less eager to localise their labour’ (Corkin Reference Corkin2012: 476) and enter into partnerships with domestic firms. Even in the housing project, and given that joint ventures create certain expectations of stronger local integration (Amendolagine et al. Reference Amendolagine, Boly, Coniglio, Prota and Seric2013), the motivation behind this business structure was to keep costs down. The choice of partnering with a state development fund as the minority partner was justified by the need to have a local associate that could minimise costs and facilitate bureaucratic procedures. The use of Chinese subcontractors was also motivated by low costs, as Chinese companies were thought to be able to meet any type of budget (CEO, housing project, Int. 10.6.13). However, as noted by Bräutigam (Reference Bräutigam2008), having a governmental body rather than a local entrepreneur as economic partner in joint ventures could limit the potential for knowledge and skills transfer to local firms.

Representatives from the business association for contractors characterised the organisation within Chinese businesses as very closed and isolated from society (Int. 14.05.13). This description fits the organisational structure of the Chinese businesses working on the three projects examined here, where closed-off areas outside main centres housed all the Chinese managers and workers, where they socialised and enjoyed separate cooking facilities and health services. According to Morris et al. (Reference Morris, Kaplinsky and Kaplan2012), one consequence of expatriate groupings in remote areas abroad is that the managers have little or no knowledge of the local economy. That complicates the creation of domestic backward linkage creation, as such relations require a certain level of trust. This aspect might be further exacerbated by general communication difficulties. In the projects examined here, language and cultural differences were seen as a major challenge, and many of the people involved in project execution had no common language to communicate in. This point is supported by Amendolagine et al. (Reference Amendolagine, Boly, Coniglio, Prota and Seric2013), who found that higher language and cultural barriers could be a plausible partial explanation for the lower impacts of backward linkages from Chinese investments in Africa.

However, this does not necessarily mean that non-Chinese foreign companies create more domestic backward linkages. Despite efforts to reduce tied aid OECD-DAC donors since the mid-1990s (Knack & Smets Reference Knack and Smets2013), most ODA contracts still go to donor country firms (Ellmers Reference Ellmers2011; Mawdsley Reference Mawdsley2012). Moreover, multinational companies tend to create follower suppliers, whereby lead suppliers locate their supply operations near those of lead firms with whom they already have network relations (Morris et al. Reference Morris, Kaplinsky and Kaplan2012). Supply becomes ‘local’ – but not from locally owned firms. Apparently, other foreign firms in Mozambique also import most of their supplies (see discussion below) or use their own established supply chains, and I was warned against seeing this as ‘typically Chinese’:

As noted by Krause & Kaufman (Reference Krause and Kaufmann2011), the impressive growth that Mozambique has experienced since the mid-1990s can be attributed mainly to massive inflows of FDI in capital-intensive mega-projects with few linkages to the local economy. Even with Mozal, Mozambique's first mega-project in 1998 and the largest foreign investment at the time,Footnote ⁵ the broader development effects have been disappointing. The results of the tailored programme for developing economic linkages with domestic companies were modest (Krause & Kaufmann Reference Krause and Kaufmann2011), and most suppliers and subcontractors have still been South African (Castel-Branco Reference Castel-Branco2008). An assessment conducted by APRM (2009, in Krause & Kaufmann Reference Krause and Kaufmann2011: 52), established that Mozal remains ‘an isolated economic enclave that uses large quantities of scarce resources without returning revenue or jobs to the economy’.

Chinese businesses do have certain attitudes towards subcontracting and supplying that challenge the development of domestic backward linkages. These attitudes relate mainly to motivations for entering the Mozambican market (e.g. ExIm Bank funding), sources of competitiveness and organisational structure. However, it is not necessarily some ‘Chinese business model’ that is the main challenge to creating domestic backward linkages, as other foreign companies also use follow-suppliers in their operations and have been active in the local construction sector since long before China's ‘Going Out’ strategy. Ownership is only one of several aspects influencing the development of linkages.

Capabilities in the Mozambican construction sector

The creation of domestic backward linkages from foreign projects is heavily dependent on the local context. Local firms must be able to absorb spillover effects and face competition; viable production and supply capacity must exist; and adequately skilled workers must be provided (Kragelund Reference Kragelund2009). If the local economy is to reap the benefits of the culture of outsourcing in the construction industry, the necessary resources (such as labour, materials, machinery and technology) must meet high standards and be readily available (Maugeri et al. Reference Maugeri, Nhabinde and Quintella2015). That is generally not the case with the Mozambican construction sector. Many Chinese firms rely on supplies from China:

The extensive use of Chinese supply sources was most evident in the housing project and the road project. This is in line with Gu (Reference Gu2009: 582): ‘the absence of local networks of specialised supplies or, where they do exist, the high cost and often poor quality of goods means that many Chinese firms simply turn to the reliable, tried and tested, and cost-competitive established suppliers back in China’ (see also Yi-Chong Reference Yi-Chong2014). In the water project, the financer (AFD) had tried to maximise local supplying by prohibiting imports from China. Sourcing from South Africa was allowed, as this was considered a developing country: thus, most supplies were brought from there. Moreover, when iron pipes of the specified quality were unavailable regionally, the Chinese company was exempted from the initial prohibition, and imported these from China (Chinese head of engineers, water project, Int. 27.5.13). This shows that the mode of financing and requirements on sourcing can have an effect on sourcing patterns, although not always benefiting the local market in Mozambique. As for the Mozambican sector, the Chinese companies did not know which local companies they could use or trust as suppliers. Transportation was too slow; and when goods and services from local companies were tested, these often lacked the capacity and technical standards to meet demands as to quality, quantity and deadlines, making it too risky to depend on local companies (Chinese head of engineers, water project, Int. 27.5.13). The Chinese project coordinator regretted that they could not involve more domestic companies, but this had proved too difficult and time-consuming (road project, Int. 19.6.13).

In both the road project and the water project, the Chinese administration and consultants noted that they preferred using South African subcontractors (Chinese Head of Project and Coordinator, Int. 19.6.13; Consultant Company, Int. 19.6.13). Such competition from South African goods and services in the Mozambican market is nothing new, nor is it connected to Chinese companies. Since its economy started liberalising in the early 1990s, South Africa has provided goods and services that have generally been ‘cheaper, of better quality, enjoyed a better reputation and were delivered more reliably and with better customer services’ than Mozambican equivalents (Castel-Branco Reference Castel-Branco2014: 29).

The Mozambican construction sector is divided between large foreign firms, on the one hand, and domestic SMEs on the other. As put by a Mozambican NGO (Int. 11.11.14) regarding the construction industry: ‘In Mozambique, we have the international – the Portuguese, the Chinese companies – who participate in ‘Champions League’, and then we have the national contractors who are in the ‘Moçambola’ [national football division].’ In Maputo province, most construction companies and producers of construction materials remain small or micro-sized. They suffer from low credibility, lack of skilled workers, technology and certification, and limited sources of credit (Lopes Reference Lopes2007; Maugeri et al. Reference Maugeri, Nhabinde and Quintella2015).

The shortcomings of domestic companies are well understood by key stakeholders, who acknowledge the lack of appropriate skills, and access to capital and equipment. Also lacking is practical experience with modern construction methods and technologies, given the dominance of foreign companies in the industry (Business association for contractors, Int. 14.5.13). The lack of appropriate skills applies also to the workforce, with a shortage of professional workers at technical and administrative levels (Lopes Reference Lopes2007). This makes it challenging for the government to push for local content beyond low-skilled positions (Ministry of Labour, Int. 14.8.13; Ministry of Public Works, Int. 25.7.13). For instance, the Mozambican subcontractor in the housing project employed Paraguayan technical workers and engineers as supervisors for Mozambican local workers (Mozambican subcontractor, Int. 10.6.13).

Small and medium-sized domestic companies have been unable to meet the growing demand from the construction sector in the country, which in turn has resulted in lead firms filling the gaps with imports (Maugeri et al. Reference Maugeri, Nhabinde and Quintella2015). The business association for contractors (Int. 14.05.13) confirmed that many construction materials must be imported from South Africa or Portugal, because they are unavailable in the Mozambican market. A 2009 World Bank report confirms this: about 90% of the construction material for high-end construction must be imported. The difference between Chinese-led projects and others is not necessarily that more material is imported – only that more is imported from China.

Interestingly, whereas imports of iron- and steel-based products continue to increase, imports of cement are declining. Maugeri et al. (Reference Maugeri, Nhabinde and Quintella2015) explain this with the establishment of Cimento Nacional in 2011, a cement producer headquartered in Dubai, and an increase in the capacity of Cimento de Mocambique, a subsidiary of a Portuguese company that has dominated the Mozambican cement market since independence in 1975. Media reports also indicate that smaller private Chinese companies will seek to establish themselves as producers of construction supplies in Mozambique as a consequence of Chinese penetration of the Mozambican construction market.Footnote ⁶ This is backed up by studies of other African countries: for example Corkin (Reference Corkin2012: 478) found that Chinese companies and entrepreneurs in Angola were working in parallel with larger Chinese companies, providing ‘complementary action’ to projects – goods and services required from larger contractors working on Chinese-financed projects. Morris et al. (Reference Morris, Kaplinsky and Kaplan2012) report similar findings. Hence, while some have predicted that labour practices and market strategies of Chinese companies may alter as they establish themselves in Africa with long-term perspectives (Anshan Reference Anshan2007; Gu Reference Gu2009), the Chinese control of supply chains could indicate that, even if chains gradually become more locally embedded, that will not necessarily benefit domestic companies. However, as evident from the Mozambican construction sector, this is not something typically Chinese: it is more that Chinese firms, like other international actors, recognise the economic potential in filling gaps in the Mozambican sector for construction material.

Social infrastructure

The low capabilities in the domestic construction sector are to a large extent connected to weak social infrastructures. Supporting social infrastructure involves efficiency and cost to the administrative and regulatory regime (Morris et al. Reference Morris, Kaplinsky and Kaplan2012). Over 90% of construction companies in the Maputo region are in the informal sector – under the radar of formal rules and regulations, and without the accompanying societal infrastructure. Further, excessively tedious bureaucratic procedures related to state ownership of land in Mozambique make access to land difficult in urban areas, boosting the informal sector (Maugeri et al. Reference Maugeri, Nhabinde and Quintella2015).

One of the major aspects highlighted by my informants was the lack of demand (Business association for construction, Int. 14.5.13; Mozambican construction company, Int. 17.12.13). As construction projects in Mozambique have mushroomed, the ‘lack of demand’ is related to the fact that foreign companies have been the main beneficiaries of this development. Several informants also mentioned that domestic companies must depend on private contacts in order to get work (Mozambican construction company, Int. 17.12.13; Mozambican subcontractor, water project, Int. 30.5.13). According to Hanlon & Mosse (Reference Hanlon and Mosse2010), those companies that have successfully linked up with foreign companies have been few, large, and with close links to the Mozambican political elite. Hansen et al. (Reference Hansen, Buur, Therkildsen and Kjær2014) report that much of the market for domestic backward linkages in Mozambique, in the form of local content, has been captured by the economic and political elite. This is further supported by Castel-Branco (Reference Castel-Branco2014: 27), who holds that economic growth in Mozambique has come about through an alliance between the Mozambican state, large domestic capital and multinational capital. SMEs are considered insignificant to the elite, and not prioritised concerning linkage development (Buur Reference Buur2014).

A second aspect highlighted by my informants was the limited access to equipment and credit. Maugeri et al. (Reference Maugeri, Nhabinde and Quintella2015) report that access to finance has remained a major challenge, despite the growing number of operators in the financial sector. This affects the technical capacity of domestic firms: a 2010 survey of the Mozambican Industrial Association (Maugeri et al. Reference Maugeri, Nhabinde and Quintella2015) showed that more than 62% of domestic companies had not upgraded their technology significantly since the 1990s, and machinery was often more than 20 years old.

Understanding policy and policy space in Mozambique

As noted by Mold (Reference Mold2012: 239), ‘the relevant question is no longer whether governments should promote investment in infrastructure to underpin development but rather how the projects are selected and designed and how are they to be financed and executed’. Several interlinked political scales influence the disappointing outcomes as to domestic backward linkages. Policies aimed at maximising gains from foreign-company activity should seek:

The Mozambican government has focused on the first of these points – how best to attract investments – but not necessarily how to guide these investments or build domestic capacity to absorb the spillover effects. ‘The government follows an open policy regarding foreign trade that eases foreign investment and prioritizes free trade over the promotion of local manufacturing and processing’ (Krause & Kaufmann Reference Krause and Kaufmann2011: 10). Mozambique has adopted unusually strong legal commitments to open its economy to foreign investment. The 1992 Constitution gives foreigners the right to invest on the entire territory and in all economic sectors, except those reserved for public ownership (UNCTAD 2012: 37). This is reflected in economic management, where the prevalent ideology has been that the government should not interfere with business decisions (Castel-Branco Reference Castel-Branco2001; Castel-Branco et al. Reference Castel-Branco, Chichava and Francisco2011). Certain political and economic practices promoted by the government tend to reflect the interests of large multinational or domestic capital, and have mostly involved incentives and facilitation of capital accumulation (Castel-Branco et al. Reference Castel-Branco, Chichava and Francisco2011). Domestic companies face sharp competition, with scant policy support: ‘What is happening is that our legislation allows the entry of anyone … [And] on the other hand you don't have the support of the national bank … You are weak, you are weak’ (Mozambican construction company, Int. 17.12.13).

The Mozambican experience shows the trade-off between long- and short-term development goals, and the perception that these cannot be pursued mutually. The complex development goals of Mozambique's strategies are transformed into simple development outcomes due to impatience as to results of visible ‘development’. Efficient and cheap infrastructure is framed as ‘development’, whereas local content and capacity building are seen as obstacles to rapid delivery. When clients insist on quick delivery and low-cost projects, there are no legal constraints that force construction companies to show local social responsibility (Zawdie & Murray Reference Zawdie, Murray, Murray and Dainty2013). When key Mozambican stakeholders highlight speed, efficiency and competitive pricing as main elements explaining Chinese success in the construction industry (Mozambican project owner, road project, Int. 20.6.13; Ministry of Public Works, Int. 27.7.13; Ministry of Finance, Int. 29.8.13), they are seen as incompatible with long-term goals of knowledge spillovers and local capacity-building in the construction sector. Chinese contractors are largely left to develop projects and manage value chains as they see fit.

USAID (2013) holds that the dialogue on industrial policy has been unclear and uncoordinated in Mozambique, with little attention to local content or backward linkages.Footnote ⁷ There is a contradiction within the Poverty Reduction Strategy Paper: the construction industry is held to play a crucial role in the country's industrialisation process and in developing micro-, small- and medium-sized enterprises, but the specific measures actually reconceptualise the construction industry into an arena for promoting temporary low-cost jobs (GoM 2011). However, debate about industrial policies has started to emerge in Mozambique, particularly in the form of local content regulations.Footnote ⁸ In the construction industry, the business association for contractors cooperated with the Ministry of Public Works in formulating a draft local content policy that would guarantee that 20–30% of contracts awarded to foreign companies should go to local businesses (Business association for contractors, Int. 14.5.13). It was hoped that this would motivate the government to be bolder in negotiations with foreign companies, including Chinese:

This informant is convinced that the main challenge for the Mozambican construction industry results from weak policies surrounding the sector. The Chinese are good at taking advantage of the liberal policy environment, but cannot be blamed for the weak support of local industry. To a large extent, the informant feels that the state is ignoring the interests of domestic companies, while facilitating the entry of anything foreign. Maugeri et al. (Reference Maugeri, Nhabinde and Quintella2015: 34) identify similar problems, finding that the state lacks ‘financial, material, and human resources, which in turn make it unequipped for rapid changes in the industry landscape’. Governmental bodies lack information on which companies operate in the country, in turn making it difficult to design effective industrial policies (Schou & Cardoso Reference Schou and Cardoso2014). Moreover, the business association is considered as being distanced from decision-making centres, which ‘reduces the chances for fruitful intervention’ (Maugeri et al. Reference Maugeri, Nhabinde and Quintella2015: 35) – further challenged by strong networks of political and economic elites (Krause & Kaufmann Reference Krause and Kaufmann2011).

Additionally, while debate on local content is welcome, the discussion on domestic capabilities shows that regulations on local content have no effect if local capacity is non-existent. In Mozambique as elsewhere (see e.g. Kragelund Reference Kragelund2009 on Zambia), strong liberalisation of the economy without building up domestic productive capacity makes it extremely difficult for local producers to link up with foreign investors. Given the state of local capabilities, policies that require lead firms to give preference to local suppliers if they can compete on the basis of quality, price and availability will have no effect. According to Lopes (Reference Lopes2007), the government has been more concerned with costs in the construction sector than with the medium- and long-term consequences of lack of professional capacity. Although education has been a continued focus of donors and the Mozambican government, these efforts have mainly targeted primary and secondary school enrolment (Lopes Reference Lopes2007). The government has supported mid-level or higher education institutions that offer studies in business and management, but the weak investment in teachers has led some to doubt the competency of graduates (Maugeri et al. Reference Maugeri, Nhabinde and Quintella2015). The benefits of low-skilled positions in terms of economic development are limited in the long run:

The Mozambican policy environment must also be understood in relation to wider policy spaces. According to Keet (Reference Keet, Cheru and Obi2010: 24), during Africa's ‘development era’ in the 1960s and 1970s, foreign investors were commonly required to include ‘significant undertakings that would increase the “gains” or improve the effects of such investment in the host countries, and minimise or reduce the costs or negative effects’. This could include requirements as to reinvestment of profit within host economies, backward linkages, training/maintenance and technology transfers, and requirements for local employment and skills transfers. Domestic support to ‘infant industries’ and various external trade strategies and national tariff policies were introduced in parallel. However, the promotion of liberal policies and structural adjustment through IFIs and bilateral donors changed the conditions under which African countries could promote industrialisation, and displaced many of these development instruments. According to Morrissey (Reference Morrissey2012: 30–31), ‘the fact that organizations such as the World Bank (and donors more generally) did not advocate industrial policy diminished the importance of such policy issues on the domestic agenda. For sub-Saharan African countries, this is one of the reasons why they failed to develop the domestic capacity to benefit from FDI.’ For decades, international institutions and major donors have promoted neoliberal economic policies in Mozambique, and the state has allowed the ‘Washington Consensus’ to direct much economic and political reform (Alden et al. Reference Alden, Chichava, Roque, Alden and Chichava2014). Selective industrial policies have not been promoted, as economic development has been understood as ‘being basically about macroeconomic stability, liberalisation, privatisation, and lowering the costs of transactions through Doing-Business-style reforms’ (Krause & Kaufmann Reference Krause and Kaufmann2011: 26).

Membership in organisations like the World Trade Organization (WTO) has further restricted governments in favouring local materials, suppliers, contractors or consultants, even for generating local employment (Lewis Reference Lewis2007). Kragelund (Reference Kragelund, Cheru and Obi2010: 208) notes: ‘the agreements covered by the World Trade Organization (WTO) have made it difficult for African countries to pursue their own (protectionist) trade, industrial and investment policies’. In the construction sector, projects have increased in size, placing them beyond the reach of domestic companies and consultants that previously could benefit from minor government procurements in construction services (Lewis Reference Lewis2007). These factors are crucial for understanding China–Mozambique economic relations today. The ways in which Chinese companies are allowed to operate, how they have penetrated the Mozambican construction market and come to dominate value chains, must be recognised as the legacy of decades of liberalisation policies pushed by the IFIs and bilateral donors, rather than any Chinese business model as such (Kragelund Reference Kragelund2009). While scholars debate whether China's growing role as a development partner has opened or will open policy space for developing countries tied into conditionality frameworks with Western donors (Oya Reference Oya2006, Reference Oya2008, Cheru & Obi Reference Cheru and Obi2010, Harman & Brown Reference Harman and Brown2013), Chinese business-driven win-win strategies appear to benefit from decades of liberalisation, not challenge them.