Abstract
This article both extends the debate on the varieties of capitalism theory beyond the existing literature to solve the ambiguous position of the variety of capitalism that is found in Turkey and brings a novel approach to the studies of the political economy of Turkey by adopting a firm-centred position using the varieties of capitalism framework. Based on a qualitative comparison with the dependent market economies (DMEs), mixed market economies (MMEs) and hierarchical market economies (HMEs), this article claims that Turkey is a hierarchical market economy with four characteristic features that are also found in Latin American economies. These core features are the dominance of the family-owned diversified business groups, state-regimented and weak industrial relations, low skills and the influence of MNCs.
Introduction
Varieties of Capitalism theory has been developed by Hall and Soskice to categorise and describe developed capitalist market economies based on the juxtaposition of Liberal Market Economies (LMEs) which is characterised by arm's length interaction among market actors and the Coordinated Market Economies (CME) in which informal networks and collaborative arrangements accompany market relations. Taking a firm-centred position, Hall and Soskice argue that firms in LMEs or CMEs are operating in an environment conditioned by complementary institutional structures, which determine their corporate strategy for innovation or investment. In the original study, six of the OECD countries are categorised as LMEs (the United States, Ireland, the United Kingdom, Canada, Australia and New Zealand) while ten as CMEs (Japan, Germany, Denmark, Sweden, the Netherlands, Belgium, Switzerland, Austria, Finland and Norway) (Hall & Soskice, 2001, pp. 1–68).
Hall and Soskice left ‘France, Italy, Spain, Portugal, Greece and Turkey in more ambiguous positions’, and indicated that these countries ‘show some signs of institutional clustering which may be described as Mediterranean Market Economy’, featured by a large agrarian sector, widespread state intervention and liberalised labour relations (Hall & Soskice, 2001, p. 21). In the last decade, researchers have used Varieties of Capitalism framework to define distinctive forms of capitalism in Southern Europe, Latin America, Central and Eastern Europe and Asia. Ben Schneider has analysed Latin American countries with the Varieties of Capitalism approach and coined the term “Hierarchical Market Economies” (HMEs) for these countries given the patronage relationship between the state and businesspeople and among firms in these countries (Schneider, 2009). Nolke and Vliegenthart have enlarged the studies of Varieties of Capitalism in their work for the Central and Eastern European countries (which had also been studied by King in a separate study previously) and they have defined these countries as “Dependent Market Economies” (DMEs) since the Central and Eastern European countries are highly reliant on the foreign direct investments (FDIs) and transnational corporations (TNCs) (Nolke & Vliegenthart, 2009) (King, 2007). Molina and Rhodes investigated the market mechanisms and informal coordination structure of two Southern European countries, Italy and Spain. Focusing on the welfare regimes and production systems of these two countries they have arrived to the conclusion that these two countries can be defined as Mixed Market Economies in which state has an extensive role to compensate the deficiencies of the institutional structure in these countries (Molina & Rhodes, 2008). Though Molina and Rhodes have abstained to embrace the term state-led capitalism for Italy and Spain (2008, p. 227), Schmidt claimed that she brought the role of the state into the Varieties of Capitalism framework by defining Italy, France and Spain as State-Influenced Market Economies (SMEs) given the distinctive influence of the state over labour and business (Schmidt, 2007, p. 4). Witt and Redding have compared 13 major Asian countries with one another and Germany, the UK, the US, France and Sweden and have concluded that the Varieties of Capitalism framework is not applicable to Asia given the heterogeneity and differing characteristics of business systems in different Asian countries (Witt & Redding, 2013).
This paper has multiple goals. Firstly, it aims to give a new impulse to the varieties of capitalism theory by extending the debate beyond the existing literature to clarify the vague presumptions regarding the variety of capitalism that is found in Turkey. Secondly, it intends to bring a novel approach to the studies of the political economy of Turkey by adopting a firm-centred approach and by analysing the institutional complementarities using the varieties of capitalism framework in order to fill the gaps that are left by the previous studies. Thirdly, its objective is to form a basis for a comprehensive overview of the institutional settings of Turkey in a comparative perspective.
Relying largely on the qualitative comparison with the DMEs, MMEs and HMEs, the fundamental claim of this paper is that Turkey is a hierarchical market economy. There are four core features of the variety of capitalism in Turkey that put it into the category of the Hierarchical Market Economy as Latin American countries. The dominance of the family-owned diversified business groups in the economy, state-regimented and weak industrial relations, low skills and the influence of the Multi-National Corporations. Similar to the grupo economicos in the Latin American economies, family conglomerates called ‘holding’ constitute the dominant form of corporate governance contrary to the DMEs, where MNCs are dominant. 28 out of the 50 largest companies in Turkey are the family-owned diversified business groups while the remaining 22 companies are either the subsidiaries of MNCs or the state-owned enterprises (Colpan, 2010). Trade unions are nearly ineffective in Turkey in contrast to the MMEs, the collective bargaining mechanism is absent, the labour union density is as low as 7% with the contribution of the large informal sector (Celik & Lordoglu, 2006). Though the investments in education and vocational training by the government and the family conglomerates has gained a momentum since the last decade, Turkey's human capital levels are below the average education levels than in DMEs and MMEs (Bartlett, 2013).
In the next section, the core features of the original study of Hall and Soskice are demonstrated briefly, followed by an analysis of the three varieties of capitalism, namely, DMEs, MMEs and HMEs, which are the subjects of the comparison in this paper. Then, in parallel with the original framework of the varieties of capitalism, five fundamental and interdependent institutions of capitalism in Turkey are compared to those in DMEs, HMEs and MMEs to reveal institutional complementarities between these elements.
Varieties of capitalism: From a dichotomy to the variety
The comprehensive approach of the Varieties of Capitalism, taking the legal, political and economic aspects, norms and institutions into its consideration reflects its novel emphasis on the institutional complementarities (Hall & Soskice, 2001). Economic performances of the developed capitalist economies are products of complementary institutional settings. Cooperative wage-bargaining process in the Coordinated Market Economies such as Germany, for example, is connected to the industry-specific training and the social security regime. Bank-based finance of the firms and the corporate governance regimes are complemented by the investment strategies and the higher wages as a result of strong labour unions which enforce companies to be competitive in diversified quality production. In the United States, as a Liberal Market Economy, weak labour unions and flexible labour markets are complemented by a robust tertiary education which provides general skills rather than industry-specific training. Shareholder-based corporate governance and the stock markets-based financial system allow firms to take greater risks with new technologies or strategies which are supported by the flexibility in labour contracts, which in the end, increases the firms’ competitiveness in LMEs in radically innovative sectors.
Founders of the Varieties of Capitalism have revealed that each type of market economy restraints the firms to target an innovative production strategy that can fit the institutional settings of the type of economy in which the firms operate. In the Liberal Market Economies, such as the United States or the United Kingdom, firms are tended to be successful in the radically innovative sectors such as pharmaceuticals or telecommunication since the weak trade union, flexible labour markets with general skills and the shareholder-based financial system allow the companies in the LMEs to make rapid decisions or take risks as well as hire and fire employees easily. In the Coordinated Market Economies, such as Germany or Japan, rigidity in the labour markets, which are also supported by strong trade unions, and the long-term interaction among workers and employers under the protection of the business law-based on a stakeholder-oriented model make the companies operating in these economies to be successful in the sectors associated with incremental innovation strategies, such as automotive and machinery (Hall & Soskice, 2001).
Departing from this understanding, Nolke and Vliegenthart argue that, Dependent Market Economies such as Czech Republic, Hungary, Poland and Slovakia have been hosting MNCs in the industries of complex and durable products due to their cheap and skilled labour base. Local subsidiaries of the MNCs are not financed through stock markets as in LMEs or domestic banks as in CMEs but instead, they are controlled and financed by the headquarters of the MNCs from abroad. Since DMEs attract FDI owing to their cheap labour base, low labour costs are used as a bargaining chip by the MNCs against labour organisations. MNCs’ need for low labour costs can also be observed in the education system of DMEs. On the contrary of LMEs (general-skill based education in combination with substantive research and development investments) or CMEs (high-quality vocational training as a result of the industrial coordination), in DME technologies are invented and transferred from the headquarters and therefore MNCs do not need to invest in human capital in the host countries. Furthermore, the low coordination between firms and state make it difficult to construct an effective vocational training system nation-wide. As a result of these features, DMEs have a comparative advantage based on the assembly of semi-standardized goods. (Nolke & Vliegenthart, 2009, pp. 672–679)
Studies of Varieties of Capitalism have been extended to the Southern Europe as well. In the Mixed Market Economies, labour unions are stronger than those in LMEs and can be considered as veto players but are not as organised as those in CMEs to facilitate coordination in collective wage bargaining. Reform process requires political leadership by the political actors in order to not end with a stalemate since the formation of coalitions is more complicated than in LMEs and CMEs (Featherstone, 2008, pp. 13–14). Labour unions, business associations and market actors resemble CMEs. However, the fragmentation between public and private sectors weakens the coordination capacity of the actors and interest groups use their power to lobby political actors for patronage or safekeeping (Hassel, 2014, p. 7). Since the weakness in the coordination capacity must be compensated by the state intervention, market actors in the MMEs have strong incentives to strive for political power through patronage relations with political actors to be granted for subsidies and support (Molina & Rhodes, 2008, pp. 227–228). As Hassel argues, ‘institutional stability in MMEs is not based on complementarities but state intervention’(2014, p. 9).
Schneider's study shows that Hierarchical Market Economies in Latin America (Argentina, Chile, Brazil, Colombia, Mexico) have four prominent features: influence of multinational corporations from manufacturing to finance and other services, low-skilled labour as a result of the low investment for human capital, dominance of the family-owned conglomerates and weak and small labour unions which is a consequence of large informal sector (2009). The distinctive character of HMEs comes from the non-market and hierarchical interaction between the business groups, MNCs and the other actors of the economy. Vocational training is not market-based as in LMEs or based on cooperation between state and corporates as in CMEs, but instead dependent on the venture of domestic firms in Latin America. Low rates of unionisation and temporariness of the labour contracts undermine the bargaining power of the labour unions and hierarchy also prevails in the relationship between employees and employers. In such a hierarchical network, the role of the government is to impose top-down regulations to control industrial relations (Schneider, 2009, pp. 556–557).
In fact, weak labour unions, the need to attract FDI, an important role for MNCs, political patronage and low investment in education or innovation are common features of almost all emerging market economies, including Turkey (Haggard & Kaufman, 1997, pp. 318–324). However, nuances in the relative weight of each component of these common features give HMEs, DMEs or MMEs their distinct character. Considering the fact that an essential aspect of the Varieties of Capitalism theory is its emphasis on the role of the institutional complementarities covering five main policy and practise areas, namely, industrial relations, vocational training, corporate governance, inter-firm relations and innovation; in the following section, Turkey's institutional structure is evaluated under these five categories in comparison with the HMEs of Latin America, MMEs of Southern Europe and DMEs of Central and Eastern Europe in order to reveal these nuances and to discover the institutional complementarities in Turkey.
Turkey as a hierarchical market economy
There have been four important developments, according to Bugra and Savaskan, in Turkey's economy during the last two decades. The first one is the loyalty of the state to the economic liberalism which helped to attract the foreign capital, the second development is the massive amounts of privatisation which fed the growth of the diversified business groups, the third is the increasing state support for the Small and Medium-Sized Enterprises through KOSGEB, the fourth is the increase in the export as a result of the innovation investments which boosted the competitiveness of the firms in Turkey (Bugra & Savaskan, 2015). These four significant developments are totally consistent with the core features of HMEs. As Schneider mentioned, the state is the main external agent that strengthen the characteristic features of HMEs by inviting MNCs to the economy and through intentionally contributing to the expansion of the family conglomerates, while also promoting vocational training and education to foster innovation. (Schneider, 2009, p. 570). In this section, five institutional components of Turkey are examined to demonstrate institutional complementarities in comparative perspective with DMEs of the Eastern and Central Europe, MMEs of the Southern Europe and HMEs of the Latin America.
Corporate governance and finance
Foreign direct investment is the main source of investment in DMEs. Nolke and Vliegenthart have measured inward and outward FDI stock and analysed financial sources of the leading export sectors in DMEs and demonstrated that foreign ownership dominates both in manufacturing and banking (with 70% of foreign ownership ratio). On the contrary, while MNCs’ presence in Latin America and Turkey is also significant, family conglomerates are more prominent than MNCs in these economies and the primary source of finance is intra-group accumulations instead of the FDIs (Colpan & Jones, 2016; Schneider, 2009, p. 560). In general, finance of the investments in DMEs is mostly provided by foreign European banks (King, 2007, p. 5). Decisions are made between managers of the subsidiary firm in the host country and the directors of the headquarters located abroad. Even small and medium-sized enterprises are dependent on the decisions of the foreign managers since they are dependent on the foreign partners for their exports (Nolke & Vliegenthart, 2009, pp. 679–82). Thus, MNCs have an indicative role in the growth performance of the economy in these countries. Since the corporate governance regulations have also been shaped by the European Union rules during the accession process, one can observe that not only the corporate governance practises but also the regulatory infrastructure of DMEs are determined by the external actors (Nolke & Vliegenthart, 2009, p. 683). In the Mixed Market Economies, a capital coalition was created between firms operating in manufacturing and services and the domestic banks to accelerate the financial deregulation process in the 1990s (Molina & Rhodes, 2008, p. 25). However, small and medium scale enterprises hardly access to the financial markets compared to the large firms and for this reason, they are financed via domestic banks, as in HMEs, including Turkey (Molina & Rhodes, 2008, p. 30).
Though MNCs’ presence in Latin America and Turkey is also significant, family conglomerates are more prominent than MNCs in these economies and the primary source of finance is intra-group accumulations instead of the FDIs (Colpan & Jones, 2016; Schneider, 2009, p. 560). Latin American HMEs are dominated by large domestic conglomerates and MNCs. These conglomerates have diversified sub-groups operating in at least four or five different sectors with dozens of firms, connected to the conglomerates with a hierarchical relationship. Where only one-third of the largest 500 companies in the US (an example of LMEs) and about 10% of the companies in the CMEs are family-controlled, in the HMEs conglomerates are family-owned, and a significant share of the economic activity is concentrated in the hands of a small number of conglomerates. The dominance of the huge conglomerates makes the playground for their competitors uneven to the behalf of the former group (Schneider, 2009, pp. 558–9) (Schneider, 2008, p. 383). Economic and political instability and inadequate property rights prevent stock markets from developing adequately, which in turn consolidate the family-ownership model in Latin America. Though liberalisation efforts during the 1990s helped stock markets to grow from 8% in 1990 to 34% in 2003 of GDP, stock markets have remained small in Latin America, compared to the other developing regions such as Southeast Asia and for this reason, financing of the investment are usually made by the savings of the companies, or through intragroup debt in conglomerates (Schneider, 2008, pp. 385–387).
In Turkey, as in Latin America, diversified large business groups are the core elements of the economy, and they account for 28 of the 50 largest companies in Turkey (Colpan & Jones, 2016, p. 70). Similar to the ratio in other HMEs, one-fourth of the largest 500 companies in Turkey are of MNCs, the rest dominantly belongs to the family-owned conglomerates while the significance of the state-owned enterprises still continues, according to the 2015 survey of Istanbul Chamber of Industry (ICI, 2016; Schneider, 2009, p. 560). Furthermore, in terms of generating employment, foreign companies constitute only 8% while domestic diversified business groups account for 57% with regard to the numbers of employees of the 50 largest companies (Colpan, 2010, p. 495). These family-owned large conglomerates operate in a broad range of areas stretching from manufacturing to finance and services (Gökşen & Üsdiken, 2001, p. 338). For example, although established as a pharmaceutical company in 1952, the Eczacibasi Group has evolved into paper products, investment banking, cosmetics and IT services instead of concentrating the R&D investments on pharmaceuticals (Colpan, 2010, pp. 498–499). Founding family members manage the entire business groups through conglomerates with a hierarchical and complex inter-firm shareholding structure (Colpan, 2010, p. 511). In order to sustain the family control over the subsidiaries while the scope of the business group activities is expanding, Koç family, for example, established the first holding company of Turkey in 1963 to centralise the decision-making mechanism (Bugra, 1994, p. 79).
Diversified business groups in Turkey can be divided into two categories: those which emerged between the 1920s and 1950s such as Koç, Sabancı, Eczacıbaşı, Çukurova, Borusan, Yaşar and those established after 1980s, for example, Zorlu, Sanko, Ciner, and Fiba. The former group invested heavily in large-scale manufacturing sector during the Import Substitution Industrialisation period in 1960s and 1970s and crowded-out the latter group from the manufacturing sector. Due to the high entry barriers in manufacturing sector, the latter group have only been able to compete with the earlier group in the financial sectors and services. After the market liberalisation of the 1980s, both the former and latter groups have followed a similar expansion pattern through internal investment, partnership with MNCs for licensing or joint venture and purchasing privatised state-owned businesses (Colpan, 2010, pp. 501–504). These groups mainly rely on the accumulated earnings of their individual group companies for investments, as well as their own banks, which is a feature also observed in HMEs and other emerging market economies (Colpan, 2010, p. 522; Colpan & Jones, 2016, p. 80). Their ability to establish political connections and their knowledge of the domestic specificities comprise a comparative advantage for these groups against MNCs (Khanna & Rivkin, 2001, pp. 50–51; Schneider, 2008, pp. 385–387). The requirement for a balanced representation of the shareholders, especially after the stock market operations gained importance, led the grupos in HMEs such as in Chile and Brazil and holdings in Turkey to institutionalise the professional management system by hiring managers from out of the family since the 1990s, however, those heirs are still sent abroad to get management degrees to continue to intervene daily decision-making through executive boards (Colpan, 2010, p. 515; Schneider, 2008, p. 384). Boards are inadequate to represent the rights of minority shareholders as the family members are heavily involved in the board nomination and appointment process in many cases (Oba, Ozsoy, & Atakan, 2010, p. 609). As a result, CEO performance in Turkey is not influenced by stock market achievements but evaluated based on income before tax statistics (Oba et al., 2010, p. 613).
Industrial relations
Though the labour unions are weak, and the collective bargaining mechanism is absent in both DMEs and HMEs as well as MMEs, the causes are different. In DMEs, trade unions are not incorporated into the bargaining process and given the competitive pressures caused by the Eastern Asian economies where the cheap labour is abundant, they do not have strong positions for wage bargaining. However, even though high wages, high union density or strong labour unions would not be attractive for the FDIs, MNCs have to keep their workers satisfied once they get involved in the costly greenfield investment since they cannot easily find skilled labour. Instead of the sectoral or national level agreements, in DMEs wage bargaining is made at the firm level, which is similar to the Southern European MMEs (Nolke & Vliegenthart, 2009, pp. 685–686) (King, 2007, p. 4).
Collective bargaining systems in Mixed Market Economies are configured much more efficiently than in Turkey (Celik & Lordoglu, 2006, pp. 18–19). However, coordination among labour unions and employer organisations are not as strong as those in CMEs. For this reason, in MMEs, collective bargaining is not sufficient to produce the desired outcome but stronger than in HMEs and DMEs since it works to give veto power to the actors for change or request compensation from the state (Hassel, 2014, p. 7; Nolke & Vliegenthart, 2009, p. 677). In the absence of coordination, Italy, Spain and Portugal relied on social pacts on the road to the Economic and Monetary Union during the 1990s to lower wages to satisfy convergence criteria (Hassel, 2014, p. 27). Disparities in unionisation rates increase the conflict between public and private sector unions and therefore raison d'etre of these unions in MMEs is to achieve political control (Hassel, 2014, p. 11). While conflict among trade unions is also a prominent feature of industrial relations in Turkey, the underlying reason of the conflict is rather ideological (Buğra, 2002).
In Latin American HMEs, as in Turkey, trade unions are weak and fragmented because of the high turnover rates, large informal sector and low coordination among industries. Though labour regulations in HMEs are higher than those in DMEs, the application of these regulations is limited in practise, not only because of the huge informal sector but also because of the low consent in the formal sector, leaving employees in a vulnerable position. On the contrary of the long job tenures that can be observed in DMEs, median job tenure in HMEs is three years. Another factor about the labour unions in HMEs is that state's control and influence head the labour unions off at the pass. Limitations on the bargaining conditions by laws and obstacles for the collective bargaining leads trade union leaders to seek political influence via establishing relations with political actors (Schneider, 2009, pp. 562–563).
In comparison, industrial relations in Turkey have been designed as a part of the state apparatus for the economic governance rather than a means of collective representation for employees or employers (Morley, Gunnigle, & Collings, 2006). As part of the institutionalisation process and in parallel to the requirements of the import substitution industrialisation, trade unions soared after the 1961 Constitution which paved the way for the establishment of trade unions and accorded the right to engage in collective bargaining with the enactment of a union law in 1963 as a top-down project (Keyder, 1990, pp. 122–123). Though the Confederation of Turkish Trade Unions (TURK-IS) was organised in 1952 under the first trade union law of 1947, the law was forbidding the basic union rights, and TURK-IS was rather a pro-government and accommodationist trade union centre. In 1967, Confederation of Progressive Trade Unions of Turkey (DISK) was established and took a strong militant position. In 1970, Confederation of Nationalist Workers’ Unions (MISK) was founded as a reaction to DISK. Turkish Real Trade Unions (HAK-IS) was founded in 1976 in parallel to the rise of the political Islam. In spite of the fact that the military coup of 1971 banned the union activities until the democratic elections in 1973, the number of the union members have increased during 1970s rapidly, being 42% in terms of trade union density until 1980 (Visser, 2015). However, even at their peak, labour unions could not have a strong influence in the policy making or wage bargaining (Keyder, 1990, p. 123).
The consequence of the military coup of 1980 in the political arena has been a synthesis of Turkish nationalism and Islamism as the leftist groups and organisations were suppressed (Bugra & Savaskan, 2015). There were two immediate effects of the 1980 Coup which went hand in hand: firstly, the power of the trade unions repressed severely and secondly liberalisation of the Turkish economy based on Washington Consensus (Altug, Filiztekin, & Pamuk, 2008, p. 28). As a consequence of these two effects, the use of subcontracted labour without any job security has spread from agriculture to manufacturing and services (Bugra & Savaskan, 2015). Privatisation processes have gathered momentum after the 2001 crisis. Income of the state from the privatisation has been four times higher between 2004 and 2011 than the years between 1984 and 2004. Nevertheless, as a result of the increasing flexibility requirements of the private sector for the competitiveness purposes, as part of the growth strategy which has been mentioned in the National Employment Strategy document, subcontracting and hiring temporary workers without any job security has been institutionalised with the Labour Act no. 4857 (Sayin et al., 2014).
In short, the state in Turkey, as in the other HMEs, is the main actor that continues to act as the regulator of the capital, an inviter of MNCs, an intermediary for labour and initiator of vocational training and technology. Schneider (2009) argues that it is this strong presence of the state that might have prevented the emergence of lifetime employment or strong industrial relations in HMEs (2009, p.19). Hierarchical structure has been the most characteristic feature of the trade unions in Turkey according to Dinler (2012). Her study reveals that while the power and authority are concentrated in the male-dominated management groups in trade unions, these groups are established through clientelistic networks benefiting from high salaries (2012, p. 2). Hierarchy in the labour unions also deters members to take active roles. Furthermore, ideological and religious cleavages between the trade unions undermined the coordination and further weakened the labour movements in Turkey. State's substantial influence in industrial relations encourages leaders of the trade unions to seek political patronage, similar to the HMEs in Latin America. Informal employment is around 35% in Turkey, and union density levels are also as low as 7%, which makes labour unions even more atomistic than in Latin American countries (Celik & Lordoglu, 2006; TUIK, 2014).
Inter-firm relations
In MMEs, the state still continues to protect national firms from the foreign competition despite the liberalisation in recent years with the anchor of the European Union (Molina & Rhodes, 2008, p. 227). Firms in these economies rely on state intervention for compensation in face of economic uneasiness so that the loss of competitiveness and rising unemployment during economic shocks are compensated by an increase in private or public sector wages, which raise the amount of social spending (Hassel, 2014, pp. 10–13). The existence of the high number of state-owned business sector contributes to the public–private schism. Over-regulation for the large companies and under-regulation for the small and medium enterprises increase the tension between the small and large firms and among different sectors. The deep schism between the business actors in MMEs impedes coordination and prevents consensus-based bottom-up guidance. Thus state's intervention for top-down conflict governance becomes mandatory (Molina & Rhodes, 2008, pp. 240–245).
In HMEs, MNCs are operating in high-technology manufacturing industries while domestic conglomerates invest in natural resources and services since these areas do not require high skills (Schneider, 2009, p. 565). In Turkey, for example, business groups invest mostly into financial services, energy and mining industries (Colpan, 2010, p. 24). Political and economic domestic instability impede cooperation between MNCs and domestic business groups for long-term investments (Schneider, 2009, p. 566). The dominance of the MNCs in DMEs is also shaping the inter-firm relations in Central and Eastern Europe. For example, according to the study of King, more than 70% of the shareholders in the manufacturing sector in Hungary are of foreign origin, significantly higher than the foreign shareholders in Turkey in the same sector, which is less than 20% (2007, p. 7). For this reason, the interaction between the customers of the export market in the European Union and the subsidiaries of the MNCs is the striking feature of the inter-firm relations in DMEs (King, 2007, p. 8).
Despite the membership to the chambers of commerce was mandatory in Turkey, and the formation of such interest groups was supported by the state, the interaction between the large companies and the state was conducted by political patronage outside of the formal institutions. Though the formation of the voluntary associations of the business entrepreneurs such as Turkish Industrialists’ and Businessmen's Association (TUSIAD) after the 1970s has made the business groups less dependent on the government support and more integrated into the global system, the influence of the state on the market actors have continued (Bugra & Savaskan, 2015). The polarisation among the market actors has been a reflection of the political cleavages in Turkey. The interaction between the government and certain groups of businessmen has deepened the competition in the economy. For example, as a counterweight to TUSIAD, Independent Industrialists and Businessmen Association (MUSIAD) was established by the religious capital holders, which is followed by the formation of other religiously conservative businessmen associations like Anatolian Lions Businessmen Association (ASKON). One can observe the importance of gaining political support for the business associations in the interaction between MUSIAD and AKP (Turkish abbreviation for the Justice and Development Party) government. While the devout bourgeoisie actively supported the foundation of AKP, AKP in turn, provided convenience for those groups once came into power (Gümüşçü & Sert, 2009).
Non-market coordination through such associations is an important feature in Turkey as a hierarchical market economy, where the aim of the cooperation of the firms is to utilise financial and human resources, as well as the information resources (Colpan, 2010, p. 497). Patterns of the formal and informal coordination in Turkey can be tracked by an examination of the member lists of the business associations since the member firms are clustered in the business associations based on the ideological and religious identities of their owners. Given the strong influence of the state on the markets in Turkey, an increasing number of businessmen take roles in politics and the interdependence between the governments and market actors continues despite the liberalisation after the 1980s, in a similar way that has been observed not only in HMEs but also in MMEs.
Vocational training and education
The education system and the vocational training in DMEs are designed specifically according to the requirements of the MNCs while in Turkey it is influenced by the requirements of the diversified business groups (Colpan & Jones, 2016, p. 82; King, 2007, pp. 4–5). Because of the unwillingness of the MNCs in DMEs and family conglomerates in Turkey to seriously invest in human capital as a result of the employee poaching concerns, vocational training has been taking place in public institutions, in general, instead of the on-the-job training as could be observed in CMEs (Colpan & Jones, 2016, p. 83; Nolke & Vliegenthart, 2009, p. 686). Compared to the DMEs, domestic firms take more initiative in Turkey regarding investment in human capital, as will be mentioned later. Since an upgrade of the skill base in DMEs would trigger wage increases, instead of heavy investment in human capital, the main target of the education system is not to increase innovation capabilities but to produce adequately skilled labour for the assembly plants. On the contrary of the nationally owned businesses that are dominant in Latin America or Turkey, MNCs in DMEs are not concerned with the development of the human capital in the long run (Nolke & Vliegenthart, 2009, pp. 686–687). Similar to the that of in DMEs, MNCs’ reluctance for investment in human capital of Turkey was already confirmed in the study of Ozyigit and Eminer (2011) in which they have not found any long-term effect of the MNCs or FDI for increasing human capital in Turkey (2011, p. 565). In MMEs, the lack of coordination and high industrial conflict prevents firms to invest effectively in strike-prone human capital (Molina & Rhodes, 2008, p. 28) while in Hierarchical Market Economies, firms are generally abstaining to invest in training employees in the climate of high turnover rates since rival firms may be able to dissuade employees from the investor firm and since politico-economic instability constitute an impediment to the firms to invest in vocational training (Schneider, 2009, pp. 564–670). Attainment of the secondary education and vocational training is low, and state's investment in human capital is lagging behind the developed countries in HMEs, as in Turkey.
According to the survey of the of the Ministry of the National Education published in 2015 in Turkey, 3955 of 9061 secondary schools provide general education while 5106 provide vocational and technical education. Among these 5106 secondary schools providing vocational and technical education, 1149 are Imam Hatips, which trains Imams and Preachers (Presidency of Strategy Development, 2015, p. 128). In 2007, in the European Human Capital Index Ranking for the Central and Eastern Europe, Turkey was shown in the second position after Slovenia for its demographic prospects (Ederer, Schuller, & Willms, 2007, p. 2). It was also specified that without the consideration of the demographic aspect, Turkey would rank on the 9th position, on the top of Poland, Bulgaria and Croatia. Particularly, in terms of human capital endowment according to the report, Turkey was considered as a laggard, because of the low public expenditures in education and poor performance on the international math and science tests scores comparisons (Ederer et al., 2007, p. 8).
Findings of Education at a Glance report of OECD that was published in 2014 were also similar. First, despite the increases in the expenditure on education over the last decade, which currently amounts to the 4% of its GDP, education spending in Turkey is still lagging behind of the OECD average, which is 6%. Second, participation in the tertiary education has been increasing since 2000 with a significantly high growth rate of 5.2% compared to the 3.2% of OECD average. Third, women attending in universities have three times higher employment rates compared to those without tertiary education (OECD, 2014a). However, the findings of the OECD report must be carefully analysed. For example, though the quantity of the universities and participation to the tertiary education have increased over the last two decades in Turkey, the quality has remained poor because of the inadequate planning (Sallan-Gül & Gül, 2014). In spite of its young workforce, Turkey has a largely unskilled labour base due to the low government spending on education and the low upper secondary graduation rates (Owings, Kaplan, & Pirim, 2012). The school-industry mismatch, which is a consequence of the poor education quality, prevents secondary school and tertiary graduates from finding an appropriate job. On the contrary of the DMEs, where about 10% of the total unemployment consist of tertiary graduates, university graduates in Turkey and many HMEs and MMEs, such as Mexico; Chile; Greece; Spain and Portugal, have serious unemployment problems varying between 16% and 22% of total unemployment (The World Bank, 2016).
Studies about the quality of vocational education and skill mismatches in Turkey are limited. Bartlett's (2013) research on skill mismatches in the European Neighborhood Countries and Turkey is one of the remarkable works that would reveal some key findings concerning the human capital in Turkey. Using the EBRD and World Bank Survey of BEEPs he demonstrated that firms’ expectation for skilled labour is not satisfied by workers with general or vocational secondary education, while the skills of the graduates of tertiary education and those with primary level education are well matched to the industrial or sectoral expectations (Bartlett, 2013, pp. 21–22). Furthermore, the need for the skilled labour is 5 or 6 times higher than the need for unskilled labour in Turkey (Vivarelli, Srour, & Taymaz, 2013, p. 21). In another study, Hirshleifer et al. (2014) analysed the impact of the vocational training programs for the unemployed people provided by Turkish National Employment Agency (ISKUR) as part of the active labour market program for three years. Their study revealed that those who attended the vocational training programs experienced 2% higher chances of employment, which is statistically insignificant (Hirshleifer et al., 2014, p. 2). The reason for this inefficiency is both due to the low education quality of the programs and poor design of the curriculum which certified the skills that the participants already had (Hirshleifer et al., 2014, pp. 26–28). In one of the most recent studies on the human capital of Turkey, Yilmaz and Saracoglu have compared Turkey's labour productivity rates to the middle-income countries including Latin American countries such as Brazil, Mexico and Chile (2016). They have found that Turkey's labour productivity growth rate (2.69) surpasses the average labour productivity growth rate of middle-income countries in recent years (1.69) (Yilmaz & Saracoglu, 2016, p. 396).
The current shape of the education and vocational training system in Turkey, therefore, is quite related to the further expansion of the family conglomerates in Turkey which focused on investing in its human capital. State's attempts to enhance vocational training through promotive programs are complementary to the requirements of the domestic firms. Compared to the HMEs of Latin America, leading family conglomerates in Turkey are more prone to invest in the training programs and education of their employees in the recent years (Kim, Kandemir, & Cavusgil, 2004, p. 17). The survey of Colpan (2010) reveals that diversified business groups show the quality and the qualifications of the people they employed a reason for the strengths and capabilities of their groups. Firms in the innovative sectors in Turkey are training their employees up to 40 hours/year as they opened up to the world since the beginning of the 2000s, which has resulted in an increase in the innovation capabilities of these firms as will be seen later in this paper (Colpan, Bonaglia, & Goldstein, 2008, p. 12).
Innovation
While in Turkey research and development activities are dominated by the domestic family conglomerates, in DMEs, innovation activities are conducted at the headquarters of the MNCs and then transferred to the subsidiaries of these MNCs in Central and Eastern Europe with the strict confidentiality (Colpan & Jones, 2016; Nolke & Vliegenthart, 2009, p. 688). As a result, instead of creative innovations, in DMEs, imitation of the established technologies became prominent (Nolke & Vliegenthart, 2009, p. 690). In HMEs, as in DMEs, multinational corporations do not have an incentive to invest in human capital and research and development. While research and development expenditures in average equal to the 0.5% of GDP in HMEs, the biggest investors to the research and development activities in these countries are states while MNCs’ contribution to the expenditures is not significant (Schneider, 2009, p. 567).
In Turkey, contrary to many HMEs (except Mexico and Brazil) and DMEs, business groups’ investment expenditures to the research and development activities have increased since 2001 significantly and doubled the government's expenditures in 2014 (Turkish Statistical Institute, 2016). However, though remaining below the OECD average, state's contribution to the research and development activities has also been gaining importance since 2004, as part of the ‘Vizyon 2023’ project (TUBITAK, 2004). This state-led project resembles the German government's strategy to create radically innovative technology clusters. In the late 1990s, German government provided funds for the start-ups in the innovative sectors and created Neuer Markt, a stock market for the venture capital in 1997, to foster radical innovation (Lange, 2009, p. 198). In the same fashion, in order to facilitate innovation activities, state institutions have started to provide venture capital schemes for the start-ups and small and medium scale enterprises in the innovative sectors. For example KOSGEB (an institution of the Ministry of Science, Industry and Technology) offers financial support for the small and medium scale enterprises for their R&D activities and provides Capital Markets Support Program which subsidises the listing costs on the stock market (OECD, 2014b, p. 97). While the budget of KOSGEB was TRY 11 million in 2010, it has reached to TRY 349,809,242 until 2015 and tripled in 2016 to TRY 991 million in 2016 (KOSGEB, 2015).
The impacts of the investments in human capital on the innovation in Turkey can also be verified by the PISA scores of Turkish children. In a well-known analysis of the impacts of educational quality and quantity on the long-term growth, Barro shows that science and mathematics tests scores have a strong positive correlation with the long-term growth rates (Barro, 2001). PISA study reveals that, whereas mathematics score of Turkish children was lowest among the OECD nations in 2003, following Mexico (Altug et al., 2008, p. 30), mathematics test results of Turkish children were higher in the PISA scores of 2008 (OECD, 2014c). In fact, one can observe that both HMEs and DMEs as well as MMEs are clustered in ‘below-average mathematics performance’ and ‘below-average equity in education outcomes’ categories together with Turkey, though Turkey is positioned slightly towards ‘above-average equity in education outcomes’ category, which is a surprising result given the regional and gender-based inequalities in Turkey (OECD, 2014c, p. 196).
There is a crucial point that needs attention concerning the innovation targets of Turkey. According to the Strategy Paper for 2003–2023, Turkey have focused on eight new technology areas for innovation investments in order to gain international competitiveness by 2023: Information and Telecommunication, Biotechnology and Genetics, Nanotechnology, Energy and Environment, Mechatronics, Manufacturing Process, Material and Design (TUBITAK, 2004, p. 12). These sectors are associated with the radical innovation, which is supported by the institutional settings of the LMEs, and would not be compatible to an HME such as Turkey 1 . Though Hall and Soskice explain the development of the radically innovative sectors in the US as a result of the market dynamics, there are evidence demonstrating the role of the government support in the formation of the Silicon Valley (Lazonick, 2010, pp. 676–677). Leading universities in Turkey have started to establish business incubator centres with the support of the government funds imitating the Silicon Valley model in Turkey in order to form a cluster of high technology. SUCOOL of Sabanci University, Cyberpark of Bilkent University, and Teknokent of METU are among incubator centres in Turkey that bring start-ups and business networks together. These developments may signal a transformation into an LME for Turkey, as Schneider has also been suspected for HMEs in Latin America in the conclusion of his study (2009).
For further information regarding the characteristics of innovation in different varieties of capitalism, see Hall & Soskice (2001, pp. 1–70).
On the other hand, as part of the National Science, Technology and Innovation Strategy 2011–2016 (UBTYS), the Scientific and Technological Research Council has supported Turkey's competitive sectors such as automotive, manufacturing and information technologies and the business sector is the key part of this target since the expectation is that two-thirds of the research and development expenditures should come from private sector (Wedekind, 2013). The target for the innovation investments was identified as 3% of the GDP in the Strategy Paper for 2003–2023 while the total spending for the research and development activities has remained at 1.01% of GDP in Turkey, according to the latest available data (OECD, 2016). However, even with its current levels, research and development expenditures have increased approximately 14 times since 2001 (TurkStat, 2015).
Though the efforts of the state to cultivate innovation in industries associated with the radical innovation continue, over 45% of the international patent applications are filed by the diversified business groups and are related to the household appliances. White goods are characterised by incremental innovation and the leading firms in the sector have been investing in the research and development activities for more than 50 years. Arçelik, which is a subsidiary of the Koç Holding and the third largest manufacturer of household appliances in Europe, is the biggest investor to the research and development activities in Turkey. It firstly established an R&D centre in Turkey and later founded research centres in the high-income countries to get closer to the latest technologies. As an example to the institutional complementarities in HMEs, two features of Arçelik's growth pattern must be underlined here. Firstly, the strategy of the company is decided at the group level, together with the CEOs and family members of the Koç Holding in a top-down and hierarchical way. Secondly, as an advantage of being a part of a diversified business group, the resources for the innovation investments of Arçelik are derived from the intra-group capital markets.
Hierarchy is the prevailing feature in the variety of capitalism that is enforced through the institutional complementarities in Turkey. Turkey, as an HME, has four core features: the dominance of the family-owned diversified business groups in the economy, state-regimented and weak industrial relations, low skills and the influence of MNCs. Patronage relationship between the state and the business groups is complemented by the state's intervention to the industrial relations through formal and informal ways in a top-down manner which in turn vitiate the influence of the trade unions. Dominance of the diversified business groups in the economy, the interaction between the family conglomerates and their subsidiary firms, and the control of these firms by the family members through the executive boards despite the presence of the CEOs are also hierarchical features that are complemented by the absence of the strong stock exchange markets in Turkey. In other words, family members do not share their executive control with the external shareholders since the subsidiary firms are primarily financed by the intra-group accumulations instead of the equity markets. The influence of the diversified business groups in innovation investments and activities is also complementary to the public education system, which is also shaped based on the requirements of the business groups in Turkey.
However, there are some points to take into consideration for further studies. Firstly, diversified business groups are not peculiar only to Turkey or the other HMEs. As has been observed by Kim et al. (2004), family conglomerates have significant roles in the emerging market economies from Latin America to the East Asia, and they have been formed to counter problems arising from the underdevelopment of institutions in the economy they exist within. Though family conglomerates have continued to consolidate their powers both in Turkey and in Latin America even after the liberalisation efforts (Colpan & Jones, 2016), it is unclear whether these family conglomerates can sustain their dominance as their rasion d'etre disappears with the further institutionalisation and development of the economy. Secondly, considering together with the increasing importance of the stock exchange and state's efforts to create venture capital for the small- and medium-sized enterprises and start-ups, a transformation towards an LME can be predicted for the future of Turkey's variety of capitalism. Especially taking into account the significant investments into the radically innovative sectors, which constitute LMEs’ comparative advantage, such transformation would not be surprising. Thirdly, the gradual Americanisation of the Turkish diversified business groups concerning the governance and financing methods, as Colpan (2010, p. 38) observed, may enforce the probability of such transformation from an HME to an LME.
Conflict of interest
The authors whose names are listed immediately below certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.