Mapping specialisations in the international trade of automotive components and parts: a multilayer network analysis

Keywords:

Abstract:

The automotive industry has a relevant and increasing impact on the world econo-my's growth (Helper & Sako, 2010). The fragmentation of global value chains (De Backer & Miroudot, 2012) and the emergence of regional specialisations (Amighini & Gorgoni, 2014a; Gorgoni et al., 2018) are crucial issues to be addressed to understand the dynamics of change in localisation of production, in the growth of final demand of different types of vehicles and the changing demand of components by car manufactur-ers. Operating in an oligopolistic market, dominated by 14 firms controlling globally 62 brands (Hunkar, 2019) , companies’ strategies have been changing due to trade agree-ments, new suppliers in the past 25 years,e.g., Mexico who entered North America Trade Agreement, NAFTA, in 1994, and China, who entered the World Trade Organi-zation, WTO, in 2001 . In addition, environmental constraints embedded in EU and China policies (DRC, 2019; European Commission, 2019) urged changes not only in products (cars and their components and parts) and their technologies of production but also in mobility technologies and infrastructures and patterns of environmentally sus-tainable mobility across countries. These structural changes require broadening the anal-ysis to the entire automotive and mobility sectors (as opposed to limiting it to the car sector).
To fully understand those changes, one should rely on several sources of infor-mation: OEMs' strategies and changing locations of their suppliers in the global value chains; technological changes in products (cars, automotive components and parts) and in production technologies and materials; industrial and environmental policies; current individual and collective mobility infrastructures and plans for their development; pref-erential regional trade agreements; dynamics of consumers preferences and needs con-cerning individual mobility and the type of vehicles satisfying their needs, ranging from new to used vehicles, from individual vs shared vehicles, from compact cars to sport utility vehicles (SUV); changes in preferential trade agreements and their impact on trade networks of automotive components and parts.
All these features are strongly related to each other and it is not possible to address them in a single paper. We specifically focus on the changing composition of interna-tional trade of automotive components and parts, analysed by using bilateral trade flows. A specific strand of economic literature explores international trade in the auto-motive global value chains. In particular, there is an increasing number of papers using network analysis [Fagiolo, Reyes, & Schiavo (2009), Hausmann et al., (2011), Piccardi & Tajoli (2018), Piccardi and Tajoli (2014), Gorgoni et al., (2018) add other references ] [refs.]. Trade data fits perfectly in a network perspective: datasets from official sources (such as United Nations) are easily available on a yearly basis for all countries over long time-spans, at a very detailed level of commodity specification. Contributions in the lit-erature on international trade adopt several methods and models to analyse trade net-work statistics' structure and dynamics.
While the literature on international trade networks provides insights into the cen-trality of different geographic areas and on countries connencting those trade areas, less is known concerning the contribution of countries bilateral trades in structuring clusters of trade that are independent from the spatial location of countries and of the various commodities in structuring the relative position of countries. Both features are of utmost importance. The analysis by cluster of countries is needed to study the effects of trade agreements. The analysis by of trade by component is intended to shed light on the ef-fects of the present transition towards electric vehicles with a large impact on trade of internal combustion engines, and on the emergence of trade of new components and parts, such as high performing batteries and related components.
The analysis aims at identifying the meso level entities (intermediate between the individual countries and the entire network) that characterise the international trade of automotive components and parts over the last 25 years. In particular, the paper ad-dresses three main research questions. First of all, how can we identify cluster of coun-tries that characterize the trade network, without a reference to the conventional geo-graphical areas but referring to the recurring pattern of interactions occurring in their bi-lateral trade flows of various components? Secondly, having identified those clusters, what is the contribution of countries and of automotive components in determining the relative importance and the structure of the various clusters? Thirdly, what are changes in the structure of those clusters over time, i.e. of the relative positions of countries and their specialisations in multilateral trades?
The paper is structured as follows. Section 2 presents the survey of the literature on network analysis on trade data and analyses in international trade of automotive compo-nents and parts trade. Section 3 describes the methodology of multilayer cluster detec-tion adopted in our analysis. Section 4 presents the data source of bilateral export trades of automotive components and parts with regard to the years 1993, 2003, 2013, 2017. Section 5 illustrates the results of the Infomap multilayer cluster analysis. Section 6 builds on those results to focus on patterns of trades within and between clusters and on the relative importance of clusters and countries in the international trade of ICE com-ponents and parts. Section 7 concludes the paper with a focus on USA, Canada and Mexico, and on Germany and Central Eastern European countries (Czech Republic, Hungary, Poland, Slovakia) and with further developments of this research strand.