Geopolitical decoupling and global production networks: the case of Ukrainian industries after the 2014 Crimean annexation

Abstract

This study investigates the decoupling of Ukrainian aerospace, defense and electro-engineering industries resulting from the Russian Crimean annexation in 2014. Conceptually, we contribute to global value chain/global production network research by developing the notion of geopolitical decoupling, thus augmenting the existing 2-fold typology. Moreover, the article elaborates a typology of recoupling. Empirically, we investigate patterns of decoupling from Russia and recoupling via alternative production networks as well as patterns of decoupling/recoupling according to the position of companies in the production hierarchy. We found a neat pattern of decoupling from Russia according to tier but profoundly different dynamics of recoupling with the European Union and Asia.

1. Introduction

This article seeks to contribute to the literature on global production by examining the impacts of geopolitical factors on the restructuring of production networks. Despite the often-central role played by geopolitical factors in the (re)constitution of these networks, their explicit consideration remains scarce within this stream of research (Glassman, 2011). Empirically, this article examines the profound restructuring of international linkages among companies in the Ukrainian aerospace, defense and electro-engineering industries resulting from the Russian annexation of Crimea and occupation of part of Donbas in 2014. The commencement of hostilities in 2014 in eastern Ukraine led not only to depopulation and severe economic decline in the affected regions (Mykhnenko, 2020) but also to a fundamental reconsideration of the geoeconomic positioning of Ukrainian industries in the global context (Ilchenko et al., 2021). This resulted in the decoupling of the investigated Ukrainian industries from the Russian production networks. At the same time, an intensive effort to find alternative customers (markets) and suppliers was exerted. Thus, this mode of decoupling motivated by the aggravated geopolitical context as well as the subsequent recoupling to European and Asian production networks differs fundamentally from cases of structural and strategic decoupling and recoupling so far documented in the literature (Bair and Werner, 2011a; Yang, 2013; Horner, 2014; Gereffi, 2020).

This article, therefore, aims to unravel the variegated nature of Ukrainian industrial decoupling and recoupling following the abrupt worsening of relations with Russia that began in 2014. We investigate both the overall pattern of decoupling and recoupling as well as the scale of decoupling and recoupling according to the position of companies in the global value chain/global production network (GVC/GPN)¹ hierarchy (i.e., we differentiate between lead firms and Tiers 1–3 suppliers). The Ukrainian aerospace, defense and electro-engineering industries have a strong tradition and, in certain market segments, have achieved international excellency. For example, the world’s largest cargo plane, Mriya (sadly destroyed by Russia in 2022), was developed and manufactured by the company Antonov located in Kyiv. Nevertheless, the exports and imports of Ukrainian companies were for decades strongly oriented toward the former Soviet Union and its successor states.

Thus, this article brings the following novelties. First, conceptually, in addition to a structural and strategic decoupling that has already been established in the literature, we introduce notion of geopolitically motivated decoupling and recoupling and develop an overall typology of decoupling and recoupling. Second, empirically, to our knowledge, this article is the first to apply the analytical lenses of GVC/GPN to the Ukrainian aerospace, defense, and electro-engineering industries.

This article is structured as follows. We first introduce the conceptual framework of the study, represented prominently by decoupling and recoupling with GVCs/GPNs. The histories of the Ukrainian aerospace, defense and electro-engineering industries are then briefly introduced. An outline of the methodological approach follows. Finally, the empirical section presents the main results of our analysis, and the concluding section summarizes the main conceptual and empirical contribution of this article and underlines the need for further research into geopolitical decoupling and recoupling types.

2. Theoretical framework

Over the last decades, closely intertwined and mutually inspiring GVC/GPN theories became established as the key approach to studying economic development at local, regional, national and global levels with important policy implications (Parrilli et al., 2013; Yeung and Coe, 2015; Coe, 2021; Blažek and Steen 2022; Crescenzi and Harman, 2022). GVC/GPN theories represent a comprehensive framework for explaining the organization and functioning of vertically disintegrated production in present today’s wide-ranging globalization phase (Dicken, 2015; Yeung 2015). One of the key attributes of global production networks is their hierarchical structure, orchestrated by lead firms that typically perform the main technological, research and marketing activities in a particular industry (Pananond, 2016; Shin et al., 2017). Lead firms engage numerous suppliers, which can be arranged into several tiers reflecting their competences, capabilities and production sophistication level (Thoburn and Takashima, 1992; Bamber and Gereffi, 2013; Tokatli, 2013; Szalavetz, 2017; Raj-Reichert, 2019; Blažek and Holická, 2022), while their governance mode may differ substantially (Gereffi et al., 2005).

One of key tenets of the GVC/GPN stream is that various companies, regions and countries are plugged-in into global production by variegated modes of strategic coupling (Yeung and Coe, 2015), which has profound impacts on their value-creation and especially value-capture (Pavlínek and Ženka, 2016). GPN literature defines strategic coupling as a mutually dependent, constitutive and dynamic process that presumes the existence of shared interest and cooperation between two or more groups of actors acting jointly to achieve their strategic goals (Yeung, 2014; Coe and Yeung, 2015). Yeung (2014) defined three basic modes of strategic coupling: organic, functional and structural. Strategic coupling should not be viewed as a one-off process but as a dynamically evolving process driven especially by continuous changes of cost–capability ratios and the variegated development of key assets among concerned actors, that is, firms, governmental organizations and other organizations on regional, national and global scales (Yeung and Coe, 2015; Coe, 2021).

The dynamism of strategic coupling also implies the possibility for a company or region to decouple from a given production network (Yang, 2013; Horner, 2014). Horner (2014, 1121) defined decoupling as ‘a reduction or break in the prevailing form of linkage between a particular GPN(s) and a territory’ and distinguished between structural and strategic decoupling. Structural decoupling is typically induced by a loss of competitiveness among regional companies and/or technological or market changes. This involves a gradual or sudden demise of revenues and jobs with dire consequences for the workers and the region at large (Bair and Werner, 2011a, 2011b). According to Yang (2013), structural decoupling is manifested by a loss in new investment rounds, the exit of foreign companies and/or the loss of foreign markets.

Structural decoupling is frequently followed (or can even be preceded) by the strategic coupling of lead firms with a new region that commands new technology, is closer to the market being served or offers yet lower production costs as epitomized by the ‘race to the bottom’ notion (Bair and Werner, 2011a, 2011b; Coe and Hess, 2011; MacKinnon, 2012; Yang, 2013). For example, Bair and Werner (2011a), in their study of textile manufacturing in Mexico’s La Laguna region, documented how internal as well as external factors drove the development of the regional industry to a relatively short-lived boom that was then followed by a severe decline as the center of textile production moved to Asia.

In contrast, strategic decoupling represents a targeted effort to escape the existing, unfavorable mode of coupling resulting in various detrimental effects, such as limited value capture and precarious working conditions. Consequently, as a result of a national government policy imperative, regional companies (often in concert with regional or national business or governmental organizations) seek to enhance positive regional or national development outcomes via dedicated efforts to upgrade their firm’s assets as well as the territorial assets and thus pave the way toward more symmetrical forms of recoupling (Horner, 2014). Thus, strategic decoupling should not be seen as a form of disengagement from economic globalization but rather as a strategy to leverage the benefits of globalization (Horner, 2014). Using the case of the Indian pharmaceutical industry, Horner (2014) demonstrated how a sequence of strategic decoupling and subsequent recoupling created opportunities for imitative learning and functional upgrading while, at the same time, rebalancing the former asymmetrical power relationships within production networks. Thus, decoupling and recoupling represent important mechanisms of uneven socioeconomic development (Yang, 2013).

Recently, however, yet another type of strategic decoupling—motivated by a broad array of geopolitical motives—has gained momentum. In particular, the current GVC/GPN literature is preoccupied with the push from various governments and companies to backshore/reshore, nearshore or friend-shore production and/or regionalize production on a continental basis at least within the most critical industries (Maihold, 2022). Behind these shifts in the global organization of production is, of course, a multiplicity of reasons. The rationale for this restructuring of GVCs/GPNs might be due to an effort to ensure quality and reliability in the supply of essential goods (e.g., for the healthcare system; Gereffi, 2020); growing irregularities, especially in maritime transport; fears of technological leakages to competitors or technological dependency upon them or straight geopolitical considerations (Gong et al., 2022; Pegoraro et al., 2022). Frequently, under such conditions, the key impetus for decoupling represents incentives and a regulatory framework imposed by national governments that might be enforced in liaison with international organizations or agreements.

An extreme case manifested by geopolitical factors is production network restructuring following international hostilities, conflicts or even wars. In such cases, decoupling from existing production networks can be wide-ranging and have profound negative impacts on both sides of the conflict, at least from a short-term perspective (Gereffi, 2020; Gong et al. 2022). Consequently, and in addition to the structural and strategic types of decoupling already established in the literature, we propose adding a third type—geopolitical decoupling. To acknowledge the wide spectrum of aforementioned geopolitical factors and their variegated impacts, we propose that geopolitical decoupling should be further divided according to prevailing motive into resilience-driven and warfare-driven types. Inevitably, these types of geopolitical decoupling represent ideal types in a Weberian sense and, consequently, it is common to encounter scenarios that fall within the spectrum between these ideal types. The key differences among these types of decoupling are summarized in Table 1.

It is worth noting that geopolitical factors not only lead to the restructuring of existing production networks via coupling, decoupling and recoupling but also play a pivotal role in constitution of such networks (Glassman, 2011). In fact, as pointed out by Glassman, global production networks are shaped and reproduced by social, political, institutional and cultural contexts. As an example of this phenomenon Glassman (2011) provides the ascent of Samsung as a prominent player in the global electronics industry. This accomplishment was significantly facilitated by the post-World War II (WWII) geopolitical alliance formed between S. Korean, USA and Japanese states which successfully overcome the rupture of diplomatic relations between Korea and Japan at the end of WWII (Glassman, 2011).

It follows from the above text and from the literature (Pegoraro et al., 2022) that recoupling is likely to be more challenging than decoupling. Moreover, while the key impetus of strategic and geopolitical decoupling tends to come from national governments, the role of governments in recoupling is much more modest, and recoupling mostly rests on the shoulders of company managers. Recoupling following structural decoupling induced by unfavorable cost–capability ratios, technological change or shifts in market demand (Bair and Werner, 2011a, 2011b) is, at best, difficult and often impossible. In principle, recoupling would require substantial upgrades in production or an alternative, less demanding, customer(s). The first option would require major effort and investment. The second is likely to provide only meager prospects in terms of value capture and the overall socioeconomic perspective of the affected region.

In contrast, strategic decoupling and resilience-driven type of geopolitical decoupling tend to be carefully considered, including the potential for future recoupling and are, consequently, well-prepared (Yang, 2013; Horner, 2014; Vanchan et al., 2018; Pegoraro et al., 2022). However, in its abrupt forms, geopolitical decoupling resulting from open international hostilities or war (such as the Russian aggression against Ukraine since 2014) is much more challenging and might be driven more by national government policies and/or international agreements on sanctions and can have profound negative impacts on both sides of the conflict (Yale, 2022). We therefore distinguish between resilience-driven and warfare-driven types of geopolitical recoupling (Table 2). However, it is important to recognize that these classifications, akin to those of geopolitical decoupling, embody ideal types in a Weberian sense, and therefore, in-between cases occur frequently. International conflicts resulting in sanctions or, in extreme cases, wars imply that many companies might suddenly lose not only their customers and important markets but also their key suppliers simultaneously, especially as regards closely interconnected economies (MacKinnon, 2012). This was exactly the case for Ukraine and Russia prior to 2014.

While the current strongly globalized world offers seemingly vast opportunities for finding replacements for former suppliers in other countries, in reality, pursuing this at the global scale tends to be cumbersome and lengthy and is, in addition, likely to proceed under sharply unequal conditions as the party affected by the war is under extreme pressure during negotiations. In these cases, recoupling into alternative production networks is highly challenging, and its success is strongly variegated at the level of individual companies as it is contingent not only upon various contextual but also company-specific factors. Obviously, an alternative strategy might be followed, namely, searching for domestic companies able to replace the original foreign customers and especially foreign suppliers. However, this strategy is also accompanied by substantial challenges, such as the limited size of domestic production capabilities and the market, a lack of technologies and know-how, shortages of highly skilled personnel, overall economic malaise and so on (Vanchan et al., 2018; Pegoraro et al., 2022). Thus, recoupling under strongly unfavorable conditions, as epitomized by warfare-driven type geopolitical recoupling, tends to result in adverse economic impacts on the companies and territories concerned.

Nevertheless, recoupling is likely to be a function not only of macro-level dynamics, such as geopolitical shifts and contests, but also factors specific to individual companies. One might argue, generally, that producers of generic (semi)products (i.e., typically, lower-tier suppliers) might recouple by finding new customers easier than more specialized higher-tier suppliers whose production tends to be highly customized to the needs of lead firms. Nevertheless, experience also shows that companies with own R&D capacities (i.e., mostly higher-tier suppliers) are often able to respond more quickly and more creatively to sudden changes and challenges compared with companies lacking their R&D (Glogar, 2013). Therefore, in our view, the ability of suppliers at various tiers to recouple with alternative production networks is conditioned by a multiplicity of contradictory factors, and thus, any patterns in recoupling are difficult to foresee—hence why this issue deserves an empirical investigation.

Consequently, in this article, we would like to answer the following research questions: First, what has been the overall pattern of decoupling from Russia and recoupling with other production networks after the 2014 Crimean annexation? Second, to what extent do the decoupling and recoupling patterns differ according to company position in the production hierarchy?

3. Evolution of the Ukrainian aerospace, defense and electro-engineering industries

Due to its rich natural resources (fertile soils, iron ore, coal, gas, oil, etc.) and ready access to the sea, the Ukrainian economy quickly became one of the key industrial centers of the Russian Empire during the Second Industrial Revolution. The abolition of serfdom in 1861 enabled a release of labor from the dominance of agriculture toward heavy labor-intensive industries, such as coal and ore mining, metallurgy and mechanical engineering. Thus, labor mobility and large injections of European and American capital led to the rapid development of industries, especially heavy branches and particularly in eastern Ukraine. Consequently, according to estimates, over 20% of the production in the Russian Empire was concentrated in Ukraine before the First World War (WWI) (Reyent and Serdiuk, 2004).

The origins of Ukrainian defense, aerospace and other branches of heavy industry were therefore founded before WWI. Several of the established companies continue to operate, despite the territory’s turbulent history, today. For example, Motor Sich (established in 1907) began to produce aircraft piston motors in 1915, Luhansk Cartridge Works (1885) established itself as a key supplier of ball cartridges for the Imperial Russian Army and the Kharkiv Locomotive Factory (1895)—today known as the Malyshev Plant—became the largest producer of tanks in Ukraine. Military equipment production, generally, represented an attractive business opportunity thanks to its high profitability (Lozynskiy, 2015). However, in contrast to the rapid development of various heavy industries, electrical engineering was weakly developed in Ukraine during the pre-Soviet period; for example, 90% of steam turbines were imported from Europe and the USA (Matsevatyi, 2009).

The loss of the Ukrainian War of Independence (1917–1921) and the establishment of Soviet governance brought profound changes, such as the abolition of private property, massive political repression, extensive industrialization and the introduction of general education to eliminate illiteracy (Holovko, 2003). Thus, starting in 1921, the development of the Ukrainian economy was orchestrated via a command economy under the Soviet Union’s state-socialism model until its collapse in 1991. Under state socialism, existing companies were nationalized and the involvement of foreign capital was forbidden. In Ukraine, this period is characterized by even stronger heavy industries development motivated by a state policy of industrialization to form the economic base of socialism and, later, communism (Holovko, 2003; Matsevatyi, 2009). Thus, it was during the interwar period (i.e., 1918–1939) that the Ukrainian electrical-engineering industry was established. In 1934, for example, turbine production started in Kharkiv (now known as JSC Ukrainian Energy Machines²), which, in cooperation with other electro-engineering companies, gradually satisfied the domestic demand for energy equipment (Matsevatyi, 2009). The interwar period also witnessed substantial further development of the defense industry. By the late 1920s, the Kharkiv tank cluster had been formed by restructuring the locomotive plant, and, inter alia, a special design team was established and later transformed into the independent company Design Bureau (Holovko, 2003). Likewise, the company Motor Sich launched the production of radial aircraft engines in this period, the modifications of which served Soviet military aviation until 1952 (Motor, 2023).

After evacuated industrial plants were relocated back to Ukraine in 1944–1945 (key industrial plants were evacuated from Ukraine further east after the Nazi Germany invasion of the USSR in 1941), Ukrainian industries experienced a new development boom owing to the geopolitical conditions of the Cold War and vast advances in science and technologies after the WWII, especially in the aerospace and defense industries (Dean, 2000). During the post-WWII period, several important companies, such as Ivchenko-Progres (1945), producing aircraft and spacecraft engines and the aerospace company Antonov (1952), were established and merged into large conglomerates. The rocket industry was also established at this point—PA Pivdenmash in 1951 and Pivdenne Design Office in 1954, jointly with other companies in Dnipro, Zaporizhia, Kharkiv and Kyiv (Holovko, 2003).

The collapse of the Soviet Union in 1991 and subsequent break-up of traditional industrial linkages led to a deep economic and social crisis in Ukraine. Many factories ceased production and/or went bankrupt. To stabilize the situation, 5 years after achieving independence, the Ukrainian government established its national currency (UAH) and employed a policy of industry and (some) infrastructure privatization. However, enterprises that were considered strategic to national security were exempted from this privatization. Therefore, most companies in the defense and aerospace industries remained state-owned; in 2010, these were integrated into the huge state concern, Ukroboronprom. Nevertheless, not all defense and aerospace companies were considered strategic. Moreover, the political power of some ‘red directors’ of state-owned companies enabled them to privatize even some of these strategic enterprises (e.g., Motor Sich). What remained of the companies in the aerospace and defense industries, as well as most electro-engineering firms, were subject to voucher privatization. Spread among employees and other small shareholders, these shares were, however, swiftly acquired by the top management of these companies, enabling them to take control of the plants during the 1990s (Dubrovskiy, 2007).

Thus, foreign capital role privatizing these companies was only minor in the investigated Ukrainian industries. Even though, in theory, foreign investors were allowed to participate in the privatization process, the Ukrainian state created informal barriers that discouraged foreign capital investment. In particular, the absence of a clear legal and institutional framework for foreign capital participation, unclear corporate legislation and a lack of reliable information on the Ukrainian regulatory system discouraged most potential foreign investors (Dubrovskiy, 2007). Thus, the role of foreign capital in Ukraine is much smaller compared with the former state-socialist countries in Central-East Europe and the Baltics.

Consequently, after 1991, the strategies of Ukrainian defense, aerospace and electro-engineering companies typically revolved around cooperation with Russian companies and firms from other post-Soviet countries with the aim of continuing production based on technologies developed in the USSR (Romanukha, 2014; Kyzym and Miliutin, 2018). An important part of their business activities, especially in aerospace and defense, represented repairs and upgrades to military or aviation products exported to developing countries before the break-up of the Soviet Union (Shevtsov and Bondarchuk, 2013; Salnikova, 2014). Nevertheless, at least some aerospace companies succeeded in establishing cooperation with advanced Western countries, including major ventures such as the American Antares and European Vega projects. Electro-engineering companies meanwhile successfully expanded their exports all over the world, especially of water turbines and related products (Ukrainian Energy Machines [UEM]: Geography of Supplies 2023).

The 2014–2015 Russian annexation of the Crimean Peninsula and occupation of key cities in the Donbas region led not only to dramatic socioeconomic impacts for the occupied territories but badly affected the whole Ukrainian economy (Mykhnenko, 2020). The scale of the impacts of this undeclared Russo-Ukrainian war in eastern Ukraine is documented by the fact that by July 2016, around 3.3 million people had fled their homes (1.8 million to other parts of Ukraine and 1.5 to Russia and Belarus) and tens of thousands of enterprises were closed, looted, dismantled and smuggled into Russia (Mykhnenko, 2020). This, along with the widespread destruction of the technical and social infrastructure, led to more than a 60% drop in Donbas’s GDP during the first 4 years of the conflict compared with a 10–15% drop across the whole of Ukraine (Mykhnenko, 2020).

In response to the Russian aggression, Ukraine banned the export of military and dual-use products to Russia (since 27 August 2014), terminated an agreement on military and technical cooperation with Russia (27 May 2015) and passed a law prohibiting the import of most products from Russia (30 December 2015). For industries, such as aerospace, defense and electro-engineering, the impacts of this sudden loss of the key Russian export market as well as key Russian suppliers were particularly profound. The state defense concern Ukroboronprom ceased all operations with Russia immediately after the occupation of Crimea. Since then, the investigated Ukrainian industries have physically lost only a few enterprises (thanks to the occupation of part of Donbas, esp. the Luhansk Cartridge Plant), but the mutual sanctions of Ukraine and Russia have been the key factor in the sharp decline of overall production volume. For example, a key Ukrainian producer of marine military engines has lost over 40% of its export revenues (Ilchenko et al., 2021). Thus, the main strategy of Ukrainian companies after 2014 has been to focus on finding new markets and suppliers outside Russian production networks (Ilchenko et al., 2021).

4. Methodological approach

The first step in our effort was to build a database of companies performing business activities in any of the three investigated industries. Using the Ukrainian database service YouControl,³ we selected firms with the following National Classification of Economic Activities codes (which is harmonized with EU NACE and US ISIC classifications): 27.1 (manufacture of electric motors, generators, transformers and electricity distribution and control apparatus), 27.2 (manufacture of batteries and accumulators), 28.1 (manufacture of general-purpose machinery), 30.3 (manufacture of air and spacecraft and related machinery), 30.4 (manufacture of military fighting vehicles) and 25.4 (manufacturing of weapons and ammunition). Next, we investigated the webpages of relevant industrial associations and Ukrainian trade fair reports and included in our database companies with relevant products but classified under a different industrial code(s); we identified approximately a thousand firms using this approach. Subsequently, each company was scrutinized using all available data (official statistics, company webpages and their annual reports when available) to verify the relevance of the company for our study. Inactive, bankrupt or insolvent companies as well as micro-enterprises with less than 10 employees were excluded. We also excluded companies located in the occupied parts of Donbas and Crimea. Consequently, the list was reduced to 329 firms. Of these, only 161 companies had complete information in the YouControl database about their economic indicators and foreign trade for the investigated period (2012–2020) and were thus selected for our empirical analysis.

For these companies, we extracted firm-level data covering ownership, types of goods produced, key economic indicators (e.g., turnover, revenues and cost of goods) and their export and import according to individual countries. Based on the case-by-case analysis of firm webpages, the companies were classified into lead firms and Tiers 1–3 suppliers using a methodology developed by Pavlínek and Janák (2007), which bases company classification on their most sophisticated/complex product(s). Particularly, according to Pavlínek and Janák (2007) Tier 1 suppliers typically produce pre-assembled subsystems or modules and/or the most technically complex parts to the final assembler. Tier 2 suppliers assemble less complex components using smaller parts, which subsequently become integral parts of a specific module. Tier 3 suppliers produce simple components with low value added such as simple plastic or rubber parts, simple castings, sheet steel and other semi-finished products, which are then used for the assembly of more complex parts (Pavlínek and Janák 2007). Examples of products typical of particular tiers in each of the investigated industries are provided in Table 3. Additionally, five state-owned specialized traders that secure a very specific niche in the Ukrainian defense and aerospace industries as they control most military and dual-use exports were classified as the fifth category.

In order to investigate the decoupling and recoupling processes in Ukrainian defense, aerospace and electro-engineering industries, we compared the export and import data from two distinctive periods: The first period (2012–2013) covers the phase preceding the annexation of Crimea (2014) and occupation of part of Donbas (2014–2015). The second period (2016–2020) represents a period in which substantial decoupling from Russia took place, and recoupling with alternative (especially Asian and European) production networks was achieved. As these periods are of unequal length, our comparison is based on yearly averages for both periods. While analysis of the decoupling from Russia is straightforward, to analyze patterns of recoupling, we grouped countries into six categories: (i) post-Soviet countries, except the Baltic states, (ii) the European Union (EU), (iii) other advanced countries, (iv) Asia, (v) other countries and (vi) export/import not specified. Finally, given the substantial devaluation of the Ukrainian currency and high inflation during the investigated period (Table 4), we provide both absolute and relative trade data and focus our interpretation on shifts in the shares of particular groups of countries or macro-regions in Ukrainian export/import.

5. Patterns of Ukrainian industrial decoupling and recoupling after the 2014–2015 annexation of Crimea and occupation of part of Donbas

In this section, we present the main results of our empirical analysis of the company-level decoupling from Russia in the three investigated Ukrainian industries and their subsequent recoupling with alternative production networks.

5.1. Decoupling from Russia

As explained above, the violation of international law by Russia via its annexation of Crimea and occupation of part of Donbas in 2014–2015 resulted in the disruption of business ties between Ukraine and Russia, especially in sensitive industries such as aerospace, defense and electro-engineering. Thus, the decoupling of these Ukrainian industries from the Russian economy epitomizes warfare-driven type geopolitical decoupling. This type of decoupling is generally highly unfavorable and challenging, but, in the case of Ukraine, its impacts were further aggravated by the fact that Russia was by far the most important trading partner in terms of both export and import in all three industries. Consequently, decoupling from Russia led to severe impacts on these industries. On average, in the second period (2016–2020), the turnover of companies in the examined industries dropped to nearly half the 2012–2013 values, indicating the scale of challenges faced by companies after the 2014–2015 Russian aggression (Figure 1).

Table 5 documents the immense scale of decoupling from Russia and recoupling with other countries. However, the following main trends should be emphasized: While the overall value of Ukrainian imports in the investigated industries fell by a staggering −81.1% between the two periods, exports dropped less (−56.9%). This implies Ukrainian suppliers replaced a substantial share of imports from Russia. The value of imports from Russia also dropped drastically (−93.5%). In contrast, exports to Russia declined profoundly, yet less severely (−81.7%). This is attributable to the fact that Ukrainian companies were (under the new geopolitical and regulatory framework; see above) forced to minimize their dependency on imports from Russia. However, the export of goods of a non-military nature to Russia (such as most electro-engineering products) was not restrained to the same degree as import. Nevertheless, counterintuitively, despite a dramatic decline in both imports and exports, trade between Ukraine and Russia has not ceased completely as might be expected, especially in such sensitive industries as aerospace, defense and electro-engineering. As expected, our data confirm that after the 2014 Russian aggression, export/import to/from Russia by aerospace and defense companies fell more sharply than in the electro-engineering industry, reflecting the strategic nature of the former industries. However, we admit that we were not able to distinguish precisely the scale of decoupling (and recoupling) in each of our investigated industries separately, as many companies manufacture products that fit several industries or have a multipurpose nature (cf. Table 3). This limitation also applies to Figure 1.

One can also reasonably speculate that in the second period (2016–2020), non-negligible parts of mutual trade between Ukraine and Russia were mediated by third-country middlemen companies. For example, exports to Latvia soared from annual average 975,000 USD in the first period to average of 41,717,000 USD in the 2016–2020 period—Latvia hosts a sizeable Russian minority (about a quarter of its population). Likewise, Ukrainian exports to Cyprus (from 5,078,000 to 12,647,000 USD) and Bulgaria (from 2,955,000 to 16,814,000 USD) also multiplied between both periods. All together this implies a substantial level of persistence in the relationships between companies, which were not severed completely even after the 2014 military conflict.

Trade with the EU underwent major changes as well. While imports from the EU dropped severely (−80.7%), exports soared (+32.9%). This is truly a substantial increase given the fact that overall Ukrainian exports dropped profoundly (−56.9%). These shifts are attributable to two main factors: First, the economic crisis in Ukraine resulted inter alia in the severe devaluation of the Ukrainian currency, and consequently, imports from EU countries became prohibitively expensive for badly affected Ukrainian companies. Conversely, in January 2016, the EU facilitated the export of Ukrainian goods to its single market via the EU–Ukraine Association Agreement. Despite this surge in Ukrainian exports to the EU, in absolute terms, export to Russia was still higher in the second period (219 mil. USD versus 240 mil. USD).

Finally, substantial shifts were also recorded in trade with Asian countries. Imports from Asia represent the only case where the value of trade remained practically unchanged between the two periods (−1.7%). This is remarkable given the vast drop in the value of overall imports (−81.1%). In relative terms, the Asian share of Ukraine’s imports soared from 4.0% to 20.9%. These data reflect an intense search among Ukrainian companies for cost-efficient suppliers caused by economic and geopolitical hardships. In the case of exports, despite a moderate drop in the value of exports between the two periods, in the second period, Asia established itself as the most important export territory for Ukrainian companies—in the 2016–2020 period, Ukrainian exports to Asia exceeded exports to Russia and the EU combined.

Table 6 presents the evolution of the average Russian share of total Ukrainian exports and imports between 2012–2013 and 2016–2020 periods and provides another view of the scale and nature of the decoupling of Ukrainian companies from Russia in selected industries. Overall, the share of Ukrainian imports from Russia dropped more profoundly (−65.4%) than the share of Ukrainian exports to Russia (−57.6%). This decline in both exports and imports to/from Russia was, of course, even more profound in absolute terms given the vast drop in overall production value (see Figure 1). As regards both imports and exports, a neat pattern can be observed in accordance with company position in the production networks. First, and in line with expectations, the largest relative drop in trade with Russia is seen among lead firms. Specifically, in the case of imports, the share of Russian supplies to Ukrainian lead firms dropped from 9.24% to only 0.79% (−91.5%), while the share of Ukrainian lead firm exports to Russia in overall trade dropped from 4.70% to 0.96% (−79.6%). Second, the share of imports as well as exports from/to Russia decreased more profoundly among Tier 1 companies compared with Tier 2 suppliers. In particular, the share of imports by Ukrainian Tier 1 companies in the 2016–2020 period dropped by −58.9%, whereas imports by Tier 2 companies dropped by ‘only’ −33.9%. This finding supports the argument that less specialized components needed by Tier 2 suppliers can be sourced from alternative suppliers more easily than more sophisticated and specific components purchased by Tier 1 suppliers and especially by lead firms. Finally, imports by Tier 3 companies tend to be much more generic and are, hence, relatively easy to replace by alternative (mostly Ukrainian) suppliers. Likewise, Tier 3 suppliers, due to the mostly generic nature of their production, can find new customers relatively easily compared with high-tier suppliers.

Nevertheless, by far the largest relative drop recorded in both exports and imports to Russia is among those companies outside GVCs/GPNs (−88.7% in case of export and −98.4% in case of import). These companies typically trade with a wide spectrum of various, largely one-off customers who provide them with better knowledge of the market and, thus, higher flexibility compared with suppliers involved in GPNs. As envisaged, in contrast to other territories, the role of ‘specialized traders’, that is, companies specialized in the export (and import) of products manufactured by other companies, was marginal in the first period and practically nonexistent in the second period. Overall, these results indicate that even severe international tensions and conflicts (such as the annexation of Crimea and occupation of part of Donbas) do not necessarily lead to a complete cessation of mutual trade, which demonstrates the unusual vigor and perseverance of international trade in general and of production networks specifically. This fact was also documented after the Russian invasion of Ukraine in February 2022, when, for example, despite many rounds of sanctions, trade between the EU and Russia has still not been severed completely. Accordingly, limited trade links still exist even between Ukraine and Russia even after the full-scale Russian invasion in 2022. For example, Russia pays Ukraine for the transit of its oil and gas to third countries via middlemen companies.

5.2. Recoupling of Ukrainian industries into alternative production networks

5.2.1. EU

Given the long-term dominance of Russia both in Ukrainian imports and exports, the ability to recouple with alternative production networks gained paramount importance after 2014. Two major markets—the EU and Asia—became an obvious priority for Ukrainian companies. Before the Crimean annexation, the EU share of total imports significantly surpassed the share of Ukrainian exports directed toward the union (Table 7). However, as a result of the sharply worsening geopolitical relationship with Russia, Ukrainian companies searched intensively for new markets for their products. The overall increase in Ukrainian exports to EU countries was facilitated by the association agreement, which free trade provisions came into force at the beginning of 2016. Thus, while in the second period (2016–2020) the EU share of Ukrainian imports remained practically unchanged (arguably due to the cost considerations of Ukrainian companies, which were operating in a truly challenging economic and geopolitical situation after the 2014 Crimean annexation), the share of Ukrainian exports to the EU more than tripled. In both periods, the UK represented the key export market for Ukrainian companies within the union—Brexit was in January 2020. In the first period, the UK was followed by Czechia and France; in the second period, this was Latvia and Bulgaria (see above).

The intensity of EU recoupling was strongly differentiated according to company position in GVCs/GPNs, as documented in Table 7. Lead firm exports increased by 57.1%, whereas Tier 1 and 2 suppliers expanded their exports approximately 3-fold. The highest rise in exports was recorded among Tier 3 suppliers (+740%). Again, these findings suggest that it is relatively easier to find alternative markets for more general and less sophisticated products than to penetrate highly saturated markets with final products, even under the more favorable regulatory framework offered by the association agreement. Nevertheless, a profound increase in exports by Tier 3 suppliers is also a consequence of a very low base level in the 2012–2013 period. The share of exports to the EU mediated by the specialized traders increased nearly six times (+578.4%), reflecting the need of certain Ukrainian companies for specialized support to enter the demanding EU market. As indicated above, Ukrainian exports to some EU countries (e.g., Cyprus, Latvia, Bulgaria and Lithuania) rose sharply by up to 50 times. In these cases, one cannot exclude the possibility that part of the Ukrainian exports to these countries was subsequently (most likely indirectly) re-exported to Russia. Consequently, Russia’s share of Ukrainian exports, as reported above, should be read with caution as, in reality, it is likely to be higher.

Compared with exports, the change in imports from the EU is less profound; still, there are significant changes. Lead firms and Tier 1 supplier companies recorded a relatively moderate decline in their share of EU imports. In contrast, Tier 2 and 3 suppliers substantially increased the share of imported components from EU countries in the second period, which might be related to their soaring exports to the single market, for which they needed certified components as inputs. Despite the abovementioned differences according to tiers, overall, the share of total Ukrainian imports from the EU increased only moderately despite the association agreement. This is not surprising given the high prices in most EU countries.

5.2.2. Asia

Despite the decline in absolute export value from Ukraine to Asia (by about a quarter; Table 5) and a minor decline in imports, the Asian market gained huge significance between the two investigated periods for Ukraine (Table 8). Whereas the share of Ukrainian exports to Asia increased by ‘only’ 81%, the Asian share of Ukrainian imports increased more than five times (by 519%). Consequently, the Asian share of all Ukrainian exports reached nearly 47%, making Asia a dominant market for the investigated Ukrainian industries. Unsurprisingly, the two largest export territories for Ukraine in Asia were China (167 mil. USD in the first period and 209 mil. USD in the second) and India (281 versus 128 mil. USD), followed by Thailand (in both periods), Iran (first period), and Turkey (second period). By far the highest increase in the Asian share of Ukrainian exports is recorded among the lead firms (more than 33 times); however, this is largely attributable to a very low benchmark value. In both periods, the largest share of Ukrainian exports was among Tier 1 suppliers, whereas the share of Tier 2 and 3 suppliers remained—despite increases in the share—negligible. A substantial share of Ukrainian exports to Asia was mediated by specialized traders, which is not surprising given the vast and profoundly differentiated nature of the Asian market requiring highly specialized knowledge.

6. Conclusions

This article aimed to contribute to the existing literature on global production by investigating often crucial, yet frequently disregarded impacts of geopolitical factors on the restructuring of production networks (Glassman, 2011). Therefore, conceptually, in this article, we further unpacked the notion of decoupling and recoupling by introducing a typology of both decoupling and recoupling. In addition to structural decoupling and strategic decoupling, which are already well-established in the literature (Yang, 2013; Horner, 2014), we coined the terms ‘geopolitical decoupling/recoupling’ and specified two subtypes: resilience-driven and warfare-driven. Resilience-driven geopolitical decoupling is done so as to enhance national industries, prevent know-how leakages and safeguard the resilience of key production. This type of decoupling tends to be carefully considered and prepared to moderate possible negative impacts. Warfare-driven geopolitical decoupling results from a sharp worsening in the geopolitical situation, which can lead to various sanctions or, in the worst-case scenario, military conflicts and wars, such as that launched by Russia against Ukraine in 2014. Therefore, warfare-driven geopolitical decoupling leads to a sudden rupture of existing linkages resulting in adverse impacts on all sides involved. Consequently, not only rationale, but also impacts of geopolitical decoupling differ substantially from structural and strategic decoupling (Horner, 2014).

Empirically, we investigated the shifting patterns of both the exports and imports of lead firms and suppliers from different tiers in three Ukrainian industries highly sensitive to the geopolitical situation—aerospace, defense and electro-engineering—over two periods (2012–2013 and 2016–2020). The results of our empirical analysis were both expected and surprising. In the 2016–2020 period, Ukrainian lead firms—as expected—recorded a much sharper drop in export to Russia than Tier 1 or 2 suppliers. Lead firms manufacture final goods which are much more easily targeted by sanctions compared with, often, multipurpose components produced by the suppliers. Likewise, the share of imports from Russia decreased most profoundly in the case of lead firms, followed by Tier 1 and Tier 2 companies. Nevertheless, the largest drop in imports from Russia was recorded in Tier 3 suppliers, which we attribute to the fact that these companies produce relatively simple components of a generic nature, and therefore, they can easily find alternative suppliers, mostly directly within Ukraine. Consequently, with the exception of Tier 3 suppliers, the intensity of decoupling tends to be positively related to the position of a company in GPNs, that is, the higher the position of a supplier in a production network, the more intense the decoupling.

However, despite a sharp drop in exports to Russia and, especially, imports from Russia, the country still served as an important trade partner for Ukrainian companies in the 2016–2020 period, even in these sensitive industries. Thus, these results indicate that even severe international tensions and military conflicts do not necessarily lead to a complete cessation of mutual trade, which demonstrates the unusual vigor and endurance of international trade in general and of production networks specifically. Moreover, the real volume of mutual trade between Ukraine and Russia after the 2014 annexation is bound to be higher than what the official statistics suggest, as individual firms and traders in both Ukraine and Russia intensively searched for ways to continue their business using various middlemen in third countries.

In terms of recoupling, Ukrainian companies succeeded within a very short period of time in recoupling with both EU and Asian production networks and markets. Nevertheless, the nature of this recoupling differs substantially. In the second period (2016–2020), the EU’s share of total Ukrainian exports more than tripled. In contrast, as regards their imports, Ukrainian companies were forced, due to the adverse economic situation after 2014, to intensely seek cheaper suppliers, especially in Asia. Thus, the Asian share of total imports within the investigated industries increased more than five times, while the EU’s share of the total imports remained practically unchanged.

Overall, we have to stress again that companies of all three industries have been affected badly by the Russian decoupling. The recoupling with alternative production networks was not able to compensate fully but only moderate the adverse impacts following the 2014 events. Despite the fact that Ukrainian industry has been severely damaged by yet another wave of Russian aggression launched in February 2022, the substantial adaptability and resilience of the investigated Ukrainian industries manifested after 2014 allow us to hope that Ukrainian industries and its economy at large will rise again when the war is over.

Unfortunately, given the current worsening of the global geopolitical situation as well as growing economic nationalism, one can foresee cases of geopolitically induced restructuring of production networks becoming more frequent in the near future. A prominent example of production networks restructuring is the ongoing struggle of both the EU and USA to rapidly enhance their semiconductor capacities, despite significantly higher costs, due to tensions in the Taiwan Strait. Presently, the Taiwanese population views their extensive and unique chip-making capabilities as a ‘silicon shield’ that protects the nation (Hsu, 2023). Any weakening of this shield could further escalate geopolitical tensions in the region. It is crucial to note that the potential (partial) decoupling of Taiwanese chip producers, as new factories and capabilities are established in Europe and North America, could have adverse economic and geopolitical consequences that extend far beyond Taiwan’s economy, even though, as Russian war against Ukraine suggests, total decoupling is unlikely.

Furthermore, significant restructuring of global production can be expected also in connection with decarbonization to mitigate the impacts of climate change. In this regard, Jim Glassman’s (2011) call for a full-fledged GPN analysis of global energy regimes can be considered farseeing. Therefore, the need to address grand societal challenges such as climate change, circular economy, digitalization and inequality (Mazzucato, 2018) will inevitably induce a new wave of restructuring in global production driven inter alia by the quest to secure access to key technologies, products and rare elements (including the associated processing technologies and capacities). This is likely to result not only in decoupling from existing value chains and recoupling with alternative ones but also in a constitution of new production networks. Consequently, coupling, decoupling and recoupling of companies and regions vis-à-vis GVCs/GPNs are likely to be frequent phenomena in the years to come. This underscores the need for further investigation into decoupling, recoupling and the constitution of new production networks motivated by geopolitical factors. Such research will enhance our understanding of the multitude of drivers shaping the current global economy and shed light on the diverse impacts of decoupling and recoupling on particular regions.

Footnotes

Acknowledgments

The authors extend their sincere appreciation to the editor and both referees for their exceptionally encouraging comments and profound insights, which have significantly enhanced the quality of the article.

Funding

Financial support of Grant Agency of Charles University (No. 219218) is greatly acknowledged.

Conflict of interest statement

No potential conflict of interest was reported by the authors.

References