Introduction

1Data presented below shows how the four east-central European countries of the Visegrad group (Czechia, Hungary, Poland and Slovakia, referred to as CEECs) narrowed the economic gap with western Europe after 1990. Romania is also included as a comparator that followed a similar growth model with even lower relative wage levels. In all cases, rapid convergence coincided with an inflow of investment by foreign multinational companies but slowed down after the 2008 crisis. A number of authors have seen this slowdown as more than a temporary blip after which rapid growth and catching up can be resumed (Galgoczi & Drahokoupil, 2017). The EBRD, referred, in relation to the former state socialist countries as a whole, to a possible “middle-income trap” (EBDR, 2017, p. 7), arguing that restoration of upward convergence would depend on the emergence of a new growth model based on domestically-generated innovations and technological progress. An investigation of CEECs economic prospects requires going beyond drawing conclusions from simple extrapolation of trends and understanding the sources of dynamism and possible barriers to further development within that growth model. To do this the current article applies an adaptation of the varieties of capitalism approach embodying elements of the framework developed from dependency theory in Latin America.

2The first section sets the concept of dependent development in the context of theories of varieties of capitalism. This points to the need for an investigation of the strategies of, and relationships between, key economic actors as a basis for identifying sources of, and barriers to, economic dynamism. The second follows with a discussion of the failure of domestic-owned firms that leaves CEECs dependent on incoming MNCs. The third and fourth sections show, first, the extent of CEEC growth and the narrowing of the gap in economic levels relative to western Europe and, secondly, the place of MNCs within that process, showing the importance of export-oriented activities but also some of the dangers brought by dependence on this form of growth. A key issue for the future prospects of dependent development is the strategies of incoming MNCs for which CEECs are only a minor part of global operations. The fifth section approaches this with a discussion of what has attracted MNCs and what past research has revealed of their activities. The sixth section combines quantitative indicators to elucidate MNC strategies in terms of the kinds of products they export from CEECs. The concluding section discusses the barriers to the catching up process implied by these strategies.

1. Is it dependent capitalism?

3Transformations of CEEC economies since 1989 have led to capitalist economic systems as normally understood, with market systems, predominance of private property and integration into the global economy. However, they do not fit with the ideal types used to characterise advanced market economies by Hall and Soskice (2001), Amable (2003), as discussed by Lane (2007) and Myant (2016). Thus, in contrast to the two very precise varieties postulated by Hall and Soskice, CEECs lack the financial systems required for their notion of liberal market economy and also the coordination mechanisms between firms and with trade unions required for a coordinated market economy. Nor has any CEEC established an economic system that can generate autonomous innovative activity as a basis for any particular kind of international competitiveness, which is the dependent variable in the Hall and Soskice framework. Moreover, that framework is built around an assumption of reasonable stability for a variety of capitalism, a questionable starting point for CEECs where the striking feature has rather been transformation and change.

4Nölke and Vliegenthart (2009) proposed a solution with a new variety, a “dependent market economy”. Following the logic of the Hall and Soskice framework, CEECs are seen as having an institutional environment that makes them attractive to multinational companies for the transfer of routine manufacturing activities. Innovations behind this are developed elsewhere by those companies and finance for investment also comes from the incoming MNC. There is thus no need for domestic innovative activity or for a sophisticated financial infrastructure. This still remains essentially a static model. The principal aim was to show that a dependent variety could be added to Hall and Soskice’s two varieties and not to assess its long-term potential. The authors do not analyse possible barriers or limitations to growth and consider it “still too early to judge” prospects of the variety (Nölke & Vliegenthart, 2009, p. 677).

5It also remains unclear how far international integration through inward FDI determines other features associated with a variety of capitalism over which domestic actors have considerable discretion (Myant & Drahokoupil, 2011, chap. 16). State welfare spending, and hence also the total tax burden, as well as employee protection, are reasonably high by international standards in CEECs, as recognised notably in Bohle & Greskovits’s (2012) interpretation of “imbedded” neo-liberalism. This has not deterred incoming MNCs which, in fact, generally played limited roles in domestic politics. A more complete characterisation of CEEC capitalism would need to take note also of the roles of varied domestic actors, including the state, civil society, representatives of social interests and domestic capital, and of how they all interact with foreign capital.

6This is consistent with the development of the concept of dependency in relation to Latin America and other developing economies. At one time, drawing on Frank’s (1967) analysis of raw-material based economies, this was seen as a status that permanently hampered industrialisation. Subsequently, the term “dependent” was used to characterise development that did take place through the 1970s with a crucial input from inward investment by MNCs. Thus, a dependent economy was defined not by its stagnation, nor by a complete subordination to another economy. It was one in which accumulation and expansion could not find “its essential dynamic component inside the system” (Cardoso & Faletto, 1979, p. xx), or one in which “the rate and direction of accumulation are externally conditioned” (Evans, 1979, p. 27). Local business and a state with some developmental aims played significant roles such that dependence was a matter of degree rather than an absolute condition (Suau Arinci et al., 2015).

2. Dependent capitalism and domestic-owned firms

10The low level of innovative activity from domestic business is confirmed by evidence from individual countries (Weresa, 2017). Data on research and development spending in the business sectors, in Table 1, shows low shares of value added in comparison with Germany and even lower shares among domestic-owned businesses. MNCs do not undertake much of their research in CEECs, but they do more than domestic businesses. This is consistent with a low level of dynamism from domestic-owned firms while MNCs can rely on research and innovations undertaken elsewhere.

Table 1. Gross expenditure on research and development as % of value added

		foreign owned	domestic owned
Czechia 2013	Manufacturing	3.28	2.50
	All business activities	1.72	0.74
Hungary 2009	Manufacturing	2.08	2.90
	All business activities	0.95	0.83
Poland 2013	Manufacturing	1.39	1.10
	All business activities	0.73	0.44
Slovakia 2009	Manufacturing	1.57	1.06
	All business activities	0.68	0.20
Germany 2013	Manufacturing	9.44	9.37
	All business activities	3.55	3.80

Source: Calculated from https://stats.oecd.org/​Index.aspx?DataSetCode=AMNE_IN

12Despite its limited role in enhancing international competitiveness, domestic business was important politically as part of an emerging higher-income group and hence in influencing the policies that shaped the environments in which different kinds of businesses operated. It contributed to the consolidation of the right-wing of political life around particularly strong pressure for a small state and low taxes (Drahokoupil & Myant, 2017). Tax policy shifted from the early emphasis on following the then western European models – all CEECs had maximum personal income tax rates of 40% or more up to 2000 (Eurostat, 2015, p. 142) – towards cutting rates on companies and on higher incomes. All taxes on income and wealth in CEECs in 2015 were equivalent to only 7% of GDP as opposed to 12.9% for the EU-28. That difference is roughly the same as the average difference in state spending as a percentage of GDP.

13The low tax agenda, and also an agenda for labour-market deregulation, were pressed primarily by domestic business and politicians. MNCs gained substantially from it, but they generally kept a lower profile in domestic politics, pressing only quite specific issues that served their immediate interests, such as more scope for flexibility in working hours, rather than any general agenda. They appeared satisfied with the institutional environment and preferred not to risk becoming involved in unnecessary conflicts (Drahokoupil & Myant, 2017). Thus, and perhaps paradoxically, foreign capital was crucial to the economy, but local capital was more important in driving a political agenda that created favourable environments for inward investment seeking low wages and that left little scope for a developmental state role that could have helped the development of innovative domestically-owned businesses.

14In fact, any enhancement of the state’s developmental role was heavily dependent on EU financial support. Net annual capital transfers from this source over the 2007-2013 period were the equivalent of 3% of GDP for Hungary and 2% for Czechia, the highest and lowest respectively among CEECs (ETUI-ETUC, 2017, p. 17). The bias is strongly towards physical investment, even for spending classified as supporting research and development. This indicates an immediate contribution to GDP, but no contribution to business activity that could yield longer-term benefits has been demonstrated (Ferry, 2017). The weakness of domestic business and the limited role of the state mean that development and growth, outlined in the next section, remain heavily dependent on incoming MNCs.

3. Results of the CEEC growth model

15Figures 1 and 2 show levels of per capita GDP relative to the EU-15 – the pre-2004 EU members – using both nominal levels and purchasing power parity (PPP). The picture is similar for comparisons of productivity and earnings levels. From very low starting points in nominal terms, reflecting substantial devaluations at the start of the transition period, CEECs made up considerable ground, particularly in the period 2000-2008. Had that rate of catching up continued, nominal levels of the EU-15 would have been reached by the Czech Republic and Slovakia in 2017, by Hungary in 2024 and by Poland in 2028. Instead, only Czechia had passed 50% of the EU-15 level by 2017.

16Access

17Per capita GDP in PPP shows a smaller gap at the start of the period. Much of the catching up in nominal terms was therefore due to increases in nominal income levels and in currency values, partially countering the effects of the earlier devaluations and despite warnings that international competitiveness could be threatened. The PPP measure also shows rather less catching up and less of a slowdown after 2008. Indeed, continuing the post-2008 catch up rate would mean reaching the EU-15 level in 2034 for Czechia, 2050 for Hungary, 2031 for Poland and 2042 for Slovakia. These very different implications of the two ways of measuring relative GDP levels illustrate the inadequacy of relying on extrapolation of past trends in aggregate indicators. Fully catching up for these economies would depend on choices made by MNCs which cannot be analysed without a different kind of indicator, as discussed below.

4. FDI and the CEEC economies

18As this section will demonstrate, MNCs became absolutely dominant in key sectors of export-oriented manufacturing, making them crucial for economic growth and for supporting the external balance. Evidence also shows sectors in which foreign ownership was less ubiquitous and had less clearly positive effects.

20High inflow years broadly match the best years of catching up to the EU-15 GDP level, passing the equivalent of 5% of GDP in all CEECs by 2000 and 10% of GDP in some years in Hungary, Czechia and Slovakia. From around 2005 the pace of inward FDI slowed. The picture also became progressively more complicated with financial movements classified as FDI that were not related to productive investment, including significant outflows as well as inflows, and were closer in effects to portfolio investment. Despite continuing fluctuations, there was no consistent recovery to pre-2005 levels in any country, implying either that overall growth would be significantly slower or that some other source of an impulse would be needed (Hunya, 2015, p. 68). That role was partially filled for a time in terms of level of investment by the inflow of EU funds, as indicated above.

21Inflows of FDI came by purchases of enterprises during privatisation, by acquisition of previously privatised enterprises, by new greenfield investment and, increasingly, from further investment by MNCs already established in the countries. The privatisation route was particularly important in the early years. Hungary was the pioneer, showing a greater willingness to welcome foreign ownership. Poland was always more reluctant to welcome foreign ownership and retained a larger domestically-owned sector. Some exceptional figures in the inflows in later years were due to privatisations, mostly of banks, public utilities and some industrial enterprises, in the years around 2000. However, mergers and acquisitions, as recorded by UNCTAD, which also include purchases of previously privatised enterprises, contributed only around 25% of total inflows in each of these countries for the period 1991-2016. The great bulk of the inflow was therefore investment in new or expanding facilities.

22The place of MNCs in economies is shown by data on the role of foreign-controlled enterprises, defined as those with foreign control of more than half of shareholder voting power, covering the business sector. Table 2 shows CEEC levels compared with Germany, a country with a low level of dependence on foreign MNCs in European terms. They are near the top among EU members both in terms of employment share - only Estonia had a higher percentage than Czechia - and also in value-added share. Data on export shares are less complete, leaving a significant proportion unclassified, and may therefore understate the role of MNCs.

Table 2. Percentage shares of foreign-controlled companies in total, 2010

	employment	Value added	Exports (2014)
Czechia	26.9	42.9	76.6
Hungary	24.7	49.3
Poland	24.3	33.7	43.4
Slovakia	22.4	36.5	74.8
Romania	25.4	42.7	67.3
Germany	9.9	20.3	24.9

Source: Eurostat, tps00004, tec00024, ext_tec07

Table 3. Percentage share in value added of foreign-controlled companies by economic sector, 2015

	All business activities	Mining and quarrying	Manufacturing	Electricity, gas, heating	Retail	Telecommunications
Czechia	41.8	32.5	55.5	19.3	45.6	44.9
Hungary	52.5	54.8	63.4	59.6	41.0	90.5
Poland	29.9	2.3	41.0	9.0	35.6	69.2
Slovakia	45.8	66.1	60.8	35.2	38.8	87.9 (2014)
Romania	43.9		59.9	40.1	34.5	68.9
Germany	22.6	32.6	28.6	36.4	14.4	27.1

Source: Calculated from Eurostat, fats_g1a_08, sbs_na_ind_r2, sbs_na_dt_r2, sbs_na_sca_r2

Table 4. Percentage share in value added of foreign-controlled companies by branch of manufacturing, 2015

	Manufactu -ring	Food	Chemicals	Computers electronic and, optical products	Motor vehicles	Wearing apparel	Furniture
Czechia	55.5	32.8	54.9	75.1	84.6	21.1	15.8
Hungary	63.4	42.2	39.8	85.7	83.7	35.3	39.9
Poland	41.0	28.7	30.6	43.6	79.0	10.8	29.6
Slovakia	60.8	38.4	47.3	76.9	81.3	35.3	26.5
Romania	59.9	43.0	58.8	69.9	82.4	41.7	33.2
Germany	28.6	14.7	29.3	28.6	12.9	18.2	7.8

Source: Calculated from Eurostat, fats_g1a_08, sbs_na_ind_r2.

23The importance of foreign-controlled companies varies between economies – it is significantly lower in Poland - and between branches in individual economies. Data limitations prevent a comprehensive comparison, but Tables 3 and 4 show the MNCs’ dominance in export-oriented branches where modern manufacturing technology and product design were most important. They were less consistently important in branches oriented more towards domestic markets or where production processes were more accessible, as in food, wearing apparel and furniture. Foreign ownership was also important in other parts of the economy, including retail, mining and utilities, but the wider variations between countries suggest that it was less essential for adequate performance.

25The benefits and costs of FDI are likely to vary over time. As argued notably by Mencinger (2007), it could even become an embarrassing burden if financial effects follow a plausible cyclical pattern. Thus, an initial investment should bring a positive effect on the balance of payments. As the operation matures it brings growth and, if it leads to exports, further benefits for the external balance. Over time such benefits reduce as investment declines, as profits are increasingly repatriated and as exports decline with obsolescence of the product, or as alternative locations are used for production.

27Table 5, using 2014 as an illustrative year for which adequate comparative data are available, shows how repatriated profits reduced the net benefits to the external balance. The outflow of income associated with FDI was in all cases greater than the FDI inflow. A more complete measure of the contribution of MNCs to the external balance requires taking account also of exports and imports. These show a continuing net positive contribution to the external balance from foreign-owned industrial enterprises, with motor vehicles particularly important. Positive contributions were very much smaller for other branches. In all cases, Poland showed less dependence on MNCs than the other CEECs and, in consequence, less income outflows but also less of a positive balance of payments effect from their activities.

Table 5. Effects of FDI on external balances, indicators as % of GDP, 2014

	FDI	Income outflow	Net exports	Industry	Motor vehicles
Czechia	2.6	7.7	2.30	11.81	5.05
Hungary	5.6	9.4	-
Poland	3.2	3.7	-2.06	3.88	1.62
Slovakia	-0.5	4.7	5.27	11.76	7.13

Source: Calculated from OECD Foreign Direct Investment statistics, http://www.oecd.org/​corporate/​mne/​statistics.htm and Eurostat, ext_tec07

28The combined slowdown in inward FDI and increasing financial outflows points to a loss of dynamism for the economies and to dangers for their external balances which could have been masked during a period of slower growth and slower upward convergence. New dynamism might come from two sources, domestic business or new kinds of inward FDI. With the base for the first of these particularly weak, prospects for the near future would seem to depend on the activities of MNCs. Clarifying prospects therefore requires an examination of their strategies.

5. How MNCs take advantage of low wages

29CEECs provide encouraging environments for MNCs as countries with liberal trade and investment environments, physical and institutional frameworks adequate for large-scale manufacturing industry and substantially lower labour costs than in western Europe. Figure 4 shows pay levels relative to Germany, with some reduction in the gap between 2000 and 2008 followed by a widening up to 2016, as with nominal GDP levels. In addition to lower wages, MNCs could also, in some cases and to some extent, find laxer labour regulation and weaker employee representation making existing regulations easier to evade. The wage difference may therefore understate the full benefit for an MNC from setting up in a CEEC. The differences among CEECs were small when set against the gap with western Europe and, particularly when labour was often a very small part of total costs, other factors – quality of infrastructure, availability of labour, other support and incentives provided by governments – were likely to be more important for the choice between countries. Even these became increasingly similar as all CEECs did their best to facilitate inward investment.

30MNCs can benefit from lower wages in a number of ways. The most obvious is to undertake the same production processes as in their home bases, yielding higher profits. The share of personnel costs in value added in manufacturing in CEECs is significantly below the EU average, and even further behind the German level, as shown in Figure 5. This gap is small in comparison with the gap in nominal wage levels and further benefits come in less obvious ways; through lower prices of outsourced components, which reduces costs of production in the home base, through high prices for components imported into the CEEC for assembly operations, through returns of credits or through fees for patents or consultancy services that similarly flow back to the company’s home base.

31A substantial literature exists on the activities of export-oriented MNCs in CEECs, much of it based on case studies with the emphasis on electronics and, especially, motor vehicles (including Pavlínek, 2008; Pavlínek et al., 2017; Krzywdzinski, 2017; Sass & Szalavetz, 2013; Sass, 2017). They generally show a contribution from inward FDI to modernisation of industrial structures with an integration into global value chains. However, there are also biases towards activities requiring lower skill levels which can in turn often move on to lower-wage locations. There is also evidence of research and development activity, but largely that associated with current production. The highest-level research and, as indicated in Table 1, the overall bulk of research activities, remains in higher-income countries. Various barriers to catching up in full with western Europe are noted, such as shortages of skilled labour that discourage MNCs from undertaking the most demanding activities in CEECs, but questions remain about the nature of upgrading that still leaves nominal wages far behind western European levels and about the future potential for a growth model dependent on inward FDI.

6. What do MNCs export from CEECs?

32To provide a basis for answering these questions, a study was undertaken of individual products exported by MNCs, using the shares of personnel costs in value added in branches of manufacturing, taken from Eurostat data, and export and import data, taken from the United Nations Comtrade database. Volumes and kilogram prices of exports and imports were calculated for selected products, defined to the 4-digit and 6-digit levels of HS codes (Harmonized Commodity Description and Coding System) for the period from 1990 to 2016. The products chosen for analysis covered around 20% of exports for individual CEECs, including categories from electronics, garment manufacture, consumer goods and, above all, motor vehicles and their components. Selection was restricted to manufacturing industry and within that based on the presence of adequate volumes of clearly similar products in countries’ exports to make comparison meaningful. The units throughout are value in US dollars divided by weight in kilograms as a percentage of the figure for Germany, taken as the standard.

33The kilogram price measure was widely used in the state socialist period as an indicator of the relative quality of exported products. Thus, the kilogram price of passenger cars exported from Czechoslovakia into the European Economic Community (EEC) in 1966 was 68% of the level for intra-EEC trade (Myant, 1989, p. 234). There were even lower figures in other years and for a number of other manufactured products, all reflecting the need to sell products, often poorly made and with obsolete designs, at prices well below those of western-European made alternatives.

34There are limitations to the use of the kilogram price indicator in an open economy. A high value may reflect high prices of imported components with little further value added. It was therefore important to follow import prices and volumes of components where possible and also to seek confirmation of MNC activities from media and company sources. This made the case of computers and their components unusable due to an enormous volume of component imports. Thus, net exports in 2016 from Hungary, a major assembler of computers over a few years before the activity moved on to lower wage countries, were only 37% of total exports. The measure is very specific to individual products, as some are naturally heavier than others, so that there is little value in comparing kilogram prices between products or in seeking an average for a particular country. The categories also contain, to a varying degree, a mix of products, but this is often of little importance at the 6-digit classification level. A finer classification comes down to the level of specific products made by particular firms. Finally, the measure has little meaning for undifferentiated products, especially raw materials, and no meaning for services which have no recorded weight. However, the main driver for growth has been export-oriented manufacturing, and especially motor vehicles, for which the kilogram price measure, albeit with some differences between countries referred to below, is generally meaningful.

35Table 6 shows comparative figures for a sub-sample of products. No figure is recorded where the value of exports failed to reach 0.01% of that country’s total: low figures should be considered meaningless as they need not indicate production in that country. Remarkably, Germany remains a substantial exporter across all of these product groups, usually increasing its volume of exports up to 2016. The kilogram price of its exports is in the great majority of cases the highest of any country considered here. Comparisons over time show that there are some cases of a narrowing of the gap, but others where it barely changes.

Table 6. Price per kilogram of selected exports as percent of German level, 2016

HS code	Product description	Czechia	Hungary	Poland	Slovakia	Romania
401110	Car tyres	110	98	70	99	77
840734	Petrol engines, over 1000cc	65	91	-	64	39
840820	Vehicle diesel engines	122	109	84	154	43
846721	Hand-held electric drills	111	46	98	46	68
8471	Computers and components	65	68	45	42	21
850811	Vacuum cleaners	78	53	88	79	27
851140	Starter motors	66	68	57	-	57
851290	Vehicle lights and other electrical parts	79	147	28	80	61
854430	Wiring sets for transport equipment	71	84	31	87	64
8703	Passenger cars	60	73	54	64	43

Source: Calculated from UN Comtrade database, https://comtrade.un.org/​data/​

36In the early or mid 1990s CEEC exports still included products designed in, and inherited from, the state socialist period. In later years it is reasonable to assume that exported manufactured goods overwhelmingly followed designs brought by MNCs, even if they sometimes subcontracted production to locally-owned firms. Quality is satisfactory for their needs, or they would not have located production in those countries. The continuing low kilogram prices of exports from CEECs therefore does not mean that the products are not internationally competitive. It reflects a choice by an MNC to locate production of a product with a low kilogram price in that country. Understanding what this means requires an analysis of individual cases, a number of which are set out below within three types plus a separate discussion of passenger cars as a uniquely important case.

371. MNCs can outsource parts of their production chains that are the least demanding in terms of labour skill levels, benefiting from a lower price for the product and possibly also higher profit.

38Wiring systems for transport equipment (HS 854430), meaning mostly motor vehicles, are a prime example of this strategy. Part of the production process involves simple, manual tasks. The MNCs that dominate this branch produce to designs from vehicle manufactures. Production expanded in Czechia for a time, but much then moved on to even lower-wage countries (Pavlínek et al., 2017). Figure 7 shows that exports from Romania grew rapidly from 1999, passing the Czech level in 2001 and the German level in 2006. It would be hard to doubt that quality is maintained for a car component manufactured by an MNC for other MNCs but, as shown in Figure 8, the Romanian kilogram price remained some way behind the German level, in fact falling further behind as exports increased. This is consistent with production shifting through lower wage countries, enabling the manufacturer to sell the same product at a lower price. Evidence on labour shares, which would indicate whether higher profits also accrue from CEECs, is not available for anything closely matching this product group.

392. MNCs can transfer production of a product, developed in their home country, gaining from a lower price thanks to cheaper labour. There may also be a higher profit.

40The study revealed a number of cases from the motor-vehicle industry in which production was more demanding than for wiring systems and had not moved on to Romania or other lower-wage countries. This included vehicle lights and other electrical components (851290) and petrol engines (840734), in both of which the kilogram price of exports from CEECs was significantly below the German level after their exports had increased. Another similar case was starter motors for cars (851140). Exports from Hungary grew rapidly from 2004, following investment by Bosch. Exports from Germany declined and imports increased until that country became a net importer, with Hungarian exports to Germany reaching 5.7 million units in 2014, more than the latter country’s total automobile production. As Hungary, and also Poland, increased their exports, so their relative kilogram prices declined, as shown in Figure 9, suggesting again that at least part of the benefit to the MNC was in lower prices they could charge thanks to lower labour costs.

41A similar process of transfer took place with vacuum cleaners, as the Swedish company Electrolux moved is complete production process to Hungary in 2007, leading to an increase in exports shown in Figure 10. The kilogram price of exports from Hungary, shown in Figure 11, fell below the level of German, and also Swedish, exports, indicating again that lower wages were making possible lower prices, helping to maintain the products’ competitive positions. In fact, Swedish exports increased again slightly in later years, but in the shape of more expensive products and the kilogram prices of exports from Sweden and Hungary therefore diverged quite strongly in the following years.

423. MNCs can undertake production of products developed elsewhere with benefits accruing as higher profits and not lower prices.

43There were a number of cases where gaps in kilogram prices relative to Germany were insignificant or, in some cases, even to the advantage of the CEEC. This applies to electric drills, manufactured in Czechia by Black and Decker, and to diesel engines for vehicles, as shown in Figures 12 and 13. For this latter product, Germany’s key partners were Hungary and Poland where Audi and Volkswagen respectively produced for export. The big increase from Hungary came after 2001 and the kilogram price remained slightly above the German level. In fact, Germany’s exports peaked in 2007 and it was a net importer of diesel engines almost every year from 2001 onwards with 27% of its imports in 2016 coming from Hungary. The other major source was Austria with a very similar kilogram price. It would appear that products of very similar types were being traded between these and a few other countries.

44Another striking example, this time with a growing role for Romania, is the production of car tyres, shown in Figures 14 and 15. All the major tyre-producing MNCs became active in CEECs, in some cases with several plants in different countries. They typically claimed to be bringing their most modern production methods. The established Czech tyre manufacturer was taken over by the German company Continental in 1992 and at first largely supplied domestic producers. Exports increased only slowly, gaining pace from 2001. Hungarian production took off from 2008 after a big investment by Bridgestone. Polish output grew from investment by three MNCs and Romanian exports took off after construction of a new plant by Pirelli in 2006. Tyre manufacture requires substantial equipment but uses established technologies, controlled by MNCs, with relatively little innovation. This is therefore a production process that does not need close contact with research and development facilities and which can move to where wages are lower. Exports continued to grow from Germany, but their volume was only 64% of that of the four CEECs plus Romania in 2016, having been 179% in 2001. Remarkably, kilogram prices in CEECs were lower than those in Germany only in Poland and Romania, having in all cases caught up even as exports increased. The Romanian kilogram price fell below the Italian level, always somewhat behind Germany (87% in 2016), as Romania became a bigger tyre exporter than Italy from 2010. In this case, the lower price could have helped Pirelli maintain its competitive position.

464. Passenger cars represent a more complex case as MNCs have a wide product range with varying profitability levels and can choose where to locate assembly plants. There are clear business reasons for locating those with lower kilogram prices in CEECs, although this need have nothing to do with production technology.

47Passenger cars (8703), as indicated above, were a major driver of economic growth, accounting in 2016 for over 10% of exports from Czechia and Hungary, fully 20% from Slovakia, but only 3% and 5% from Poland and Romania respectively, and over 10% from Germany. MNCs entered from the early 1990s, mostly through privatisation of enterprises established in the state socialist period and then by greenfield investment. The expansion in exports is shown in Figures 17 and 18. As Figure 18 shows, German exports continued to expand, remaining above the total for all CEECs plus Romania.

48However, as indicated in Figure 19, the expansion in exports coincided with stagnation, or decline, in the level of relative kilogram prices, after an initial catch-up phase, and all were well below the German level in 2016. This gap cannot be explained by low product quality. Nor is it a result of obsolete design as MNCs typically produce new models in their CEEC plants, as they do in western Europe. The gap is the result of the kinds of vehicles they produce in their different plants.

49The category 8703 contains 7 sub-categories at the 6-digit level that have significant representation in the exports of CEECs or Germany. Table 7 shows the bias in CEECs’ exports towards smaller vehicles on which profits have generally been lower. The first column in Table 8 confirms that these have lower kilogram prices, a quarter the level for the lowest relative to the highest. Table 8 also shows that kilogram prices are below the German level in most sub-categories. The row showing the kilogram price for individual countries if they had the German export structure shows that, for the most part, this is of less importance in explaining the overall gap for 8703 as a whole than the mix of sub-categories.

50The strategy of producing cheaper cars in lower wage countries makes sense for an MNC with multiple production locations. Labour is a relatively small part of product value in car assembly - 5-10% depending on the location – but the saving is most significant for the smaller vehicles for which profitability is the most likely to be threatened. This applied across all CEECs as large-scale production took off. Thus, the fall in kilogram price in Slovakia after 2004 followed investment in new plants to produce small cars by Kia and PSA, while the previously high level reflected assembly, with a high share of imported components, of larger vehicles by Volkswagen.

Table 7. Percentage share of sub-categories in total value of exports of passenger cars, 2016

	Germany	Czechia	Hungary	Poland	Slovakia	Romania
870321	1.9	9.6	0.9	6.4	8.4	17.8
870322	9.3	24.6	17.0	46.6	14.9	15.5
870323	37.4	16.5	52.8	8.9	27.6	16.9
870324	8.3	0.4	0.4	0.8	3.6	0.3
870331	3.7	3.3	3.0	6.1	0.9	44.7
870332	30.0	44.8	24.8	30.4	28.2	3.0
870333	8.4	0.8	1.0	0.8	15.9	1.6

Note: the HS codes for categories of passenger cars are; 870321, petrol engine, to 1000cc; 870322, petrol engine, 1000-1500cc; 870323, petrol engine, 1500-3000cc; 870324, petrol engine, over 3000cc; 870331, diesel engine, to 1500cc; 870332, diesel engine, 1500- 2500cc; 870333, diesel engine, over 2500cc

Source: Calculated from UN Comtrade database, https://comtrade.un.org/​data/​

Table 8. Kilogram price of passenger car exports, CEECs plus Romania, 2016

	As % of average for all exports from the six countries	As % of German level
		Czechia	Hungary	Poland	Slovakia	Romania
8703	100	60	73	54	64	43
Outcome with Germany’s structure		68	95	53	61	60
870321	59	90	128	110	76	77
870322	73	78	93	74	76	58
870323	115	51	65	47	63	35
870324	243	77	161	97	29	89
870331	70	86	144	83	82	63
870332	94	75	86	60	72	93
870333	151	90	106	82	72	106

Note: the HS codes for categories of passenger cars are; 870321, petrol engine, to 1000cc; 870322, petrol engine, 1000-1500cc; 870323, petrol engine, 1500-3000cc; 870324, petrol engine, over 3000cc; 870331, diesel engine, to 1500cc; 870332, diesel engine, 1500- 2500cc; 870333, diesel engine, over 2500cc

Source: Calculated from UN Comtrade database, https://comtrade.un.org/​data/​

6. Conclusion and discussion

52If the varieties of capitalism approach is to be of lasting value, then it should ideally be able to provide a basis for understanding and predicting economic development that is superior to a simple extrapolation of past trends. It should be possible to show sources of, and barriers to, economic dynamism, including, where appropriate, political and institutional factors that are not taken into account in purely economic analyses. The kind of capitalism that has been established in CEECs, it is argued above, can reasonably be characterised as dependent. Key domestic actors have proved incapable of providing economic dynamism, leaving that role to incoming MNCs for which CEECs are but a small part of global operations. Without a substantial change from the established variety of capitalism, economic prospects for the foreseeable future will depend on how these MNCs choose to use their CEEC operations.

53It is striking the extent to which these CEEC operations remain peripheral, duplicating what has been proven elsewhere. Those products and activities that require close contact with R&D remain based in higher-income countries. Indeed, among the strategies covered above, a notable absence was the complete migration of an MNC. Without that, and/or without the development of their own large companies, CEECs will be unable to catch up to the western European economic level. However, there are substantial barriers inhibiting any such process which can be summarised under three points;

541. MNCs already have substantial facilities in higher-income countries and moving would carry substantial financial costs. They would lose benefits from a qualified and adequate labour force, built over many years, and from established contacts and networks. Moving could also incur an enormous political cost. They do move gradually, by closing plants more quickly in higher-income countries and opening new ones in CEECs. An illustrative example is the French car manufacturer PSA which closed a factory in France in 2013 while expanding production, making newer models, in its newer plant in Slovakia. Even this carries political risks and caused prolonged protests.

552. With a multi-plant operation, it makes the most sense to keep activities in higher-income countries for which profitability is the most certain, irrespective of pay levels. In the case of passenger cars, that means larger and more expensive vehicles. It clearly is not impossible to assemble in lower-wage countries. As the evidence above showed, large cars are also produced in Hungary and Slovakia. For production alone, if MNCs were choosing without any sunk costs, they could well place much more in CEECs, or possibly even lower-wage countries.

563. It is particularly difficult to transfer research facilities. This is a significant part of spending in manufacturing, as indicated in Table 1 which shows it equal to over 9% of value added in German manufacturing. Barriers to moving to CEECs include the need to attract adequate employees and they are unlikely to be impressed by low pay levels and poorer social service provision. Indeed, thanks to provisions for free movement of people, such employees are likely to be tempted into higher-income countries, accentuating difficulties also for any domestic business that might envisage building a substantial research base. The labour issue can also extend beyond research into the employees who could put new methods into practice, testing and adjusting as problems arise. The effect, again, as argued by Krzywdzinski (2017, p. 261), is to keep the most advanced activities in higher-wage countries. Labour shortages are a much smaller issue for less skilled employees and these can even be recruited from other countries, as is common in motor-vehicle production in Czechia (Drahokoupil, Myant & Domonkos, 2015).

57Part of the explanation for low relative wages is the weaker bargaining power of labour. This would seem the obvious explanation where the product and production processes are broadly identical but the share of labour in value added is very much lower in CEECs, as in the case of tyres. This weakness is a consequence of past history, bequeathing a less militant labour force often, but not in all workplaces, with low trade union membership. Employees are also likely to be impressed by pay in MNCs that is generally above the levels of domestic firms. Above all, however, their bargaining position is weakened by the strength employers acquire from having multiple plants. Threats to transfer production, to concentrate investment elsewhere or, ultimately, to close a plant in total, figure persistently in collective bargaining, encouraging moderation from employees. This applies even in the motor vehicle sector where investment appears among the most permanent (Drahokoupil, Myant & Domonkos, 2015).

58Bargaining strength in relation to shares in the current the product explains only part of the gap in earnings levels between CEECs and western Europe. Nor can the gap be explained in simple terms, as orthodox economics might predict, from differences in worker productivity. As the evidence above has indicated, low wages are a cause of low productivity as much as the converse. If wages are low then MNCs will choose that location for activities that do not require highly-paid workers, meaning less-skilled-labour-intensive activities that show a lower level of measured productivity. Some of the examples covered above seem to illustrate a very logical method for an MNC to benefit from cheap labour. Rather than drawing a higher profit directly, they reduce a product’s price when it is outsourced. Productivity is then measured as lower in the lower wage country simply because wages are lower, even if, as can be the case, the production process and product are the same. This is reflected in the observed lower productivity levels in CEECs, both by current exchange rates and also by PPP. Persistent advice from international agencies that wages should rise no faster than productivity, and that holding down wages is the key to maintaining competitiveness, therefore have the perverse effect of helping to keep those countries at a lower economic level.

59Wage levels for MNCs are in fact not determined in any precise way in CEECs. A lower limit is set by going rates in the country, reflecting past history and conditions in other parts of the economy. An upper limit is set by the level at which an MNC will be attracted to invest and/or dissuaded from leaving. That decision also depends on many other considerations – infrastructure, quality of labour, forms of government support – and wages are often a small share of total costs. This leaves a substantial zone of freedom for wage levels. They are pulled upwards where MNCs seek to recruit and stabilise their labour forces by paying somewhat more than domestically-owned firms. They are held in check by the relatively weak bargaining position of labour. Indeed, a slightly more aggressive bargaining approach from Czech trade unions, plus government acceptance of the case for higher public sector pay and higher minimum wages, were important factors in rising relative nominal wages from 2014 (Myant, 2018, p. 33) and hence also in higher relative per capita GDP levels. Evidence on the low labour shares in CEECs suggests that there could be substantial scope for further pay increases across the region before MNCs would consider transferring production elsewhere. Indeed, higher wages could help reduce barriers to transfer of higher-level activities into CEECs, although that would also need very substantial changes in government policies to promote development of skills and innovation and to encourage the most qualified to stay in, or come to, CEECs.

60The implication is that this form of dependency sets limits to catching up with western Europe. CEECs are left some way behind, even threatened with further instability as MNCs may move on to still lower-wage countries. Using the typology of Porter’s stages of competitiveness (Porter, 1990), they cannot move into the “innovation-driven” stage. Innovation always comes from outside, embodied in products and processes designed elsewhere. Production in CEECs is based always on established methods and technologies so that these countries can never lead with the newest. That does not preclude quite rapid growth. It does imply that they are left in a “middle-income trap”, holding down wages to make themselves attractive to MNC, but in turn leaving themselves unattractive to the higher-level activities that would raise their economies to the western European level.