The Impact of Outsourcing to Middle and Eastern Europe

Content

II. List of Figures

III. List of Appendix

IV. List of Abbreviations

1 Introduction

2 Theoretical framework for the prediction of outsourcing effects
2.1 Outsourcing as international trade in intermediate inputs
2.2 Effects of outsourcing in conventional trade theory: the Heckscher-Ohlin model
2.3 Characteristics of outsourcing to Eastern Europe and model assumptions
2.4 New trade models focusing on outsourcing
2.4.1 Feenstra and Hanson (1997)
2.4.2 Arndt (1997)
2.4.3 Kohler (2001)
2.5 Dilemma of theory in explaining outsourcing effects

3 Effects of outsourcing in an empirical perspective
3.1 Elements of an empirical analysis of outsourcing
3.2 Labor market trends in Western Europe since 1990
3.3 Magnitude and structure of outsourcing
3.4 Size and direction of outsourcing effects
3.5 Empirical links between outsourcing and the labor market

4 Policy Implications

5 Conclusion

V. References

VI. Appendix

II. List of Figures

Figure 1: Minimum cost combinations

Figure 2: Production stages within a sector

Figure 3: Outsourcing scenario

Figure 4: Production points before and after outsourcing

Figure 5: Growth of EU 12 outward processing exports

III. List of Appendix

Appendix 1: Labor productivity in OECD countries in growth rates

Appendix 2: EU 12 outward processing export shares in 1999

Appendix 3: Direction of trade of CEEC in 1993, 1997, and 1998

Appendix 4: Composition of EU-oriented exports and imports by CEEC (1993, 1997)

Appendix 5: List of examined empirical studies

Appendix 6: Skill structures in German industries

Appendix 7: Definition and source of variables

Appendix 8: Results of testing the wage bill as dependent variable

Appendix 9: Results of testing wages and employment separately

Appendix 10: Share of outsourcing in the considered industries (average annual changes in % between 1995 and 2002)

Appendix 11: Regression results for parent labor demand in Germany and Austria

Appendix 12: Regression results for the different sectors

Appendix 13: Share of high-skilled jobs in subsidiaries

Appendix 14: Criteria for separation of skill groups

Appendix 15: Differences between the broad and the narrow outsourcing measure (in %)

Appendix 16: Estimation results for the different skill groups

IV. List of Abbreviations

Abbildung in dieser Leseprobe nicht enthalten

1 Introduction

The EU integration process since 1990 has markedly lifted trade barriers between Western and formerly communist Eastern European countries. As a reaction, trade flows between these countries intensified in recent years. For instance, the share of German exports directed to Central Eastern Europe (CEE) amounted to 7.8 % in 2001 while imports from CEE were as high as 8.8 %. This compares to 3.0 % and 2.8 %, respectively in 1990.¹ By outsourcing parts of their operations to Eastern European countries in search for cost-efficiency domestic firms account for an increasing part of these trade flows.. In this respect, many Western European countries today may be perceived as “bazaar-economies”.² The political as well as academic discussion on benefits and costs of outsourcing is intensive and particularly focuses on labor market consequences.³ A popular - however often only anecdotally-evidenced⁴ - view is that outsourcing adversely affects low-skilled workers⁵ in the outsourcing countries.

This paper therefore intends to explore the effects of outsourcing with an emphasis on labor market consequences in Western European countries. The analysis abstracts from welfare considerations of outsourcing and exclusively concentrates on distributional effects.⁶ In order to comply with this paper’s objective, the impact of outsourcing is examined by referring to both theoretical frameworks and empirical evidence. On this basis, policy implications on how to respond to the identified trends and challenges in the ongoing European integration process are derived. Unless stated differently this paper argues from the perspective of a Western European country whose firms outsource parts of their value chain, e.g. certain pro- duction elements, to Eastern Europe.

Outsourcing⁷ is defined here as the geographical split-up of a firm’s value chain over national boundaries.⁸ It incorporates the procurement of intermediate inputs from a foreign location via imports that either takes place within a firm (intra-firm) or outside the firm in the form of arm’s length transactions (inter-firm).⁹ Thus outsourcing may (optionally) be accompanied by vertical foreign direct investment.¹⁰ The intermediate inputs receive a value-added in the outsourced stage of the value chain. Subsequently, they require final assembly upon their return to the outsourcing entity.¹¹ Therefore the phenomenon of outsourcing can be clearly separated from buying raw materials and trading final goods.

The paper proceeds as follows: Section 2 relates the characteristics of outsourcing to trade theory. Different theoretical frameworks predicting the effects of outsourcing on factor-prices are presented along with their central assumptions. Additionally, they are put into perspective with regard to a conventional Heckscher-Ohlin framework. In section 3 the model-outcomes are confronted with reality by reviewing relevant empirical studies in the context of outsourc- ing to Eastern European countries. The findings of the previous sections are then - in section 4 - transformed into suggestions for the political agenda dealing with the labor market challenges resulting from a further EU integration. Section 5, finally, concludes on central insights of the paper and indicates directions for further research.

2 Theoretical framework for the prediction of outsourcing effects

2.1 Outsourcing as international trade in intermediate inputs

Outsourcing can be associated with international trade in vertical fragments of value-chains.¹² With the intention to reach a cost-efficient production disaggregated parts of a firm’s value- chain are allocated to countries in which the specific operations are performed at the lowest costs while delivering the demanded quality. This allocation of production becomes neces- sary, as the required input factors are not tradable themselves.¹³ The production cost differen- tials between countries are related to technological factors (productivity) and differing factor- endowments (factor proportions). Countries specialize in those fragments of the value-chain that correspond to their comparative advantages and import the corresponding intermediate inputs.¹⁴ However, it is important to distinguish trade related to outsourcing from traditional trade: While outsourcing induced trade deals with intermediate inputs and takes place within a certain sector (intra-industry trade) conventional trade is characterized by its focus on final goods, which are exchanged between sectors.¹⁵

In the context of outsourcing to CEE countries it is - as a first conventional approach¹⁶ - assumed here that the receiving countries are relatively labor-abundant and that their workforce is comparatively low-skilled.¹⁷ Due to these endowments costs for low-skilled labor are relatively cheap in Eastern Europe.¹⁸ The more developed Western sending countries enjoy a relatively higher capital stock in combination with a more advanced level of technology, which is closely related to their relatively higher reservoir of high-skilled labor. Therefore Western firms outsource low-skilled labor-intensive production stages to Eastern Europe while carrying out more skill-intensive activities domestically.

2.2 Effects of outsourcing in conventional trade theory: the Heckscher- Ohlin model

In order to explore the effects of trade flows related to outsourcing from a traditional economic point of view the central implications of conventional trade theory - with main reference to Heckscher-Ohlin¹⁹ (termed “H-O” thereafter) - are briefly outlined as groundwork for further considerations. The model is first presented in its general form and then applied to the specific setting of outsourcing to CEE countries.

International price differentials for production factors as the cause for external trade have been explained in two ways in the neo-classical trade theory.²⁰ Ricardo²¹ assumes the factor endowments of the involved countries to be identical and focuses on productivity differences as the reason for comparative cost advantages. In contrast, the H-O model rests on the as- sumption of identical production functions and views different factor proportions as the cause for trade: Countries export goods that extensively require factors which the country relatively abundantly disposes of, whereas these countries import those goods that mainly use factors which are relatively rare at home. Hence the model allows for explaining changes in factor prices in addition to changes in product prices, which is of value for an analysis of outsourc- ing effects on the labor market.

H-O examine the effects of trade on factor-prices such as capital and labor: Assuming a two countries, two factors, two goods - world in which capital-rich country A specializes in the capital-intensive sector 1 and labor-abundant country B in the labor-intensive sector 2, the rents to capital in country A will increase and wages will decrease after the establishment of trade between the two countries. This is due to the fact that the output increase in the capital- intensive sector 1 after specialization demands an over-proportional amount of capital while the output reduction in sector 2 releases more labor than capital. Therefore capital becomes relatively scarce and demands higher rents. The inverse outcome holds true for the labor- abundant country B.

In general, international trade lowers the return to the relatively scarce factor in an economy (due to the adjusted factor demand after specialization). There is a tendency for factor prices to converge at the international level, as those factors whose prices increase after entering trade have been relatively cheap before (because of their relative abundance). H-O state that a complete equalization of international factor-prices can be expected under certain condi- tions.²² The predicted convergence of factor-prices induces at the same time a convergence of international factor-intensities, since the changed factor prices represent an incentive for firms to modify the input relations of production; if technically possible, they will increase the use of the now cheaper production factor. On this account, the Stolper-Samuelson theorem pro- vides a link between goods- and factor-prices: If the relative price of a good increases through external trade, the price of that factor intensively used in the fabrication of the product also rises at the expense of other factors. Thus the effects on relative factor-prices stemming from external trade are magnified.

If applied to the context of outsourcing to Eastern Europe as sketched above, H-O-type differentials in countries’ factor-proportions in terms of the labor market are represented by discrepancies in high-and low-skilled labor endowment²³: In this scenario Western European countries enjoy a comparatively big reservoir of high-skilled labor whereas this factor is scarce in the CEE countries. This setting implies a relatively higher wage gap between highand low-skilled labor in the Eastern European countries compared to the Western European countries. Given international factor-price convergence, wages for high-skilled workers in Western countries will rise and low-skilled workers’ fortunes will decline. Therefore labordemand and wages develop negatively for low-skilled workers in the respective Western countries, as they face increased import competition. Blue-collar workers can only choose between accepting lower wages and facing unemployment.

As will be shown in section 3, these predictions are not necessarily consistent with empirical findings. Given the outlined characteristics of outsourcing and the specific economic envi- ronment when analyzing outsourcing to Eastern Europe, traditional trade theory has some shortcomings.²⁴ These shortcomings are either rooted in the assumptions of the model or its mechanics. The following paragraphs will first comment on the assumptions made by H-O and then present the most influential modifications of the model proposed by various authors in order to be better suited for the analysis of the effects of outsourcing induced trade.

2.3 Characteristics of outsourcing to Eastern Europe and model as- sumptions

The H-O model is based on multiple assumptions of which some are inappropriate in an analysis of outsourcing to Eastern European countries: They either do not reflect the specifics of outsourcing or they misrepresent the economic conditions of the involved countries.

Identical production functions across countries are a central pillar in the H-O framework. However, equal levels of technology are unrealistic in the case of Eastern and Western Euro- pean countries. Instead Western European countries display a more advanced technological level, which is exemplified by their higher labor productivity. It is essential to include techno- logical aspects in the analysis of outsourcing effects since “exogenous technological progress biased against unskilled workers”²⁵ is mentioned in the literature as a second central driver for the deteriorating situation of low-skilled workers besides trade effects. Accordingly, shifts in technology affect demand for labor in two ways: on the one hand directly, as labor is substi- tuted by capital; on the other hand indirectly over trade, as the country experiencing the tech- nology shift intensifies its capital intensive production (thereby further reducing the demand for labor). Technology and trade display comparable consequences in terms factor-price effects. Technological progress and trade are not mutually exclusive determinants of factor- demand. Instead, both can be interrelated. Compare in this context, for instance, Wood²⁶: “[…] the pace and direction of technological change may be influenced by trade […].” Ander- ton/Brenton/Oscarsson²⁷ refer to the term of “defensive innovation”, which describes a coun- try’s efforts to achieve higher productivity when confronted with import competition. It is often difficult in reality to determine the relative strength of each of the two co-existing de- terminants.²⁸

Inter-industry trade is the standard case in the H-O framework, which, as already mentioned, is not suited to analyze the trade effects of outsourcing: “Traditional trade theories primarily explain movements in relative wages across industries, whereas what needs to be explained is the actual dramatic fall in the relative wages and employment of unskilled workers that has occurred within sectors”.²⁹ This distinction is crucial because an analysis in a one-sector setting allows for isolating the factor bias of outsourcing, as movements of factors between sectors can be excluded.

In the conventional H-O model the object of analysis is trade in final goods. Outsourcing, though, refers to the import of intermediate goods which re-enter the domestic production process. This is a decisive distinction since domestic prices for goods are not only influenced by worldwide competition in final goods but also by prices of imported inputs that substitute for domestic inputs³⁰. The employed foreign intermediate inputs in the final product imply changes in the factor-intensities of the respective production fragments that stay domestic. This exercises forces on domestic factor prices, as the domestic demand for production factors adjusts according to the modified factor-intensities in production. It is essential to notice that the factor-intensity of an industry is always an average value of the production fragments that remain located domestically.³¹

International factor mobility is excluded in the H-O set-up. With regard to the constantly deepening European integration process this assumption is clearly at odds with reality. In particular, capital mobility is advanced in the European context and also labor migration is possible.³² Moreover, while denying movements of factors across borders, factor mobility is assumed to be perfect between sectors within a country in the H-O framework. However, in reality it can be observed that many factors are - to a certain degree - bound to their actual use and are thus sector-specific.³³ Accordingly, inter-sectoral mobility of factors might only come into effect after costly adjustments.

The H-O model necessarily demands unequal factor-endowments between countries together with diverging factor-proportions to initiate trade. This assumption can be problematic in the context of outsourcing to Eastern Europe because, first, it is not per se given that the Western European countries have a higher ratio of high-skilled workers. This aspect will be further explored in section 3 of the paper. Second, in the light of cross-border factor-mobility devel- opment and convergence processes are enabled that potentially alter the proportions of factor- endowments in Eastern and Western European countries in a dynamic perspective. In particu- lar, the speed of capital accumulation in the Eastern European countries, for instance through foreign direct investments, has increased as the integration with the EU proceeds.³⁴ Moreover, Wood³⁵ points out that “alterations in relative world supplies of skilled and unskilled labor” potentially act as external force to changing the equilibrium factor prices under an H-O re- gime.

2.4 New trade models focusing on outsourcing

In reaction to the outlined deficiencies of the H-O assumptions with regard to explaining the outsourcing-related trade effects several modified frameworks have been developed to evalu- ate the effects of outsourcing. Roughly, the newly developed frameworks can be segmented into two groups: (i) those that predict a worsening situation for blue-collar workers in the sending country (same result as in the H-O model), and (ii) those that state that low-skilled workers will benefit from outsourcing. Furthermore some models derive predictions that are not clear-cut and condition their results on different case distinctions. In the following, the main tendencies in the literature on the topic will be highlighted by introducing a representa- tive model for each category. Emphasis is put on the central assumptions driving the different model outcomes.

2.4.1 Feenstra and Hanson (1997)

A first approach in the literature to the explicit modeling of outsourcing was developed by Feenstra and Hanson in the context of US firms outsourcing unskilled-labor intensive produc- tion fragments to Mexico.³⁶ They consider a two-country (denoted as North and South) world- economy where there is only one final good (Y) . Therefore their model can be linked to fac- tor-biased technological change since it excludes any cross-sector movements of factors and allows for analyzing intra-industry trade in isolation.³⁷ Both countries are endowed with capital (K_i) and skilled (H_i) as well as unskilled (L_i) labor. Moreover, it is assumed that factor endowments differ to an extent where no complete factor-price equalization takes place.³⁸ The endowments are such that the South enjoys a higher relative return to capital and skilled labor in reflection of the relative scarcity of these factors.³⁹ Labor mobility between skill categories is allowed so that supply of skilled and unskilled labor becomes elastic and reacts to changes in relative wages.⁴⁰ This reflects the above stated critique regarding the possibility of changing factor-endowments in a dynamic perspective. International factor mobility, however, is excluded in a first step.

The production of the (single) good (Y ) takes place using a continuous range of intermediate inputs represented by z ∈ [0,1]. These inputs have different requirements of skilled and un- skilled labor: A unit of z demands α H (z) units of skilled labor and α L (z) units of unskilled labor.⁴¹ The resulting unit costs of production are expressed by [Abbildung in dieser Leseprobe nicht enthalten]where B is a constant. Figure 1 displays this framework:

Figure 1: Minimum cost combinations⁴²

Abbildung in dieser Leseprobe nicht enthalten

The minimum cost combinations are indicated by the[Abbildung in dieser Leseprobe nicht enthalten]lines in the graph. It can be seen that[Abbildung in dieser Leseprobe nicht enthalten]is below[Abbildung in dieser Leseprobe nicht enthalten]in the area where production requires relatively more unskilled labor as input. Therefore the point[Abbildung in dieser Leseprobe nicht enthalten]reflects the trading equilibrium up to which the South can benefit from its comparative advantage in low-skilled labor-intensive inputs. The result- ing intra-product specialization implies that the North outsources those parts to the South that have a skill-intensity below[Abbildung in dieser Leseprobe nicht enthalten]. This creates a relatively higher demand for skilled labor in both countries. The reason for this is that outsourcing alters the skill-intensity in both the domestic and foreign economy. Key to this argumentation is the consideration of the average skill-intensity of production after outsourcing has taken place. For instance, in the North average skill-intensity rises since the low-skilled parts from the lower end of the production range are transferred abroad. This is comparable in its effects to factor biased technological progress on the remaining domestic fragments. In turn, it leads to an increased relative de- mand for high-skilled labor along with increased wages in the North.

If in a next step the restriction on international capital mobility is relaxed, firms in the North will have an incentive to invest in the South, as capital rents are higher there. This relaxation of capital immobility approaches the reality in terms of Eastern European integration.⁴³ An immediate effect is the reduction of relative capital rents in the South, as capital scarcity is diminished. Holding wages constant, the curves indicating the minimum cost loci of production are shifted as depicted in Figure 1 by the dashed lines.⁴⁴ This leads to an outward shift of the optimal value[Abbildung in dieser Leseprobe nicht enthalten] so that the area of outsourced inputs to the South is pushed towards higher skill-intensities. International capital mobility thus enhances the positive effect of outsourcing on skilled-labor wages in both involved countries.

The model by Feenstra and Hanson supports the derived predictions of the traditional H-O framework with regard to outsourcing: Production relocation across national borders is at the expense of unskilled workers in the Western sending country. Capital mobility additionally worsens their situation. The inclusion of different skill-levels is helpful in analyzing labor market effects. However, this result is mainly driven by the one-sector property of the model, which is certainly not a valid assumption in the case of outsourcing to Eastern Europe, as it impacts several different sectors.

2.4.2 Arndt (1997)

In his framework for the analysis of outsourcing impacts Arndt depicts a two-countries, two factors (capital (K) and labor (L)) and two-goods (X, Y) situation, which reflects the H-O set- up.⁴⁵ In contrast to Feenstra and Hanson, Arndt models an economy with two sectors, which is similar to analyzing sector-biased technological progress.⁴⁶ The countries are small, which translates into a price-taker position in the world market for final goods. Arndt assumes fur- ther that sector X is relatively labor-intensive, while Y is capital-intensive. These intensities, however, are only average figures, as the single stages of the production process differ in their individual factor-intensities. This can be graphically clarified by looking at Figure 2, where the average expansion path of Industry X is split-up into two productions stages[Abbildung in dieser Leseprobe nicht enthalten]that display inverse relative factor-intensities:

Abbildung in dieser Leseprobe nicht enthalten

Figure 2: Production stages within a sector⁴⁷

The effects of outsourcing can now by evaluated with the help of Figure 3.⁴⁸

Abbildung in dieser Leseprobe nicht enthalten

Figure 3: Outsourcing scenario⁴⁹

Starting from the non-outsourcing state with expansion paths [Abbildung in dieser Leseprobe nicht enthalten]and[Abbildung in dieser Leseprobe nicht enthalten] and the accompa- nying relative wage ratio [Abbildung in dieser Leseprobe nicht enthalten]it is assumed that labor-intensive stage[Abbildung in dieser Leseprobe nicht enthalten]of sector[Abbildung in dieser Leseprobe nicht enthalten] is out-sourced, as its production is cheaper abroad. Thus only stage[Abbildung in dieser Leseprobe nicht enthalten]stays domestic.⁵⁰ The domestic factor-intensity of sector X is then determined by[Abbildung in dieser Leseprobe nicht enthalten] alone. At unchanged final prices for good X (tenable due to the small country assumption) the production costs are lowered, which is represented by an inward shift of the X -isoquant to X ' . This constellation induces producers to increase the output of X. The increased output under outsourcing, however, intensifies relative labor demand because sector X is still more labor-intensive than sec- tor Y although its labor-intensity has been reduced through outsourcing. Consequently, the ratio of factor prices changes to [Abbildung in dieser Leseprobe nicht enthalten]. An additional effect is realized, as the relatively higher wage rates provide an incentive to producers to substitute capital for labor, which leads to the dashed expansion paths for[Abbildung in dieser Leseprobe nicht enthalten]and[Abbildung in dieser Leseprobe nicht enthalten] . It turns out that outsourcing has a positive effect on 1 wages throughout the whole economy (of the sending country). At the same time also em- ployment is expanded. To show this, Figure 4 outlines the production points before [Abbildung in dieser Leseprobe nicht enthalten]and after outsourcing has been realized [Abbildung in dieser Leseprobe nicht enthalten]):

Abbildung in dieser Leseprobe nicht enthalten

Figure 4: Production points before and after outsourcing⁵¹

It is observable that output in sector X increases whereas sector Y shrinks. By comparing the distances on the axes it can be concluded that the expansion of sector X requires more inputs (capital and labor) than is freed by the contracting sector Y.⁵²

Hence the model acknowledges that the factor labor relatively gains from outsourcing in the sending country. So far, though, labor has not been differentiated according to skillcategories. If the equilibrium mechanism of the model is run with skilled and unskilled labor as factor inputs - where skilled labor substitutes for capital - it can be inferred that outsourcing is beneficial for unskilled workers.⁵³ This derivation requires inter-sectoral mobility of labor so that factor-specificities are not taken into account. Therefore the framework by Arndt predicts the exact opposite effects of the previous models, i.e. that skilled workers lose with outsourcing while their low-skilled colleagues profit from it.⁵⁴

2.4.3 Kohler (2001)

Kohler’s model examines a small open economy with two sectors and given prices for final goods due to perfect competition.⁵⁵ Sector 1 is subject to fragmentation (two distinguishable fragments) while sector 2 cannot be disaggregated. Both sectors employ technologies with constant returns to scale and linearly homogeno[Abbildung in dieser Leseprobe nicht enthalten]us production functions. The economy dis- poses of a constant endowment of two production factors, capital (K)and labor (L) , of which capital is sector-specific ([Abbildung in dieser Leseprobe nicht enthalten]) whereas labor is equally usable in both sectors. The production function for sector 1 can be written as:^{56 57}

Abbildung in dieser Leseprobe nicht enthalten

[Abbildung in dieser Leseprobe nicht enthalten]is the fragmentation technology and[Abbildung in dieser Leseprobe nicht enthalten]stand for the amounts of the two compo-nents that enter final assembly under fragmentation. Assuming capital in sector 1 is mobile across fragments factor constraints are[Abbildung in dieser Leseprobe nicht enthalten]Profit- 1 2 1 2 maximization then leads to the first-order conditions where p is the relative price of good 1:

Abbildung in dieser Leseprobe nicht enthalten

These conditions state that the relative wages and capital rents have to be equal to the mar- ginal value productivities of each fragment in a sector. Because of the assumed linear homo- geneity, the production function of sector 1 may - with reference to the relative price of good 1 - also be formulated as:

Abbildung in dieser Leseprobe nicht enthalten

The expressions [Abbildung in dieser Leseprobe nicht enthalten] and[Abbildung in dieser Leseprobe nicht enthalten]can be interpreted as effective prices of the fragments since they reflect the value contribution of each of the fragments to the final revenue achieved with good 1. The effective prices are denominated by [Abbildung in dieser Leseprobe nicht enthalten] respectively. Applying the 1 2 first-order conditions (2) and (3) yields:

Abbildung in dieser Leseprobe nicht enthalten

Therefore the opportunity costs of production - given by relative factor prices and employed factor quantities - have to equal the effective price of a fragment in order to satisfy the zeroprofit-condition of perfect competition.⁵⁸

The outsourcing scenario is now given by a neighboring country with lower wages given by [Abbildung in dieser Leseprobe nicht enthalten].⁵⁹ It is assumed that the only part of domestic production amenable to outsourcing is fragment 2 of industry 1. However, outsourcing to a foreign country is costly. Thus Kohler models fixed costs Z that are expressed in terms of foreign labor and variable costs; these are given by a labor cost surcharge in form of the multiplier [Abbildung in dieser Leseprobe nicht enthalten] . Foreign direct investment (i.e. sector-specific capital[Abbildung in dieser Leseprobe nicht enthalten]may be invested abroad) is allowed here. In this environment ~ 1 the effective value of fragment 2, p F, may be procured at costs below the initial domestic 2 2 opportunity costs given by (7). For a single firm the costs savings potential is expressed by [Abbildung in dieser Leseprobe nicht enthalten] So, if this cost advantage is positive firms have an incentive to out- source fragment 2 to the low-wage country.

The outsourcing case cannot maintain the equilibrium values of factor prices and employed quantities that have initially been derived for the domestic economy. In particular, firms will increase production of fragment 2 as long as its procurement costs are lower than its effective price. This also implies a production expansion of fragment 1 (this is due to the linearly- homogenous production function). Keeping everything else constant, outsourcing of fragment 2 would lead to an excess supply of labor at home. In response, the domestic wage rate will be driven down to the foreign wage level adjusted for outsourcing costs (effective foreign wage rate). This leads to the modified equilibrium condition:

Abbildung in dieser Leseprobe nicht enthalten

The effective foreign wage level applies to the whole domestic economy since capital and labor are substitutes. Labor market clearing conditions have to be restated as:

Abbildung in dieser Leseprobe nicht enthalten

Hence, as can be seen from (10), sector 2 as well as the production of fragment 1 of sector 1 completely rely on domestic labor - however, their sum not necessarily adds up to the domes- tic labor supply. Equilibrium is restored if the remaining workers are placed within domestic production of fragment 2. Moreover, condition (11) expresses that the labor demand for frag- ment 2 in excess of domestic supply has to be satisfied by using foreign labor. Thus the equi- librium amount of outsourcing is determined in terms of foreign labor use. Those labor market equilibrium conditions imply that outsourcing is not necessarily complete in the sense of transferring a whole sector abroad so that only partial outsourcing might be realized - this is bound to the model-inherent demand of domestic labor market equilibrium. Due to the linearly-homogenous production technology the necessary level of accompanying foreign direct investment can be expressed as:

Abbildung in dieser Leseprobe nicht enthalten

Hence the equilibrium level of investment is determined by the ratio between foreign labor in fragment 2 and (purely domestic) labor in fragment 1 and the capital employment in fragment 1. Equality between the marginal value products in both fragments and the capital rent is ensured. This equilibrium is stable, as capital mobility always works to equalize the marginal value productivities in all fragments.⁶⁰ Potential outsourcing of fragment 2 has driven down the domestic wage level and at the same time increased the rent to the sector-specific capital in sector 2. It has to be highlighted that factor-intensities did not have an influence on the presented results.

Kohler’s model shows in the case with foreign direct investment, i.e. the more appropriate case for the considered context, that labor loses from outsourcing to the advantage of sector- specific capital. Capital mobility is the important lever to ensure this outcome. The essential improvement of the model is represented by incorporating sector-specificities that are also observable in reality. As criticized in Arndt’s model before, labor is considered to be a non- differentiated factor. However, when interpreting skilled labor as the sector-specific factor it can be concluded that outsourcing deteriorates the situation of unskilled workers in the send- ing country.⁶¹ This result, given capital mobility, resembles the predictions reached by H-O as well as Feenstra and Hanson, as shown above.

[...]

---

¹ Compare Marin (2004), p. 4.

² Horn/Behncke (2004), p. 1.

³ Compare Wood (1995); Arndt (1997); Abraham/Konings (1999); Feenstra/Hanson (1999); Kohler (2000); Anderton/Brenton/Oscarsson (2002); Egger/Stehrer (2003).

⁴ Many participants in the discussion observe an increasing degree of integration between countries and declin- ing wages or unemployment of unskilled workers in recent years. However, the parallel existence of both phe- nomena does not necessarily reflect a causal relationship, which is often automatically assumed in the discussion.

⁵ Defined here as “workers with no more than a basic education” (Wood (1995), p. 58).

⁶ This restriction is motivated by the observation that there is a large consensus in the literature about the (positive) welfare effects of outsourcing. In contrast, the effects on factor prices are interpreted ambiguously.

⁷ Other terms frequently used in the literature: “slicing-up of the value-added chain” (Krugman (1995)), “vertical trade” (Jones 1996), “fragmentation” (Deardorff (1998)).

⁸ When referring to the international dimension of outsourcing some authors use the term “offshoring“ (Kirkegaard (2005); Amiti/Wei (2005)).

⁹ Compare Diehl (1999), p. 3; Egger/Stehrer (2003), p. 61.

¹⁰ Compare Kohler (2001), p. 38.

¹¹ Compare Marin (2004), p. 7.

¹² Compare Jones/Kierzkowski/Lurong (2004), p. 10.

¹³ Compare Kohler (2001), p. 32.

¹⁴ Note the inherent assumption that final assembly is done domestically, which appears reasonable in case the home country offers some specific advantages with regard to final assembly (i.e. know-how).

¹⁵ Compare Markusen/Venables (2000), p. 210.

¹⁶ This assumption reflects a conventional point of departure in the literature. However, in the next section of this paper the factor-endowments of Western and Eastern countries will be explored in more detail with the help of empirical findings.

¹⁷ The assumption of a low-skilled workforce, however, requires a qualification. Egger and Stehrer (2003, p. 64) write on this behalf: “Evidence […] suggests that these countries [Eastern Europe] have well educated workforces. But, on the other hand, (i) there seems to be a skill-shortage with respect to Western technologies, and (ii) the capital stock of these countries is relatively outdated. Both of these arguments justify the assumption that the CEE countries are relatively unskilled labour abundant.”

¹⁸ It should be highlighted that cost advantages cannot correctly be measured by only looking at wages. Instead labor productivity has to be taken into consideration. Accordingly, unit costs are the relevant variable for comparison. Marin (2004, p. 8), for example, quotes evidence that CEE affiliates of German firms operate at a unit cost level of 27.6 % compared to domestic production.

¹⁹ Compare Heckscher/Ohlin (1991).

²⁰ The following overview of neoclassical trade theory, in particular the Heckscher-Ohlin model, is inspired by Rose/Sauernheimer (1999).

²¹ Compare Ricardo (1817).

²² These conditions include among others incomplete specialization in the involved countries, clear separability of capital- and labor-intensive sectors, and not too extremely divergent factor proportions in the countries.

²³ The consideration of different qualities within one factor-class can be related back to Leontief (1953) who found in an empirical study that exports of the (capital-rich) US were labor-intensive when regarding human capital. The integration of this “paradox” in the H-O framework leads to the distinction of labor in a skilled and an unskilled component.

²⁴ See Anderton et al. (2002, p. 2): “However, the impact of globalization appears to be more complicated than is allowed for within the confines of standard factor proportions trade theory.”

²⁵ Diehl (1999), p. 2.

²⁶ Wood (1995), p. 62.

²⁷ Anderton et al. (2002), p. 21.

²⁸ Studies dealing with this issue among others: Berman/Bound/Machin (1998); Krugman (2000); Anderton/Brenton/Oscarsson (2002).

²⁹ Anderton/Brenton/Oscarsson (2002), p. 1.

³⁰ Compare Diehl (1999), p. 3.

³¹ Compare Arndt (1997), p. 72.

³² Compare Abraham/Konings (1999), p. 587.

³³ Compare Kohler (2001), p. 35.

³⁴ Compare Eckert/Rossmeissl (2005), p. 56.

³⁵ Wood (1995), p. 59.

³⁶ Compare Feenstra/Hanson (1997).

³⁷ Compare Egger/Stehrer (2003), p. 61.

³⁸ It is stipulated in the H-O model that factor price equalization only occurs if factor-endowment do not extremely diverge between countries.

³⁹ If the return to capital is termed r and the return to skilled and unskilled labor are termed q and w, respec- tively, this statement can be expressed algebraically as follows: r > r; q S q N S N > w S w N

⁴⁰ Higher education levels can transform unskilled workers into skilled employees in the medium- to long-run.

⁴¹ Note in addition that the ratio α H (z) is increasing with z. α L (z)

⁴² Source: Feenstra/Hanson (1997), p. 375.

⁴³ In particular the increasing levels of foreign direct investment should be considered in this regard.

⁴⁴ The effects on wages are not taken into account in order to clarify the point. However, if they were taken into consideration wage adjustments would change the slopes of the cost curves.

⁴⁵ Compare Arndt (1997).

⁴⁶ Compare Egger/Stehrer (2003), p. 62.

⁴⁷ Source: Arndt (1997), p. 72.

⁴⁸ The subsequent analysis is done from the point of view of a relatively capital-abundant country that faces import competition in labor-intensive goods.

⁴⁹ Source: Arndt (1997), p. 74.

⁵⁰ By assumption, stage x contains the final assembly since otherwise re-importation of intermediate inputs 1 would become obsolete.

⁵¹ Source: Arndt (1997), p. 74.

⁵² Specifically, that is: Regarding the[Abbildung in dieser Leseprobe nicht enthalten]. For the L -axis the constellation is [[Abbildung in dieser Leseprobe nicht enthalten]In both cases the X, Y -coordinates are measured according to the relevant axis.

⁵³ Compare Egger/Stehrer (2003), p. 62.

⁵⁴ The following presentation only considers the case in the model where foreign direct investment is allowed, as this scenario is deemed most realistic for the outsourcing to Eastern European countries.

⁵⁵ Compare Kohler (2001).

⁵⁶ The numbering of the formulas here corresponds to the numbering in Kohler (2001), p. 36.

⁵⁷ For sector 2 the production function is given by

⁵⁸ Note that the effective price of a fragment is contingent on the input quantities used in other fragments so that the effective prices have to be interpreted as derived values here - they do not exist independently from each other.

⁵⁹ The superscript f indicates the foreign country and * refers to the initial equilibrium in the domestic setting without outsourcing.

⁶⁰ In a scenario where there is a difference in marginal value productivities of labor between fragments, capital will move internationally to increase the capital-intensity of the less labor-productive fragment. This process leads to a convergence of marginal value productivities of both labor and capital in both countries.

⁶¹ Compare Egger/Stehrer (2003), p. 62.