Journal of Industry, Competition and Trade

, Volume 15, Issue 1, pp 69–85 | Cite as

Slicing Up Global Value Chains: a Micro View

Open Access
Article

Abstract

Global value chains, GVCs, have had a transformative impact on the world economy since the early 1990s. We study 45 specific GVCs with company-confidential invoice-level data. We find that the case companies’ headquartering functions capture a large share of the overall value added, 27 % on average. The value added shares of other functions are as follows: distribution 21 %, final assembly 16 %, and logistics 5 %. The remaining 30 % of the value added goes to vendors. Upon considering value added by country, we find that the home economy’s share is 47 % on average. This share is reduced with offshored—as opposed to Finnish—final assembly: 2 percentage points for a high-end smartphone, over ten percentage points for a low-end feature phone, and 27 percentage points for machinery and metal products. We attribute the latter large drop to co-location of non-assembly functions, intellectual property issues, transfer pricing, and profit allocation. We conclude that GVCs are complex and heterogeneous; value chains of basic products and services are not nearly as global as those of advanced ones; and the value added share of wholesaling and retail is large in consumer products. We nevertheless argue that value added is less tied to assembly—and other tangible aspects of GVCs—than conventional wisdom suggests; the intangible aspects—market and internal services, and creation and appropriation of intellectual property—are more important. The increasing presence of GVCs brings about several thorny policy issues that are yet to be addressed.

Keywords

Global value chains Case studies Finland Transfer pricing Economic policy 

JEL Classification

D23 F23 L14 M11 

1 Introduction

On 22 September 1996, the Los Angeles Times provided a shocking piece of news: only 3.5 % of the $10 overall value added of a “Tea Party” Barbie doll (Mattel Inc.) remained in the location suggested by the “made in…” label of mainland China (Tempest 1996). Subsequently, this line of inquiry has largely concentrated on high-end electronics (see, e.g., Ali-Yrkkö et al. 2011; Dedrick et al. 2010; Linden et al. 2011, 2009), although other products have been analyzed as well (for reviews, see Kenney 2012; Sturgeon et al. 2013).

Since the Barbie incident, scholarly discussions have been significant in disciplines ranging from economics and business studies to geography and sociology. The impact of global value chains (GVCs) has been transformative. Cattaneo et al. (2010, p. 7) note that “… GVCs have become the world economy’s backbone and central nervous system.” Backer and Miroudot (2013) note that over half of global trade currently consists of transactions in the context of GVCs and yet, this perspective has not been fully recognized in the public debates about labor market and trade policy, as many politicians and even experts still presume that goods and services are produced in one well-defined location. In reality, however, all but the most basic products are an assemblage of inputs sourced from around the world.

Timmer et al. (2014) provided a macro-view of the “made in the world” phenomenon and used the World Input–Output Tables (Dietzenbacher et al. 2013) to document the rapid fragmentation of production and increasing value-added share being captured by the owners of capital and highly skilled labor since the early 1990s. In addition, these researchers also noted that the developed countries increasingly specialize in functions carried out by highly skilled workers and that the developing countries have increased their specialization in capital-intensive functions. Our current work is a micro-equivalent of the empirical work performed by Timmer et al. (2014); whereas they observed the aggregate outcomes of GVCs operations over time, we observe the operation of specific GVCs for individual products with actual internal firm data; a feature that none of the other “bottom-up” studies has. These macro and micro approaches are complementary and yield similar conclusions. However, many important features—e.g., the role of transfer pricing—are lost with the macro approach.

The increasing geographic and organizational dispersion of production raises the question of where the value added is created and by whom. Our approach can provide insight into this question. We derive the value added by the companies within GVCs, by the various business functions involved (e.g., assembly and logistics), and the geographical locations of the functions (for example, the value added share of distribution largely resides in the country of final sale).

Our analysis is informed by 45 case studies of GVCs, each of which uses company-confidential invoice-level data with respect to specific products and services. This is the largest, most rigorous (in terms of using actual data), and most diverse (in terms of the industries covered) set of GVC case studies conducted to date. To highlight the rigorousness of the data, we are able to see not only where profits are taken in the internal trade of these firms, but also logistics and inventory costs—a significant advantage over other studies that simply tear down a product, estimate its component costs and allocate profit to the firm’s headquarters. Obviously, for a GVC of a physical product, logistics and inventory are serious issues and costs.

Our GVC cases show considerable variation across dimensions within and especially between industries. Thus, current discussions, which are largely based on generalizations, often do not appreciate the subtleties of the real world. Moreover, most of these generalizations are based on a few iconic case studies and do not capture the realities in understudied sectors such as services and raw material intense process industries. While all of the attention has gone to GVCs, many value chains are still best characterized as regional or even local. With these caveats, we nevertheless conclude that value added is dominated by the intangible aspects of GVCs, i.e., market and internal services, as well as the creation and appropriation of intellectual property. Value added shows a tendency to migrate to either the earlier or later stages of the value chain at the expense of assembly/processing or service provision toward the middle (akin to Shih’s smiling curve, http://v.gd/g3EJcd). The high value-added functions still tend to be performed by highly skilled labor located in the developed countries, even though the role of the developing countries is on the rise in this respect.

Although the GVC-induced shift in policy thinking is rather profound, many previously cherished principles of good policy conduct remain intact. However, in this work as well, the devil is in the details, and micro-level and practical considerations reveal thorny issues. For example, although laissez-faire is the applicable rule-of-thumb and although there appears to be little scope for active or targeted enterprise policies (in the developed countries), we identify situations in which a nation-state might seek to apply more hands-on policies. Furthermore, the prevailing principles of good conduct are at times directly in doubt, e.g., with respect to competition policy. On a few occasions, we observed an increased need for certain policy measures, i.e., social transfers, but less scope with which to implement them.

2 Context

The rise of GVCs has been fueled by technological progress, labor cost arbitrage, access to resources and markets, trade policy reforms (Backer and Miroudot 2013), and—more recently—a relocation of important parts of the supplier base in industries such as electronics since the mid-1980s. The GVCs may be viewed as real-time flows of information, material, capital, and labor inputs. These flows require instantaneous transfer of instructions, quick and cheap movement of intermediate inputs and final outputs, and a modularity of functions that do not occur within one organizational structure in a specific location (Grossman and Rossi-Hansberg 2008). The operation of GVCs benefits from reasonably coherent contractual, governance, and legal principles, which are shaped by national policies, in the home and host locations of multinational enterprises (MNEs). Neilson et al. (2014, p. 3) note that, while “… explicit theorization of the state’s role has been somewhat lacking in the GVC [literature]…”, it is nevertheless central: “… state action and inaction creates the enabling conditions that shape whether and how firms, regions and nations are able to engage with global markets, and their capacities to upgrade these engagements.”

The organization of GVCs is primarily driven by the MNEs’ expected net present value of corporate profits, which involves operational efficiency, cost arbitrage, and the ability to repatriate returns. On one hand, MNEs are forced to address often conflicting demands from the national authorities in different countries; on the other hand, they often play countries or localities against one another to obtain lower tax rates, higher investment subsidies, or other benefits.

The GVCs enable deepening specialization, which (in principle) leads to global welfare gains. However, it remains unclear how these gains are distributed across countries, among businesses, and between individuals within countries.

3 Data

Our empirical material consists of 45 in-depth case studies conducted in 2008–2014. All of the cases involve a specific product or a service offered by a particular company, i.e., the brand holder (or the coordinator of the GVC), at a certain point in time. The brand holder companies are headquartered in Finland—with three exceptions, in which cases the headquarters are located in other high-income countries. Each value chain spans multiple countries. In 15 cases, we observe the same company coordinating multiple but functionally overlapping GVCs that churn identical products or services typically for geographically distinct markets. The names of these four case companies are declassified—Helkama, a bicycle manufacturer; Nokia, still manufacturing mobile phones at the time; Ponsse, a provider of forest harvesting equipment; and WhiteVector, social media intelligence services—but for the most part, the case companies must remain anonymous and we discuss the particular product/service in rather general terms.

Data are collected from both primary and secondary sources. Primary sources consist of semi-structured workshops with executives of the case companies (typically involving the chief financial officer, the senior vice president of supply chain management, a business unit controller, a purchasing director, and/or a product designer/engineer; 3–8 workshops per case, each lasting 2–4 h) as well as interviews with various GVC stakeholders (six interviews per case on average, each lasting 1–2 h). The process was iterative because face-to-face meetings invariably required elaborations via email and telephone. Most of the work was carried out in (or from) Finland, but extensive field work was also conducted in China. Typically, the management of each brand holder company contributed 40–80 h of its time to provide sufficient data.

Secondary sources were used in all cases, but they were of particular importance in the case of mobile phones as well as in tracing cost structures and other information on higher-tier vendor companies. Secondary sources involve the following (cf. Ali-Yrkkö et al. 2011): physical reverse engineering of the product; publicly and privately (personal contacts) available information on, e.g., industry profit margins; commercial “teardown” reports in the case of electronics (e.g., Portelligent 2007); interviews with industry experts; and financial reports and press releases from the companies involved and those of their direct competitors (differences in reporting in various geographies and officially required further information, e.g., 20-F reports in the United States, are particularly useful).

4 Definitions

In our terminology, a value chain is composed of the entire range of activities involved in providing a product or a service. Global value chains (GVCs) span multiple countries. These activities include both arm’s-length relationships and exchanges within an enterprise and begin with conceptualization and design, evolve to provision and distribution, and end only when the product or service in question is completely consumed and disposed (although, for the most part, our analysis ends with the final delivery of the product or service). Although the discussion on value chains is somewhat more concerned with immediate flows of information, variable inputs, and goods, we also account for equally important tangible and intangible capital and fixed inputs.

Each function or link in the value chain—performed either in-house by the brand holder or purchased from an outside vendor—conducts its own value-added activities and forwards its output to the next phase. The starting point of the analysis is to identify each of these links in the value chain. We follow through by defining the value added of each link with an eye on performing companies (profits, ownership of and compensation for tangible and intangible capital), individuals (labor income) as well as host locations (to determine national outcomes). We observe all the relevant details within the case company but are mostly forced to extrapolate product-specific profit margins of vendor companies.

We employ the conventional measure of value added, although we are forced to define it at a fine level of aggregation. Value added is the difference between gross output and intermediate consumption at each link. Thus, value added is equal to the net output of each link and reflects the amounts that employed labor and capital inputs jointly contribute. We proxy the value added with the sum of the following three elements: wages and other labor income, operating profits (if unobserved at a given link, we would apply a typical industry rate of profit that accounts for the type and the risk profile of the function), and costs attributable to the employed capital inputs (the depreciations of the performing companies’ own tangible and intangible capital stocks as well as costs of other capital inputs, e.g., rents of capital equipment and licensing fees).

Throughout our analysis, we record the actions that actually take place in the GVC. Thus, our calculations reflect the transfer pricing practices employed by the involved corporate entities and also reflect other complexities, e.g., inability to repatriate profits from China in certain cases—despite its legitimacy in light of the internationally accepted guidelines (OECD 2010).

Because we employ the conventional measure of value added and are able to re-aggregate the value added in the context of each product/service by country, our measures correspond directly to gross domestic product because GDP may be understood as the sum of value added within national borders.1 Thus, our calculations directly indicate what goes into the actual measurement of GDP (by country/year) for each product/service.

5 Analysis

Throughout this section, the tables summarize certain standard aspects of our case studies, and in the text, we highlight some dimensions. Although the 45 cases extend over five tables, we assign a running number to each case (the numbers of the cases with more than one geographic incarnation are followed by an identifying letter).

In Tables 1, 2, 3 and 4, the case company’s value-added share is on average 41 % (for services in Table 5, the corresponding share is 60 %).2 If we exclude the value added shares of any in-house assembly, logistics, and distribution, the case company’s broadly understood headquartering functions capture 27 % of the value added. Vendors of raw materials and intermediate inputs—including all goods and certain services (other than any outsourced assembly, logistics, and distribution)—capture 30 % of the value added (obviously the vendors’ value added has its own functional structure). The remaining value added shares are: distribution 21 %, final assembly 16 %, and logistics 5 %. As we will show, however, these averages conceal huge variation.
Table 1

GVC case studies: electronics

   

By company/function

By location

   

Case

Vend.

Assembly

Logistics

Distrib.

Fi

EU*

US*

Asia

Oth.

No

Description

Year

%

%

Loc.

Org.

%

Org.

%

Org.

%

%

%

%

%

%

1a

Nokia N95

2007

49

36

Fi

In

2

Ex

0.0

Ex

15

41

27

14

14

5

1b

Nokia N95

2007

49

36

Lo

In

2

Ex

0.0

Ex

15

39

12

28

16

5

2a

Nokia 3310

2000

43

40

Fi

In

9

Ex

2.5

Ex

15

39

33

8

14

6

2b

Nokia 3310

2000

43

40

Hi

In

9

Ex

2.5

Ex

15

26

46

8

14

6

2c

Nokia 3310

2000

43

40

Lo

In

9

Ex

2.5

Ex

15

26

33

8

27

6

3a

Nokia 1100

2003

43

38

Fi

In

9

Ex

4.9

Ex

15

41

33

7

13

6

3b

Nokia 1100

2003

43

38

Hi

In

9

Ex

4.9

Ex

15

21

49

7

17

6

3c

Nokia 1100

2003

43

38

Lo

In

9

Ex

4.9

Ex

15

21

33

7

34

6

4a

Nokia 1200

2007

21

54

Lo

In

5

Ex

6.4

Ex

18

8

46

10

26

9

4b

Nokia 1200

2007

21

54

Lo

In

5

Ex

6.4

Ex

18

8

37

10

36

9

5

Device, low

2011

62

15

Fi

In

15

Ex

0.3

In

23

17

4

4

3

3

6

Device, mid l.

2011

67

10

Fi

In

14

Ex

0.1

In

23

16

3

1

1

2

7

Device, mid h.

2011

61

12

Lo

Ex

6

Ex

0.2

In

23

6

2

6

6

2

8

Device, high

2011

52

8

Fi

In

4

Ex

0.1

In

40

5

1

4

1

1

Average

  

46

33

  

8

 

2.6

 

19

22

26

9

16

5

Average, Ass. Fi

  

53

24

Fi

 

9

 

1.3

 

22

26

17

6

8

4

Average, Ass. Lo

  

40

39

Lo

 

6

 

3.4

 

17

18

27

12

24

6

Notes: The case company’s value-added share (“Case”) includes in-house assembly/logistics/distribution; accounting for these categories, the shares under “By company/function” add up to 100 % (save for rounding). The column “Vend.” includes all vendors and vendors of vendors, which is composed of all purchased parts and components as well as services (excluding assembly/logistics/distribution). Assembly reflects its direct costs, but the geographical breakdown also reflects the remainder of the value added (profits, etc.) attributable to it. The alternative locations (“Loc.”) are: Finland (“Fi”), another high cost country (“Hi”, e.g., Germany), and certain low cost locations (“Lo”, mostly China). The method of organizing a specific function (“Org.”) refers to either in-house provision (“In”) or provision by another company either through outsourcing or via a market transaction (“Ex”). Logistics consists of inbound and outbound transportation. Distribution refers to wholesaling and retailing (or comparable functions). In the “By location” geographical breakdown, “EU*” refers to the EU-27 excluding Finland, “US*” refers to North America (here dominated by the United States), “Asia” to the continent, and “Oth.” to the rest of the world; the columns under “By location” add up to 100 % (save for rounding). The “Average” below the horizontal line refer to the arithmetic mean of the numbers above it; the “Average, Ass. Fi” refers to the cases in which the final assembly is located in Finland and to the “Average, Ass. Lo” to the cases in which the final assembly is in a low-cost country. Case notes. 2a–4b: Sale in EU-27 outside Finland (logistics include warranty costs); 2a–3c: Year of market introduction (the Nokia 3310 was analyzed in 2003 and the Nokia 1100 in 2004 (these GVCs were not significantly adjusted in the course of the respective models’ life cycles); 5–8: Logistics excludes inbound; 5–7: Wholesaling by the case company, external retailing

Sources: Case 1a–b, Ali-Yrkkö et al. (2011); Cases 2a–4b, Seppälä and Ali-Yrkkö (2014); Cases 5–8 Ali-Yrkkö (2013)

Table 2

GVC case studies: machinery and metal products

   

By company/function

By location

   

Case

Vend.

Assembly

Logistics

Distrib.

Fi

EU*

US*

Asia

Oth.

No

Description

Year

%

%

Loc.

Org.

%

Org.

%

Org.

%

%

%

%

%

%

9a

Bicycle

2011

36

33

Fi

In

16

Ex

5.0

Ex

25

67

13

5

11

5

9b

Bicycle

2011

28

42

Lo

Ex

2

Ex

3.0

Ex

25

58

21

5

11

5

9c

Bicycle

2011

27

37

Lo

Ex

2

Ex

9.0

Ex

25

59

9

5

21

5

10a

Drilling equip.

2011

44

51

Fi

In

23

Ex

4.0

In

0

54

14

19

2

11

10b

Drilling equip.

2011

47

46

Lo

In

17

Ex

7.6

In

0

26

15

2

42

14

11a

Filter system, hi

2011

27

65

Fi

In

6

Ex

4.0

Ex

5

32

51

6

6

6

11b

Filter system, hi

2011

21

70

Lo

In

1

Ex

4.0

Ex

5

22

53

6

14

6

12a

Filter system, lo

2011

34

56

Fi

In

14

Ex

5.0

Ex

5

40

44

5

6

5

12b

Filter system, lo

2011

22

67

Lo

In

2

Ex

6.0

Ex

5

22

50

6

16

6

13a

Handling equip.

2011

64

26

Fi

In

50

Ex

10.0

In

0

64

19

2

11

4

13b

Handling equip.

2011

50

34

Hi

In

38

Ex

16.0

In

0

19

8

50

18

5

13c

Handling equip.

2011

75

20

Lo

In

58

Ex

5.0

In

0

15

6

0

76

3

14a

Heating eq., hi

2011

37

21

Fi

In

12

Ex

28.0

Ex

14

35

8

10

3

44

14b

Heating eq., hi

2011

42

15

Lo

In

12

Ex

30.0

Ex

13

2

2

46

2

48

15a

Heating eq., lo

2011

43

16

Fi

In

6

Ex

28.0

Ex

14

40

7

7

2

44

15b

Heating eq., lo

2011

45

12

Lo

In

6

Ex

30.0

Ex

13

1

2

47

2

48

16a

Special vehicle

2011

56

41

Fi

In

39

Ex

2.0

In

0

61

9

24

2

4

16b

Special vehicle

2011

57

39

Lo

In

40

Ex

3.9

In

0

45

10

22

14

9

17a

Steel structure

2011

48

46

Fi

In

22

Ex

6.0

In

0

46

33

7

7

7

17b

Steel structure

2011

48

38

Lo

In

19

Ex

6.4

In

0

5

73

7

7

7

18a

Paper mach. part

2010

54

37

Fi

Ex

1

Ex

1.0

Ex

7

75

19

1

3

1

18b

Paper mach. part

2010

54

37

Hi

Ex

1

Ex

1.0

Ex

7

55

15

4

22

4

19a

Tool tip

2011

52

45

Fi

In

6

Ex

3.0

In

0

90

5

0

5

0

19b

Tool tip

2011

57

41

Hi

In

6

Ex

2.0

In

0

51

0

0

17

32

20

Forest tractor

2011

38

60

Fi

In

10

Ex

2.0

In

7

48

Average

 

44

   

16

 

8.9

 

7

41

20

12

13

13

Average, Ass. Fi

 

44

 

Fi

 

17

 

8.2

 

6

54

20

8

5

12

Average, Ass. Lo

 

41

 

Lo

 

16

 

10.5

 

9

26

24

15

20

15

Notes: For definitions of the column headings, see Table 1. Case notes. 10a–b, 17a–b and 18a–b: The headquartering country is not Finland but rather another high-income country; 15a: Logistics includes customs and tariffs; 17a–b: Assembly consists of only direct labor costs (i.e., excludes depreciation etc.); 20: Due to the confidentiality agreement with Ponsse, the value-added share by location is only reported for Finland

Sources: Case 9a–b, Kalm et al. (2013); Case 18a–b, García (2011); all others, Ali-Yrkkö (2013)

Table 3

GVC case studies: foodstuffs and textiles

   

By company/function

By location

   

Case

Vend.

Assembly

Logistics

Distrib.

Fi

EU*

US*

Asia

Oth.

No

Description

Year

%

%

Loc.

Org.

%

Org.

%

Org.

%

%

%

%

%

%

21

Bread, dark

2011

57

12

Fi

In

36

Ex

6.8

Ex

24

95

4

0

0

0

22

Bread, white

2011

54

11

Fi

In

25

Ex

4.4

Ex

32

97

1

1

1

1

23

Chicken, proc.

2011

37

29

Fi

In

38

Ex

3.2

Ex

30

98

2

0

0

0

24

Cold cut

2011

38

14

Fi

In

18

Ex

1.5

Ex

47

94

6

0

0

0

25

Packaged meal

2011

34

16

Fi

In

15

Ex

2.6

Ex

47

96

4

0

0

0

26

Bag of candy

2011

50

17

Fi

In

18

Ex

1.4

Ex

32

0

0

0

0

0

27

Jar of jam

2011

26

29

Fi

In

16

Ex

0.7

Ex

44

84

16

0

0

0

28a

Chocolate bar

2011

33

20

Fi

In

9

Ex

0.6

Ex

46

88

1

4

4

4

28b

Chocolate bar

2011

19

12

Fi

In

6

Ex

1.1

Ex

66

27

66

2

2

2

29a

Textile, tiny

2011

40

4

Lo

Ex

7

Ex

0.0

Ex

50

90

0

0

10

0

29b

Textile, tiny

2011

40

4

Lo

Ex

7

Ex

0.0

Ex

50

24

66

0

10

0

30a

Textile, small

2011

30

11

Lo

Ex

8

Ex

0.0

Ex

51

81

1

1

19

0

30b

Textile, small

2011

30

11

Lo

Ex

8

Ex

0.0

Ex

51

15

67

1

18

0

31a

Textile, mid-s.

2011

29

11

Lo

Ex

10

Ex

0.0

Ex

50

79

1

1

20

0

31b

Textile, mid-s.

2011

29

11

Lo

Ex

10

Ex

0.0

Ex

50

13

66

1

20

0

32a

Textile, large

2011

29

13

Lo

Ex

9

Ex

0.0

Ex

50

78

1

1

20

0

32b

Textile, large

2011

29

13

Lo

Ex

9

Ex

0.0

Ex

50

13

66

1

20

0

Average

 

36

   

15

 

1.3

 

45

63

22

1

8

0

Average, Sales Fi

 

32

   

9

 

0.1

 

49

83

1

1

14

1

Average, Sales EU*

 

29

   

8

 

0.2

 

53

18

66

1

14

0

Notes: For definitions of the column headings, see Table 1. Case notes. In this table, a and b refer to the country of final sale rather than to the assembly location; a cases refer to the final sale in Finland and b cases to elsewhere in the EU-27. Case 23: Due to losses associated with the product, the value added share of the case company is less than the value added of its internal assembly

Sources: Cases 21, 22, 26, and 28a–b, Langenskiöld (2011); Cases 23–27, Karjalainen (2011); Cases 29a–32b, Alakoski (2012); all others, Ali-Yrkkö (2013)

Table 4

GVC case studies: other products

   

By company/function

By location

   

Case

Vend.

Assembly

Logistics

Distrib.

Fi

EU*

US*

Asia

Oth.

No

Description

Year

%

%

Loc.

Org.

%

Org.

%

Org.

%

%

%

%

%

%

33

Sawn timber

2012

18

47

Fi

In

4

Ex

2.0

Ex

33

100

0

0

0

0

34

Cardboard, lo

2011

26

34

Fi

Ex

40

In

0.0

In

9

89

3

2

2

4

35

Cardboard, hi

2011

32

27

Hi

Ex

40

Ex

0.0

In

10

43

52

1

0

3

36

Tissue roll, lo

2011

29

8

Hi

Ex

49

Ex

0.0

Ex

13

28

65

2

1

3

37

Tissue roll, hi

2011

24

9

Hi

Ex

54

Ex

0.0

Ex

13

22

71

2

2

4

Average

 

26

25

  

37

 

0.4

 

16

56

38

1

1

3

Average, Ass. Fi

 

22

41

Fi

 

22

 

1.0

 

21

95

2

1

1

2

Average, Ass. Lo

 

27

9

Lo

 

52

 

0.0

 

13

25

68

2

2

4

Notes: For definitions of the column headings, see Table 1. Case notes. 33: Logistics include only outbound sales in Finland; 34: Sales in Finland; 35: Sales in the EU-27 outside of Finland; 36–37: Sales in the EU-27 outside of Finland

Sources: Cases 34 –35, Leppänen (2012); Cases 36 – 37, Sortti (2012); all others, Ali-Yrkkö (2013)

Table 5

GVC case studies: services

   

By company

By location

   

Case

Vend.

Fi

EU*

US*

Asia

Oth.

No

Description

Year

%

%

%

%

%

%

%

38

Sw. platform

2013

29

71

61

10

11

10

8

39

Digi serv., B2C

2013

30

70

56

14

10

10

10

40

Digi serv., B2B

2013

81

19

72

19

4

3

3

41

SaaS analytics

2010

93

7

97

1

1

1

1

42a

Logistics, Finl.

2013

60

40

85

0

0

0

15

42b

Logistics, Russia

2013

44

56

2

0

0

0

99

43

Translation s.

2011

77

23

72

9

4

8

8

44

Machine parts

2012

56

44

13

30

22

2

33

45

Rock band

2011

68

32

100

Notes: For definitions of the column headings, see Table 1. Case notes. For discussion regarding 42a and 42b, see text. Case notes. 45: Upon considering the geography of value added, we abstract from fuel, instrument, and some other minor costs

Sources: Cases 38–41, Kalm et al. (2014); Case 43, Serbessa (2012); Case 44, Seppälä and Kalm (2013); Cases 42a–b and 45: unpublished manuscripts by Matias Kalm

In Tables 1, 2, 3, 4 and 5 Finland, the dominant host country, captures on average 47 % of the value added generated globally. In several cases, the location of final assembly changes from Finland to a lower cost location—mostly China—while other aspects of the GVCs remain intact. For a high-end smartphone (Table 1, Case 1a vs 1b), such a change reduced Finnish value added by two percentage points; for a low-end feature phone, the difference is over ten percentage points (Table 1, Case 2a vs 2c or 3a vs 3c). Across the eleven relevant machinery and metal products comparisons in Table 2, offshored final assembly reduces Finnish value added from 8 to 49. On average, the reduction is 27 percentage points.

5.1 Electronics

To illustrate the process of conducting GVC case studies, we first summarize certain aspects of our original Nokia N95 smartphone case study (Ali-Yrkkö et al. 2011). In early 2007, the unbundled pre-tax retail price of the phone was €546, which represents the total value added that we break down. In our analysis, we individually evaluated all 600 of the N95’s components, as well as the software and intellectual property it embodies. We determined the value added by companies, functions, and location for all direct and indirect tangible and intangible inputs (including capital expenses and contributions of such supporting functions as top management), from raw materials and initial idea generation to the final sale. Depending on the input, 1–8 stages were required before the final assembly by Nokia, either in Salo (Finland, Table 1, Case 1a) or in Beijing (China, Table 1, Case 1b), and 2–4 stages occurred after.3 In the case of the N95, the assembly location had little impact on the value captured by Finland, the country that hosts Nokia’s headquarters. For a phone assembled in China, the headquartering country’s value-added share was 39 %, whereas for a phone assembled in Finland, this value was 41 %—only two percentage points more. This case suggests that value added is largely detached from the physical flows, and the location of final assembly is not of great importance. Internal services, returns on intellectual property rights, and other intangible aspects of the GVC largely determine where the value added is accounted for.

The N95 provides insights for a point in time. However, how does the provision of phones evolve over time? To study this issue, we considered three feature phones at different points in time (Seppälä and Ali-Yrkkö 2014): the Nokia 3310 in 2000 (with an unbundled pre-tax retail price of €78.60, Table 1, Case 2a–c), the Nokia 1100 in 2003 (€62.70, Table 1, Case 3a–c), and the Nokia 1200 in 2007 (€27.00, Table 1, Case 4a–b). The chosen models embody identical technologies, features, and functionality. Nokia’s overall value added fell from over €30 per 3310 (43 % of the total in 2000) to under €6 per 1200 (21 % of the total in 2007). Although outside vendors faced similar price erosion, their share of the value added increased over time. In terms of business functions, price erosion was the least rapid in services (distribution, logistics, trade, and warranty provision). Finland captured 39 % of the 3310 value added and 8 % of the 1200 value added, and the corresponding shares for Asia grew from 14 to 36 %. The three feature phones illustrate both the plummeting price of a given feature set and the gradual shift of tasks towards developing countries, which also involves intangible aspects of the value chain. For instance, the Nokia 3310 was designed end-to-end in Denmark and Finland, whereas the Nokia 1200 was designed in China, although Denmark still assumed responsibility for the hardware and software platforms.

In addition to mobile phones, we consider four other consumer electronics products (made by the same case company, Table 1, Cases 5–8). These cases reveal the significant role of transfer pricing (for a discussion in the context of precision machinery, see Seppälä et al. 2014): The brand holder’s subsidiaries worldwide buy products for sale at prices that roughly reflect the costs of components and final assembly. As a result, the locations that host the sales subsidiaries appear to be quite profitable in our calculations, and the headquartering country of Finland commands 5–17 % of the value added. However, the case company has misinterpreted the established transfer pricing principles (OECD 2010). If we correct this error, Finland captures 42–66 % of value added.

Table 1 suggests that in all but the most basic electronics products, the value added by the brand holder is quite large, whereas the share of assembly is quite small. The share of distribution is invariably larger than that of assembly; in the case of a low-end feature phone, even the share of logistics dwarfs that of assembly. At least at the high end, the headquartering country does quite well, provided that the employed transfer pricing practices conform to internationally accepted guidelines.

5.2 Machinery and Metal Products

We studied a women’s bicycle in 2011 (Kalm et al. 2013), at which point our case company of Helkama was considering three options: in-house assembly in Finland (Table 2, Case 9a, assembly location until 2006) and outsourced assembly either in Lithuania (Table 2, Case 9b, assembly location since 2012) or in Indonesia (Table 2, Case 9c, assembly location in 2009–2011). In Finland, the final assembly is 16 % of the total value added and in either outsourcing location, this value is 2 %. For a relatively simple yet somewhat labor-intensive product such as a bicycle, the difference between a low-cost and a high-cost assembly location is large. However, because logistics and certain other costs are higher with outsource offshoring, the cost disadvantage of Finland amounts to 8 %. In the case of outsourcing to Lithuania, Finland still captures 58 % of the value added, which reflects Finland’s role as the headquartering country as well as the significant share of local distribution (including wholesaling and retailing).

For ten mechanical and precision engineering products (Table 2, Cases 10a–19b), our case companies place the assembly in multiple locations, mostly in Finland and China. With local assembly, Finland’s value-added share ranges from 31 to 90 %, and with offshore assembly, the share ranges from 2 to 55 %. The difference between onshore and offshore is between 11 and 49 percentage points, which appears large in light of the intuition gained from studying consumer electronics. We attribute this observation to three factors. First, unlike in electronics, intermediate inputs are commonly locally sourced and localized support functions may be required; the company might employ local sales and marketing staff and perform location-specific development due to, e.g., national idiosyncrasies in building codes. Furthermore, in selected cases, the factory floor is also the breeding ground for innovations embodied in future offerings, which calls for the presence of local research staff. Second, in contrast to electronics, intangible capital tends to be more tacit and embodied in the production process. Third, the assembly plant is more commonly defined as a risk carrying entity, in which case a larger fraction of profits is allocated to that location. The ten engineering cases also suggest that with more complex and customized engineering products, the overall cost of operation may be roughly the same or lower in a (seemingly) high-cost assembly location (Table 2, Cases 13a–19b).

Ponsse assembles its forest tractor in Finland (Table 2, Case 20), and a large share of components are manufactured by local companies or companies located in Western Europe. Our analysis reveals that 48 % of the product’s total value is created in Finland (when sold outside Finland). A significant share of the value added captured in Finland is created by local suppliers.

Table 2 suggests that in machinery and metal products, assembly is quite important, although in most cases its value-added share is still dominated by the case company’s (other) value added, which is attributable to broadly understood headquartering functions (the case company value added excluding any in-house assembly, logistics, and distribution). The role of logistics and distribution varies greatly but, on average, the value added share of these two components is roughly equal to that of assembly.

5.3 Foodstuffs and Textiles

A chocolate bar and a bag of rye bread are among the eight analyzed foodstuffs (Table 3, Cases 21–28b). These cases highlight the considerable role of wholesalers and retailers, who capture 38 % of the total value added on average, which is approximately the same as the brand owner’s share.

In the case of the four non-apparel textile products (Table 3, Cases 29a–32b, by the same company), the brand owner outsources production to a contract manufacturer, capturing 7 % of the total value added. Wholesalers and retailers capture half of the value added, and the brand holder captures one-third.

Table 3 on foodstuffs and textiles highlights the dominant role of wholesaling and retailing in certain consumer products, which reflects both the significant associated costs and market power, which is driven by the case companies’ inability to directly reach their clientele in a cost-efficient manner. It is nevertheless the case that successful branding and distinctive design features earn considerable returns. In Table 3, the letters a and b following certain case numbers refer to the country of final sale rather than to the assembly location (as in Tables 1, 2, and 4): a refers to the final sale in Finland and b to elsewhere in the EU-27. A comparison of a and b (the last two lines in Table 3) suggests that the country of final sale makes a 65 percentage point difference in the value-added capture in Finland.

5.4 Wood-Based Products

A piece of two-by-four inch sawn timber (Table 4, Case 33) is interesting in two respects. First, 100 % of the value added is captured in Finland. If the tree trunks were imported from Russia, which was a common practice in the early 2000s, Finland’s share would decline to 55 %. Second, the sawn timber itself is provided at a loss, and saleable side products, e.g., woodchips, make the process economically viable.

Our cases also include other wood-based products (Table 4, Cases 34–37). The analyses of these products reveal that the value created in physical activities, including raw material supplies and processes towards the final product, varies between 56 and 67 % of the value added, whereas the value created in the remaining and more intangible activities varies between 33 and 44 %.

Table 4 suggests that in certain industries, raw materials and assembly still play a dominant role. The geographic breakdown hints that these GVCs often remain quite regional and even local.

5.5 Services

In addition to products, we consider four (mostly) digital and four other service offerings.

A business-to-business software platform (Table 5, Case 38, Kalm et al. 2014) offered by a venture-backed startup consists of a developers’ toolkit and a library of software elements that may be embedded into clients’ products. The company charges a fixed fee for the developers’ toolkit, and the variable revenue stream is tied to the sales volume of the product in which the component is embedded. The case company’s value-added share is approximately 29 %, and the company spends a substantial amount on overhead services (i.e., law, travel, and consulting services). The share of Finnish value added is 61 %.

A large multinational company delivers a digital consumer service (Table 5, Case 39, Kalm et al. 2014) via a delivery platform and the physical infrastructure that it owns, but the actual content is purchased from outside vendors. The revenue model consists of monthly purchases and additional amenities. The case company’s value added share is 30 %, which is considerably less than that of the content providers. The share of Finnish value added is 56 %. The same company also provides a business-to-business digital communications service (Table 5, Case 40, Kalm et al. 2014). This service relies heavily on the underlying distribution network owned by the case company, but the key component is a software layer that establishes secure private communication “pipes” for client companies operating in multiple locations. Revenue is mostly based on a monthly service fee. The case company’s value added share is 81 %. The share of Finnish value added is 72 %.

Whitevector, which was a small startup at the time and is now a component of M-Brain, provides Software-as-a-Service (SaaS) social media analytics (Table 5, Case 41, Kalm et al. 2014). Revenues are based on a monthly service fee. The provision of the service is almost purely algorithmic. Because Whitevector obtains its “raw” content for free and the cost of leasing sufficient cloud computing capacity and licensing software that it uses as a component of its in-house software is rather marginal, its value added share is 93 %. The share of Finnish value added is 97 %.

In Tables 1, 2, 3 and 4, on average 5 % of the value added is attributable to logistics. In order to study the role of logistics in further detail, we consider an international logistics company based in Finland that first stocked and then delivered—on behalf of a multinational client—a running shoe imported from Asia either to Finland (Table 5, 42a) or to Russia (Table 5, 42b).4 A comparison of the two cases suggests that Finland being the headquarters country makes little difference—value added is largely captured in the country where the delivery takes place.

Even though most of the working hours are done by freelancers, the case company providing a translation service (Table 5, Case 43) creates 77 % of the value added. Consequently, Finland captures 72 % of the value added.

While the analysis of Seppälä and Kalm (2013) suggests that over the life-cycle of the product the provision of after-sales spare parts (Table 5, Case 44) generates more income that the initial sale, it is also demanding when it comes to handling global inventory and distribution—these functions, here handled internally by the case company, account for 38 % of the value added (this share is included under “Case” in Table 5). The direct manufacturing cost of the spare part accounts for 1 % of the value added. Geographically value added largely shows at the distribution centers throughout the world.

Our final analysis concerns a local group providing live performances of popular music (Table 5, Case 45). Two-thirds of the value added is captured by the band; for the remainder, it relies on outside vendors for, e.g., booking, logistics, and accommodation. Since the band does not perform or sell recordings outside Finland, effectively all of the value added is domestic.

6 Discussion

From our data we conclude: (1) the term GVC disguises considerable variation within and especially between industries; (2) value chains for raw-material-intense process industries, for example, are not nearly as global as the value chains of industries that provide more advanced products; (3) in many consumer products the value added share of wholesaling and retail is quite large; (4) accounting for where profit is allocated inside a MNE is difficult; (5) inventory carrying costs and logistics are significant but often ignored even by MNEs themselves; and (6) industries vary in their functional “stickiness”. In addition, services have scarcely been analyzed in this context: Our few examples of digital services suggest that the case companies and their headquarters locations can capture large shares of the value added, with little “dribbling” to locations where the service is consumed. In non-digital services, however, value added has a certain tendency to be captured at locations of provision and consumption.

Our case studies suggest three main ways for a company to obtain an “over-sized” chunk of the total value added (in order of importance): acting as the orchestrator and/or brand owner, controlling the customer/user interface, and retaining a gate-keeping position (e.g., by cornering the market for a key input). For an individual, these attractive company positions imply job assignments in high-level service tasks in the broadly understood headquartering functions as well as in the creation and management of intangible assets.

Particularly in selected precision-engineered industrial products, assembly plays a prominent role, but generally, value added is dominated by the intangible aspects of GVCs, i.e., market and internal services as well as the creation and appropriation of intellectual property. Value added shows some tendency to migrate to either the earlier or later stages of the value chain at the expense of assembly/processing or service provision toward the middle.

From the viewpoint of a national economy, the impact of the offshoring of final assembly varies significantly. This variation depends primarily on four issues: (1) the impact of relocation on non-assembly functions and tasks, (2) the role of intellectual property in the product, (3) transfer pricing practices employed after relocation, and (4) the juridical location of the company’s profit center.

With the rise of GVCs, national competitiveness begins to depend on citizens’ abilities, both as individuals and as organizations, to position themselves well within GVCs. With this understanding, the policy objectives of a nation-state are as follows: (1) to engage nationally active businesses and individuals in attractive GVCs; (2) to place these businesses and individuals in strong bargaining positions; (3) to ensure that value added shows up as national income, returns, and wealth; (4) to ensure sufficient private and public tangible and intangible investments; and (5) to optimally deploy all nationally available resources. These five objectives amount to stating that a nation-state seeks sustainable high and rising labor productivity and full employment within its borders.

Perhaps surprisingly, the OECD (2013) suggests that in the presence of GVCs, the core features of modern economic policy are intact (in the spirit of Schumpeterian or the “new-new growth theory”, cf. Aghion and Howitt 2009). Thus, the OECD’s policy prescription remains: promote market competition; encourage creative destruction; make public investment in education, research, and infrastructure; and foster a financial and business environment that is conducive to private investment in tangible and intangible assets.

In our understanding, the OECD’s (2013) policy prescription is indeed compatible with how GVCs operate in practice. However, our micro-analysis also points toward several hard-to-resolve paradoxes. First, whereas good policy conduct builds on laissez-faire principles and maintains symmetry across businesses and their functions, there is no denying that certain functions, e.g., headquarters, are nationally more desirable than others. Second, with the rise of GVCs, the relative importance of innovative activity has also risen. However, as knowledge flows have grown more international, the core motivation for providing public support for innovative activity, i.e., local knowledge spillovers, has weakened. Third, due to both consumer surplus and the smooth operation of GVCs, keen competition in domestic markets is to be applauded. For the same reasons, any national government wants foreign governments and international organizations to promote competition in global markets. However, a national government also wants businesses and individuals within its national borders to have market power and thus a lack of competition when such power is exercised on foreign entities. Fourth, GVCs give rise to several features that suggest a need for relatively extensive social safety nets, but the political support for and the scope of financing such safety nets might simultaneously be reduced. Fifth, the scope of trade policy broadens to include considerations of all cross-border flows. In fact, in the context of GVCs, any market malfunctioning becomes a trade policy issue. Although a good rule-of-thumb is to treat domestic and foreign provenances equivalently with respect to goods, services, investment, capital, ownership, technology, or workers, it is sometimes rational to depart from this over-arching principle. For instance, if and when ownership acts to choose locations for headquarters and other business functions, controlling local ownership is desirable. Sixth, GVCs make setting of wages increasingly global. In most developed countries, this circumstance poses a challenge for existing wage setting mechanisms, not least because it has simultaneously become more difficult to form sensible unions on both sides of the labor market, because large and homogeneous employee and employer groups with aligned interests are less prevalent.

7 Conclusions

GVCs are less tied to their tangible aspects, e.g., the assembly location, than conventional wisdom suggests. The intangible aspects, market and internal services, and creation and appropriation of intellectual property are more important. However, this generalization is not universally true—GVCs are amazingly complex and heterogeneous. Upon comparing value-added of domestic and foreign final assembly, we find significant variation in national outcomes. For example, in cases of offshored final assembly, the value added of the home economy is only two percentage points lower for a high-end smartphone, while it is already ten percentage points lower for a low-end feature phone, and nearly thirty percentage points lower on average for machinery and metal products. We also find that value chains for basic products and services are not nearly as global as those of advanced products.

In the course of our analysis, we observed changes in how corporations operate their GVCs. Sometimes the case companies implemented changes as a response to our analysis, since we explicitly highlighted the roles of inventory carrying costs and logistics. Our observations nevertheless suggest the dispersion of global value chains at ever-finer resolution is not reversing but it is becoming more precise and smarter. Once the longer-term costs and benefits are evaluated, seemingly high-cost locations, such as Finland, often appear quite competitive.

With the advent of GVCs, policymakers have found themselves in a difficult situation. Just following economic development is challenging. Implicit assumptions in decision making—such as aligning interests of the country and its champion corporations as well as virtues of high-tech industries—are challenged. While policy thinking needs to be adapted, some of the received wisdom on sound policy principles remains valid. Especially high-income European countries nevertheless urgently need new strategies for retaining higher value-added activities and for capturing associated returns.

Footnotes

  1. 1.

    Plus taxes, minus subsidies; ignoring several complexities, e.g., computational rents in housing and illegal activity.

  2. 2.

    Note that the sum of all value-added activities is equal to the final sale price of the product/service. We consider pre-tax prices without subsidies.

  3. 3.

    For the most part, we normalize the country of final sale to reflect the geographic distribution of the overall sales for the product/service in question. This choice has consequences in the actual calculations, since for example value added by wholesalers and retailers largely resides in the country of final sale, which in turn has influences on the value added per country.

  4. 4.

    Should we study the value chain of the running shoe in question, logistics would account for 0.35 % of the value added when delivered to Finland and 0.41 % when delivered to Russia.

Notes

Acknowledgments

This research is a part of the ongoing collaboration of BRIE, the Berkeley Roundtable on the International Economy at the University of California at Berkeley, and ETLA, the Research Institute of the Finnish Economy. Matias Kalm and Timo Seppälä have been key members of the team performing global value chain analysis at ETLA and have greatly contributed to this manuscript.

References

  1. Aghion P, Howitt P (2009) The economics of growth. MIT Press, CambridgeGoogle Scholar
  2. Alakoski M (2012) Who creates value added in a global value chain? Master’s thesis, Aalto UniversityGoogle Scholar
  3. Ali-Yrkkö J (2013) Mysteeri avautuu: Suomi globaaleissa arvoverkostoissa. Taloustieto, HelsinkiGoogle Scholar
  4. Ali-Yrkkö J, Rouvinen P, Seppälä T, Ylä-Anttila P (2011) Who captures value in global supply chains? Case Nokia N95 smartphone. J Ind Compet Trade 11:263–278CrossRefGoogle Scholar
  5. Backer KD, Miroudot S (2013) Mapping global value chains. OECD Trade Policy Papers 159Google Scholar
  6. Cattaneo O, Gereffi G, Staritz C (2010) Global value chains in a postcrisis world: resilience, consolidation, and shifting end markets. In: Cattaneo O, Gereffi G, Staritz C (eds) Global value chains in a postcrisis world: a development perspective. The World Bank, Washington, DC, pp 3–20CrossRefGoogle Scholar
  7. Dedrick J, Kraemer KL, Linden G (2010) Who profits from innovation in global value chains? A study of the iPod and Notebook PCs. Ind Corp Chang 19:81–116CrossRefGoogle Scholar
  8. Dietzenbacher E, Los B, Stehrer R, Timmer M, de Vries G (2013) The construction of world input–output tables in the WIOD project. Econ Syst Res 25:71–98CrossRefGoogle Scholar
  9. García M (2011) Creating value from innovation: the analysis of an innovative product value chain. Master’s thesis, Aalto UniversityGoogle Scholar
  10. Grossman GM, Rossi-Hansberg E (2008) Trading tasks: a simple theory of offshoring. Am Econ Rev 98:1978–1997CrossRefGoogle Scholar
  11. Kalm M, Pajarinen M, Rouvinen P, Seppälä T (2013) The rise of Baltic Sea value chains—A bicycle producer’s world tour. In: Ketels C (ed) State of the region report. Baltic Development Forum, Copenhagen, pp 126–135Google Scholar
  12. Kalm M, Seppälä T, Ali-Yrkkö J (2014) Who captures value in digital services? ETLA Briefs 27Google Scholar
  13. Karjalainen J (2011) Where is value created in the Finnish food industry? Case: Finnish food company. Master’s thesis, Aalto UniversityGoogle Scholar
  14. Kenney M (2012) Where is value in value networks? In: Breznitz D, Zysman J (eds) 21st century manufacturing. United-Nations Industrial Development Organization (UNIDO), Vienna, pp 13–36Google Scholar
  15. Langenskiöld S (2011) Where is value created: value chain mapping and analysis. Master’s thesis, Aalto UniversityGoogle Scholar
  16. Leppänen S (2012) Value creation in global value chains of paperboard-packed consumer goods. Case: Finnish packaging board manufacturer. Master’s thesis, Aalto UniversityGoogle Scholar
  17. Linden G, Dedrick J, Kraemer KL (2011) Innovation and job creation in a global economy: the case of Apple’s iPod. J Int Commer Econ 3:223–240Google Scholar
  18. Linden G, Kraemer KL, Dedrick J (2009) Who captures value in a global innovation network? The Case of Apple’s iPod. Commun ACM 52:140–144CrossRefGoogle Scholar
  19. Neilson J, Pritchard B, Yeung HW-C (2014) Global value chains and global production networks in the changing international political economy: an introduction. Rev Int Polit Econ 21:1–8CrossRefGoogle Scholar
  20. OECD (2010) Transfer pricing guidelines for multinational enterprises and tax administrations. Organisation for Economic Cooperation & Development, ParisGoogle Scholar
  21. OECD (2013) Interconnected economies: benefiting from global value chains. Organisation for Economic Cooperation & Development, ParisGoogle Scholar
  22. Portelligent (2007) Nokia N95: GSM/EDGE 850/900/1800/1900MHz +WCDMA/HSDPA 2100MHz UMTS Cellular PhoneGoogle Scholar
  23. Seppälä T, Ali-Yrkkö J (2014) Changing geographies of value creation in global supply chains: Evidence from mobile telecommunications. In: Seppälä T (ed) Contemporary determinants and geographical economy of added value, cost of inputs, and profits in global supply chains: an empirical analysis. PhD dissertation, Aalto University, Department of Industrial Engineering and Management. ETLA A 48, Helsinki, pp 105–136Google Scholar
  24. Seppälä T, Kalm M (2013) Profiting from product innovation: a product life analysis of the economic geography of value added. Industry Studies Working Papers 1Google Scholar
  25. Seppälä T, Kenney M, Ali-Yrkkö J (2014) Global supply chains and transfer pricing: insights from a case study. Supply Chain Manag 19:445–454CrossRefGoogle Scholar
  26. Serbessa T (2012) Distribution of value added in a global service production. The case of a Finnish service provider. Master’s thesis, Aalto UniversityGoogle Scholar
  27. Sortti I (2012) Exploring the distribution of value in the global nonwovens value chain: where is sustained value created? Master’s thesis, Aalto UniversityGoogle Scholar
  28. Sturgeon TJ, Nielsen PB, Linden G, Gereffi G, Brown C (2013) Direct measurement of global value chains: collecting product- and firm-level statistics on value added and business function outsourcing and offshoring. In: Mattoo A, Wang Z, Wei S-J (eds) Trade in value added: developing new measures of cross-border trade. The World Bank, Washington, DC, pp 289–319Google Scholar
  29. Tempest R (1996) Barbie and the world economy. Los Angeles Times, September 22. http://articles.latimes.com/print/1996-09-22/news/mn-46610_1_hong-kong
  30. Timmer MP, Erumban AA, Los B, Stehrer R, de Vries GJ (2014) Slicing up global value chains. J Econ Perspect 28:99–118CrossRefGoogle Scholar

Copyright information

© The Author(s) 2015

Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.

Authors and Affiliations

  1. 1.ETLA, The Research Institute of the Finnish EconomyHelsinkiFinland

Personalised recommendations