Beruflich Dokumente
Kultur Dokumente
Lockström2007 Book Low-CostCountrySourcing
Lockström2007 Book Low-CostCountrySourcing
Deutscher Universitäts-Verlag
Bibliografische Information Der Deutschen Nationalbibliothek
Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der
Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über
<http://dnb.d-nb.de> abrufbar.
D 1540
Foreword
Low-cost country sourcing has for long been one of the hottest buzz words in the
purchasing and supply management community, especially among practitioners.
The lack of coverage in the academic field can potentially be explained by the
complexity and the breadth of the topic. Although companies in most industries
have sourced on a global level for decades or even centuries in order to reduce
cost, the subject have received increasing attention over the last few years in a
sense that the concept has become more and more institutionalized as part of the
overall sourcing strategy.
The developments are welcome – especially since the field of purchasing and
supply management traditionally has been lagging behind other, more
“acknowledged” disciplines such as finance, accounting, and marketing, among
others. This lag is glaring if one considers the absence of faculties involved in the
discipline among many universities. Therefore, the initiative of Supply
Management Institute to build a worldwide research and training network has
proven to be highly attractive for both scholars and practitioners.
Moreover, the thesis part investigating implication puts the topic in perspective,
by probing into an area which has yet been covered to a very low extent. This part
contrasts the long tradition of heavily focusing on external factors as key strategic
drivers, instead looking at the internal factors of the firm and how these contribute
to the performance of low-cost country sourcing initiatives.
The results of the thesis show that there is currently a tremendous momentum in
the field of low-cost country sourcing. Companies that do not leverage their
organizations in low-cost regions will miss opportunities to reduce cost while
maintaining quality levels, get access to new suppliers, get closer to new customer
bases and so forth. Those that do not exploit opportunities in emerging supply
markets and protect their strategic assets will most likely suffer from competitive
disadvantages and deteriorating profits on a long term. Those who do will be the
future market leaders.
Finally, I would hereby like to congratulate Dr. Martin Lockström for a great
effort. As part of the first generation of doctoral students at the Chair for
Purchasing, Logistics and Supply Chain Management at European Business
School, he has understood that performance is not only measured through
successful delivery of research projects, scientific publications and high-quality
lectures, but more importantly, through a very strong and amiable personality and
mindset. Not to mention the least is Dr. Lockström a couple of months after his
doctoral degree ceremony appointed as Associate Professor at SMI China to build
up the BMW-SMI Endowed Chair for Purchasing and Supply Management at
Tongji University in Shanghai. I can simply congratulate to the successful career
of Dr. Lockström and express my honor to let him reach his full potential.
Acknowledgements
This thesis is an embodiment of a journey in space and time. Coming from
Sweden as a newly hatched M.Sc. graduate and starting at the Chair of Purchasing,
Logistics and Supply Chain Management, ebs European Business School, in the
small town of Oestrich-Winkel, Germany, I knew little about what was to come.
With an initial headcount of four, the Chair organization now includes almost
fourty staff members in three countries. Through the growth and success of the
Chair, I have gathed knowledge and experience that I will benefit from for the rest
of my life. Topics such as entrepreneurship, strategy, human resources
management, and internationalization, are nowadays things I have not only read
about in text books, but dealt with on a daily basis at the Chair.
The research of this thesis would not have been possible without support from
other organizations and individuals. Although I risk omitting people that should
have been mentioned, I would like to name a few of those who have contributed
most to the emergence of this thesis. As a start, I would like to thank the faculty at
the Chair of Purchasing, Logistics and Supply Chain Management of ebs
European Business School. First of all, I would like to thank my first supervisor,
Prof. Christopher Jahns, who has truly inspired me to work hard, and he has given
me the authority and autonomy to realize my dreams. I would also like to thank
Dr. Evi Hartmann for sharing her long experience in practice, and her invaluable
input regarding technical issues and methodology, but also for providing helpful
feedback on my research.
There are also a number of doctoral students at the Chair, with whom I have had
fruitful discussions. These include Roger Moser, Gerhard Trautmann, Tobias
VIII
There are also a number of people outside the Chair I would like to thank. Firstly,
I would like to thank Mr. Peter Bernard, Mr. Paul Hampton and Mr. Steve
Muddiman at Ariba, Inc., for providing funding and support that enabled the
empirical research of my thesis. I would also like to take the opportunity to and
thank Mr. Alexander Vukotic at Confield Research who always was very
cooperative and friendly during the field study execution. I also owe thanks to
DAAD for the financial support in terms of a scholarship during my doctoral
studies, which has greatly facilitated my personal situation. Another institution
that should not go unmentioned is CDHK of Tongji University, Shanghai, at
which I had the opportunity to experience the world’s number one supplier market
from within. My stay there greatly served as a source of inspiration when working
on my thesis in the end phase.
I would also like to thank my parents, and my wife Veronica. Without their
kindness, support and patience, I would not have been able to accomplish my
goals. They have also helped reminding me that there is a life besides and beyond
writing a dissertation. Finally, I would like to express my sincere apologies for
anyone forgotten to be mentioned. Therefore, I hereby collectively thank all of
those who deserve being mentioned but unintentionally were omitted.
Martin Lockström
IX
Table of Contents
1. Introduction..........................................................................................................................1
2.1. Terminology.........................................................................................................................10
2.1.1. Purchasing, Procurement, and Supply Management............................................................11
2.1.2. Strategic Sourcing................................................................................................................15
2.1.3. Global Sourcing and Low-Cost Country Sourcing ..............................................................18
2.2. Internationalization ..............................................................................................................21
2.2.1. Globalization and International Trade .................................................................................24
2.2.2. Internationalization of the Firm ...........................................................................................29
2.2.3. Internationalization of Purchasing and Supply Management...............................................40
2.3. Concept of Low-Cost Country Sourcing..............................................................................43
2.3.1. Conceptual Frameworks ......................................................................................................43
2.3.2. Motivations and Barriers......................................................................................................49
2.3.3. Contingency Models ............................................................................................................53
2.4. Positioning the Topic: A Morphological Analysis...............................................................55
2.4.1. Methodological Background................................................................................................55
2.4.2. The Morphological Approach..............................................................................................56
2.4.3. Morphological Box For Low-Cost Country Sourcing .........................................................57
2.5. Interim Summary .................................................................................................................65
References ....................................................................................................................................227
List of Figures
Figure 1. Thesis outline....................................................................................................................8
Figure 2. The purchasing and supply management continuum .......................................................13
Figure 3. The three dimensions of globalization .............................................................................25
Figure 4. The four drivers of globalization .....................................................................................28
Figure 5. Modes of internationalization ..........................................................................................32
Figure 6. Global leverage points from internationalization.............................................................34
Figure 7. Factors influencing the attractiveness of nations .............................................................36
Figure 8. Configuration and coordination of operations .................................................................37
Figure 9. Coordination and integration along the value chain.........................................................38
Figure 10. Kaufmann and Hedderich’s (2005) global sourcing framework...................................44
Figure 11. Arnold’s (1989) global sourcing framework.................................................................47
Figure 12. Global sourcing stages ..................................................................................................48
Figure 13. Morphological box for ”purchasing and supply management”......................................62
Figure 14. Morphological box for low-cost country sourcing.........................................................65
Figure 15. Graphical representation of research design parameters................................................74
Figure 16. Rresearch design configuration......................................................................................76
Figure 17. Key decisive research design factors .............................................................................78
Figure 18. Research method............................................................................................................80
Figure 19. The theory as an instrument for predictions...................................................................82
Figure 20. Industry representation...................................................................................................92
Figure 21. Philosophical research dimensions ................................................................................95
Figure 22. Graphical representation of philosophical considerations..............................................98
Figure 23. Driving forces behind globalization.............................................................................102
Figure 24. The Supply Management Navigator ............................................................................107
Figure 25. Vision for low-cost country sourcing activities ...........................................................108
Figure 26. Low-cost country sourcing strategy.............................................................................110
Figure 27. Use of key performance indicators ..............................................................................111
Figure 28. Use of cross-functional purchasing teams....................................................................111
Figure 29. Who initiates and drives LCCS initiatives? .................................................................112
Figure 30. The purchasing channel ...............................................................................................113
Figure 31. LCCS knowledge level ................................................................................................115
Figure 32. Top ten sourcing countries for direct spend among French companies .......................116
Figure 33. Top ten sourcing countries for indirect spend among French companies ....................117
Figure 34. Top ten sourcing countries for direct spend among German companies .....................118
Figure 35. Top ten sourcing countries for indirect spend among German companies ..................118
Figure 36. Top ten sourcing countries for direct spend Italian companies....................................119
Figure 37. Top ten sourcing countries for indirect spend Italian companies.................................120
XIV
Figure 38. Top ten sourcing countries for direct spend among Spanish companies......................121
Figure 39. Top ten sourcing countries for indirect spend among Spanish companies ..................121
Figure 40. Top ten sourcing countries for direct spend among UK companies ............................122
Figure 41. Top ten sourcing countries for indirect spend among UK companies .........................123
Figure 42. Importance of various low-cost regions for direct spend .............................................124
Figure 43. Importance of various low-cost regions for indirect spend..........................................124
Figure 44. Direct LCCS spend as percentage of total purchase spend ..........................................125
Figure 45. Indirect LCCS spend as percentage of total purchase spend........................................126
Figure 46. Direct LCCS spend as percentage of total spend by industry ......................................127
Figure 47. Indirect LCCS spend as percentage of total spend by industry....................................128
Figure 48. Suitable supplies for LCCS..........................................................................................130
Figure 49. Sourcing of goods vs. services 2005 and 2010 ............................................................133
Figure 50. Different supply categories sourced in LCCs 2005 and 2010......................................134
Figure 51. The ten most important supply risks ............................................................................135
Figure 52. LCCS Process phases...................................................................................................139
Figure 53. The LCCS roadmap .....................................................................................................140
Figure 54. Business impact from LCCS........................................................................................143
Figure 55. Scheme of organizational, economic and management theories..................................150
Figure 56. The causal chain of the resource-based view (Barney, 1991)......................................156
Figure 57. Constructs and interrelated hypotheses........................................................................181
Figure 58. Exogenous measurement model with parameters ........................................................196
Figure 59. Endogenous measurement model with parameters ......................................................198
Figure 60. Structural model with parameters ................................................................................208
Figure 61. Structural model with path coefficients .......................................................................213
XV
List of Tables
Table 1. Terms related to purchasing and supply management.......................................................15
Table 2. Definitions of sourcing and strategic sourcing in literature...............................................18
Table 3. Definitions of global sourcing and low-cost country sourcing.........................................21
Table 4. Internationalization topics .................................................................................................24
Table 5. Globalization topics in literature .......................................................................................29
Table 6. Internationalization topics at firm level............................................................................39
Table 7. Internationalization of purchasing and supply management topics...................................43
Table 8. Concepts related to low-cost country sourcing .................................................................52
Table 9. Chi-square test for non-response bias ...............................................................................93
Table 10. Key contributors to the resource-based view of the firm ..............................................155
Table 11. Resource classification schemes (adapted from Fahy, 2001)........................................167
Table 12. Univariate statistics before and after normal scores procedure .....................................193
Table 13. Standardized solution for exogenous measurement model ...........................................197
Table 14. Standardized solution for endogenous measurement model .........................................198
Table 15. Fit indices for the exogenous side of the measurement model......................................200
Table 16. Fit indices for the endogenous side of the measurement model....................................200
Table 17. Principal component analysis of LCCS resources.........................................................202
Table 18. Principal component analysis of purchasing process performance. ..............................203
Table 19. Structural coefficients with significance levels .............................................................210
Table 20. Fit indices for structural model .....................................................................................211
Table 21. Summary of hypothesis tests.........................................................................................217
XVII
List of Abbreviations
CATI Computer-Aided Telephone Interviewing
CEO Chief Executive Officer
CFO Chief Finanial Officer
COO Chief Operations Officer
CPO Chief Procurement Officer
DMNC Diversified Multinational Corporations
e.g. exempla gratia (for example)
ERP Enterprise resource planning system
EU European Union
FDI Foreign direct investment
GDP Gross domestic product
HCC High-cost country
i.e. id est (this means)
IPO International Procurement Office
IMP International marketing and purchasing paradigm
IT Information Technology
LCC Low-cost country
LCCS Low-cost country sourcing
LISREL Linear Structural Relationships
MRO Maintenance, repair and operations
NAISC North American Industry Standardization Code
PLS Partial Least Squares
PSM Purchasing and Supply Management
R&D Research & Development
RBV The resource-based view
SCA Sustainable competitive advantage
SCM Supply chain management
SEM Structural Equation Modeling
SWOT Strengths, weaknesses, opportunities and threats
TCT Transaction cost theory
VRIO Valuable, rare, inimitable and organization support
WFOE Wholly foreign-owned enterprise
1
1. Introduction
The aim of this chapter is to introduce the reader to the topic of low-cost country
sourcing (LCCS) and to provide relevant background information as well as the
main problem to be solved within the scope of the thesis. It also contains the setup
of research objectives and definition of research questions as guidance for the
research. It also contains an overview of the thesis structure in order to facilitate
navigation throughout the thesis.
In order to reduce costs and thus satisfying internal customer demand (Jahns &
Bänziger, 2002; Koppelmann, 2003), an important objective of the corporate
supply function is to source goods and services in a way that yields reductions in
total cost while delivering adequate quality. The strategic importance of this task
has increased significantly over the last few years, as the degree of value added
has decreased to less than fifty percent in many industries (Jahns, 2005).
One of the most effective ways to achieve cost-saving targets is to source from
countries with a comparatively lower price level, so called low-cost country
2
sourcing (LCCS). Although the term does not specify which costs are actually
lower, one is usually referring to the vast difference in labor costs between the
high-cost country (HCC) and the low-cost country (LCC). Typically, the labor
cost difference ranges between a factor of two to twenty, meaning that a wage rate
including benefits in a HCC of €20 could translate to as little as €1 in an LCC
(Battacharya et al., 2004). As a result, the opportunity cost of not sourcing from
LCCs is very high.
Reducing costs is not the only reason for LCCS, however. LCC markets are often
characterized by rapid growth and absolute market size, and increasing customer
demand has turned these regions into important sales markets as well. Hence,
corporate LCC presence is important not only for sourcing and manufacturing, but
also for serving local consumer markets. In other words, companies that can
acquire significant market shares in these regions are more likely to breed cash
cows in the future.
The low labor cost level has a number of secondary implications, especially in
manufacturing and assembling operations. Firstly, it is possible to increase the
amount of manual labor content and hence reduce capital requirements in terms of
expensive automation equipment such as welding robots, conveyor belts etc.
Whereas lower labor costs translates directly to the bottom line, lower capital
investments can significantly reduce the amount of tied-up capital on the balance
sheet (Battacharya et al., 2004). An increase in manual labor content also means
increased flexibility and reduced costs for exiting legacy assets, since less
machinery and tooling equipment need to be sold or moved when operations are
relocated.
However, LCCS is a crucial task that needs to be planned and executed with grate
care. Despite anecdotal evidence of vast cost-savings opportunities, a company
can easily end up with increased instead of decreased total costs if not doing
LCCS operations properly. Due to the many challenges imposed by globalization
(Jahns, 2001), there are numerous challenges and obstacles to overcome, for
3
example cultural barriers, fluctuating currencies, increased distances for travel and
transportation, political instability among others (Battacharya et al., 2004).
The aim of the sourcing initiatives is to identify suppliers that can deliver
maximum amount of value at the lowest possible total cost, at adequate quality
levels. Firms must learn how to develop strategies that can ultimately lead to
adequate organizations and processes that can cope with challenges in the
environment, in other words creating a “strategic fit” (P. Lawrence & Lorsch,
1967; Zajac, Kraatz, & Bresser, 2000). The main problem for the strategist (i.e.
the CPO) is therefore to obtain critical information and knowledge to craft sound
and viable sourcing strategies.
Interestingly, less attention has been paid to the role of intra- and inter-firm
resources as a source of competitive advantage and their potential influence on
performance. Although somewhat contradictory empirical evidence (e.g.
Schmalensee, 1985), there seems to be empirical evidence from general research
settings in favor of the importance of intra-firm factors (Cool & Schendel, 1988;
4
Cubbin, 1988; Hansen & Wernerfeldt, 1989; Roquebert, Phillips, & Westfall,
1996; Rumelt, 1991). Hence, the first objective of this thesis to examine these
effects in a global sourcing setting, in effect also providing a novel and different
perspective on the topic compared to the dominating TCT-based research on
global sourcing done during the 1990s and early 2000s.
From a practical perspective, anecdotal evidence has shown that companies have
so far primarily relied on the external assessment, or market-based view (“outside-
in perspective”), in order to identify attractive supplier markets, while paying less
attention to the internal assessment (“inside-out perspective”) during the strategy
development phase. Therefore, it is likely that many companies miss opportunities
to adapt their organizations to a highly dynamic and complex sourcing
environment and thus not exploiting comparative advantages to the fullest extent
possible.
This motivates further research on how internal factors on the buying side affect
business performance. Academic research in this field is highly needed, as many
firms are experiencing increasing competition on a global level and are therefore
more or less forced to source globally. Since the research stems from actual needs
among companies, it is very likely that the findings from the thesis will have a
significant importance to business executives as well as scientists in the field.
view of the firm (RBV) and parallel streams of research will be applied as a
theoretical lens onto the topic. The theoretical framework will also be briefly
discussed in the light from other theories from organizational science, economics
and strategic management. Examples of such theories are systems theory,
international trade theory, transaction cost theory and theory of foreign direct
investment. In line with the purpose of the study in combination with the research
gaps, five overarching research questions can be formulated:
The ultimate goal of the thesis is of course to answer the research questions above.
The research questions themselves also provide guidance for the structure and
contents of the thesis. As such, the thesis is structured in a way that it fulfils three
types of requirements as defined by Gleich (2001). These are as follows:
From a structural point of view, the thesis is divided into six chapters (Figure 1).
Chapter 1 serves as an overall introduction to the thesis. More in-depth, chapter
1.1 introduces the reader to the topic by providing relevant background
information and defining the main problem to be solved by the thesis. Based on
the background information and problem definition, chapter 1.2 elaborates the
objectives that are to be accomplished over due course of the research within the
6
thesis. Finally, in chapter 1.3 (i.e. this chapter) explicit research questions that are
to be answered are formulated, and the structure of the thesis itself is explained.
Chapter 2 contains the literature review, which lays the groundwork for
understanding the LCCS topic itself, as well as the reasoning and argumentation
in later chapters. Chapter 2.1 deals with explanation and definition of important
terms used throughout the dissertation (cf. Gleich, 2001, p. 18). The terms are also
compared in order to highlight crucial nuances and differences that are essential
for the understanding of the subject of study. In chapter 2.2 internationalization of
firms are discussed. This chapter helps understanding why and how firms are
expanding into new geographical regions. The knowledge gained from this
section is very helpful in order to broaden the resource-based perspective on
which the theoretical fundament in later chapters is based. Chapter 2.3 builds on
the previous chapters by further developing terms and defining the concept of
LCCS. The LCCS term is then positioned with respect to literature through the
deployment of a morphological analysis in chapter 2.4.
Thus, it tries to give answer to the “what” and “when” questions. Chapter 4.1
contains a general overview of LCCS trends in terms of improvement
requirements, importance, organizational developments, system development,
performance measurement and responsibilities. Chapter 4.2 deals with strategic
LCCS management, based on the Supply Management Navigator (Jahns, 2005).
Next, chapter 4.3 examines trends at the operational level, i.e. activities that are
performed on a day-to-day basis. Then, chapter 4.4 summarizes key activities into
a strategic guideline that helps managers and other practitioners structure
activities in a timely manner. The chapter ends by providing an interim summary
in chapter 4.4.
Chapter 6 is a relatively brief and contains conclusions from the research as such,
but also a discussion in a broader sense. Chapter 6.1 discusses the theoretical
contributions of the research to the existing body of research and it is thus of
primary interest for academics. Chapter 6.2 deals with the managerial implications
of the research, i.e. the practical considerations and recommendations (cf. Gleich,
2001, p. 18) to CPOs and equivalent senior purchasing managers that have
ultimate responsibility for sourcing operations within firms. Finally, in chapter 6.3
limitations of the research is provided and suggestions for future research is
provided.
8
1. Introduction
1.1 Research
1.1 Background and 1.2 Research
Questions and
Problem Definition Objectives
Thesis Outline
3. Research Methodology
The aim of this chapter is to provide a foundation for the LCCS concept by
presenting work done by scholars and practitioners from the past. In other words,
the chapter serves as a review of the literature on the topic. Moreover, a
subsection is devoted to clarify the differences and nuances in terms used in order
to facilitate the reading of this thesis. Finally, some general conclusions are drawn
about the status quo of the research in the field along with some guidance of
future directions. This is essential for the understanding of the topic and to be able
to juxtaposition the concept of LCCS with other streams of research. Few
literature reviews with respect to LCCS have been performed in the past,
potentially indicating an underdeveloped body of research.
In order to get a clear picture of what low-cost country sourcing is, it makes sense
to dichotomize the word into its constituent parts and analyze them separately.
First of all, the word sourcing needs to be clarified. It is a word that has gained a
lot of popularity over the last decade and become a commonly used “buzz word”
in business-related contexts. This has also led to a situation where it is partly
misused and used in too many different contexts. Alas, it is difficult to give the
word a precise definition. Very often, it is used to overarch the sourcing process,
which, put simply, involves the search, evaluation, and selection of suppliers.
10
With this in mind, the word can be used to represent sourcing from one or many
suppliers for a given supply category, so called single and multiple sourcing,
respectively. It can also be used to represent specific regions to source in, for
example local sourcing for domestic suppliers, global sourcing for suppliers
regardless of geographic location, or low-cost country sourcing, which will be
further elaborated later in this chapter. Some people even use the sourcing term
synonymously with the term outsourcing, a discussion that is omitted in this thesis.
2.1. Terminology
In order to successfully explain and investigate the low-cost country sourcing
phenomenon, a number of underlying terms and concepts need to be elaborated.
The meaning and use of the various terms are dependent on the evolutionary paths
that firms have taken throughout the history of acquisition of factor inputs. The
vocabulary used also greatly varies from scholar to scholar, and most terms
overlap to some extent in terms of their meanings. As a result, many of the terms
are used interchangeably, which sometimes can cause confusion. The situation
gets even more complicated as the terms and concepts are not only related to each
other, but also to other fields such as material handling, logistics and supply chain
management. Consequently, the aim of this section is to clarify frequently used
terms as a support to the reader.
about their strategic importance. Ellram and Carr (1994) gave a brief historical
overview where Ammer’s (1974) was one of the earliest writers to directly
address the role of purchasing in business organizations. This view was also
shared by Ansoff (1965) who viewed purchasing as an administrative rather than
strategic task. Purchasing did not receive much attention as a strategic weapon
until the 1980s, when Porter’s five-force model established itself as the number
one strategy framework (Porter, 1980). In this framework, two of the market
forces are “bargaining power of suppliers” and “bargaining power of customers”,
respectively. Consequently, a stream of research emerged in which theories and
concepts were elaborated on how to cope with and balance these forces. During
the 1990s, the paradigm behind purchasing and supply management shifted from
isolated and individual purchasing scattered across organizations, to a more
integrated and holistic approach for the entire organization (Robert J. Trent &
Robert M. Monczka, 2002).
unit of the firm that simply takes orders from internal customers and performs
transactions with external suppliers. As a result, there is little or no coordination
or integration of purchase requirements among the various internal customers of
the firm. A summary of the scholarly work described in this chapter can be found
in at the end of this chapter.
The definitions of purchasing and procurement by Arnold (1997) are very similar
to those of Dobler and Burt (1996), but the main difference is that Arnold uses the
term supply management as an overarching term that also includes procurement,
materials management and logistics. Here, logistics includes inbound logistics,
internal logistics and outbound logistics (Arnold, 1997). In a similar fashion,
Corsten (1994) defines supply management as proactive procurement with a
strategic orientation with the aim of making significant contribution to corporate
performance.
they also claim that German companies exclusively focus on materials. Another
argument is that many resources are interrelated (Arnold, 1997; Pieper & Pfohl,
1993). As an example, the use of a certain type of production material may need a
certain type of tooling or machinery, which in turn needs personnel for operating
and capital for financing.
In contrast, scholars such as Dobler and Burt (1996), Monczka et al. (1998) and
van Weele (1995) restrict their definitions of purchasing and supply management
to materials, services and capital equipment, thus excluding personnel and
financial resources.
Of course, the purchasing processes for personnel and capital are very different
regarding the competencies needed to execute them; this is why such purchasing
processes are handled by other departments (Kaufmann, 1999). Although not
named as such explicitly, “purchasing processes” for the acquisition of personnel
and marketing campaigns are almost exclusively handled by the human resources
department and marketing department, respectively.
Kaufmann (1999) strongly argues that the acquisition of services should be part of
the purchasing department’s task list. He claims that purchasing of machinery and
equipment is closely related to other purchasing processes and such purchases
often are associated with purchases of indirect material and services such as
maintenance, repair and operations (MRO).
itself, but instead a bundle of products and services, such as customer support via
telephone, access to web services, and so on.
A third argument for including services is the fact that the degree of value added
among companies is constantly decreasing and is on average lower than fifty
percent in some industries (Jahns, 2005). Thus, companies are focusing on core
competencies to a greater extent than ever, and this focus not only includes
products, but also services (Monczka & Trent, 1993). As the degree of value
added decreases, the need for professionally managed procurement of products
and services increases (Monczka & Trent, 1993).
The sourcing term has been defined in numerous ways in literature. Vollman et al.
(1984) and van Weele (1994) define it as “researching the market for potential
input sources, securing the continuity of these sources, searching for alternative
sources and keeping knowledge up to date”. Kraljic (1983) on the other hand,
positions sourcing management activities to what he refers to as “bottleneck
supplies”. That is, supplies characterized by high uncertainty/complexity but low
strategic importance to the buying firm. In contrast, Monczka et al. (1998) defines
it as “a cross-functional process that involves members of the firm other than
those who work in the purchasing department. The sourcing management team
may include members from engineering, quality, design, manufacturing,
marketing, accounting, strategic planning and other departments”.
This definition clarifies that sourcing is not just about finding suitable suppliers,
but rather seeing the acquisition of supplies as an ever-changing and continuous
task that goes beyond sole transactions, involving a complex interplay between
suppliers and buyers. Contrary to purchasing, it is also evident that sourcing is
1
In line with the discussion about moving away from a goods-dominant view (Vargo & Lusch,
2004, p. 2), the term more generic term “value creation mechanism” is used instead of
“production” for the definition of “sourcing”.
17
2
Noteworthy, the strategic sourcing definition by Kotabe (1994), is not limited to external
suppliers, but also includes so called internal suppliers of other business units and this
phenomenon is consequently called internal sourcing.
18
that encompasses all steps from opportunity and progress assessment to strategy
institutionalization (Smeltzer et al., 2003).
3
Comparative price level: A comparative price level is defined as the purchasing-power parity
divided by the exchange rate. Purchasing-power parity (PPP) in turn is defined as the number of
currency units required to buy goods equivalent to what can be bought with one unit of the
currency of the base country or with one unit of the common currency of a group of countries.
Definition from United Nations Statistics Division (http://unstats.un.org/unsd/methods/
icp/gdp/gdp06_htm.htm)
20
Under the waves of globalization and outsourcing, the literature review reveals
that large manufacturing companies often develop their products in Europe and
the USA, manufacture in Asia and Latin America, and sell worldwide. This
strategy is the starting point for the definition of global sourcing by Monczka and
Trent (1991) and implies the integration and coordination of procurement
requirements across worldwide business units, looking at common items,
processes, technologies, and suppliers.
The definition of LCCS as it is used throughout this thesis is derived from the
global sourcing term defined by (Monczka & Trent, 1991) and the definition of
LCC above. By combining these two terms as a starting point, the term “low-cost
country sourcing” can be defined as follows:
Hence, from this point of view, LCCS can be seen as a subset of global sourcing,
specifically focusing on the conditions in the low-cost sourcing markets. On the
other hand, as discussed above, global sourcing deals with sourcing in any country
around the globe, thus disregarding country borders. In order to preserve the terms
originally used in the sources, various terms are used interchangeably in the
literature review.
21
Kotabe et al. (1994) Global sourcing Global sourcing as the management of R&D,
manufacturing and marketing interfaces on a
global basis
Mol (2003) Global sourcing Global sourcing as the decision-making
process through which firms find and manage
inputs for final production
Monczka and Trent Global sourcing Global sourcing as the integration and
(1991) coordination of procurement requirements
across worldwide business units
2.2. Internationalization
The concept of internationalization plays an important role for LCCS. By
understanding why and how internationalization takes place at a macro-economic
level, one can effectively translate the developments to firm- and business-unit
level. As a result, it is possible to narrow down the focus so that one can assess
the impact from internationalization on supply-side activities. Accordingly, this
chapter addresses the globalization and international trade in general as a start,
then discussing internationalization of firms, finalized by elaborating the
internationalization of purchasing and supply management. A summary of the
scholarly work described in this chapter can be found in Table 4 at the end of this
chapter.
Since the use of the word ”global” was popularized by influential scholars such as
Levitt (1983) among others, there has so far been relatively little consensus about
its real meaning. Often, the word has been used in juxtaposition with other words
to highlight a world-wide reach. For example, Douglas and Craig (1989) used the
term ”global marketplace” to denote the international product and services market.
In a similar fashion, Particelli (1990) denotes any kind of presence in a foreign
market as ”global participation”. Levitt (1983) and Porter (1986) use the
term ”global strategy” for denoting strategies that involve integration of business
operations across nation borders. Aaker (1998), on the other hand, defines the
term as a means for getting access to strategic markets, bypassing trade barriers
and taking advantage of foreign investment incentives such as tax and interest
rates. In line with Bartlett and Goshal’s (1987; 1989) organization taxonomy (later
elaborated in chapter 2.2.2), ”global firms” are those which pursues a single
strategy that is scaled up, encompasses markets, and exploits commonalities
across nations.
(1989), Hibbert (1997), Rugman and Hodgetts (1995) consider the global business
environment to be a subset of the more general international business environment.
Based on the information above, the aim the following sections is to examine the
drivers and effects from internationalization by starting at a global level and
subsequently narrowing down the scope to firm level.
Economic, social
and political inter-
dependencies
Increased
convergence
Geographic scope
and highly uncertain (Fahy, 2001). With this in mind, Naisbitt (1994) claims there
is a so called ”global paradox”, implying that many large-sized companies today,
despite their market power and economies of scale, are lacking the flexibility to
remain strong in the future. Therefore, it is hypothesized that smaller, more
flexible and agile firms are about the gain market share on the expense on the
industry titans. Ohmae (1985) and Peters (1990) stress the need for speed in order
to keep up with the pace currently characterizing the current business environment.
Globalization would probably not take place if there were not an economic
impetus for promoting it. Already in the 18th century, British economist Adam
Smith (1776) proved that specialization leads to increased industrial output.
Furthermore, he considered international trade as a special case of specialization
where resources are assumed to be scarce and consumer demand in a nation
cannot be independently satisfied. To solve the equation, he proposed that each
nation should engage in activities that use resources with which the nation is
relatively well endowed with. For example, a nation endowed with inexpensive
labor should engage in labor-intense activities. The production surplus should
then be exported to other nations in exchange for goods that the nation cannot
produce as efficiently.
Although Smith did not develop the ideas extensively, another famous classical
economist David Ricardo developed them into what is known as “the principle of
comparative advantage” (Ricardo, 1817). Although one nation can be better at
producing every kind of good or service, and thus have an “absolute competitive
advantage”, it is still beneficial to trade goods as long as the production efficiency
ratios of different goods are different for the two countries. In its original shape,
the principal of comparative advantage is restricted to a two-nation, two-good
system; however, with some modification, it can be extended to include multiple-
country, multiple-good systems.
commodity that extensively uses factors of production that it is well endowed with.
The theory relies on a several assumptions in order to be valid, namely (1)
countries have perfect goods and factor markets, (2) the factors of production are
locationally immobile, (3) production units from country to country are identical,
(4) companies are price takers operating under conditions of atomistic competition,
and (5) there are neither trade barriers nor any transaction costs. Clearly, none of
these prerequisites are completely fulfilled; as a result, its validity is limited.
However, the concept is still valuable, as it can still explain the driving forces of
the internationalization and globalization phenomenon that is taking place.
The difference between the two waves is two-fold. Firstly, in the first wave of
globalization, trade flows were centered on goods, whereas the second wave is
characterized by flows not only of goods, but also of services, information and
knowledge. Secondly, during the first wave, globalization was limited to the
developed countries of the world whereas the second wave also involves
developing countries as well as emerging and transition economies (Baldwin &
Martin, 1999).
According to Yip (1992), there are four primary groups of driving forces behind
globalization: market factors, economies of scale, comparative advantage, and
regulatory factors (Figure 4). Market factors refer to the characteristics of
geographical markets. An example can be incumbent competitors that enter new
markets or new competitors enter incumbent markets. Another example can be the
homogeneity of the markets in terms of consumer tastes and requirements.
Economies of scale implies taking advantage of the expansion of operations into
new markets, while utilizing experience from the past in order to reduce the
28
Scale/Scope
Comparative Advantage
firms, thus switching attention from trade patterns to modes of foreign direct
investment (FDI). In the footsteps of the Bain/Mason industrial organization (IO)
paradigm (Bain, 1968; Mason, 1939), forthcoming scholars gave rise to the
greatest and most influential body of research ever seen in the fields of business
science and strategic management science. During the 60s and 70s, most research
was centered on entry modes. That is, how to enter and expand in new markets,
through exports or FDI (wholly-owned or joint venture). Hymer (1960) suggested
that FDI was driven by the objective to compete in foreign markets and to reinvest
profits earned domestically.
Much of the literature on international management stem from the seminal work
of Chandler (1962), which stressed the importance of fit between structure and
strategy. In a seminal paper, Vernon (1966) claims the primary reason for
internationalizing is to take advantage of comparative advantages among different
nations. In a later paper, Vernon (1979) states that product life cycle stages had
31
Stopford and Wells (1972) studied 187 large-sized U.S. firms in order to shed
light on the question how firms are internationalizing their organizations. It was
found that, depending on the relative mix of foreign product diversity and share of
sales abroad, firms took four distinct internationalization paths. Firms with low
levels of foreign product diversity and foreign sales were characterized by an
international division organization. When foreign product diversity increased,
firms arranged their organizations according to a worldwide product divisional
organization. On the other hand, firms that increased their share of sales abroad,
tended to arrange themselves according to a area division structure. In those cases
where foreign product diversity and share of foreign sales were high, a matrix-like
organization structure was adopted. Subsequent studies supported these empirical
findings (Daniels, Pitts, & Tretter, 1985; Franko, 1976).
Area
Division Matrix
Organization Organization
Share of
foreign sales
International Product
Division Division
Organization Organization
From the centralized global firm, the decentralized multinational firm, and the
intermediate international firm, a fourth form known as the transnational firm can
34
Efficiency
Economies of scale,
comparative
advantage
Knowledge
Leverage
Responsiveness
Deploy knowlege and
Adapt to local
skills world-wide
consumer demands
What is more, not only the environmental conditions influence the organization,
but also certain unique time-paths, that is, administrative heritage in terms of
strategic assets or liabilities. Bartlett and Goshal (1988) claim that business
opportunities often arise at locations where a firm is not very well endowed with
the necessary strategic resources. Thus, the ability of a company to create and
maintain a flexible structure while continuously utilizing this ability to
continuously reconfigure itself to environmental changes requires a transition to
the transnational organization type. This is somewhat similar to the evolutionary
paradigm of comparative advantage proposed by Nelson and Winter (1982).
(1981) and Prahalad and Doz (1987) mapped industries along two dimensions in
their seminal papers – global integration and local responsiveness. This was the
starting point of the as of today old expression “think global – act local”. With this
framework, they captured the benefits of integrations (such as economies of scale),
with the competencies required to reap those benefits. Both dimensions are
important for determining how a firm can compete in a given industry, the global
dimension is most important in order to define the geographic scope of the
industry to compete in.
The first factor is factor conditions, which concerns the position of a nation in
terms of factors of production, such as skilled labor or infrastructure, which can
nurture innovativeness, and bring down certain costs. The second is demand
conditions, which involves the nature of the demand in the home market. It is
hypothesized that complex and demanding consumers force companies to work
hard in order to satisfy them, thus those companies become more consumer-
oriented; this could lead to improved competitiveness which then can be
leveraged abroad. The third component of the diamond is related and supporting
industries, which potentially can improve innovativeness and technology transfer
across industries. Fourth, and last, is firm strategy, structure, and rivalry; this is
basically the determinants of industry attractiveness (Porter, 1980). Companies
who have learnt how to survive in an “unattractive industry” in the home market
is said to be better prepared for competing abroad, where market forces are
weaker.
Firm strategy,
structure, and
rivalry
Factor Demand
conditions conditions
Related and
supporting
industries
Porter (1986) states there are two dimensions of configuration for each activity in
the value chain, namely asset configuration and coordination. Asset configuration
means the degree of dispersion, i.e. where activities are located, whereas
coordination means the degree to which activities are coordinated across country
borders. The dimensions form a two-times-two matrix, with the following options:
From a supply-side perspective, this implies choosing from where to source and
how to integrate procurement needs of geographically dispersed business units
across countries. Porter’s view extends the Ricardian paradigm of comparative
advantage to apply not only on the corporate demand side, but to each echelon of
the value chain. Following this view, it is hypothesized that global firms are
expected to increase globally coordinated activities and expanding geographical
coverage.
Home-Based Global
Degree of Global
Integration
Multi-Domestic
Degree of
Integration
In the first stage, “domestic purchasing only”, as the name implies, the company
on purchases from domestic suppliers (Robert J. Trent & Robert M. Monczka,
2002). Although the company might have international operations, such as sales
41
The fifth and final stage, “integration and coordination of global sourcing with
other functional groups”, is similar to the fourth stage except for the important
difference that integration and coordination takes place not only among business
42
units, but also between functional units, such as finance and accounting,
production and marketing (Robert J. Trent & Robert M. Monczka, 2002). In other
words, integration and coordination of global sourcing activities are realized
through a matrix organization.
At the time of doing the study, the majority of the surveyed companies was found
to be in the third stage (31.0%), followed by the second stage (21.3%), and the
fourth stage (18.1%). The first and fifth stages were represented by 13.4 and 16.1
of the surveyed companies, respectively. Over the three to five years following the
year of the survey, companies were expected to traverse rapidly from the lower
stages to level five. The fact that a minor decrease in the number of companies in
level four was expected reveals that many companies basically are leapfrogging
this stage and enters the fifth stage immediately.
Bozarth et al. (1998) used applied a somewhat different approach among 108 US
companies, and arrived at four distinct development stages. This study also
showed that companies in the lower development phases primarily focused on
short-term benefits in terms of cost, quality and technology, potentially foreseeing
more long-term advantages that can be eliminated from the “system”, such as
tied-up capital, shorter lead-times.
43
One of the most recent contributions to the existing body of literature is a global
sourcing framework by Kaufmann and Hedderich (2005). This framework
consists of two dimensions: location of own value creating facilities and location
of supply sources. Thus, it is very similar to Fayerweather’s (1969) global
sourcing framework, which deals with what suppliers are serving which
production units, and which production units are serving which consumer markets.
44
In other words, the two dimensions concern foreign value creation from a global
business strategy perspective, and imports from a sourcing perspective,
respectively. Furthermore, Kaufmann and Hedderich (1969) claim that foreign
value creation can be performed through franchises, joint ventures, and
subsidiaries. Imports include intra- and extra-firm trade.
Abroad
Location of
Value Creation
Domestic Local
Sourcing Import Orientation
Domestic
Domestic Abroad
Location of Sourcing Activities
Local sourcing abroad features low intra- and extra-firm imports. Basically, it can
be viewed as a replication of domestic local sourcing in an offshore market. The
key difference between local sourcing abroad and global sourcing is the fact that
cross-border flows of supplies only occur to a very limited extent. Thus, it lacks
the extent of international coordination and integration that is needed to be
categorized as global sourcing. The key driver behind local sourcing abroad is
often the access to inexpensive labor. Furthermore, local content requirements in
many foreign sales markets make local sourcing abroad an essential and viable
approach. However, difficulties among local suppliers to meet quality or quantity
requirements of foreign companies led to a development where producers
“forced” their suppliers to relocate (Kaufmann, 1995). Often, local sourcing
abroad is restricted to sourcing of commodities due to the challenges above,
whereas strategic supplies are still being sourced from incumbent supply markets.
(Kaufmann/Carter, 2000).
Finally, global sourcing implies high levels of import sourcing and foreign value
creation. In other words, global sourcing can be seen as a means to integrate and
coordinate globally dispersed value creation and purchasing locations The key
objective is to coordinate company-wide demand to maximize bundling effects
while receiving the best quotes possible from the most important supply markets
(Hausmann/Kaufmann, 2002). In line with the discussion by Monczka and Trent
(Robert J. Trent & Robert M. Monczka, 2002), global sourcing requires both
cross-functional and cross-business-unit integration and coordination.
Earlier studies have shown that firms cannot harness the full advantages of higher-
level global sourcing unless a more strategic approach is adopted (Samli,
Browning, & Busbia, 1998). Although companies have started to realize the
46
Operational
Traditional International
Procurement Purchasing
Location of
Value Creation
Supply Global
Management Sourcing
Strategic
Domestic International
Location of Sourcing Activities
This form of collaboration between functions and business units facilitates spend
consolidation and early supplier involvement. Figure 1 summarizes the possible
levels of Worldwide Sourcing and shows the current positioning of 162 mainly
large companies (average annual sales of $1.5 billion) who participated in the
exploratory research carried out by Monczka and Trent (2003).
Kaufmann and Hedderich (2005) Location of value creation and location of sources as two
determinants of global sourcing activities
Trent and Monczka (2002) Identified five distinct development stages toward global sourcing
49
However, the real growth in research on global sourcing did not take place until
the late 1980s and early 1990s. According to Trent and Monczka (2002),
international sourcing was the primary research focus during this period.
Furthermore, they state that the growth in research was correlated to the decline in
competitiveness of many Western firms and that international sourcing was a
solution to this problem. Most research within the field have addressed potential
effects of international sourcing when competing against companies acting in a
global environment (Rajagopal & Bernard, 1991).
Most research during the 1990s showed that purchase price reduction and not total
cost reduction was the most important outcome from global sourcing initiatives
(Petersen, Frayer, & Scannel, 2000). Moreover, most global sourcing initiatives
were reactive, as a response to the increasing global competition, in order to find
50
suppliers offering the lowest unit prices (Alguire et al., 1994). A review of global
sourcing literature done by Murray (2001) showed that many managers primarily
view global sourcing as means to realize short-term cost savings while neglecting
the sustainable competitive advantages global sourcing can provide on a long term.
Until recently, sourcing decisions have been done at comparatively low
organizational levels and with little collaboration with other functional
departments; thus, sourcing has had a more operational or tactical rather than
strategic nature (Guinipero & Monczka, 1990).
Some scholars have had contradictory views of the overall cost benefits of global
sourcing (Levy, 1991). Technological and organizational developments have
facilitated global sourcing initiatives while logistics and inventory still remain as
hurdles; hence, total cost estimations have often been underestimated. As
secondary outcomes of global sourcing initiatives, researchers have stated greater
access to product and process technology, higher quality and ability to introduce
competition to the domestic supply base (Trent & Monczka, 1998).
It has also been widely acknowledged that suppliers in LCCs can produce goods
with similar or even better quality than their European or US counterparts (C. R.
Carter & Narasimhan, 1990; Handfield, 1994; Min & Galle, 1991). The suppliers
51
It is nowadays not necessary to first launch a product in the home market, and
then launch the same product in foreign countries. Due to improved
communication technologies and global sourcing, products can now be launched
in multiple geographic consumer markets simultaneously. LCCS plays an
4
Six sigma quality here refers to the requirement of deviating six standard deviations from
requirements, which equals a defect rate of 3.4 ppm.
52
increasingly important role here since many LCCs, such as China, are important
not only as supplier markets, but also as consumer markets, due to the large
populations and improved standards of living (Battacharya et al., 2004). Since
technology dissipates at a high rate, parallel sourcing, production and marketing
worldwide is essential in order to secure first-mover advantages and prevent
reverse engineering (Bozarth et al., 1998).
Alguire et al. (1994) Global sourcing drivers Increased global competition as key driver
Bozarth (1998) Technology as global sourcing Study showing that global sourcing can be a
driver means for technology acquisition
Carter and Narasimhan Quality as global sourcing Empirical studies showing that LCCS can
(1990); Handfield driver nowadays yield adequate quality of goods and
(1994); Min and Galle services
(1991); Bhote (1987)
Frear et al. (1992); Global sourcing to access Studies showing that global sourcing can be
Kotabe and Murray customer markets used to produce and sell to consumers in
(1990) supplier markets
Hahn et al. (1986) Sustainable competitive Global sourcing as a means for obtaining
advantage sustainable competitive advantage by looking
beyond national boundaries
Handfield (1994); Global sourcing drivers Purchase price reduction as key driver
Spekman (1991);
Petersen et al. (2000)
Kogut (1985); Porter Global sourcing Developed the notion of integrated sourcing
(1986)
Monczka and Guinipero Global sourcing barriers Logistics, inventory, distance, nationalism and
(1984) a lack of knowledge about foreign business
practices are the most important problems
Murray (2001) Global sourcing drivers Study indicating that managers perceive
global sourcing as a means for realizing short-
term benefits
Starr (1984) International sourcing drivers Identified complexity as an important driver for
international sourcing
Trent and Monczka Globalization developments Decline in competitiveness among Western
(2002) firms as driving force
53
In 1990, another study by Kotabe and Murray (1990), the relationship between
market performance and process and product innovation was investigated.
Somewhat surprising, this study showed that outsourcing led to decreased market
performance. Murray concluded that “there may not be one best way to source;
one should investigate the nature of business environments before making strategy
decisions” (Murray et al., 1995).
The relationship between asset specificity and channel integration has also been
examined through transaction cost theory, hypothesizing that the higher asset
54
specificity of goods sourced (i.e. the higher strategic importance), should lead to
higher channel integration (i.e. larger degree of ownership of the procurement
channel) (Aulakh & Kotabe, 1997). Somewhat counter-intuitive, the results
showed an opposite relationship than what was hypothesized. Moreover, it was
concluded that channel integration itself was not related to performance, but the
control mechanism instead turned out to have positive performance implications.
Kotabe and Omura Global sourcing strategy and Performance positively correlated with internal
(1989) performance sourcing and negatively correlated with
product adaptation.
Kotabe and Murray Market performance and Outsourcing led to decreased market
(1990) process and product innovation performance
Aulakh and Kotabe Asset specificity and degree of Results opposite than hypothesized, asset
(1997) channel integration specificity negatively related to channel
integration
Kotabe et al. (1998) Core and supplementary The results indicated that innovation in so
services and performance called core and supplementary services was
important for strategic performance
Hult (2002) Sustainable performance and Culture encourages entrepreneurship,
cycle time performance innovativeness and organizational learning, in
turn positively influencing business
performance
provides an audit trail, which makes it possible to trace the path from the initial
problem formulation to specific results and solutions.
The approach starts with identification and definition of the various dimensions of
the problem. Subsequently, each dimension is assigned a set of values that each
dimension can hold. The various states can then in turn serve as dimensions that
can hold values. This leads to an n-dimensional matrix known as a morphological
box or Zwicky box (Zwicky, 1966). Zwicky defines five iterative steps in his
morphological approach as shown below (Zwicky, 1969):
2. All dimensions that might be important for the solution of the given problem
must be localized and analyzed.
4. All solutions contained in the morphological box are closely scrutinized and
evaluated with respect to the purposes to be achieved.
Since this case deals with purchasing activities, it is logical to set the delimitations
of the problem domain to the purchasing/supply department of the firm and
consequently set the subject of study to purchasing and supply management as
defined in section 2.1.1. By doing so, the problem can be solved by accomplishing
the following goal:
In order to arrive at a morphological box that fit the scope of this thesis, the
discussion is limited to large-sized firms.
58
In line with the definition in section 2.1.1, supply management is defined as “the
identification, acquisition, access, positioning and management of resources that
an organization needs or potentially needs in the attainment of its strategic
objectives” (Cavatino, 2001).
ranges from a couple of years up to half a decade. The long-term planning deals
with more abstract dimensions such as formulating visions and changing
corporate culture (Bleicher, 1999). This time horizon stretches far away into the
future, up to as much as twenty years ahead (Shrivastava, 1994).
Although the paradigm of the transnational firm mainly focuses on the demand
side of the firm, the supply side needs to be aligned as discussed in the previous
paragraph. Consequently, we use the four categories above to specify
geographical scope of the purchasing activities.
In some cases, the outcome was the matrix organization that tried to reap benefits
and eliminate weaknesses from the functional and divisional organizations,
respectively. Some companies, such as Unilever and ABB succeeded to
implement this organization type, whereas others such as Citigroup and Dow
Chemical failed. Consequently, we choose to divide the type of organization into
three groups: centralized, decentralized and matrix, where the latter can be
considered as a hybrid of the two former ones.5 Depending on the organizational
orientation of the firm as a whole, the purchasing department must be aligned
accordingly.
5
Although these terms represent ideal types, the terms are still valuable for illustrative purposes.
61
must be made; this is evidenced for example in the IT department, which activities
also span across the whole enterprise (Jarvenpaa & Ives, 1991). Hence, we define
the leadership level in three categories: first-level management, middle
management and top management.
Finally, three dimensions can be identified depending on the nature of the supplies
in question. The first is asset specificity, which specifies the criticalness and
uniqueness of the goods sourced. The higher asset specificity, the more difficult it
is to use the good in another situation (Williamson, 1975). Supplies with high
asset specificity are usually complex. This leads to higher transaction costs due to
difficulties to find and evaluate the right suppliers, longer and more complicated
set-up costs and contracts (Williamson, 1975).
State
Dimension
Internal/External
Internal External
sourcing
However, since low-cost country sourcing involves many strategic issues, such as
fulfilling needs for other functional units, considerations for overall inventory
levels etc., we can already now conclude that low-cost country sourcing is closely
interrelated with supply management concepts. As an implication to this, it is
clear that low-cost country sourcing activities needs to be managed by senior
management in order to succeed; low-cost country sourcing is thus aligned with
63
Few companies acting world-wide today are centralized. The most centralized
companies are probably found in the pharmaceutical and chemical processing
industry, where products are identical throughout the world. Here, marketing is
the activity that is different from region to region. Thus, centralized management
is least likely. Since most large-sized companies today have extensively divested
products and are acting in numerous geographical markets, a wide range of
supplies from many different markets are being purchased. Therefore, the most
probable management type to find is the matrix organization.
On a mid term, low-cost country sourcing can be seen a strategic weapon, made
possible primarily by significant total cost savings. Therefore, it is logical to
believe that the first-movers in such undertakings are the ones that have cost
leadership as a primary marketing strategy. This does not exclude other
companies from sourcing in low-cost countries for other reasons though, but they
64
The nature of the supplies sourced in low-cost countries, have until recently been
dominated by low-value and non-critical components (Battacharya et al., 2004).
This is due to the fact that time-critical supplies are difficult to source because of
the long distances and the increased inventory levels incurred. Higher value here
means larger amounts of tied-up capital; this is highly unwanted. Hence, it makes
sense to believe that supplies with low asset specificity are dominating, and will
continue to dominate, since one cannot do much about the geographical distance.
In the same manner, due to the dominance of supplies with low asset specificity, it
is logical to conclude that sourcing is mainly conducted with external suppliers,
using predominantly transactional relationships.
Selecting precise states for each of the dimensions is not a stringent science;
instead, it is done by making use of earlier findings in the field (empirical
assessment) in combination with sound reasoning (logical assessment). Although
there were some ambiguities in the previous chapter, we the choices made so far
are sufficient from a qualitative perspective. This means that we have reached the
destined goal, namely to position “low-cost country sourcing” in the context of
purchasing and supply management. The final morphological box with the
selected states is shown in Figure 14.
State
Dimension
Internal/External
Internal External
sourcing
encompasses the whole supply chain, from the raw materials extractor, to the end
customer.
At firm level, it was concluded that there are several reasons for internationalizing,
for example entering new consumer or supplier markets, realizing economies of
scale and scope across borders, by leveraging competitive and comparative
advantages. Here, globalization differs from internationalization in a sense that
globalization is a more generic term, not only including business, but also political,
social and technological aspects. Increased cultural openness, deregulation and
privatization of markets, technological advancements and decreasing
communication costs were said to influence the process. Empirical studies have
also shown that companies configure themselves in different ways depending on
internal and external factors. A common taxonomy includes four categories,
namely international, multinational, global, and transnational firms.
Finally, chapter 2.4 concluded chapter 2 by position the LCCS topic in the
literature landscape through a morphological analysis. Thereby important
dimensions that characterize the LCCS concept were also identified. Based on the
morphological analysis in combination with the previous literature review, all in
all nine characteristic dimensions were distilled out of the process.
69
3. Research Methodology
This chapter presents an overview of the methodological perspective of the
research. After a general discussion about scientific research, section 3.1 describes
the research design based on the research questions and theoretical models
hypothesized in this study. With the research design as a starting point, the
research strategy is presented in section 3.3. In section 3.3, the perspectives are
broadened by a discussion about the philosophical aspects of the research design.
Finally, an interim summary of the chapter is given in 3.4.
Rigor Objectivity
Testability Generalizability
Replicability Parsimony
Rigor implies having a solid theoretical base and a sound methodological design
(Sekaran, 2003). Moreover, a research methodology that is carefully selected, well
thought-through, and commonly accepted in the scientific community provide
rigor to the research. Therefore, a minimal requirement for rigorousness is well-
anchored hypotheses, evaluation techniques and drawn conclusions. Ultimately,
rigorous research planning enables research with high reliability and validity.
criterion for being a scientific theory. As Hume (1978 [1739]) puts it: “just
because the sun has risen every day for as long as anyone can remember, doesn”t
mean that there is any rational reason to believe it will come up tomorrow”.
Anyhow, in practice, science is not black or white, but rather a grey zone of what
is true or false. A theory that has withstood testing repeatedly has a higher
credibility than a theory which has never been tested. Therefore, replicability
becomes an important requirement for scientific research; if it does not allow
repeated test, it is deemed unlikely to gain credibility.
According to the old adage, “it is better to be approximately right than exactly
wrong”. An important measure here is therefore precision, which means how
accurate a measurement is. The logic is that it does not do much harm if one is
only approximately right as long as it is possible to determine how far one is
from ”reality”. Here, the concept of confidence plays a central role. It tells the
researcher about the range, within which the ”true” measurement is lying with a
certain probability. This confidence interval is commonly 95 percent, meaning
that there is a five-percent chance that the measured value is outside the range.
The narrower range (in terms of upper and lower bounds), the greater confidence.
Sekaran (2003, p. 25) is truly realistic, by stating that “the more objective the
interpretation of the data, the more scientific the research investigation gets”.
In essence, parsimony means selecting the shortest (and, thereby, most efficient)
path to the production of true theorems, given a set of facts (observations) and a
set of theories. In other words, given two equally predictive theories, choose the
simpler one. The French mathematician and philosopher Poincare (1903) thought
that nature itself, this metaphysical entity which encompasses all, is parsimonious.
He believed that mathematical simplicity must be a sign of truth. Interestingly,
Dirac (1939) almost equates simplicity (parsimony) with beauty, by stating
In much research, models are overly complex. As a result, they are often virtually
impossible to visualize in mind and understand from a logical perspective. With
this in mind, it is not by accident that the majority of the most popular
management frameworks consists of two-times-two matrices, such as the Kraljic
matrix in purchasing and supply management (Kraljic, 1983). Mol (2001) asks
what sense an extremely complicated model makes. Furthermore, he states that “a
model with 25 independent variables that are all deemed relevant might lead to an
outcome where everything matters a little bit. However, what really matters in
research is not what matters a little bit, but what matters most” (Mol, 2001 p. 57).
objectives, one cannot determine when the destined goal has been reached; it can
be compared with a skipper on boat not knowing where to go. Moreover, if the
procedures of the research project are not elaborated, the objectives will certainly
be very difficult to accomplish on time and on budget, if accomplished at all that
will say. With the same analogy, a research project without predetermined
analysis tools is like navigating the sea without a compass. Therefore, before
starting the research activities, four dimensions of design parameters have to be
considered before starting the actual research itself, namely (1) pure/applied
research, (2) primary/secondary research, (3) theoretical/empirical research, and
(4) descriptive/explanatory research (Figure 15). These are discussed below.
Pure research implies that there are no explicit practical implications of the
research conducted. In other words, there is no apparent use of the knowledge
except for the contribution itself to the particular field of science. However, this
does not exclude the research from being applied in practice at a later stage (Riley,
Wood, Clark, Wilkie, & Szivas, 2003). In contrast, applied research aims directly
at solving a problem found in practice. Often, as is the case with this thesis,
applied research may be sponsored by external clients to academia, such as
companies and other industrial organizations. As a result, applied research puts a
lot of emphasis on pragmatism and applicability, meaning that the research
outcome can be easily translated into action with little hassle. In the case of this
thesis, the research is without doubt applied, as it seeks to answer a number of
75
In general, primary research concerns research that deals with collecting first-
hand empirical data by using commonly accepted research procedures. Secondary
research, on the other hand, does not involve collection of first-hand data, but
instead relies on analysis of already existing empirical data sets (i.e. ”secondary
sources”). Usually, secondary research is also undertaken in primary research in
order to examine what research has been conducted in the past in a particular field
of interest. Thus, secondary research can serve as a precursor to research design
(Riley et al., 2003). In essence, primary data comprise new and original data from
first-hand sources whereas secondary data includes the rest, such as books, journal
papers, conference proceedings, online databases, and so on. The research
conducted in conjunction with this thesis is predominantly of primary nature,
since a great extent of the data used for the analysis comes from first-hand sources.
Nevertheless, a significant amount of data is gathered from second-hand sources
in order to discuss the various concepts and terms, for the literature review, and
for the creation of the theoretical framework.
As is the case with primary and secondary research, empirical research does not
automatically exclude theoretical research. This is due to the fact that that in order
76
The fourth and final research design dimension is the distinction between
descriptive research and explanatory research.6 In essence, descriptive research
mainly concerns the “what”, “when”, “where” and “who” questions (Riley et al.,
2003, p. 148). In other words, descriptive research only aims at providing a clear
picture of the data itself, without any claims of generalizability. Exploratory
research, on the other hand, tries to uncover the relationships between variables
(i.e. causal relationships), thus trying to answer the “why” questions. Moreover,
the exploratory research also aims at generalizing the results to the population
from which the sample was drawn. With this in mind, as chapter 4 primarily
involves descriptive research whereas chapter 5 mainly deals with explanatory
research, it is clear that the research conducted for this thesis is somewhere in the
middle of the descriptive–exploratory research continuum.
After the four design dimensions have been determined, the standpoint of the
research done as a basis for this thesis can be summarized in a chart as depicted in
Figure 16.
6
In statistical contexts (depending on authors), the terms descriptive and inferential research are
frequently being used, respectively.
77
Firstly, one has to clarify what the different terms and concepts really mean and
their relevance for companies and industries as a whole (cf. 2.1).
In essence, there are two decisive factors when developing a research strategy.
The first decisive factor is to choose whether to conduct quantitative or qualitative
research. The second decisive factor is whether to follow an inductive or a
deductive approach. Although the different alternatives are not mutually exclusive,
that will say, a study can for example contain both quantitative and qualitative
elements, they can somewhat simplified be represented as a 2x2-matrix. The
different approaches have their advantages and disadvantages, as elaborated
below.
Quantitative- Quantitative-
Inductive Deductive
Qualitative- Qualitative-
Inductive Deductive
sample, but also drawing inferential conclusions about the population from which
the sample was drawn.
Induction has been subject to a lot of criticism, however. The main problem is the
way in which hypothesis are considered true or false. In order to prove the
inductive example above to be true, one would need to examine all swans in
universe. To solve this dilemma, Popper (1934) replaced the inductive reasoning
with falsifiability, which means that instead of trying to prove a hypothesis to
be ”true”, one tries to reject it by proving it to be ”false”. The approach taken for
the explanatory research in this thesis is known as the hypothetico-deductive
method, which implies starting with a theoretical framework, followed by
subsequent deduction of hypotheses that are finally tested and evaluated.
Moreover, the sources include both primary data and secondary data. As Yin
(1994) notes, different research methods are useful for different purposes. In this
study, different methods are seen as complements and no method is seen as
fundamentally superior over another method.
Moreover, as is the case of the research conducted for this thesis, scientific
research is usually carried out iteratively, through theory development and theory
testing (Yin, 1994). An example of such an iterative that was applied for this
research in this thesis is depicted in Figure 18.
1. Observation
2. Identification of problem
area/Literature review
3. Theoretical Framework/
Development of hypotheses
4. Operationalization of
Constructs/Research model
6. Interpretation of data/
Hypothesis testing
Normally, research starts by making some sort of observation(s) (step one). In the
context of strategic management and business science, it could be a shift of
strategies pursued, such as increasing degrees of outsourcing, increasing degree of
internationalization and so on. In the case of this thesis, the observation done was
an increasing extent of sourcing in low-cost countries. Already at this stage it is
possible to decide the forthcoming level of analysis, for example individual, group,
business unit, corporate, or even regional/country level.
81
The observation step was then followed by identifying associated problems (step
two). This was achieved by probing deeper into the subject of study, primarily
through an extensive literature review and discussions with experts from industry
and practice. In this way, a picture of what was happening and why started to
emerge. In the research of this thesis, the main “problems” identified were two-
fold. First, a strong trend towards increased sourcing in low-cost countries could
be identified from literature. As a consequence, the natural question was: “will the
trend persist, or will it level off in a near future?” Secondly, another problem area
that emerged was which factors that influence the outcome of LCCS initiatives.
The influence from factors in the external environment has been covered
relatively extensively – as a consequence, the aim of this research is to shift focus
to the interior of firms.
makes it possible to predict the outcome with a certain probability, given a known
set of input data.
Operationalization of
theory/Research model
Step five involves the data collection, which is elaborated more in-depth in
section 3.2.3. This is a crucial step of the research process in which the
appropriate data collecting method is chosen. For the study in this thesis, a
quantitative survey methodology was carried out. Moreover, a careful selection
and execution of sampling procedure ensures high reliability and validity of the
results and high generalizability of the sample to the population. During this phase,
data on every variable in the theoretical framework was collected. This data then
serves as the input for the analysis phase, in which the data is statistically
analyzed. There are two dimensions in this phase. First, a descriptive analysis was
done to give an overall picture of the data. Second, a more exploratory analysis
was conducted in order to test relationships between variables, in other words
testing the hypotheses.
In the sixth and final step, an overall evaluation of the hypotheses is performed in
order to get a holistic picture of the outcome. The aim of the interpretation of the
hypotheses evaluation is to finally provide answers to the research questions and
discuss the theoretical and practical implications of the results. The theoretical
implications give rise to a feedback loop (see Figure 18) in which suggestions to a
modified theory is provided. The practical implications are manifested in a set of
“rules” and guidelines that managers should consider in order to succeed with
their low-cost country sourcing initiatives. A more elaborate discussion on the
most important steps of the research is dedicated to the rest of this subsection.
83
Moreover, the purpose of the literature review is to identify critical variables that
need to be included in the survey process. Therefore, it serves as a complement to
the expert interviews done in this study, since one cannot guarantee that the expert
interviewees articulate all of them. For example, it could be the case that they
cannot articulate them, or are simply not aware of their impact. Another important
goal of the literature is of course to avoid “inventing the wheel again”, by
addressing research topics that have already been investigated.
The literature review for this thesis started by identifying various published and
unpublished materials on the topic of low-cost country sourcing and closely
related fields. In a second step, the relevant information was gathered by getting
access to needed secondary sources. In a third step, the literature was completed
by writing a summarized version of the collected materials.
The material for the literature review in terms of books, journals, magazines,
academic theses and conference proceedings were primarily gathered from online
resources. The online resources used for this thesis are EBSCO, ProQuest, JSTOR,
ScienceDirect and Emerald. These online databases provide access to an
enormous wealth of information, and contain a number of scientific journals that
were extensively used especially for the literature review.
Moreover, statistical data from governments such as the US Census Bureau and
agencies like The International Monetary Fund and The World Bank were also
used. More general search engines such as Google Scholar were used, especially
for searching academic pdf documents from other universities around the world.
In most cases, the most recent data and sources as possible were used. However,
the ”freshness” of sources was not an exclusive criterion, but also the factors such
as commonly accepted importance of authors in the particular fields. For example,
the works of economists such as Adam Smith and David Ricardo dates back to the
18th and 19th century, respectively, and deserves being mentioned in any
publication discussing globalization and internationalization.
The descriptive part of the thesis investigating status quo and trends is not
theoretically elaborated in itself, but is indirectly based on systems theory through
86
According to Punch (2000), theory validation aims at testing a theory, or more in-
depth, testing hypotheses derived from one or several existing theories.
Quantitative research normally starts with the theories, deduces hypotheses from
them, which are then tested. Since the research questions of this thesis are
87
inherently inferential in nature, with the objective to draw conclusions about the
population domain, a quantitative survey technique was selected as research
instrument; In other words, an objectivistic, nomothetic survey methodology. As
the research concerns a mix of theory validation and prediction, the type of
research is strongly positivistic. Only when the objective of the study is clear and
not complex, like in this case, survey techniques can be used (Borque & Fiedler,
1995). Survey techniques have been widely applied in closely related fields of
purchasing and supply management, such as global sourcing (see for example
Alguire et al., 1994; Kotabe, 1994; Narasimhan & Jayaram, 1998; Robert J. Trent
& Robert M. Monczka, 2002). Commonly, questionnaires are used to generate
extensive databases with low level of details. For this study, the questionnaire was
designed to gather data about trends, accessibility to resources, and different
performance measures.
Before the full-scale survey was launched, a pilot study was conducted in order to
test whether the articulation of the research instruments was sufficient to meet the
research objectives. Thus, the pilot study helped clarifying the effectiveness and
relevance of the questionnaire questions. Moreover, it provided opportunities for
updating and improving the questionnaire so that it maximizes the opportunities
for securing the views of the respondents (Riley et al., 2003). The pilot study was
performed on a relatively small sample (N = 30) in the same way as the full-scale
survey was planned to be executed. An important difference, however, was the
fact that the interviewers at the call center performing the interviews also had a
qualitative, on-going discussion between the questionnaire questions in order to
detect unclear issues. These issues were then discussed with the project leader at
the call-center and necessary adjustments were done to optimize the questionnaire
further.
The questionnaire, on which the survey was based, was designed through three
key guiding principles, namely (1) wording of questions, (2) principles of
measurement, and (3) visual design. These are all very important to consider, as
they can minimize the bias of the data (Sekaran, 2003).
88
The wording of the questions should be aligned to the level of knowledge of the
respondents (Sekaran, 2003). Factors that were considered here are estimated
level of knowledge, business culture, and language spoken. The questions were
formulated in way that supposedly maximizes the chances to tap the dimensions
and elements of the concepts. Since the survey was done through telephone
interviews, the interview process must be very efficient and smooth in order not to
waste time. Therefore, most questions (with only a few exceptions) were designed
as closed, meaning that the respondents cannot answer the questions freely, but
rather choosing from sets of predefined alternatives.
The questions were also scrutinized to eliminate the risk for unclear, ambiguous
or lengthy questions. The initial questionnaire was pre-tested for time length,
clarity and other problems. It was also translated into the corresponding language
of the countries subject of study by native speakers at the call center together with
a research assistant who could support with expertise. The questionnaire was then
translated back to the original language. Thus, it could be assured that no meaning
was lost during the translation and that the terminology was consistent and correct.
Since the questionnaire was implemented through telephone interviews, the need
for visual appearance is of course redundant. Nevertheless, the telephone
interviews had to be initiated in a professional and friendly manner. Consequently,
the interviewers at the contracted call-center were briefed on how to introduce
themselves to the respondents. Upon initiation, the interviewers presented
89
The collected data was then analyzed statistically with two de facto standard
software packages, namely SPSS 12.0 and LISREL 8.57. The reason why two
different software packages were used is because they both have their strengths
and weaknesses, and each of them have some unique features not found in the
other. The simpler descriptive analysis was primarily performed in SPSS through
Ȥ2-testing, t-testing, and calculations of mean and variation values. The
exploratory analysis and inferential tests, on the other hand, were predominantly
conducted in LISREL by using maximum likelihood path analysis. More in-depth
explanations of the statistical analysis techniques applied are provided as they are
used, in chapter 4 and 5, respectively.
3.2.4. Sampling
The core of the data collection process is sampling, which means taking a part of a
population to represent the population as a whole (Alreck & Settle, 1995).
Another method is enumeration, meaning that one investigates the whole
population. Since this approach is very costly in terms of time, money and
personnel, it is not very feasible. Therefore, due to economic reasons, sampling is
much more frequently used. With sampling, only a small fraction of the entire
population needs to be investigated. While it is a very cost-effective approach, it
has to be done with great care in order not to introduce bias or error in the data set
(Alreck & Settle, 1995). In other words, if the sampling is not done correctly, it
will not represent the intended population and results will ultimately be biased.
The first step of the sampling procedure was to identify the population. The main
criterion for choosing the sample is to assure that the population consists of
90
people that actually possess the information sought (Alreck & Settle, 1995). In the
case of the research for this thesis, the expert interviews revealed that Chief
Procurement Officers (CPOs) and equivalent senior purchasing managers at
business unit level were appropriate for representing the population since these
tend to be involved in decision-making regarding localization of sourcing
activities. Furthermore, they are frequently positioned at board-level where
strategic sourcing decisions normally are taken.
With respect to the high-level resources investigated in this study, this category of
respondents can be considered the most suitable for this study and hence key-
informant bias is minimized (Kumar, Stern, & Anderson, 1993). Furthermore, the
expert interviews that were conducted before the survey indicated only a very
small discrepancy between the academics and the practitioners. This indicates a
high degree of trustworthiness to the survey informants and thus risk of single
informant bias, which has been pointed out in literature as a potential problem
(Premkumar & Ramamurthy, 1995), was considered low.
The next step was to define the sample frame, which consists of all sample units
representing the population. The sample frame can be a list of names with
addresses for a mail survey, or a map showing residential areas for face-to-face
interviews with residents. In this study, the sample frame consisted of an
electronic database obtained from Dun & Bradstreet, which contained 5,147
entries of names and corresponding telephone numbers of potential respondents.
The actual respondents were randomly selected from the database by using a
computerized random number generator, thus fulfilling the criterion of being a
probability sample (Riley et al., 2003). This implies that each sample unit has a
known, non-zero chance of being selected (Arber, 1993; Chisnall, 1991).
91
Probability sampling provides the best means for drawing a sample that can be
generalized to the population through inferential extrapolation.
All in all 1,273 phone calls had to be made in order to obtain the two hundred
interviews. Out of the total number of calls, about 15 percent resulted in no
answer, and another 14 percent resulted in an answering machine. Three percent
of the numbers were invalid. The remaining 68 percent equals 866 firms which
yields an effective response rate at 23.1 percent. The response rate is comparable
to those of mail surveys and poses no problem as sometimes have been pointed
out in literature (Fitti, 1979; Massey, Barker, & Hsuing, 1981). The sample
represented a wide range of industries, distributed as depicted in Figure 16. The
industries correspond to NAISC codes 11, 21, 23, 32, 33, 44, 45, 48 and 52.7
Mean direct and indirect spend volume were 651 and 262 million euros,
respectively. The questionnaire used for the survey is shown in Appendix B.
7
The NAISC used in this thesis were obtained from the NAICS Association web site
(http//www.naics.com/search/), by extracting the two first digits from the respective industry codes
returned by the search engine.
92
Financial Services
3%
Transportation 4% Other 2%
Food & Agriculture
Retail 6%
16%
Telecom 7%
Pharma, Chemical
Machinery 16%
8%
Electronics 8%
Energy/
Resources/
Automotive 8% Material 12%
Construction 10%
throughout the research process, are no less important when doing quantitative
research. They are as follows:
At a first glance, these questions may sound trivial and obvious. However, after a
second thought, one realizes that they are indeed very pragmatic and useful. The
questions are also in line with what Saunders (1997) states as important issues to
address in a methodology chapter. Simply put, a conscious mind is essential when
doing research. Without it, there is no possibility to find an optimal path towards
the destined goal. Preferably, the research process starts with a set of concrete
research questions that together serve as a beacon for navigation. Furthermore,
careful planning of the different steps in the research process is also very
important.
In other words, the research process pretty much resembles of classical project
planning and execution. Quantitative research is generally highly focused and the
methodology is predetermined. Qualitative research, on the other hand, has often a
more probing and exploratory nature, and the research process is often redefined
and modified throughout the course of the research (Hussey & Hussey, 1997).
Ontology
Nominalism Realism
Epistemology
Anti-positivism Positivism
Human nature
Voluntarism Determinism
Methodology
Idiographic Nomothetic
3.3.1. Ontology
Ontology, which comes from the Greek word for “being”, addresses the very
essence of the phenomenon subject of study in the research (Hollis, 1994).
Ontology aims at answering whether there is an objective reality, or if the reality
is in the eye of the beholder.
In social sciences, relationships among entities are many times very diffuse and
extremely complex. The difficulty to measure many properties, such as
intelligence or trust implies that social science contains a significant degree of
subjectiveness is referred to as “nominalism”. This means that the “truths” are just
created by individual cognition and social communication of the ideas. On the
other side of the scale, one could for example put natural sciences such as
chemistry or physics which measure quantities that are relatively simple to assess.
Consequently, natural sciences belong to what is referred to as “realism”.
Business science, on the other hand, to which the LCCS topic belongs, probably
lies somewhere in between on the scale, as there is a mix of quantities to measure.
Some are easy to quantify, for example return on investment, cash flow, or profit.
Others, such as innovativeness, flexibility, or competitiveness, are much more
difficult to grasp, and are thus of a more nominalistic nature. With this in mind,
the research in this thesis is ontologically anchored relatively far towards the
realism endpoint, as objective facts are to be obtained.
96
3.3.2. Epistemology
Epistemology deals with the production and or acquisition of knowledge, what is
true and false, and how knowledge is communicated. Here, the two extremes
argue either that the knowledge about the world can be proved or rejected,
identifying laws of cause and effect (“positivism”), or identifying subjective
knowledge of individuals (“anti-positivism”) 8 . Positivism tries to produce
knowledge through verifying or falsifying hypotheses; in the end, observations
made will ultimately distill down to what is considered true knowledge.
Positivism has close ties to natural sciences where the observer is passive and
does not interfere with the phenomenon. The heavy focus on cause and effect
therefore stresses the ”how” questions rather than the ”why” questions. Through
reduction, the phenomenon subject of study can be simplified and easier to
investigate. The logic boils down to the conclusion that there is no difference
between something that cannot be observed and something that does not exist at
all. Relativism, in contrast, aims at probing a problem from within, rather than
being a passive observer. What is more, not only the problem itself is of
importance, but also the surrounding environment. Consequently, a holistic
approach is preferable when conducting qualitative research.
For the research in this study, the epistemological dimension is primarily towards
the positivistic extreme, as the research specimen is modeled through deduction of
a number of hypotheses that are tested quantitatively. It is not purely positivistic
though, as chapter 4 is relatively holistic and tries to give a complete rather than
reduced picture of the LCCS phenomenon.
8
It is worth noting that the terminology used varies from author to author. Other terms for
“positivism” are for example “representativism” or “objectivism”, while for “anti-positivism”
terms such as “relativism”, “constructivism” or “relativism” have been used. See for example
Rüegg-Stürm (2003) for an alternative terminology.
97
In a business setting, this dimension is very difficult to assess. If the time frame
through which environmental factors are considered is extended to infinity, one
could of course argue that the situation is completely deterministic (i.e. behavior
inherited genetically is a result of evolution). With the time frame equivalent to
the life span of an average individual, the truth probably lies somewhere in
between. That means that individual actions are to some extent determined by the
surrounding environment, and to some extent to free will.
3.3.4. Methodology
The three philosophical dimensions above have all significant impact on the
research methodology since they determine for example how to formulate
research questions, apply theory, collect empirical data, as well as analyze and
interpret empirical data. For example, if one relies on subjectivist approaches to
social science, the case study method, which involves a small group of research
subjects that are qualitatively investigated, should be highly suitable (Larsson,
1993). On the other hand, if an objectivist approach is applied, where a large
sample is examined quantitatively, the nomothetic survey method would be
preferred (Larsson, 1993).
Without doubt, the research in this thesis belongs to the latter category, especially
chapter 5, where the number of variables is small in comparison to the sample size.
Although the methodology applied in Chapter 4 is more of a probing nature, it is
yet quantitative and performed through the same survey, thus making it
98
Nominalism Realism
Anti-positivism Positivism
Voluntarism Determinism
Idiographic Nomothetic
The research of this thesis concerns low-cost country sourcing, which repeatedly
has been claimed to have a significant impact on strategic and operational
performance (Alguire et al., 1994; Kotabe, 1994; R.J. Trent & R. M. Monczka,
2002). The phenomena are trends and implications of low-cost country sourcing,
which in themselves set the focal point of the research. These two phenomena
consist of several parts that need to be broken down and analyzed systematically
while not losing the big picture. Despite its status as “buzz word” in business
contexts, it cannot survive and gain acceptance without substantial scientific
underpinnings over the long term. Thus, this study aims at investigating low-cost
country sourcing as a strategic and viable concept for generating sustainable
competitive advantage.
First, the research design was elaborated in chapter 3.1. The research design is
centered on four design parameters, each of which forms a continuum between
two extremes. These are pure/applied research, primary/secondary research,
theoretical/empirical research, and descriptive/explanatory research. None of the
design parameters were configured at either extreme, but rather somewhere in
between, i.e. the research utilize several techniques and approaches. In sum
though, it can be concluded that the research in this thesis is more applied than
theoretical, uses more primary than secondary data, more empirical than
theoretical, and approximately equally descriptive as explanatory.
Chapter 3.2 laid the groundwork for how to go about the research by developing a
research strategy. The strategy was developed according to a generally accepted
approach, which consists of six process steps, namely (1) observation of a real-life
phenomenon, (2) literature review and identification of a problem, (3) hypothesis
and theoretical framework development, (4) operationalization of constructs and
research model development, (5) data collection and analysis, and finally (6)
hypothesis testing and interpretation of results. A feedback loop from the last
phase ensured that the theoretical underpinnings could be altered depending on
the outcome of the analysis. The empirical study was conducted in the shape of
expert interviews which served as a pre-study, followed by a large-scale survey
involving 200 CPOs at large-sized firms from France, Germany, Italy, Spain and
the UK.
In chapter 3.3 philosophical aspects of the research were also considered, namely
ontology, epistemology, human nature, and methodology. They form four distinct
continua, similar to the research design parameters in chapter 3.1, and the
extremes on each continuum are nominalism/realism, anti-positivism/positivism,
voluntarism/determinism, and idiographic/nomothetic research method,
respectively. By configuring the research along these four continua, fundamental
questions such as how the world is perceived, how knowledge can be obtained,
and what the most appropriate research method is, can be answered.
101
There are many driving forces behind globalization, but they can generally be
condensed to four main categories, namely political factors, socio-economic
factors, technological factors, and macro-economic factors (Figure 23). Political
factors refer to regulations and policies set by local and national governments,
such as trade barriers, tax rates and labor laws. Socio-economic factors deal with
characteristics of the population in a nation, for example educational levels,
language proficiency, cultural openness, and so on. Technological factors includes
use of new innovations that can help reducing communication and transportation
costs, such as wireless and information technologies, new and improved means of
102
On the supply side, this has led to a situation where goods and services are
sourced more and more on a global scale, with less attention to national borders.
In other words, a shift in sourcing patterns is taking place. Most notably, we will
see that China is more important than the home country to French companies, and
this pattern will be even more pronounced by 2010. Moreover, according to the
statistics obtained, historical trade patterns still remain, at least to some extent.
Tied to geography, culture and language, data shows that companies still source
extensively in their home countries and neighboring countries, with a few
important exceptions.
103
So how are the trends manifested in the purchasing and supply management
literature? A study by Trent and Monczka (1998) indicated that the trends take
place at the following levels:
Clearly, the key drivers can be found at many levels throughout firms, industries,
and the economy as a whole. The following sub sections deal with the different
sub trends more in-depth.
organization, but also through cooperation with external suppliers. While quality
and cost have always been perceived as important, time-related factors (i.e. speed)
have become the next target for efficiency improvements.
In line with the old adage “You cannot manage what you cannot measure”
(Kaplan & Norton, 1992), performance measurement is necessary for gauging the
success rate of functional and team-based strategies and plans. More in-depth,
purchasing managers should deploy performance measurement to assess the
following (Monczka et al., 1998):
The current trend aims at focusing more on effectiveness rather than efficiency, i.e.
contributions to top-line results through sourcing of innovations etc. This also
indicated a paradigm shift where the perceived strategic importance of purchasing
and supply management is on the rise.
Not very long time ago, it was common practice for buying firms to play suppliers
off against each other. Moreover, it was also common to switch suppliers on a
105
regular basis and only offer short-term contracts in order to induce competition
among them. Today, practices have significantly changed and buying firms have
more diverse range of relationships with their suppliers. The trend is indicating
that most companies will continue to reduce their suppliers and instead
developing the relationships with preferred suppliers (Trent & Monczka, 1998).
Hand in hand with this trend is another related trend which implies reliance on
larger full-service suppliers that have capabilities to design and build entire
subsystems, contrary to the alternative approach to have many smaller suppliers
providing components of the subsystem (Trent & Monczka, 1998). Often, supplier
base reductions involved picking a smaller group of suppliers out of the original
supplier base. Potentially, the supplier base reduction could have been performed
more effectively if it also included involvement of new suppliers, especially from
other geographic regions, such as LCCs. Supplier base reduction can facilitate the
focus on core competencies, ultimately improving effectiveness and efficiency
(Jahns, 2005).
While these strategy types are closely interrelated, they are unique in terms of
their execution and impact on the purchasing and supply function. At the first
level, the purchasing and supply function develops and executes its own strategies
and, whereas at a higher level the strategies might turn out to be dysfunctional for
the organization as a whole. In other words, what is optimal for the purchasing
106
and supply function might not be optimal for the company. At the second level the
purchasing and supply strategies are designed to support the corporate strategy to
the greatest extent possible. Finally, at the third level purchasing and supply
management is an integral part of the corporate strategy. The alignment of the
sourcing strategy is an important aspect when sourcing in LCCs, as will be shown
in chapter 4.2.
4.1.3. Organization
Having an adequate organizational structure is essential for successfully
implementing a viable sourcing strategy. Today this usually implies collaboration
not only between functional units and business units within the firm, but also with
preferred suppliers. The need to achieve bundling effects for purchase spend has
led to a concentration of buying centers, resulting in hybrid organizational
structures. This reflects an increased need for purchasing and supply management
to become more integrated with other parts of the organization.
As such, this chapter uncovers how companies are formulating visions and
strategies in order to maximize the value of their end-customer products and
services through low-cost country sourcing. As a starting point, the Supply
Management Navigator (SMN) was used as an analytical framework (Jahns,
2005). It contains the necessary information for effective and efficient
purchasing/supply management practices. Since the target group for this study
primarily consists of CPOs, the areas of investigation focused on the management
module (Figure 24).
A chief procurement officer at senior management level has the power to set a
corporate vision for global sourcing activities that can later be effectively
translated into strategies for individual business units (Monczka et al., 1998), and
to serve internal customers (D. F. Lawrence, Hopkins, & Bhonsle, 2005). In this
context, the vision itself serves as a starting-point for definition of important
sourcing markets, and how these can be exploited in order to optimize the
corporate value chain (C. R. Carter, 1998). A clearly articulated vision is essential
for the long-term orientation of supply side activities and for the conformity to the
overarching corporate philosophy (Module 1 in the SMN).
According to the LCCS strategy results (Figure 26), it is clear that the most
widely taken action is to align the LCCS strategy to the corporate strategy.
Companies are also adapting the LCCS strategy to threats and opportunities in the
external environment to an equal extent. Implemented to a slightly lesser extent,
companies have derived strategies for how to source individual supply categories
in LCCs. The majority of participating companies have also aligned their LCCS
strategies with other functional strategies such as R&D, marketing etc. This
indicates that functional departments within the organization are collaborating to
some extent. In second to last place, we find integration of strategic suppliers into
business processes such as early supplier involvement only implemented to a high
or very high extent among 30 percent of the respondents. Lastly, organizational
changes to support LCCS are the action that is taken to the least extent.
110
Align LCCS
Adapt LCCS strategy
strategy to the
to threats and
corporate
opportunities strategy
of sourcing markets 8% 16% 28% 31%
31% 17%
AlignAdapt LCCStostrategy
LCCS strategy to
the corporate
threats and opportunities
strategy 8% 15% 27% 34%
34% 16%
Not at all
Derive
Derive strategies for individual
special strategies for individual Not at all
11% 18% 28% 30%
30% 13%
LCCS
LCCS supply
supplies andcategories
services To some extent
To some extent
Link LCCS
Integrate strategy
strategic to other
LCCS suppliers into ToTomoderate
a moderate extent
extent
business processes 17% 19% 28% 25%
25% 11%
functional strategies ToTohigh extent
a high extent
To very high extent
Do organisational changes toLCCS
Integrate strategic support To a very high extent
LCCS initativesprocess 21% 24% 24% 22%
22% 8%
suppliers into business
Do
Linkorganisational changes
LCCS strategy to other to
functional
9% 22% 31% 32%
32% 6%
supportstrategies
LCCS initiatives
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Share of respondents in percent
As can be seen in Figure 27, it is evident that companies are primarily focusing on
the quality of goods and services sourced (i.e. conformity to specifications); 56
percent of the companies do this to a high or very high extent. Total cost savings
is measured to a slightly lesser extent, but is still frequently evaluated among the
majority of the respondents. The reason why this number is not higher might be
due to difficulties in assessing enterprise-wide spend data. Delivery reliability is
measured more or less to the same extent as order lead time (measured to a high
or very high extent by 35 percent of the responding companies). Percentage of
spend sourced in LCCs is found in last place (measured to a high or very high
extent by 31 percent of the respondents).
111
Percentage of spend
sourced in low-cost1 20% 18% 31% 21% 10%
countries
4.2.4. Organization
Having the right organization is a crucial aspect of any company (Module 6 in the
SMN) (Barney, 1997; Barney & Hesterly, 1996). This holds especially true when
sourcing in LCCs – it does not matter how good the strategy is if the organization
is not able to execute it. In order not to sub-optimize overall company
performance, it is important that functional departments can cooperate and
coordinate activities at an appropriate level (Dobler & Burt, 1996; Leenders et al.,
2002).
R&D 61%
Marketing 47%
Manufacturing 42%
Warehousing/Distribution 27%
As can be seen in Figure 28, the quality assurance department and R&D
department are, on average, involved in 61 percent of cross-functional purchasing
teams in the participating companies, closely followed by the marketing and the
finance/controlling department at 47 and 45 percent, respectively. The
manufacturing department is slightly less frequently involved with 42 percent
involved in cross-functional purchasing teams.
So now we know that quality assurance and R&D are most commonly involved in
cross-functional purchasing teams. But who is initiating and driving LCCS
initiatives within the business? A closer look at Figure 29 shows that 34 percent of
the respondents claim that the CPO is most commonly initiating LCCS activities.
This comes with no surprise, since the CPO usually has the ultimate responsibility
for the purchasing/supply function (Dobler & Burt, 1996; Leenders et al., 2002).
Next follow the CEO and COO, which are also relatively frequently initiating and
driving the LCCS operations. The CFO is involved less often, as is the case with
internal and external customers. Although not a common practice, LCCS activities
can sometimes be driven by other stakeholders such as staff at the R&D
department.
t
(CPO) 34%
r
(CEO) 21%
r
(COO) 14%
r
(CFO) 8%
s
Customers 7%
Other
r 16%
individual sourcing market. Which type of purchasing channel is most suitable for
a given situation is such a highly complex issue; a whole book could be written on
the topic. Consequently, as this part of the thesis is a descriptive study, the aim is
to examine to what extent different purchasing channels are being used by large-
sized companies today, thus focusing less on the underlying determinants.
Nevertheless it can be stated that one of the most important decisive factors for
the choice of purchasing channel is the purchase volume from a specific sourcing
market (Carduck, 2000). Normally, with a purchase volume exceeding 100
million Euro per year, an international procurement office (IPO) can be justified.
Figure 30 shows the current situation among the responding companies.
Direct
Direct purchasingfrom
purchasing from LCC
LCC
25%
supplier supplier
Purchasingthrough
Purchasing through 3
3rd party
party
rd
19%
intermediary
intermediary
Purchasingthrough
Purchasing through IPO
IPO 16%
As can be seen in Figure 30, the predominant way to purchase from LCCs is
through direct purchasing from suppliers in the LCC. The second most popular
approach is to make use of another subsidiary of the parent company for
purchasing activities. Although associated with the highest risks, these two
alternatives provide the best opportunities for close collaboration and
development of suppliers (Carduck, 2000). Closely followed on third place, we
find the alternative to purchase from sales subsidiaries belonging to the supplier in
the home country of the buying firm. Next follows purchasing through third-party
intermediaries such as trade houses, brokers and agents. Although a less risky
114
Figure 31 shows that the overall knowledge level of various topics associated with
LCCS is relatively moderate among the participating companies. Only knowledge
about cross-functional collaboration stands out of the crowd. Moreover, as an
example, for knowledge about LCCS strategy development, 41 percent of the
responses belongs to the “moderate” category; such an important core activity is
expected to score much higher.
If one is not sure whether a newly developed strategy is optimal, then there is
naturally a risk of sub optimal results. In line with the data obtained, more non-
strategic topics about supportive systems, such as supply systems, were expected
to score relatively low; these are often subject to outsourcing to external service
providers. As other research has shown that fifty percent of all offshoring projects
do not meet expectations (Cohen & Young, 2005), a lack of knowledge could
potentially be one of the sticking points.
115
8 6%
cross-functional collaboration 17% 26% 40%
40% 12%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Share of respondents in percent
The sourcing pattern for France is very interesting (Figure 32 and Figure 33). As
shown in the chart, China is currently considered significantly more important
than the home country (scoring 49 vs. 32 percent for direct spend and 45 vs. 37
percent for indirect spend). Moreover, the gap between the two countries will
increase even further over the next five years, with a score of 62/65 percent (from
now on written as “direct/indirect spend”) for China and a small drop from 32/37
percent to 27/35 percent for France. Italy, Poland and Thailand will remain
relatively important, whereas Germany is expected to fall in importance for
sourcing of direct supplies.
70%
62%
60%
Percentage of total responses for top 3
most important sourcing countries
49%
50%
40%
32%
30% 27%
25% 24% 23%
22% 21%
20%
16%
15%
13%
12% 11% 11%
8% 8% 8% 8%
10% 7%
0%
nd
a
nd
na
a
ly
y
K
ry
ce
si
an
di
U
Ita
ga
la
la
hi
an
ay
In
m
Po
ai
C
un
Fr
al
er
Th
M
H
G
Figure 32. Top ten sourcing countries for direct spend among French companies
117
60%
Percentage of total responses for top 3 53%53%
most important sourcing countries
50%
40% 37%
33%
0%
y
K
ia
n
y
d
a
ep
a
ke
nc
an
U
ai
n
n
d
si
la
hi
In
Sp
R
r
us
a
Tu
Po
C
Fr
ch
er
R
ze
C
2005 2010
Indirect spend
Figure 33. Top ten sourcing countries for indirect spend among French companies
The situation among the German companies (Figure 34 and Figure 35) is
interesting in the sense that domestic suppliers have a very dominant position
today with 53/69 percent. Somewhat surprisingly, the situation is expected to
remain until at least 2010 (scoring 50/67 percent). Hence, the discussion about an
offshoring wave in Germany may be exaggerated (Baily & Farrell, 2004). The
domestic suppliers seem to remain internationally competitive on a medium term.
In line with expectations, China will drastically increase in importance, apparently
at the expense of Southern European and Central European suppliers. A relatively
high gain in importance is expected for India as well. To sum up, it seems that the
supplier base currently located in Europe, outside Germany, will partially shift
towards low-cost regions in Asia. Interestingly, as opposed to other European
countries, Germany sources a comparatively high number of goods and services in
Turkey. Although industrial structures may play an important role, one cannot
neglect the large share of Turks who live in Germany. Most likely, their
knowledge about the Turkish language and culture facilitates sourcing operations
there.
118
80%
60% 56%
53%
50%
50%
40%
34%
31% 32%
28% 28%
30% 26%
19% 18%
20% 15% 16%
13% 13%
9% 10% 9%
10% 6% 6% 6%
3%
0%
nd
a
.
na
ly
y
ia
ey
il
ce
ep
an
di
az
ak
Ita
U
la
rk
hi
an
In
R
m
Br
Po
ov
C
Tu
Fr
ch
er
Sl
G
ze
C
Figure 34. Top ten sourcing countries for direct spend among German companies
80%
Percentage of total responses for top 3
69%
70%
most important sourcing countries
67%
60%
50%
39%
40% 36%
31% 30%
29% 28% 27%
30%
23% 23%
20% 18%
12% 13%
10% 10% 10%
10% 6%
4% 4% 3% 3%
0%
ly
K
e
ia
.
y
y
ep
ke
nc
an
It a
ar
ki
U
n
d
la
hi
In
g
R
r
a
m
Tu
ov
Po
un
C
Fr
ch
er
Sl
H
G
ze
C
2005 2010
Indirect spend
Figure 35. Top ten sourcing countries for indirect spend among German companies
119
Among the Italian companies, only minor changes are expected (Figure 36 and
Figure 37). The importance score of the home country is expected to decrease
from 49/65 percent to 40/60 percent. A similar development is expected for
Germany and France, although a sharper decline is estimated for German
suppliers. Furthermore, it is interesting to note that German suppliers have a very
strong position among Italian companies as the most important sourcing country
today. However, this situation will change as China climbs to the number one
sourcing country for direct supplies by 2010. The estimated developments for
indirect spend are less dramatic, with slight declines for suppliers in Western
Europe and increasing importance for Central European suppliers.
80%
Percentage of total responses for top 3
most important sourcing countries
70%
60%
56%
49%
50% 45%
41% 40%
40%
30% 29%
30% 25%
20% 17%
13%
9% 8%
10%
4% 5% 4% 3% 4% 5% 4% 5%
0%
nd
a
ly
na
.
y
ey
il
ce
ep
di
an
az
U
Ita
la
rk
hi
an
In
R
Br
m
Po
C
Tu
Fr
ch
er
G
ze
C
Figure 36. Top ten sourcing countries for direct spend Italian companies
120
70%
65%
40%
30%
30%
22% 22% 21%
20% 19%
20%
8% 7% 9% 8% 8%
10% 6% 6% 6%
0%
nd
na
ly
.
y
s
K
ey
ce
ep
an
nd
ai
U
Ita
la
rk
hi
an
R
Sp
rla
m
Po
C
Tu
Fr
ch
er
he
G
ze
et
C
N
Figure 37. Top ten sourcing countries for indirect spend Italian companies
The most dramatic change to the sourcing pattern is estimated to take place in
Spain (Figure 38 and Figure 39). Today, domestic suppliers are considered most
important and China is ranked as number two and four for direct and indirect
supplies, respectively. By 2010, however, China is expected to take the lead as the
most important sourcing country for both spend categories. Here, it is also
interesting to see that Mexico has significant importance, verifying the fact that
language acts as a facilitating factor when sourcing. Interestingly, India scores
higher in Spain than in any of the other countries where the study was conducted.
121
50%
46%
45%
Percentage of total responses for top 3
most important sourcing countries
40%
35%
35% 32%
30%
30%
26%
24%
25% 23% 23%
20%
20%
14% 14%
15% 13% 13% 13%
12% 12% 12%
10% 8% 8% 8%
5%
0%
nd
a
na
.
y
n
ey
ce
ep
an
di
ai
ic
U
la
rk
hi
an
In
R
ex
Sp
m
Po
C
Tu
Fr
ch
M
er
G
ze
C
Figure 38. Top ten sourcing countries for direct spend among Spanish companies
50%
45%
Percentage of total responses for top 3
45%
most important sourcing countries
40% 38%
35%
35% 32%
30% 27%
25% 24%
22% 22% 23%
20% 18%
15%
15% 13%
11% 11% 11%
10% 10% 10%
10%
6%
5%
5%
0%
ico
K
ce
ia
.
il
n
d
a
ep
az
an
U
ai
n
n
d
an
ex
la
hi
In
Sp
Br
m
Po
C
M
Fr
ch
er
G
ze
C
Figure 39. Top ten sourcing countries for indirect spend among Spanish companies
The sourcing pattern in the UK is also interesting (Figure 40 and Figure 41). As in
the case with Germany, it seems that UK suppliers will manage to fend off global
122
50%
45%
45%
Percentage of total responses for top 3
most important sourcing countries
41%
40% 37%
35% 32%
30%
25% 25%
25% 24%
20% 20%
20% 17%
0%
SA
ey
K
ia
.
y
d
a
ep
a
nc
an
U
n
n
d
si
rk
la
hi
In
R
us
a
Tu
Po
C
Fr
ch
er
R
G
ze
C
Figure 40. Top ten sourcing countries for direct spend among UK companies
123
60%
53% 53%
Percentage of total responses for top 3
most important sourcing countries
50%
40% 37%
33%
0%
y
K
ia
n
y
a
ep
a
ke
an
nc
U
ai
n
n
d
si
la
hi
In
Sp
R
r
us
a
Tu
Po
C
Fr
ch
er
R
ze
C
2005 2010
Indirect spend
Figure 41. Top ten sourcing countries for indirect spend among UK companies
When looking at the importance of specific low-cost regions (in this case sub-
continents), it is evident that the importance of these will increase until at least
2010 (Figure 42 and Figure 43 show respondents rating of regions on a five-grade
scale). Among the companies studied, the average importance score for low-cost
regions will increase from 2.4 to 2.8 for direct supplies and from 2.0 to 2.4 for
indirect supplies. The biggest relative increase in importance is expected to occur
in East Asia, South Asia and Eastern Europe. Less importance increases are
expected for Central Europe, Middle East and Africa. Although East Asia is
estimated to have the biggest increase in importance, Eastern Europe will be
perceived as most important when considering both direct and indirect spend. This
is probably due to the low comparative price level in combination with the
relatively low geographic distance to Europe, making it a viable option for
supplies with moderate time requirements. China, here included in the East Asia
category, would probably score much higher on its own. Most likely, the same
applies to countries such as India, which is included in the South Asia category
for the purposes of this study.
124
5,0
4,5
4,0
3,0
2,9
3,0 2,8 2,8
2,7
2,5 2,4 2,4 2,4
2,5
2,1 2,1
1,5
1,0
Central Eastern East Asia South South- Central Middle Africa Average
Europe Europe Asia East Asia and Latin East
America
2005 2010 Direct spend
5,0
4,5
4,0
3,5
Rating Score
3,2
3,0
3,0 2,8
2,6
2,5 2,4
2,5 2,4
2,2
2,1 2,0 2,0 2,0
2,0
2,0 1,8 1,8
1,7
1,6
1,5
1,5
1,0
Central Eastern East Asia South South- Central Middle Africa Average
Europe Europe Asia East Asia and Latin East
America
40%
35% 34%
31%
30% 29%
27%
Percentage LCCS Spend
25%
22%
19% 19%
20% 18%
14% 14%
15%
10%
5%
0%
UK Italy Germany Spain France
2005 2010
Direct spend
40%
35%
30%
Percentage LCCS Spend
25%
22% 22% 22%
21% 21%
20%
16% 16%
15% 14%
10%
10%
7%
5%
0%
France Italy UK Germany Spain
Indirect spend
France, on the other hand, is expected to increase its share of direct LCCS spend
by more than one hundred percent (from 14 to 29 percent) over the next five years,
but only very little when it comes to indirect spend. A similar pattern applies to
the other countries, where most of the increases in LCCS spend stem from
sourcing direct supplies and less from indirect supplies. Italy and the UK, on the
other hand, with a relatively higher degree of low-cost country sourcing today, are
estimated to increase their degree of low cost country purchases to a lesser extent,
both expressed in absolute and relative terms.
to increase their direct and indirect LCC purchase spend by 64 and 34 percent,
respectively.
Comparing direct and indirect spend, it is worth noting that the difference
between the two spend categories is not significant today, but direct low-cost
country spend is expected to experience a higher relative growth until 2010. One
would normally expect indirect spend to be more suitable for sourcing in LCCs
due to its relatively low strategic importance, but that argument does not seem to
hold in this situation. A possible explanation is that direct supplies usually are of
higher value and hence it is possible to realize larger relative cost savings for that
type of supplies (Battacharya et al., 2004).
Direct spend
Energy/Resources/Material
40
35
Pharma/Chemical Food and agriculture
30
25
20
15
Machinery Telecom
10
5
0
Construction Electronics
Retail Transportation
Then follows food and agriculture (22.5 percent), telecom (21 percent),
electronics (19 percent), transportation (18 percent), automotive (17 percent),
128
retail (16.5 percent), construction (15 percent), machinery (12.5 percent) and
pharma/chemical (10 percent).
Indirect Spend
Energy/Resources/Material
40
35
Pharma/Chemical Food and agriculture
30
25
20
15
Machinery Telecom
10
5
0
Construction Electronics
Retail Transportation
Automotive
2005 2010
On the other hand, taking the retail industry as an example, the end-product
quality is of secondary importance. In combination with the relative ease to enter
markets (Bresnahan, Reiss, Willig, & Stigler, 1987), entry barriers are low, which
is reflected in the relatively low profit margins. As a result, economies of scale,
129
Similar to the retail industry, we find industries such as the electronics, telecom,
machinery and automotive industries, all of which are experiencing high
competition from domestic producers in emerging markets (Nellore, Chanaron, &
Söderquist, 2001). Consequently, companies in these industries have been forced
reshape their global strategies over the last decade and source in LCCs to an
increasing extent in order to bring down cost (Alguire et al., 1994; Bozarth et al.,
1998). The moderate LCCS growth in the telecom sector can be explained by the
high product and service complexity, where Western incumbents on the global
arena such as Cisco Systems, Nortel and Ericsson still hold a relatively strong
market position against green fielders such as Chinese Huawei (Nissen, 2005).
The LCCS pattern for indirect supplies is very different from the direct supplies,
both in absolute and relative terms. Not only the amount of spend sourced today is
lower, but the growth over the next five years is also much lower. The most
important explanation to this is the huge diversity and complexity of indirect
spend, since this category frequently involves different types of services. The
often intangible and tacit nature of services makes the whole sourcing process less
efficient. The difficulty is due to the fact that it is much more difficult to asses the
quality of services than of goods, since one cannot easily order a “service sample”
for evaluation. Instead, one often has to rely on past track records and business
reputation of service providers.
This effect is even more pronounced when sourcing in LCCs where the business
environment is characterized by cultural and language barriers, time zone
130
For obvious reasons, standardized supplies are also considered suitable for
sourcing in LCCs. Firstly, seen from a market perspective, a company can achieve
large economies of scale if it can sell final products on several consumer markets
with only minor modifications. This characteristic goes hand in hand with the
characteristic found in forth place, namely supplies that can be used as inputs for
manufacturing for the local market. Secondly, standardized supplies are easier to
specify and specifications are easier to interpret, which speeds up the whole
sourcing process and facilitates high quality. The pharmaceutical and beverage
industries are examples in which production inputs are standardized and
transportation costs are relatively high thus yielding extensive local production
through a decentralized organization (Brainard, 1993; Horstmann & Markusen,
1992; Pontes, 2004).
In third place, we find supplies where a large local supplier base already exists.
The advantages are quite obvious. With a large local supplier base, competition is
higher, which in turn facilitates lower purchase price and better quality due to
higher effectiveness and efficiency of the suppliers (Porter, 1990). It also renders
better bargaining power of the buyer and less risk, since multiple sourcing and
supplier shifting can be done more easily.
Supplies with high technological contents prove not to be among the most suitable
supplies for sourcing in LCCs for a number of reasons. For example, supplies
9
During a guest lecture for the course “Strategic Value Chain Management” at European Business
School, Oestrich-Winkel, Germany, summer semester 2005, Siemens Com CPO Dietmar Dresp
claimed that quality is no longer an issue; it is only a matter of having the right quality
management systems in place.
132
with high technological content are more dependent on timely delivery due to
their higher value and shorter life cycle. Moreover, high-tech supplies contain
proprietary knowledge, and, for obvious reasons, the buyer often does not want
such knowledge to be dispersed uncontrolled. This holds especially true in many
LCCs, where intellectual property rights and copyrights have little support from
local jurisdiction (Haley et al., 2004).
This statement is further supported by the two final supply characteristics, namely
time-critical supplies and supplies with high intellectual property content, scoring
a meager 2.65 and 2.48, respectively. High-tech supplies are also more difficult to
specify, which makes supplier relationship management a crucial task. Supplies
for mature products (late in product life cycle) prove not to be very suitable for
LCCS; most likely, this is due to the low profit margin of such products, which
cannot offset initial overhead costs in the sourcing process when setting up a
buyer-supplier relationship (Battacharya et al., 2004).
30%
28%
25%
5%
0%
Goods Services
As can be seen in Figure 49, goods are sourced to a 70 percent greater extent than
services. This is most likely due to the complex nature of services as discussed in
the previous paragraph. Today, 17 percent of goods are currently sourced in low-
cost countries, with an estimated increase of 64 percent over the next five years.
Hence, by 2010, the amount of goods sourced in low-cost countries is estimated to
reach a staggering 45 percent. Sourcing of services in low-cost countries, on the
other hand, is estimated to increase from 10 percent today, to more than 25
percent by 2010.
If taking a closer look at individual commodity categories (Figure 50), one can see
that electronics is most widely sourced in LCCs today. It is also the supply
category that will be most widely sourced by 2010; with spend for electronics
estimated to increase from 13 percent today to 16.5 percent in five years as a
percentage of total category spend. Purchase spend for the packaging, metals and
transportation categories will approximately amount to 10 percent by the year
2010. However, the plastics/rubber, engineering/construction and MRO/services
categories are all estimated to decrease by roughly 5-10 percent.
134
13.0%
Electronics
16.5%
11.9%
Plastics/Rubber
10.9%
11.7%
Transportation
12.5%
11.4% 2005
Metals
12.9% 2010
10.7%
Packaging
11.7%
8.5%
MRO
8.3%
0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% 18.0%
Percentage of total category spend sourced in LCC
Figure 50. Different supply categories sourced in LCCs 2005 and 2010
assess and pose a very high impact on business operations if the associated events
would occur.
y
Political instability 3.72
Increased comparative
price levelss 3.58
n
Criminality and corruption 3.52
Currency Risks
s 3.51
Prioritization of other
s 3.27
business initiatives
Costs for exiting
s 2.86
legacy assets
s
Quality problems 2.78
At the top, we find political instability as the most important risk to consider when
sourcing in low-cost countries. It includes any action at a governmental level that
can cause uncertainty and interruptions to the business operations – ranging from
raising tariff and non-tariff barriers to the event of civil or international war. The
most effective way to mitigate this risk is through risk avoidance, i.e. limiting the
operations in such regions where political instability is prevalent. As an example,
China’s openness to business over the last decade, along with the country’s entry
in the WTO has significantly improved political stability and reduced uncertainty
for foreign companies (Chow, 2001).
Increased price levels in LCCs are ranked as the number two risk factor by the
respondents. This risk has a very clear financial impact on operations, since high
inflation rates can erode profit margins in the long term. Fortunately, price level
increases are relatively easy to assess thanks to the vast amount of macro-
136
economic data readily available.10 Taking China as an example again, its high
inflation rate of 5 percent on a year-to-year basis should not pose any threats as
most of the inflation stems from increases in grain prices (31.8 percent last year)
(Dashan, 2004). Moreover, as three quarters of the Chinese population still live in
rural areas (Ravallion & Chen, 2004), the pool of low-wage labor will remain
decades ahead.
Ranked as number three, criminality and corruption can of course have severe
impact on at several levels of society such as economic, social and political
(Shelley, 2005). The probability of criminality is difficult to assess in specific
cases although statistics are easy to obtain at a regional level. The impact usually
ranges from frauds and scams to theft and burglary in corporate properties.
Corruption, on the other hand, can be assessed by examining common business
culture. While corruption is generally less prevalent in the western hemisphere, it
is considered more of a risk in less-developed regions as a result of poverty and
extensive government involvement in the public (Goolsarran, 2006).
Currency risks (i.e. currency volatility) at fourth place can have very big impact
on businesses, especially for companies with a high degree of foreign trade. Since
currency volatility can be easily assessed through extensive economic statistics
publicly available and mitigated through hedging mechanisms such as currency
futures or taking positions in the currency spot market (Dales & Meese, 2001), the
risk was actually perceived to be more important than expected.
Lack of supplier skills and knowledge in fifth place includes all levels of the
supplying organization: from the staff working in manufacturing and assembly or
production of services to the management level. The buying firm should strive to
select high-performing suppliers that can provide knowledge and complementary
competencies (Goffin, Lemke, & Szwejczewski, 2006; Johnsen, Phillips,
10
There are several excellent online resources with macroeconomic data readily available, for
example United Nations Economic Commission for Europe (UNECE, http://w3.unece.org), the
Statistical Bureau of Statistics (Eurostat, http://epp.eurostat.ec.europa.eu) or US Census Bureau
(http://www.census.gov/)
137
Caldwell, & Lewis, 2006), preferably those who have a proven track record.
While cost savings and quality cannot be expected to be realized immediately
(Dobler & Burt, 1996), long-term supplier development is dependent more on the
capabilities of the buying firm than the supplier skills. Moreover, on-site supplier
visits and continuous quality inspections are a must for success.
Intellectual property risk is a hot topic in the context of low-cost country sourcing.
Although it did not rank higher than sixth place, it is likely that the participating
companies have already mitigated this risk through risk avoidance, i.e. not
sourcing supplies with high intellectual property content, which has also been
noted in literature (Battacharya et al., 2004). It could of course also mean that
companies have not yet paid attention to this risk. This hypothesis is further
supported by the low sourcing suitability rating depicted in Figure 50.
Price erosion, caused by shifting the supplier base to low-cost regions, means that
achieving competitive advantage through lower purchase price will become more
and more difficult over time. Instead, internal capabilities such as supplier
development, inventory control and effective quality management will play an
increasingly important role, as well as product strategies such as product bundling
(Lancioni & Gattona, 1992). Multinational enterprises that today pursue proactive
supply management are very well aware of this and develop strategies to cope
with these challenges, which is most likely the reason for the relatively low rating
(3.48).
At second from bottom in Figure 11, costs for exiting legacy assets prove to be
not a very serious issue either, although they could imply a financial burden
(Battacharya et al., 2004). The risk assessment can be made easier since the risk is
138
completely in the hands of the company itself, thus only imposing a specific
risk. 11 Moreover, the business impact (such as costs for laying off workers,
shutting down production plants etc.) can be calculated relatively easily from
internal data.
As shown in the previous chapter, sourcing for quality in low-cost countries has
actually proven to be a viable approach. With this in mind, it is no surprise that
quality problems pose a low threat for the outcome of low-cost country sourcing
initiatives. To summarize, it is evident that external risks are perceived to be more
important. Is this due to psychological or rational reasons? Clearly, external risks
are more difficult to assess which makes them more important if associated events
should occur. However, without proper assessment, rational decisions cannot be
made.
11
In line with general risk management literature, a distinction is here made between systematic
and specific risk, in a sense that the latter is only influenced by factors with the own organization
and can thus be hedged away. The former denotes market risk that the organization virtually
cannot influence.
139
1. Enable – Set up Have implemented LCCS-focused spend analysis 30% 44% +47%
necessary infrastructure.
Have dedicated staff specifically to LCCS 22% 35% +59%
2. Plan – Develop Have formulated strategies at corporate, purchasing organisational 29% 46% +59%
strategies through internal and supply category level to facilitate LCCS
and external environment
assessment Use integrated global capacity planning 25% 36% +44%
3. Execute – Test the Use international procurement office(s) 13% 19% +46%
waters and ramp up
Have contracted 3rd party global sourcing services 14% 17% +21%
4. Evaluate – Assess Benchmarks LCCS initiatives against other companies 19% 25% +32%
impact and improve
Use total cost management approaches in sourcing process 32% 42% +31%
The four process steps are labeled “Enable”, “Plan”, “Execute” and “Evaluate”. In
the first step, Enable, the necessary infrastructure is built. On the corporate supply
side, the most important issue here is the ability to identify and analyze enterprise-
wide purchase spend. Without it, the possibility to achieve economies of scale
through bundling is very limited. The study reveals that less than one third of the
participating companies today are able to analyze enterprise-wide purchase spend.
With an estimated growth of 47 percent over the next five years, 44 percent of the
companies will be able to analyze enterprise-wide spend by 2010.
is that companies might not have realized the importance of dedicated staffing,
but that by 2010 LCCS will clearly become a strategic focus.
4. Evaluate:
Benchmark
3. Execute: and assess
impact
2. Plan: Test the
waters and
1. Enable: Develop ramp up
strategies 26%Ö34%
Set up
necessary
infrastructure
14%Ö21%
23%Ö33% 25%Ö37%
Business process redesign and product redesign are two powerful actions that can
be taken to increase the benefits from LCCS. Business process redesign is
especially important, since the domestic sourcing process cannot just be replicated
in LCCs (Battacharya et al., 2004). For example, onsite supplier management and
development is often needed in order to build up relational capital assure quality
(Cousins, Handfield, Lawson, & Petersen, 2006b); thus, the sourcing process
needs to be redesigned in a way that it facilitates adequate face-to-face
communication. This is especially important in countries such as China, where
trust is built between individuals, and not organizations (Haley et al., 2004).
The study shows that 22 percent of the participating companies have redesigned
the sourcing process to support LCCS, which is a surprisingly low number. The
low-cost labor in LCCs provides opportunities to redesign products in order to
increase the degree of labor content and reduce the need for capital-intense
automation equipment (Battacharya et al., 2004). The result is decreased tied-up
capital as well as improved flexibility. However, only 16 percent of the
respondents have redesigned processes and products and this number is only
expected to increase to 18 percent over the next five years.
141
The second LCCS process phase is planning. The key issue here is to make an
internal assessment of the organization and match it to conditions in the external
environment in order to exploit opportunities and neutralize threats, in line with a
classical strategy formulation approach (e.g. Shrivastava, 1994). 29 percent of the
study participants say they have formulated strategies at corporate, purchasing,
organizational and supply category level. Although estimated to increase to 33
percent by 2010, this number is considered to be relatively low. 25 percent of the
respondents make use of global capacity planning to forecast future needs of
internal customers. This number is expected to reach a penetration rate of 36
percent within five years. 25 percent of the respondents assess the sourcing
environment by on-site supplier visits in LCCs and 21 percent have contacted
business development agencies there.
The third phase is about executing the LCCS strategy. By starting out at a
controlled pace by using for example balanced scorecard (Kaufmann, 1999), thus
“testing the waters”, initial trials can be carried out while keeping risk impact low.
During this time, valuable knowledge can be obtained while most cost saving
gains coming from reduced purchase price. However, this is not sufficient on a
long-term basis (Kaufmann, 1999). LCCS operations must be integrated in the
global business model in order to reap as many benefits as possible.
If the purchase volume is high enough (Carduck, 2000; Reinelt, 1999), a steady
purchasing channel in the shape of an international procurement office (IPO) can
be set up. This enables removal of costly intermediaries and facilitates supplier
development in the sourcing country. Only 14 percent of the respondents use IPOs
and this number will increase to a meagre 17 percent by 2010. As an example,
retail industry leader Wal-Mart is currently purchasing merchandise amounting to
10 percent of China’s total exports to the US through its IPOs and the company is
achieving enormous bundling effects in this way (Shenkar, 2005).
has only been done among 14 percent of the respondents and is expected to
increase to 17 percent by 2010.
The fourth and last process step is centered on evaluation. The purpose is to assess
the outcome of previous actions and set an agenda for improvements. One of the
most powerful means is to benchmark against other companies, not only against
competitors in one’s own industry, but also against companies in other industries
(Jahns, 2005). This helps the company to determine the market position and
identify strengths where performance is excelling as well as weaknesses that need
to be addressed. Benchmarking of LCCS activities is only done by 19 percent of
the companies surveyed; however this number is expected to grow to 25 percent
over the next five years. To capture landed cost-savings from LCCS, 32 percent of
the respondents are applying total cost management approaches. With as strong a
growth as benchmarking, this number is estimated to increase to 42 percent over
the same period.
After the end of the LCCS management process, an agenda for improvements
must be developed and implemented so that effectiveness and efficiency gains can
be realized. In this way, businesses can adopt a closed loop approach which spurs
continuous improvements and long-term success, similar to the concept of total
quality management (Lillrank & Kano, 1989)
Quite surprisingly, labor costs scored lower than expected, as they often pointed
out as the main lever for overall cost savings. Roughly one-third of the
respondents claim to have experienced positive impact on quality, capital
investment costs and delivery reliability. However, now the picture starts to
become ambiguous as about one-fifth of the respondents say they have
experienced negative impact on the quality and capital investment costs, and more
than one-quarter say they have experienced negative impact on delivery reliability.
This result is also in line with earlier findings. The fact that order lead time,
inventory costs and transport costs are negatively affected was also expected and a
natural consequence of longer geographical distances between buyers and
suppliers, as well as a result from more complex supply chains.
Looking at the chart, there is a natural explanation to the question why the impact
from LCCS is higher on some performance indicators and lower on others. The
144
performance indicators at the top are closely related to so-called factor advantages,
which is the nature of very fundamental input resources, such as labor, raw
material and infrastructural factors, which virtually any company can exploit just
by its presence on the LCC market. When traversing towards the bottom of the
chart, the factors behind the performance indicators become more and more
dependent on internal capabilities of the buying company.
Taking delivery reliability, order lead time, inventory costs and transportation
costs as examples, it is evident that these are all part of global supply chain
management, a highly complex issue that is dependent on many factors. Such
factors could be how the buying company is working with its suppliers, what
information they are sharing with suppliers and how they do this. It is also highly
dependent on the sourcing location in order to weigh transportation and inventory
costs against other cost elements. At the end of the day, these actions are all
dependent on rational and sound decisions made by internal staff, which
highlights the importance of educated, skilled and talented purchasing and supply
managers.
The initial discussion was based on globalization as the fundamental driver to the
LCCS developments, which in turn is affected by a wide range of factors that
somewhat simplified can be boiled down to four categories of drivers, namely
political factors, socio-demographic factors, technological factors, and macro-
economic factors. It was also concluded that they are not nurturing the
globalization process in isolation however, but interplay in a complex manner. For
145
The first sub section is chapter 4.1 which contains a general overview of trends in
purchasing and supply management from secondary sources, as a point of
reference for further analysis. The general discussion about trends described in the
literature are primarily focusing on issues such as requirements for improvements
in business performance, increased importance of suppliers as a result of
decreasing degree of value added, and a changed paradigm of supplier
relationship management.
The descriptive analysis in the chapter was structured according to the so called
Supply Management Navigator (Jahns, 2005), which is a “knowledge landscape”
that contains the necessary bits and pieces for successfully carrying out sourcing
activities. By applying the navigator in an LCCS setting, relevant aspects for the
research could be identified. The first part of the descriptive analysis concerned
strategic LCCS management and was discussed in chapter 4.2. Aspects
investigated in this section was the degree to which firms have articulated visions,
developed strategies, are measuring performance, set up organizations, and
dedicate human resources to support LCCS initiatives.
Next, in chapter 4.3 operational aspects were elaborated. Here, the lion’s share of
the analysis focused on important sourcing countries today and in five years, the
suitability of certain supply categories, differences in sourcing patterns between
goods/services and direct/indirect supplies, and supply risk management. It was
evident that there is currently a high momentum in the field at present, where the
amount of supplies sourced in LCCs will increase by more than sixty percent over
a five-year period. Another interesting result was that the home country of the
buying firm seems to remain in high importance also over the next five-year
period, but neighboring countries will lose market shares in favor of suppliers in
low-cost countries.
Finally, in chapter 4.4, the extent to which certain LCCS-related activities are
carried out today and expected extent in five years were investigated. These
activities were identified through expert interviews with eight practitioners and
146
The results from chapter 4 indicate that it is possible to realize cost savings while
preserving quality. However, this raises a number of questions. For example, what
makes some firms more successful than others in terms of low-cost country
sourcing? Are they just fortunate to do business in an attractive industry (Porter,
1980), or is it just plain luck that leads to success? Much research has focused on
factors in the external environments as explanatory variables for the success of
sourcing initiatives. The question is, with the hype regarding important sourcing
regions such as China and India, does this mean that any firm starting sourcing
there will automatically be successful? The answer is likely no, but the question is
then where key to competitive advantage lies?
context is given by Ulrich and Barney (1984, p. 472), by stating that organizations
are “altering their structure and patterns of behavior to acquire and maintain
needed external resources”. With this as a starting point, the rest of this chapter is
devoted to the application of RBV in an LCCS setting and empirically test its
validity.
In times with an increasing share of value created outside the firm (Jahns, 2005),
with highly permeable firm boundaries, some scholars have questioned the
distinction of the firm with its environment (Perrow, 1967). This highlights the
strong decency between the firm and its environment suggested by numerous
management gurus and researchers (A. Chandler, 1962; P. Lawrence & Lorsch,
149
1967). Many attempts have been made to develop an integral theory that capture
the different idiosyncrasies at different analysis levels, alas, with limited success
(Astley & Van de Ven, 1983; Barney & Ouchi, 1986; Hrebiniak, 1985).
In this instance, top-down means that the theory is holistic and integral; thus the
focus is on the big picture rather than the constituent parts. Bottom-up, on the
other hand, focuses on the small details; the components and interrelations that
forms an “aggregation of specifics” (Astley & Van de Ven, 1983).
The horizontal axis, on the other hand, distinguishes between strategic choice
(voluntaristic) or environmental determinism (deterministic). This dimension is
rather philosophical and concerns the question whether free will exists in reality
(see the discussion about human nature in chapter 3.3). In order to depict a
continuum along the axes rather than dichotomous situations, the scheme is not
divided into a two-times-two matrix.
150
Macro
level Organizational ecology
Human ecology theory
Bain/Mason IO framework
Level of
analysis Schumpetarian economics
Resource-dependency
Institutional theory theory
Chamberlinian
Micro
Structural contingency theory economics/RBV
level
Deterministic Voluntaristic
Role of choice
which was later popularized by Porter (1980). In line with the modern version of
the resource-based view, Chamberlin (1933) found evidence that key capabilities
could stem from technical know-how, reputation, brand awareness, team work,
patents and trademarks. Due to the heterogeneity of products, a market that would
normally be considered as perfectly competitive can contain ”pockets” of
monopolistic-like firms, which have some influence on the market. An example is
for example Coca-Cola, which has obtained a superior market position through
branding and advertising, despite many attempts from rivals to imitate the product.
The ideas from Chamberlin and Robinson were later developed by Penrose (1959)
in the 1950s, by arguing that the resources themselves are not inputs to the
production process, but instead the services they can generate. Simply put,
services are dependent on the underlying resources, which can give rise to
different services depending on the configuration and combination with other
resources in other contexts. Her views have surprisingly similar arguments as
have been proposed in the dynamic capabilities view developed by scholars such
as Teece et al. (1998; 1997) and Eisenhardt and Martin (2000). It is this
heterogeneous distribution of resources that has the potential to make firms differ
from each other (Penrose, 1959). In sum, Penrose (1959) claims the firm to be “a
pool or resources, the use of which is organized in an administrative framework”.
The interesting thing about the historical developments of the RBV is the
relatively little importance it received, especially during the 1980s. In the early
years of strategic management, the schools of thought primarily emerging at
Harvard Business School, gave equal weighting to the external vis-à-vis the
internal environment.
The attention of what later was to become the RBV became even less during the
1980s when Porter (1980) launched his five-forces industry attractiveness model.
The model, which had its roots in the Bain/Mason industrial organization, led to a
paradigm shift where the perceived importance of firm resources declined in favor
to the industry quadrant of the LCAG framework (Hoskisson, Hitt, Wan, & Yiu,
1999).
However, during the late 1980s, an increasing body of empirical evidence was
starting to speak against the ideas of Porter. Not because there was a belief that his
model was wrong, but rather that it did not tell the whole story. For example,
studies by Cubbin (1988), Hansen and Wernerfeldt (1989) and Rumelt (1991)
showed that performance between firms was different within the same industries.
Furthermore, Cool and Schendel (1988) and Lewis and Thomas (1990) even
observed performance differences within the even narrower strategic groups. Even
more interesting, in a 2,800-company sample, Rumelt (1982) observed that
business-unit factors explained 44 percent of the variance in performance,
compared to a mere 4 percent from industry factors. With such an extensive body
of empirical evidence, the strategic management pendulum started to swing back
to the neutral position.
During the same period, in parallel with the empirical body of research that was
conducted, a new stream of theoretical research emerged. Beginning with scholars
such as Wernerfeldt (1984) and Barney (1986a), a growing number of researchers
started to believe that a great deal of market success can be attributed to internal
factors. Wernerfeldt (1984) was the first scholar to use the term “resource-based
view”, and indeed indicated the high potential of the theory by describing the
theoretical findings as “the first cut at a huge can of worms”. In the following
years, a truly huge body of research on the RBV emerged, ranging from insights
in both economics and management. A summary of the most influential
contributions are shown in Table 12.
Mahoney and Pandian (1992, p. 373) argue that RBV provides an “illuminating
generalizable theory of the growth of the firm” and that competitive advantage is a
function of industry structure, organizational governance and firm effects (i.e.
firm resources and strategies). Furthermore, they claim that RBV has the potential
to coalesce with the other cornerstone of competitive advantage (i.e. market
154
Conner (1991) has made a comparison of RBV with five major schools of thought,
namely neoclassical economics, industrial organization, the Chicago school,
Schumpetarian economics and transaction cost economics. The conclusions from
this comparison are that RBV shares at least one commonality and one difference
with these five schools of thought. However, this does not by any mean say it
serves as a unifying theory, but rather that it supports other theoretical viewpoints
while contributing with its own unique perspective.
Although RBV has been widely criticized for being tautological (Priem & Butler,
2001), this problem is effectively solved in the empirical study of this thesis by
not defining resource characteristics and competitive advantage in the same terms.
Instead, the questionnaire investigated the possession of key resources separately
from the performance assessment, thus also avoiding retrospective bias (March &
Sutton, 1997).
155
With this in mind, it is assumed that resource heterogeneity and the possession of
key resources are a result of strategic choice (i.e. managerial action). The
156
implication is that the key tasks for managers are to identify, obtain/create, protect
and deploy key resources in order to arrive at the optimum resource bundle
(Penrose, 1959). In sum, the RBV can be represented as a causal chain with four
constructs, namely (1) the resource pool which is the whole domain of resources,
both free resources and resources in possession of firms, (2) key resources in
possession by firms, (3) sustainable competitive advantage generated by firms
possessing key resources, and (4) superior performance in the shape of Ricardian
rents (Figure 56). The next subsection will elaborate on the constructs in more
detail.
Key Resources
SCA
Resource Pool Valuable
Customer service Superior
Resource Rare Performance
Heterogenity Cost leadership
Imperfectly Ricardian rents
Imperfect imitable Differentiation
resource mobility Monopoly rents
Imperfectly Innovativeness
substitutable
The resource-based view seeks to explain why firms differ. The theory departs
from the assumption of perfect competition and then seeking to identify the
dimensions that do not apply in the particular case (cf. Chamberlin, 1933). When
perfect competition holds, the market is in equilibrium and the market players are
too many and too small in comparison with total market size in order to have any
influence on it. Moreover, the equilibrium state implies that producers are ”price
takers” and hence cannot earn any profits. According to the resource-based view,
successful firms that are in possession of key resources will earn so called “rents”
(Ricardo, 1817). The value of a resource is thus the sum of the discounted future
cash flows it can render over its life time (Lippman & Rumelt, 1982).
The concept of rents can be traced back to the era of classical economics in the
19th century. The first economist to use the term was Ricardo (1817), whose ideas
157
was later further developed by British economist Alfred Marshall (1920). ”Rents”
are defined as earnings that do not induce additional competition. This is
according to the neoclassical assumption that the possibility to earn profits (i.e.
profitable industry segment) will attract new firms until no more profits can be
made and a state of equilibrium is achieved. According to Penrose (1959), rents
are defined as the prices of the flow of services emanating from the underlying
resources, referring to the idea that resources is the origin of the services a firm
can produce. Broadly speaking, literature distinguishes between five types of rent:
Ricardian rents
Monopoly rents
Entrepreneurial rents
Quasi-rents
supply. Those resources which are only temporarily scarce are known as quasi-
rents.12 In sum, the Ricardian-Marshallian concept of rent therefore hinges on the
fixed/temporary scarcity of the resource, as well as its value compared to its next
best alternative. According to the latter criterion, rents are considered to
be ”supra-normal profits”, or ”above-average profits”.
12
The term quasi-rents is also sometimes denoted as Marshallian rents after its “founder”. See
(Peteraf, 1993) for a in-depth discussion on the topic.
159
The extensive attention given to the word ”competitive advantage” can be traced
back to the fields of economics, and most notably, military studies (Whittington,
1993). The idea of seeking a strong and attractive position dates all the way back
to the ancient Greek wars. This explains the original meaning of the Greek word
for strategy, namely strategos, meaning “generalship” (Cummings, 1993). von
Neumann and Morgenstern (1944), on the other hand, see strategy as a pattern of
action with the aim of move and counter-move.
Although the term ”competitive advantage” is one of the most renowned ones in
strategic management literature, most research has revolved around how to
achieve it rather than what it is (for a good example, see Coyne, 1986). Most
likely, the reason to this is the fact that it is a relative term that is very difficult to
define. Firstly, it is a relational term that only makes sense when the focal firm is
compared with one or more competitors along a number of dimensions that are
relevant in particular situation of competition (Hao, 2000; Hu, 1995; Kay, 1993,
1994).
(Coff, 1999), nor that superior performance is a result from competitive advantage
(Powell, 2001). For example, competitive advantage along one dimension might
be offset by another competitive disadvantage and will thus not be reflected in
market performance (Ray, Barney, & Muhanna, 2004). Furthermore, it is possible
that economic rents are being generated after a competitive advantage has ceased
to exist (Hao, 2000).
A common approach to get around the above mentioned problems has been to
view the competitive advantage intermediary either as a black-box, where superior
performance is assumed to be an implication of competitive advantage (i.e. the
assumption that superior performance always and only follows competitive
advantage), or by considering competitive advantage and superior performance to
be equivalent. The latter case clearly imposes a risk for tautology (Priem & Butler,
2001). Others have attempted to measure other variables, such as differentiation
or cost leadership, which have been suggested to be ”effects” caused by
competitive advantage (Bharadwaj, Varadarajan, & Fahy, 1993; Day, 1994; Day
& Wensley, 1988; S. D. Hunt & Morgan, 1996; Porter, 1980, 1985).
The fundamental problem here is that such market-oriented factors only applies
when analyzing rents generated by industry-related factors. In other words, this
means that one would try to assess the impact from internal factors while
measuring the actual effect stemming from industry-related factors. An analogy
here is to measure the temperature in the refrigerator when trying to determine
whether it is a good day for ice cream. The various industry-related factors
proposed in literature as a measure for resource-based competitive advantage
might prove highly reliable, but validity is certainly the opposite. As a
consequence, Ray et al. (2004) suggest to choose business process effectiveness
as the independent variable when testing the resource-based view empirically.
As such, firms can sustain a competitive advantage only if the key resources do
not diffuse among firms. For this to hold true, heterogeneously distributed key
resources must be imperfectly mobile. Put simply, imperfect mobility implies that
a certain resource cannot be traded in a strategic factor market (R. Caves, 1980;
Dierickx & Cool, 1989). Resources of which the property rights are vaguely
defined or lacking “bookkeeping feasibility” belongs to this category (Bator, 1958;
Dierickx & Cool, 1989; Meade, 1952). Moreover, some resources are inherently
imperfectly mobile due to its idiosyncratic value to a particular firm; outside the
firm, it would be less valuable. In other words, resources are imperfectly mobile
when they are specialized to a particular firm (Williamson, 1985). Heterogeneity
is a necessary condition for imperfect mobility, but the opposite does not
necessarily need to hold. However, the latter condition is relatively unlikely in
reality, meaning that imperfect mobility implies heterogeneity as well (Peteraf,
1993).
13
Initially, the framework proposed by Barney (1991) considered imperfect imitability and
substitutability separately, the so called VRIN framework (valuable, rare, inimitable and
163
Therefore, possession of valuable (but not rare) resources can be said to avoid an
opportunity disadvantage. 14 This is supported by McKelvey (1982) and Porter
(1980), who claim that firms do increase their probability of economic survival
under conditions of competitive parity. According to Hirschleifer (1980), a
nonsubstitutable), but were later merged and complemented with the organizational focus
dimension (cf. Barney, 1996, 1997).
14
Analogous with the term opportunity cost, we introduce the term competitive opportunity
disadvantage, to denote the competitive disadvantage of not being a possession of a valuable, but
neither rare, nor imperfectly imitable nor imperfectly substitutable resource. An example of
resources with an opportunity disadvantage is an ERP. As such, it can yield tremendous efficiency
improvements across the enterprise. Not having an ERP implemented thus implies a competitive
disadvantage relative to those firms who have one. But it cannot generate sustainable competitive
advantage per se, because a vendor like SAP or Oracle can (and do) implement the same ERP to a
wide range of its clients.
164
Although valuable and rare resources can generate competitive advantage, it is not
for sure that the competitive advantage generated is sustainable. So called ex post
limits to competition provides the necessary requirement in order to make the
competitive advantage sustainable (Peteraf, 1993). Without sustainability,
competitive advantage, and ultimately rents, would erode over a short period of
time. Hence, resources must remain scarce in supply. In the resource-based view,
the first ex post limit to competition in order to render sustainability is imperfect
imitability.
Put simply, imperfect imitability means that competing firms that do not possess a
certain rare and valuable resource, are not able to obtain them (Barney, 1986a,
1986b; Lippman & Rumelt, 1982). The reasons for imperfect imitability can be
one or several of the three following factors (Dierickx & Cool, 1989): (1)
dependency on unique historical conditions, (2) causal ambiguity, meaning that
the link between a resource and sustainable competitive advantage is not clearly
understood, and (3) social complexity which obscures the situation. Clearly,
imperfect imitability is especially important for companies seeking to compete
through differentiation and innovativeness, since first-mover advantages
otherwise would erode.
When the needed resources have been acquired, they need to be developed,
protected, and deployed. By developing a resource, it is meant that it is combined
with other resources in order create a configuration which has a value that is
greater than its constituent parts. Resources also have to be protected, either by
mechanisms hindering them from diffusing to competitors directly or indirectly
through imitation or substitution. As discussed in chapter 5.1.3, the inherent
nature of supplies can help a great deal to protecting the resources from being
166
The number of resources in every firm is virtually endless. Many times, different
terms are used interchangeably which can cause a great deal of confusion. For
example, the term “competency” is often used interchangeably with terms such as
“skill” or “capability”. Moreover, they are often used in combination with
attributes such as “core” or “distinctive” in order to highlight their strategic
importace (Fahy, 2001). As a first simple typology, Black and Boal (1994)
differentiates between contained resources and system resources, which is pretty
much in line with the discussion above. Contained resources are resources that are
in themselves “atomic” in a sense that they are fundamental resources themselves
and cannot be broken down further. System resources are, as the name implies,
systems of resources that are built up contained resources. A common
classification scheme is to distinguish resources depending on their physical
characteristics, such as financial resources, physical resources, human resources,
organizational resources and technological resources (Hofer & Schendel, 1978).
Resource Type
As discussed in the previous section, Ray et al. (2004) suggest to choose business
process effectiveness as the dependent variable when testing the resource-based
view empirically. Positive impact on purchasing process effectiveness from
supply-side activities has been observed earlier, for example through supply chain
integration (Berry, Towill, & Wadsley, 1994), early supplier involvement (J. R.
15
Michael Porter coined the concept of strategic groups in his studies about industry structures in
the late 1980s. Following his work, there have been numerous studies considering strategic groups
either partly or completely. The existence of strategic group has been widely debated in literature.
Strategic groups defined as groups of firms that have same strategic dimensions (Hunt, 1972;
Porter, 1979). According to this definition, firms in the group follow similar strategies in terms of
the key decisive criteria. Strategic groups can consist of a single firm or all the firms in the
industry.
169
Carter & Ellram, 1994) and strategic supplier development (Narasimhan &
Jayaram, 1998).
Since this approach has proved to be successful in the past, purchasing process
performance was selected as a measure of the impact of firm resources in
possession or acquired through low-cost country sourcing in this particular study.
Furthermore, the performance measures were also dichotomized into two
subcategories, namely effectiveness and efficiency. Here, effectiveness refers to
the extent to which a pre-defined objective has been achieved, whereas efficiency
refers to the extent of inputs needed to achieve those pre-defined objectives
(Monczka, Carter, & Hoagland, 1979).
Tangible resources are basically entities with physical properties that one can see
and touch. They are also characterized by fixed long-term capacity (Wernerfelt,
1989). Moreover, tangible resources have the property of being owned by a legal
institution and the value can often be measured by means of accounting and thus
affects the balance sheet (Hall, 1992). Another important property is that tangible
assets are tradable in strategic factor markets (Barney, 1986a; R. Caves, 1980;
Dierickx & Cool, 1989), meaning that the resource ownership may very well
change over time.
opportunity disadvantage (cf. section 5.1.3 about the resource pool and key
resources). Thus, we hypothesize:
Intangible resources, on the other hand, have a more vague nature, and can be
entities such as product and process patents, brand names, reputation etc (Hall,
1992; Williams, 1992). Contrary to tangible resources, intangible resources cannot
be valued efficiently in terms of supply and demand. However, the value of such
resources can be assessed by observing the discrepancy between the balance sheet
valuation and market capitalization of publicly listed firms (Grant, 1991; Hall,
1989, 1992; Rumelt, 1987). Such discrepancies have been empirically observed in
industries such as the pharmaceutical industry, consumer goods industry and
service industry (Fahy, 2001; Grant, 1991).
Also trust and reputation generated from formal and informal socialization have
proven important for successful supplier relationship management (Burt, 2000;
Cousins, Handfield, Lawson, & Petersen, 2006a; Johnston, McCutcheon, Stuart,
171
& Kerwood, 2003), especially in LCCs through supplier on-site visits (Handfield
& Nichols Jr., 2004). With this in mind, the second set of hypotheses is:
Capabilities have also been frequently been pointed out as the strategic factor that
makes the firm dynamic in a sense that it lets the firm “renew competences”
(Teece et al., 1997), and is in line with the knowledge-based view in a sense that
existing knowledge can be reconfigured into new knowledge sets that can render
new competencies (Grant, 1996). This property is becoming increasingly
important in today’s volatile business environment where competition is
increasing, product life cycles become shorter and shorter and the requirements
for time-to-market and timing are crucial (Teece et al., 1997). Moreover, several
scholars claim capabilities to be the ultimate source of competitive advantage, as
they make firms truly distinguishable and unique (Learned et al., 1969; Prahalad
& Hamel, 1990).
172
Based on the four criteria postulated by Barney (1991), we can assess the relative
importance of internal resources. As discussed earlier, tangible resources are often
purchased from strategic factor markets; such a market is defined as “a market
where the resources necessary to implement a strategy are acquired” (Barney,
1986a). As a result, such resources can virtually be acquired by any firm that has
sufficient financial resources to perform the transaction, thus questioning its
strategic importance (Ramsay, 2001a).
Intangible resources, on the other hand, are more difficult to trade in strategic
factor markets due to their idiosyncratic nature (Williamson, 1979). Moreover,
they are predominantly developed within the firm, as a result of “choosing
appropriate time paths of flows over a period of time” (Dierickx & Cool, 1989).
Although some intangible assets such as copyrights and patents can be bypassed
by making appropriate modifications to product designs, short-term rent
173
In short, these arguments show that intangible resources are likely to be more
difficult to imitate and substitute as the causal ambiguity is higher. As was
discussed earlier, intangible resources are also characterized by an ability to be
scaled up and reused with little or no efforts, implying that the value can be
“inflated” through asset mass efficiencies and potentially creating synergies with
other resources, so called asset stocks interconnectedness (Dierickx & Cool,
1989). Consequently, the potential value of intangible resources is likely to be
much higher compared to tangible resources. This leads to the following
hypotheses:
As discussed in the previous section, many scholars have claimed that capabilities
are the most important source of competitive advantage. The rationale is very well
grounded in RBV. Firstly, capabilities (i.e. know-how) are created through
“learning-by-doing”, meaning that knowledge about how to carry out a task or
solve a problem is accumulated over time (Argote & Epple, 1990; Arrow, 1962;
Senge, 1994). Furthermore, the current and future capability development is also
claimed to be contingent on past experience (B. Levitt & March, 1988). In other
174
The least common denominator of the various research streams is the fact that
higher-order resources are considered to be the source of dynamism and
adaptability within firms (Helfat & Peteraf, 2003; Teece & Pisano, 1998; Vargo &
Lusch, 2004; Winter, 2003). For example, Vargo and Lusch (2004) claim that is
not goods themselves that comprise customer value, but rather the knowledge that
were used to create them. As such, a distinction between operand and operant
resources is done, implying that operant resources are “employed to act on
175
operand resources or other operant resources” (Vargo & Lusch, 2004, p. 2). In
other words, operant resources are here represented by capabilities.
It can be concluded that capabilities are very valuable since they essentially
represent the value-creating mechanism and are rare due to the fact that they
cannot be purchased in factor markets. Morover, several scholars claim
capabilities to be the ultimate source of competitive advantage, as they make
firms truly distinguishable and unique (Learned et al., 1969; Prahalad & Hamel,
1994). Finally, they are difficult to imitate and substitute due to the tacitness and
causal ambiguity. Based on the discussion above, the next set of hypotheses is:
seriously questioned (Ramsay, 2001a). On the other hand, as all supply markets
suffer from market failure to a certain extent, the requirements postulated by RBV
can still hold in practice (Ramsay, 2001b). A growing body of research seems to
be in favor of the increasing strategic importance of purchasing due to factors
such as the decreasing degree of value-added and possibilities to set up alliances
and strategic supplier relationships (P. H. Andersen & Christensen, 2000; Mol,
2003).
gain competitive advantage, as firms expand their sourcing markets into new
geographic regions. Such resources have two characteristics: they are (i) external
to the firm and (ii) imperfectly mobile. (i) implies that the resources are not in full
control of the firm, but can be at least partially controlled through management
decisions. An example is a joint venture with a manufacturing firm or an
exclusive strategic relationship with a LCC supplier. (ii) implies that local
presence is necessary to some extent in order to take advantage of the resource,
since it cannot be moved to the home country of the buying firm.
Advanced factor resources, on the other hand, can only be obtained through
resource stock accumulation, analogous to the earlier discussion about intangible
resources and capabilities (Dierickx & Cool, 1989). An example of such an
advanced resource is the transfer of technologies and competencies from suppliers;
178
it has been shown that preferred suppliers not only deliver goods at the right time,
at adequate quality and at lowest price possible, but increasingly delivering
“competencies” and “knowledge” (Goffin et al., 2006; Johnsen et al., 2006).
Value does not solely stem from the supplies being purchased, but also from the
buyer-supplier relationship itself (Lindgreen & Wynstra, 2005). Moreover,
managing these relationships requires capabilities that are difficult to imitate and
also require supportive organizations (Barney, 1997). Based on the discussion
above, the next set of hypotheses is:
resources needed to execute its strategy in order to maximize the resource stock
and ultimately generation of rent.
Naturally, since basic factor resources can only be protected by the buying firm to
a very limited extent, they can basically be appropriated by any firm that has the
internal resources needed such as manpower and capital needed to set up sourcing
activities in a foreign supply market. Consequently, the value and rareness of such
resources are limited. Although the substitutability of basic factor resources
potentially can be highly imperfect, the imitability is certainly not, due to the lack
of protective barriers.
Advanced factor resources, on the other hand, differs from basic factor resources
in a number of ways. Firstly, they have a more tacit and intangible nature which in
turn yields a higher degree of causal ambiguity and the degree of imperfect
imitability and substitutability is thus higher. As an example, special relationships
to key market actors such as strategic suppliers or government officials may not
be known to competitors, nor is it for sure that competitors are able to obtain such
information.
This in turn leads to a situation where the buying firm has the potential to
influence decision-making of its partners (Gaski, 1986; Sheth & Parvatiyar, 1992).
Thus, the buying firm has the possibility to influence the external environment by
exerting power on other market actors (Griffith & Harvey, 2001), and especially
create imperfections in factor markets (Barney, 1986a). Based on this discussion,
180
The relationship between effectiveness and efficiency has not been extensively
investigated in literature. Therefore, it is hypothesized that there exist a non-
recursive relationship, positive relationship between purchasing process
effectiveness and efficiency. That is, the ability to accomplish pre-defined goals
can lead to less resource consumption and vice versa. Therefore, without making
any claims on causal direction, the last hypothesis is:
Tangible
Resources
H4a/b H1a
H6a
Purchasing
Basic H6b
Process
Factor Efficiency
Resources H7a H7b
H8a/b
Advanced
Factor
Resources
All items were measured on a five-grade semantic differential scale (Ajzen &
Fishbein, 1980) ranging from one to five. The construct items were not rotated
182
from interview to interview, but were randomly ordered. The questionnaire was
pre-tested during a purchasing and supply management seminar with eight
sourcing experts comprising a mix of researchers and industry practitioners. Items
that were not immediately understood or found ambiguous were changed in their
wording or dropped altogether unless it would detrimentally affect construct
validity.
Out of the original set of items, the constructs (i.e. resources and purchasing
process performance) could be operationalized through three items each. The
fifteen items representing the resources were then grouped into five factors
through principal component analysis. In this way, potential problems with
collinearity were eliminated (Backhaus, Erichson, Plinke, & Weiber, 2003). The
same approach was applied for the three items representing purchasing process
effectiveness and efficiency, respectively, in order to generate the corresponding
components.
In order to consolidate purchase spend to the greatest extent possible, firms must
coordinate and integrate sourcing activities both across business units as well as
functional units (Robert J. Trent & Robert M. Monczka, 2002). Here, various
supportive systems play a key role as enablers for increasing transparency to
metrics such as total cost savings, quality improvements etc. Therefore, it was no
surprise that the first measure representing tangible resources identified was
implementation and use of performance measurement systems. Such systems can
be any sort of IT system such as ERPs, e-sourcing solutions, spend management
solutions etc., where internal corporate data can be gathered, processed and
presented as key performance indicators.
when sourcing in emerging and transition economies, quality is one of the most
important metrics, and the high importance has been repeatedly acknowledged in
earlier research (Cross & Gordon, 1995; Turnbull, Delbridge, Oliver, &
Wilkinson, 1993).
Through the expert interviews, it also became clear that sufficient resource
allocation to affiliations in emerging and transition economies were very
important for successful operations. Therefore, the second item, free capital, as
the name implies, is basically readily available cash for investments and
expenditures, and directly affects solvency. Thus, it was selected as the second
item representing tangible resources.
The interviews also revealed that LCCS operation plants and facilities (i.e.
physical entities which enable and facilitate local operations) were regarded as
highly important. The overall consensus was that sourcing activities in emerging
and transition economies cannot be performed “remotely”, implying that facilities
such as international procurement offices (IPOs) need to be located in the
sourcing market, in proximity to suppliers, to support on-site supplier visits and
similar activities.
This is also in line with Caves and Porter’s (1977) concept of mobility barriers,
which themselves are a further development from Bain’s (1956) idea of entry
barriers. Patents are not limited to only include the physical characteristics of an
invention (what it is), but can also include processes (i.e. the way in which a good
is produced). Thus, the first item measuring the extent of intangible resources
available to the firm is the amount of product/process patents held by the firm. Of
course, in order to have effect, intellectual property regulations in the host country
are essential.
The expert interviews also revealed that reputation of the firm was important, as it
was claimed that this property can increase the trust in buyer-supplier
relationships in emerging and transition economies, thus facilitate bidding and
negotiation processes (Hall, 1992). Therefore, the second item is the reputation of
the firm in the LCC where sourcing is taking place.
Also all kinds of codified knowledge about sourcing in emerging and transition
economies, here referred to as LCCS know-how, was considered highly important.
The conclusion from the interviews is that it is actually not the knowledge itself
that is of importance, but they act as precursors to higher-order capabilities
(Eisenhardt & Martin, 2000; Teece & Pisano, 1998; Teece et al., 1997; Winter,
2003). Thus this measure was selected as a third item of the construct.
185
Extent of Capabilities
The expert interviews made it clear that the ability to integrate strategic suppliers
in business processes was one of the most important capabilities when sourcing in
emerging and transition economies; thus it was selected as the first item of the
capabilities construct. The importance is confirmed in literature where leveraging
knowledge with local suppliers can help the buying firm decreasing the so called
“market knowledge gap”, which is defined as “the difference between
international partners related to the local market” (Griffith & Harvey, 2001).
Empirical studies have indicated its positive impact on performance (Carr &
Pearson, 1999).
The second item that was distilled from the expert interviews is the ability to
utilize cross-functional teams. This implies that staff from various functional units
such as production, purchasing, marketing, finance and R&D are brought together,
often together with key suppliers in order to develop an optimal product.
Empirical studies have shown that this is the most effective way to reduce product
cost; the later in the product development cycle one tries to influence costs, the
more difficult it gets (Jahns, 2005). These two items pointed out as the two most
critical factors for successful sourcing in emerging and transition economies are
also consistent with supply management literature (Dobler & Burt, 1996).
This claim is further supported by Wu and Choi (2004), who claim that good
relationships with governments can help firms to obtain valuable information
about issues such as future economic policies, taxation and import/export duties,
thus reducing uncertainty and increasing the financial planning horizon. For
186
As most studies point out that cost reduction is the key driver behind low-cost
country sourcing, it is no surprise that the access to low-cost labor was considered
the most important basic factor resource by the interviewees. This resource
provides the highest leverage for total cost savings in many instances, since labor
cost is often one of the biggest cost elements of purchased supplies (Battacharya
et al., 2004). Moreover, it was claimed that the choice of low-cost manual labor in
favor of expensive automation equipment can reduce the amount of tied-up capital
and thus strengthen the balance sheet.
The second important basic factor resource identified through the interviews was
access to the LCC supplier base with a certain critical mass (i.e. sufficient size), as
it was claimed that this facilitates competition and supplier selection, both for
supplies that require strategic collaboration and for supplies that require arms-
length relationships. This is also an effect from the supplier base consolidation
taking place in many firms; as the number of tier-one suppliers are reduced and
increased focus on single sourcing in order to increase bargaining power and
economies of scale, the risk of supply chain disruptions also increases (Atkinson,
1989). Therefore, the second item selected is access to LCC supplier base.
Another interesting outcome from the expert interviews was the fact that the
access to LCC customer markets also was considered very important. The
rationale was the fact that vast synergies can be achieved if the supplier market
and customer market are closely located to each other. The synergies will be even
higher in case the markets are overlapping, thus effectively shifting the entire
187
The first item of the advanced factor resource construct obtained through the
expert interviews was access to educated/skilled local workforce, which is
frequently stated as a crucial aspect of LCCS. Many interviewees claimed that
excellent workforce can help reducing the cultural distance (i.e. the difference
between market actors along a set of cultural dimensions) between the buying
firm and its environment, which is also supported by literature (Gomez-Mejia &
Palich, 1997; Kogut & Singh, 1988).
The second item obtained through the interviews was access to new technologies.
In other words, through strategic sourcing in emerging and transition economies,
not only goods can be supplied to the buying firm, but also technology and
knowledge embedded in the supplies sourced, similar to the marketing paradigm
suggested by Vargo and Lusch (Vargo & Lusch, 2004). This resource can provide
competitive advantage, but only under the prerequisite that it cannot be easily
obtained by competitors. This item has proved to be an important driving force to
buying firms, as it helps them to improve innovativeness and quality and
ultimately improve competitive advantage (Alguire et al., 1994).
The last item provided by the expert interviewees was first-mover advantages in
LCCs, which was claimed to give opportunities for locking up scarce resources
and thus preempting competitors from acquiring them in a “first come, first serve”
manner (Lieberman & Montgomery, 1988, p. 1112). Examples of first-mover
advantages mentioned in this context was hard-to-imitate exclusive buyer-supplier
188
relationships and joint ventures, but also more tangible advantages such as tax
rates and climate, which potentially may erode more rapidly; therefore, the timing
of entry is a crucial factor (Kevin, Lam, Karakowsky, & Gongming, 2003).
Although the life span of first-mover advantages have been widely debated, it is
clear that it is contingent on the type of first-mover advantage, which eventually
translates into (at least) short-term rents (Boulding & Christen, 2003).
There are two forms of factor analysis, namely exploratory factor analysis and
confirmatory factor analysis. According to Hair et al. (1992), there is continued
debate concerning the appropriate role of factor analysis. Many researchers
consider it only exploratory, useful in searching for structure among a set of
variables, or as a data reduction method.
Exploratory factor analysis (EFA) attempts to discover the nature of the constructs
influencing a set of responses. The aim of exploratory factor analysis is to
examine the underlying structure of a relatively large set of variables. The key
feature of EFA is that the researcher has no à priori assumption of the underlying
factors (i.e. latent variables) and their corresponding indicator variables. This is
the most common form of factor analysis. Hence, no prior theory is applied, and
one uses factor loadings to determine the factor structure of the data.
The aim of this section is to examine and validate the underlying factor structure
of the empirical set of variables for the model. This is achieved in two steps. First,
the theorized factor structure was put to test through a confirmatory factor
analysis with the statistical software package LISREL. In this way, a rigorous
assessment of convergent and discriminant validity could be performed
(Campbell & Fiske, 1959). Finally, a semi-exploratory factor analysis was
performed through principal component analysis (PCA) in order to generate
normally distributed and non-correlated factors for the principal component
regression analysis.
The reason why the first stage of the factor analysis is denoted as “semi-
exploratory” is that we do not introduce any constraints to the analysis, but let the
indicator variables arrange themselves freely according to the PCA algorithm. At
a first glance, one would suggest that this makes the analysis qualified as a
exploratory factor analysis. However, since an à priori establishment of the
underlying factor structure has been done, thus the criterion for being an
exploratory factor analysis is not fulfilled.
The fact that the indicator variables are not constrained to any hypothesized
factors implies that it cannot be categorized as a confirmatory factor analysis
either. As Jöreskog (1974 p. 2) puts it, “Many factor analyses are to some extent
both exploratory and confirmatory, since they involve some variables of known
and other variables of unknown composition”. In other words, exploratory and
confirmatory factor analysis should not be seen as mutually exclusive, but rather
as a continuum of ordered progression (J. C. Anderson & Gerbing, 1988).
In order to assess convergent and discriminant validity for the various constructs
used in this study, confirmatory factor analysis was applied. Confirmatory factor
192
the data frequently exhibits a hierarchical nature (Micceri, 1989; Schafer &
Graham, 2002; Yuan & Bentler, 2002).
There are several approaches that can cope with this problem. One is to
standardize the data and thus rely on correlation matrices instead of covariance
matrices (Gerbing & Anderson, 1984). This is generally not advisable as the
variable scaling is lost. Another approach is to use the asymptotically distribution
free (ADF) approach proposed by Poon and Lee (1994). However, this procedure
can only be applied for very large samples (Fouladi, 2000; Muthén & Kaplan,
1992). Consequently, this approach is not feasible for this study as the current
sample size is N = 200.
194
Table 14. Univariate statistics before and after normal scores procedure
Capabilities (ȟ3)
x7 SUPPINT 3.105 0.993 -0.063 -0.281 -0.062 -0.257
x8 CROSSFNC 3.278 1.052 -0.073 -0.330 -0.095 -0.306
x9 GOVMGT 2.621 1.188 0.207 -0.739 0.173 -0.704
The problems associated with nonnormal data can be avoided by transforming the
data into a more normal distribution shape (Babakus, Ferguson, & Jöreskog, 1987;
Curran, West, & Finch, 1996; West, Finch, & Curran, 1995). The so called
normal scores technique was applied to achieve the transformation while
preserving means and standard deviations (Jöreskog, 1999; Jöreskog, Sörbom, du
Toit, & du Toit, 2000).
After the normal scores transformation, the data fulfills a rule of thumb saying
that absolute univariate skewness and kurtosis should not exceed 1.0 (Hair et al.,
1992). In comparison with Monte Carlo simulations done in the past, the
deviations from normality are considered “slight” (Boomsma & Hoogland, 2001;
Curran et al., 1996). To be on the safe side, in order to decrease the possible
impact from the small deviations, so called robust maximum likelihood (RML)
estimation was applied in which Ȥ2 values are scaled to account for the degree of
multivariate non-normality (Satorra & Bentler, 1988, 1994). Since RML takes
third- and fourth-order product moments into account, it has proven to be very
stable against multivariate non-normality.
Convergent Validity
The empirical data was analyzed through structural equation modeling (SEM)
using the statistics software package LISREL 8.72. First, the measurement model
for the exogenous side of the model was set up and analysed. The configuration
along with parameters is shown in Figure 58.
As can be seen in Table 15, there is evidence for convergent validity, as all factor
loadings are highly significant. Moreover, Construct reliability was highly
satisfactory, as the parameters for each corresponding latent variable exceeds 0.7
(Nunally, 1978). The amount of variance extracted were below but relatively close
to the threshold for all constructs but one (Homburg, 2000).
197
Basic factor resources (ȟ2): % Variance extracted: 44.3%, Construct reliability: 0.70
Ȝx10 0.630 8.34 0.349 į10 0.651 8.97
Ȝx11 0.651 9.86 0.424 į11 0.576 7.24
Ȝx12 0.714 10.44 0.510 į12 0.490 6.10
Advanced factor resources (ȟ2): % Variance extracted: 50.5%, Construct reliability: 0.75
Ȝx13 0.742 10.35 0.551 į13 0.450 5.31
Ȝx14 0.711 10.87 0.506 į14 0.495 6.02
Ȝx15 0.678 10.66 0.460 į15 0.541 7.39
Inter-factor Correlations
ĭ12 0.857 17.22 ĭ24 0.773 9.49
ĭ13 0.839 15.89 ĭ25 0.683 9.31
ĭ14 0.873 13.71 ĭ34 0.873 13.66
ĭ15 0.599 8.23 ĭ35 0.779 11.65
ĭ23 0.963 20.91 ĭ45 0.767 11.50
Notes:
All parameters are significant at the 0.001 level
Next, the measurement model for the endogenous side of the model was set up
and analyzed according to the same procedure as with the exogenous side. The
configuration along with parameters is shown in Figure 59.
198
Although the item reliability of some parameters were relatively low, all
parameters proved to be highly significant (Table 15). Thus, evidence for
convergent validity is adequate. The parameters for Ȝy4 were exceptionally low; as
a result, it was altogether dropped from the model in order to improve reliability
and validity. In this way, variance extracted and construct reliability of Ș1
increased from 38.2% to 44.6% and from 0.77 to 0.80, respectively.
Purchasing process efficiency (Ș1): Variance extracted: 44.6%, Construct reliability: 0.80
Ȝy1 0.793*** 12.34 0.629 İ1 0.371*** 5.81
Ȝy2 0.657*** 9.46 0.432 İ2 0.569*** 8.41
Ȝy3 0.787*** 12.07 0.619 İ3 0.381*** 6.16
Ȝy4 0.259*** 3.386 0.067 İ4 0.933*** 9.12
Ȝy5 0.537*** 7.19 0.288 İ5 0.712*** 9.21
Ȝy6 0.510*** 5.97 0.260 İ6 0.785*** 9.48
Purchasing process effectiveness (Ș2): Variance extracted: 51.3%, Construct reliability: 0.75
Ȝy7 0.679*** 9.41 0.461 İ7 0.539*** 7.36
Ȝy8 0.554*** 9.08 0.307 İ8 0.572*** 7.78
Ȝy9 0.879*** 12.06 0.773 İ9 0.227** 2.64
Inter-factor Correlations
Ȍ12 0.856*** 25.07
Notes:
** Significant at the 0.01 level, *** Significant at the 0.001 level
199
Discriminant Validity
Discriminant validity was evaluated through chi-square difference tests for each
pair of constructs in the measurement model. Here, discriminant validity is said to
exist when an unconstrained model (correlation coefficient between constructs are
set free) vis-à-vis a constrained model (correlation coefficient set to 1.0) has
significantly lower chi-square value. This indicates that the constructs are not
perfectly correlated and thus discriminant validity is acceptable (J. C. Anderson &
Narus, 1984; Bagozzi, Yi, & Philipps, 1991; Jöreskog, 1971). On the exogenous
side, the chi-square difference test for intra-firm resources and inter-firm
resources also showed acceptable discriminant validity (Satorra-Bentler Ȥ21df =
230.25 – 162.01 = 68.24, p < 0.001) (Table 17).
In order to assess the fit of the measurement model, two nested models were
compared for the exogenous and endogenous side, respectively (one oblique and
one restricted where the correlation is fixed to 1.0) (J. C. Anderson & Gerbing,
1988). The results are shown in Table 17. As can be seen, the oblique
measurement models provide a better fit to the data than the restricted ones.
Consequently, the restricted measurement models were omitted for further
examination. The model parameters proved to be highly significant.
Also the chi-square values associated with the measurement models were
significant; these values are dependent on sample size and are inflated due to non-
normal properties of the data, the Satorra-Bentler Ȥ2/df ratio is applied in this
study (Bagozzi & Yi, 1988; Cagli, 1984; Hulland, Chow, & Lam, 1995; Satorra &
Bentler, 1988). According to Carmines and McIver (1981), this ratio should be
less than 3.0. In a similar manner, goodness-of-fit (GFI), root mean square error
of approximation (RMSEA), normed fit index (NFI), comparative fit index (CFI),
parsimonious goodness-of-fit index (PGFI) were used for this study. With the
exception of RMSEA for both oblique measurement models (RMSEA > 0.10), all
fit indices were within the range that indicates good fit to the data (i.e. Ȥ2/df 3.0,
200
GFI 0.80 NFI, CFI > 0.90) (Carmines & McIver, 1981; Hoyle, 1995;
Marcoulides & Schumacker, 1996).
Based on this, the conclusion that can be drawn is that the hypothesized
measurement models provide a better fit to the data than rival nested models. The
PGFI is usually around 0.50 for most models. In this case, PGFI for the two
oblique exogenous and endogenous measurement models were 0.55 and 0.63,
respectively. Hence, the parsimonious fit was relatively high.
Table 17. Fit indices for the exogenous side of the measurement model
2 2
Model S-B Ȥ df S-B Ȥ /df GFI RMSEA NFI CFI PGFI
Notes:
Null model: Ȥ2 = 3,673.36, S-B Ȥ2 = 2898.71, df = 105, S-B Ȥ2/df = 27.61, GFI = 0.29, RMSEA = 0.37, NFI = 0.21, CFI = 0.22,
PGFI = 0.25
On the endogenous side, the chi-square difference test for purchasing process
effectiveness and purchasing process efficiency indicated acceptable discriminant
validity (Satorra-Bentler Ȥ21df = 119.43 – 56.17 = 63.26, p < 0.001) (Table 18).
Table 18. Fit indices for the endogenous side of the measurement model
2 2
Model S-B Ȥ df S-B Ȥ /df GFI RMSEA NFI CFI PGFI
Notes:
Null model: Ȥ2 = 930.30, S-B Ȥ2 = 635.14, df = 28, S-B Ȥ2/df = 22.68, GFI = 0.46, RMSEA = 0.33, NFI = 0.32, CFI = 0.33,
PGFI = 0.36
According to Hair et al. (1992), there are two methods of exploratory factor
analysis: Principal component analysis and common factor analysis. Principal
component analysis is appropriate when the researcher is seeking to find the
201
minimum number of factors needed to account for the maximum portion of the
variance represented in the original set of items. In contrast, common factor
analysis is appropriate when the researcher is seeking to identify the latent
dimensions or constructs represented in the original items.
In line with the first stage of the factor analysis in this study, principal component
analysis was the preferred choice as it can determine how and to what extent items
are linked to their underlying factors (Byrne, 1998). For this study, the principal
component analysis was configured in a way that prevents an item from loading
on more than one factor in order to assure unidimensionality (Hughes et al., 1986).
In this case, it is hypothesized that three items are selected for measuring each
latent construct representing resources.
Factor loadings are used to present these relations. According to Hair et al. (1992),
factor loadings greater than 0.30 are considered significant; loadings of 0.40 are
considered more important; if the loadings are 0.50 or greater, they are considered
very significant. For this study, a factor loading of 0.50 was used as cut-off point.
The most commonly used method of determining whether items are loading on
one factor is the so called latent root criterion,16 implying that only the factors
having eigenvalues greater than 1 are considered significant; those with
eigenvalues less than 1 are considered insignificant and are omitted for further
analysis (Hair et al., 1992).
First, the principal component analysis was applied for all indicator variables on
the exogenous side of the research model using varimax rotation and eigenvalues
greater than one. The outcome was very successful, resulting exactly in the the
hypothesized factor structure. The results from this factor analysis are
summarized in Table 19.
16
The latent root criterion is frequently referred to as the Kaiser criterion in statistics literature (cf.
Backhaus et al., 2003)
202
Notes:
a
Based on rotation of sums of squared loadings
All significance levels < 0.001
Next, the three indicator variables on the endogenous side of the research model
were subject to a principal component analysis, also using varimax rotation and
egigenvalues greater than one. The nine items were reduced to two factors,
accounting for 68.2 and 55.6 percent of the total variance in the data, respectively
(depicted in Table 20).
203
Notes:
a
Based on rotation of sums of squared loadings
Significance level < 0.001
The program of SPSS 12.0 was used to perform the factor analysis, each scale
being factor analyzed separately. From Table 19 and Table 20, it was clear that all
items had factor loadings greater than 0.50. In order to prevent an indicator
variable to load on more than one factor, the varimax rotation was used. The
factor analysis showed that each indicator variable loaded according to the
hypothesized structure.
Reliability
Reliability refers to whether one gets the same result by using an instrument to
measure something repeatedly (Bernard, 2000). It measures the extent to which an
experiment, test or measuring procedure renders the same results over several
trials (Carmines & Zeller, 1979). In other words, it is a statistical measure of the
reproducibility of the data gathered by a survey instrument (Litwin, 1995).
Generally, four different methods are available for measuring reliability: (1) the
test-retest method, (2) the alternate-form method, (3) the split-halves method and
(4) the internal consistency method (Nunally, 1978).
204
The test-retest reliability method implies conducting a survey repeatedly using the
same set of respondents for each of the surveys. Subsequently, correlation
coefficients are calculated to compare the sets of responses. Thus, the correlation
coefficients give an estimate of the test-retest reliability of the survey instrument.
In this way, reliability can be assessed by measuring the reproducibility. Here, a
correlation coefficient equal or exceeding 0.70 is usually set as a lower threshold
for sufficient reliability (Litwin, 1995).
The second method, the alternate-form method, is a means for assessing the
reliability of a survey instrument. By rewording questions and responses, or
change their order, the objective is to create items that are similar but not identical;
the items are different only in wording or order. The items or scales of the survey
instrument are then used for the same population repeatedly in different points of
time. Then, the correlation coefficients are calculated in a similar fashion to the
test-retest method. Again, correlation coefficients equal to or exceeding 0.70 are
considered to have adequate reliability (Litwin, 1995).
The third method, the split-halves method, uses a slightly different approach than
the two former ones. Here, the sample is divided into two halves, where reliability
is estimated for each halves then laterally compared with each other. It is
important to remember that the halves are randomly selected in order to eliminate
group differences. Furthermore, the sample must also be large enough to allow
reliability assessment of each half (Litwin, 1995).
The fourth reliability method, internal consistency method, is the most commonly
applied psychometric measure for survey instruments and scales. Internal
consistency measures how well a set of items or scales measure a common latent
construct. Thus, the items or scales measuring the construct must be clearly
focused on that particular construct intended to be measured. The most well-
known statistic to estimate internal consistency is Cronbach’s alpha coefficient
(Cronbach, 1951; Nunally, 1978). The coefficient measures the internal
consistency among a set of items that are conjoined to form a common construct
205
or latent variable. In other words, it reflects the homogeneity of the scale. As with
the other reliability methods, a coefficient equal to or higher than 0.70 is
considered adequate (Nunally, 1978).
Item Analysis
The values of item to scale correlations should be greater than 0.50; those lower
than 0.50 do not share enough variance with the rest of the items in that scale.
Therefore, it is assumed that the items are not measuring the same construct and it
should be deleted from the scale (Kemp, 1999). As an example, Saraph et al.
(1989) used this method to evaluate the assignment of items to scales in
developing their instrument for measuring the critical factors of quality
management. It was judged that item analysis should be performed in order to
understand whether items were assigned appropriately. As is the case with the
206
Validity
In this research, however, it was argued that the fifteen scales for measuring
accessibility to resources and the nine scales for measuring purchasing process
performance had content validity since the development of these measurement
items was based on an extensive literature review and expert interviews with a
group consisting of both academics and practitioners.
A common critique against this type of causal analysis methods is the prevalence
of retrospective and self-serving bias (March & Sutton, 1997). However, since
each respondent most likely is equally biased in every question, the bias is
automatically eliminated through the inherent way correlation coefficients are
calculated. a common critique against these types of quantitative analysis methods
is the prevalence of retrospective and self-serving bias (March & Sutton, 1997).
However, the questions being asked were done in isolation, i.e. first the
availability of resources was assessed, then purchasing performance was assessed
separately. Thus, questions like “to what extent did resource X help you achieving
cost reductions” were avoided in order to minimize negative effects from self-
serving and retrospective bias.
What also comes into question is of course causality, which can be established
either deductively (i.e. based on theory) or empirically through longitudinal
studies (Mol, 2003). Since this study is cross-sectional, the causal directions are
deductively supported from the extensive body of research presented earlier in
this paper. Furthermore, from a nomological perspective, the effects in the
208
“reverse causal direction” are considered negligible, i.e. it is much more likely
that cost savings follow effective supplier development rather than vice versa.
į1 x1 ʌȟ1
į2 x2 ʌȟ2
ȟ1 ȗ1
ʌȟ3
į3 x3 Ȗ11
į7 x7 ʌȟ7 Ȗ31
ʌȟ8
į8 x8 ȟ3 Ȗ32 ȗ2
ʌȟ9
į9 x9 Ȝy5 y5 İ5
Another issue that needs to be addressed is the choice of indicator mode, i.e. the
question whether to use a reflective or a formative modeling approach. Indicator
mode is primarily determined by the theory behind the model; the way in which
unobservable (latent) constructs are conceptualized (Fornell & Bookstein, 1982).
Reflective indicators mean that the causal link is going from the underlying latent
variable towards each of the indicator variables. The indicators are said “to be
caused by the latent construct”. In other words, the latent variable is the cause,
whereas the effect is reflected in the indicators. An example of a reflective latent
209
In contrast, formative indicators are totally opposite. In this case, the indicator
variables are said “to cause the latent variable”. An example here could be socio-
economic status (SES), which can be formatively represented by indicator
variables such as education, income and occupational prestige (Chin, 1998).
Clearly, if a person would lose his/hers job, a decline would be observed in the
SES latent variable. However, the opposite is not necessarily true; a decline in
SES does not necessarily mean that a person has lost his/hers job. Furthermore, a
change in an indicator (e.g. income) does not necessarily imply a similar
directional change for the other indicators (e.g. education or occupational prestige)
(Chin, 1998). Therefore, such a latent construct is without doubt of formative
nature.
With this in mind, the next question is whether to apply partial least squares
regression or maximum likelihood path analysis with observed variables. Both
methods have their respective advantages and disadvantages. Partial least squares
regression is advantageous since it does not require normally distributed data, it is
210
stable against multicollinearity, and performs well under circumstances where the
number of indicator variables is large in comparison with the sample size (Abdi,
2003). However, the drawback is that there is limited ways of statistically testing
overall model fit. Maximum likelihood path analysis, on the other hand, generates
a test statistic which is distributed as Ȥ2. The test statistic gives an indication on
how well the maximum likelihood estimation has succeeded to reconstruct the
covariance matrix of the variables. In other words, it indicates the discrepancy
between the original covariance matrix and the regenerated covariance matrix.
Thus, the Ȥ2 should be as low as possible and insignificant which indicates no
significant difference (Bollen, 1989).
This test statistic can in turn be used to determine absolute, relative, as well as
parsimonious fit. The disadvantage though is the fact that data is assumed to have
multivariate distribution in order not to bias parameter estimates (Boomsma &
Hoogland, 2001). Fortunately, as the principal components calculated earlier are
inherently normally distributed, they could be used to perform a maximum
likelihood observed variable path analysis in LISREL (e.g. Blacock, 1964; Bollen,
1989; Kelloway, 1998). The model along with parameters is depicted in Figure 60.
The initial model, with 21 structural paths, is exactly equal to the number of
equations, namely k(k – 1) / 2 = 21, where k = 7 (i.e. the number of variables).
This yields a so called just-identified (or saturated) model which fits to the data
perfectly. Hence, no indices of fit are generated as they would all show ideal
values. The parameter estimates are shown in Table 21.
As can be seen in Table 21, two path coefficients proved to be insignificant (Ȗ11
and Ȗ42). In line with the argument to remove paths that do not contribute to total
explained variance (Gujarati, 1995), we obtain a model with two degrees of
freedom after fixing the insignificant paths to zero, thus making the model
overidentified. The good thing now is that it is possible to obtain a quantitative
assessment of absolute, relative and parsimonious fit.
211
Parameter Standard
(Hypothesis) coefficient t-value p-value
Notes:
R2(Ș 1) = 0.185, R2(Ș2) = 0.183, * Significant at the 0.05 level, ** Significant at the
0.01 level, *** Significant at the 0.001 level
It is common practice to report several indexes of overall fit when examining the
results from LISREL analyses, since there is currently little to no consensus
regarding which fit indexes produce the most reliable results (Marsh, Balla, &
McDonald, 1988; Raykov, Tomer, & Nesselroade, 1991; Tanaka, 1993). The
corresponding fit indices are shown in Table 22, which clearly show that the
model fits very well to the data. The chi-square value at 1.648 is highly
insignificant (p = 1.00, df = 17).
2
Ȥ p df GFI AGFI RMR NFI CFI PNFI PGFI
Comparative fit indices are used to compare the fit of one model with the fit of
another. The normed fit index (NFI) is calculated by determining how much better
the hypothesized model is compared to a so called null model, which essentially is
a model where all possible paths are constrained to zero (i.e. no relationships at
all). Like GFI and AGFI, the value ranges between zero and one, and the value
should exceed the cut-off point of 0.90 to indicated good fit (Bentler & Bonett,
1980). NFI for the model in this study was 0.998 which indicates a strong
comparative fit. Similar to the AGFI, the so called comparative fit index (CFI)
proposed by Bentler (1990) also considers the degree of freedom. CFI in this case
was 1.00, which indicates a perfect fit. Like the earlier fit indices, this value
should exceed 0.90.
The root mean squared residual (RMR) is measuring the mean of the squared
discrepancies between the observed and the hypothesized covariance matrices.
Values below 0.05 are considered as good fit; for this study, the value is 0.0178
and hence indicates a very good fit ( 0.05) (Kelloway, 1998).
The parsimony goodness-of-fit index (PGFI) takes the complexity of the model
into account when estimating its overall fit (Byrne, 1998). This is done by
considering the number of estimated parameters when assessing the fit of the
model. Both the goodness-of-fit of the model (i.e., the GFI) and the parsimony of
the model are used to generate the PGFI (Mulaik et al., 1989). Although there is
no cut-off point recommended for PNFI and PGFI in the literature (Kelloway,
1998), it is typically around 0.50 (Byrne, 1998; Mulaik et al., 1989). For this
model, the parsimony normed fit index (PNFI = 0.800) and parsimony goodness-
of-fit (PGFI = 0.606) indices can therefore be considered relatively high. The
213
intuitive reason to the relatively high value is the low complexity of the model
structure.
17
The actual test was performed by using a small software utility called “FZT”, which was
developed by Prof. Dr. Karl Wuensch and colleagues at the Department of Psychology at East
Carolina University.
215
Hypothesis 8a dealt with the relative impact between advanced factor resources
and basic factor resources on purchasing process effectiveness. Analogous to the
test of hypotheses 4a and 4b, Hotelling’s t-test for “correlated correlations” was
used to test this hypothesis. The difference between the beta coefficient of
advanced factor resources (Ȗ51 = 0.176) and basic factor resources (Ȗ41 = 0.164)
was observed but insignificant (one-sided test, df = 197, t = 0.1227 and p =
0.1217). Hypothesis 8b investigated the relative impact between advanced factor
resources and basic factor resources on purchasing process efficiency. Since the
relationship between basic factor resources and purchasing process efficiency (Ȗ41)
was insignificant, hypothesis 8b was trivially accepted.
†
1a. Purchasing process effectiveness is positively influenced by the extent of 0.313
tangible resources in possession by the firm sourcing in low-cost
countries.
1b. Purchasing process efficiency is positively influenced by the extent of 0.0140**
tangible resources in possession by the firm sourcing in low-cost
9
countries.
†
5a. Capabilities have a higher impact on purchasing process effectiveness 0.0981
than intangible resources when sourcing in low-cost countries. (wrong sign)
5b. Capabilities have a higher impact on purchasing process efficiency than 0.0671*
intangible resources when sourcing in low-cost countries.
9
6a. Purchasing process effectiveness is positively influenced by the extent of 0.0127**
basic factor resources in possession by the firm sourcing in low-cost
9
countries.
†
6b. Purchasing process efficiency is positively influenced by the extent of 0.481
basic factor resources in possession by the firm sourcing in low-cost
countries.
8a. Hypothesis 8a: Advanced factor resources have a higher impact on 0.1217†
purchasing process effectiveness than basic factor resources.
8b. Hypothesis 8b: Advanced factor resources have a higher impact on N/A
a
9
purchasing process efficiency than basic factor resources.
An overview of scholarly work about RBV was provided in chapter 5.1, and also
put in perspective by comparing it with other plausible theories from economics,
organizational theory and social science. It was also concluded that, despite a
consideration of internal factors around the middle of the 20th century, external
factors were almost totally dominating strategic management literature during the
1980s and major part of the 1990s. The chapter also explained the fundamental
logic of RBV and the causal chain with corresponding constructs. Another
conclusion was that RBV is only one (although dominating) stream of research
out of several others such as dynamic capabilities view (Eisenhardt & Martin,
2000; Teece et al., 1997; Winter, 2003), operand/operant resource view (Vargo &
Lusch, 2004), and the resource-advantage theory of competition (S. D. Hunt &
Morgan, 1996).
Chapter 5.2 dealt with the build-up of a theoretical framework by applying RBV
as a theoretical lens onto the LCCS topic. A review of they key contributions led
to a resource classification according to what a company has and what it does,
namely the five resource categories tangible resources, intangible resources,
capabilities, basic factor resources, and advanced factor resources. The possession
of each of these resources was hypothesized to positively influence purchasing
performance. Depending on their value, rareness, inimitability, and organizational
support, it was also hypothesized that intangible resources have a higher impact
on purchasing performance than tangible resources, capabilities have a higher
219
impact than intangible resources, and advanced factor resources have a higher
impact than basic factor resources. The hypotheses, all in all seventeen of them,
were then connected to form a research model. As a last step, the various
construct in the research model were operationalized.
In chapter 5.3, the research model was tested by using empirical data collected
according to the description in chapter 3.2. The analysis started by performing a
confirmatory factor analysis in order to validate the theorized factor structure. The
factor analysis confirmed the factor structure, although one item had to be
dropped due to its low item-to-total correlation. Then, the same factor structure
was used to generate non-correlated and normally distributed factors through
principal component analysis. After assessing reliability and validity of the
constructs, they were used for the operationalization of the research model
through path analysis, which in this case essentially is a principal component
regression model. The analysis showed that twelve out of the total number of
seventeen hypotheses were accepted. Therefore, the theoretical model is
considered to be sufficiently validated. Overall, the research model has a
relatively high explanatory power in comparison with similar empirical studies
deploying regression models and path analyses.
221
Second, the review of earlier empirical studies show that only a minor share of
firm performance can be contributed to external factors, thus leaving a knowledge
gap regarding the importance of intra-firm effects when sourcing in LCCs. RBV
proved to be a plausible theory for explaining the effects of intra-firm factors and
is one stream of research together with several others that acknowledge strategic
assets and competencies as the ultimate source of competitive advantage. It is
claimed that these must be kept away from competitors in order to preserve their
uniqueness. By combining the more classic approaches of RBV with other
streams of research that acknowledge inter-firm resources, e.g. mutually
222
Third, the results from the empirical study support the hypothesized strategic
importance of purchasing and supply management that has been under scrutiny
for several years (Mol, 2003; Ramsay, 2001a). The results were in line with
expectations, where all resource types but one (i.e. basic factor resources) were
significant and positively related to purchasing process performance. Thus, the
results support the paradigm of the firm as a seeker of an increased research pool,
not only through purchased goods, but more importantly through similar and
complementary higher-order resources (Chung et al., 2000; Wilkinson et al.,
2005), inter-firm resources (Dyer & Sing, 1998), and through mutual adaptations
(Hallén et al., 1991).
The relative importance of the individual resource types still remains somewhat
unclear due to the fact that only the hypothesized higher importance of intangible
resources over tangible resources was significant. Theoretically, capabilities were
expected to score highest, as this category has frequently been claimed to be the
most important source of competitive advantage (Teece & Pisano, 1998).
Although the capabilities investigated in this study are indeed more difficult to
acquire than the tangible resources, it is unlikely that they are core competencies
of any firm, considering the definition in literature (Prahalad & Hamel, 1990).
This can explain why these two resource categories had approximately the same
impact on purchasing process performance.
The importance of intangible resources and advanced factor resources also scored
high, which was also in line with expectations. Hence, it is evident that valuable,
difficult-to-imitate resources, such as technology transfer (Goh, 2005) and the
higher-order capability to perform the actual transfer (Winter, 2003), play
important roles for the outcome of sourcing initiatives in emerging and transition
economies.
223
Somewhat surprising, the importance of basic factor resources was not significant.
This leads to the conclusion that such resources can easily be obtained by any
company that is locally present in the sourcing market and any competitive
advantages that could potentially arise will quickly erode, a claim that has been
noted for long by several scholars (Barney, 1986b; Ramsay, 2001b).
To sum up, the reason why three pairwise comparisons of resource was significant,
leading to rejection of hypothesis 4b, 5a and 8a, was due to the fact that the
observed regression coefficient was too small to yield significant results.
Therefore, sources of error such as inadequate models or data can be excluded.
The key insight here is the fact that the internal and external assessments should
not be seen as two isolated entities, but rather two integral process steps within the
sourcing strategy development. By bridging the gap between the traditional
market-based positioning school with the resource-based view, the role of the
purchasing and supply function as a mediator between supply markets and the
internal customers becomes apparent. Hoopes et al. (2003) proposed a promising
generic framework based on resource heterogeneity to bridge this gap.
Second, as activities at operational level are derived from strategy, this highlights
the importance of purchasing and supply management, where managerial skills
play a key role in locking up scarce resources in the external environment,
224
Third, this study has also shown that basic factor resources (i.e. traditional factor
inputs such as land and labor) played an insignificant role. This appears to be
counterintuitive at a glance, as practitioners often state short-term benefits such as
inexpensive labor and tax incentives to be the primary reasons for sourcing in
emerging and transition economies. The lifespan of such benefits does not seem to
be more than temporary, as they can be appropriated by virtually any firm. Hence,
sole search for such short-term benefits would support the thesis of purchasing’s
strategic irrelevance (Ramsay, 2001a).
This implies that the strategic focus should be placed on developing the firm’s
organization in general and the know-how of the employees in particular. These
strengths can then be used in the every-day challenges that the purchasing staff are
facing, such as negotiations, setup of joint ventures and strategic supplier
relationships, among others, which ultimately lead to competitive advantages
stemming not only from the own firm, but also from an extension of the firm in
terms of inter-firm resources (Dyer & Sing, 1998), that are hard for the
competitors to duplicate.
Fourth, the fact that the share of goods and services sourced in LCCs over the next
five-year period is estimated to increase by more than sixty percent indicates a
very high momentum in the field of LCCS. This also means that indicates that
companies will continue to search for suppliers in which they can reduce cost
even further while assuring quality. Overall, the home country of the buying firms
will play a decreasing role in the future, where the “center of gravity” of the
business landscape is slowly shifting towards more pronounced low-cost regions.
Interestingly, the home country of the buying firms will not decrease as much in
importance as the neighbouring countries, potentially indicating a certain degree
of national protectionism. The rising star on the LCCS sky is without doubt China,
which is experiencing tremendous economic growth due to massive foreign direct
225
Second, only large-sized firms were included in this study. Since large-sized firms
generally have more resources available to deploy through subsidiaries in foreign
countries, they can potentially overcome barriers associated with sourcing in
emerging and transition economies. As a result, the results from this study may
not be generalizable to small or medium-sized firms. This opens up a window of
opportunity for future research where SMEs are examined in order to see whether
differences in performance implications from sourcing in emerging and transition
economies exist.
Third, as the sample in this study represents a wide range of industries, one has to
be cautions about drawing conclusions at industry-level as the number of firms in
some industry categories (e.g. financial services) is very low and thus cannot
provide any statistically significant results. For the larger industry categories, such
as food and agriculture, machinery, and energy/resources/material, a more
detailed analysis may be worthwhile.
Fourth, as the aim of this study is to validate a theoretical model for sourcing in
emerging and transition economies, the number of predictors in this study is
limited to fifteen, in order to preserve parsimony. In case a normative model was
to be developed, a more exhaustive range of predictors would have to be
considered to capture the idiosyncrasy and complexity that characterize real-life
226
sourcing contexts. However, this is beyond the scope of this study. Fifth, as many
firms have their own specific performance objectives, it is likely that the
performance measures used in this study do not perfectly correspond to the
measures of the responding firms.
227
References
Aaker, D. A. (1998). Strategic Market Management (5 ed.). New York: John
Wiley & Sons.
Abercrombie, N., Hill, S., & Turner, B. S. (1994). Dictionairy of sociology (3 ed.).
London: Penguin Books.
Ajzen, I., & Fishbein, M. (1980). Understanding Attitudes and Predicting Social
Behavior. Englewood Cliffs, NJ: Prentice-Hall, Inc.
Alreck, P. L., & Settle, R. (1995). The Survey Research Handbook. New York:
Irwin/McGraw-Hill.
Anderson, H., Havila, V., & Salmi, A. (2001). Can You Buy a Business
Relationship?: On The Importance of Customer and Supplier
Relationships in Acquisitions. Industrial Marketing Management, 30(7),
575-586.
Andrews, K. R. (1971). The Concept of Corporate Strategy. New York, NY: Dow
Jones-Irwin.
Astley, W. G., & Van de Ven, A. H. (1983). Central perspectives and debates in
organisation theory. Administrative Science Quarterly, 28(2), 245-273.
Babakus, E., Ferguson, C. E., Jr., & Jöreskog, K. G. (1987). The Sensitivity of
Confirmatory Maximum Likelihood Factor Analysis to Violations of
Measurement Scale and Distributional Sampling. Journal of Marketing,
24(2), 222-228.
Backhaus, K., Erichson, B., Plinke, W., & Weiber, R. (2003). Multivariate
Analysemethoden (10 ed.). Heidelberg: Springer Verlag.
Baghai, M., Coley, S. C., & White, D. (1999). Turning capabilities into
advantages. McKinsey Quarterly(1), 100-109.
Bagozzi, R. P., & Yi, Y. (1988). On the Evaluation of Structural Equation Models.
Journal of the Academy of Marketing Science, 16(1), 74-94.
Bagozzi, R. P., Yi, Y., & Philipps, L. W. (1991). Assessing Construct Validity in
Organizational Research. Administrative Science Quarterly, 36(3), 421-
459.
Baily, M. N., & Farrell, D. (2004). Exploding the Myths About Offshoring. San
Francisco, CA.: McKinsey Global Institute.
Barney, J. B., & Tyler, B. (1990). The attributes of top management teams and
sustained competitive advantage. In M. Lawless & L. R. Gomez-Mejia
(Eds.), Strategic Management in High Technology Firms. Greenwich, CT.
Bartlett, C. A., Doz, Y. L., & Hedlund, G. (1990). Introduction: The changing
agenda for researchers and practitioners. In C. A. Bartlett, Y. L. Doz & G.
Hedlund (Eds.), Managing the Global Firm. London: Routledge.
231
Bartlett, C. A., & Goshal, S. (1987). Managing across borders: New strategic
requirements. Sloan Management Review, 28(4), 7-17.
Battacharya, A., Bradtke, T., Hemerling, J., Lebreton, J., Mosquet, X., Rupf, I., et
al. (2004). Capturing Global Advantage. Boston: Boston Consulting
Group.
Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in
the analysis of covariance structures. Psychological Bulletin, 88, 588-606.
Berry, D., Towill, D. R., & Wadsley, N. (1994). Supply chain management in the
electronics products industry. International Journal of Physical
Distribution & Logistics Management, 24(10), 20-32.
Black, J. C., & Boal, K. B. (1994). Strategic resources: Traits, configurations and
paths to sustainable competitive advantage. Strategic Management Journal,
15(131-148).
Bollen, K. A. (1989). Structural Equations With Latent Variables. New York, NY:
Wiley.
Bozarth, C., Handfield, R. B., & Das, A. (1998). Stages of global sourcing
strategy evolution: an exploratory study. Journal of Operations
Management, 16(2-3), 241-255.
Bresnahan, T. F., Reiss, P. C., Willig, R., & Stigler, G. J. (1987). Do Entry
Barriers Vary Across Markets? Comments and Discussion. Brookings
Papers on Economic Activity, 1987(3).
233
Buvik, A., & Reve, T. (2001). Assymetrical deployment of specific assets and
contractual safeguarding in industrial purchasing relationships. Journal of
Business Research, 51(2), 101-113.
Caves, R. E., & Porter, M. E. (1977). From entry barriers to mobility barriers:
Conjectural decisions and contrived deterrence to new competition.
Quarterly Journal of Economics, 91(2), 241-262.
Chandler, J., A.D., Sölvell, Ö., & Hagström, P. (1998). The Dynamic Firm: The
Role of Technology, Strategy, Organisation, and Regions. Oxford
University Press: Oxford.
Chow, G. C. (2001, March 22-24, 2001). The Impact of Joining WTO on China's
Economic, Legal and Political Institutions. Paper presented at the
International Conference on Greater China and the WTO, Hong Kong.
Chung, S., Harbir, S., & Lee, K. (2000). Complementarity, Status Similarity and
Social Capital as Drivers of Alliance Formation. Strategic Management
Journal, 21(1), 1-22.
Cool, K., & Schendel, D. (1988). Performance differences among strategic group
members. Strategic Management Journal, 9(3), 207-233.
Cooper, H., & Hedges, L. V. (1994). The Handbook of Research Synthesis. New
York: Sage.
236
Cooper, M., Lambert, D. M., & Pagh, J. D. (1997). Supply Chain Management:
More Than a New Name for Logistics. International Journal of Logistics
Management, 8(1), 1-14.
Cousins, P. D., Handfield, R. B., Lawson, B., & Petersen, K. J. (2006a). Creating
supply chain relational capital: The impact of formal and informal
socialization processes. Journal of Operations Management, In Press,
Corrected Proof.
Cousins, P. D., Handfield, R. B., Lawson, B., & Petersen, K. J. (2006b). Creating
Supply Chain Relational Capital: The Impact of Formal and Informal
Socialization Processes. Journal of Operations Management, In press.
Cross, B., & Gordon, J. (1995). Partnership Strategies for Market Success.
Business Quarterly, 60(1), 91-96.
Cummings, S. (1993). The first strategists. Long Range Planning, 26(3), 133-135.
Curran, P. J., West, S. G., & Finch, J. F. (1996). The Robustness of Test Statistics
to Nonnormality and Specification Error in Confirmatory Factor Analysis.
Psychological Methods, 1, 16-29.
Dales, A., & Meese, R. (2001). Strategic Currency Hedging. Journal of Asset
Management, 2(1), 9-21.
237
Daniels, J. D., Pitts, R. A., & Tretter, M. J. (1985). Organising for dual strategies
of product diversity and international expansion. Strategic Management
Journal, 6(3), 223-237.
Dierickx, I., & Cool, K. (1989). Asset stock accumulation and sustainability of
competitive advantage. Management Science, 35(12), 1504-1511.
Dillman, D. (2002). Mail and Internet Surveys: The Tailored Design Method.
New York: Wiley.
Dirac, P. (1939). The Relation Between Mathematics and Physics. Paper presented
at the Proceedings of the Royal Society, Edinburgh.
Dobler, D. W., & Burt, D. N. (1996). Purchasing and Supply Management: Text
and Cases (6 Edition ed.). New York: McGraw-Hill/Irwin.
Doz, Y., Bartlett, C. A., & Prahalad, C. K. (1981). Global Competitive Pressure vs.
Host Country Demands: Managing Tensions in Multinational
Corporations. California Management Review, 23(3), 63-74.
Dyer, J. H., & Sing, H. (1998). The Relational View: Cooperative Strategy and
Sources of Interorganizational Competitive Advantage. Academy of
Management Review, 23(4), 660-679.
239
Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic Capabilities: What are They?
Strategic Management Journal, 21(10/11), 1105-1121.
Ellram, L. M., & Carr, A. (1994). Strategic Purchasing: A History and Review of
the Literature. International Journal of Purchasing and Materials
Management, 20(2), 9-18.
Fawcett, S. E., & Birou, L. M. (1992). Exploring the Logistics Interface Between
Global and JIT Sourcing. International Journal of Physical Distribution &
Logistics Management, 22(1), 3-14.
Fitti, J. E. (1979). Some results from the Telephone Health Interview Survey. In
Proceedings of the American Statistical Association (pp. 244-249).
Fornell, C., & Bookstein, F. L. (1982). Two Structural Equation Models: LISREL
and PLS Applied to Consumer Exit-Voice Theory. Journal of marketing
Research, 19(4), 440-452.
Frear, C. R., Metcalf, L. E., & Alguire, M. S. (1992). Offshore Sourcing: Its
Nature and Scope. International Journal of Purchasing and Materials
Management, 28(1), 32-34.
Gadde, L.-E., & Snehota, I. (2000). Making the Most of Supplier Relationships.
Industrial Marketing Management, 29(4), 305-316.
Goffin, K., Lemke, F., & Szwejczewski, M. (2006). Exploratory study of 'close'
supplier-manufacturer relationships. Journal of Operations Management,
24(2), 189-209.
Goldberg, A. (1994). Fast times, fast cycles. Marketing Computing, 14(3), 20.
241
Hahn, C. K., Kim, K. H., & Kim, J. S. (1986). Cost of Competition: Implications
for Purchasing Strategy. Purchasing and Materials Management, 22(4), 2-
7.
242
Hair, J. F., Jr., Anderson, R. E., Tatham, R. L., & Black, W. C. (1992).
Multivariate Data Analysis With Readings (3 ed.). New York: Macmillan
Publishing Company.
Haley, G. T., Haley, U. C. V., & Tan, C. T. (2004). The Chinese Tao of Business:
The Logic of Successful Business Strategy. Chichester: John Wiley & Sons.
Handfield, R. B., & Nichols Jr., E. L. (2004). Key issues in global supply base
management. Industrial Marketing Management, 33(1), 29-35.
Harvey, M. G., Novicevic, M. M., Hentch, T., & Myers, M. (2003). Global
account management: a supply-side managerial view. Industrial Marketing
Management, 32(7), 563-571.
243
Hedges, L. V., & Olkin, I. (1985). Statistical Methods for Meta-analysis. New
York: Academic Press.
Held, D., McGrew, A., Goldblatt, D., & Perraton, J. (1999). Global
Transformations: Economics, Politics and Culture. Cambridge: Polity
Press.
Helper, S., & Sako, M. (1995). U.S. and Japanese Supply Systems: Are They
Converging? Sloan Management Review, 96(3), 105-113.
Hirshliefer, J. (1980). Price theory and applications (2 ed.). Englewood Cliffs, NJ:
Prentice-Hall.
Hoopes, D. G., Madsen, T. L., & Walker, G. (2003). Guest Editors' Introduction
to the Special Issue: Why Is There a Resource-Based View? Toward a
Theory of Competitive Heterogeneity. Strategic Management Journal,
24(10), 889-898.
Hoskisson, R. E., Hitt, M. A., Wan, W. P., & Yiu, D. (1999). Theory and research
in strategic management: Swings of a pendulum. Journal of Management,
25(3), 417-456.
Hout, T., Porter, M. E., & Rudden, E. (1982). How global companies win out.
Harvard Business Review, 60(5), 98-108.
Hughes, M. A., Price, R. L., & Marrs, D. W. (1986). Linking Theory Construction
and Theory Testing: Models with Multiple Indicators of Latent Variables.
The Academy of Management Review, 11(1), 128-144.
Hulland, J., Chow, Y. H., & Lam, S. (1995). Use of Causal Models in Marketing
Research: A Review. International Journal of Research in Marketing,
13(1), 181-197.
Hussey, J., & Hussey, R. (1997). Business Research. Hampshire: Macmillan Press
Ltd.
Irvin, R. A., & Michaels, E. G. (1989). Core skills: Doing the right things right.
McKinsey Quarterly(Summer issue), 4-19.
Johnsen, T., Phillips, W., Caldwell, N., & Lewis, M. (2006). Centrality of
customer and supplier interaction in innovation. Journal of Business
Research, 59(6), 671-678.
Johnston, D. A., McCutcheon, D. M., Stuart, F. I., & Kerwood, H. (2003). Effects
of supplier trust on performance of cooperative supplier relationships.
Journal of Operations Management, 22(5), 23-38.
Jöreskog, K. G., Sörbom, D., du Toit, S., & du Toit, M. (2000). LISREL 8: New
Statistical Features. Chicago: Scientific Software International.
Kaplan, R. S., & Norton, D. P. (1992). The Balanced Scorecard - Measures That
Drive Performance. Harvard Business Review, 70(1), 71-79.
Kay, J. (1993). The structure of strategy. Business Strategy Review, 4(2), 17-37.
Keiding, N., Morgan, B., Speed, T., & van der Heijden, P. (Eds.). (2004).
Generalized Latent Variable Modeling: Multilevel, Longitudinal and
Structural Equation Models. New York: Chapman and Hall/CRC.
Kevin, J. L., Lam, C. K., Karakowsky, L., & Gongming, Q. (2003). Firm
resources and first mover advantages: A case of foreign direct investment
(FDI) in China. International Business Review, 12(5), 625-645.
Kim, J.-O., & Mueller, C. W. (1978). Factor Analysis: Statistical methods and
practical issues. Thousand Oaks, CA: Sage Publications.
Klein, J. A., Edge, G. M., & Kass, T. (1991). Skill-based Competition. Journal of
General Management, 16(4), 1-15.
248
Kogut, B., & Singh, H. (1988). The Effect of National Culture on the Choice of
Entry Mode. Journal of International Business Studies, 19(3), 411-432.
Kotabe, M., & Murray, J. Y. (1990). Linking product and process innovations and
modes of international sourcing in global competition: a case of foreign
multinational firms. Journal of International Business Studies, 21(3), 383-
408.
Kotabe, M., Murray, J. Y., & Javalgi, R. G. (1998). Global Sourcing of Services
and Market Performance: An Empirical Investigation. Journal of
International Marketing, 6(4), 10-31.
Kotabe, M., & Omura, G. (1989). Sourcing strategies of European and Japanese
multinationals: a comparison. Journal of International Business Studies,
20(Spring), 113-130.
249
Kumar, N., Scheer, L. K., & Steenkamp, J.-B. (1993). Powerful Suppliers,
Vulnerable Resellers, and the Effectiveness of Supplier Fairness: A Cross-
National Study. University Park, PA: Institute for the Study of Business
Markets, Penn State University.
Lancioni, R., & Gattona, J. L. (1992). Strategic Value Pricing: Its Role in
International Business. International Journal of Physical Distribution &
Logistics Management, 22(6), 24-27.
Lawrence, P., & Lorsch, J. (1967). Organization and Environment. Boston, MA:
Division of Research, Harvard Business School.
Learned, E., Christensen, C., Andrews, K. R., & Guth, W. (1969). Business Policy:
Text and Cases. Homewood, IL.: Irwin.
Leenders, M. R., Johnson, P. F., Flynn, A. E., & Fearon, H. E. (2002). Purchasing
and Supply Management. Homewood, IL: Richard D. Irwin Inc.
Lessler, J., & Forsyth, B. (1996). A Coding System for Appraising Questionnaires.
In N. Schwarz & S. Sudman (Eds.), Answering Questions: Methodology
for Determining Cognitive and Communicative Processes in Survey
Research (pp. 259-292). San Francisco: Josey-Bass.
Lewin, P., & Phelan, S. E. (2000). An Austrian Theory of the Firm. Review of
Austrian Economics, 13(1), 59-79.
250
Lewis, P., & Thomas, H. (1990). The linkage between strategy, strategic groups,
and performance in the UK retail grocery industry. Strategic Management
Journal, 11(5), 385-397.
Lewis, W. W., & Harris, M. (1992). Why globalisation must prevail. McKinsey
Quarterly, 2(2), 114-131.
Lieberthal, K., & Lieberthal, G. (2003). The Great Transition. Harvard Business
Review, 81(10), 70-81.
Mahoney, J. T., & Pandian, J. R. (1992). The Resource-Based View Within the
Conversation of Strategic Management. Strategic Management Journal,
13(5), 363-380.
Mason, E. (1949). The Current State of the Monopoly Problem in the U.S.
Harvard Law Review, 62, 1265-1285.
Micceri, T. (1989). The Unicorn, the Normal Curve, and Other Improbable
Creatures. Psychological Bulletin, 105, 156-166.
Miles, M., & Huberman, A. (1984). Qualitative Data Analysis. London: Sage.
Monczka, R. M., Trent, R. J., & Handfield, R. B. (1998). Purchasing and Supply
Chain Management. Cincinnati, OH: South-Western College Publishing.
Mulaik, S. A., James, L. R., van Altine, J., Bennett, N., Lind, S., & Stilwell, C. D.
(1989). Evaluation of Goodness-of-Fit Indices for Structural Equation
Models. Psychological Bulletin, 105, 430-445.
Murray, J. Y., Kotabe, M., & Wildt, A. R. (1995). Strategic and financial
performance implications of global sourcing strategy: a contingency
analysis. Journal of International Business Studies, 26(1), 181-202.
Muthén, B., & Kaplan, D. (1992). A Comparison of Some Methodologies for the
Factor Analysis of Non-Normal Likert Variables: A Note on the Size of
the Model. British Journal of Mathematical and Statistical Psychology,
45(1), 19-30.
Naisbitt, J., & Aburdene, P. (1990). Megatrends 2000. London: Sidgwick and
Jackson.
254
Nellore, R., Chanaron, J.-J., & Söderquist, K. E. (2001). Lean Supply and Price-
Based Global Sourcing - The Interconnection. European Journal of
Purchasing & Supply Management, 7(1), 101-110.
Nelson, R. R. (1991). Why do firms differ, and how does it matter. Strategic
Management Journal, 12(1), 61-74.
Nissen, K. (2005). New World Telecom: A Survival Guide for Global Equipment
Suppliers. Business Communications Review, 35(9), 57-61.
Ohmae, K. (1985). Triad Power: The Coming Shape of Global Competition. New
York: The Free Press.
Paulson, D. S. (2003). Applied Statistical Designs for the Researcher. New York:
Marcel Dekker, Inc.
255
Penrose, E. (1959). The Theory of the Growth of the Firm (2 ed.). Oxford: Oxford
University Press.
Piore, M., & Sabel, C. (1984). The Second Industrial Divide. New York: Basic
Books.
Poincaré, J. H. (1903). La science et l' hypothese [The science and the hypothesis].
Paris: E. Flammarion.
Poon, W.-Y., & Lee, S.-Y. (1994). A Distribution Free Approach for Analysis of
Two-Level Structural Equation Models: A Meta-Analysis. Journal of
Behavioral Statistics, 26(1), 105-132.
Porter, M. (1990). The Competitive Advantage of Nations. New York, NY: The
Free Press.
Prahalad, C. K., & Doz, Y. (1987). The Multinational Mission: Balancing Local
Demands and Global Vision. New York: The Free Press.
Prahalad, C. K., & Hamel, G. (1990). The core competence of the corporation.
Harvard Business Review, 68(3), 79-91.
Prahalad, C. K., & Hamel, G. (1994). The core competence of the corporation.
Harvard Business Review, 68(3), 79-91.
Priem, R. L., & Butler, J. E. (2001). Tautology in the Resource-Based View and
the Implications of Externally Determined Resource Value. Academy of
Management Review, 26(1), 57-66.
Ravallion, M., & Chen, S. (2004). China's (Uneven) Progress Against Poverty
(No. 3408): World Bank.
Raykov, T., Tomer, A., & Nesselroade, J. R. (1991). Reporting structural equation
modeling results in psychology and aging: Some proposed guidelines.
Psychology and Aging, 6(4), 499-503.
Reed, R., & Philippi, R. (1990). Causal ambiguity, barriers to imitation and
sustainable competitive advantage. Academy of Management Review,
14(1), 88-102.
Riley, M., Wood, R. C., Clark, M. A., Wilkie, E., & Szivas, E. (2003). Research
and Writing Dissertations in Business and Management. London:
Thomson Learning.
Samli, A. C., Browning, J. M., & Busbia, C. (1998). The status of global sourcing
as a critical tool of strategic planning. Journal of Business Research, 43(3),
177-187.
Satorra, A., & Bentler, P. M. (1988). Scaling Corrections for Chi-Square Statistics
in Covariance Structure Analysis. ASA Proceedings of the Business and
Economic Section, 308-313.
259
Satorra, A., & Bentler, P. M. (1994). Corrections to Test Statistics and Standard
Errors in Covariance Structure Analysis. In A. von Eye & C. Clogg (Eds.),
Latent Variables Analysis: Applications to Developmental Research (pp.
399-419). Thousand Oaks, CA: Sage Publications.
Schafer, J. L., & Graham, J. W. (2002). Missing Data: Our View of the State of
the Art. Psychological Methods, 7(2), 147-177.
Sekaran, U. (2003). Research Methods for Business. Hoboken, NJ: John Wiley &
Sons.
Shenkar, O. (2005). The Chinese Century: The Rising Chinese Economy and Its
Impact on the Global Economy, the Balance of Power, and Your Job.
Philadelphia, PA.: Wharton School Publishing.
Smith, A. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations.
Oxford: Clarendon Press.
Srivistava, R. K., Shervani, T. A., & Fahey, L. (1998). Market-Based Assets and
Shareholder Value: A Framework for Analysis. Journal of Marketing,
62(1), 2-18.
Stopford, J. M., & Wells, L. T. J. (1972). Managing the MNE: Organisation of the
Firm and Ownership of the Subsidiaries. New York: Basic Books.
Sungmin, R., & Nermin, E. (2006). The environment and its impact on
satisfaction with supplier performance: An investigation of the mediating
effects of control mechanisms from the perspective of the manufacturer in
the U.S.A. Industrial Marketing Management, In Press, Corrected Proof.
261
Teece, D., Pisano, G., & Shuen, A. (1997). Dynamic Capabilities and Strategic
Management. Strategic Management Journal, 18(7), 509-533.
Thurstone, L. L. (1931). The theory of multiple factors. Ann Arbor, MI: Edwards
Brothers.
Tomson, B., & Daniel, L. G. (1996). Factor Analytic Evidence for the Construct
Validity of Scores: A Historical Overview and Some Guidelines.
Educational and Psychological Measurement, 56(2), 197-208.
Trent, R. J., & Monczka, R. M. (1998). Purchasing and supply management trends
and changes throughout the 1990s. International Journal of Purchasing
and Materials Management, 34(4), 2-11.
262
Turnbull, P., Delbridge, R., Oliver, N., & Wilkinson, B. (1993). Winners and
Losers - The "tiering" of Component Suppliers in the UK Automotive
Industry. Journal of General Management, 19(1), 48-63.
Vargo, S. L., & Lusch, R. F. (2004). Evolving to a New Dominant Logic for
Marketing. Journal of Marketing, 68(1), 1-17.
West, S. G., Finch, J. F., & Curran, P. J. (1995). Structural Equation Models With
Nonnormal Variables: Problems and Remedies. In R. H. Hoyle (Ed.),
Structural Equation Modeling: Concepts, Issues, and Applications (pp. 56-
75). Thousand Oaks, CA: Sage Publications.
Wilkinson, I., Young, L., & Vagn Freytag, P. (2005). Business mating: Who
chooses and who gets chosen? Industrial Marketing Management, 34(7),
669-680.
Willis, G., & Lessler, J. (1999). Question Appraisal System QAS-99. Rockville,
MD: Research Triangle Institute.
Wind, Y., & Douglas, S. P. (1981). International portfolio analysis and strategy:
The challenges of the 80s. Journal of International Business Studies, 12(2),
69-80.
Vollman, T. E., Berry, W., & Whybark, D. (1984). Manufacturing Planning and
Control Systems. Homewoods, IL.: Richard D. Irwing Publishing, Inc.
Womack, J., Jones, D., & Roos, D. (1990). The Machine That Changed the World.
New York: Harper Perennial.
von Neumann, J., & Morgenstern, O. (1944). The Theory of Games and Economic
Behaviour. Princeton, NJ: Princeton University Press.
Wu, W., & Choi, W. L. (2004). Transaction Cost, Social Capital and Firms'
Synergy Creation in Chinese Business Networks: An Integrative Approach.
Asia Pacific Journal of Management, 21(1), 325-343.
Yan, R. (1999). Short-Term Success: The Litmus Test for Success in China.
Harvard Business Review, 76(5), 61-75.
Yin, R. K. (1994). Case Study Research: Design and Methods (2 ed.). Thousand
Oaks, CA: Sage.
Yuan, K.-H., & Bentler, P. M. (2002). Eight test statistics for multilevel structural
equation models. Computational Statistics & Data Analysis, 44, 89-107.
Zajac, E., Kraatz, M., & Bresser, R. (2000). Modelling the Dynamics of Strategic
Fit: A Normative Approach to Strategic Change. Strategic Management
Journal, 21(4), 753-773.
1. Introduction
Dissertation background information
Purpose of the interview
Confidentiality of responses
2. For practitioners
What is your job position?
Which industry does your company belong?
What products and services are offered to the customer?
How many employees does your company have?
What is the annual turnover of your company?
To what extent are you sourcing in LCCs?
Since when are you working at your current position?
3. For academics
To which faculty and chair do you belong?
What is your field of research?
How long have you been active in the field of research?
4. Trends
What are the most important LCCs today and in five years?
What supplies do you consider most suitable for sourcing in LCCs?
What are the most important barriers/facilitators behind LCCS?
What are the most important activities for day-to-day LCCS?
5. Implications
What are the most important internal factors (resources) for the success of
LCCS initiatives?
How do these internal factors influence the outcome of LCCS decisions?
269
Part I: General
1. In which country is the headquarters of your company located?
[Belgium, France, Germany, Italy, Luxemburg, Netherlands, Spain, UK]
3. To your nearest estimate, what was the size of the total direct purchase spend
of your company in 2005?
[Verbatim input (million Euros)]
4. To your nearest estimate, how large was the total indirect purchase spend of
your company in 2005?
[Verbatim input (million Euros)]
7. Regarding vision for LCCS initiatives, to what extent does your company
involve corporate senior manage- [1 – Not at all, 2, 3, 4, 5 – Very large extent]
ment in the development of an
LCCS vision?
align strategic LCCS goals to the [1 – Not at all, 2, 3, 4, 5 – Very large extent]
corporate goals?
9. How do you rate the following geographical areas for sourcing direct supplies
today (2005) and in the future (2010)?
10. How do you rate the following geographical areas for sourcing indirect
supplies today (2005) and in the future (2010)?
a. Central Europe [1 – Not important at all, 2, 3, 4, 5 – Very important]
b. Eastern Europe [1 – Not important at all, 2, 3, 4, 5 – Very important]
c. South Asia [1 – Not important at all, 2, 3, 4, 5 – Very important]
d. South-East Asia [1 – Not important at all, 2, 3, 4, 5 – Very important]
e. East Asia [1 – Not important at all, 2, 3, 4, 5 – Very important]
f. Central and Latin America [1 – Not important at all, 2, 3, 4, 5 – Very important]
g. Middle East [1 – Not important at all, 2, 3, 4, 5 – Very important]
h. Africa [1 – Not important at all, 2, 3, 4, 5 – Very important]
11. What are the three most important sourcing countries for direct supplies at
your company today (2005) and in the future (2010)?
12. What are the three most important sourcing countries for indirect supplies at
your company today (2005) and in the future (2010)?
14. To what extent are other departments besides the purchasing department
involved in cross-functional LCC purchasing teams at your company?
[1 – Not at all, 2, 3, 4, 5 – Very high extent]
16. To what extent are the following stakeholders driving LCCS initiatives at your
company?
CEO [1 – Not at all, 2, 3, 4, 5 – Very large extent]
CFO [1 – Not at all, 2, 3, 4, 5 – Very large extent]
COO [1 – Not at all, 2, 3, 4, 5 – Very large extent]
CPO [1 – Not at all, 2, 3, 4, 5 – Very large extent]
Customers [1 – Not at all, 2, 3, 4, 5 – Very large extent]
Other:__________ (verbatim input)
18. How do you rate the level of knowledge of the purchasing staff at your
company, regarding
LCCS strategy development [1 – Very low, 2, 3, 4, 5 – Very high]
(entry/exit/operations strategy)
LCCS process management [1 – Very low, 2, 3, 4, 5 – Very high]
cross-functional collaboration [1 – Very low, 2, 3, 4, 5 – Very high]
enterprise-wide LCCS information [1 – Very low, 2, 3, 4, 5 – Very high]
transparency (spend data etc.)
LCCS trade (economics, culture, [1 – Very low, 2, 3, 4, 5 – Very high]
politics, barriers)
supply systems [1 – Very low, 2, 3, 4, 5 – Very high]
(e-Sourcing platforms etc.)
273
20. To your nearest estimate, what percentage of your total commodities spend is
sourced in low-cost countries today (2005)?
[Verbatim input]
21. To your nearest estimate, what percentage of your total commodities spend is
sourced in low-cost countries in the future (2010)?
[Verbatim input]
274
22. To your nearest estimate, what percentage of your total services spend is
sourced in low-cost countries today (2005)?
[Verbatim input]
23. To your nearest estimate, what percentage of your total services spend is
sourced in low-cost countries in the future (2010)?
[Verbatim input]
25. To what extent are the following resources accessible to your company in
order to gain competitive advantage through LCCS?
LCCS operating plants and facilities [1 – Not at all, 2, …, 5 – Highly
accessible]
cash in hand/bank [1 – Not at all, 2, …, 5 – Highly
accessible]
performance measurement systems [1 – Not at all, 2, …, 5 – Highly
(cost, quality, delivery) accessible]
firm’s reputation in LCC [1 – Not at all, 2, …, 5 – Highly
accessible]
process/product patents [1 – Not at all, 2, …, 5 – Highly
accessible]
innovative demands of buyers [1 – Not at all, 2, …, 5 – Highly
accessible]
cost-reduction demands of buyers [1 – Not at all, 2, …, 5 – Highly
accessible]
LCCS know-how [1 – Not at all, 2, …, 5 – Highly
accessible]
ability to integrate strategic suppliers [1 – Not at all, 2, …, 5 – Highly
in design accessible]
275
27. To what extent are you using different types of purchasing channels with LCC
suppliers at your company, regarding
direct purchasing from LCC [1 – Not used at all, 2, 3, 4, 5 – Extensively
supplier used]
purchasing from LCC supplier [1 – Not used at all, 2, 3, 4, 5 – Extensively
through foreign subsidiary in LCC used]
purchasing from LCC supplier’s [1 – Not used at all, 2, 3, 4, 5 – Extensively
sales subsidiary in home country used]
purchasing through 3rd party [1 – Not used at all, 2, 3, 4, 5 – Extensively
intermediary (e.g. trading company, used]
commission house, import broker etc.)
purchasing through international [1 – Not used at all, 2, 3, 4, 5 – Extensively
procurement office (IPO) located used]
in LCC
276
28. How do the following characteristics of supplies impact on the suitability for
LCCS, regarding
supplies sourced from an LCC in [1 – Not suitable at all, 2, 3, 4, 5 – Highly
which the same supplies can be suitable]
used for products to a large extent
(i.e. large local sales market)
supplies for which a large local [Same as above]
supplier base already exists
standardised supplies [Same as above]
(i.e. uniform supplies that can be
used in a wide range of products)
time-critical supplies [Same as above]
quality-critical supplies [Same as above]
supplies with high intellectual [Same as above]
property content
supplies with high [Same as above]
technological content
supplies for mature products [Same as above]
(late stage of product life cycle)
277
Degrees of Freedom = 80
Minimum Fit Function Chi-Square = 212.98 (P = 0.00)
Normal Theory Weighted Least Squares Chi-Square = 210.92 (P = 0.00)
Satorra-Bentler Scaled Chi-Square = 162.01 (P = 0.00)
Chi-Square Corrected for Non-Normality = 511.00 (P = 0.0)
Chi-Square Difference with 25 Degrees of Freedom = 3462.45 (P =
0.0)
Estimated Non-centrality Parameter (NCP) = 82.01
90 Percent Confidence Interval for NCP = (49.50 ; 122.29)
Minimum Fit Function Value = 1.07
Population Discrepancy Function Value (F0) = 0.41
90 Percent Confidence Interval for F0 = (0.25 ; 0.61)
Root Mean Square Error of Approximation (RMSEA) = 0.072
90 Percent Confidence Interval for RMSEA = (0.056 ; 0.088)
P-Value for Test of Close Fit (RMSEA < 0.05) = 0.00
Expected Cross-Validation Index (ECVI) = 1.46
90 Percent Confidence Interval for ECVI = (1.05 ; 1.42)
ECVI for Saturated Model = 1.21
ECVI for Independence Model = 18.61
Chi-Square for Independence Model with 105 Degrees of Freedom =
3673.36
Independence AIC = 3703.36
Model AIC = 290.92
Saturated AIC = 240.00
Independence CAIC = 3767.84
Model CAIC = 462.85
Saturated CAIC = 755.80
Normed Fit Index (NFI) = 0.96
Non-Normed Fit Index (NNFI) = 0.97
Parsimony Normed Fit Index (PNFI) = 0.73
Comparative Fit Index (CFI) = 0.98
Incremental Fit Index (IFI) = 0.98
Relative Fit Index (RFI) = 0.94
Critical N (CN) = 138.98
Root Mean Square Residual (RMR) = 0.073
Standardized RMR = 0.054
Goodness of Fit Index (GFI) = 0.88
Adjusted Goodness of Fit Index (AGFI) = 0.81
Parsimony Goodness of Fit Index (PGFI) = 0.58
278
Degrees of Freedom = 90
Minimum Fit Function Chi-Square = 307.527 (P = 0.0)
Normal Theory Weighted Least Squares Chi-Square = 304.788 (P = 0.0)
Satorra-Bentler Scaled Chi-Square = 230.254 (P = 0.00)
Chi-Square Corrected for Non-Normality = 683.736 (P = 0.0)
Chi-Square Difference with 0 Degree of Freedom = 15.600 (P = 1.000)
Estimated Non-centrality Parameter (NCP) = 140.254
90 Percent Confidence Interval for NCP = (99.266 ; 188.923)
Minimum Fit Function Value = 1.545
Population Discrepancy Function Value (F0) = 0.705
90 Percent Confidence Interval for F0 = (0.499 ; 0.949)
Root Mean Square Error of Approximation (RMSEA) = 0.0885
90 Percent Confidence Interval for RMSEA = (0.0744 ; 0.103)
P-Value for Test of Close Fit (RMSEA < 0.05) = 0.000
Expected Cross-Validation Index (ECVI) = 1.833
90 Percent Confidence Interval for ECVI = (1.253 ; 1.703)
ECVI for Saturated Model = 1.206
ECVI for Independence Model = 18.610
Chi-Square for Independence Model with 105 Degrees of Freedom =
3673.364
Independence AIC = 3703.364
Model AIC = 364.788
Saturated AIC = 240.000
Independence CAIC = 3767.839
Model CAIC = 493.737
Saturated CAIC = 755.798
Normed Fit Index (NFI) = 0.937
Non-Normed Fit Index (NNFI) = 0.954
Parsimony Normed Fit Index (PNFI) = 0.803
Comparative Fit Index (CFI) = 0.961
Incremental Fit Index (IFI) = 0.961
Relative Fit Index (RFI) = 0.927
Critical N (CN) = 108.272
Root Mean Square Residual (RMR) = 0.0906
Standardized RMR = 0.0674
Goodness of Fit Index (GFI) = 0.830
Adjusted Goodness of Fit Index (AGFI) = 0.774
Parsimony Goodness of Fit Index (PGFI) = 0.623
279
Degrees of Freedom = 19
Minimum Fit Function Chi-Square = 72.80 (P = 0.00)
Normal Theory Weighted Least Squares Chi-Square = 70.59 (P = 0.00)
Satorra-Bentler Scaled Chi-Square = 59.17 (P = 0.00)
Chi-Square Corrected for Non-Normality = 73.24 (P = 0.00)
Chi-Square Difference with 9 Degrees of Freedom = 859.75 (P = 0.0)
Estimated Non-centrality Parameter (NCP) = 40.17
90 Percent Confidence Interval for NCP = (20.69 ; 67.27)
Minimum Fit Function Value = 0.37
Population Discrepancy Function Value (F0) = 0.20
90 Percent Confidence Interval for F0 = (0.10 ; 0.34)
Root Mean Square Error of Approximation (RMSEA) = 0.10
90 Percent Confidence Interval for RMSEA = (0.074 ; 0.13)
P-Value for Test of Close Fit (RMSEA < 0.05) = 0.00013
Expected Cross-Validation Index (ECVI) = 0.53
90 Percent Confidence Interval for ECVI = (0.37 ; 0.60)
ECVI for Saturated Model = 0.36
ECVI for Independence Model = 4.76
Chi-Square for Independence Model with 28 Degrees of Freedom =
930.35
Independence AIC = 946.35
Model AIC = 104.59
Saturated AIC = 72.00
Independence CAIC = 980.73
Model CAIC = 177.66
Saturated CAIC = 226.74
Normed Fit Index (NFI) = 0.94
Non-Normed Fit Index (NNFI) = 0.93
Parsimony Normed Fit Index (PNFI) = 0.64
Comparative Fit Index (CFI) = 0.96
Incremental Fit Index (IFI) = 0.96
Relative Fit Index (RFI) = 0.91
Critical N (CN) = 122.72
Root Mean Square Residual (RMR) = 0.064
Standardized RMR = 0.079
Goodness of Fit Index (GFI) = 0.92
Adjusted Goodness of Fit Index (AGFI) = 0.85
Parsimony Goodness of Fit Index (PGFI) = 0.48
281
Degrees of Freedom = 20
Minimum Fit Function Chi-Square = 131.99 (P = 0.0)
Normal Theory Weighted Least Squares Chi-Square = 151.93 (P = 0.0)
Satorra-Bentler Scaled Chi-Square = 119.43 (P = 0.00)
Chi-Square Corrected for Non-Normality = 76.05 (P = 0.00)
Chi-Square Difference with 1 Degree of Freedom = 81.33 (P = 0.0)
Estimated Non-centrality Parameter (NCP) = 99.43
90 Percent Confidence Interval for NCP = (68.70 ; 137.67)
Minimum Fit Function Value = 0.66
Population Discrepancy Function Value (F0) = 0.50
90 Percent Confidence Interval for F0 = (0.35 ; 0.69)
Root Mean Square Error of Approximation (RMSEA) = 0.16
90 Percent Confidence Interval for RMSEA = (0.13 ; 0.19)
P-Value for Test of Close Fit (RMSEA < 0.05) = 0.00
Expected Cross-Validation Index (ECVI) = 0.92
90 Percent Confidence Interval for ECVI = (0.61 ; 0.95)
ECVI for Saturated Model = 0.36
ECVI for Independence Model = 4.76
Chi-Square for Independence Model with 28 Degrees of Freedom =
930.35
Independence AIC = 946.35
Model AIC = 183.93
Saturated AIC = 72.00
Independence CAIC = 980.73
Model CAIC = 252.70
Saturated CAIC = 226.74
Normed Fit Index (NFI) = 0.87
Non-Normed Fit Index (NNFI) = 0.85
Parsimony Normed Fit Index (PNFI) = 0.62
Comparative Fit Index (CFI) = 0.89
Incremental Fit Index (IFI) = 0.89
Relative Fit Index (RFI) = 0.82
Critical N (CN) = 63.59
Root Mean Square Residual (RMR) = 0.073
Standardized RMR = 0.086
Goodness of Fit Index (GFI) = 0.84
Adjusted Goodness of Fit Index (AGFI) = 0.71
Parsimony Goodness of Fit Index (PGFI) = 0.47
282
Degrees of Freedom = 28
Minimum Fit Function Chi-Square = 577.16 (P = 0.0)
Normal Theory Weighted Least Squares Chi-Square = 930.35 (P = 0.0)
Satorra-Bentler Scaled Chi-Square = 635.14 (P = 0.0)
Chi-Square Corrected for Non-Normality = 163.62 (P = 0.0)
Chi-Square Difference with 0 Degree of Freedom = 0.0 (P = 1.00)
Estimated Non-centrality Parameter (NCP) = 607.14
90 Percent Confidence Interval for NCP = (528.82 ; 692.88)
Minimum Fit Function Value = 2.90
Population Discrepancy Function Value (F0) = 3.05
90 Percent Confidence Interval for F0 = (2.66 ; 3.48)
Root Mean Square Error of Approximation (RMSEA) = 0.33
90 Percent Confidence Interval for RMSEA = (0.31 ; 0.35)
P-Value for Test of Close Fit (RMSEA < 0.05) = 0.00
Expected Cross-Validation Index (ECVI) = 4.76
90 Percent Confidence Interval for ECVI = (2.88 ; 3.70)
ECVI for Saturated Model = 0.36
ECVI for Independence Model = 4.76
Chi-Square for Independence Model with 28 Degrees of Freedom =
930.35
Independence AIC = 946.35
Model AIC = 946.35
Saturated AIC = 72.00
Independence CAIC = 980.73
Model CAIC = 980.73
Saturated CAIC = 226.74
Normed Fit Index (NFI) = 0.32
Non-Normed Fit Index (NNFI) = 0.33
Parsimony Normed Fit Index (PNFI) = 0.32
Comparative Fit Index (CFI) = 0.33
Incremental Fit Index (IFI) = 0.33
Relative Fit Index (RFI) = 0.32
Critical N (CN) = 16.13
Root Mean Square Residual (RMR) = 0.30
Standardized RMR = 0.36
Goodness of Fit Index (GFI) = 0.46
Adjusted Goodness of Fit Index (AGFI) = 0.31
Parsimony Goodness of Fit Index (PGFI) = 0.36
283
Degrees of Freedom = 17
Minimum Fit Function Chi-Square = 1.65243 (P = 1.00)
Normal Theory Weighted Least Squares Chi-Square = 1.64779 (P =
1.00)
Estimated Non-centrality Parameter (NCP) = 0.0
90 Percent Confidence Interval for NCP = (0.0 ; 0.0)
Minimum Fit Function Value = 0.0083037
Population Discrepancy Function Value (F0) = 0.0
90 Percent Confidence Interval for F0 = (0.0 ; 0.0)
Root Mean Square Error of Approximation (RMSEA) = 0.0
90 Percent Confidence Interval for RMSEA = (0.0 ; 0.0)
P-Value for Test of Close Fit (RMSEA < 0.05) = 1.00
Expected Cross-Validation Index (ECVI) = 0.20103
90 Percent Confidence Interval for ECVI = (0.20103 ; 0.20103)
ECVI for Saturated Model = 0.28866
ECVI for Independence Model = 0.79522
Chi-Square for Independence Model with 21 Degrees of Freedom =
140.27329
Independence AIC = 154.27329
Model AIC = 23.64779
Saturated AIC = 56.00000
Independence CAIC = 184.36152
Model CAIC = 70.92929
Saturated CAIC = 176.35289
Normed Fit Index (NFI) = 0.98822
Non-Normed Fit Index (NNFI) = 1.15895
Parsimony Normed Fit Index (PNFI) = 0.79999
Comparative Fit Index (CFI) = 1.00000
Incremental Fit Index (IFI) = 1.12450
Relative Fit Index (RFI) = 0.98545
Critical N (CN) = 4024.65017
Root Mean Square Residual (RMR) = 0.017812
Standardized RMR = 0.017799
Goodness of Fit Index (GFI) = 0.99764
Adjusted Goodness of Fit Index (AGFI) = 0.99611
Parsimony Goodness of Fit Index (PGFI) = 0.60571
284