Gerrefi - Governance Global Value Chains

The Governance of Global Value Chains
Author(s): Gary Gereffi, John Humphrey and Timothy Sturgeon

Source: Review of International Political Economy, Vol. 12, No. 1, Aspects of Globalization (Feb.,
2005), pp. 78-104
Published by: Taylor & Francis, Ltd.
Stable URL: http://www.jstor.org/stable/25124009
Accessed: 04-05-2015 18:29 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Review of International Political
Economy.
http://www.jstor.org
This content downloaded from 130.88.0.131 on Mon, 04 May 2015 18:29:33 UTC
All use subject to JSTOR Terms and Conditions
Review
Political
of International
The
12:1 February
Economy
2005:
of global
governance
78-104
value
?VTay|or&Franci$Group
chains
Gary Gereffi
Duke University
John Humphrey
Institute
Studies
of Development
and
Sturgeon
Timothy
Massachusetts
Institute
of Technology
ABSTRACT
This
in
value
global
costs
economics,
action
and
in
are:
a theoretical
builds
article
terns
framework
It draws
chains.
to help
explain
streams
of
on
three
determining
(1) the complexity
of
of global
value
chain
and market
which
(2) the ability
transactions,
in the supply-base.
-
and (3) the capabilities
governance
from
range
high
studies:
global
bicycles,
value
hierarchy,
to low
chain
apparel,
governance
and
horticulture
to
codify
transactions,
relational,
modular,
coordination
The theory generates
and power asymmetry. The article highlights

of
pat
trans
and
networks,
technological
capability
production
a
to
that play
role
three variables
identify
large
learning
are
and
These
value
chains
how
global
governed
change.
firm-level
nature
governance
literature
captive,
of
levels
explicit
the dynamic
four
through
electronics.
five types
and overlapping
brief
industry
case
KEYWORDS
Global
value
chains;
governance;
networks;
transaction
costs;
value
chain
modularity.
in significant ways during the past sev

The world
economy has changed
areas
in
of international
trade and industrial
the
eral decades,
especially
new
most
Two
of
the contemporary
features
the
of
important
organization.
are
which
have fueled
of
and
the
trade,1
economy
globalization
production
Review of International Political Economy
ISSN 0969-2290 print/ISSN 1466-4526 online ? 2005 Taylor & Francis Ltd
http://www.tandf.co.uk
DOI: 10.1080/09692290500049805
GEREFFI ET AL.: GLOBAL VALUE CHAINS

coun
in a wide
range of developing
which
of transnational
corporations,
to focus on innovation
and product
and
the
value-added
of manufac
segments
strategy, marketing,
highest
over
'non-core'
their
while
direct
and
services,
reducing
ownership
turing
these
functions such as generic services and volume production.
Together,
two shifts have laid the groundwork
for a variety of network
forms of gov
on the one hand, and large
ernance situated between arm's length markets,
on
the
other.
The
purpose of this article
vertically
integrated corporations,
is to generate a theoretical framework
for better understanding
the shifting
structures
in
sectors
structures
for
markets,
governance
producing
global
we refer to as 'global value chains'. Our intent is to bring some order to the
in the field.2
forms that have been observed
variety of network
the growth of industrial capabilities
tries, and the vertical disintegration
are redefining
their core competencies
of global-scale
affects not only
The evolution
industrial organization
the fortunes of firms and the structure of industries, but also how and why
- in the
countries advance - or fail to advance
global economy Global value
inwhich global
chain research and policy work examine the different ways
are
and
distribution
and
the
for
systems
integrated,
production
possibilities
to enhance their position
in global markets.
firms in developing
countries
One of our hopes is that the theory of global value chain governance
that we
will
for
to
here
useful
the
of
be
effective
tools
related
develop
crafting
policy
industrial upgrading,
economic development,
and
creation,
employment
poverty alleviation.
1. FRAGMENTATION,
COORDINATION,
AND NETWORKS
IN THE GLOBAL ECONOMY
For us, the starting point for understanding
the changing nature of inter
national
trade and industrial organization
is contained
in the notion of a
value-added
business
scholars who
chain, as developed
by international
have focused on the strategies of both firms and countries
in the global
In its most basic form, a value-added
chain is 'the process by
economy.
which
is combined with material
and labor inputs, and then
technology
are
and
A single firm
distributed.
assembled, marketed,
processed
inputs
one
or
consist
in
of
link
it
this
be
verti
may
process,
may
only '
extensively
are
The
in
1985:
issues
this
literature
15).
cally integrated...
(Kogut,
key
a firm
which activities and technologies
in-house
and
which
should
keeps
be outsourced
to other firms, and where
the various activities
should be
located.
are also concerned with how

Trade economists
global
Arndt
Kierzkowski
and
(2001) use the term
ganized.
describe
the physical
of different parts of
separation
that the international
dimension
of this
cess, arguing
in
allows
different
countries
Fragmentation
production
is or
production
to
'fragmentation'
a production
pro
is new.
separation
to be formed into
79
REVIEW OF INTERNATIONAL
POLITICAL
ECONOMY
or between
cross-border
networks
that can be within
firms.
production
Feenstra
(1998) takes this idea one step further by explicitly
connecting
the 'integration
of trade' with
the 'disintegration
in the
of production'
of world markets
global economy. The rising integration
through trade
has brought with it a disintegration
of multinational
firms, since companies
are finding it advantageous
to 'outsource' an increasing share of their non
core
and service activities both domestically
and abroad.
manufacturing
in
This has led to a growing proportion
of international
trade occurring
and other intermediate
components
goods (Yeats, 2001 ).3
across geographic
If production
is increasingly
space and
fragmented
are
between
then
these
activities
coordinated?
For
how
firms,
fragmented
Arndt and Kierzkowski,
the options are clear: 'Separability of ownership
is an important determinant
structure of cross-border
of the organizational
is not feasible, multi
production
separation of ownership
sharing. Where
are likely to play a
direct
national
and
investment
corporations
foreign
are possible
it is feasible, arm's-length
role. Where
dominant
relationships
is less important'
and foreign direct investment
(Arndt and Kierzkowski,
2001:4).
This binary view of how global production might be organized,
either
or
transac
is
markets
within
transnational
firms,
by
through
explained
tion costs economics
in terms of the complexity
of inter-firm relationships
and the extent to which
specific to a particular
they involve investments
re
market
transaction - asset specificity
1975). Arm's-length
(Williamson,
are
described
lations work well for standard products because
they
easily
their ease
and valued. Coordination
problems are reduced not only because
contracts simple to write, but also because
standard
of description makes
as
can
At
needed.
the same
for
stock
and
be
supplied
produced
products
are made by a variety of suppliers
and
standard products
time, because
problems
arising from asset specificity
bought by a variety of customers,
are
low.
reasons why
costs approach offers various
the transaction
Conversely,
more
customized
the
the
in-house. First,
firms will bring certain activities
or
more
in
to
it
is
involve
the
service,
transaction-specific
likely
product
out
out
either
rules
of
vestments.
the
risk
which
This raises
opportunism,
or makes
itmore costly because
safeguards have to be
sourcing altogether,
even
transaction costs increase
without
opportunism,
put in place. Second,
For example,
when
inter-firm relationships
require greater coordination.
involve
architectures
non-standard
inputs and integrated product design
in
intense
more
and
therefore
information
transfers of design
complex
are
across
architectures
teractions
Integral product
enterprise boundaries.
in the design of
more
inputs, and changes
likely to require non-standard
areas of the
in
other
to
changes
design
parts tend
precipitate
particular
coordina
and
1995). Similarly,
Robertson,
(Fine, 1998; Langlois
system
as
is
tion costs increase for parts whose
time-sensitive,
separate
supply
80

in order to synchronize
the flow of
processes have to be better coordinated
inputs through the chain.
the importance
of transaction
costs need not
Nevertheless,
recognizing
that complex and tightly coordinated
lead to the conclusion
production
systems always result in vertical integration. Rather, asset specificity, op
at the inter-firm level
costs can be managed
and coordination
portunism,
a
actors
in
methods.
Network
instances control
many
variety of
through
and
through the effects of repeat transactions,
opportunism
reputation,
or
so
in particular
social norms that are embedded
locations
geographic
cial groups. Network
theorists
1988; Powell,
(e.g., Jarillo, 1988; Lorenz,
and mutual
1986) argue that trust, reputation,
1990; Thorelli,
dependence
behavior, and in so doing they make possible more
dampen opportunistic
inter-firm divisions
of labor and interdependence
than would
be
complex
costs
transaction
theory.
predicted by
the literature on firm capabilities
and learning, which has
Furthermore,
its roots in the resource view of the firm pioneered
(1959), pro
by Penrose
to buy key inputs in the face
vides other reasons why firms are prepared
and therefore construct
of asset specificity
inter-firm
complex
relatively
to
can
how
and
whether
firms
Penrose,
capture
relationships.
According
in part on the generation
and retention of competencies
value depends
to replicate. In practice,
(that is, resources) that are difficult for competitors
even the most vertically
firms
internalize
all the techno
integrated
rarely
are
to
and
that
management
logical
capabilities
required
bring a product
or service to market.
Transaction
cost economics
this fact
acknowledges
an
even
an
the
If
of
variable
by employing
input,
frequency.
important
then itwill likely be acquired externally. This
one, is required infrequently,
an argument
is essentially
about scale economies.
The literature on firm
and
that
the
contrast,
argues
capabilities
learning, by
learning required to
to
in
the
certain
value
chain activities
engage
capability
effectively develop
and effectively
for some firms
may be difficult, time-consuming,
impossible
to acquire, regardless
of frequency or scale economies.
Thus, firms must
on external resources. The doctrine
in certain instances depend
of 'core
takes this a step further, arguing that firms which
competence'
rely on the
of other firms and focus more intensively
on
complementary
competencies
their own areas of competence will perform better than firms that are ver
diversified
(Prahalad and Hamel,
1990).
tically integrated or incoherently
These issues, while often discussed
at the local or national
level, or in the
context of 'adense network of social relations' (Granovetter,
1985:507), can
to
be
the
of
and distri
applied
equally
structuring
global-scale
production
bution. The recent work of geographers
such as Hughes
(2000), Henderson
et al. (2002) and Dicken
et al. (2001) has emphasized
the complexity
of
inter-firm relationships
in the global economy. The key insight is that co
ordination
and control of global-scale
their
production
systems, despite
can be achieved without
direct ownership.
complexity,
81
POLITICAL
ECONOMY
discussed
The theories of industrial organization
here, when considered
of
suggest that different ways of dealing with the problem
cumulatively,
firm
for
asset specificity, and different motivations
constructing
complex
in the face of asset specificity,
result in three modes
to-firm relationships
of industrial organization:
market, hierarchy, and network. But empirical
are alike. In the next section we
us
not
all networks
tells
that
observation
a
can
to
and
explain this variation.
help
specify
theory that
develop
2. TYPES OF GOVERNANCE
IN GLOBAL VALUE CHAINS
is to be useful to policymakers,
If a theory of global value chain governance
It has to simplify and abstract from an extremely
it should be parsimonious.
that play a large
the variables
body of evidence,
identifying
heterogeneous
while
chain
of
value
role in determining
governance
holding oth
patterns
and
ers at bay, at least initially. Clearly, history,
institutions,
geographic
mat
rules of the game, and path dependence
social contexts, the evolving
influence how firms and groups of firms are
factors will
ter; and many
a simple framework
is useful
linked in the global economy. Nevertheless,
a clear view of fundamen
and provides
it isolates key variables
because
be
situations
that might otherwise
tal forces underlying
specific empirical
that gener
Our intention is to create the simplest framework
overlooked.
ates
results
relevant
to real-world
outcomes.
a framework,
called 'global
In the 1990s Gereffi and others developed
chain
the
value-added
of
the
that
tied
chains',
directly
concept
commodity
and Korzeniewicz,
of industries
to the global organization
(see Gereffi
of coordination
the importance
1994). This work not only highlighted
of new global
across firm boundaries,
but also the growing
importance
as key drivers in the forma
brand
retailers
and
marketers)
buyers (mainly
and
and organizationally
tion of globally dispersed
fragmented production
networks. Gereffi
distribution
(1994) used the term 'buyer-driven
global
chain' to denote how global buyers used explicit coordination4
commodity
to help create a highly competent
pro
upon which global-scale
supply-base
direct ownership.
duction and distribution
systems could be built without
in dis-integrated
chains and con
explicit coordination
By highlighting
or
within
contained
vertically
integrated,
trasting them to the relationships
drew
chains
framework
'producer driven' chains, the global commodity
of cross-border
in driving the co-evolution
attention to the role of networks
chains framework
the global commodity
industrial organization.
However,
forms that more recent
did not adequately
specify the variety of network
on
indus
the horticulture
research
field research has uncovered. While,
2000) and the footwear
industry (Schmitz and
try (Dolan and Humphrey,
that
(retailers,
2000) reinforced Gereffi's notion
global buyers
Knorringa,
and traders) can and do exert a high degree of control over
marketers,
own production,
value chains even when
they do not
spatially dispersed
82

has
facilities, recent research on global production
transport or processing
of
coordination.
forms
other
highlighted
important
on the electronics
and contract manufacturing
Work
industry
by
Lee
three types
and
and
(2001) contrasted
(2002)
by Sturgeon
Sturgeon
of supply relationships,
of prod
based on the degree of standardization
uct and process:
that
standard
(1) the 'commodity
supplier'
provides
(2) the 'captive sup
relationships,
products
through arm's length market
to the
dedicated
products using machinery
plier' that makes non-standard
and
the
that
customized
(3)
needs,
'turn-key supplier'
buyer's
produces
to pool capacity
for buyers
and uses flexible machinery
for
products
customers.
This analysis
of infor
different
the complexity
emphasized
in
between
firms and the degree of asset specificity
mation
exchanged
(2002) referred to production
equipment.
Sturgeon
systems
production
that rely on turn-key suppliers as 'modular production
networks'
because
from the global
suppliers could be added and subtracted
highly competent
on as as-needed
basis. Around
the same time,
arrangements
production
in
and Schmitz
between
(2000, 2002) distinguished
suppliers
Humphrey
with
whose
situation
quasi-hierarchical
relationships
buyers,
corresponds
to 'captive suppliers',
and network
between
firms that coop
relationships
erate because
and
complementary
they possess
competences.5
Humphrey
in determining
Schmitz emphasized
the role of supplier competence
the
extent of subordination
to buyers.
of suppliers
If global buyers needed
to
invest in supplier competence,
would
to
need
both
the
they
specify
prod
uct and process parameters
to be followed by suppliers and to guard this
in the supplier by remaining
investment
the dominant,
if not exclusive,
customer.6
the approaches
outlined
above and empirical
reference points
Using
taken from many studies of global value chains,7 we propose amore com
as do most
We acknowledge,
governance.
plete typology of value-chain
- from trans
other frameworks
that seek to explain industry organization
- that
actions costs to global commodity
chains to organizational
theory
market-based
firms
among firms and vertically
relationships
integrated
(hierarchies) make up opposite ends of a spectrum of explicit coordination,
and that network
of value
comprise an intermediate mode
relationships
chain governance. What we add to this conceptualization
is an extension
of the network
into three distinct
relational, and
category
types: modular,
identifies five basic types of value chain gover
captive. Thus, our typology
nance. These are
types, although
analytical, not empirical,
they have been
in part derived
are:
from empirical observation.
They
1. Markets.
Market
linkages do not have to be completely
transitory, as is
over
can
of
with
markets;
time,
spot
typical
they
persist
repeat transac
tions. The essential point is that the costs of
new partners
to
switching
are low for both parties.
83
POLITICAL
ECONOMY
in modular
2. Modular
value
chains
value chains. Typically,
suppliers
more
or
to a customer's
be
which
make products
may
specifications,
services'
when
less detailed. However,
suppliers
'turn-key
providing
technol
take full responsibility
for competencies
process
surrounding
that limits transaction-specific
investments,
ogy, use generic machinery
on behalf of
and materials
and make
for components
capital outlays
customers.
3. Relational value chains. In these networks we see complex interactions be

and
tween buyers and sellers, which often creates mutual
dependence
through reputa
high levels of asset specificity. This may be managed
the role
authors have highlighted
tion, or family and ethnic ties. Many
in
value
chain
relational
of spatial proximity
linkages, but
supporting
networks
trust and reputation might well function in spatially dispersed
on dispersed
or
over
are
are
time
based
where
built-up
relationships
1992).
family and social groups (see for example, Menkhoff,
small suppliers are transaction
4. Captive value chains. In these networks,
on much
face significant switch
larger buyers. Suppliers
ally dependent
are frequently char
networks
Such
costs
and
are, therefore, 'captive'.
ing
and control by lead firms.
acterized by a high degree of monitoring
form is characterized
5. Hierarchy. This governance
by vertical integration.
ismanagerial
form of governance
The dominant
control, flowing from
or
to subsidiaries
and
to
from
subordinates,
managers
headquarters
affiliates.
3. A THEORY OF VALUE CHAIN GOVERNANCE

the
laid out this typology, our next step is to develop an operational
Having
we
would
which
conditions
Under
chain
value
of
governance.
ory
global
or
relational,
vertically
integrated global
captive,
expect market, modular,
on the work
cited above, we
to arise? Building
value chain governance
of value chain governance
will identify and discuss three key determinants
and ca
of information;
of transactions;
codifiability
patterns: complexity
so doing, we acknowledge
of asset
the problem
In
of
suppliers.
pability
but also give em
cost economics,
as identified by transaction
specificity
costs - the costs
transaction
'mundane'
termed
been
to
what
have
phasis
in coordinating
activities along the chain. It has been argued that
involved
costs rise when value chains
or mundane,
transaction
these coordination,
ar
are producing
non-standard
integral product
products with
products,
and
time
is
sensitive
and products whose
(Baldwin
chitectures,
output
Clark, 2000).
on the
new demands
Lead firms increase complexity when
they place
when
and
seek
value chain, such as when
they
supply
just-in-time
they
lead firms also adopt strate
increase product differentiation.
However,
of these transactions. One important way of
gies to reduce the complexity
84

of technical and process standards.
doing this is through the development
transmitted between firms can be reduced
The complexity
of information
that codify information
and
of technical standards
through the adoption
in
activ
the
Where
flow
of
allow clean hand-offs between
trading partners.
ities these standards apply goes a long way toward determining
the orga
standards for the hand-off
nizational break points in the value chain. When
are widely
the value chain gains many of
of codified specifications
known,
in the realm of modular
the advantages
that have been identified
product
the conservation
of human effort through the re-use of
design, especially
system
elements
or modules
as
new
products
are
brought
on-stream
and Steensma,
1995; Schilling
2001; Sturgeon,
(Langlois and Robertson,
can
customers
and
2002). In the realm of value chain modularity,
suppliers
a
in
be easily linked and de-linked,
fluid
and
flexible
network
very
resulting
are market-like,
structure. While
the system remains qualita
the dynamics
different
of
the
volumes
of
because
information flow
non-price
large
tively
across
in
the
inter-firm
albeit
codified
form.
Furthermore,
ing
boundary,
a high-level
can be accommodated
of product differentiation
with limited
as long as differentiation
is defined by a set of un
information
exchange
and widely
Institutions, both public and
ambiguous
accepted parameters.
private, can both define grades and standards and (in some cases) certify
that products
of process standards
comply with them.8 The development
in relation to quality, labor and environmental
and certification
outcomes
perform similar functions.9
At the same time, the integration
of new suppliers
into global value
chains also increases coordination
and Lall (1992) ar
Keesing
challenges.
are
in
that
to meet require
countries
gue
producers
expected
developing
ments
that frequently do not (yet) apply to their domestic markets.
This
creates a gap between
the capabilities
required for the domestic market and
those required for the export market, which raises the degree of monitoring
and control required by buyers.
These considerations
lead us to construct a theory of value chain gover
nance based on three factors:
A. The complexity of information
and knowledge
transfer required to sus
tain a particular
transaction, particularly with respect to product and
process specifications;
B. the extent to which
can be
this information
and knowledge
codified
and
without
and, therefore, transmitted
efficiently
transaction-specific
investment between
the parties to the transaction; and
C. the capabilities of actual and potential
in relation to the re
suppliers
of
the
transaction.
quirements
If these three factors
there are eight possible
then
only two values
high or low
of which
five are actually found.10
combinations,
are allowed
85
POLITICAL
ECONOMY
are easily codified, product

transactions
specifications
to make
the
have the capability
simple, and suppliers
asset
in
with
little
from
input
buyers,
products
question
specificity will
can be expected.
In market
and market
fail to accumulate
governance
set
to
and
sellers.
Be
prices
by
specifications
exchange buyers respond
cause the complexity
trans
is
of information
low,
relatively
exchanged
actions can be governed with little explicit coordination.
1. Markets. When
are relatively
2. Modular
value chains. When
the ability to codify specifications
extends
can arise. This can come
to complex products,
value chain modularity
and technical standards
is modular11
about when product architecture
and by unifying
variation
interactions
by reducing component
simplify
also
when
and
and
process specifications,
suppli
component,
product,
ers have the competence
to supply full packages
and modules,
which
reduces asset specificity
hard to codify (tacit) information,
internalizes
and control. Link
and therefore a buyer's need for direct monitoring
on
the
benefits of arms
of
codified
based
ages
knowledge
provide many
access
to
low-cost inputs
and
market
speed, flexibility,
linkages
length
but are not the same as classic market exchanges based on price. When
a
from a lead firm to a supplier,
design file is transferred
computerized
across the inter-firm link than
there ismuch more flowing
for example,
information
about prices. Because of codification,
information
complex
can be exchanged with
and so, like simple
little explicit coordination,
to new partners remains low.
the cost of switching
market exchange,
cannot be codified,
3. Relational value chains. When product
specifications
are
are complex,
and supplier capabilities
transactions
high, relational
can be expected. This is because tacit knowledge
value chain governance
must be exchanged between buyers and sellers, and because highly com
for lead firms to outsource
petent suppliers provide a strong motivation
The mutual dependence
to gain access to complementary
competencies.
social and spatial
that then arises may be regulated
through reputation,
can
It
like.
also be handled
and
the
ethnic
ties,
family and
proximity,
on
that
breaks a con
costs
the
that
party
impose
through mechanisms
commitments
of
credible
as
in
Williamson's
discussed
tract,
analysis
tacit infor
of complex
and hostages
1983). The exchange
(Williamson,
interaction
face-to-face
most
often
is
mation
by frequent
accomplished
the
which makes
and governed
by high levels of explicit coordination,
new
to
costs of switching
partners high.
- in the form of detailed
the ability to codify
4. Captive value chains. When
are both high
of product specifications
and the complexity
instructions
are
chain
value
then
but supplier capabilities
low,
governance will tend
in
low
is
because
This
the
toward
competence
supplier
captive type.
a
deal
and specifications
the face of complex products
great
requires
and control on the part of the lead firm, encouraging
of intervention
as lead firms seek to lock-in
of transactional
the build-up
dependence
86

in order to exclude others from reaping the benefits of their
suppliers
costs and are
the suppliers
face significant
efforts. Therefore,
switching
are
a
narrow
to
confined
range
frequently
'captive'. Captive
suppliers
- and are
in simple assembly
of tasks - for example, mainly
engaged
on the lead firm for complementary
activities such as design,
dependent
and
process
logistics, component
technology
purchasing,
upgrading.
inter-firm linkages control opportunism
Captive
through the dominance
resources and
of lead firms, while
at the same time providing
enough
access to the subordinate
market
firms to make
exit an unattractive
option.
5. Hierarchy. When product specifications

cannot be codified, products are
cannot be found, then lead
and
complex,
highly competent
suppliers
firms will be forced to develop and manufacture
in-house. This
products
form is usually driven by the need to exchange
tacit knowl
governance
as
as
to
between
value
activities
chain
well
the
need
edge
effectively
manage
complex webs of inputs and outputs and to control resources,
especially
intellectual
property.
The five global value chain governance

types, along with the values of the
three variables
that determine
them, are listed in Table 1. These five types of
arise from ascribing different values to the
global value chain governance
three key variables:
of inter-firm transactions;
(1) complexity
(2) the degree
can be mitigated
to which
this complexity
and (3)
codification;
through
the extent to which
have
the
to
meet
the
necessary
suppliers
capabilities
a
Each
different
trade-off
governance
buyers' requirements.
type provides
between
the benefits and risks of outsourcing.
As shown in the last column
a spectrum
of Table 1, the governance
types comprise
running from low
levels of explicit coordination
and power asymmetry
between buyers and
Table
Governance
determinants
Complexity
transactions
type
Market
Low
Modular
Relational
Captive
Hierarchy
There
1 Key
High
High
High
High
of
of global
Ability
to codify
transactions
High
High
Low
value
chain
Capabilities
in the
Degree of explicit
coordination and
supply-base
power
High
High
Low
combinations
of the three variables.
eight possible
chain types. The combination
of low complexity
of
to occur. This excludes
is unlikely
two combinations.
codify
transaction
is low and the
to codify
is high,
then low
ability
to exclusion
from the value chain. While
this is an important
governance
type per se.
value
asymmetry
Low
High
High
are
governance
Low
Low
High
Five of them generate

global
transactions
and low ability
to
if
the
of the
Further,
complexity
would
lead
supplier
capability
a
it does not generate
outcome,
87
POLITICAL
ECONOMY
to high levels of explicit coordination

in the case of markets,
and
suppliers,
in the case of hierarchy.
between buyers and suppliers,
power asymmetry
The fact that the governance
here can be used to il
types developed
in
luminate how power operates
elaboration.
global value chains merits
In captive global value chains, power
is exerted directly by lead firms
on suppliers, which
to the direct administrative
is analogous
control that
at
exert over subordinates
in an off
top management
headquarters
might
or affiliate of a vertically
shore subsidiary
firm
(or 'hierarchy'
integrated
in our framework).
Such direct control suggests a high degree of explicit
coordination
and a large measure
of power asymmetry with the lead firm
the
dominant
(or top management)
party. In relational global value
being
is more symmetrical,
the
balance
firms
between
the
chains,
power
given
There is a great deal of explicit co
that both contribute key competences.
a
in relational global value chains, but it is achieved
ordination
through
to the more
close dialogue between more or less equal partners, as opposed
unidirectional
flow of information
and control between unequal partners
as in captive global value chains and within hierarchies.
Inmodular
global
customers
is relatively
value chains, as inmarkets,
and
suppliers
switching
remain relatively
low because both suppliers and
easy. Power asymmetries
with
work
partners.
multiple
buyers
in graphic
Figure 1 (next page) illustrates much of the above discussion
the five global value chain types arrayed along the dual
form, showing
The small line
and power asymmetry.
of explicit coordination
spectrums
arrows represent exchange based on price while
the larger block arrows
and control, regulated
thicker flows of information
through ex
represent
a
more
instructions
from
coordination.
This
includes
power
coming
plicit
as in
to a less powerful
ful buyer (or manager)
(or
subordinate),
supplier
as
a
as
or
well
within
the
confines
of
value
chains
hierarchy,
captive global
the behavior of more or less equal partners, as in
social sanctions regulating
relational global value chains. In the case of modular
global value chains,
down to a codified hand off at the
thick information
flows are narrowed
its
tacit information within
inter-firm link, leaving each partner tomanage
some other form of global
or perhaps by combining
own firm boundaries,
for part of the
such as captive or market-based,
value chain governance,
the relational and modular
between
chain. While
suppliers
relationships
are dis
their material
and the firms providing
inputs and components
as market-based
other
forms.
in
take
the
could
equally
figure, they
played
4. DYNAMIC VALUE CHAIN ANALYSIS:

SECTORAL CASES
and provid
the main
types of global value chain governance,
Identifying
are
a
for
theoretical
steps and
important
they arise,
why
ing
explanation
88
ket
Modular
High
Firm
Integrated
Suppliers
Captive
Suppliers
Component
andMaterial
Captive
Hierarchy
Relational
Coordination
Explicit
Degree
of
Degree
of
Power
Asymmetry
Suppliers
andMaterial
Component
Suppliers
Customers
Low
<*
Figure
Five
global
governance
chain
value
Materials
End
Use
>
oo
Table
Governance
2 Some
dynamics
Complexity
value
of global
of
Ability
transactions
type
POLITICAL
to codify
Capabilities
the
Low
High
Modular
Relational
High?]
Hieh
High
Captive
Hierarchy
HighHigh
High
in governance:
of changes
Dynamics
?
of transactions
Increasing
complexity
lation to new demands.
?
of transactions
Decreasing
complexity
? Better codification
of transactions.
? De-codification
of transactions.
Increasing
Decreasing
governance
transactions
Market
?
?
chain
ECONOMY
supplier
supplier
in
supply-base
High
^tHi&h?l
?l
Low
iir
?lLow
+HiSh
High?
Low
>
Low Low
also
reduces
and greater
supplier
ease
competence
in re
of codification.
competence.
competence.
of the contem
this work will lead us to a better understanding
hopefully
it a useful tool for policy, a
to make
economy. Nonetheless,
porary world
should allow us to do more than
theory of global value chain governance
we must
try to an
just generate different forms of inter-firm coordination;
in particular,
clearly
ticipate change in global value chains. Case studies,
structures
evolve over time. In the following
show us how governance
structures have
how global value chain governance
section, we highlight
in four distinct industries: bicycles, apparel, fresh vegetables,
and
evolved
Some trajectories of change are identified on Table 2, and we
electronics.
refer to these trajectories as we discuss each of the cases.
4.1. The bicycle
industry:
From hierarchy
to market-based
coordination
in the twentieth
of the bicycle
The evolution
century provides
industry
can evolve
a good example of how hierarchies
toward inter-firm gover
on market mechanisms.12
It shows how mar
nance that relies primarily
is enabled not only by low transaction costs
ket governance
particularly
costs associated with coordination
of component
design with final product
of scale and production
enabled by the rise of
and the economies
design
but
also
the
of
standards,
competencies
development
specialist
by
industry
in
5
Table
3
and
numbers
2).
among suppliers
(trajectories
In the early years of the bicycle industry (the 1890s), vertically
integrated
soon became fragmented. To
but
firms manufactured
bicycles,
production
each segment of the value chain, such as
day, there are large firms within
in drive-train
and several large branded bicycle man
Shimano
components
than one segment
but very few firms that span more
(Galvin
ufacturers,
2001: 40). The different bicycle components
and Morkel,
require different
90

limits economies
of scope. An integrated
which
bicycle
competencies,
or
different
manufacturer
would
competences,
technological
require many
firms.
coordinate
the activities of many different
would
need to explicitly
After the initial stage of the industry's development,
specialist firms be
came more competitive
that made
than vertically
integrated
companies
between
Well-defined
various
interfaces
components
bicycles.
complete
mean
of scale through
have the advantages
that specialist manufacturers
in
of scale occur upstream
demand pooling. To the extent that economies
for market
and
the value chain, there are strong incentives
coordination
to make
this possible.
the development
of the institutional mechanisms
of the suppliers also gives them a greater capac
The specialist knowledge
their specific product
ranges, as long as this does
ity to innovate within
not require changes
in other components.
Where
these specialists
domi
in
nate a market
drive
Shimano
(for example,
segment
they can
systems),
innovate within
than others, and if extremely
this area more successfully
a new de facto standard applicable
across the
successful, may establish
industry.
such specialization
The industry standards
and divi
required to make
sions of labor work can arise in a variety of ways. They can be imposed by
a dominant
in bicycles and IBM in personal
firm, as in the case of Shimano
can
as with
arise
computers;
they
through inter-firm networks,
informally
the emergence
of regional standards
in the early days of the bicycle indus
can
or they can be regulated
be
try; they
managed
by industry associations;
as
case of the develop
in
and
international
the
by
negotiations,
agencies
ment of new standards for mobile phones. The establishment
of standards
is often contentious
and part of the competitive
firms.
of
positioning
4.2. The
apparel
industry:
From
captive
to relational
value
chains
The apparel
and
industry has been characterized
by global production
trade networks
since at least the middle
of the twentieth
and
the
century,
and
its
of
have
expansion
capabilities
growing
global supply-base
per
it to move
mitted
relational value
rapidly from captive to more complex
chains over the span of just a few decades. The epicenter of export-oriented
has been East Asia, as Japan in the 1950s and 1960s,
apparel production
the 1970s and 1980s, and
Hong Kong, South Korea, and Taiwan during
as
in the 1990s emerged
China
textile and ap
world-class
sequentially
success was
(Bonacich et al., 1994). The key to East Asia's
parel exporters
to move
from captive value chains - i.e., the mere
of
imported
assembly
zones - to a more
in export-processing
in
inputs, typically
domestically
form of exporting broadly known
in the
tegrated and higher-value-added
the assembly-oriented
industry as full-package
supply.13 Whereas
captive
model
in
coordination
the
form
of cut fabric and detailed
required explicit
full package production
involved
the more
forms
instructions,
complex
91
of coordination,
knowledge
relational
chains.
value
exchange,
POLITICAL
and supplier
ECONOMY
autonomy
typical
of
in which
for
captive networks,
foreign firms take responsibility
full
local
the
used
all
contractors,
component
parts
by
package
supplying
to inter
contractors
the capability
requires offshore
develop
production
source the needed
pret designs, make
inputs, monitor
product
samples,
on-time delivery. From a
the buyer's price, and guarantee
quality, meet
of the full package export
the main advantage
development
perspective,
to simple assembly,
is that it allows local firms to learn how
role, compared
consumer
sub
to make
competitive
goods and generates
internationally
to
the
domestic
stantial backward
economy.
Increasing supplier
linkages
has been the main driver behind the shift from captive to rela
competence
tional value chains in the apparel industry (trajectory number 5 in Table 2).
international
of overseas
The establishment
offices and frequent
buying
Unlike
tacit infor
the intense interaction required for exchanging
travel supported
and
between
and building personal
mation
buyers
suppliers.
relationships
Trade rules have had an important
impact on global value chain gover
nance in the apparel industry, and this provides
just one example of how
to shape the archi
variables other than the three we have identified work
US
tecture of cross-border
economic
by
import quotas established
activity.
fueled
the
theMulti-Fiber
(MFA)
spread of global production
Arrangement
in the early 1970s.The existence of quotas
in apparel beginning
networks
rise
of
value-chain
the
intermediaries,
including East Asian trad
prompted
as
such
Li
&
such
Fung and manufacturers
Hong Kong's
ing companies
as the Fang Brothers, to coordinate
the flow of orders from US and Euro
around the
pean buyers to a large numbers of apparel factories established
in places with available quota (Gereffi, 1999: 60-63; Magretta,
world
1998).
the MFA ismostly phased out in 2005 in accordance with theWorld
When
on Textiles and Clothing,
Trade Organization's
global apparel
Agreement
more concentrated
to
far
become
is
among the most capa
likely
production
ble firms in a handful of low-cost production
sites, including China, India,
2003: 12). Such
and Turkey (Gereffi and Memodovic,
Indonesia, Mexico,
firms.
of intermediary
the position
could well undermine
concentration
im
to
be
we
in
continue
this paper
have highlighted
Still, the variables
is
to
increased
by
codify transactions
portant. To the extent that the ability
to improve,
continue
this concentration
process, and supplier capabilities
in apparel to become more
we would
expect the relational value chains
modular
(trajectory number 3 in Table 2).
From market
4.3. Fresh vegetables:
to explicit coordination
coordination
and the
trade between
nature of fresh vegetables
The changing
Kenya
a shift from market-based
value chain
United Kingdom
global
highlights
92

to more explicit coordination,
and it reveals the importance
governance
in driving
this change.14
of
UK
of the competitive
supermarkets
strategies
use
to
the
in
mid-1980s
UK
the
quality and
supermarkets
began
Beginning
a main source of competitive
as
differ
of
their
variety
offerings
produce
several distinct forms of governance
entiation, and in doing so generated
at different stages in the chain.
trade was handled
the fresh vegetables
the mid-1980s,
Until
through
a series of arm's-length
in Kenya bought
Traders
market
relationships.
or at the farm gate and exported
it to the
in wholesale
markets
produce
as
it was sold in wholesale
markets.
where
United Kingdom,
However,
an
in
took
the
United
chains
Kingdom
increasing
gradually
supermarket
share of fresh food sales and therefore became more powerful
actors, they
more explicit coordination
in the chain. Supermarket
began to introduce
as strategic because
saw fresh produce
it was one
(fruit and vegetables)
consumers
to shift from one
lines that could persuade
of the few product
In order to attract customers,
the super
chain to another.
supermarket
new items, emphasized
consistent
markets
introduced
quality, provided
to provide
of products
the processing
supply, and increased
year-round
that required
little or no preparation
fresh produce
prior to cooking or
were
same
to respond to an
At
forced
the
the
time,
eating.
supermarkets
to
environment
related
food safety, par
regulatory
complex
increasingly
as
and
residues
conditions
for
post-harvest
processing,
ticularly pesticide
well as environmental
and labor standards.
these strategic goals by increasing explicit coor
Supermarkets
pursued
mar
in the value chain. Instead of purchasing
dination
through wholesale
closer relationships
with UK importers and African
they developed
to
contracts with suppliers whose
moved
and
renewable
annual
exporters,
were
to
and
and audit. Su
systems
subject
capabilities
regular monitoring
to
to
in the chain,
incorporation
began
inspect suppliers prior
permarkets
to the
at
all
in
and made
checks
the
down
chain, right
points
regular spot
field. The interaction of the firms in the chain also became more complex
and relational. Suppliers and buyers worked
together on product develop
ment, logistics, quality, and the like. This created new value chain relation
kets,
Over time, relationships

between
ships and competencies.
supermarkets
and UK importers
took new forms, with
the recent trend moving
value
in the direction
chain governance
of modularity.
The supermarkets
have
reduced
the number of UK suppliers/importers
for each product
range
and given the remaining
for supply chain
suppliers greater responsibility
and consumer
research. These im
management,
product development,
and food retailers, although
porters work for a range of UK supermarkets
the three largest supermarket
chains (Tesco, Asda, and Sainsbury)
do try
to avoid using the same suppliers.
Further back along
the chain, organizational
has de
fragmentation
creased and inter-organizational
have
become
The
relational.
relationships
93
POLITICAL
ECONOMY
risks of this have been contained by the development

of exclusive bilateral
one UK importer,
A
will
deal
with
exporter
relationships.
Kenyan
only
it
to
sell
other
markets
other
and a UK
channels,
may
through
although
some
even
There
has
been
importer will only have one Kenyan
supplier.
forward and backward
integration between African exporters and UK im
or equity participation.
This bilateral de
porters, with outright ownership
of African
exporters and UK importers has not created captive
pendence
First, importers and exporters do change partners from time
relationships.
to time. Second,
of mutual
there is a situation
and power
dependence
an
to
the UK market, but importers also
outlet
symmetry. Exporters need
in
need an assured supply of produce. Third, the exporters have become
as
and
additional
functions
competent,
processing
creasingly sophisticated
were
to Africa where
costs are lower (trajectory number 5 in
transferred
as
Table 2). In Kenya,
the industry has become much more concentrated
costs of processing
the investment
have risen.
to
the
from Africa
Within
Kenya,
largest exporter of fresh vegetables
led
the United Kingdom,
have
leading exporters
increasing requirements
at the expense of purchasing
to increase own-farm production
vegetables
from both smallholders
and large contract farmers. This can be seen as a
case of increasing complexity
it is not
leading to vertical integration when
or
either
codification
by
higher supplier competence.
accompanied
4.4. The US
industry: From hierarchy

chains, and beyond
electronics
to modular
value
For most of the twentieth

States has been dominated
in the United
century, the electronics
industry
first in the
firms,
by large, vertically
integrated
out
the
radio
of which
and
then
(ATT)
(RCA),
industry
telephone
industry
as
consumer
sectors
and
television
such
electronics
other
grew
eventually,
computers
(e.g., IBM). In the 1960s and 1970s, with the push for better semi
or
an independent,
and aerospace
for military
conductors
applications,
steam
Texas
Instruments)
'merchant', components
gathered
industry (e.g.,
Aeronautics
and Space Administra
with
the Air Force and the National
the role of 'lead firm'. In the 1980s, as the civilian electronics
tion playing
a range of
the personal
computer,
industry began to grow rapidly with
with
other value chain functions were outsourced,
beginning
production
and circuit board assem
fabrication
for both semiconductor
equipment
such as disk drives
to specialized
sub-components
bly, and then spreading
itself in a
and most recently to the manufacturing
and monitors,
process
practice
called
'contract
manufacturing'.15
elec
the 1990s nearly all major North American
product-level
During
as well, made the
tronics firms, and several important European companies
to get out of manufacturing.
Plants were closed or sold off to con
decision
electronics
share of the world's
tract manufacturers,
driving a significant
94

production
manufacturers.
capacity
The
into a handful
contract
of huge
manufacturer
globally
Solectron,
operating
for
contract
example,
grew
and $256 mil

from a single Silicon Valley location with 3,500 employees
in 1988 to a global powerhouse
with more
than 80,000
lion in revenues
in revenues
in 2000.
in 50 locations and close to $20 billion
employees
its service offerings beyond
the same period Solectron
extended
During
to include, among other things, product
circuit-board
(re)design
assembly
for-manuf acturability, component purchasing
and inventory management,
test routine development,
final product assembly, global logistics, distri
service and repair. Global contract manufacturers
bution, and after-sales
such as Solectron
introduce a high degree of modularity
into value chain
create
the large scale and scope of their operations
or
of
value
that
chain
activities
bundles,
modules,
comprehensive
generic
can be accessed by a wide variety of lead firms. Standardized
protocols
for handing-off
and stan
computerized
design files and highly automated
it
to
firms
dardized
for
lead
and
made
switch
process
easy
technologies
and inhibited the build-up
of specific assets.
share contractors,
seek new sources of revenue by providing
addi
Today, as contractors
tional inputs to lead firm design and business processes,
and new circuit
board assembly
appear on the scene, such as those for boards
technologies
governance
because
with
is becom
the hand-off of design
optical components,
specifications
more
it
less
and
harder
for
lead
firms
standardized,
ing
complex
making
to switch and share suppliers. Closer collaboration
in the realm of product
to receive fully blown computer-aided-design
design requires contractors
new products;
files for their customer's
files that can contain core intellec
tual property As contractors
take over more distribution
lead
functions,
firms must
reveal critical knowledge
about end-customer
requirements
are being embedded
and pricing. All of these interactions
in elaborate
in
formation technology
that
the
of
lead
and
firms
span
systems
organizations
their key contractors,
creating new areas of risk for lead firms in the areas
of intellectual property
lock-in. Shared infor
leakage and buyer-supplier
mation
in two directions
systems are evolving
technology
simultaneously:
toward proprietary
systems that increase asset specificity and lock-in, but
better protect key intellectual property;
and toward open standards
(e.g.,
that better support value chain
RosettaNet)
and/or
systems
third-party
but that leave the door open for intellectual property
modularity
leakage.
The question of which direction
the industry will take - toward proprietary
used standards
systems and relational value chains, or toward commonly
and modular
value chains - is still open, and its answer will help to deter
mine the future shape of the electronics
industry.
case shows value chain
The electronics
is enabled by the
modularity
codification
of complex
information
(for example,
through computerized
process
product design and automated
at the inter-firm
the hand-off
simplifies
because codification
technologies)
link. But the case also shows that
95
POLITICAL
can be undermined
by 'de-codification'
modularity
Table 2), spurred either by technological
change, as
circuit
of
board
gence
assembly
technology,
optical
a
supplier activities in such way that suppliers reach
to assist with lead firm activities
that remain tacit or
or both,
such as product
design
4.5. The dynamics
and customer
of global
ECONOMY
(trajectory
in the case
or by the
across the
are highly
number 4 in
of the emer
of
bundling
codified link
proprietary,
contact.
value
chains
to highlight
the dy
in this section are meant
The case studies presented
nature of global value chains. Value chain gover
namic and overlapping
nance patterns are not static or strictly associated with particular
industries.
on
actors
between
value
chain
of
how
details
interactions
the
They depend
are applied to design, production
are managed,
and
and how technologies
of the value chain itself. Nor are value chain governance
the governance
Even in a particular
industry in a particular place and
patterns monolithic.
one stage of the chain to an
from
time, governance
may
vary
patterns
can largely be
and variation
that this dynamism
other. While we believe
in this article,
variables presented
for by the three explanatory
accounted
characteris
more work will be needed
their dynamic
to fully understand
the ability to codify
of information,
tics. How and why do the complexity
and supplier competence
information,
change?
com
We can at this stage offer only a partial answer. First, information
as lead firms seek to obtain more complex outputs and ser
plexity changes
This can reduce the effective level of supplier
vices from their supply-base.
as
the new requirements
existing
capabilities may not meet
capabilities
reduced
1
in
Table
number
2). Alternatively,
complexity may
(trajectory
increase the ability to codify transactions
(trajectory 2 in Table 2). Second,
and
codification
tension between
there is a continuing
industries
within
and
As
4
in
Table
3
and
numbers
innovation
(1995)
2).
Storper
(trajectories
can restart the clock
David
(1995) have both pointed out, new technologies
on the process of codification.
Third, supplier competence
changes over
introduce
time: increasing as suppliers
learn, but falling again as buyers
or
come on-stream,
new suppliers
into value chain, as new technologies
for existing suppliers
as lead firms increase the requirements
(trajectories
5 and 6 in Table 2).
numbers
When we look broadly at the evidence provided
by global value chain
to
it is tempting
research across a variety of industries and time periods,
In all of the case
about trends in the global economy.
make generalizations
studies presented
here, and many other industries as well,
increasing ca
of global
to
the
architecture
have
in
the
push
helped
supply-base
pabilities
and toward the
and captive networks
value chains away from hierarchy
seems to
and market
relational, modular,
types. Value chain modularity
levels of value
firms
lead
offer
when
be especially
greater
likely
suppliers
96

chain bundling
services), which has the ad
(e.g., turn-key and full-package
tacit
and
of
vantages
knowledge
pooling capacity utilization
internalizing
will
for greater economies
of scale. However,
organizational
fragmentation
not lead to value chain modularity
is extremely difficult. For
if codification
in the organization
of the US
example, a strong shift toward fragmentation
motor vehicle
in the mid-1980s
in value
has resulted
industry beginning
chains with strong relational elements. This can be partly explained by the
systems (Fine, 1998), which has
difficulty of codifying
complex mechanical
inhibited the rise of industry-wide
standards and kept the complexity
of
the transactions between
lead firms and suppliers high even as the capabil
ities of suppliers, driven in part by the consolidation
of first tier suppliers,
has increased dramatically
(Humphrey, 2003; Sturgeon and Florida, 2004).
As standards,
information
and the capabilities
of suppliers
technology,
an
to
the
modular
form
be
central
appears
improve,
playing
increasingly
role in the global economy.16 Again,
the general shift toward value chain
cost and risk advantages
has
been
the
driven
of out
by
fragmentation
a
can
to
that
asset
solution
the
of
sourcing
problem
(assuming
specificity
as our starting point,
be developed).
When we take relational networks
- and
a shift to modular
to market - forms
however,
perhaps eventually
can be expected
as standards and codification
schemes
improve because
more fluid value chains offer additional
in cost and risk. Still,
decreases
we resist the
overly simplistic notion that global value chains are evolving
a
that enable the codification
along
single trajectory. First, the standards
are different across industries
of product and process
and
specifications
are constantly
standards
for
and
Second,
pro
evolving.
codifying product
cess specifications
can become obsolete
as technologies
change or when
there is a drive to bundle value chain activities
in new ways.
This can
as
we
drive market and modular
in
be
the case
may
relationships,
seeing
of the electronics
back
toward
relational
and,
governance,
industry today,
severe enough,
if the problem of asset specificity becomes
the hierarchi
cal form. Third, knowing
the standard and adopting
the protocol may
not be straightforward,
or
for all actors
inexpensive,
possible
immediately
in an industry, and there may be competing
in use that make
standards
choosing and investing difficult and risky. Since standards and protocols
are dynamic, major
accrue to those actors that
advantages
actively par
in
the
favors established
actors and
process, which
ticipate
rule-setting
locations
to
(Sturgeon, 2003). Finally, there is clearly no single best way
organize global value chains. In some product
categories, where
integral
it difficult
architecture makes
to break the value chain, vertical
product
most
be
the
to value chain gov
integration may
competitive
approach
ernance.
despite,
garment
bi-weekly
Sony
and
Samsung's
success
in
consumer
electronics
has
or perhaps
come
because of, high levels of vertical

In the
integration.
success with
Zara's
industry,
extremely
rapid product
cycles
in some cases - has been
the
in-house
supported by
company's
97
textile manufacturing
subsidiary
POLITICAL
and captive
ECONOMY
sewing workshops
(Bonnen,
2002).
5. CONCLUSIONS
a
In this article we have developed
typology of global value chain gover
nance and
some
for why these patterns
theoretical
presented
justifications
occur.
structure
We
the
that
of
argue
might
global value chains depends
three
variables:
the
of transactions,
the ability to
critically upon
complexity
and
in
the
the
These
variables
transactions,
codify
capabilities
supply-base.
are sometimes
determined
the
characteristics
of
by
technological
products
are inherently more
and processes
and dif
(some transactions
complex
on the
ficult to codify than others, for example)
and they often depend
effectiveness
of industry actors and the social processes
the
surrounding
and
of
standards
and
codifi
other
dissemination,
adoption
development,
cation schemes.
in particular,
It is the latter set of determinants,
that opens
the door for policy interventions
and corporate strategy
The global value chains framework
focuses on the nature and content of
the inter-firm linkages, and the power that regulates value chain coordina
tion, mainly between buyers and the first few tiers of suppliers. However,
it is important not to ignore the actors at both ends of the value chain.
can wield
a
On the upstream
and equipment
end, component
suppliers
in the personal
great deal of power. For example,
computer
industry two
set parameters
that most other value chain ac
firms, Intel and Microsoft,
tors must to adjust to. The power of such 'parameter-setting'
firms, such as
in semiconductors,
in bicycles and Applied Materials
is not ex
Shimano
but through their market dominance
erted through explicit coordination,
in key components
On the downstream
and technologies.
end of the chain,
users
can
a
in
role
the
significant
play
determining
highly knowledgeable
services
of
the
and
that
attributes and innovative
products
global
trajectory
value chains churn out, as they do inmany complex service industries such
as
enterprise
computing.
Even
consumers
average
are
far
from
passive,
as
it emerges
Leslie and Reimer (1999) point out. Consumer
culture, whether
intention
from the home, street, school, or park, can subvert the original
to
in
that
and
of producers
ways
products
ascribing meaning
by altering
never intended.
and
marketers
designers
structures that
Our primary concern in this article iswith organizational
in
that
those
have a global
borders
and
international
span
particularly
structures
also
and institutions
local and national
reach. Nonetheless,
us with
into
have provided
matter. Geographers
and planners
insights
of local industrial agglomerations
the spatial and social propinquity
innova
to buoy organizationally
and often highly
disaggregated,
and
and
activities
1988;
Scott,
Walker,
tive, economic
Storper
(e.g., Storper
of tacit
stressed the spatial embeddedness
1989). This work has usefully
how
work
98

between
of tight interdependencies
and the importance
geo
knowledge
and
firms
clustered
1995).
(Maskell
1999;
Storper,
Malmberg,
graphically
that such ag
We acknowledge
these points, and have argued elsewhere
are
most
of global
relational
where
the
the
portions
places
glomerations
value chains might be found (Sturgeon, 2003). The varieties of capitalism
literature, coming largely from political science (e.g., Berger and Dore, 1996;
rules
that national-level
1992), similarly
1999; Streeck,
Soskice,
argues
education
and
in
and institutions
finance, corporate
governance,
(e.g.,
stud
affect the character of industries. Other
and training) profoundly
that
show
ies (Borrus et al, 2000; Florida and Kenney,
1993; Lynch, 1998)
as
are
carried
rooted
characteristics
abroad,
many geographically
foreign
onto the global stage.
direct investment projects local and national models
can and do have profound
effects on value chain gover
These variations
even when
nance. For example,
for emergent
conditions
the underlying
are well established,
as value chain modularity
such
forms
organizational
as they are in the Japanese personal
computer
industry, large-scale out
to
be
antithetical
corporate
strategies and
long-standing
sourcing might
as
in
which
make rad
lifetime
such
institutions,
large firms,
employment
difficult
and
slow.
ical industry reorganization
extremely
the 'rules of the game' as it
It is also clear that global-scale
regulations,
effect on the shape and direction of change in global
were, have a profound
to apparel
from electronics
value chains. In a wide
range of industries,
on
in
for
duties
value
added
to household
selective
exemptions
goods,
as
status
807
and
most-favored-nation
such
section
locations,
particular
for Europe,
States and outward processing
for the United
arrangements
of
value
the geographical
have encouraged
chains,
global
fragmentation
as we have seen in the apparel case study. Yet political pressures
in both
to retain (or gain) apparel
nations
and developing
jobs, and
developed
are
to
risk
desires
diversification,
through geographical
managerial
spread
be
than itwould
likely to keep the apparel value chain more fragmented
criteria alone.
if production
decisions were based on economic
of the
of factors that affect the evolution
While
there are a multitude
we feel confident
that the variables
internal to our model
global economy,
the shape and governance
of global value chains in important
influence
context within which
of
the
institutional
ways,
regardless
they are situ
framework
that we propose
takes us part of the way
ated. The governance
of global value chains, but much
toward amore systematic understanding
areas is the development
remains to be done.17 One of the most pressing
that are consistent with the frame
of policy tools for industrial upgrading
work. One of the key findings of value chain studies is that access to devel
on participating
oped country markets has become increasingly dependent
in global production
networks
countries.
led by firms based in developed
is
the
of
value
essential
chains
for
Thus,
governance
global
understanding
countries can gain access to global markets, what
how firms in developing
99
POLITICAL
ECONOMY
the benefits of access and the risks of exclusion might be, and how the net
in global value chains might be increased. While
gains from participation
in the global economy
is
the search for paths of sustainable
development
an inherently difficult and elusive objective, our task is greatly facilitated
inwhich global value chains
by having a clearer sense of the various ways
are governed,
and the key determinants
that shape these outcomes.
ACKNOWLEDGEMENTS
to thank the Rockefeller
for its gener
Foundation
wish
Initiative
of
Global
Value
Chains
the
(see:
support
http://www.
In preparing
this article, we have drawn upon dis
globalvaluechains.org).
on value chains held in Bellagio,
cussions at workshops
Italy, in September
in April 2002. How
2000 and April 2003 and in Rockport, Massachusetts,
The
ous
authors
ever,
any
errors
or
in
shortcomings
the
article
our
remain
own.
NOTES
1 While
'internationalization'
across
tivities
national
to
refers
boundaries,
the
of
spread
geographic
implies
'globalization'
the
ac
economic
inte
functional
gration and coordination of these internationally dispersed activities (Dicken,

2003:12).
in this paper can explain all
2 We do not suggest that the theory developed
governance
patterns
as a
used
complement
can be observed
in
and
3
sectoral
imported
(1998:
to the amount
is a term is used by Clemons
'Explicit coordination'
non-market
use
the term
80-81)
'vertical-specialization
in goods
that
embodied
of imported
inputs
a
when
of
trade occurs
'Vertical
country
global
specialization'
it later exports.
to
intermediate
parts
goods
produce
to refer
exported.
uses
specificity.
et al.
Hummels
Similarly,
based-trade'
are
in
be
The
chains.
should
value
theory
global
that
and complexity
to, not a substitute
for, the rich detail
their historical,
value
chains,
geographical,
especially
global
observed
drew
This work
competences
of economic
of coordination
forms
on
the analysis
in value
chains.
of Palpacuer
et al. (1993) to refer to
activity.
(2000) on core
and
complementary
6 Work on the apparel industry (Gereffi, 1999) and on commodity exports from
Africa (Gibbon, 2001) also showed a variety of contracting arrangements.
7 An indication of the range of studies is provided by the collection edited by
Gereffi and Kaplinsky (2001).
8
For
a discussion
of grades
a more
For
(2001).
general
and
see Reardon
et al.
industry,
and its
architectures
product
in the food
standards
of modular
discussion
implications for industry structure, see Baldwin and Clark (2000).

9 The development of product and process standards and their implications
value chain governance are discussed by Nadvi and W?ltring (2002).
10
Low
informational
tions
that
are
Furthermore,
and
suppliers
buyers,
unlikely
complexity
to occur
without
codification
regardless
if there is low complexity

still
do
not
have
the
of
generates
supplier
competence,
and a high possibility
capabilities
then it is likely that they will be excluded
to meet
the
two
for
combina
high
or
low.
for codification,
requirements
from the chain. While
100
of
this

does
not
generate
and with
common,
global
value
per se, it is a situation
chain
type,
for suppliers
requirements
increasing,
that
is quite
increasingly
perhaps
case is important insofar as it

likely to occur (Sturgeon and Lester, 2004). This
a
of
the
discussion
country suppliers
up
opens
problems facing developing
and policies for industrial upgrading.
11
architectures
Product
to modular.
integral
In integral
prod
of a product are tightly linked and
the functional elements
for a particular
optimised
from
vary
generally
uct architectures,
In modular
configuration.
architectures,
product
by
contrast, the physical building blocks (or sub-systems) of a product are loosely
coupled and designed to be relatively independent of one another because of
and
some
rules, which
compo
permit
and recombined
into a large num
and visible
design
to be disaggregated
interfaces
standardized
nents
sub-systems
ber of product variations (see Baldwin and Clark, 2000; Schilling and Steensma,
2001; Ulrich, 1995).
12 This discussion is based on Galvin and Morkel (2001).
13
In the Asian
ment
the full-package
(OEM).
context,
manufacturing
was
model
also
as
known
original
equip
14 Kenya is the largest exporter of fresh peas and beans from Africa to the Eu
ropean Union and by far the most important supplier to the UK market. This
section is based on the work of Dolan and Humphrey
(2000,2004).
15 This discussion is based on Sturgeon (2002).
16 This process is not driven solely by the efforts of suppliers. Value chain actors
to simultaneously
firm strategies
increase
outsourcing
a set of
have
created
sup
highly
supply-chains
capable
more
attractive
for lead firms
that have
that, in turn, make
outsourcing
pliers
and Lee, 2001).
the evo
yet to take the outsourcing
(Sturgeon
Similarly,
plunge
or local
from one national
value
lution
chains
of global
context,
emanating
Lead
co-evolve.
clearly
and consolidate
especially
value
chains
their
an
if successful,
provides
example
in other
rooted
places.
that often
a reaction
generates
17 A high priority for the future will be the development
of methods
formeasuring
in the model.
transactional
complexity,
the key
variables
Effective
for
proxies
in
and supplier competence must be identified and tested in
level of codification,
the field.
REFERENCES
S. and
Arndt,
Kierzkowski
omy, Oxford:
C. and
Baldwin,
H.
Kierzkowski,
(eds),
Oxford
Clark,
Fragmentation:
K.
(2001)
New
'Introduction',
Production
in
Patterns
Press,
pp. 1-16.
University
(2000) Design
Rules, Cambridge,
MA:
S.
H.
and
Arndt,
in the World
Econ
MIT
Press.
Berger, S. and Dore, R. (eds) (1996) National Diversity and Global Capitalism, Ithaca,
NY:
Cornell
Press.
University
Bonacich, E., Cheng, L., Chinchilla, N., Hamilton, N. and Ong, P. (eds), (1994) Global
Production: The Apparel Industry in the Pacific Rim, Philadelphia, PA: Temple
Press.
University
Bonnen, A. R. (2002) The

Phenomena7.
European
Fashion
Planning
Borrus, M., Ernst, D. and Haggard,

in Asia,
London
and New
York:
Industry in Galicia; Understanding

Studies,
10(4):
the 'Zara'
519-27.
S. (eds), (2000) International Production Networks

Routledge.
demons,
E., Reddi, S. and Row, M. (1993) The Impact of Information Technology
on the Organization of Economic Activity: The 'Move to theMiddle' Hypoth
esis7,
journal
ofManagement
Information
Systems,
10(2):
9-35.
101
David, P. A. (1995) 'Standardization
between
and Order
Freedom
Policies
POLITICAL
ECONOMY
for Network
Technologies:
in R. Hawkins,
Revisited',
R. Mansell,
The Flux
and
J. Skea
(eds), Standards, Innovation and Competitiveness: The Politics and Economics of

Standards inNational and Technical Environments, Aldershot, UK: Edward Elgar,
pp.
15-35.
4th
edition,
Dicken, P. (2003) Global Shift: Reshaping theGlobal EconomicMap
in the 21st Century,
London:
Sage.
K. and
P., Olds,
Towards
Scales:
P., Kelly,
Dicken,
ritories
and
Global
Economy',
and Networks,
Ter
the Global
for Analysing
H. W.-C.
(2001)
Yung,
a Relational
Framework
1(2):
Networks,
'Chains
89-112.
Dolan, C. and Humphrey, J. (2000) 'Governance and Trade in Fresh Vegetables: The
Industry', Journal of
Impact of UK Supermarkets on the African Horticulture
Development Studies, 37(2): 147-76.
Dolan, C. and Humphrey, J. (2004) 'Changing Governance Patterns in the Trade in
Fresh Vegetables between Africa and the United Kingdom', Environment and
Planning A, 36(3): 491-509.
of Production in the
Feenstra, R. (1998) 'Integration of Trade and Disintegration
Global Economy', Journal of Economic Perspectives, 12(4): 31-50.
Fine, C. H. (1998) Clockspeed: Winning Industry Control in theAge of Temporary Ad
vantage,
MA:
Reading,
Perseus.
Florida, R. and Kenney, M.

to the U.S.,
its Transfer
Galvin,
A.
P. and Morkel,
The
Structure:
(1993) Beyond Mass Production: The Japanese System and

York:
New
Case
Oxford
of
the World
8(1): 31-47.
Gereffi, G. (1994) 'TheOrganization
How
Retailers
U.S.
Korzeniewicz
pp.
Press.
University
on Industry
(2001) 'The Effect of Product Modularity
Shape
Industry',
Bicycle
of Buyer-Driven
Overseas
Production
Industry
and
Innovation,
Global Commodity
inG
Networks',
Gereffi
(eds), Commodity Chains and Global Capitalism, Westport:
Chains:
and M.
Praeger,
95-122.
(1999) 'International Trade and Industrial Upgrading
Gereffi, G
Chain',
Commodity
Gereffi, G
32(3),
Gereffi, G
Journal
of International
Economics,
in the Apparel
48: 37-70.
and Kaplinsky, R. (eds) (2001) 'TheValue of Value Chains', IDS Bulletin,
special
issue.
and Korzeniewicz,
ism, Westport:
M.
(eds) (1994) Commodity Chains and Global Capital
Praeger.
O. (2003) 'The Global Apparel Value Chain: What

Gereffi, G and Memodovic,
Indus
for
Upgrading by Developing Countries?' United Nations
Prospects
trial Development Organization
(UNIDO), Sectoral Studies Series, available
at
for downloading
<http://www.unido.org/doc/12218>.
Gibbon, P. (2001) 'Upgrading Primary Products: A Global Value Chain Approach',

World Development, 29(2): 345-63.
Granovetter, M. (1985) 'Economic Action and Social Structure: The Problem of
Embeddedness', American Journal of Sociology, 91: 481-510.
J., Dicken,
Henderson,
P., Hess,
M.,
Coe,
N.
and
Yeung
H. W.-C.
'Global
(2002)
Review of
Production Networks and the Analysis of Economic Development',
International Political Economy, 9(3): 436-64.
and Changing commodity Networks:
Hughes, A. (2000) 'Retailers, Knowledges
The
Case
Hummels,
D.,
Changing
Policy
of
the Cut
Flower
D.
Rapoport,
Nature
of World
Review,
31:175-90.
Geoforum,
'Vertical
(1998)
Yi, K.-M.
Specialisation
York
Reserve
Bank of New
Federal
Trade',
Trade',
and
and
the
Economic
79-99.
Humphrey, J. (2003) 'Globalisation and Supply Chain Networks:

in Brazil and India', Global Networks, 3(2): 121-41.
The Auto Industry
102

Humphrey, J. and Schmitz, H. (2000) 'Governance and Upgrading: Linking In
dustrial Cluster and Global Value Chain Research', IDS Working Paper, 120,
Institute
Brighton:
of Development
Studies,
of Sussex.
University
Humphrey, J. and Schmitz, H. (2002) 'How Does Insertion in Global Value Chains
Affect Upgrading in Industrial Clusters?', Regional Studies 36(9): 1017-27.
Jarillo, J. (1988) 'On Strategic Networks', StrategicManagement Journal, 9(1): 31^1.
D.
Keesing,
and
ing Countries:
Trade Policy,
from Develop
S. (1992)
Manufactured
'Marketing
Exports
in G. Helleiner
and Public
(ed.),
Learning
Support',
Sequences
and
Industrialisation
Oxford:
Oxford
Press,
University
Development,
Lall,
176-93.
pp.
Kogut, B. (1985) 'Designing global Strategies: Comparative and Competitive

Added Chains', SloanManagement Review, 26(4): 15-28.
R. and
Langlois,
P. (1995)
Robertson,
Firms,
and Economic
Markets
Routledge.
Leslie, D. and Reimer, S. (1999) 'Spatializing Commodity

Geography, 23(3): 401-20.
nor
Friends
'Neither
(1988)
Strangers:
in D. Gambetta
in French
Industry',
E.
Lorenz,
tracting
Basil
Oxford:
Relations,
Cooperative
T. (1998)
Home:
'Leaving
Lynch,
Automotive
Informal
Trust
(ed.),
An
Style
P. and
Malmberg,
tiveness',
Cambridge
Decades
with
Interview
A.
Victor
Business
and
Relations
23(1-2):
Asienforum,
of
by U.S.
Internationalization
of Economics,
58, Duisburg:
Social
Supply Chain Management,
Fung',
Harvard
and
Learning
167-85.
23(2):
Business
Industrial
Review,
Competi
to Trust Trust? Chinese Non-Contractual
Structure:
The
Case',
Singapore
Internationales
261-88.
K. and W?ltring,
port
'Localised
(1999)
Journal
T. (1992) 'Xinyong or How
Menkhoff,
of Subcon
and Breaking
Center.
Kong
Maskell,
Networks
Making
Blackwell.
Three
76(5): 103-40.
Nadvi,
Chains', Progress inHuman
J. (1998) 'House, Global and Entrepreneurial:

-
Hong
London:
Change,
Firms', IPCWorking Paper, 98-007, Cambridge, MA: MIT Industrial
Performance
Magretta,
Value
F. (2002) 'Making Sense of Global Standards',

Institut
f?r
und
Entwicklung
INEF Re
Gerhard-Mercator
Frieden,
University.
F. (2000)
'Competence-Based
A Discussion
of Current
Palpacuer,
works:
ment',
Competition
and Change
Strategies
and
Changes
4(4): 353-^100.
and
Their
Global
Net
Production
Implications
for Employ
Penrose, E. (1959) The Theory of theGrowth of theFirm, Oxford: Basil Blackwell.

W.
Powell,
Research
Prahalad,
Harvard
Reardon,
nor
'Neither Market
Network
(1990)
Hierarchy:
in Organizational
295-336.
Behavioural:
C. and G. Hamel
'The Core
(1990),
Competence
Business
T, Codron,
Review,
J.-M.,
68(3):
Busch,
Forms
of
of Organization',
the Corporation'.
79-91.
J. and Harris,
L., Bingen,
C.
(2001
) 'Global
Change
inAgrifood Grades and Standards: Agribusiness Strategic Responses inDevel

oping Countries', International Food and Agribusiness Management Review, 2(3):
421-35.
Schilling, M. A. and Steensma, H. K. (2001) 'The Use of Modular

Forms:
An
Industry-Level
Analysis',
Academy
of Management
Organizational
Journal,
44(6):
1149-68.
Schmitz, H. and Knorringa, P. (2000) 'Learning from Global Buyers', Journal of

Development Studies, 37(2): 177-205.
Soskice, D. (1999) 'Divergent Production Regimes: Coordinated and Uncoordinated
Market Economies in the 1980s and 1990s', inH. Kitschelt, P. Lange, G. Marks,
103
and J. D. Stephens
(eds), Continuity
POLITICAL
ECONOMY
and Change in Contemporary Capitalism,
Press,
University
Cambridge:
Cambridge
M.
of Regional
(1995) The
Storper,
Resurgence
pp. 101-34.
Economies,
Ten Years
Euro
Later',
pean Urban and Regional Studies, 2(3): 191-221.

Storper, M. and Scott, A. (1988) The Geographical Foundations and Social Regu
in J.Wolch, and M. Dear (eds), The
lation of Flexible Production Complexes',
Power
Boston:
of Geography,
Allen
and Unwin,
21-40.
pp.
Storper, M. and Walker, R. (1989) The Capitalist Imperative: Territory, Technology and
Industrial Growth, Oxford: Basil BlackweU.
Streeck, W. (1992) Social Institutions and Economic Performance, London: Sage.
Sturgeon, T. (2002) 'Modular Production Networks: A New American Model of
Industrial Organization',
Industrial and Corporate Change, 11(3): 451-96.
T.
'What
(2003)
Sturgeon,
Really Goes on in Silicon Valley? Spatial Clustering Dis
inModular
persal
225.
Production
T. and Florida,
R.
in the Motor
Vehicle
Sturgeon,
ment
Networks',
Journal
of Economic
Deverticalization
'Globalization,
inM. Kenney,
and B. Kogut
(2004)
Industry',
Global Advantage, Stanford, CA: Stanford University

T. and
Co-Evolution
Lee, J.-R. (2001)
'Industry
for Electronics
Manufacturing',
Paper
Supply-Base
Winter
Conference,
June.
Aalborg,
Sturgeon,
in East
Suppliers
Asia',
Global Production Networking

Oxford
Thorelli,
University
H.
(1986)
agement
Journal,
Research
Policy,
in S. Yusuf,
A.
Altai,
and
and
the Rise
and
for
(eds),
and Technological Change in East Asia, New York:
'Networks:
Between
Markets
and
Hierarchies',
Strategic
Man
7: 37-51.
in theManufacturing
Firm',
24: 419-40.
O. (1975)Markets andHierarchies, New York: Free Press.

O.
(1983)
American
'Credible
Economic
Yeats, A. (2001) 'JustHow

(eds),
Oxford
and
Challenges
K. Nabeshima
Williamson,
Oxford:
of a Shared
at Nelson
Presented
Williamson,
Kierzkowski
Creating
Press.
Ulrich, K. T. (1995) The Role of Product Architecture
change',
Employ
(eds),
Press.
Sturgeon, T. and Lester, R. (2004) The New Global Supply-base:

Local
3:199
Geography,
Commitments:
73(4):
Review,
Using
519-40.
Big is Global Production
Fragmentation:
Press,
University
New
pp.
Production
Hostages
Sharing?'
Patterns
to Support
in S. Arndt
in the World
Ex
and H.
Economy,
108^13.
104

Gerrefi - Governance Global Value Chains

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gerrefi - Governance Global Value Chains

Uploaded by

Copyright:

Available Formats

The Governance of Global Value Chains

Author(s): Gary Gereffi, John Humphrey and Timothy Sturgeon

(2) the ability

and (3) the capabilities

The theory generates

and power asymmetry. The article highlights

in significant ways during the past sev

GEREFFI ET AL.: GLOBAL VALUE CHAINS

are also concerned with how

GEREFFI ET AL.: GLOBAL VALUE CHAINS

IN GLOBAL VALUE CHAINS

GEREFFI ET AL.: GLOBAL VALUE CHAINS

3. Relational value chains. In these networks we see complex interactions be

3. A THEORY OF VALUE CHAIN GOVERNANCE

GEREFFI ET AL.: GLOBAL VALUE CHAINS

are easily codified, product

GEREFFI ET AL.: GLOBAL VALUE CHAINS

5. Hierarchy. When product specifications

The five global value chain governance

Five of them generate

to high levels of explicit coordination

4. DYNAMIC VALUE CHAIN ANALYSIS:

4.1. The bicycle

GEREFFI ET AL.: GLOBAL VALUE CHAINS

GEREFFI ET AL.: GLOBAL VALUE CHAINS

Over time, relationships

risks of this have been contained by the development

industry: From hierarchy

For most of the twentieth

GEREFFI ET AL.: GLOBAL VALUE CHAINS

and $256 mil

4.5. The dynamics

GEREFFI ET AL.: GLOBAL VALUE CHAINS

because of, high levels of vertical

GEREFFI ET AL.: GLOBAL VALUE CHAINS

gration and coordination of these internationally dispersed activities (Dicken,

is a term is used by Clemons

et al. (1993) to refer to

implications for industry structure, see Baldwin and Clark (2000).

if there is low complexity

and a high possibility

then it is likely that they will be excluded

from the chain. While

GEREFFI ET AL.: GLOBAL VALUE CHAINS

per se, it is a situation

case is important insofar as it

of a product are tightly linked and

the functional elements

17 A high priority for the future will be the development

and supplier competence must be identified and tested in

Bonnen, A. R. (2002) The

Borrus, M., Ernst, D. and Haggard,

Industry in Galicia; Understanding

S. (eds), (2000) International Production Networks

(eds), Standards, Innovation and Competitiveness: The Politics and Economics of

Dicken, P. (2003) Global Shift: Reshaping theGlobal EconomicMap

in the 21st Century,

Florida, R. and Kenney, M.

(1993) Beyond Mass Production: The Japanese System and

(2001) 'The Effect of Product Modularity

(eds), Commodity Chains and Global Capitalism, Westport:

(1999) 'International Trade and Industrial Upgrading

and Kaplinsky, R. (eds) (2001) 'TheValue of Value Chains', IDS Bulletin,