Durban Commodity Chain
Durban Commodity Chain
Durban Commodity Chain
Working Paper No 46
2006
ISBN : 1-86840-607-5
In the closing sentence of his call for a new paradigm in port studies,
Robinson issues the following challenge: “the role of ports and the way in
which ports position themselves in the new business environments beyond
2001 must be defined within a paradigm of ports as elements in value-driven
chain systems, not simply as places with particular, if complex, functions”
(2002: 252). Recent research on the maritime shipping industry has focused
upon the emergence of more integrated logistics chains, both within the
maritime industry itself and between maritime- and land-based
transportation modes (Slack at al, 2002; Notteboom, 2002). Greater
integration of logistics chains raises difficult questions for ports and port-
cities seeking to secure or maintain dominant positions within global trade
flows (Notteboom & Winklemans, 2001; Heaver et al, 2001). The attention
to integration along logistics chains parallels and indeed draws upon the
wider literature on global commodity chains (Gereffi, 1994; Gereffi et al,
2005) and supply chains (Cox et al, 2002).
The logistics chain perspective indicates that ports should seek to insert
themselves as privileged nodes within particular logistics chains. While we
do not disagree with this general policy conclusion, in this paper we show
that the potential for conflict and uncertainty over goals, roles and actions
with respect to supply chain insertion increases significantly when we
rescale the analysis in two dimensions; namely, when move from the port to
the port-city and the nation, and from the logistics to the whole supply chain.
While recognising the close connections between them, for clarity we
distinguish between logistics chains which deal with the distribution of
goods in physical space, and value chains which deal with inter-firm
relationships in economic space. We use the term supply chain to encompass
both. However, with some exceptions, few attempts have been made to nest
the analysis of logistics chains within the larger supply chains in which they
are located. We argue that this analytical bracketing may potentially lead to
incorrect policy advice.
Using a case study of the attempts since 1994 to insert Durban into global
automotive supply chains, we trace the debate and conflict over what
logistics functions to serve in which value chain. We show that this conflict
results not only from the tensions within or between competing value and
logistics chains, but also from the tensions between parallel local and
national decision-making arenas. At the local level, port planning and day-
to-day port management are not coordinated with city planning and
Inserting Durban into global automotive supply chains 2
economic development functions (Hall & Robbins, 2002). At the national
level, port policy is caught up in the privatisation and transformation of the
national transport agency, while automotive sector policies reflect national
industrial goals. This decision-making environment seriously complicates
the goal of rapid agreement and alignment around supply chain insertion.
The goals of actors operating within supply chains, if not the precise
strategies they may employ or the nature of the regimes governing specific
chains, are relatively clear. First, actors seek to insert themselves into supply
chains. Supply chains provide actors with access to technology, capital,
supplies, expertise, and markets, and most importantly, to knowledge about
these critical resources (Gereffi, 1994). For ports, the goal of logistic chain
insertion has been pursued through improved landside connections,
incentives, leases and concessions to attract more port callers, and port
networking (see Heaver, 1995; Notteboom & Winklemans, 2001). National
and local economies seek to insert themselves into what are, at least in the
automobile sector, value chains that dominated by a decreasing number of
global players (Sturgeon & Florida, 2000). With respect to the case study
that follows, Morris at al wrote that “(i)f Toyota SA is to compete in the
long term, it needs better access to Toyota Japan’s global networks, and it
similarly needs to help facilitate significant export contracts for its domestic
component suppliers” (2002: 125).
It is also important to recognise that we have been using the term supply
chain to apply to what are in fact a diverse group of historically contingent
and unique formations (Hall, 2004). As indicated above, we have reduced
this diversity in supply chains to two ideal types; the logistic chain and the
value chain. In part, we do this is because of the historic functional
separation between production / manufacturing and distribution /
transportation, although this is an epistemological divide that has been
superceded by economic globalisation (Hall et al, forthcoming). The
question for ports, and the city-regions that host them, is which chain or
chains might they seek to insert themselves into, recognising that there may
be trade-offs between different chains. For example, greater port throughput
(in a logistics chain) need not necessarily translate into greater local
economic activity (in a value chain); indeed, modern ports may impose
negative externalities on their host cities without providing commensurate
local employment and other economic benefits (Campbell, 1993).
The over-arching point is that recognising that ports are elements in global
supply chains severely complicates the scalar and bounding dimensions of
the analytical and policy task. To begin answering the question, which link
in which chain, we propose an ideal-typical framework that cuts across three
spatial scales (port, port-city and nation), and identifies various actors
seeking to insert themselves into, integrate activities along, and secure
dominance within two types of value chain (logistic and value chains). The
framework is summarised in Table 1; we have included a series of examples
of the strategies that might be employed with respect to each. The case study
that follows will illustrate how a variety of actors acting at different scales
have used these strategies to insert Durban into global automobile supply
chains. We show that conflict results not only from the tensions within or
between competing value and logistics chains, but particularly from the
tensions between parallel local and national decision-making arenas.
By 1981, TSA had established its dominance in the domestic market as the
primary producer of passenger and light commercial vehicles, dominance in
total new vehicle sales and total production volume that it maintains to this
day. Both Toyota’s scale of output (in relative South African terms) and the
considerable range in vehicle platforms that it produced further encouraged
development of a component supplier base, primarily in Durban, but also in
other regions of the country. However, despite the steady growth of
localised components suppliers during the period of the 1970s and 1980s,
TSA’s growing production volumes remained very closely linked to its
ability to import the bulk of its components through the nearby Port of
Durban. These inputs supported Toyota’s scale and range of production that
was central to the company’s ability to maintain its domestic dominance in a
variety of product sub-markets.
What then were the logistic and value chain issues in this “pre-global”
context? In the first instance the highly protected production environment,
including various forms of transport and production subsidies, allowed for
considerable protection for producers from the price, quality and delivery
pressures that are de rigueur in today’s globalised production and market
environments. This orientation towards relatively uncompetitive domestic
Inserting Durban into global automotive supply chains 7
markets reduced the attention paid to logistics and supply chain management
within firms. For the most part, the transport infrastructure at the Port of
Durban, provided sufficient capacity and acceptable service. Of course, the
trend towards containerisation, and attendant increases in the size of vessels,
was beginning to signal the need for new investment and new approaches to
port management. Nevertheless, for much of this period there was limited
pressure from port users and low responsiveness from port decision making
structures with regard to performance. A similar lack of responsiveness in
rail and falling performance standards led many businesses to shift long-haul
land transport from rail to road (MSA, 1998).
The combined effect of domestic recession during the late 1980s and the
process of economic reform initiated by the Apartheid government in this
period, in a desperate attempt at securing its survival, thrust South African
businesses into an environment of rapidly escalating competition and
reduced government protection. Even in sectors which were not directly
affected by rapid tariff reductions, the extended domestic recession required
that they seek opportunities in export markets. This soon led to a much
greater appreciation, at least in the private sector, of the importance for firms
of inserting themselves into, achieving integration efficiencies along, and
securing dominance within global supply chains.
Figure 1: South African Gross Value Added (GVA) annual average growth, 1996-03
9 Community services
8 Finance
7 Transport
6 Trade
5 Construction
4 Electricity
3 Manufacturing
2 Mining
1 Agriculture
-1.0% 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% 8.0%
Percentage
30.0%
25.0%
Percen tage
20.0%
15.0%
10.0%
5.0%
0.0%
1996 1997 1998 1999 2000 2001 2002 2003
Y ears
These policy adjustments were deemed favourable by the OEMs and they
began, one-by-one, to take advantage of the opportunity to build South
Africa into their global operations. According to Barnes et al (2003), the
automobile sector grew rapidly between 1994 and 2002, doubling its export
to output ratio and accounting for an increasing share of output value, gross
value added and manufacturing employment. As a result of the combined
effects of the MIDP and the depreciation of the Rand, the share of the
Inserting Durban into global automotive supply chains 10
automobile sector of total manufacturing sales grew from 9.7 percent in
1994 to 12.8 percent in 2003.” (Lorentzen et al, 2004: 8) Growth in
automobile sector exports and imports was accompanied by rapid growth in
automobile logistics. While vehicle production overall increased after 1998,
production for the domestic market has continued to fall: a greater portion of
local demand is being met by imports, at the same time that exports have
grown (see Figure 3).
350000
300000
250000
200000
150000
100000
50000
0
1995 1996 1997 1998 1999 2000 2001 2002 2003
Source: Reproduced from Barnes J, Kaplinksy R and Morris, 2003: p 9 – originally from NAAMSA.
25,000
2 1 ,8 8 9 2 2 ,1 5 0
20,000
1 7 ,5 6 6
1 4 ,8 9 3
15,000
Rand (Million)
1 0 ,7 7 0
8 ,8 5 7 9 ,3 3 9
10,000
8 ,6 4 5
7 ,8 9 2
6 ,0 6 7
5,000
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
Ye a r
One notable early MIDP beneficiary was VWSA, which had been exporting
Jetta’s to China since 1991, but perhaps the most prominent MIDP success
was BMW, the German luxury-car manufacturer. BMW were first
assembled in South Africa from CKD ‘kits’ from 1968; in 1973 BMW South
Africa established the first BMW assembly plant outside Germany in
Rosslyn, north-west of Pretoria. In 1994, the first exports were sent to
Australia, and in 1996, with a R1bn investment, the plant became “BMW
World Plant, Rosslyn”, an integrated part of the global production system. In
1999, a new vehicle distribution center opened at the plant, and in 2002,
approximately 80 percent of output was being exported. Whilst both VW
and Daimler-Chrysler export through Eastern Cape ports, the Port of Durban
focuses on serving a role as the main import channel and the primary export
channel for BMW and Toyota. Ford also now exports the Focus to Europe
through Durban.
Hence, by 2000 Toyota South Africa (TSA) had announced that a majority
of the local business had been acquired by Toyota Manufacturing
Corporation of Japan (TMC) and that the Durban plant would begin to
feature in Toyota’s global production system.1 This insertion of the local
Toyota operation in the firms’ global supply chain – achieved by adjusting
the licensing and ownership arrangements - was accompanied by a series of
actions to integrate, secure and upgrade both value and logistics chains. To
take full advantage of the MIDP restructuring, TSA had to export. To export
they had to renegotiate license agreements with TMC of Japan.2 In turn,
TSA and TMC also began working together to integrate the supply chain by
improving quality and delivery standards on all sorts of levels – in plant
design, in the feasibility study for the catalytic converter plant (which
opened in 2000 as the majority TMC-owned Catalyer), and in background
work with component suppliers. The work with suppliers was helped
significantly by the formation by local components firms of the
Benchmarking Club, and subsequently, the Durban Auto Cluster. This
initiative was the result of proactive intervention by local actors, including
suppliers, the local university, and municipal government (see Lorentzen at
al, 2004; and Morris and Robbins, 2004).
The agenda of the logistics working group revolved around both maritime
and landside issues. The landside concerns are an example of logistics chain
integration, and entailed greater coordination between Toyota and the
suppliers on road haulage for parts delivery. Toyota was sending trucks to
Port Elizabeth to collect parts, while Durban-based suppliers were sending
parts to Eastern Cape OEMs (VW and Mercedes). Collaboration resulted in
significant back-haul cost savings.
On the maritime side, while some concerns did revolve around eliminating
inefficiencies through greater logistics chain integration, they also concerned
the uncompetitive local port services industry; in other words, they also
revolved around issues of logistics chain dominance. Both Toyota and its
suppliers felt they were getting poor service and highly variable pricing. The
logistics working group put some collective pressure on the South African
Revenue Service (customs), and local clearing and forwarding agents to
simplify documentation and make transactions cheaper. However, many
smaller shippers also felt that they were paying a premium on containers,
and so they established a consortium (shippers association) to improve
prices for small shippers. The group negotiated improved and consistent
rates, in turn, for the European, North American, Asian, and Latin American
routes. Finally, the group tried, without much success to pressure the port
Inserting Durban into global automotive supply chains 14
management to improve container terminal operations overall in terms of
productivity, lost boxes, damage, delays, and so on.
The group was supported by lots of media attention on the port’s problems,
and even succeeded in having City government pass a resolution in 2000
calling on the port to improve its operations, and to work with city. This
pressure did not result in much improvement.
The planning, location, design and scale of the initial automobile terminal
reflected BMWs needs, but certainly not those of the soon to be exporting
Durban-based Toyota. The decision to build the automobile terminal was
taken at exactly the same time as two joint port-city planning processes were
under way. The Port-City forum, begun erratically in 1997, and this was
followed by the associated South Industrial Basin6 (SIB) planning process
which was actively developing a grander vision for an automotive logistics
park with supply chain value adding activities (ie. customisation,
accessorisation and associated space for parts suppliers). The SIB process
was funded in part by the national government Department of Trade and
Industry, yet Portnet bypassed this process to go ahead with the new
automobile terminal. The case for locating an auto terminal at the SIB end of
the port is strong, since the SIB accounts for 29.2 percent of all the
automotive components employment in the Province of KwaZulu-Natal, and
half of all the firms with turnover of more than R300m per annum (Barnes
and Johnson, 2004).
Instead Portnet chose for the automobile terminal a location on the northern
side of the port, adjacent to the Central Business District. Cars for export
from Toyota have to go to far side of port, through downtown and on public
roads, facing insurance concerns, congestion and delays. While the terminal
is served by rail, this involves shunting from the rail head adjacent to the
SIB. Yet BMW is the only auto terminal shipper that makes use of rail
(Arikin, 2004: 57). Finally, the auto terminal soon ran into capacity
constraints, necessitating some very expensive additional investments. These
involved doubling the number of parking spaces to 7 000 by building a
multi-storey parking garage, and the building of a dedicated bridge over the
rail lines. These were completed in 2004.
However, it has become clear that Toyota was simply testing the waters and
ensuring the robustness of its systems as during the course of 2004 the
company announced that it intended to introduce two other export-focused
assembly lines. According to an industry expert, the decision has been made
to increase Toyota production in Durban to the order of world status, or
approximately 250,000 units per year by 2010 (Barnes, 2005). Of these,
between 120,000 and 140,000 units will be exported, with most going to
Europe. It is highly unlikely that the existing automobile terminal port
facility will be able to accommodate this growth.
Sadly, this is hardly out of character for the NPA. While the NPA has spent
the better part of the last four years developing a Port Masterplan for
Durban, it has only been during the course of 2005 that external stakeholders
have had access to the plan details. The fact that it was developed in
isolation from other key stakeholders is reflective of ongoing governance
challenges of a parastatal system which still carries the baggage of
hierarchically imposed solutions that were the order of the day under
Apartheid. At the same time the intransigence of the Durban port authorities
reflects, in no small measure, the ports’ dominance within the supply chains
that are routed through it. The distances between ports in South Africa are so
great and the capacity of the entire system is so constrained, that the Port of
Durban enjoys an effective monopoly that remains unchallenged. For now,
and for the foreseeable future, there is no alternative to Durban for the
shipment of either parts or vehicles with an origin or destination from
Durban to Gauteng.
We began this paper by embracing the notion that ports and port-cities
should seek to insert themselves into supply chains. However, what our case
study of inserting Durban in global automobile supply chains has shown is
that this is a complex process, involving multiple actors and chains. For this
reason we asked which link, in which chain? We have argued for an
approach to supply chain insertion which recognises that actors in chains
face both internal and external competitive / dominance and cooperative /
integrative pressures. We have also argued that supply chain insertion plays
out across multiple scales. Hence our conceptual framework differentiates
In the case of Durban, there has been some success in inserting the port into
the logistics chain of one exporter, while improving facilities for automobile
importers. However, this supply chain insertion was driven by a non-local
shipper taking advantage of a successful national industrial strategy, and was
implemented on the docks by a centrally-controlled port authority. In other
words, the port authority did not make it happen, and could not have made it
happen, alone. And it has come at a significant local cost, most notably in
the lost opportunities to accelerate the insertion of Durban into Toyota’s
global value chain.
In a general sense, we may conclude that there are two forms of conflict
here. First, we have supply chains in competition with each other to
dominate particular (critical) supply chain resource, in this instance the
attention and facilities of the national port authority. Second, we have a port
seeking to insert itself into particular extra-local logistics chains, and a port-
city seeking prominence for value chains embedded in the local economy.
The shift to a world dominated by supply chains will not eliminate the
problems of highly localised tensions between port and city; indeed in some
cases it may exacerbate them.
The situation we have described might have been very different if we were
dealing with a port authority more responsive to local economic actors. For
this reason we conclude by noting that the thorny issues of port governance
Inserting Durban into global automotive supply chains 19
and of improved city-port coordination remain unresolved. In large part, the
scalar tensions playing out on the Durban waterfront are a consequence of
the flip-flopping of national ports policy. The post-Apartheid government’s
first attempt to address transport infrastructure investment backlogs, Moving
South Africa (1998), failed to take root. Similarly, government has now
backed away from the National Commercial Ports Bill which contained an
initial commitment to separate the ports from the national transportation
parastatal, Transnet, as well as to allow terminal concessions. Instead, since
the early 2000s, government has endorsed what is today the status quo,
namely the NPA and SAPO operating as two separate divisions under
Transnet. Most recently government unveiled a National Logistics Strategy,
which recommends reconfiguring and harmonising institutions and
regulations, and allowing more competition in transport sector. However,
this national policy has very little to say about effective action at the local
level to address logistic and value chain integration. From our perspective,
this strategy is unlikely to do enough to allow the port and the city of Durban
to answer, in a timely fashion, the next time they are asked, which link, in
which chain.
ENDNOTES
2. TSA stopped producing the Camry, which was now imported from
Australia. It also began importing the RAV4 and Prado, and began
components exports with MIDP tariff benefits. TSA still has the
greatest model range of all South African OEMs, producing
commercial trucks, a low-end hatch-back (Tazz), mid-sized sedan
(Corrola), higher end hatch-back (RunX), minvan / taxi (HiAce), and
4. These figures apply for firms participating in the initiative from both
KwaZulu-Natal and the Eastern Cape.
5. This is not to say that BMW was spared the efficiency concerns faced
by other Transnet users. The firm still uses road rather than rail to
bring containerised parts from Durban to bypass the City Deep inland
port, and even use Port Elizabeth and Cape Town to avoid the delays
at the Durban container terminal.
6. The Southern Industrial Basin extends from the south end of the port
to south of the Durban International Airport, and includes Prospecton,
Jacobs, Mobeni and Seaview. Prospecton is home to Toyota SA.
REFERENCES