Paper presented at the Conference on
Clusters, Industrial Districts and Firms: the Challenge of Globalization. Conference in honour of Professor Sebastiano Brusco
Modena, Italy. September 12-13, 2003
POLICY LESSONS IN ORGANISING VERTICAL AND HORIZONTAL COOPERATION IN VALUE CHAINS AND INDUSTRIAL CLUSTERS
Mike Morris, Justin Barnes1 School of Development Studies University of Natal Durban 4041 South Africa morrism@nu.ac.za
June 2003
KEYWORDS: Clusters, Global Value Chains, South Africa, Policy, Automotive, Competitiveness.
Not to be quoted without the authors permission
1
Prof Mike Morris is also currently a Visiting Fellow at the Institute of Social Studies at the Hague – morris@iss.nl; Dr Justin Barnes is also managing director of Benchmarking and Manufacturing Analaysts (www.bmanalysts.com) which operates as the chief facilitation service provider of the automotive clusters discussed in the paper (justin@bmanalysts.com).
INTRODUCTION The international literature on the industrial clustering shows how clustering has enhanced the economic development of local economies and the collective efficiencies derived from cooperation (Pyke 1992; Rosenfeld 1996; McCormick 1999; Amin 1999; Schmitz 1995). Likewise, the global value chain literature argues strongly for the linkage role that lead firms along the chain play in promoting and assisting firms to engage in industrial upgrading (Gereffi 1999; Kaplinsky and Morris 2002; Humphrey and Schmitz 2002;). However, the two sets of literature (apart from Humphrey and Schmitz 2002) tend not to talk to each other. As if industrial clusters do not lock into global value chains; as if global value chain networking linkages do not encompass cooperation between sets of firms at the same level of linkage. In both these different sets of literature, the research and the discussion also tends to examine the past - how upgrading and learning has taken place within already existing clusters or already established value chains (Bair and Gereffi 2001; Dolan and Humphrey, 2000; Humphrey and Schmitz, 1998; Nadvi, 1997; Schmitz, 1995, 1997). However there seems to be little research, or at least a dearth of publication, of how vertical or horizontal linkages can be purposively created. Building industrial clusters and value chain linkages is often propagated within the discourse of industrial policy as of critical importance. The crucial role of the state, private institutions and the private sector within the national system of innovation is also emphasised in accessing and maintaining networking linkages. However, setting out government policy aimed at developing value chain linkages and/or facilitating clustering of firms is one thing. It is quite another implementing such policy and overcoming the numerous institutional obstacles hidden behind the fine words and programmes. This emphasis within the realm of policy seldom extends to issues of implementation, and studying how purposive action can implement policies to produce vertical and/or horizontal networking and learning tends to disappear from the policy discourse. There are also very few studies that record, analyse and draw lessons on the role of intermediaries and local product champions in facilitating these processes of horizontal and vertical cooperation. This is strange since there seem to be numerous attempts to experimentally create learning networks, foster trust and collective efficiency taking place in various parts of the globe, only some of which is reported on (Cleveland 1995; Kuper 1997; Doner & Schneider 2000; Bessant and Francis 1999; Barnes and Morris 1999; Morris 2001; Bessant and Tsekouras 2001; Lundequist and Power 2002; Meyer-Stamer 2003). Mostly, the opportunity to document these organisational experiences, analyse the dynamics at play, conceptualise the policy implications, and draw out some lessons to be learnt for the design and operation of other learning networks has tended to be neglected. The literature also tends not to distinguish the different dynamics between developed and developing countries in setting up learning networks, creating clusters, or value chain cooperation. However, learning networks in developing countries encounter particular problems, for these economies share a number of common characteristics which inhibit learning processes (Bessant et al 2003). Developing country markets tend to be supply-constrained and favour low- income goods. Their labour markets are also generally characterised by low levels of skill. The incentives to technical change are often weak and are overly-biased towards process changes. These economies exhibit weak early warning systems for picking up signals about rate, direction and strength of technological challenges. Developing country environments also often tend to be low-trust, riven with conflict and suspicion. Institutional failure and under-preparedness is common, and they are characterised by poorly articulated national and regional systems of innovation. Their economies are often also dominated by
2
firms with low degrees of specialisation and innovation capacity. The irony is that the need to upgrade onto a trajectory of international competitiveness is overwhelming, but the capacity and ability (from a variety of sources) to do so is often very limited. The need to co-operate to learn and face the hostile international challenges is obvious, but the fear of doing so is incapacitating. The objective of this paper is to address the issue of the dynamics of implementation, of the processes at play in building co-operation between enterprises; and it does so by showing what can be achieved in developing country environments in setting up learning networks, clustering and value chain co-operation. The paper is essentially divided into two parts. In the first we tell the process story of setting up, maintaining and running such networks in South Afric a. This section deals in some detail with the experience of setting up five networks of firms in South Africa, which have engaged in horizontal and vertical co-operation. Three of these (KwaZulu-Natal, Eastern Cape and Gauteng Auto Benchmarking Clubs) are location specific involving continuous improvement cooperation between auto component firms, and would fall into classic cases of industrial clusters of geographically proximate firms co-operating horizontally to learn and upgrade their process manufacturing activities. The other two involved both horizontal and vertical cooperation through organising the value chain as a network. One was in the furniture value chain (the Saligna Value Chain Group) encompassing firms across the entire country. The other (the Durban Automotive Cluster) is geographically specific but encompasses the entire automotive value chain in this region. The analysis of cooperation between these groups of firms is important for a number of reasons. Firstly they have been set up in an extremely low trust environment. South Africa, emerging from a long history of extreme conflict under apartheid, is hardly a society that would rank high on any international country list of trust, harmony and social peace. If functioning, successful vertically and horizontally cooperative clusters of firms can be set up in this environment, then clearly there may be generalisable lessons for other societies. Secondly, these examples have gone beyond classic horizontal cluster cooperation between similar sized firms and have also involved vertical cooperation along value chains. In separate sub sections devoted to them each of these networks is described, the strengths and weaknesses analysed and the successes and failures elaborated. In the concluding part we move away from the empirical story by focusing on what can be generally learnt from these experiences and applied in other cases. The focus here is drawing out the implications of these experiences so as to address the gap in current policy and strategy work in this area of industrial policy – i.e. the implementation task of setting up and running networks of clusters and value chains to achieve horizontal and vertical co-operation. BUILDING COOPERATION IN FURNITURE AND AUTOS The Saligna Furniture Value Chain2 The South African furniture and timber products industry had a long history of little trust and lack of cooperation. The Saligna value chain group was catalysed into being as a result of external crisis impacting on the industry and internal need for change in order to survive. The external forces arose from a massive reduction in domestic market demand of Saligna (a species of eucalyptus hardwood) from its traditional source as cheap tough mine stopes, as well as the rise of global environmentalism opening up higher value added export
2
. See Morris 2001, Kaplinsky, Morris and Readman 2002 for an extended discussion of this cluster.
3
opportunities for Forestry Stewardship Certified (FSC) hard wood furniture. A key market driver in the global timber products industry is the stress on environmental responsibility. For most developing countries this threatens their hardwood furniture exports which traditionally use indigenous rain forests. In South Africa, however, Saligna was commercially grown in large plantations as cheap tunnel stopes for the mines. However mining restructuring resulted in a radical decline in market demand, creating saligna surpluses. This provided the internal stimulus, for the plantation companies and sawmills dominating this value chain were seeking new markets to realise their sunken plantation investments. Their existing pulp and paper processing operations was one alternative but so was the expansion of the higher value added furniture sector. The timber surplus, combined with the fact that saligna finishes can simulate threatened hardwoods, offered unexpected possibilities for exporting furniture to Europe and North America. It also provided the potential to move furniture producers into new market niches, with higher unit prices. This was not lost on some timber manufacturers and suppliers who were also experiencing raw material shortages. Saligna furniture offered a low-cost, environmentally acceptable alternative to increasingly scarce and highly-priced traditional hardwoods. The changing perspective of the sawmills had a huge impact. Previously, through controlling the quantity and quality of timber supplied to the manufacturers, they held the rest of the value chain to ransom. Now it was their desire, and need, for change which provided the foundation for the development of a strong sense of the interdependence of players along the value chain. However, overcoming longstanding barriers to trust and translating this into actual co-operation with mutual benefits proved more complex. It was a combination of international and local university based external intermediaries and internal change agents from the sector championing co-operation that overcame the impasse. The external intermediaries became the facilitators and were accorded this role because they were seen as being able to bring international status and esteem to the group, maintain a necessary level of neutral credibility, and provide potential leverage with government. They brought together representatives from all the major sectors in the value chain –growers, sawmills, furniture manufacturers, the furniture export council, and two key government departments – who, in mapping their value chain, brought out the key problems between each link. The external intermediaries facilitated the group’s meetings, provided important information at the workshops, and played a useful information linking role between meetings. A key role producing concrete, practical results was performed by technical task teams. These were set up to tackle the two key issues identified of maximising the quantity as well as quality of the timber supply, and upgrading of the product’s design, marketing and branding for the export market. These technical task teams were comprised by, and led by, the enterprise participants. Most of them (especially those pertaining to quantity and quality of timber) were co-ordinated, and dominated, by enterprise participants from the various sawmilling operations, who were the lead firms in the value chain. In fact the quantity and quality of supply dominated the entire life of the cluster, and the manufacturing upgrading issues were left on the back burner. The technical task teams focussing on supply achieved some notable successes: improved knowledge flows, standardised measurement systems, greater wood recovery rates, utilisation of younger trees for certain products, suitable densities for different manufacturing applications, a better grading system. The co-operative ventures of the group had limited success with respect to upgrading challenges. However it generated substantial efficiency and upgrading information knowledge flows, improved supply chain efficiency from sawmills to
4
manufacturer, developed potentially new product opportunities but which were never fully realised, and contributed to upgrading technical abilities within firms. On the issue of the sawn log saligna timber supply that brought them together in the first place the network however achieved little success. The manufacturing component of the network was too small to guarantee a critical mass of continuous drawoff to ensure favoured supply status to the group’s members. The other initiative to solve the supply problem also revealed the network’s inherent weaknesses. The large state forests were due for privatisation and it was clear that much of this pristine mature timber would disappear into chipping, export and pulping activities. The small independent growers and sawmills were keen to bid for small lots and the one large corporate sawmiller that did not own much plantation land was concerned at not being able to gain access to this newly available timber resource. The external intermediaries used their political contacts with the relevant government department which was positive, but raised two issues. Unless the manufacturers could upgrade their products, would they be able to pay the price required to bid saligna away from chipping and pulping? Furthermore, could this network speak for the mass of manufacturers on a price increase? This created a paralysis. The group was too small to speak for the industry, and even within their own ranks, manufacturers were divided between higher value adding producers manufacturing high quality furniture able to pay more, and producers using the resource to make lower value added cheap garden furniture. However the biggest failure lay in the group’s inability to focus on co-operation between the manufacturing enterprises to ensure greater production efficiency. Hence the key challenges of process, product and functional upgrading were left largely untouched. Instead of trying to foster horizontal cooperation to tackle these issues and create collective efficiencies in performance, the manufacturers used the supply issue to avoid opening up the black box of firm inefficiency, treating it as purely a logistics value chain problem and not as a price and efficiency utilisation issue. Consequently they lobbied for cheap wood, railing against timber being sold at higher prices to the pulp and paper industry, and in the end focused on final products with little high quality design, hardly any value adding branding, and insufficient emphasis on finishing. Despite the initial successes the cluster collapsed when the firms were asked to pay membership fees, the external intermediaries resourcing the cluster withdrew, and the two key internal change agents amongst the manufacturers were bought out. The Auto Component Benchmarking Clubs 3 Relationships in the automotive sector in South Africa in the mid 1990s, like most industries (and groups of people for that matter), were characterised by fierce competition, little cooperation and lack of trust. The initial spark for a change of this environment and the creation of these continuous improvement clusters came from a variety of internal and external sources. The shift from import substituting industrialisation (ISI) to trade liberalisation, a major drop in tariff protection and rapid integration into the world economy in the 1990s, meant that the writing was clearly on the wall for the auto component sector. Either they became internationally competitive quickly or the assemblers would source most components internationally and they would go to the wall. The potential crisis facing them was being researched by university academics assisting industrial restructuring of selected industrial sectors. The research results, comparing local firms against global lean production performance indicators, were disseminated widely through numerous workshops and firm visits, and the quantitative data shocked a number of firms in the industry. It catalysed some
3
See Barnes and Morris 1999 and www.bmanalysts.com for information on the Benchmarking Clubs..
5
of the firms’ CEO’s to push the academics involved to play a completely different role – as neutral external intermediaries (i.e. business service providers) to form and operate a continuous improvement cluster. With partial government financial support and firm membership fees (on a 65:35 ratio) the KwaZulu-Natal Benchmarking Club was formally launched in January 1998. The KZN Benchmarking Club was founded with 12 member firms – 11 component companies and one large assembler – and an executive of 2 member firm representatives plus service provider representatives. The operational basis of the cluster was a continuous improvement, competitiveness driver model against which each firm’s performance was benchmarked to facilitate learning and upgrading. This is summarised in Figure 1:
Figure 1: The market driver matrix
Market drivers
1.Cost control 2.Quality
Operational performance measures
Inventory use (raw materials, work in progress, finished goods) Customer return rates, internal reject, rework and scrap rates, return rates to suppliers Time from customer order to delivery, delivery frequency of suppliers and supplier delivery reliability, delivery frequency to customers and delivery reliability Manufacturing throughput time, machine changeover times, batch sizes, inventory levels, production flow Literacy & numeracy levels, employee development/training, suggestion schemes, labour & management turnover, absenteeism rates, employee output R&D expenditure (process and product), Contribution of new products to total sales
Linked organisational practices
Single unit flow, quality at source, cellular production, kanbans Quality control structures, statistical process control, quality circles, team working, multi-skilling Business process engineering, cellular structures in order processing and dispatch, value chain relationships and supply chain management Production scheduling, JIT, single minute exchange of dies, multi tasking and multi-skilling, cellular production in manufacturing Continuous improvement (kaizen), work organisation, worker development and commitment programmes, industrial relations Concurrent engineering, R&D
3.Lead times
Value chain flexibility
4. Flexibility
Internal operational flexibility
5. Capacity to change
Human resource development
6. Innovation capacity
The Club was based on providing the following services to its members: • a comprehensive (strictly confidential) diagnostic report which: Ø measured the operational performance levels of each firm member according to the above market drivers and key performance indicators; Ø surveyed their ten major customers perception of the firm’s performance as well as ten major suppliers; • a confidential annual benchmark comparing each firm against a ‘like for like’ international competitor; • a monthly newsletter outlining aggregated benchmark findings; • quarterly workshops discussing generic findings, common problems and various solutions to competitiveness problems; • encouragement of information sharing through visits etc. However lack of trust, fear of sharing information, and old ISI mindsets looking to blame anyone else (e.g. government, assemblers, suppliers etc) for their problems, government inefficiencies and bureaucratic stupidity, meant that it took some time to take hold as a proper functioning cooperating cluster generating collective efficiencies. The key moment occurred when the firm members took ownership of the cluster – for example, when the quarterly workshops were shifted from a neutral (university) venue to one of the firm premises, and the 6
firm representatives (as opposed to the external intermediaries) gave lead presentations focussing on their own experience in sorting out various problems. From then on the previous reluctance to share information very rapidly faded, and the lack of trust dissipated. The success of the KZN Club led, within a couple of years, to two other sister Clubs being formed in the other heartlands of the South African automotive component industry – first in the Eastern Cape (Oct 1999) and later in Gauteng (mid 2001). These Clubs operate along the same operational and governing principles as the original KZN Club. What had started off as a small single regional cluster, by June 2003 had grown to three nationally linked clusters covering the entire country and comprising close to 60 automotive firms. Linked not only by a common service provider and a common information newsletter with minor changes for club specific ‘news’, but also because member firms often attended each others quarterly workshops thus spreading knowledge sharing between the Clubs. The success of these clusters can best be measured along four criteria: • increasing knowledge sharing, • significant learning and spontaneous firm visits by members; • major improvement in their operational performance as reflected in a variety of competitiveness indicators; • and finally, the spread of the Clubs as new members were attracted. The clusters are primarily based on firms wanting to ensure continuous improvement and operational performance enhancement in order to make them more internationally competitive through their membership. The best quantitative indicator of the impact of these three clusters on the member firms is the impact they have had on the competitiveness of the cluster members as measured by improvements in the operational performance of Club member firms. Although a wealth of information is available and is shared with firms on a regular basis, only the overall summary (table 1) of progress in process upgrading of Club member firms both i relation to their own improvements and relative to an international n sample of benchmarked firms is presented below.
Table 1: Learning and operational performance change of firms in clusters Critical Success Factors Key Performance Indicators South African Firms
Improvement 1998/9 2001 1998/99-01 62.6 42.0 32.8% 3270 1240 62.0% 4.9 3.9 20.7% 21989 18518 16.0% 19.9 17.9 9.9% 78.7 82.2 4.5% 92.2 92.7 0.6% 1.3 2.0 56.2% 4.4 1.64 4.0 2.12 9.4% 29.5% Western Europe 31.2 549 1.9 8319 16.8 92.2 96.1 1.3 4.2 1.83
4
Comparator Firms
Emerging economies 38.6 624 3.5 13213 12.0 92.3 93.5 3.1 5.7 2.90
Cost control Total inventory (days) Quality Customer return rate (ppm) Internal reject rate (%) Supplier return rate (ppm) Flexibility Lead time (days) Supplier on time (%) On time to customers (%) Capacity to Training spend as % total change remuneration Absenteeism (%) Innovation R&D expenditure (%) capacity
Source: Benchmarking and Manufacturing Analysts.
4
Time series data only exists for 32 South African based component firms (3-4 year period). Performance in 2001 is matched by a sample of 26 international firms, for which we do not have time-series data.
7
From the above table it is clear that the Benchmarking Clubs, as continuous improvement clusters, have had a major impact on the internal operational performance of the member firms. With the exception of delivery reliability to customers, progress for all of the measures in South Africa has been significant. Despite improvement, the South African components sector has in most respects some way to go before it reaches the global frontier, but given the relatively strong performance of the emerging country competitors relative to those in Western Europe, there is every prospect of the South African firms making up a good deal of this competitiveness gap. Generally speaking, South African component firm’s performance increase is better where internal factors (work- in-progress control, training, absenteeism) are involved than where they are dependent on external factors (raw material inventories, supplier performance). From a value chain perspective this suggests that the growth of learning is still predominantly in the first-tier components suppliers and has not yet diffused widely up the value chain. The relatively poorer performance in terms of inventory control is accounted for by logistic problems along the value chain, especially with regard to incoming materials (minimum-sized import quantities; problems at the ports) and distance to the export market (for stocks of finished goods). In general there are clear indications that significant process upgrading has occurred and a substantial movement towards international competitiveness amongst these firms is under way. Clearly a number of factors apart from the Club’s operations have played their part in this, but certainly from all accounts the horizontal cooperation, trust building, knowledge sharing and inter firm learning embedded in the operations of these three clusters have played a major and critical role in the process. The area where improvement has been sub-optimal has been between firms and their external environment – i.e. in areas where the Benchmarking Clubs have had little direct control and influence. This threw up the limits of horizontal cluster co-operation between manufacturing enterprises and forced firms to look at their vertical linkages as well as the broader institutional supporting environment. It is therefore not surprising that pressure built up for the clusters to tackle problems both along the value chain and in relation to the broader environment as the next arena of operation. The Durban Auto Cluster5 By 2001, although horizontal cluster cooperation, through the Benchmarking Clubs, aimed at raising the internal operational performance of auto firms in KwaZulu-Natal was well and truly embedded, other aspects of the industry threatened its international competitiveness. These pertained to issues that impacted on the operational performance of the firms but seemingly lay outside their sphere of direct control. They were par excellance value chain issues, either located in the external institutional environment or in the industry’s supply chain. Although firms had been raising these issues informally, the stimulus for the creation of a larger cluster with the mandate to tackle these bigger issues did not come from them. The initial stimulus came from the head of the economics department of the Durban Metropolitan Council. With Toyota SA securing a sizeable export order the local authority official, seeing the vibrant synergies between a leading assembler striving to export, an already established cluster of auto component firms, and the growth potential for the region, provided the financial resources and institutional backing over a six month period to set up a new auto cluster encompassing vertical and horizontal cooperation, with a mandate to tackle the key issues blocking the reaping of collective efficienc ies.
5
See www.dbnautocluster.org.za for more information.
8
The already established trust networks comprising the KZN Benchmarking Club were used to initiate this new cluster. External intermediaries from the Benchmarking Club service provider with the trust of the firms acted as facilitators. The method utilised to set it up over a six month period from June to December 2001 was the following: • • • • • • • • all forty auto firms in the local value chain were identified, visited, appraised of the initiative, and invited to participate, with no financial obligation, in the set up process; a number of workshops involving all the major actors in the local value chain were held; international research on cluster success was undertaken to provide the required knowledge and information to participants, local, national and provincial government was made visibly present at these workshops so as to provide political legitimacy; a local needs survey was undertaken on the key issues affecting the local industry; participants voted to identify the four key focus issues – these were logistics, human resource development, supply chain development, and operational competitiveness; technical task teams comprising firm level expert representatives were set up to manage these focus areas identifying the issues to concentrate on, as well as writing a clear one year business plan with program goals, activities, and designated budgets; a workshop in November 2001 discussed these business plans, a firm based governance structure, and an overall budget providing for government support as well as firm membership fees, and then firms were required to vote on whether the cluster should be formally launched, and whether they would commit to sign up as fee paying members.
The Durban Auto Cluster as a public-private initiative was formally launched in Jan 2002,6 run by an executive committee comprising representatives from the firms, the government sections providing funds, and the two facilitators from the service provider. It encompasses four programmes with clear business plans, run by four technical steering committees, with a firm representative as chair and the rest seconded by their firms plus a designated service provider member as technical support, with control over designated program activities and responsibility for expenditure on its bud get line items. These are Supplier Development, Human Resource Development, Logistics, and Operational Competitiveness 7 . The cluster operates in terms of providing key services to members, financial saving through joint activities, knowledge sharing through workshops and newsletter, joint research disseminated in a user friendly manner, access to an on line data base. The value chain vertical linkages play a crucial role in ensuring horizontal cooperation reaps collective efficiencies. In this respect, the 1st tier suppliers have played a critical role using their resources, industry muscle and knowledge to provide assistance to the 2nd and 3rd tier members of the cluster in the operations of the programmes. Logistics: The driving force in this programme is the chairman who comes from a key 1 st tier supplier and who has used his extensive knowledge and resource base to maximum effect. The broad objective of the logistics programme is to share information, create price transparency, secure fair pricing and reduce costs through consolidation and load balancing across the three forms of logistics - sea, road and air. Given estimates that logistics accounts for between 5-15% of selling costs, coupled with the rapidly increased growth in import and export trade, this has become a critical factor affecting competitiveness. A survey of member firms logistics costs showed that sea freight charges for some smaller firms were more than
6 7
The DAC currently has 28 participating firms paying membership fees. This programme will not be discussed here since it is simply the incorporation and expansion with new members of the KZN Benchmarking Club, which we have already discussed, into the DAC.
9
double those of the larger firms. Since these 2nd and 3rd tier suppliers were also feeding into the larger firms, this had a competitiveness impact all along the value chain. Hence a key strategic initiative of the logistics programme has been to radically cut freight costs. Thus far, apart from work in progress, the following has been achieved. Firstly, benchmarked freight rates have been distributed to member firms so they can see how they relate to average prices. Secondly a ‘cluster freight rate’ has been negotiated for the Europe-SA route for all cluster participants that wish to participate 8 - $1300 per 20 ft equivalent container as compared to the $2500 some smaller firms were paying. Thirdly, the variety of logistics services associated with freight movement have been benchmarked and made available to the firms so that they can negotiate better individual prices. The fundamental aim of these processes is not to form a cartel against the freight service providers and drive minimum charges down, but to cut costs for the smaller firms so that these impact on the systemic competitiveness of the value chain. Finally, a programme is currently being initiated to coordinate road and air freight movements into and out of the regional automotive industry in order to cut costs on this front. Supplier Development: The activities of this programme focus on four areas – raising purchasing skills, aligning purchasing functions, developing a supplier awareness database, and information sharing – with respective workgroups set up to tackle these issues. Key interventions in skills development have been identifying skill and training requirements, increasing knowledge sharing and holding training workshops on best practice purchasing. The workgroup on aligning purchasing functions is attempting to bring about synergy in the ways in which suppliers are dealt with by the OEMs and larger 1st tier firms. A generic supplier development template has been set up to create commonality in the industry. This is linked to a supplier evaluation form and checklist to monitor supplier performance, as well as a common costing tool. Pilot projects in selected firms are scheduled to be launched soon. A comprehensive database of all suppliers is currently being developed for use by firms seeking suppliers and vice versa. Information sharing has been aimed at creating enhanced knowledge of the dynamics at play in the supply chain, publicise the demands and supplier requirements of the assemblers to other tiers in the supply chain, educate the lower tiers through workshops on how to meet these challenges, and thereby stimulate a greater level of vertical knowledge integration in the value chain. Finally the programme has also embarked on an ambitious plan to market the capabilities of suppliers to international firms within the global value chain. Human Resource Development: The broad aim of this programme is to raise operational capabilities. It has thus far engaged in the following activities. A forum has been set up to interface with the authority structure responsible for training – the Manufacturing, Engineering and Related Services Sector Education Training Authority (MERSETA). Since most firms are either unaware of grants for training or are unable to follow the administrative procedures to successfully access their claims, a key objective is to assist firms in all administrative, bureaucratic dealings with the MERSETA. This interface is also enabling firms to involve workers in a ‘learnership’ scheme – an on and off the job qualification programme – as well as learning how to claim tax deductions for every learnership registered. Finally this programme is coordinating a joint pilot adult basic education training programme for a number of firms in conjunction with a local training institute. POLICY IMPLICATIONS AND GENERALISABLE LESSONS The five case studies clearly demo nstrate that, even under historically inherited low trust environments, purposive action can create horizontal and value chain co-operation. Secondly
8
At least one of the larger firms with huge volumes pay even less than the negotiated cluster rate.
10
they demonstrate the interaction and mutually reinforcing relationship between value chain cooperative linkages and more classic collective efficiency gains from horizontal cluster cooperative activities. Purposive action has played an important role in facilitating, or ‘driving’ through, the transfer of knowledge down value chains – this is especially the case in terms of the role of 1st tier firms in the DAC - and ensuring learning gains between the various firms comprising the clusters. Empirically this is quite clear and there is clearly a lesson here. If clusters embodying vertical value chain and horizontal cooperation can be created and sustained in such a historically inherited, conflict ridden, low trust environment as South Africa then surely it must be possible to replicate them elsewhere? This is the analytic and policy problem - how to generalise from these case studies? The policy challenge is to go beyond the specific conditions and forces at play in each of these clusters so as to provide generic lessons that can be applied to other situations. The following factors seem to have a crucial role in unpacking the analytic and policy issues at stake:
q
External crisis is critical in focusing firms to look beyond narrow sectarian interests towards the larger picture promoting cluster and value chain co-operation. However, firms are risk averse, and cooperation is risky, particularly in low-trust environments. Crisis has therefore also to contain viable market opportunities, for firms to overcome their naturally risk averse tendencies in order to be open to cooperation. Many firms recognised that without doing things differently they were going to go out of business. However, they also realised that a go it alone strategy is unlikely to work and that more can be gained from working and learning together. Trust is critical for the functioning of a cluster both in terms of setting it up and maintaining its continuation. Trust can be created even in societies with high levels of antagonism. South Africa is riddled with mistrust, and cynicism about the benefits of cooperation is rife. If trust can be created here, there is a role for purposive action in more conducive social environments. However trust takes time to be established and one has to proceed at the pace of the most cautious. In the end trust is established through concrete activities demonstrating the clear benefits of cooperation. The best example comes from the auto firms in the benchmarking clubs. The first few meetings were characterised by extreme fearfulness by some firms of even discussing particular issues, especially those who perceived themselves to be direct competitors. Hence the more forward looking firms had to be cautioned by the external intermediaries, restrained from pushing these firms beyond their limits, and hence the cluster proceeded at the pace of the slowest. Yet after a year these same firms were holding meetings in each others firm premises, taking everyone along on a tour of their production operations, sharing information about their best practice operations, and encouraging firm level ‘learning by visiting’. Neutral external intermediaries with a real knowledge of the industry can act as facilitators overcoming internal conflicts, jealousies and mistrust. Acting as neutral brokers, mediating cooperation and drawing together disparate interests is critical. The importance of maintaining neutrality vis a vis other members of the value chain and cluster cannot be overemphasized. The external intermediaries have to be seen to be pushing for the common good, and hence need to restrict their activities to areas where there is real observable potential for achieving common collective good. In particular they should never be seen to be acting on behalf of, favouring, or being the agent of, the dominant lead firms, the assemblers, other governors of the value chain, or indeed any
q
q
11
individual participants. This can be quite difficult, since these facilitators also have to ensure that the drivers of the value chain remain on board. For example, there was always a great temptation to use the muscle of the local auto assemblers to persuade firms in their supply chain to join the Benchmarking Clubs. However, if the component firms in these clusters had ever felt that the assemblers’ agendas were being pushed, it would have been the death knell for trust and cooperation.
q
Dominant lead firms exercising a governance, or coordinating role, within a specific value chain are important in creating and sustaining cooperation. These firms may not play the leading change agent role, but without their sanction, legitimisation and cooperation it is very difficult to sustain vertical cooperation. The major auto assembler in the KwaZulu-Natal region, Toyota, has never played a dominant role in the Benchmarking Club nor the Durban Auto Cluster. Indeed it is often a passive and even absent participant. However everyone knows that it wholeheartedly supports both these initiatives and this has given massive legitimacy to both clusters. In the same way, without the active participation of the sawmills, that controlled key aspects of the furniture value chain, the Saligna cluster could never have operated, and would not have recorded any successes. Internal change agents, willing and able to play a catalytic function propagating cooperation, are essential. However the actors playing this role shift and change over the course of time, and recognising shifts in position and place of these internal change agents is important as the process unfolds and new actors come to the fore. Internal change agents do not necessarily come from those exercising governance within the value chain. Both examples of the furniture and automotive clusters demonstrate this. However when key coordinators of the value chain are brought on board as critical change agents then their substantial resources, experience and muscle can have a fundamental impact on making collective gains for the cluster. The best example of this is the role of the major 1st tier supplier as the chair of the logistics technical steering committee in DAC. It is no exaggeration to say that he has single-handedly been responsible for negotiating the massive change in the freight rate costing regime for all cluster participants. Inherited mindsets have to be broken. Firms have histories and cultures. They are located in particular path dependencies, which condition the mindsets of their managers and workers. They do not always understand why, and more importantly, what they need to learn. They have a natural tendency to blame operating difficulties on outside pressures and seek solutions in external forces. This is particularly an issue when emerging from an import substituting industrialisation trajectory where the local market was everything and demand side measures ensured economic success. Government becomes both the problem and the solution, if only the right people can be lobbied. This can be used to advantage though. Firms often initially participated in the various cluster creating activities in the automotive and furniture sectors because of the presence of highly respected, internationally renowned, external intermediaries with esteemed academic status, since the firms assumed that they must be the repositories of easy solutions. A ‘critical mass’ of relevant firms at key levels of the value chain is essential for successful horizontal cooperation. Relevance and criticality may be defined by position in the value chain, size of firm, sheer weight of numbers or simply interest in finding solutions to a particular problem. However, unless enough stakeholders are involved to affect change, the network is likely to remain abstract. A major weakness of the Saligna
q
q
q
12
cluster was that there were simply too few and irrelevant furniture manufacturers involved. Their vociferousness did not compensate for their insufficient industrial clout and this impacted deleteriously on the long term sustainability of this value chain cluster. Conversely, the substantial support in the auto sector by the component manufacturers firms, both in terms of criticality (the 1 st tier suppliers and two key assemblers in KwaZulu-Natal and the Eastern Cape) and in terms of sheer weight of numbers from 2nd and 3rd tiers, means that these have become self reproducing and recruiting clusters.
q
Government’s facilitative role through providing financial and institutional support is very important, particularly at the initial stages of the cluster when firms are still hesitant about the benefits that will accrue from their participation. Knowing that the cluster has government backing plays an important psychological and legal role at the start of a cluster initiative. They are more likely to chip in their own money and pay membership fees if they also know that government is participating on some matching grant basis. Furthermore accessing government funding imposes its own discipline, for governmental financing requirements comes with the need to have a clear legal framework, executive governance structures, business plans, and accountability. However receiving government support is not without its own built in difficulties. Government also has its own sectarian interests, and the cluster has to be guarded from being diverted from its main course into fulfilling tangential political agendas. Although developing country government’s are rapidly changing their rhetoric from an import substituting industrialisation framework towards export orientated, liberalised more open economies, this does not mean that their changed policies can be backed up by meaningful implementation strategies. In our auto cluster cases, notwithstanding very impressive new government rhetoric and policies, the inherited bureaucratic structure of government remained generally intact. There was insufficient re-conceptualisation of the deployment implications of these new policies, of retraining departmental staff from being “paper pushers” to “change agents” fully attuned to the competitiveness demands being placed on South African firms, and hence reconfiguring institutional arrangements to ensure that implementation followed the policy shift. As such, the new policy support measures that replaced the previous protective regime were immersed in bureaucratic red tape – to the extent that it was extremely difficult for the clusters to access them. Using the language of lean production we would argue that the government failed to reconfigure its internal operations in line with new market demands. The long lead times associated with its previous operations were no longer adequate in the face of the international competitiveness demands being placed on the firms relying on some level of government support. The over emphasis of bureaucratic reporting structures used up valuable skills, time and resources of the service providers and created massive cash flow problems. For the problem with tying up a cluster in too much red tape is that it is all too easy to push it into the financial red. Collectively accessing the ear of government is an important incentive in generating and sustaining interest in value chain cooperation. The interest in, and hope of successfully, accessing government seems sustainable even when government is not wholeheartedly reciprocal. As long as the lines of communication to government are open, this seems to be viewed with some significance. Our experience in all five clusters is that member firms carry an infinite hope, despite little being actually realised, that government will be more accessible, as a result of their support for the cluster, the connections that the facilitators may have to particular government officials, and the financial resources from different levels of government. The truth is that, in our experience, government listened very infrequently to industry and even when it did seemed incapable of acting
q
13
appropriately and urgently. Hence, the ability to resolve problems remained locked within industry, which is the principal reason why co-operative initiatives (whether they be horizontal or vertical) are so important in developing countries, despite the difficulty of doing so in these environments. Without sounding cynical, and as long as other collective benefits accrue to cluster members, this incentive of accessing the ear of government, whether realised or not, is a powerful tool in maintaining the sustainability of a cluster, and it is not in the interests of the cluster to disabuse the members of their hopes. The pretence that government has the capacity to be involved is useful to maintain, but the reality is that not too much time should be devoted to working with government.
q
Practically oriented activities focusing on selective incentives and definite benefits with scaleable and achievable targets are crucial in creating the sustainability of such cluster cooperation. Building value chain systemic efficiency and firm level competitiveness occurs through a series of linked improvements, rathe r than a single big jump. Hence reaping tangible relatively immediate benefits quickly creates credibility and breaks down barriers to trust. Without sustained real improvements, the incentive to maintain ongoing participation declines rapidly. Ultimately the cluster has to become seen as the place to draw expertise from, as the centre of excellence, as a place to provide real services aligned with and larger than the particular firms resource base. A key moment occurs when individual firms start to incorporate the cluster into their strategic planning, in the sense that they stop allocating their own resources separately from the cluster, but instead start to integrate these with the cluster activities in order to reap additional benefits that would not be available from their own individual activities. 9 The logistics programme in the DAC is an excellent example of this, as are some of its collective procurement activities and shared human resource development training programmes. Once this moment of alignment occurs, as long as the cluster carries on delivering services, it is guaranteed the continued support of the participating firms. For the business plan of the individual firm in respect of these activities becomes inseparable from the business plan of the cluster programme. Involvement in the cluster is not something different from what the firm is originally always doing. In this respect the Benchmarking Clubs were constructed on a much weaker and more difficult foundation. For buying in a new service such as benchmarking requires a different kind of commitment and is sustained on a different basis. Once benchmarking becomes regarded as actually an essential part of a firm’s business activities, then the cost of disengagement from the cluster escalates. Information flows and knowledge sharing create and solidify co-operation. This goes beyond simply lowering the transaction costs of information for any particular member of the cluster or value chain, and often actually creates information flows where none had previously existed. Firms also need to be made constantly aware that the cluster has a particular agenda and that all of its activities are geared towards a specific objective. This is why the monthly auto newsletters are so important. Whilst the firm- level benchmark undertakings are the key to the success of the Benchmarking Clubs, they only take place on an annualised basis at the firms. They may therefore have been the “big hit” that tied the firms into the Club and may have been the criterion by which firms initially evaluated whether they should join, but they are too dispersed to generate a sense of collective ownership of the clusters’ objectives. The newsletter through its monthly dissemination to
q
9
This can be translated into a simple formula: CBA = +1, i.e. any firm engaging in a cost benefit analysis must gain +1 from their normal activities as a result of participating in a cluster, otherwise there is no point in their membership. The key to creating a successful cluster is to realise that firms don’t participate from an altruistic view and hence the facilitators have to produce as many +1s as possible.
14
the management of each of the Club members ensures that firms are constantly locked into the cluster as a learning network. This is then reinforced through the quarterly workshops where meaningful firm- level debates and firm-to- firm networking takes place. As the firms make more direct contact with one another, the interface between them becomes less of a mediated activity and the information and knowledge flows through the cluster more directly. Our experience is that the firms have a wealth of information and knowledge, which can be harnessed and utilised as a resource for the benefit of other cluster members. The 1st tier firms in the DAC play such a role in regard to its different programmes, and in so doing serve their own particular interests, for it is in effect engaging in another form of supply chain management and learning. The key moment when the Benchmarking Club moved from being and incipient learning network to a mutually reinforcing learning network was when the flows of information ceased to be primarily an information stream between external intermediary and the firms and became a reciprocal knowledge sharing stream between the cluster firms themselves.
q
Quantifiable assessment is a crucial part of knowledge sharing. Especially in a continuous improvement cluster, one has to be able to quantify the extent to which firms need to change and where the changes have to occur. Management (especially those who feel fingers may be pointed at them) often find this difficult to accept. Having hard measurement tools and real data makes it possible for the facilitators/service provider to maintain neutrality, and make the necessary point without seeming to be attacking those involved. It also allows the management to identify those areas, which have to be targeted and to plan strategies. Such quantifiable data is very important in also projecting an approach that the cluster is there for the common good. If facilitators don’t have hard data it is much easier to be persuaded by sectional interests regarding the magnitude of particular problems or solutions and hence find themselves acting in a biased manner. Quantifiable assessment through internal surveys of cluster members problems, situation, needs, views, etc also plays a role in presenting the common perspective of the cluster. It allows members to situate themselves relative to the worst, average, and best performers, or to place their own needs etc in perspective. In the final analysis, knowledge sharing is the oil that maintains trust and allows collective strategies to be marshalled and deployed. Hence the importance of the credibility of the information, and the trust, that firm members have in it. Research that is well grounded academically but which can be accessed and understood by cluster members therefore has a very important role to play in achieving the diverse objective of the clusters. 10 Financial obligations and membership fees are extremely important for locking firms into the cluster. This plays an important role in stopping the cluster degenerating into a talk shop, a place of moaning, or a fantasy world of unrealisable expectations. If the members are paying money for concrete services then they will rightly demand tangible forms of benefit are delivered. In the auto clusters the monetary fee that members pay on a monthly basis is no t done out of good will or altruistic intentions. The firms want material gain from the relationship. The monthly newsletters, quarterly workshops, firm- level reports, presentations, and other collective services are all consequently very important as part of a set of material benefits that secure and maintain firm- level buy-in.
q
10
There are also major academic and policy spin offs for the facilitators and external intermediaries involved in providing such research information. The five clusters have provided the most extraordinary research laboratory and rich data for the academics cum facilitators cum service providers involved with them.
15
q
Blueprints do not exist, and external intermediaries, facilitators and /service providers need to be aware that there is no set template for running a network of firms. There may be basic do’s and don’ts, such as ensuring firm confidentiality and providing a quality service to members, but it is clear from our experience that networks evolve out of the particular sets of experiences of the various constituent parts. Service providers therefore need to be flexible in terms of adapting to this evolutionary process. Given issues of path dependency and the fact that trust levels may be low in liberalising economies, service providers need to be sensitive to firm apprehension regarding the sharing of information. These are not issues that can be attacked with a sledgehammer. They need to be treated sensitively and chipped away at with an ice-pick, whilst at the same time ensuring that the broader strategic objectives of the network are rigidly adhered to. Mobilising and deploying resources is very important in setting up and maintaining the sustainability of the cluster. Once the cluster starts running it becomes very complex and requires dedicated resources. The energy and resources required for the simple administration and co-ordination tasks to establish and maintain sustainable co-operation cannot be underestimated, and usually this requires transforming external intermediaries into some form of formal fee receiving service provider. The danger here is that the service providers involved may simply be adding the cluster to their portfolio of other income generating activities and lose focus. The cluster is a collectivity and, like the member firms, the service provider has to differentiate (and often subordinate) its own interests to those of the common good of the cluster. Fundamentally the service provider is there to deliver cluster level services, not to favour its own consulting services or those of another firm. This is the reason why the service provider to the Benchmarking Clubs only provides diagnoses of problems and not consulting services to rectify them. For, once the service provider starts to do that, it is open to being accused of punting its own interests, and its precious neutrality goes out of the window. Ownership and a firm based governance structure is incredibly important in sustaining a cluster. It is critical that firms take ownership of the cluster at some stage and the intermediary agent needs to step away from being the central focus of the network. This is often a turning point, a critical moment when the network becomes a truly self-contained learning one, with the intermediary agent a mere facilitator. Buy- in is absolutely crucial for setting up and maintaining a cluster, and one cannot underestimate the long term benefits that are secured through the involvement of firms in practical activities, pilot projects, technical task teams, and technical steering committees. Ultimately these form the foundation for transference of real control and power to a governance structure rooted in the firms themselves. However one should also avoid the dangers of populist idealism. Enormous power is always concentrated in the facilitators/service providers, and necessarily so, for they literally make the cluster run on a day-to-day basis. They do things that no individual firm is able or indeed willing to do. But they have to do so within the tight regimen of a business plan and clear definition of focused activities. In short the business plans and defined activities keep the firms focused on the ball and the facilitators honest. The trick is thus to create a governance structure that locates and transfers real power to the executive structure and programme steering committees and in so doing maintains an appropriate balance between these two power centres.
q
q
In conclusion, the analytic strength of these case studies, and the lessons they have generated must lie in whether these generic factors that have been identified explain the creation, sustainability, and in the one case the eventual demise, of the South African clusters under
16
review. Externally induced crisis coupled with opportunity, as well as the purposive action of external intermediaries who were able to bring to bear the influence of government played a critical role and provided the necessary conditions. Key internal forces, through buy-in of lead firms, as well as the role of internal change agents driving the process, provided the sufficient conditions for creating the various cluster. The stress on practical activities, providing a clearly defined role for firm level actors, through various technical working groups and steering committees, combined with defined, realisable and measurable outcomes to meet stated needs, and a constant diffusion of information and knowledge sharing, maintained support for these clusters. The real analytic and policy test however lies in whether these experiences can be replicated in other countries, conditions and situations. In this respect the list of lessons arising from these case studies needs to be organised into a framework of steps, core processes and questions which can be generalised to other network situations. Its is here that the five case studies yield their most powerful analytic and policy insights. For they allow us to reinforce and add to the ‘template’ of core process and underlying questions for the design and operation of learning networks, suggested on the basis of other cases (Bessant et al 2003).
Figure 2: Core processes in inter-organisational/firm networking and learning
Process
Network creation Decision-making/governance
Underlying questions
How the membership of the network is defined and maintained; how the external stimulus is identified How (where , when, who, etc.) executive structures are set up and decisions get taken How to harness and limit the role of external intermediaries, internal change agents and government How (and if) conflicts are resolved How (who and in what form) tangible benefits are delivered and lead firm services/resources are utilized How information flows and is managed How knowledge is articulated, captured and made available for the whole network How members are motivated to join/ remain in the network – e.g. through active facilitation, shared concerns for development, fee payment, mobilizing of selfish concerns and benefits etc. How the risks and benefits are shared How relationships are built and maintained between individual representatives in the network (adapted from Bessant et al 2003)
Exercise of leadership
Conflict resolution Deployment of resources and benefits accruing Information processing Knowledge capture
Motivation/ commitment
Risk/benefit sharing Integration
17
Refere nces:
Amin A. (1999), ‘An Institutionalist Perspective on Regional Economic Development’, International Journal of Urban & Regional Research, 23, 2 Bair J., and Gereffi G (2001) ‘Local Clusters in Global Chains: The Causes and Consequences of Export Dynamism in Torreon’s Blue Jeans Industry’ World Development 29, 11 Barnes J and Morris M. (1999), ‘Improving Operational Competitiveness Through Firm-Level Clustering: A Case Study of the Kwazulu-Natal Benchmarking Club’, Working Paper 24, School of Development Studies, Durban: University of Natal. Bessant J and Francis D. (1999) ‘Using Networks To Help Improve Manufacturing Competitiveness’, CENTRIM. Bessant J. and Tsekouras G. (2001) ‘Developing Learning Networks’. A.I. and Society, 15, 2 Bessant J., Barnes J., Kaplinsky R., Morris M. (2003), ‘Building and Sustaining Learning Networks’, presented at Euroma/POMS conference: Operations management research and practice - one world?, Lake Como, Italy, 15-18 June Cleveland J. (1995) ‘West Michigan Manufacturers Council: A Case Study in Interfirm Collaboration’ and ‘The Northeast Indiana TQU Network: Quality through Collaboration’ in Common Purpose, Common Sense: Case Studies in Interfirm Collaboration. USNet, Regional Technology Strategies Dolan C and Humphrey, J (2000) ‘Governance and Trade in Fresh Vegetables: The Impact of UK Supermarkets on the African Horticulture Industry’, Journal of Development Studies, 37,2. Doner R. F. and B. Schneider (2000) ‘Business Associations and Economic Development’, Business and Politics , December. Gereffi G. (1999) ‘International Trade and Industrial Upgrading in the Apparel Commodity Chain’, Jopurnal of International Economics, 48. Humphrey J & Schmitz, H (1998) ‘Trust and Inter-Firm Relations in Developing and Transition Economies’ Journal of Development Studies, 34, 4. Humphrey J. and Schmitz, H. (2002). ‘How does Insertion in Global Value Chains Affect Upgrading in Industrial Clusters’, mimeo. Kaplinsky R., and Morris, M.L. (2001). ‘A Handbook for Value Chain Research’, Institute of Development Studies, University of Sussex and School of Development Studies, University of Natal, (www.ids.ac.uk/global, and www.nu.ac.za/csds) Kuper K (1997) ‘“Cooperating to Compete”: Soft and Hard Business Networks’ Trade and Industry Monitor, Vol. 1 – April Lundequist P. and Power D. (2002) ‘Putting Porter into Practice? Practices of regional cluster building: evidence from Sweden’, European Planning Studies, 10 , 6. McCormick D (1999) ‘African Enterprise Clusters and Industrialization: Theory and Reality’ World Development, 27, 9 Meyer-Stamer J. (2003) ‘Obstacles to Co-operation in Clusters and How to Overcome Them’, Development Alternatives, 9,1. Morris M (2001) ‘Creating Value-Chain Cooperation’ IDS Special Bulletin: The value of Value Chains, 32, 3 Nadvi K. (1997). ‘ The Cutting Edge: Collective Efficiency and International Competitiveness in Pakistan’, IDS Discussion Paper 360. Brighton: IDS. Pyke F (1992) Industrial Development through Small-Firm Cooperation: Theory and Practise, Geneva: International Labour Office Rosenfeld S. (1996) Overachievers: Business Clusters that Work, Prospects for Regional Development, North Carolina: Regional Technology Strategies Schmitz H. (1995). Collective efficiency: Growth path for small firms, Journal of Development Studies, 31, 4. Schmitz H. (1997) ‘Collective efficiency and increasing returns’, Working Paper No 50. Institute of Development Studies, University of Sussex.
18