Location, network economies and supply chains

The new business practices that became widespread during the late 1980s and 1990s contain two, seemingly contradictory, elements. First, as transactions and communications costs have come down, companies have become increasingly focused on core businesses while outsourcing the production of non-core inputs and selling off non-core businesses. This amounts to increased specialization or division of labor in the economy. Second, as competition has become fiercer and profit margins have been squeezed, transaction and coordination costs have become the target for cost cutting efforts.

One way of economizing on transaction costs, is to reduce the number of transactions. There is thus a trade-off between benefiting from increased specialization, which inevitably increases the number of transactions on the one hand, and economizing on transaction costs on the other hand. One solution to this trade-off is the development of closely knit supply chains. Production is highly specialized, but each supplier of inputs enters a long-term relationship with the downstream customer and produces made-to-order or standard inputs delivered at the company gate at a high frequency. Another option is to further reduce transaction costs through continuous innovation and diffusion of information technology and thereby allow for a larger number of transactions and looser relations to suppliers while keeping down costs. The least attractive option is to lower the price of inputs that can not be supplied according to just-in-time delivery systems in order to compensate the downstream customer for additional storage facilities and inventory management. Finally, if quality control is poor on the part of the supplier, the downstream customer will probably have to be compensated for additional quality control as well. In some industries these problems may constitute a prohibitive barrier to trade. Nevertheless, such barriers may also represent business opportunities for local well-educated entrepreneurs providing logistics services, laboratory testing and other quality control services.

Location theory

According to classical location theory, the larger the distance from a center a producer is located, the lower the price he receives on sales to the center, and the higher the price he pays for his purchases from businesses located at the center. Consequently, remote places tend to have lower income per capita than more central places. A study by Venables and Liamo (1999) analyzes the impact of reduction in transport costs in a context where sectors differ according to transport intensity and factor intensity. They argue that diminishing transport costs is equivalent to moving all locations closer to the center, and thus the level of income tends to increase everywhere. However, as transport costs come down, relative prices of transport-intensive goods decline and regions producing the most transport-intensive goods experience a deterioration in terms of trade. The net effect may well be that regions close to the center experience a welfare loss as a result of globalization, while more remote regions gain.

Research on location observes that companies tend to locate close to each other. The theory seeks to explain why this is so, and derive patterns of trade and investment as a function of technology and transport- and communication costs. When transport costs are significant, producers would tend to locate close to their major markets. Economies of scale strengthen this tendency, while ready markets for differentiated intermediate inputs and a multi-skilled workforce also induce firms to cluster in central locations. However, as more companies locate in a region, problems of congestion appear, and cost related to congestion may outweigh the benefits of being close to other businesses and customers. The net advantage of locating in the center may eventually be smaller than moving to a more remote location.

An empirical study by Ke and Luger (1996) finds that such a point has been reached in the computer, telecommunication equipment and electronic components industries in the US. Excessive demand for certain skill categories, office space and housing, and traffic jams outweigh the benefits of a common pool of specialized skills, infrastructure and sources of inputs and the tendency is geographical decentralization of production. Furthermore, the proliferation of the Internet has made it possible to trade skills without actually moving persons physically. Skills can in some cases even be sourced internationally. Developing countries involved in such trade, for example India, have experienced an improvement in returns to investment in higher education partly as a result of engaging in trade in human capital intensive services (Pantagariya 2000).

A recent paper by Quah (2000) shows that even when transport costs are totally eliminated, clustering may still take place if productivity spillovers are time-sensitive. Spillovers will then take place between locations in the same time zone, among businesses having the same office hours. His analysis is most relevant for information- and human capital intensive industries such as finance, computer software development and other services that can be traded over the Internet. Close links between European and emerging African companies should therefore be expected in these industries.

Network economies

Network economies are present if the value of a product to each individual or firm increases with the network size. Although closely related, network effects are different from economies of scale. In telecommunications for example, economies of scale stem from substantial fixed costs related to investment in transmission lines and other infrastructure. The more customers and the larger the sales of telecommunication services, the lower the average cost of each minute of telephone conversation. The network externality is of a somewhat different nature. When the number of people who makes and receives calls increases, each individual can communicate with more people. Each new subscriber confers a benefit to all other users, even if the cost of a minute of conversation stays the same.

The impact of network effects on technology adoption, pricing and other economic activities have been studied extensively. A general observation is that technologies subject to strong network effects exhibit long lead times followed by explosive growth. It is also commonly found that new technologies with network properties are adopted earlier the larger is the sales volume of the provider (Saloner and Shepard 1992). The cost of adopting the new technology declines over time due to learning effects and incremental improvements of the technology on the supply side. At the same time the benefit from adopting the technology increases over time due to learning effects and new applications on the consumer side. These benefits are bigger in a large network, which is why large companies are able to introduce new technology earlier than smaller companies.

A recent development is that networks such as railway lines, gas pipelines, electricity transmission and distribution lines and broadcasting networks may serve as channels for transmission of information and communication services. For some networks little additional investment is necessary, while for others new cables, fiber-optic or otherwise, need to be rolled out. In developing countries such synergies could create a critical mass of possible Internet users earlier than what would otherwise be possible. It should however be noted that the technology is still in its infancy and transmission of information over alternative networks is still inferior to telecommunications. Nevertheless, simultaneous liberalization and privatization of the electricity and telecommunications sectors in many developing countries has opened new opportunities for reaping the benefits of synergies. Furthermore, it has been argued that energy services, including electricity, should be included as a separate sector in the GATS in a similar way as telecommunications, among other things in order to reap the benefits of synergies between electricity and telecommunications.

A recent study on telecommunications infrastructure and economic development (Røller and Waverman 2000) finds that there is a relatively strong causal relation between telecommunication infrastructure and economic growth. They find evidence that the causal link runs both ways – e.g., telecommunications infrastructure induces growth, which in turn creates demand for further telecommunication infrastructure developments. The study does, however, find evidence of a threshold or critical mass phenomenon. In fact, countries with universal service achieve a twice as high growth effect of additional telecommunication investment than countries with a lower penetration rate according to the study. Since telecommunications account for only a small share of GDP, it is not employment and investment in the telecommunications sector per see that create growth. Rather it is the quantity, diversity and quality of information and communication services provided over the networks, improving productivity in all sectors of the economy, that accounts for the acceleration in growth rates as soon as universal provision has been reached. If the study’s conclusions are correct, this indeed poses a dilemma for developing countries. Marginal investments in telecommunications infrastructure may have little impact on growth, while providing universal services may be far beyond the means of the countries in question.

The data set applied by Røller and Waverman (2000) only covers the period up to 1990, e.g., before the Internet became widespread. There is therefore the possibility that a broader range of services provided over the Internet and synergies between different networks contribute to obtaining a critical mass at an earlier stage of telecommunication penetration than before. This is, however, a hypothesis that needs to be further investigated before any conclusions can be made.

A final point worth mentioning is that rapid diffusion of telecommunications and electricity in the developed world about a century ago created a tremendous growth spurt partly as a result of network economies. Most SADC countries have not even reached the stage of industrial development that follows from electrification and traditional analog telecommunications. They now face the possibility of experiencing two technological revolutions during a short period of time: i) electrification and access to telecommunication and ii) the diffusion of the Internet. Under what circumstances this will actually materialize and lead to a rapid growth spurt, and under which conditions it will lead to widening digital- and income gaps, or neither, is an important research area where not much is known as of yet.

Supply chains

The supply chain encompasses all activities associated with the flow and transformation of goods from the raw material stage, through to the end user, as well as the associated information flows. Materials and information flow both up and down the supply chain. Supply chain management is then defined as the science of optimizing a company’s methods of manufacturing, storing and shipping the products it sells. 1Definition provided by the Supply Chain Management Research Centre http://www.cio.com/forums/scm/ Partly as a result of the diffusion of Japanese management techniques, supply chain management also increasingly involves contractual relations with suppliers and subcontractors. The transition from "just-in-case" inventory management to "just-in-time" supply has also reduced the average size of orders significantly. Combined with demand for customized products this has increased the need for cooperation and coordination, both in order to synchronize production and in ensuring quality control at source.

Traditionally, transportation is a high cost component in a supply chain, accounting for approximately 25% of the overall operating costs. 2Lancioni et al. (2000). Other logistics costs such as inventory in pipeline and safety stock costs add to this, and explain why cost reducing process innovation has increasingly focused on logistics. 3The Council of Logistics Management has adopted the following definition of logistics:

As is well known, any chain is as strong as its weakest link. Failure to deliver the required quantity and quality on time on the part of one supplier in the chain has disastrous consequences for all other supply chain members. There are at least two ways of preventing such disasters. The first is extensive screening combined with incentive schemes in order to ensure reliability, competence, and loyalty. This screening procedure may constitute a substantial barrier to entry into international supply chains for small firms in developing countries. 4The so-called o-ring theory explores this feature of production networks and identifies possible poverty traps in such an industrial environment (Kremer 1993). The second way of reducing risk in supply chains is to introduce competition at each level in the chain. Each input is then sourced from several competing subcontractors who in turn compete on price and quality.

The latter practice is common for suppliers of standardized inputs in South Korean supply chains. A case study by Cho (1997) finds that during the early 1990s there has been a substantial transfer of production activities from large companies, the so-called chaebols, to small firms. Furthermore, more than 70% of all small firms are linked to the chaebols through subcontracting arrangements. 5196 firms of 30 chaebols transferred business to 2796 small firms during the 1990s (Cho 1997). Most of them are located close to the downstream customer. Cho also finds that the big firms on top of the supply chains spend more time and resources on supervising procurement and subcontractors and interfirm cooperation than they do on direct production activities.

A study of supply chains in Mexican in-bond assembly plants, the so-called maquila, finds that local firms only to a very limited extent are part of the supply chains of the maquilas (Brannon et al. 1994). The study suggests that high price, inadequate quality and unpredictable delivery are possible causes. A study of supply chains in Eastern Europe (Bateman 1997) argues that the precondition for developing supply chains is the establishment of a sufficient number of small firms. This is important because the dynamic process of industrial clustering and vertically and horizontally integrating the production process will not be possible without the flexibility that an underwood of small firms represents. Other preconditions are obtaining large-firm support for small-firm development and the building of trust that facilitate technology transfer and development. Finally, several studies have found that local potential suppliers’ difficulties in meeting the supply chain management’s standards have been linked to unavailability or prohibitive cost of financial capital, scarcity of skills and shortages of materials. All these problems should be at least partly alleviated through a more liberal trade regime allowing for trade in financial services and the movement of natural persons with the required skills. We finally note that Bateman’s study supports the findings of Cho (1997) from South Korea, where small firms indeed were conducive in the development of local supply chains.

We can conclude that in order for local firms to participate in supply chains managed by internationally competitive large firms, whether these are local or foreign, requires a minimum level quality control, delivery reliability and ability to process and act upon information quickly and effectively. Location theory predicts that remote or poor countries such as the SADC countries, may gain substantially from declining transaction costs, but the gains would mostly appear when transaction costs have come down sharply. The cost of transferring information related to telecommunications is only one aspect of transaction costs. Others are transportation costs, tariffs and other barriers to trade and factor mobility. We have seen that modern supply chains and production networks have a high degree of specialization and are built around effective communication and transport networks. In order to assess the SADC countries’ ability to benefit from trade liberalization in the network services sectors, we therefore present some figures on the quality and cost of their telecommunication and transport services in the next section.