On Information Technology and Competitiveness in Latin America by Alberto Chong Alejandro Micco

Transcripción

1 Inter-American Development Bank Banco Interamericano de Desarrollo Departamento de Investigación Research Department On Information Technology and Competitiveness in Latin America by Alberto Chong Alejandro Micco Prepared for the seminar Towards Competitiveness: The Institutional Path Annual Meetings of the Board of Governors, Inter-American Development Bank and Inter-American Investment Corporation Santiago, Chile March 16, 2001 The authors would like to thank Eduardo Lora and César Calderón for comments, and Christian Daude and Natalia Pérez for valuable assistance. The views expressed in this document are the authors and do not necessarily reflect those of the Inter-American Development Bank. This paper is also available online at 1

2 INFORMATION TECHNOLOGY AND COMPETITIVENESS IN LATIN AMERICA 1. WHAT IS THE DEAL WITH INFORMATION TECHNOLOGIES? Information technologies, understood as those technologies that help produce, gather, distribute, consume and store information have, more than ever, come to the fore. The reason is simple. They have become ubiquitous in every day life in a relatively short period of time. This is particularly true of the Internet whose access has increased by hundreds of times in recent years (Figure 1). Terms that were non-existent just a few years ago such as world wide web, , Intranet, and many others, are now part of people s every day vocabulary. Consumers can now go on-line and comparison shop around the world among hundreds of vendors with little effort. They can download music, photographs, and film from the web in a matter of minutes. Complex banking and other financial transactions can be made at home. People can listen or watch the news live from pretty much anywhere in the world. And this is just the tip of the iceberg. Figure 1 Estim ates of internet access, (Thousands of subscribers) 250,000 United States Western Europe Japan Developing Countries 200, , ,000 50, Source: Pyram id Is this for real? Is the Internet just a somewhat different way to communicate, a technological curiosity not very different from traditional methods, such as the telephone, the fax, or snail-mail, or are we entering a new era, a global economy on steroids? In fact, while the Internet does seem to be something more than just a fancy way to provide and receive information, the extent to which it contributes to an economy is unclear. To this date, this debate is very much open. Some commentators argue that the world is posed to enter a third industrial revolution that will transform the economy in such a way that the old laws of economics will no longer apply. Sooner or later, it has been claimed, the law of supply and demand will cease to exist. A more conservative view, shared by 2

3 most economists, is that while the laws of economics are not the problem, specific particularities of information technologies are. Box 1. The Origins of the Internet What is now the Internet began in the late 1960 s as a project of the US Department of Defense s Advanced Research Projects Administration (DARPA). DARPA wanted to link the Department of Defense, defense research centers, such as government laboratories and universities, and the Department s contractors. The network, initiated in 1969, was called ARPANET and had four linked mainframe computers. As a result, the Internet s basic characteristics reflect the early needs of the US military and the research community. The desire to optimize transmission speed. Since the network would inevitably have segments consisting of low-speed telephone lines, the protocol had to be designed so that the maximum use of available bandwidth could be obtained. The need for a system that would allow universal access by a wide variety of computers, operating systems, local networks, and communication media with different technical standards. The need for survivability. If war broke out and part of the network was damaged, it would be necessary to route transmissions around the damaged portion so that messages could reach their destination. The need for high-quality communication, since distortion of the signal could corrupt the transmitted data or scientific files. As a network of networks, the protocol had to remain simple so that all users would easily understand it. The Internet s protocol for communicating differs significantly from that of conventional telephone calls. The most fundamental difference is that when a user makes a phone call, and end-to-end circuit is opened, and kept open, between the origination and the destination of the call. Even if there is a silence in the conversation, the line cannot be used by somebody else. The Internet, on the other hand, is connectionless as there is no end-to-end set-up and packets share a line with many other users. While packet switching is a very efficient and frugal method of communicating and one that satisfies most of the needs of its original military mission, its shortcomings become apparent as larger numbers of users engage in a broader range of uses. Source: OECD (1990) As Shapiro and Varian (1998) explain, the same well-known economic principles applied to the traditional economy are applicable to this economy on steroids, as exemplified by previous technological breakthroughs provided by the historical evidence. The three key particularities of the information technologies era are related to (i) pricing 3

4 information; (ii) lock-in problems and switching cost issues; and (iii) network externalities and standards. Pricing Information. As Shapiro and Varian explain, information is costly to produce but cheap to reproduce. Books that cost hundreds of thousands of dollars to produce can be printed and bound or a dollar or two, and one-hundred million dollar movies can be copied on videotape for one or two cents! Production of information goods involves high fixed costs but low marginal costs. The cost of producing the first copy of an information good may be substantial, but the cost of producing or reproducing additional costs is negligible. This cost structure leads to substantial economies of scale. The more someone produces information-related products, the lower the average cost of production. Moreover, the dominant component of the fixed costs are sunk costs while the marginal costs of additional copies of the product do not tend to increase as in other commodities. Lock-In Problems. Sometimes new technologies are linked with lock-in effects so that once a path is chosen the costs of switching become extremely difficult. In fact, lock-in effects are not absolute as new technologies do displace old ones, but their existence can affect a firm s ability to compete, its strategy, and options. The extreme historical example of a lock-in problem is the case of the layout of a computer keyboard, the so-called QWERTY arrangement. Why is this slower arrangement still in use, even when others, such as the Dvorak (1932) system appear to be more efficient? The problem is that it is difficult for any individual to get out of this system because the return to each person depends on what everybody else is doing. One simply cannot ask the question QWERTY or Dvorak? in a vacuum. The answer depends on how others have answered the question. This divergence between individual cost and social gain occurs whenever a system of production, or organizational form, exhibits externalities, so that the cost or benefit of adopting that system by an individual depends on how many other individuals have adopted that system. The adoption costs of a system may be reduced by the number of existing adopters (Ray, 1999). Network Externalities. A third feature of many information technology-related products is the fact that they tend to exhibit network externalities. Communication technologies are a prime example. Telephones, electronic mail, Internet access, fax machines, and modems, all exhibit network externalities. Technologies subject to strong network effects tend to exhibit long lead times followed by explosive growth. The pattern results from positive feedback, as the installed base of users grows, more and more users find adoption worthwhile. The key challenge is to obtain a critical mass so that the market can build itself. A nice example is the fax machine. The basic technology was patented in 1843, and AT&T introduced it in the United States in However, faxes remained a niche product until the mid-1980s. During a five-year period, the demand for and supply of machines exploded. Before 1982, almost no one had a fax machine; after 1987, the majority of businesses had one or more. The Internet exhibits the same pattern. The first message was sent in 1969, but up until the mid-1980s was used only by techies. Internet technology was developed in the early 1970s but didn t really take off until the late 1980s. But when Internet traffic did finally start growing it doubled 4

5 every year from 1989 to After the Internet was privatized in April 1995, it started growing even faster (Shapiro and Varian, 1997). 1 In a similar vein, recent research shows that the use and impact of new technologies follows an S-shaped path. This kind of expansion resembles the way an infectious epidemic spreads among the population. In the first stage there is a slow rate of contagion and a small, relatively stable number of infected individuals. Once a critical number catch the infection, the rate of subsequent infection accelerates rapidly. In the third stage, there are so many victims that the number of cases tends to stabilize. Similarly, new technologies require an incubation period before they build a user base quickly, and often have little impact on growth rates and output for some time. They might require additional training of the workforce, reorganization of the production process or company structure, replacement of obsolete machinery and so on. After this period, which can be very long, productivity and growth can rocket. In addition to the direct impact of the technology, there are often indirect spillovers into other industries. In the final stage, the technology will be fully exploited and growth will slow again. (Chong and Zanforlin, 1999; Coyle, 1999). Box 2. B Information Technologies and Productivity: Why Now? In a widely cited paper, David (1990) highlights the striking parallels of information technologies and previous technological breakthroughs. He mentions the steam engine and the combustion engine, but chooses to concentrate on how the dynamo came to conquer US industry around the turn of the last Century. The process took longer than one might imagine: almost half a century. It took time to expand the capacity of the electric system it took time to tailor the technology as best possible to its potential applications in industry. And it took time for the organization of the workplace to adapt to the opportunities opened up by the new technology (in the case of the dynamo, switching from huge steam engines to a series of smaller electrical machines and so making factories more flexible). It also took time for the workforce to come to grips with the new technology (learning by doing) and in some cases, it took time before it became profitable to replace cheap labor with electrically powered machinery. During the early days of electrification, the productivity gains were not particularly large, and in some cases dropped. But once the adaptation process gained momentum and higher volumes of electrical power began to push down prices, there was explosive growth. However, it should be mentioned that some skeptics have pointed to the fact that there are also patterns that are dissimilar between information technologies and electricity. In particular, computer related equipment has experienced a much quicker price drop and unlike IT-related products, electricity managed to coexist with the Source technologies it replaced. 1 However, having the superior technology does not guarantee success, as agreeing upon standards is also important. 5

6 In fact, economic growth has three basic sources. First, increased labor input. More hours worked, more workers or better quality workers all constitute increased labor input. Second, increased capital input, that is, more physical machinery. These first two involve increased inputs of the two basic factors of production. The third source of growth is increased efficiency in using these inputs. Economists refer to this as total factor productivity or TFP. We can break growth in an economy into these three sources using an approach known as growth accounting. Within this, TFP has a precise mathematical meaning. If we write growth as y, then with a few basic assumptions we can decompose growth as follows: y = w L L + w K K + A where L represents labor input, K capital input, w L and w L represent the respective shares of labor and capital in total income. indicates change and A represents the change in efficiency, the change in total factor productivity. Thus, it measures the improvement in productivity that cannot be explained simply by increased inputs of labor or capital. Figure 2 Output Grow th for Selected Regions 0.05 Growth in GDP per worker Latin America East Asia Indus trial Higher TFP growth is the holy grail of modern growth economics. Any economy can grow by increasing its labor inputs, so long as available unemployed labor exists. Similarly, any economy can grow by increasing its use of physical capital, but investing in capital carries a cost and requires people to cut back on consumption today. If an economy can achieve higher TFP growth, real incomes can increase over time without the necessity of incresed use of such inputs (Goldman Sachs, 2000). 6

7 Figure 2 details growth by region, and Figure 3 decomposes growth using the simple Total Factor Productivity approach described above. Figure 3 Total Factor Productivity Growth for Selected Regions Growth in Total Factor Productivity (TFPG Latin America East Asia Industrial Total factor productivity increases when existing capital and labor are combined to produce more output, that is, when productive efficiency increases. Factory redesign, organization efficiency, better methods, all contribute to this. The use of the Internet (additional to the purchase of the equipment, which enters as physical capital) is part of this, too. Thus, as much information technologies amplify brain power in the same way that the technologies of the industrial revolution amplified muscle power (The Economist, 2000) the ultimate test of its benefit is the potential impact on productivity, either by creating new products or by making the existing ones more efficiently. After all, faster productivity growth is the key to higher living standards. Recent developments in industrial countries have fueled the conventional wisdom that information technologies do affect productivity and thus, economic growth. This idea has been reinforced by the fact that the United States is, still, experiencing its most prolonged expansion in its history. In fact, not only is productivity performance exceptionally strong, but also the rate of unemployment has gone below what was thought to be the natural rate of unemployment. All these while there are little signs of inflationary pressure. The major new boost to the growth potential and productivity so widely touted by the media is the technological breakthrough spearheaded by the United States. In fact, Oliner and Sichel (2000) show that the contribution to productivity growth from the use of information technology, including computer hardware, software, and communications equipment, surged in the second half of the 1990s. In addition, technological advance in the production of computers appears to have contributed importantly to the speed-up in productivity growth. These researchers estimate that the use of information technology 7

8 and the production of computers account for about two-thirds of the one percentage point step-up in productivity growth between the first and second halves of the last decade. In summary, they claim, information technology is the story. Box 3: On Growth and Technology: The Recent Evidence The poor performance of (physical and human) capital accumulation on explaining growth differentials has motivated the debate on whether labor productivity growth differences stem from the accumulation of capital or technology catch-up. The evidence shows that differences in TFP growth explain about 90 percent of the variation in growth rates of output per worker across 98 countries over the period (Klenow and Rodriguez-Clare, 1997). This result implies that technology catch-up plays a dominant role in the growth process. Recent evidence has given a major role to technological change in the dynamics of growth. For example, Greenwood, Hercowitz and Krusell (1997) show that there has been significant technological change in the production of new equipment, and that these improvements have made equipment less expensive, thus triggering a more rapid accumulation of equipment both in the short and long run. These facts have highlighted the role of investment-specific technological change as a source of economic growth and economic fluctuations, and the relevance of vintage capital models as an appropriate way to model productivity (Hulten, 1992, Greenwood and Yorukoglu, 1997; Greenwood, Hercowitz and Krusell, 1997).In addition to differences in types of capital, differences in the speed at which countries adapt their policies and structures to match those of leader countries play an important role in explaining country growth differences. For example, Parente (1994) finds that even though all countries grow at the same rate in the long run, countries are located at different points of the growth spectrum due to policies and institutions that affect how fully they can benefit from the world technology frontier. According to their framework, faster than average growth could be attributed to the adoption of better policies and an improvement in institutions that allow countries to benefit more from the technology frontier. However, skeptics point to the fact that it is difficult to know the baseline or benchmark against which to measure the impact of information technologies. How can one be certain how productive firms would be in the absence of information technologies? Along these lines, Triplett and Bosworth (2000) compare the intensity of investment in information technologies in various industries in the United States with their growth rates of total factor productivity (or TFP), which measures the increase in the ratio of output to the sum of capital and labor inputs. They find essentially no correlation. Indeed, certain industries where information technologies were especially strong in relation to total output, such as education and banking, had either low or negative growth in total factor productivity. However, these researchers acknowledge possible measurement problems (Litan and Rivlin, 2000). The most recent study available on this issue is that by Nordhaus (2000). 8

9 By using new data and a new methodology, this researcher estimates that productivity growth in the new economy sectors has made a significant contribution to economy-wide productivity growth. In the business sector over the last few years, labor productivity growth excluding the new economy sectors was 2.24 percent per year, as compared to 3.19 percent per year including the new economy. Of the 1.82 percentage point increase in productivity growth in the last few years relative to the earlier period, 0.65 percentage points was due to the new economy sectors. The contribution to the new economy was slightly larger for well-measured output because that sector is smaller than the business economy. Moreover, Nordhaus finds that the major contributors have been manufacturing electric and non-electric machinery, the major subsectors of which are computers and semiconductors. These two sectors, which constituted under four percent of nominal Gross Domestic Product, contributed to 0.6 percentage points of the 2.4 percent per year GDP productivity growth in the period. Being a latecomer to the information technology dance, can the region still benefit from this revolution? In fact, there are reasons to be optimistic about Latin America s chances of taking advantage of the Internet revolution. For one, being a latecomer in Latin America may work to the region s advantage, as the dissemination of e-commerce and other Internet applications will likely be compressed into a shorter time frame and the spillover benefits for efficiency improvements may be absorbed faster. As described above, there has been a delay between investment in information technologies and productivity growth (the S-shaped curve described in the previous section). However, a latecomer does not have to reinvent the wheel; by emulating the best practice or application of technology he may be able to realize the stream of benefits with a shorter gestation period. Although the use of the Internet in Latin America lags with respect to other regions, the speed with which the new technology has spread has been nothing less than remarkable. In fact, it has been shown that information technologies spending in developing economies has been growing more than twice as fast as in developed countries over the past decade, though admittedly from a low base (The Economist, 1999; Goldman Sachs, 2000). This is shown in Figure 4, which summarizes a hypothetical diffusion pattern between Internet and productivity (also, see Appendix 2). 2. HOW CAN LATIN AMERICA BENEFIT FROM THE INTERNET REVOLUTION? It is expected that information technologies, such as the Internet, and electronic commerce will generate productivity gains by reducing transaction costs. The rapid dissemination of information, the substitution of digital for paper record keeping, and the networking capabilities of the Internet will improve flexibility and responsiveness, encourage new and more efficient intermediaries, increase the use of outsourcing, reduce time to market by linking orders to production, and improve internal coordination (World Bank, 2000). Productivity gains for firms can be expected through improved procurement and inventory control and reduced cost of intermediation and of sales transactions. Consumers also can benefit through reduced search costs, thus increasing competition and reducing prices. 9

10 In fact, the most important attribute of the Internet is, perhaps, the most obvious. Internet allows the transmission of information quickly, conveniently, and inexpensively. Routine transactions, including making payments, processing and transmitting financial information, and maintaining records can be handled less inexpensively with web-based technology. Using information technologies many firms can reduce the costs of production (Litan and Rivlin, 2000). Box 4: How Industries in Latin America Can Benefit From the Internet Researchers argue that the potential for transaction cost savings from transition to the Internet is especially high in the health care sector because it is large, information intensive, and highly dependent on paper records. Moving health claims processing to the Internet would require aggressive efforts to standardize claims formats but savings could be huge. It has been claimed that savings can amount to more than fifty percent per claim in the United States. The Internet also offers great potential in the area of managing medical records, not only for cutting costs, but for improving the quality and effectiveness of care. Assuming that privacy concerns can be adequately addressed, patients and providers would benefit enormously from converting the paper medical records into electronic forms. As in industrial countries, the potential of Internet use in the financial sector is very large. In the United States, however, retail banking has become a major source of tension. Financial services based on customer-provider relationships tied to geography and the provider s knowledge of the customer contrast with savings offered by on-line markets for standard financial products. In the lending industry, customers may use the Internet to shop for information and compare rates. As consumers grow more comfortable with on-line transactions it would not be surprising if loans start originating on-line, too. With the now legal digital signature law signed in the United States recently, this may occur sooner rather than later. If so, consumers will save through lower margins plus lower costs in processing applications. Source: Brookings Institution (2000) Additionally, firms in Latin America can use the Internet to achieve the kinds of procurement and inventory savings now enjoyed only by firms from industrial countries, for example, by purchasing out of the box electronic commerce applications. Goldman Sachs (1999) estimates that 30 percent or more of the total cost of intermediate goods typically comprises process costs, or the costs of administering transactions and maintaining inventories. The potential for savings can be divided into reduced processing costs of procurement transactions, reduced price of inputs owing to increased competition, and improved inventory control. Keeping an electronic inventory and transferring information on replenishment needs over the Internet enables producers and retailers to reduce the time that components and raw materials spend at each processing stage. Even relatively small reductions in inventory holding time in retail trade can mean substantial increases in profits because the average cost to retailers of holding inventory for a year is at least 25 percent of the price, and margins may average only 3 4 percent (OECD 1999). Improved inventory control will enable firms to become more integrated with suppliers, which saves time and allows greater production specialization. Increased 10

11 production integration has led to a boom in specialized manufacturing firms producing components for more well-known companies (World Bank, 2000). Procurement in Latin America is slower, less efficient, and more labor-intensive than in industrial countries, so the technical efficiency gains from transferring procurement systems to the Internet could be relatively large (although the lower cost of labor in the region means that the economic gains could be more limited than in industrial countries). Also, the savings in working capital from reduced holding of inventories would be significant in Latin America, where the cost of capital is high and credit is often rationed or unavailable. However, the lack of reliable telecommunications networks and complementary services for example, transport facilities may limit these gains. Some limited survey evidence indicates that North American firms that were better at supplychain management to begin with are cutting these costs by a larger amount when using Internet-based inventory systems. This may be because an adequate supply of highskilled workers and a flexible organization are required to reap the full benefits of these systems. Productivity gains can be derived from eliminating or improving the efficiency of intermediaries involved in marketing and distribution. Middlemen often charge substantial mark-ups because of their knowledge about and contacts with suppliers. By greatly expanding access to information, the Internet has enabled the elimination of retailers, wholesalers, and (in the case of intangible products) even distributors in some sectors. More commonly, existing middlemen have been replaced by new approaches to intermediation made possible by the technology for example, online auctions and aggregators (firms that represent collections of buyers that can demand lower prices for bulk purchases). The Internet also can generate significant cost savings in transport. The advertisement and trading of empty truck space over the Web is reducing costs per ton in the U.S. trucking sector (The Economist 1999). According to one industry estimate, $15 to $20 billion annually in cost savings (4 to 5 percent of output in the U.S. trucking industry) may be realized. Eliminating or transforming middlemen functions will enable developing country producers to access both domestic and foreign markets at lower cost (see below). By contrast, Latin American trading firms whose main purpose is to help domestic companies trade with international markets will be at particular risk. Network externalities, combined with low marginal cost of adding new users, mean that the market for providing intermediary services offers considerable advantage to the first company on the scene. Thus the later arriving and less technologically sophisticated firms in many developing countries may have difficulty competing with industrial country firms as Internet-based intermediaries (UNCTAD 2000). Also, in some sectors, such as the export of primary commodities, developing countries may not be able to capture the cost savings from reduced intermediation. The Internet offers the potential for savings in retail transactions compared with traditional systems. OECD (1998) suggests that the greater availability of information to the consumer and savings on providing services could increase the productivity of sales staff in OECD countries by a factor of 10. However, the evidence on the sales of goods over the Internet so far does not show large savings. Preliminary studies found that goods sold on the Internet were priced the same or higher than in stores (Goldman Sachs 1997; 11

12 OECD 1998). Other studies estimated that books and compact discs (CDs) were 10 percent cheaper on the Internet (Oliner and Sichel 2000; The Economist 2000). The potential savings in service transactions are more impressive. For example, the total cost (including investment) of bank transfers over the Internet is half that of existing automated systems and one-eighth that of transactions using tellers. Note that a portion of these savings reflects efficiency gains, while a portion reflects the transfer of costs from producers to consumers in the form of time spent searching the Internet. The impact of lower cost of service transactions is likely to be less significant in developing than in industrial countries because lower wages mean that firms have less incentive to undertake the fixed costs involved in setting up electronic systems. Also, poor distribution systems, inadequate protection against fraud for credit card purchases, and limited consumer Internet access limit the potential for business to consumer commerce in many developing countries. Easier access to knowledge through the Internet will speed technology diffusion, which is of critical importance to developing countries because they tend to operate within the technological frontier. Electronic commerce can reduce the costs of communication between geographically distant partners and lower the search and compare costs involved in finding potential business partners and technologies. Also, the Internet provides a radial structure for interpersonal communication networks. Bulletin boards and news servers allow individuals to exchange information faster and within a wider environment than with networks based on telephone and fax. Connolly (1997) found that differences in communication and transportation infrastructure were significantly related to differences in the rate of product imitation encouraged by foreign direct investment (although this does not necessarily mean that electronic commerce has an independent positive effect). Grossman and Helpman (1991) argue that international contacts enable a country to obtain foreign technologies and adjust them to domestic use, an important channel through which the productivity levels of developed and developing countries are interrelated. Such international networking is greatly facilitated by the Internet. Harris (1998) quotes a Neilson survey that found business s primary use of the Internet was for gathering information. By opening markets to a wider range of potential buyers and sellers, the Internet is likely to foster a greater volume and variety of trade. The Internet could erode an important advantage now enjoyed by firms in industrial countries; that is, proximity to wealthy customers. For example, the Internet reduces the cost of producing customized products designed for distant markets; a tailor in Montevideo can now sew a suit by hand for a lawyer in Boston. The Internet will reduce barriers to the sale of services embodying skilled labor (Harris 1998). 2 The Internet s impact on Latin American firms access to multinational supply chains is uncertain. Increased information on developing country firms may improve their 2 In the Philippines, for example, companies use the Internet to provide accounting services, process insurance claims, and track credit card defaulters for industrial country firms (Jordan and Hilsenrath, 2000). In India, workers have been transcribing U.S. physicians oral records into written files since 1996, at one-tenth the cost of U.S. transcription services (Washington Post 1996). 12

13 access to multinationals, which tend to be limited by information on potential suppliers. Goldman Sachs (1999) estimates that because of poor research, firms purchasing managers tend to award 90 percent (in terms of number) of their procurement contracts to about 20 percent of suppliers. On the other hand, suppliers with poor hardware, software, and Internet transmission capabilities may be unable to compete with better-connected companies. There is some evidence that the new online auction systems have not resulted in the expansion of supply networks. General Electric, for example, has seen a reduction in the number of its suppliers since starting its online bidding site for procurement (The Economist 1999). Latin American firms may find it difficult to access online auctions because of a lack of credibility. Purchasers need to have confidence that suppliers will provide input on time and in conformance with specifications, and product quality may not be known ex ante. More than half of 35 large firms using online auction or exchange sites said that they would not do business through online Web sites with firms they did not know. Interview results indicate buyers typically industrial country firms see the risk in purchasing from developing country firms as especially high. Over time, greater use may be made of certification agencies (e.g., the International Standards Organization and the International Electrotechnical Commission) to independently assess the quality of new firms products and services. However, relatively few small firms, even in developed countries, use the certification services these bodies provide, because of certification costs and concerns that certification may not fully address buyers concerns in the markets where small firms compete (Callaghan and Schnoll 1997). Box 5: The Case of Siderar in Argentina In March 1996, Siderar a steel production firm, undertook a process of communication improvement with its business partners. The idea was to provide and receive information by electronic means. Initially, electronic mail was employed. Later, more sophisticated operations followed. From simple requests to providers, and order confirmations to clients, Siderar used information technology in auctions, international bids, and others. To this day, this firm is in the process of setting up a system called e-procurement by which inventory control and total integration with providers is expected. By using information technologies, Siderar encourages its 732 providers to access some specifically designed web pages, in order to allow information exchange and operational coordination. By doing this, providers are able to check on the status of payments, documents (bills, receipts, and others) as well as day-to-day reports. The advantages are that administrative costs for both providers and Siderar are much lower. The use of paper has diminished, and standardization among internal departments and between Siderar and its providers is now widespread. Additional software applications have been developed in SAP, a German control system. On the side of the clients, the system has advantages too. Among others, there is now the possibility of knowing, at any given moment, the status of the request. Is it in the manufacturing stage? Has it been delivered? When is the purchase expected? The system has cost Siderar about $1.2 million. Source: Clarín Newspaper (2000) and Fundes (2000) 13

14 3. WHERE DOES LATIN AMERICA STAND? The idea of Latin America as a latecomer to the Internet revolution is clear. In fact, the distribution of Internet access among countries is severely unequal. Despite rapid growth in Internet access in developing countries, industrial countries still account for the majority of Internet subscribers. More than 30 percent of U.S. residents had access to the Internet in 1999 compared with about 0.5 percent in Latin America. Electronic commerce is also relatively small in most developing countries. For example, electronic commerce in Latin America is estimated at 459 million dollars in 1999 compared with a GDP of about 2 trillion dollars (World Bank, 2000). In fact, Internet access is highly differentiated in the developing world. 35 Figure 4 Figure 4. Regional internet access, 1999 (per cent of population with access to Internet) Industrial Countries Source: Pyramid US Other MENA SSA Eastern Europe LAC Asia There are important differences across regions in the world in terms of economic creativity, a measure created by Warner (2000). He tries to bring together under one measure several important aspects of innovation, technology transfer, and diffusion with the institutions that facilitate innovation and diffusion. Figure 5 summarizes how the index is constructed. Reading from left to right, Warner first constructs an index of pure innovative activity, which is based on eight survey questions. He also constructs an index to measure which countries are most active in international technology transfer. This is what Warner calls the technology transfer index. Since countries can get technology either by investing it themselves or by importing it, he measures an overall technology index by whichever of these components is largest, after they have been scaled to have similar units. The key idea is that countries get credit on the technology index for either innovation or technology transfer. From an economic point of view, what 14

15 is important is that the country participates in the newest technologies and innovation not whether it innovates itself. To raise Gross Domestic Product through technology related activities, a country needs to achieve value-added at some stage of the process, not necessarily the inventive stage. Next, Warner constructs an index of startups, or ease of starting new enterprises. This index is an average of two parts. Whether financing is available and whether it is easy overall to start a new business. The former is measured by averaging responses to two questions: whether venture capital is available for risk taking entrepreneurs, and whether it is easy to get a loan with a good business plan but with little collateral. In the same Figure, the final economic creativity index is an average of these two indices, namely access to technology and startups (Warner, 2000). Figure 5 Innovation Index (Based on eight survey questions) Technology Transfer Index (Based on two survey questions) Ease of Activating New Business Venture Capital Financing Available Possible to Obtain Loan with Little Collateral Technology Index (Equals the Innovation Index or the Technology Transfer Index, whichever is greater Startup Index (Average of activating new business and the two financial questions to the left) Economic Creativity Index (Average of Technology and Startups) The average index of economic creativity for the industrial countries is 0.92, whereas the index for developing economies is The gap is observed in all the categories involved in economic creativity, although it is more significant in the case of innovation (0.89 for industrial countries vs for developing countries), although start-up of businesses is not that far behind. In addition, note that differences within the group of developing countries are also notorious. For example, compare the performance of Latin America and East Asia. The economic creativity index for East Asia is 0.32, whereas the creativity index of Latin America is Although both economies have poor performance in innovation, the advantage clearly goes to East Asia. If we inspect the sources of the advantage of East 15

16 Asia over Latin America, we would seek explanations in the areas of starting-up new businesses and technological transfer (see Figure 6). Figure 6 Economic Creativity across Regions Economic Creativity Innovation Tech Transfer Start-Up Industrial Latin America East Asia On the other hand, it is always interesting to assess what forces are driving economic creativity across the world. For the sample of full countries, we find that innovation has a higher correlation with economic creativity (0.85) than technology transfer (0.64). In addition, start-up of new business has a high correlation with economic creativity (0.95). However if we breakdown these correlations across regions, the results are even more appealing. Innovation and start-up of new businesses are major forces behind economic creativity, and their correlation with the latter is positive and significant across the regions, with only marginal differences in regional correlation. However, the behavior of technological transfer is quite consistent with the conventional wisdom. Technological transfer does not seem to play a major role in driving economic creativity, whereas it plays one of the most important roles in the economic creativity of developing nations (see Figure 7). 16

17 Figure 7 Correlation with Economic Creativity Innovation Tech Trans fer Start-Up Industrial Latin America East Asia In a world with international trade of goods and services, foreign direct investment, international exchange of information and dissemination of knowledge, the productivity of a nation depends both on domestic and foreign efforts of research and development (Coe and Helpman, 1995). While both have proven to be significant, the impact of foreign R&D stocks on TFP growth depends on the ratio of domestic imports to GDP and foreign direct investment. Recent evidence (Eaton and Kortum, 1996) has shown that ideas from the five leading research economies (United States, Japan, Germany, France, and the United Kingdom) have contributed more than 90 percent of the productivity growth for the rest of the OECD. The geographical proximity of some developing nations to the leading research economies has explained the importance of technological transfer in their growth process. In this respect, two Latin American economies made the top 10 in this category (Mexico is fourth, whereas Brazil is ninth), and the leading economies in technological transfer (top 3) are Singapore, Ireland and Luxembourg. Note that Canada is the only G-7 economy in the top 10 (in the fifth spot). Innovation plays a major role in the economic creativity of the industrial economies. Thirteen of the Top Fifteen economies belong to this group (with Israel and Singapore in the 6 th and 14 th being the exceptions). The performance of Latin America is poor, with all countries displaying negative scores. Costa Rica (-0.21) and Chile (-0.43) are the LAC leaders in this category, whereas Bolivia (-1.84), El Salvador (-1.35) and Ecuador (-1.31) are the poorest performers in innovation. Unlike Latin America, not all the countries of East Asia registered negative scores. Singapore (0.98) and Taiwan (0.81) are the clear leaders in the region in terms of innovation, while the poorest performers are Indonesia (-1.18), Thailand (-0.94) and the Philippines (-0.92). 17

18 Start-up of new businesses is highly correlated with economic creativity across all regions. The world leaders in this category are the United States (2.02), Hong Kong (1.63) and Luxembourg (1.51). The performance in start-up for Latin America is similar to its performance in innovation: all economies registered negative scores. Among the leaders in the region are Chile (-0.15) and Brazil (-0.59). However, the poorest performers in the region, Ecuador (-2.01), Bolivia (-1.92), and Peru (-1.51) are among the poorest performers in the world. As a matter of fact, 7 out of the 10 poorest performers in starting new businesses are Latin American countries. East Asia shows a superior performance relative to Latin America, with the leaders in the region, Hong Kong (1.63) and Singapore (1.31), being in the top 10 countries in the world. On the other hand, China (-0.78), Vietnam (-0.68) and the Philippines (-0.48) are the poorest performers in the region. Finally, the world leaders of economic creativity (which covers aspects of technology and start-up of new businesses) are the United States (2.02), Finland (1.73), and Singapore (1.63). In Latin America, only three economies register positive scores in creativity: Brazil (0.20), Chile (0.11) and Mexico (0.03) -located in positions 31, 33, and 35 (out of 59) in the world rankings. The worst performers in Latin America, Bolivia (- 1.69) and Ecuador (-1.66), are also the poorest performers in the world. On the other hand, Singapore (1.63), Hong Kong (1.10) and Taiwan (0.97) are the best performers in East Asia, with the poorest performers being China (-0.56), Indonesia (-0.32), and Thailand (-0.11). Figure 8 Corre la tion w ith Econom ic Cre a tivity Internet PCs Im ports of Equipm ent Industrial Latin Am erica East Asia Internet Hosts, Personal Computers, and Imports of Equipment We attempt to characterize technology from the evolution of Internet hosts, personal computers, and imports of equipment. 3 While the first two measures might 3 Internet hosts are defined as any computer system with an Internet Protocol address connected to the network. The data do not provide a full count of users because surveys 18

19 characterize the developments in Information Technologies (IT), the third one may give us an idea of technologies adapted/copied, somewhat similar to Warner s technological transfer measure, above. First, we find that there is a big gap in the number of Internet hosts (per 10,000 people) and personal computers (per 1,000 people) between the industrial and developing countries. Whereas the number of Internet hosts and personal computers are 173 and 191 for industrial countries, the number for developing countries are 6 and 21, respectively. Within the group of developing countries, the disparities between Latin America and East Asia are also significant. The number of Internet hosts (per 10,000 people) and personal computers (per 1,000 people) are 12 and 35, respectively, in East Asia. On the other hand, the same figures for Latin America are 3 and 29, respectively. 4 Figure 9 Internet Hosts per 10,000 people Average across the World All Countries Industrial Countries 2.6 Latin America 10.4 Eas t A s ia and the Pac if ic South Asia Middle East and North Africa 0.8 Sub- Saharan Africa 13.7 Eas tern Europe do not capture all computer systems connected to the Internet (e.g., computers behind firewalls) and thus provide an indicator of the minimum size of the Internet. 4 The data for imports of machinery and equipment were taken from the United Nations Yearbook of Trade Statistics and from the Yearbook of Trade in Engineering products corresponding to Section 7, machinery and transport equipment, of the SITC classification. 4 We construct this variable as the share of imported machinery and equipment to gross domestic product. Section 7 of the Standard International Trade Classification (SITC rev.1) includes all kinds of machinery and parts. The main subcategories are electrical and non-electrical machinery and transport equipment. Electrical machinery comprises electric power generators, domestic electrical appliances, medical appliances, telecommunications equipment, office machinery, and others. Non-electrical machinery includes nuclear reactors and power generators other than electric, and machinery for special industries such as agricultural, textile, and printing machinery etc. 19

20 Second, people with personal computers are more likely to have access to the world-wide-web. The correlation between Internet hosts and personal computers is 0.95 for the full sample of countries, with this correlation being particularly high (almost 0.99) for the sample of East Asian countries. Third, it seems interesting to evaluate the correlation between the evolution of IT (say, Internet hosts and computers) with the imports of equipment (say, a rough measure of the economy's ability to adapt or copy foreign technologies). We find that the correlation is quite high for East Asia (0.65 for Internet hosts and 0.67 for personal computers), whereas it is negligible for Latin American countries (0.01 for Internet hosts and 0.16 for personal computers). Internet Hosts, Computers and Economic Creativity The development of information technology (proxied by the number of Internet hosts and personal computers) is linked with the economic creativity of nations. Although the correlation between IT and economic creativity is positive, it is higher for industrial economies than for developing nations (0.73 vs for Internet hosts and 0.80 vs for personal computers). However, note that the lower correlation for developing countries might be driven by the sub-sample of African countries given the high correlation for both Latin America and East Asia (0.81 and 0.88, respectively, in the case of the Internet). If we inspect the correlation of IT and the components of economic creativity we find some interesting results. Internet is highly correlated with innovation (0.62) and start-up of new business (0.73), however it shows a negative correlation with technological transfer (-0.26). For Latin America and East Asia, IT (proxied by Internet and computers) has a positive correlation with all the categories of economic creativity. Specially, we note the high positive correlation between Internet and technological transfer (0.76 for Latin America and 0.79 for East Asia). Imports of Equipment and Economic Creativity First, we find that the correlation between imports of equipment and economic creativity is higher in developing countries than in industrial economies (0.57 vs. 0.30). Although the correlation of imports of equipment with each category of the economic creativity index is higher for the sample of developing countries (relative to industrial countries), we can observe that the imports of equipment have the highest correlation with innovation (0.57) and the start-up of new business (0.56). Second, the imports of equipment in Latin America have not shown the pattern observed for the developing countries. We find a negative correlation between imports of equipment and economic creativity (-0.03), which is primarily driven by the index of technological transfer (-0.07). This might imply that new equipment is not used efficiently in the production process of Latin American economies. Third, East Asian countries have shown a high correlation between imports of equipment and economic creativity (which is the highest among the 20