Executive summary

Information technology (IT) has enormous influence on the global economy.' OECD governments view IT as a key technology and have developed policies and public-private programs to promote its diffusion in private industries and public services. In developing countries, IT can enhance competitiveness of key industries, modernize basic infrastructures, and reduce costs of economic transactions.

There is growing appreciation of the role of technology in enhancing productivity and competitive advantage. Technology diffusion programs in OECD countries number over 1,000, at an average annual net cost to governments in late 1980s of $70 billion, about half of which were targeted at IT.

Recently the strategic importance of IT has been reflected in visions of a "national information infrastructure" in Japan, Europe, and the United States. The U.S. government is rewriting regulations to promote competition and private sector-led expansion of the country's telecommunications infrastructure and market-driven services. It also is examining social applications, intellectual property rights, standards, and spectrum management reform. In Japan both government and industry are pursuing "info-communications" not only to help move toward an "intellectually creative society" for the 21st century but also to address the immediate problems of aging population, urban concentration and economic restructuring. The European Union High Level Group has proposed ambitious reforms and demonstration applications continent-wide to strengthen competitiveness and address pressing Community problems in transport, education and employment.

This study reports on the experience of eight industrial countries, the rationale for government promotion of IT, and guidelines for policymakers. The countries are Canada, Germany, Ireland, the Netherlands, Sweden, the United Kingdom, the United States, and Japan. The study focuses on programs of special significance to developing countries.

IT diffusion involves more than acquiring computerized equipment and microelectronics-based product designs and related know-how. It involves the development of technical change-generating capabilities, to adapt given technology to a widening range of needs. Continuous improvement after acquisition is essential to sustain competitiveness.

Capabilities to generate and manage technologies used in production (and product differentiation) are primarily developed by firms. The IT revolution is driving the need for investment in these capabilities. This is a management responsibility. But market failures and externalities affect the diffusion of new technologies. Governments have responded with policies and programs to ensure socially efficient IT diffusion, particularly to small and medium enterprises (SMEs). There is growing evidence that such programs, when appropriately designed, have significant impact on SME use and benefits from IT.

Determinants of Effective Diffusion

Private-sector demand for new technology depends on incentives, awareness of benefits, and positive attitudes toward risk and change. Supply-side factors influencing technology adoption include:

· Information on the technological options, costs and benefits. and alternative sources of supply and know-how (information markets);

· Technical and managerial skills necessary to invest, adapt, and use the new technology (labor markets);

· Finance, particularly for intangible investments, and services and new technologies (financial markets);

· Research and development (R&D) to absorb knowledge and complex technologies; and

· Infrastructure, networks, and technical support institutions (externalities).

Managing National IT Policy Portfolio

OECD countries have developed three broad categories of policies: support for IT production, promotion of IT use, and investment in infrastructure to link users and producers. Many of the policies are IT-specific - research grants for software development, standards for public information and electronic transactions, programs to promote IT literacy, and incentives for small enterprises to adopt applications such as computer-aided design. General economic policies also have IT ramifications: competition policy, financial and trade liberalization. tax incentives to upgrade products and processes, loans for industrial R&D and small enterprise productivity enhancements.

The United Kingdom's IT policy includes programs in every category. Several government departments (Trade and Industry, Education and Science, Defense) support IT generation through direct R&D support, human resource development, awareness programs, and technology information services. The Department of Trade and Industry coordinates cross-sectoral technology diffusion programs, demonstration projects, technology consultancies.

Computer-in-education programs, technology centers, and user clubs. Government computerization, public procurement standards, infrastructural expenditures, IT standards, and regulatory policies all help link IT users and producers.

OECD governments also promote public-sector use of IT, stimulating demand for indigenous IT products and supporting governments' growing information-handling function. Historically countries with large IT suppliers have used public procurement to develop IT production. Their procurement policies have become more open (except for defense), and they influence departmental purchasing towards emerging international standards.

The United Kingdom's central computer and telecommunications agency promotes public-sector IT use by helping agencies develop information strategies, plan IT projects, improve the productivity and methods of IT staff, set standards, choose quality IT products, pilot new applications, and secure high value in contracting out IT services.

Industrial countries have learned that:

· Diffusion support is the highest priority. While it may not be necessary to develop or produce IT locally to reap its benefits, diffusion is essential to maximize returns from R&D.

· Aiding the capabilities of SMEs should be a focus of diffusion. Their adoption of IT often lags behind large enterprises, and they are constrained by the market and their capabilities. Local governments should promote IT diffusion to SMEs.

Central governments need to take the lead in IT policies infrastructure, and R&D.

· Countries may target certain sectors for IT diffusion to enhance their strategic positions and assets. Ireland uses multinationals and its low-cost, highly educated workforce to promote export of software services. Germany diffuses computers and microelectronics to the machine tools industry to consolidate its competitive position.

· Computing, communication and multimedia technologies are converging; policies and strategies in these areas should be coordinated.

· While the OECD countries have built their national IT strategies bottom-up, recent efforts include some top-down elements. The United Kingdom has developed strategic planning exercises and intelligence capabilities. such as "technology watch" and "future looks," to ensure the selectivity and coherence of its diffusion programs and to coordinate scarce public resources for priority IT use.

The shift of OECD policies from promoting the IT industry to diffusing IT applications has been difficult and incomplete. Most countries viewed the IT industry as strategic in economic and military senses. European countries were concerned by Japanese successes in semiconductor markets and the launch of its Fifth Generation Computer project in 1981. This reinforced the European tendency to subsidize "national champions" in computing. In the U.S. the substantial public R&D and procurement of IT for defense have strengthened the industry. The government mainly funded basic R&D and left diffusion to market forces. But empirical evidence indicated many constraints to IT exploitation. The slower than expected productivity increases from IT investments caused policymakers in OECD countries to reconsider the balance of their policies between IT generation and its use. The persuasiveness of IT as a generic technology, the lag in its adoption by SMEs, and common pitfalls in applying new technology reinforced the need for a shift in favor of diffusion. Some Asian NICs like Singapore also have demonstrated that countries could leapfrog in IT use.

The shift toward diffusion also has been slow and difficult for practical and ideological reasons. It has been easier to devote resources to national champions, basic R&D, and show-case projects. In contrast, diffusion-oriented programs do not require big budgets but rather coordinating many ministries and disciplines, reaching many SMEs and changing management practices in user enterprises. Unlike the "hands-off" policies supporting basic R&D for IT generation, diffusion-oriented programs often involve selectivity and direct interaction. A shift towards diffusion implies that government organize pilot projects, become a lead user of advanced IT applications, and promote contact among IT producers and users.

Analyzing IT Diffusion Programs

IT diffusion programs include:

· Technology capability programs that support R&D, provides information, track technological developments and advance strategic alliances and technology transfer:

· Diffusion programs for small firms that place technical experts at firms to promote technological education and identify improvement opportunities, stimulate the use of private consultants to help users upgrade manufacturing practices, provide technology advice through state-supported technology and productivity centers, and raise awareness through the media, conferences and demonstration centers;

· Bridging programs, in which government provides vocational and technical education, adopts best practices in public sector use of IT, sets standards, develops science and technology parks, and coordinates inter-agency programs.

Diffusion programs should be based on analysis of barriers to adoption and use. Enterprises go through different stages in adopting IT: information search for new technology, analysis of business needs and technological options, negotiation and procurement, and assimilation of the technology and redesign of business processes. As a firm progresses, IT benefits and diffusion barriers change. To capture transformation-level benefits, a firm may need "hand-holding" by more experienced practitioners. The most advanced benefits require improved infrastructures and institutional overhaul.

Programs can be matched to firms of various sizes and technological capabilities. Some simply provide information and demonstration. Others strengthen the capabilities of SMEs. yet others target the technologically competent, to eliminate infrastructural weaknesses and build network standards and technical platforms for the rest of the industry to follow.

Guidelines for Designing Diffusion Programs

Designing technology diffusion programs is a relatively new activity in most countries. OECD experience offers guidelines in four aspects for policymakers: the technology portfolio, program design, delivery mechanisms, and program management.

Portfolio of technologies. Diffusion policies should reflect technology life cycles and the needs of users. Mature technologies may be targeted for wider diffusion, especially to SMEs. Forcing the pace, however, wastes resources on immature technologies. For example, several European governments promoted expensive minicomputer-based computer-aided design (CAD) system to companies in the 1980s. A decade later, Ireland successfully diffused PC-based CAD tools to small firms as the technology matured.

Programs should also consider the technological sophistication of users and their exposure to international competition and best-practices. The Irish system for manufacturing support offers both simple and world-class technology support to different users. It allays skepticism by demonstrating incremental rather than radical changes. Demonstration projects aimed at widespread diffusion, especially among SMEs, should involve real users in real factories, rather than showpiece installations in universities or research institutes.

Program design. Programs should be packaged. An industrial extension package might include a diagnostic function (technology audit as in the advanced manufacturing technology application program in Canada and Ireland), implementation consultancy support (as in the U.K.'s and the Netherlands' regional technology centres and Ireland's manufacturing consultancy service), and other support services for those who can identify specific needs, but have limited resources (as in Germany's BMFT quality program).

Program goals should be realistic, clearly communicated, measurable, and attainable. Understanding initial conditions and the nature of market failure is crucial. For example, the Irish CSTAR program incorrectly extrapolated consumer behavior from the affluent "telephone society" of rural Denmark to rural Ireland, where good telecommunications is a recent innovation. User needs also should be carefully researched. Diffusion programs should be consistent with users' business priorities and risk-taking capabilities. Typically users of a program to encourage the application of a technology such as Computer-Aided Design (CAD) or materials requirement planning (MRP) will need to do more than buy a "black box." Program designers must understand the organizational and business changes involved. For SMEs, these risks can be life-threatening. Incentives and cost-sharing schemes should reflect the users, motivation to participate. Cost-sharing of services can encourage users to take manageable business risks. Market signals should be used to determine the levels and targets of subsidies. Testing and regular evaluation are essential.

Program delivery mechanisms. Stable delivery mechanisms are important, since frequent changes impose learning costs on both users and providers. A key factor of the Canadian and Swedish programs (IRAP and IVF) is their stability. They remain relevant because they adapt to changing technologies but retain established routines. Temporary intervention often mismatches budgets and demand. Budgetary flexibility should match the changing needs of potential adopters during the life cycle of a program. Critical resources must be available at program launch. The tension between the need for critical mass in the supplying organization and proximity to users requires appropriate spatial and administrative decentralization.

Levels of technology and user sophistication should guide the choice of delivery institutions. Development of the technical consulting market can be an important factor in the program designer's options. Universities are cheap sources of competence and young researchers but are short on implementation. Using private consultants at the diagnostic stage is particularly productive. They often "over-deliver" in the hope of winning further business. Technology transfer mechanisms embodied in people are resource-intensive. Programs should therefore "lever" diffusion experience by promoting networks and information sharing. For example, the Dutch and Swedish programs involve industry representatives and trade associations in the selection and dissemination of small, applied R&D and demonstration projects; Ireland offers information-sharing service to link past and potential users of a program. Cost-sharing and grant-matching are more likely to ensure the recipient's commitment to the project, because loans - including conditional loans for R&D and innovation activities - often are not repaid. These cost-sharing mechanisms are used in all the countries surveyed.

Program management. Application and execution should be simple for users, especially SMEs. Minimum administrative competency is necessary to ensure proper financial management, monitoring, and communication. Corporate organization, with business-like practices and transparency, is often the most effective delivery agent of extension services, as in Ireland's microelectronics application center. A combination of "light touch" monitoring and project mentorship is vital, as is the case in the Swedish advanced manufacturing technology program.

Designing National Strategies for Developing Countries

Market failures and capability barriers are more severe and pervasive in developing countries. Structural constraints, low awareness and limited technological knowledge suggest that most of their institutions will slowly adopt and underinvest in new technologies. For early adopters and small entrepreneurs in particular, learning costs and adoption risks will be extreme unless aided by information networks and support infrastructures. Targeted training, extension, and demonstration activities also are necessary.

National IT diffusion strategies should include sound incentives, institutions, and capabilities. Cost-sharing and government-business partnerships can stimulate the use of IT consultancy, strengthen the technical consulting industry, and build intermediary institutions and networks. Telecommunications and broadcasting policies and infrastructural development also are key elements.

Priority should be given to improving access to international know-how. Because IT is a premier generic technology, excessive protection could lead to serious macro-economic damage. Collaboration with global IT leaders is more promising. Networks should be developed among foreign subsidiaries, advanced local users, and the potentially large community of small local enterprises.

Technical and managerial skills and organizational learning capabilities are required. Using IT to create rapid response, flexible manufacturing, customized production and new forms of industrial organization rely on training. Executive development programs can help integrate IT into business strategies. A national strategy should consider changes in the education system and its role in technological diffusion. Flexible, multi-skilled workers, firm-based training and experimentation, and life-long learning are necessary in the information age.

The strategy should be user-oriented. IT products must be accessible to local industries most likely to use them to improve international competitiveness. For most developing countries, the development of local IT suppliers should be encouraged only as necessary to improve the competitiveness of IT users. Developing countries need early, inexpensive experience to build a critical mass of local users and start them on a cumulative learning process.

The OECD experience suggests broad directions for learning within public agencies to further technology diffusion. IT strategy emerges from intensive public-private interaction and builds on experience. Working closely with the private sector, public agencies can gradually build intimate knowledge of potential IT users. As users of IT, government agencies may contract out systems development and support services. To decentralize information systems management within government, a central agency may limit its role to setting key standards for information sharing, IT procurement, best practices in systems planning, and information resources management. It may also establish basic training for IT professionals in the civil service and make top administrators aware of their leadership role. Resources should be focused on key elements of a national information infrastructure. A public sector-wide view of computerization needs would best identify information and communication bottlenecks and opportunities.

Designing Diffusion programs for Developing Countries

Environmental analysis and market surveys are essential to tailor program design to local problems, test assumptions about potential providers and users, and establish a basis for monitoring, evaluation and adaptation. Environmental uncertainties and differences in capabilities among users and providers are likely to be pronounced in developing countries. Building demand for a diffusion program takes time as awareness, familiarity, and trust between clients and service providers develop gradually. Demand analysis, piloting, cost-sharing with users, and the use of market signals and national "technology watch" are key tools to enhance the strategic fit of programs.

Diffusion programs are likely to undergo substantial evolution and require frequent evaluation. Programs take time to fine tune. In many OECD countries, periodic independent evaluation has led to successive generations of diffusion programs, or "open program design." Formal evaluation alone is inadequate. Intensive contact among all stakeholders and agencies propel learning and dynamism. Initial information scarcity and the institutional learning process suggest that the "blueprint" approach to program design, often adopted by aid agencies, is inappropriate.

Complementarity between physical and intangible investments in IT is important in developing countries. As global competition and technological change require higher skills and organizational learning, firms must invest in intangibles. Industrial countries have learned to selectively subsidize intangible services such as consulting and training.

Low-cost diffusion approaches are especially important. Developing countries should provide information services more broadly, while limiting implementation assistance and mentoring to SMEs and to early demonstrators of innovative applications. They should target simple and packaged information technologies, based on CD-ROM or PCs, to build a critical mass of users and to initiate institutional learning processes. They also should design declining-subsidy schemes in cost-shared diffusion programs, leverage international consultants with local partners, and emphasize process reengineering and managerial improvements.

Successful diffusion programs rely on local institutions and networks to speed knowledge transfer. They should build on intermediaries such as trade and professional associations, financial institutions, universities and R&D institutes, and private consulting firms. This approach is more responsive to industrial traditions, local resources, and established client relations. At the same time, IT diffusion programs should reward institutions that respond to the demands of new technologies and users by sponsoring cost-sharing research projects and consulting services, particularly for SMEs.

Diffusion programs can help build networks and promote relationships among users and suppliers, thus enhancing information sharing and demonstration effects. Cooperative research projects among SMEs, consortia of local suppliers, and joint ventures with foreign partners for IT diffusion would accelerate diffusion. The creation of such networks should be an instrument and an objective of IT diffusion programs.

Role of Government and Private Sector

Although IT generation and diffusion should continue to be driven by the private sector, government's role should be reoriented to expand private sector choices, promote diffusion among lagging users such as SMEs and social sectors, and stimulate the development of a national information infrastructure. The balance between hands-on and hands-off programs will vary according to capabilities of public agencies, development of the basic infrastructure and the IT industry, concentration of user industries, and availability of technical and managerial education. For most countries, government should focus on promoting standards and training, and developing networks among IT users to share experience. In devising diffusion programs, the government may exploit and extend the capabilities of the private consulting industry to facilitate learning among SME users.

Implications for Aid Agencies

Aid agencies should move beyond investment in IT components of development projects and help governments become effective in using IT, in designing IT policies and strategies, and in diffusing IT throughout the economy. Aid agencies may take a dynamic, long-term view of the roles and capabilities of both public and private sectors. They should promote public-private partnership and help develop private-sector involvement in planning and implementing IT projects.

In particular, aid agencies should:

· Advise governments on the strategic implications of IT for national development and for key sectors like trade, manufacturing, logistics, and education.

· Assist in developing policies and infrastructures that support the diffusion of IT and build links between IT producers and users: public investment and private participation in telecommunications, technical education and computer literacy, incentives for R&D and IT adoption, and promotion of technical standards to facilitate diffusion and information sharing.

· Assist public-private collaboration in the design and implementation of IT diffusion programs to improve competitiveness of private enterprises, particularly the SMEs.

· Improve the capabilities of government as a user of IT and manager of public information resources.

· Support strategic alliances and recruitment of IT multinationals as partners for local IT producers and users.