Technology transfer vs. technological learning: IT--infrastructure and health care in developing countries Jorn Braa, Norwegian Computing Centre, Oslo, Norway Eric Monteiro, Dept. of Informatics, Univ. of Trondheim, Norway Erik S. Reinert, Studies in Technology, Economy and Policy, Oslo, Norway Abstract. The notion of technological ``transfer'' assumes that a piece of technology may be regarded as an isolated machinery or artifact. This, as many have argued, neglects the crucial aspects of the social and cultural context of use of a technical artifact. Technology has to be learnt and mastered rather than transferred. From this rather general point, we more specifically analyse the conditions and possibilities for technological learning about IT--infrastructure in developing countries. In so doing, we distinguish between local, contextual learning and institutional or technological infrastructure for learning. These processes of learning need to take place in areas sheltered from international competition. We argue that the health care sector is a suitable candidate for such an area for developing countries. Examples from Mongolia and South African are used to illustrate our points. 1. Introduction If there were nothing else to a piece of technology than the physical thing (the artifact) itself, technology transfer would be trivial. The problem, of course, is to simultaneously transfer appropriate and effective patterns of use of the artifact. The use of an artifact is never fully determined by the physical thing itself, it is also influenced by the social, cultural and economical context (Kerbal 91, Odedra 92, Pfaffenberger 93). With regards to technology transfer from the North to the South, this implies that in addition to transferring the artifact one has to transfer knowledge and skills which enable the artifact to be used appropriately. As ``transfer'' of knowledge and skills is a potentially misleading phrase, we prefer to talk of learning: in addition to transfer of artifacts, one has to ensure that technological learning takes place. Similarly, social studies of technology empirically document how all users of technology shape and adapt artifacts in ways which were not planned during design (Bijker et al. 92). These studies show that it is never possible to completely determine the pattern of use during design, only to enable certain patterns and constrain others. There is thus a considerable amount of evidence indicating that technology cannot be ``transferred'' and subsequently be put to use exactly as planned during design. The patterns of use of a technology have to be acquired or learnt. So far, this is a general argument which is supported by both empirical findings and conceptual considerations. Our paper addresses this issue by focusing on IT in general and IT--infrastructure in particular. If a piece of technology has to be learnt and mastered, the relevant question is this: what are the conditions and possibilities for appropriate processes of learning about IT--infrastructure in developing countries? The aim of this paper it to contribute both to suitable concepts for framing the issue and empirical material illustrating our analysis. The required technological learning in connection with IT--infrastructure is a highly heterogeneous process. It takes place on several levels and in different domains. More specifically, it takes place on a local level when adapting and adjusting a specific software application to a given, local context. This is covered in section 2. It is, however, necessary to move beyond the local level when establishing an IT--infrastructure. It also requires more than simply providing the basic, technical prerequisites in terms of tele- and data--communication services. Establishing an IT--infrastructure involves a close interplay between technical, institutional and economical factors. We discuss this in section 3. In sections 2 and 3 we draw upon examples from the challenges facing Mongolian health care information systems. An important line of argument is the need for developing countries to establish areas sheltered from international market forces for learning about IT--infrastructure. We argue in section 4 that the health care sector is such an appropriate area for developing countries. In section 5 we compare our sketch of a technology policy for developing countries for IT--infrastructure with an ongoing debate in South Africa over the same issue. Section 6 contains concluding remarks. The case from Mongolia is the result of a three month field study of the existing health information systems. (Braa et al. 95) contains a more detailed account of this specific case. The illustrations from South Africa stem from participating in the working group for a regional health information system in Western Cape province. This working group forms part of a national initiative under the Department of National Health in South Africa. 2. Context and localised, small--scale learning In systems development, as opposed to software engineering (Pressman 92, page 771), one has emphasised the importance of learning about the context of a computer system in order to make it useful (Bjerknes et al. 90, Floyd et al. 89). This knowledge does not exist before it is developed locally through praxis, and, as a consequence, it cannot straightforwardly be transferred. It is the knowledge, attitudes and preferences of all involved parties which constitute the non--technical context surrounding systems development. Systems development is seen not only as a process of designing and implementing (or purchasing) technical artifacts, but also as a process of communication and mutual, interactive learning: systems developers learn about the non--technical context, and would--be users learn about the possibilities and functionality of the hardware and software. This knowledge is unique in the sense that it did not exist prior to the development process. Viewed in this way, system development is a process during which a new understanding or ``reality'' is constructed from the knowledge about the use of IT in the local context (Floyd et al. 92). We thus distinguish between, on the one hand, systems development (Bjerknes et al. 90, Floyd et al. 89) and, on the other hand, software engineering (Pressman 92). Even though the two terms are often used synonymously, we find it illuminating to separate between them. This is because they serve to emphasise an important distinction in the development of software, namely whether or not one is sensitive to the local context of use of the system. In addition to this distinction, there is another one which more often than not is neglected when discussing development of software. This is the distinction between stand--alone and systemic (or network) systems. In connection with IT--infrastructure, this is a crucial distinction. Systemic technologies quite literally need a network of several local praxises to function. The resulting network will not be established before demand arises. It has to be developed gradually together with its exploration and mastering. A ``critical mass'' of distributed experience has to be established. There exists empirical evidence indicating that the diffusion of systemic technologies is quite distinct from stand--alone technologies (Antonelli 91). Summing up our brief discussion, we suggest that the two distinctions or dimensions identified above --- whether sensitiveness to context of use is required, whether it is systemic or not --- are relevant in understanding the challenges facing the establishment of an IT--infrastructure. The two dimensions give rise to the following classification of IT: classification of IT context--free context--sensitive stand--alone 1 2 systemic/ network 3 4 Along the top row (squares 1 and 2) we find different approaches to the development and use of basically stand--alone artifacts. This corresponds, as indicated above, to software engineering and systems development, respectively. Software engineering focuses on decomposition, isolation or de--contextualisation of tasks; systems development focuses on the mutual learning process and adaption between the design and its context of use. Along the bottom row we have technological assemblies of a communicative character. An illustration of square 3 is basic data communication, for instance the Internet protocols or OSI's reference model. We, however, focus on square 4 for two reasons. Firstly, we take the right hand side of the table (that is, squares 2 and 4) to constitute the interesting use of IT: the relevant learning processes underlying mastering of IT need to be sensitive to its context of use. Learning presupposes mutual adaption, otherwise we ``transfer'' IT instead of learning it. We accordingly argue against developing countries which set high priority on hard--core, context--free computer science, say, parallel programming or data base technology. This belongs to the left hand side of the table where little technological learning takes place. Secondly, as IT matures the real challenge will be how to integrate existing systems and supporting geographically spread work routines. A recent report by the World Bank makes a similar point when identifying how infrastructure, broadly conceived, has been neglected in favour of more stand--alone, isolated pieces of technology (World Bank 94). This amounts to favouring the bottom part of the table. The concern expressed by the World Bank to emphasise continuous, ongoing and flexible maintenance, rather than developing new technology, may readily be recognised as a greater sensitiveness to the context of use of a technology. How, then, is the issue of local use and the context of information met? We illustrate the situation by drawing on examples from Mongolia. The existing health information system in Mongolia is a centralised, vertically organised system. With complete disregard to context, statistics are gathered by aggregation of predefined entities from the local level all the way to the top. The original intention of the system was that information should only be used at the top. This is in stark opposition to our earlier remarks, that the use, interpretation and mastering of IT always have to be locally grounded. We illustrate how this disregard for the context of the information makes the existing health information system a poor basis for cost--effective health care in Mongolia. Official statistics report a more than 50% decrease in recent years in the number of patients staying in hospitals. When people partly cease to attend hospitals, due to, among other things, lack of drugs, an information system based on hospital data runs into problems. The basis for its information gets increasingly biased or misleading as an increasingly smaller part of the population is taken into account. As an example, the official, hospital based statistics report that syphilis and gonorrhoea are declining. Surveys based on the entire population, not only the few attending hospitals, indicate increases in the diseases. Despite high female literacy and extensive network of health care facilities, Mongolia has a higher maternal and infant mortality than countries with similar characteristics. Major infant killers have been respiratory and diarrhoeal diseases for the past twenty years. The maternal mortality has increased in recent years. In order to improve health care for mothers and children, three major programs have been launched: against diarrhoea, against respiratory diseases and a family planning program. There has lately been a slight decrease in infant mortality, but nobody knows why. Is it a ``mechanical" result of the rapidly decreasing fertility, or the result of the program against diarrhoea, or the one against respiratory diseases, or is it simply a result of under-reporting of deaths as some people claims? Answers to these question is vital to efficient primary health care (PHC) management. But as the officers in charge of PHC in the Ministry of Health explained, the present system does not answer any of these questions --- it simply reports the number of deaths. 3. Infrastructure and institutional frameworks of technological learning Economic theory is in the midst of a paradigm shift: the very idea of how economic growth takes place is being challenged. The previous paradigm is based on what Schumpeter called the pedestrian view that it is accumulation of capital per se that propels the capitalist engine. The new understanding of the economic growth process focuses on learning and innovation as captured by the term ``National innovation systems'' (Lundvall 92). The new ideas are becoming very influential in the OECD and in the EC, but at the same time the old neo--classical paradigm continues to be applied in policies towards the developing world (OECD 91). In the new economic theory evolving, it is recognised that occasional fundamental technological changes --- changes in techno-economic paradigm --- alter the very basis of society; steam power, electricity, rail transportation, or IT. IT plays a role today similar to that of electricity and the railway under previous shifts in techno--economic paradigms. All countries catching up with the leader --- England --- up until World War I, did so by protecting the new technology activities through tariff barriers. Japan and Korea did so until very recently. This option is presently being refused to the developing world by the policies of the World Bank and the IMF. The invisible hand of the marketplace will automatically assign the economic activities belonging to the previous techno--economic paradigms to the poor world. As an illustration, consider the following case. The largest producer in the world of balls for the American game of baseball is the poorest country of the Western Hemisphere --- Haiti. No US investment or new technology has managed to change the way these balls are produced, by hand with a basic needle--and--thread technology. The wage for the most efficient producers of baseballs in the world is 30 US cents an hour. In comparison, golf ball production was mechanised, and the average golf ball producer --- in an industrialised country --- receives a wage which is 30 times higher. Still, the industrialised world insists on believing that Haiti will get rich if she only could save more and produce more baseballs. Providers of raw materials and manufactured goods from labour intensive manufacturing goods, the Third World countries are specialised in the very products where technological progress does not take place. The catch--word ``technology transfer" does therefore work against formidably strong market forces. The poor nations are locked in a vicious circle of low internal demand and a specialisation in the world division of labour in producing mature products, not further mechanisable at present. The present U.S. and European governments strategies for domestic development purposes vary enormously from what the same OECD countries recommend for the Third World. Whereas the national strategies of the OECD countries focus on creating industries with dynamic imperfect competition, all policies towards the Third World are based on the assumption that all markets are perfect and that all government intervention therefore is harmful. Likewise, whereas the government in the rich countries play a key role in establishing a technical infrastructure, for instance the U.S.'s ``information highway'', the very same countries recommend that the creation of infrastructure in the Third World should be left to the free market. This re--thinking of economic theory thus implies a radically different role for infrastructure and frameworks --- including institutions and the role of the government --- for technological learning. We provide illustrations from Mongolia of technological learning. They are related to the issues of: infrastructure, institutional constructs and the role of governments (see (Gurbaxani et al. 90) for a discussion of the role of government in Singapore). Infrastructure: Emphasising the context of local use is necessary but hardly sufficient. A number of important health care issues only surface by comparing distinct sites or by aggregation of data. This amounts, in more technical terms, to developing a suitable IT--infrastructure also supporting gathering of non--local information. We illustrate the situation by providing examples of the need for integration and communicative infrastructure. The situation in the province of the Bayanhongor Aimak in Mongolia illustrates the need for cross--comparison among local sites. Official statistics report a decreasing infant mortality rate (IMR). This is correct but misses a vital point, namely that there are huge variations within the parts of the province. The IMR in the districts varies between 0 and 25. By analysing the reports from each district and plotting them onto a map of the province, a distinct pattern emerged: in the mountanious northern part of the province, the IMR was dramatically high and increasing. In order to take appropriate action, this information is vital. But it is ``averaged away'' in the official statistics. Again, the simple counting of deaths, without any concern for where and why the infants die, is of little help when trying to cope with the problem of infant mortality. In visiting local hospitals here, we learned that the majority of infant mortality occurred at home, a fact not reported in official statistics. Institutional constructs: Two institutions have played a major role in the history of informatics in Mongolia: the National Computing Centre (NCC) --- now reduced to computing department --- of the State Statistical Office and the computing centre of Academy of Science. NCC was involved in research both regarding technical aspects of IT and in developing practical solutions. Being the processing unit of the State Statistical Office, NCC was also a dominant user of IT. They also functioned as Computer Centre, processing data and making statistics, for ministries and other state organisations. Containing both research within IT, development of practical solutions and the use of these solutions, NCC played an important role in adapting IT to Mongolian conditions. It is, however, necessary to state that the needs of a planned economy, that is, statistics, were the driving force behind both use and development. A few years ago about 40 programmers and 30 electrical engineers were among the more than 200 employees at NCC, and they were a real centre of knowledge. Today they are only 12 IT professionals. The computer department of Academy of Science has a similar, dramatic decline over the past few years. Adaption to market economy implies that they have to sell their services in order to survive. Unfortunately, due to the economical crisis, there is no demand for their services, and hence no investment in system development. These two institutions were probably the closest one came to giving technological learning an institutional base in Mongolia --- but they have practically been dismantled. Role of government: Both economically and culturally the marmot hunt plays an important role in Mongolia. Unfortunately, the animal is also a carrier of plague. Each year during the hunting season there are outbreaks in different parts of the country. As plague is extremely dangerous and infectious, fast reporting and immediate isolation of affected population and areas are required. Communicating through non--digital telephone exchanges is not reliable and may easily take a day. Data--communication between computers are barely developed in Mongolia. But a message handling centre based on PC network will soon be established in each Aimak. These centres will have the capacity of 12 subscribers, but the price will be so high that no public sector organisation, and particularly not within health care, can afford to use it. The modern data--communication network which is being introduced, the Ministry of Transportation has decided, is required to function on a purely market basis. Only banks, the Stock Exchange and some private companies are able to use the data network about to be implemented. This lack of government support for infrastructure stands, as pointed out further above, in stark opposition to the situation in the North. It effectively blocks use of modern data communication within the health care sector where a gradual development and learning is necessary. Also non--governmental organisations (NGOs), not only governments, may stimulate the demand and production of computer systems. The development of a desk--top publishing application adopted to cater for the Mongolian Cyrillic alphabet was made possible because of projects for the WHO and DANIDA. An application to cater for the Mongolian Script is funded in a similar way. 4. Health care as a sheltered sector for learning The health care sector, we suggest, could provide a suitable sheltered sector for learning about, and diffusion of IT. There are three principal motivation underlying such a suggestion. First, health care is directly concerned with the basic needs of the population, and scarce resources to health purposes should be carefully managed. WHO strongly advocates the potential of IT for improving the quality and cost--effectiveness of health services and support activities (WHO 88). PHC is a concept to provide basic health care to the population, described in the Alma Ata declaration (WHO 78). PHC consumes more than 70 of all organised health care in developing countries and represents a key factor in the struggle for delivering ``health for all'' in the developing world. We thus focus on applications for supporting PHC, and the integration of these systems into the national level. Applications at district level strongly emphasise the issue of methodologies to ensure local adaption and sustained learning. The second reason why the health care sector is appropriate, is that this sector is sufficiently complex to make full exploitation of IT necessary. The health care sector is not one homogeneous sector. It is a complex body of institutions, praxises, professional interests and other actors. It comprises both small, geographically spread primary health units and large, centralised and more bureaucratic hospitals. There is a number of relevant federal and local political institutions involved with budgeting, planning and controlling health care. The problems related to the existence of disjoint hierarchies ordered according to profession (medical doctors, nurses and administrative staff) are well-known, arguably even notorious. It makes it reasonable to talk about the politics of health care. These features of health care indicate that its complexity matches (at least those of) any other sector. It represents an instructive arena in which to learn and adopt communication technology and open, integrated systems. The third and final reason is that the health care sector is that part of the governmental fabric which reaches the most peripheral level of society; it penetrates the entire society. In developing countries the health sector is typically among the biggest employers. The use of computers may be of two quite distinct kinds: either more or less integrated with the organisation, the health care sector in the present case, or confined to technological enclaves. As computers, ultimately, are formal, symbol--manipulating devices, we have to recognise the limits of formalisation to operate them. By keeping IT within strictly confined domains, the surrounding organisation of work routines may be sufficiently formalised as well; it makes transfer of both the technology and the restricted domain conceivable. This situation is illustrated by the fact that it is perfectly feasible to walk into British Airways' office in Daar--Es Salam in Tanzania and make a reservation on an inter--European flight using their computer based, world--spanning booking system --- while at the same time having problems making a local telephone call. British Airways is hence an example of technology transfer without any learning taking place. They simply transfer the confined context as well. But with domains, such as the health care sector, which are not readily restricted and formalised, this strategy is not viable. There is no way to avoid adaption and technological learning. Arguing for the economically opportunities of IT is rapidly becoming a cliche. There appears, however, to be a bit of confusion as to where this growth is to take place. More specifically, the prevailing view of how to boost the economy of developing countries through exploitation of IT seems to focus on software and hardware production industry directly --- to improve the trade balance by export (see for instance (Nidumolu et al. 93)). Alternatively, and more in line with our argument, one could focus on improving the use of IT and accordingly be concerned with appropriate domains of use. Recalling the two dimensions of IT from section 2, the former strategy is recognised to rely on de--contextualisation. It thus eliminates much of the learning taking place through flexible adaption to local conditions. At the level of technology policy, the difference between focusing on production vs. domains comes out quite clearly: subsidising of production vs. establishing sheltered arenas for technological learning. We have little to say with regards to the former, but our suggestion to focus on health care is an endorsement of the latter strategy. 5. An example from South Africa We discuss the African National Congress' (ANC) Reconstruction and development Programme (RDP) (ANC 1994a) and national health plan (ANC 1994b). It aims at altering this harmful effect of apartheid by developing an unified national health system that integrates all the different services. It is based on a PHC approach. The important point, for our purposes, is to acknowledge the variety and thus the context sensitive nature of the systems development that must take place. The logical and physical integration --- both horisontally and vertically --- of these semi--autonomus information systems underscore the systemic nature of the process that lays ahead. Recalling the table in section 2, local adaption and integration of information systems of this kind belong to the right hand, context sensitive side of the table. Decentralisation of technology and, as is the case here, management, tends to belong to the context--sensitive part of our table, while centralisation of technology and management tends to belong to the context--free part. South Africa has one of the most in-equitable health care systems in the world. A crucial element in the ongoing reconstruction is to establish a district health system responsible for PHC delivery. To make this process possible, the RDP and the national health plan (ANC 94b) both state that a National Health Information System (NHIS) must be introduced. A process towards developing a NHIS was initiated by a seminar in Broederstroom in March 1994. Since then, this process has been carried on by a national committee established by the Department of National Health. Working groups have been created in all provinces, work--shops are being held and plans for the development of a NHIS are made in each province. South Africa has an extremely fragmented health care system making prioritised integration very hard. The fragmentation reflects both the historical development of health services from colonial times and the effect of apartheid. Until 1993, there were fourteen departments of health at the central level: ``whites'', ``asians'', ``coloureds'' and the ``homelands'' created for ``blacks'', all had their own department of health in addition to the Department of National Health. The rather autonomous departments of health in the former provinces add to this organisational complexity. These different health structures all ran their own services (hospitals, clinics etc.), and they all had their own information systems. There were little or no coordination between them. As a result, there is a considerable geographical disparity. The public health services were, and still are, fragmented along racial lines, between curative and preventive care and between urban and rural areas. On the sub--provincial level, there are multiple authorities in the same geographical area. The fragmentation occurs on two levels: the local government level and the health service level. These two levels are not accountable to each other, and they are not accountable to the community. The district health system to be implemented is intended to solve this confusion. The RDP and the Department of National Health both emphasise the role of IT in the process towards integrating the health services, from the community and district level up to the provincial and national level. The district will be the pivotal level in the NHIS to be developed, thus underlining the bottom--up approach underlying the process. The cornerstone of the NHIS will then be a multitude of semi--autonomous, but integrated, district health information systems. The districts will differ greatly in social, economical, ethnic, political and geographical conditions, and they will range from poor rural areas of subsistence farming to urban industrial areas, from Zulu--land to Soweto and from the squatter camps to the wine yards. In each district the aim is to develop a locally adapted health information system. In order to achieve the goals of locally adapted district HISs, the process of development and implementation of these systems need to include user participation, communication and mutual learning as argued in section 2. Decentralised and user oriented approaches to technology development is important when the aim is to develop locally based, small--scale technology. South Africa has a highly centralised structure. There is little experiences with decentralised and user oriented approaches to technology development. In order to improve the quality of life for the black majority, the challenge in South Africa is to deliver a wide range of solutions at the local (and context--sensitive) level. The centralised (and context--free) approach to technology development prevailing in South Africa may form an important obstacle to this. In evaluating the achievements of the different ministries after the first 100 days of government, the Weekly Mail Guardian (August 12, 1994) illustrates this point by emphasising that the obstacles are made up by the old way of doing things: ``For example, Water Affairs is a technically superior department, proud of formidable engineering feats such as giant dams and making the Vaal river flow backwards, but has never paid much attention to the 12--million people who have no access to clean water and the 21--million who have no adequate sanitation.'' Building big dams without addressing the real needs of the population may be recognised as a centralised, technical fix approach to technological development. It belongs to the left--hand, context--free side of the table. The challenge is to move to the context--sensitive, right--hand side of the table, not by building big dams, but solving the problem of clean water and sanitation for the population. Or, as in the case with the NHIS, not only provide centralised database solution at national and provincial level, but also develop locally adapted health information systems at the peripheral district level. (Pacey 1983) labels this dichotomy in attitudes towards technology as a perspective with the focus on use (for instance, PHC, clean water and sanitation for the population) as opposed to the traditional focus on supply (such as building big dams and power stations with no regards to the use of the output). This way of focusing on use is very much in line with the strategy we advocate towards technological learning through use. ANC might be among the most computer literate governing parties in the world. The party has been a a heavy user of e-mail since the mid--80s when the communication on e-mail between the exile ANC and some universities within South Africa increased rapidly. Information from and about ANC is available on the Internet (gopher://wn.apc.org:70/11/anc/), and the history of the struggle against apartheid is available on CD--ROM. The tele--communication and IT--infrastructure policy put forward by the RDP emphasises that tele--communications is an information infrastructure that must play a crucial role in developing, enhancing and facilitating health care, education, public administration and rural development. This focus on using tele--communication and IT--infrastructure to enhance public sector is quite contrary to the market liberal policy in Mongolia with very limited focus on public sector. 6. Conclusion Our point of departure is that technology is learnt rather than ``transferred''. Furthermore, substantial, historical experience from the North strongly suggests that learning about IT needs to take place in sectors sheltered from international competition. This is what the North traditionally has done, and it is what the North is currently doing in connection with developing IT--infrastructure or ``information highways''. This implies that the South can only catch up through a planned approach --- and not by eliminating all possibilities of the required processes of technological learning. Only in this way is it conceivable for less developed countries to catch up technologically and, consequently, economically. Technological learning takes place on different levels. On the local level, in the case of the health care sector this would correspond to PHC, it entails that IT has to be used, interpreted and learnt in its proper context of use. At the same time there is a need for an infrastructure supporting integration and aggregation of information from local sites. The real challenge is to find a reasonable balance between the local and non--local needs. As we hope to have illustrated, this balance can only be reached through practical experience and learning. To sum up, based on the kind of conceptual analysis sketched earlier, together with the empirical material, we suggest that local use and processing of information is the key to improve quality of information both on the local level and centrally. The WHO calls for decentralisation of health care services. This corresponds, in technical terms, to the development of relatively independent, semi--autonomous, local systems which allow integration and communication; in short --- open and integrated systems. Institutions which form part of ``systems of innovation'' are important for all nations. We suggest that the health sector is appropriate as an arena for learning about use and development of IT: improving PHC--delivery while at the same time exploring IT. References ANC a. The Reconstruction and Development Programme, African National Congress (ANC), Johannesburg, South Africa, 1994. ANC b. A national health plan for South Africa, African National Congress (ANC), Johannesburg, South Africa, 1994. C. Antonelli. The diffusion of advanced telecommunications in developing countries. OECD, 1991. W. E. Bijker and J. Law, editors. Shaping technology/ building society. MIT Press, 1992. G. Bjerknes and B. Dahlbom, editors. Organizational competence in systems development. A Scandinavian contribution. Studentlitteratur, Lund, 1990. J. Braa, E. Monteiro and E. Reinert. Towards an information technology strategy in developing countries : the critical role of learning. In Proc. IFIP WG 9.4 The impacts of informatics on society : key issues for developing countries, Cuba, February 1994. (To appear, 1995). C. Floyd et al. Out of Scandinavia : alternative approaches to software design and system development. In Human--computer interaction, volume 4, pages 253 -- 350. Lawrence Erlbaum Associates, 1989. C. Floyd, H. Zullighoven, R. Budde, and R. Keil--Slawik, editors. Software development and reality construction. Springer--Verlag, 1992. V. Gurbaxani, K.L. Kraemer, J.L. King, S. Jarman, and J. Dedeick. Government as the driving force toward the information society: national computer policy in Singapore. The information society, 7:155--185, 1990. Ali Kerbal. The 'dilapidated mode' and technology transfer to developing countries. In Designing for everyone: Proceedings of the 11th Congress of the international Ergonomics Association, Queinnec, Y. og Daniellou, F. (Eds.), Taylor and Francis, pages 1410--1412, 1991. Bengt ke Lundvall, editor. National Systems of innovation. Towards a Theory of innovation and interactive learning. Oxford University Press, 1992. S. R. Nidumolu and S.E. Goodman. Computing in India: an Asian elephant learning to dance. Comm. of the ACM, 36(4):15--22, 1993. M. Odedra. Is information technology really being transferred to the African countries? In G. Cyranek and S.C. Bhatnagar, editors, Technology transfer for development, pages 47--58. Tata McGraw--Hill, 1992. OECD. Technology and productivity, the challenge for economic policy. OECD, 1991. Pacey, Arnold. The culture of technology. Basil Blackwell, 1983. Bryan Pfaffenberger. The factory as artefact. In Pierre Lemmonier, editor, Technical choice, pages 338--370. Routledge, 1993. Roger S. Pressman. Software engineering. A practioner's approach. McGraw--Hill, 1992. World Bank. World Bank Development Report. Infrastructure and development. World Bank, 1994. WHO. Primary health care. World Health Organization, Geneva, 1978. Health for all series, no. 1. WHO. Informatics and telematics in health. Present and potential use. World Health Organization, Geneva, 1988.