1.Structure of the economy.

Spain is the EU's second largest member country in terms of area and the fifth largest in terms of population and GDP (10.5 per cent and 8.2 per cent respectively of the EU-15 in 1998). However, in terms of GDP per capita, Spain falls to thirteenth place, higher only than Portugal and Greece. If, in general, the structure of the Spanish economy shows a similar pattern to that of the EU, it reveals some distinctive features of the Spanish economy that can be evaluated from demand and supply.

• Focusing on the demand side, we can examine the structure of the expenditures as a percentage of GDP in current prices enabling comparisons with the EEC-EU to be made. This reveals three main characteristics:

1. The proportion of public consumption has been less than that of the EEC-EU average. This trend is a result of the regressive fiscal system inherited from the dictatorship. This gap has narrowed since the end of the 1970s, with the introduction of personal income tax in 1979 and VAT in 1986.
2. The level of investment in Spain is higher than that of the EEC-EU average. While net domestic saving did not cover the domestic investment, the difference between investment and saving has been covered by foreign investment. This savings gap in relation to investment has been considered a restriction to long term growth. However the net direct investment flows has declined from 3.8 to -4.6 billion dollars between 1986 and 1998.
3. Spain exhibits less openness to foreign trade, measured by exports or imports as proportion of total demand. This pattern has changed since Spain joined the EEC-EU in 1986 since, as a consequence of the increase in exports over imports, there was an increase in the trade deficit in the 1980s. During the recession which started at the beginning of the 1990s, openness slowed down and the coverage ratio improved. Finally between 1995 and 1998 the trade openness increased again reaching 27 per cent in 1998.

• In the second part, we can examine the basic structural changes in GDP.   One widely used measure of structural change is the changing composition of economy-wide output or GDP. The industrial structure of OECD economies is shifting towards one more heavily based on services. From 1980 to 1999, the Spanish economy constitutes a good example of de-industrialisation. This corresponds to the changes in the structure of the economy of the EEC-EU and Japan, which have been characterised by an increase in the share of services and a decrease in the share of manufacturing as percentage of GDP. In contrast, in the USA and Canada, the increase in the share of services has been accompanied by a decline in agriculture, mining and non-market government activities, and by only a very slight decline in the manufacturing share. In fact, in Spain, the rising shares of both market and non-market service activities as a percentage of GDP have been more significant and reflect decreases in manufacturing (around 6 per cent).

2. Industrial Structure: Productive and Technological Specialisation

The classification of the main sectors into three or more parts is a conventional but limited method of measuring the extent of structural change. A more accurate approach would be to disaggregate the sectoral evolution of GDP and employment in relation to the different trends in each individual sector, so it becomes possible to distinguish growing, medium and declining growth activities. A complementary measurement of structural change in terms of direction could be obtained by classifying the branches according to their R&D intensity as high, medium and low technology (this will be explained in section 5.2).

The principal characteristics of the manufacturing sector since 1985 have been:

1. A significant under specialisation in high technology industries and, to a lesser extent, in medium technology activities. Both of these categories have been considered to be growing activities by the OECD (1993 & 1997). By contrast, there is an over-specialisation in low and medium-low technology activities, which are classified as medium and declining growth activities.
2. In general, the productive specialisation in high and medium high technology activities has been improving and converging to the OECD pattern. The medium-low and low technology industries have also been converging to the pattern, but decreasing their relative shares.
3. There are two clear sectoral patterns of specialisation. On the one hand, there is an increasing specialised trend in chemicals and transport equipment (Aircraft, motor vehicle and other), both of them included in high and medium technological activities. On the other hand, there is a decreasing pattern of specialisation in food, beverages & tobacco, and textiles, apparel & leather.

This aforementioned compositional approach would be useful in helping to evaluate the direction of structural change, but would fail to identify the transformations taking place between industries below the aggregate levels. Additionally, it could not connect changes in the structure due to other factors such as shifts in domestic demand, foreign trade, technical change or input productivity. An alternative analysis is based on the concept of the sub-system and the notion of a vertically integrated sector, which was introduced by Sraffa (1960) and Pasinetti (1967 & 1981) for theoretical purposes, but can also be used in applied terms (OECD: 1996a & 1997a, Díaz Fuentes 1994 & 1999).

In addition to compositional and directional structural changes, the economies are undergoing structural changes of different sorts, as manufacturing firms change the way they organise their production and source their production inputs. Such changes affect the linkage between manufacturing and services. Structural linkage of this sort can be examined by looking at the structures of production inputs in manufacturing and services (Sakurai, Papaconstantinou & Ioannidis 1997).

The expansion in the intermediate demand for services is accompanied by their increasing use in the production of manufactured goods in Spain (from 26 per cent in 1975 to 37 per cent in 1994) and in the principal European Union countries. This implies that the production of industrial goods goes beyond the industrial sector and requires increasing services. The decline in output and employment in manufacturing does not imply that services are replacing industry as the 'new engine of growth'. Manufacturing is the sector with higher productivity and the more significant forward and backward linkages . Moreover, the more dynamic and R&D intensive manufacturing branches such as computers and office and computing machinery, aircraft, drugs and medicine, and radio, TV and telecommunications equipment are linked strongly with the more dynamic service branches.

Growth in services must be explained by considering the increasing integration between services and industry. This relationship can be explained by the following four factors (Diaz Fuentes 1994). First, the growing specialisation among sectors, which requires a complex network of services such as Transport, Communications, Bank and Insurance and Business Services to link the different sub-systems. Second, the expansion in foreign trade which is another dimension of specialisation. Third, the augmenting regulations (quality standards, the environment and so on) which require specialised services such as engineering, publicity, lawyers, auditors, accountants, trainers, finances, insurance and others. Fourth, the emergence of new economies of scale in the production of services, which induce a process of externalisation of the service activities.

3. Structural Trends in Employment and Technology Intensity.

The Spanish occupational structure has undergone substantial changes during the last four decades. While the share of industry in overall employment has remained more or less constant, that of agriculture has fallen dramatically, the decline having being completely absorbed by services. In 1960, agricultural employment represented 38.7 per cent of total employment, while employment in services totalled 31 per cent. By 1998, the respective shares were 8.4 and 61.7 per cent. This represents, by any standard, a major structural change, differentiating Spain from the other more advanced European countries, and possibly influencing the overall unemployment rate.

For the Spanish economy as a whole, growth in employment between 1980-97 can be traced to consistent job creation in services (42,7 per cent), with jobs in the manufacturing industry declining between 1980-85 (by -18.9 per cent), recovering between 1986-90 (24.6 per cent), falling again between 1990-95 (-13.8 per cent) and recovering between 1995 and 1998 (3 per cent). The decline has been particularly pronounced in medium-low and low-technology industries, which have lost 22.6 per cent and 25 per cent of their workforce since 1980. However in the high-technology and medium high sectors, employment in 1995 was also 8.6 per cent and 11.3 per cent lower than in 1980.

The employment shares of each industry in total manufacturing employment have not shown a particular pattern of change in terms of technological intensity, as in the case of manufacturing production. The industries that have lost significant participation in total employment included those in which the state intervened (iron & steel, non-ferrous metals, chemical, and shipbuilding), as well as textiles and apparel & leather products.

It is important to point out that the worst feature of the Spanish economy is, without doubt, unemployment. During 1984 to1987 and 1993 to1998 the overall unemployment rate stood at over 20 per cent of the civilian labour force. Unemployment is particularly high among women, youth, and the long-term unemployed. Institutional restrictions in labour markets are not the only factors that have contributed to this dramatic feature (OECD 1995). Business and corporative barriers subsist and restrict the adaptation of the economy to a more competitive environment.

4. The Evolution of the Economy since EEC-EU Integration: Real Convergence.

In comparative terms it is interesting to evaluate the evolution of the less developed EEC-EU economies and, in particular, if they have been converging to the EEC-EU average. Real convergence is the progressive process by which less developed countries catch-up with the level of social and economic welfare of the more advanced ones. This process should occur in a context of sustainable growth and can be evaluated by: GDP per capita or hours worked; physical capital stock; R&D or technology capital stock; human capital and other indicators concerning social development.

Throughout the 1980s, in Spain, as well as Greece, Ireland and Portugal, these indicators have been consistently below the EEC-EU average. When Spain and Portugal joined the EU, their GDP per capita represented 70 per cent and 59 per cent respectively of the EU average (in comparison with Greece which accounted for 67 per cent in the same year). By 1998, Spain and Portugal grew to account for 78.7 per cent and 71.7 per cent respectively of the EEC-EU GDP average (compared to Greece which had maintained the same relative level). From 1986 to 1998, Spain and Portugal were the second and third fastest growing EEC-EU economies, with annual GDP growth rates of 2.9 per cent and 3.2 per cent respectively, only surpassed by Ireland, which has grown at some 6.6 per cent. Ireland's ranking within the EEC-EU in terms of GDP per capita has jumped from fourteenth to seventh position. Growth in Ireland, Portugal and Spain correspond with the convergence hypothesis, which states that the poorest countries catch up with the richest ones. The Greek case does not confirm completely this hypothesis yet.

Another indicator of real convergence is the level of labour productivity measured as GDP per hour worked. For Spain, this measure was 75.1 per cent and 91.4 per cent of the EU-15 average in 1986 and 1996 respectively, which shows that the gap between Spain and Europe has not been closed. The stock of capital as a percentage of labour employment was 72 per cent and 94 per cent of the EEC-EU average in the same years, which represented figures closer to the EU-15 average. In contrast, the stock of technological capital as a percentage of GDP was only 21 per cent and 45 per cent of the EEC-EU average in the same years. This reveals the yawning technological gap that exists between Spain and Europe. Spain's level of human capital (calculated as the population with higher studies) was 58 per cent in 1986 and 74 per cent of the EEC-EU average in 1996 (Bosca, De la Fuente and Domenech 1996: Human Capital and Growth: Theory Ahead of Measurement). More impressively, public capital stock per inhabitant has grown dramatically, from representing 39 per cent in 1986, to 77 per cent in 1996 of the EU-15 average. It is noteworthy that these increases were a result of public sector policies. The Spanish private sector did not play a predominant or reactive role in these investment processes, as it did in other European countries. Spain's limited progress in terms of its real convergence to EEC-EU productivity could be jeopardised if public sector investment is restricted for R&D, education and infrastructure, as a result of fiscal commitment to nominal convergence.

The Spanish economic structure has become more internationalised in terms of export/output ratios and foreign direct investment, although these are still lower than EEC-EU average levels. Exogenous shocks create asymmetrical adjustment in less developed countries that are part of the monetary union (De Grauwe 1997). This adjustment is more severe in cases of lower productivity as a result of lower levels of capitalisation in human and technological resources, and less productive and trade specialisation, in comparison with advanced developed countries which are more specialised in technology intensive and dynamic sectors. An economy with lower productivity must increase its physical, human and technological capitalisation in the long term in order to be part of the monetary union. Structural adjustment is not an automatic process and specific policies are required to correct the differential in human capital, R&D and social infrastructure.

5. Macroeconomic policies since EEC-EU accession.

The Spanish macroeconomic context in the years running up to the country's integration with the EEC-EU was affected by the turmoil of the economic and political transition from an interventionist and authoritarian regime to an increasingly liberal and democratic system. Spain's overall economic performance had been outstanding during the 1960s and in the early 1970s. The policies of the period combined a gradual economic liberalisation with strong state intervention and direct participation in public enterprises in various industrial sectors including the extractive, metallic and non-metallic mineral products, chemical, and transport (including shipbuilding, aircraft and motor vehicles) sectors. However, in the 1970s, as result of the international crisis and Spain's failure to adjust policies properly, the economic situation deteriorated substantially. At the beginning of the 1980s, measures to 'positively adjust' industries in decline were gradually introduced (including in the transport, basic metals, extractive, textile, electrical machinery and electronic components sectors). At the same time, other new key sectors were promoted. In general, policy could be described as defensive one, which consisted of cleaning up and converting to traditional industrial sectors. Recent studies have argued that these policies represented more continuity than change with the economic policies of the past (Simon 1997: 320-8).

Spain's integration to the EEC-EU in 1986 made way for a new institutional framework of economic stabilisation, liberalisation, promotion of R&D and technological cooperation. The process of European unification itself has had consequences for the Spanish macroeconomic context. From 1986 to the beginning of the 1990s, the country's economy underwent an expansive phase that corresponded to EEC-EU trends. This was followed by a recession that started in 1991 then a significant recovery after 1994. Since the integration of the Iberian countries to the EU, the European project has been largely dominated by nominal convergence to the criteria of monetary unification. This process of economic integration has increased the interdependence of the different economies, and of the political coordination of economic policies, in particular, due to the Single European Act (1987), the European Economic and Monetary Union (Madrid Council 1989) and the Treaty on European Union (Maastricht Council 1992).

II. Technology Policy and the National Innovation System

1. On the Need for Technology Policy

Technology is considered to play a vital role in shaping industrial performance since it affects productivity growth, creates and destroys jobs, changes skill requirements in the economy, and shapes the capacity of firms and industries to compete in international markets (OECD 1996b:9). An increasingly interdependent world economy only compounds the scenario whereby the introduction of new products or processes in any branch of any industry tends to render old products and processes obsolete or uneconomic. Firms wishing to survive and grow must be capable of adapting their technology based strategy to this competition. Thus, the promotion of technology, science and education has become a priority for many governments, and even a battle-field among the developed economies in the European Union, Japan and the United States (Freeman 1997:204).

Technology policy in the form of government intervention has been traditionally justified on the grounds of market-failure such as the presence of externalities, indivisibilities and risks. One principal argument is that most firms simply would not participate in R&D unless they were supported in some way and that basic R&D and related activities should be considered a public good (OECD 1995:17). Evolutionists also argue for government support of R&D and a broad research infrastructure which they claim is essential for the ongoing development of technological development (Sharp 1997:93). Since SMEs have particular difficulties in their efforts to innovate, technology policy is justified to help them overcome these diverse obstacles. Governments can chose from a variety of instruments by which they can promote technology including offering fiscal incentives, loans, grants and R&D subsidies. The focus of technology policy has broadened from its original aim to promote the introduction and acquisition of new technologies per se to also stimulating technology diffusion. This shift in focus has emerged in parallel to an increasingly broad understanding of technology itself. It has been convincingly argued that there is a circular link between technological innovation and diffusion, since when a firm's ability to absorb and use new technology is heightened, this in turn helps improve its ability to develop innovations itself (OECD 1996b:13).

One minimal definition of technology policy provided by the OECD (1996b:29) is the interpretation of a single indicator: the national R&D expenditure as a percentage of GDP, such as Gross Domestic Expenditures on R&D (GERD) or Business Enterprise Expenditures on R&D (BERD). In addition, there are other many ways of evaluating technology policy including government budget appropriations on R&D (GBARD), R&D personnel and patents.

Although patents are not necessarily the best indicator of innovation, since patent data may say more about product than process innovations (OECD 1995:56-7) they remain a useful indicator of innovation. Increasingly, however, certain analysts point to the inadequacies of considering R&D alone claiming that this is not the only method of evaluating technology policy. The initial focus on R&D as an indicator of technology policy was caused partly by a narrow conception of the process of technological innovation and partly due to the methodological problems in measuring other indicators of technological innovation. More recently, new techniques have been refined which guide the measurement of a broader set of indicators. Other factors which might be considered include incremental innovations, and the national system of innovation in each country. One role assigned to technology policy then would be to encourage the appropriate generic conditions that stimulate innovation which often means strengthening the national system of innovation by improving links between universities and other institutions' research departments with industry.

2. Technology Policy in Spain.

Real advances in technology policy in Spain became apparent at the beginning of the 1980s. At the time, Spain was undergoing huge political, social and economic turmoil and transformation, since it was leaving behind three and a half decades of the Franco dictatorship and entering a new democratic era. In technological terms, before the 1980s, national R&D expenditure in Spain was low (only one quarter of the OECD average) and there was a high degree of foreign technological dependence (Buesa 1997). Improvements in the 1980s were manifested by an increase of resources allocated to R&D and enhanced promotion of R&D, innovation and technological diffusion in enterprises by the responsible agents, namely, the National R&D Plans, as well as through tax incentives. Notwithstanding these gradual, visible improvements, an OECD report in 1987 on technological capability in Spain pointed to the way in which many aspects of Spain's technological infrastructure continued to be weak, highlighting the low levels of research spending by the universities, a reluctance on behalf of higher research institutions and businesses to cooperate, and the relatively high dependence upon the public sector for R&D expenditure accompanied by weak private sector R&D expenditure in comparison with EEC-EU averages (OECD 1987).

Spain's official membership of the EC in 1986 provided a window of opportunity for the further improvement of technology policy. First, Spain profited by the economic boom which accompanied its EEC-EU membership. Second, there was broad consensus that Spain must take strides to 'catch up' with the advanced European countries in technological terms in order to improve its competitiveness on an international level. Third, since Spain's membership of the EEC-EU coincided with the announcement of a single market programme and the development of a European Technology Community, this meant that the country could take advantage of the prospect of greater access to markets and increased chances for Spain to modernise its technology policy within the EEC-EU framework.

A range of legal and institutional reforms were introduced from 1986 which served to improve technology policy in general. These included legal reform enabling university research to improve relations with firms, the introduction of a new Spanish patent legal system and, most importantly, a bill on Science and Technology, which reorganised the national system of research centres and institutional arrangements for the coordination of technology policy. This framework was promoted as the basis for the establishment of a technology policy in Spain and it brought about a substantial reorganisation of the institutions in charge of the coordination of technology policy.

3. Institutions of the Spanish Innovation System.

The principal institutions which comprise the Spanish System of Innovation are represented in figure A. Until the creation of the Ministry of Science and Technology (MST) in April 2000, the Interministerial Commission on Science and Technology (CICYT) was the main instrument in charge of the overall coordination of scientific, technological and innovation issues. This CICYT was responsible for coordinating the policies and funds of the various relevant ministries, including that of Education and Culture, Industry and Energy, Public Health, Defence and Environment. The CICYT was assisted by an Office of Science and Technology (1998-2000) and managed by a General Council for Science and Technology which was in charge of national and regional policies. Regional technology policy varies in Spain depending on the capacity of the local authorities and the extent to which they have gained autonomy from the central government. So, while the CICYT draws up technology policy, the General Council is in charge of managing it at the national and regional levels. An Advisory Council, constituted by members of the scientific community (these members were required to be neutral experts in particular fields and worked temporarily for the government to provide advice on the planning and strategy of the National R&D Plans).

The National R&D Plans are drawn up every four years and subject to annual revisions. The National R&D Plan is divided into three main blocks: national, sectoral and regional. National programmes comprise two broad themes which are the improvement of the quality of life (policy blocks include R&D for the environment and biotechnology) and technologies of production and communications (including exploratory space missions and new communications technologies). Sectoral programmes include the promotion and finance of research which does not fall into the priority areas of the national plan (including researcher mobility, bi-lateral interchanges and research group cooperation with the EU) which fall under the Ministry of Education and Culture (MEC).

The role of the public corporation: Centre for Industrial Technological Development (CDTI) has been significantly extended since its origins in 1977. Since EC accession, it has been used increasingly to manage and finance technology policy through the financing of firms' R&D activities, the encouragement of Spanish firms' participation in international R&D programmes, and the promotion of technology transfer among agents including firms, universities and other research institutions. It is involved with three main kinds of projects. The first are "concerted" projects which must be pre-competitive research. These may be carried out in collaboration with public research centres or technological centres and may receive subsidises from the MST (formerly MINER) and finance from the Autonomous Communities (CCAA). The second and third projects are competitive research (PDT) and technological innovation projects (PIT) respectively. Both may receive MST (former MINER) subsidies and CCAA finance, and both must fit into with the strategic objectives of the national programme for the improvement of the industrial technology action plan (PATI).

The PATI is one of the key programmes in regard to technology policy, in particular, industrial technology and applied research. It operates on a horizontal and sectoral level and is designed to run in parallel with the national R&D plans. The PATI was originally introduced in 1991 with the aim of re-orientating science and technology policy in Spain towards a new and broader understanding of innovation, and especially, to improve the cooperation between scientific-technological links between the academic work and business research which were historically weak. The PATI's objectives were thus no less than to strengthen and mature Spain's National System of Innovation. Prime concerns include that there is a critical mass of agents involved in the innovation process including productive, technological, scientific and financial agents, and that these agents establish and use multiple interactions between them so as to develop long-term networks. Other key aims are that R&D activities are oriented so that results can be used in multidisciplinary activities, and that the diffusion of knowledge and technologies between the various agents are used so as to maximise the gains of their results (CICYT 1997). These objectives are accomplished using the following policies:

1. Through the development of interfaces between scientific, technological and productive agents by consolidating networks that already exist, such as university-industry foundations and technological centres, while also analysing the possibilities of creating new ones which would lead to increasing the dynamism of technological capacity. This is coordinated by the MST (former MINER) in order that an efficient national network of support of technological innovation is consolidated. This network will work closely with the Offices for the Transfer of Research Results (OTRIs). The OTRIs were established to improve the collaboration between public research institutions and private companies. The main function of the OTRIs is to improve the collaboration between public research institutions and private companies and they are located in universities, public research institutions and, more recently, in private industrial research associations. However, while the OTRIs are supply-driven, and concentrate on transferring technology from research centres to the private sector, the new network established under the PATI is demand-driven, focusing on transferring technologies to meet the specific needs of firms.
2. By encouraging cooperation between firms' R&D activities, technological centres (TCs) and public research centres (CPIs) in order to overcome inefficiencies brought about through efforts carried out in isolated or constrained by a lack of resources. The programme to stimulate the transference of research results (PETRI) has been strengthened and extra grants are available to help technology centres transfer their results. New financial support is available for R&D projects for joint development between firms and one or more technology centres. Partial funding will also go to pre-competitive projects which are multidisciplinary requiring the participation of diverse groups and firms.
3. Through the identification of technological demand by developing instruments which helped identify technological demands of the productive and social sectors so that the scientific and technological centres can pursue these demands, preferably in collaboration with firms and other interested bodies.
4. Increasing human resources with technological skills within firms, by facilitating the incorporation of technical and scientific staff into firms, as well as encouraging the management of the introduction of a technological culture into firms. New measures in the most recent national R&D plan limit the number of grant-holders by firm, and favour those applying to go to firms which have not yet benefited from the policy.
5. Encouraging the diffusion and dissemination of knowledge and technology to and between diverse productive sectors by facilitating the diffusion of R&D results from the scientific and technological agents to firms and between diverse industrial sectors, in order to ease the access of knowledge and technology by productive sectors so that they can develop products, processes or new or improved services.

One key horizontal policy is to increase the cooperation between technology policy-making and implementation within the different regions of the country. This is important, first, due to the unequal way in which R&D expenditure is spread around the country and second, to avoid a lack of coordination resulting in the overlapping of technological initiatives. In sectoral policies, one important policy is the PATI's support of SMEs, through the creation or extension of the role of Technological Centres, Centres of Cooperative Research between firms and R&D centres with the aim of establishing an advanced tertiary sector, constituted of engineering companies, R&D companies, technological consultancies and equipment companies. Policies towards SMEs are particularly important in Spain given the structure of the industrial sector since the average size of the Spanish firm is well below the EU-15 average. In Spain's industrial sector some 73 per cent of firms are micro-firms (defined as those with less than ten employees); 23 per cent have between ten and forty-nine employees; 3.4 per cent have between fifty and 249 employees and only 0.6 per cent have over 250 employees (EUROSTAT 1996a:15).

4. The National R&D Plans.

• Spain's first National R&D Plan was implemented between 1988 and 1991 and was strongly influenced by resemblances to other European R&D programmes. Among its principal aims were: to increase R&D expenditure as a percentage of GDP from 0.7 per cent to 1.2 per cent; to increase the number of R&D staff from 20,000 to 30,000 by 1991; to develop ways to support innovation on a regional basis; and finally, to increase the private sector's involvement in R&D expenditure. By the end of the first plan, some of its objectives could be said to have had considerable success. R&D expenditure as a percentage of GDP rose sharply during the 1980s, from a low of 0.4 per cent in 1981, and continued to rise while the plan was in place, reaching 0.9 per cent in 1990. This represented an increase superior to that of the average of EEC-EU and OECD countries (OECD 1997b:116) even if it did not meet the percentage set out in the national plan. Business R&D expenditure as a percentage of the total also rose from 42.8 per cent at the beginning of the 1980s to 48.1 per cent in 1991 (OECD 1997b:122). Another useful indicator, government budget appropriations or outlays on R&D (GBARD) rose in Spain from 0.34 per cent of GDP in 1986, which represented just over one third of the European average, to 0.39 per cent in 1987 to 0.55 per cent in 1990 (EUROSTAT 1997b:259). This was impressive, and contrasted with the same indicator in the EUR-15 which fell steadily through the period and beyond, from 0.98 per cent in 1987 to 0.86 per cent in 1997.         In other areas, however, the plan proved over-ambitious, and part of the problem was that the institutions lacked the experience and tradition of putting such policies into place. Indeed, the implementation of the plan brought to light inherent problems such as a shortage of technical skills, the weakness of industry-university links and poor technological diffusion. Efforts to improve technological diffusion included from 1989 the establishment of a network of technology transfer organisations, or OTRIs and an Office for Technology Transfer (OTTs) throughout the country. The main function of the OTRIs and OTT is to improve the collaboration between public research institutions and private companies. The OTT devolves to the Secretariat of the National R&D plan and is responsible for the co-ordination and assistance to OTRIs. In 1995, there were around 88 OTRIs across the country and were registered in the following way: universities (46), public research institutions (12) and private non-profit making entities (30). Among their functions are:

1. To create a database of research fields and facilities in public research institutions.
2. To identify results of publicly funded research projects, assess their transfer potential, and disseminate and diffuse this to enterprises.
3. To prepare research contracts and technical assistance contracts.
4. To provide information about EEC-EU R&D programmes and help with the preparation of proposals.
5. To assist with the exchange of research personnel between industries and the public research institutions.

• The second National R&D Plan (1992-1995) sought to tackle some of the problems identified in the first plan and included a proposal to further increase the percentage of GDP destined for R&D expenditure. However, most of the efforts embodied within the plan were negatively affected by the recession between 1991 and 1993. Financial support for innovation and diffusion programmes in the Ministry of Industry fell by 35 per cent overall from 1991 to 1995 (Sánchez 1997).

Since 1994 there has been evidence of an economic recovery, however, R&D statistics have not yet been influenced in a positive way. R&D expenditure as a percentage of GDP increased from 0.7 per cent in 1987 to 1 per cent in 1992, then fell to 0.86 per cent in 1995 and 1997. Another aim of the second plan was, mirroring EU developments, to withdraw the government's intervention in technology policy by further shifting the responsibility of the R&D financing onto the private sector. In the EU, the share of government R&D expenditure as a percentage of total R&D expenditure dropped throughout the 1980s and 1990s; this declined from around 46.7 per cent of the total national R&D expenditure in 1981 to 39.2 per cent in 1995 (OECD 1997b:122). However, in Spain, business R&D expenditure as a percentage of the total, which rose during the first national plan to 48.1 per cent in 1991, fell back to 40.3 per cent in 1997 (OECD 1997b:122). In terms of R&D performance, Spain differs from many countries where the business enterprise sector accounts for around two-thirds of R&D performance such as in the US, Canada, Japan, France, Germany and the UK. In Spain, business enterprise performance increased between 1981 to 1989 from 45.5 per cent to 56.3 per cent, but then fell to 45.9 per cent in 1995 (OECD 1997b:120, 48.8 per cent in 1997:INE 2000). Whereas around half of R&D personnel performance are found in the business sector in the EU, this figure is around 33 per cent in Spain.

Some analysts argue that the aim of reducing government intervention proved difficult because, while Spain's membership of the EEC-EU brought about increased competition within Spanish industry, it also exposed its technological backwardness, and this led to the Spanish government taking an increasingly active role in promoting technology within industry (Farrell 1997:211). In order to survive the EEC-EU integration there was more, rather than less, intervention on the part of the Spanish government in industry and the technological structure. However, another important reason of the need for government intervention is that the business sector reacted to the crisis by over-reducing its R&D expenditures. Thus, by filling this void, the importance of the government sector in R&D expenditure continues to be a feature of the Spanish technology system in common with other less developed European countries (Díaz Fuentes & Clifton 2000).

One positive result that occurred during the first and second plans was the continued growth of the number of R&D personnel as a percentage of the total labour force. In Spain this grew rapidly from 3.3 per cent in 1987 to 5.5 per cent in 1997 (Table 8), though not as rapidly as in Ireland (EUROSTAT 1997b:328). However, by 1999 the percentage in Spain still lagged far behind the EUR-15 average.

The new government which came into power in 1996 has stated in the third National R&D Plan (1996-1999) its objective to recover the efforts made in the 1980s to stimulate R&D and innovation among industrial firms particularly to SMEs, multi-national companies (MNCs) and firms in sectors with comparative advantages to export. The need is emphasised to improve the diffusion of technology which is to be accomplished in two main ways; first by broadening out the original limited set of industrial sectors targeted to include all types of manufacturing firms and second, by permiting an increasing decentralisation of policy. This latter policy is crucial in Spain where R&D expenditure is highly uneven across the country and is concentrated in Madrid, Catalonia and the Basque Country (INE 1997b:60 and 2000). One suggestion made by a senior spokesman of the CDTI is that the success of the networks of technological parks in the Basque Country should be emulated as a model in the rest of Spain (Ramón Alique 1997). Meanwhile the Secretary General of the National R&D Plan has stated that technology policies will increasingly decentralised by enabling regional governments to develop their own diffusion strategies in order to improve the technological capacity of firms within their own territory, usually in alliance with regional development agencies and financing (Sánchez 1997).

• The third National R&D Plan has further extended the role to be played by PATI in its efforts to encourage the development of the Spanish science and technology system and chiefly, to try and bring Spain's technological capacity up to the levels of other advanced countries by supporting individual and collaborative research and technological activities. In order to lessen Spain's technological dependence, the government has doubled its funding to the EU Fourth R&D programme to encourage international technology transfer. Another of the aims of this plan is to improve the access and information available to Spanish firms that wish to apply for patents. Spain obtained the second highest record in the level of increased patents to the European Patent Office between 1989 and 1997 after Portugal. However, when European patent applications per one million population are calculated, Spain has low figures (ten), ahead only of Greece (four) and Portugal (two) (EUROSTAT 1996b:102). Like R&D expenditure, patent applications are highly concentrated in three main regions (Díaz Fuentes et al 2000).                                   The re-elected government has put forward its broad aims for the period 2000-2003 in the new fourth National Plan, that is called (acquiring more adjectives) the National Plan of Scientific Research and Technological Development and Innovation" (SRTDI). Additionally, another important institutional reform was the creation of the Ministry of Science and Technology, while the Ministry of Industry and Energy has disappeared. The MINER was until recently the main manager of Spain's National R&D plans. This change implies not only a simple institutional or organisational change. It implies a new approach towards "innovation and technology policy" that goes beyond "industrial policy". However, the government's new plans have certain continuity with previous ones; on the one hand it is assumed that scientific research is a public good to be shared by different agents and, on the other hand, there is still an effort to create favourable conditions for business innovation culture.

1. The impact of innovation on international competitiveness

There is a widespread acceptance that the competitiveness of an economy depends as much on firms' decision as on the public policy on technological, physical, infrastructure and human capital. As mentioned in the first section, the competitiveness of the Spanish economy was influenced by the evolution of the real exchange rate in the commitment to the European Monetary System (EMS). The real exchange rate is commonly considered the main determining factor of the competitiveness of an economy, but there are other non-price factors which also determine the competitiveness of an economy. In this sense, certain real variables of competitiveness are analysed: first, the competitiveness based on the Spanish shares on OECD total exports, which shows the participation of certain industries for Spain in the international markets of this industry; second, the Spanish export shares of production in a given activity; third, the import penetration or imports as a proportion total domestic demand of a certain activity; fourth, the ratio between export and imports by sector, or trade balance ratio; fifth, the export specialisation that is calculated by comparing the Spanish and OECD export structure, and is designated "revealed comparative advantages"; and sixth, intra-industry specialisation ). The analysis of the export shares from 1980 to 1995 shows that Spain has increased its exports more than the EU and OECD. This has been one of the results of the trade liberalisation process and indicates a stronger participation of the country in international markets. The increases have been more important in the high technology sectors than in the medium or low technology ones.

The export shares of production indicators indicate a steady growth in manufacturing exports that overtake production growth. This trend indicates an increasing openness of the economy that has been stronger in the high and medium-high technology sectors than in the medium-low and low activities. The export shares on international trade or domestic production are convenient measures but should be contrasted with import penetration. This variable indicates the domestic capacity to provide certain activity. In the Spanish case, import penetration has increased significantly (from 12.3 per cent to 23.4 per cent), in particular after the EC integration. Although, in absolute terms, the increase was more significant in the high and medium high intensity sectors, in relative terms, import penetration increased more in the lower than in the higher intensity activities. At the same time, the economy has been increasingly exposed to competition in those sectors with lower import penetration before EC integration. Since 1986, Spain has increased the import penetration more than any other EU country, but this is partly because Spain was the EU country with the lowest level of import penetration.

The export import ratio expresses the trade balance surplus or deficit in percentiles. In general, this deteriorated after the Spanish accession into the EU. This deficit affected most industries, particularly between 1980-90. This was more dramatic in the low and medium-low intensity industries. In these activities, the deficit increased in the period 1991-95, but it improved in average in the high and medium-high technology activities. In fact, if we assume that the export import ratio indicates a trade specialisation of the economy, there was a change in the trade pattern. The low technology activities showed a ratio 1.76 in 1980-85 which passed to 0.74 in 1991-95, while high technology activities increased from 0.37 to 0.47 in the same period.

The export specialisation or revealed comparative advantages indicator confirms this pattern. The Spanish economy improved its position in medium-high and high technology activities but in medium-low and low technology activities the situation deteriorated. At the sectoral level, among dynamic sectors, there is an increasing specialisation in: motor vehicles; drugs and medicines; electrical apparatus; and aircraft. Meanwhile most of the traditional activities continue to reveal the comparative advantage of the country, specialised in labour and natural resource intensive activities. But these advantages have lost significance as a result of trade liberalisation (EC and GATT). This process exposed the Spanish economy not only to European, but also to international competition, particularly to the Third World and the NICs (which have the advantages of unskilled labour and natural resources). Finally, the intra-industry specialisation indicates if the country is simultaneously an exporter and importer of the same activity (near to 100), or a net exporter or importer (near to 0). In general, intra-industry specialisation has increased, and it indicates a higher level of imports in low and medium-low industries, which in the past were net exporters. In contrast, the high and medium-high technology activities indicate the increasing demand of imports which increases international competitiveness.

One of the main objectives of technology policy in Spain is to increase the competitiveness of its industrial sector by attempting to forge each sector's convergence with EU levels. Technology intensity is calculated by the OECD (1996a,1997a and 2000) using three indicators, which reflect to different degrees the 'technology-producer' and 'technology user' aspects. On the basis of these three criteria, the OECD classification of industries according to their technology intensity have been contrasted with our own estimation for the G-7, the EU-13 and for Spain during the period 1980-95. Overall, it was found that the Spanish technological intensity indicators correspond very closely with OECD-10 classification.

Once this framework for comparison has been established, it then becomes possible to examine the extent to which the different manufacturing categories in Spain have caught up with EU-13 technological intensities . As a general observation, it was founded that total Spanish manufacturing technological intensity represented 18 per cent of the EU-13 level in 1980-5. This increased to 38 per cent in 1991-5. In this period, however, the increase was particularly significant in the high-technologies industries which increased from 24 per cent to 62 per cent and low-technology which increased from 29 per cent to 60 per cent. The increase in the medium technologies was less overall. Low-medium technology increased from 34 per cent to 62 per cent, however, this was biased due to the shipbuilding sector. High-medium technology only rose from 22 per cent to 34 per cent. In all categories, with the exception of other transport equipment, technology intensity increased in comparison with EU-13 between 1980-5 and 1991-5. This catch-up was more accelerated between 1980-90 than between 1991-5.

4. Conclusions on Technology Policy and EU Convergence

In general terms, public support to R&D business expenditures through funds and loans prioritised government related activities and, some high and medium technology intensity industries, predominantly SMEs, like electronic components, professional goods and machinery. Regional and EU funding complements this national support, in particular to SMEs, certain low technology industries, and activities undergoing reconversion.

Structural characteristics of the Spanish industrial sector can be used to explain some of the technological deficiencies of the economic system. In particular, the predominance of SMEs means that it is difficult to perform R&D and innovation, and certain public sector companies in declining industries. Other political and institutional factors have influenced the uneven distribution of R&D resources across the country, with two thirds of Spain's R&D expenditures being concentrated in Madrid, Catalonia and the Basque Country.

Using new statistical sources, our sectoral analysis of technological policy in Spain has examined firm size, different indicators of technology intensity, and instruments by which the government promotes technology with different sectors. The evaluation of technology intensity in terms of GDP, production and investment using OECD (1996b and 1997a) data demonstrates three characteristics of the Spanish manufacturing system. First, classification by activities of Spanish manufacturing firms corresponds with EU-13 and OECD-10 classifications, and this shows that Spain has a similar pattern in terms of technological specialisation, albeit with a significant technological gap. Second, there has been some catching-up in Spain with EU levels in terms of technological gaps of the different manufacturing activities, particularly in the six years after the EC integration in 1986. Third, this convergence was more intense in the high and low technology intensity activities than in the medium ones.

Once Spanish manufacturing activities had been found to be compatible with the OECD and EU-13 classifications, this facilitated extending the examination of the technological activities and intensities of manufacturing industries using national sources particularly MINER and INE. In this way results from both sources proved the tradition assumption that large firms are more active in R&D and related activities than SMEs in nearly all sectors in Spain. At the same time, there is a general inverse correlation between the innovative and R&D enterprises and the structure of the markets by sectors. Second, both sources showed that the R&D intensities and relative employment are usually higher in large firms than in SMEs. However, when technology intensities were measured for innovative firms only, R&D relative levels were nearly always more intense in SMEs than in large firms.

INE statistics showed additionally a general inverse correlation between sectoral market structure and innovation and R&D intensities. The MINER and INE results correspond with OECD-10 technological evaluation by sectors, and support the idea that SMEs lag far behind large firms in terms of R&D, but that many innovative SMEs are more active according to specific sectors (OECD 1993 and Freeman & Soete 1997: 230). Finally, the evaluation of the Spanish technological policy based on the MINER surveys showed that medium firms receive more support than larger ones and SMEs. R&D public funding targets included activities undergoing industrial reconversion, medium and low intensity industries. The national and regional funding put in evidence a defensive industrial and technological policy, however the EU supported and favoured the cooperation in some high and technology activities.