Juan R. Cuadrado-Roura
Universidad de Alcalá

Resumen El enfoque centrado en la periferia ha generado mucha literatura que hoy en día podemos considerar superada. Generalmente se centraba en problemas de accesibilidad y sus consecuencias económicas. Sin embargo, este artículo no sigue un enfoque convencional. Pese a que toma en consideración los problemas relativos al transporte y la accesibilidad como preocupaciones constantes, existen otros factores cruciales que caracterizan la "perifericidad". Dichos factores implican, entre otros, capacidad de innovación, gasto en investigación y desarrollo e inversión en capital humano a nivel regional. Tras realizar algunas consideraciones iniciales, este artículo se centra en el análisis empírico donde evaluamos el comportamiento de regiones centrales, periféricas e intermedias de una muestra de la Europa de los 15. Los resultados muestran grandes diferencias entre estas regiones pero también una heterogeneidad considerable entre estos tres grupos. De nuestro análisis extraemos algunas conclusiones y las relacionamos con los desafíos actuales a los que tienen que enfrentarse todas las regiones pero con más esfuerzo las periféricas. El artículo concluye con unas cuantas observaciones finales sobre posibles acciones y la necesidad de establecer una cooperación entre las regiones a nivel europeo. Abstract The centre-periphery approach has generated large literature that we can now consider overcome. It was generally focused on accessibility problems and its economic consequences. However, this paper doesn´t follow a conventional approach. Although it considers transport and accessibility problems as core concerns, there are other crucial factors that characterize the ´peripherality´. These factors involve innovation capacity, research and development expenses and human capital investments at regional level, among other. After some starting general considerations, this paper focuses on the empirical analysis where we evaluate the behaviour of central, peripheral and intermediate regions of a EU-15 sample. The results show big differences among these regions but also quite heterogeneity within these three groups. From our analysis we extract some conclusions and relate them to actual challenges all the regions face but more severely the peripheral. The paper closes with a few final remarks and some considerations on possible actions and the need for cooperation both between regions and on a European scale.

*Text presented at the Uddewala Symposium, Sweden, 2004 1. Introduction    In the last fifteen years the regional aspects of economic activities have re-entered in the economic debate. Under labels like ´innovative milieux´, ´industrial districts´, and ´regional innovative systems´, as well as a consequence of the new approaches to the economic growth theory, it has been frequently analysed why certain regions are economically successful while others are not. The basis for these approaches are analysis on regional behaviour considering or a high number of regions or else case studies of several successful regions, like Silicon Valley and the Third Italy1, just to name two of the most prominent examples. On the basis of these case studies several authors have attempted to explain the specific reasons for the success of some regions compared to other not so successful.    This paper deviates from most of the approaches in the literature in two ways. First, it focuses on a redefinition of the concepts of ´centre´ and ´periphery´, which should be based not only on geographical terms but also on other elements, particularly on innovation activities and education. Second, the paper does not focus on one or a few specific peripheral regions but on 32 European regions (EU-15), not only peripheral but also intermediate and central ones, so qualified according to the conventional approach based on accessibility. The project in progress will include 6 intermediate and peripheral regions from the New Members and 3 from the two countries expected to be integrated in the EU in 2007. Unfortunately, the actual figures and indicators do not permit to incorporate such regions in this analysis.    The paper is partially based on a recent analysis which focuses on the general features of the dynamics of ´peripherality´2 and the main purpose is to show that in addition to distance and transport, other factors are so important to qualify (or not) a region as peripheral. If some indicators on innovation, R+D effort, educational level, etc. are taken into account, not all peripheral regions can be considered as strongly ´peripheral´ and, on the other hand, some regions qualified as ´intermediate´ and even ´central´ from a conventional point of view appear to have symptoms of peripherality in terms of their capacity to joint the new trends of modern economies in an increasing competitive atmosphere.    The actual challenges facing the periphery of Europe as well as the risks and opportunities for regions are also considered in this paper, all this giving room to some suggestions on the innovation as a response, the challenge of competing with central regions, and some final reflections on interregional cooperation and future perspectives.    The text is organised as follows. First, an EU territorial picture is presented (section 2), to give way to some reflections on ´peripherality´, the limits of the conventional approach centre-periphery and some aspects which seem to be actually crucial for peripheral regions (section 3). To approach the degree of ´peripherality´ an empirical analysis is proposed (section 4) taking 32 EU regions as a sample: 10 central, 10 intermediate and 12 peripheral according to the European Commission assessment. The results of the analysis are shown in section 5, paying special attention to an approach in terms of economic and social cohesion and considering also some specific indicators on innovation and human capital. The final part of the paper (sections 6 and 7) set up some challenges facing the periphery of Europe and the possible responses and actions. 2. The EU territorial picture    The emerging picture of the EU-15 is one of a very high concentration of economic activity and population in the central area or pentagon (which stretches between North Yorkshire in England, Franche-Comté in France, Hamburg in northern Germany and Lombardy in the north of Italy) (Third Report on Economic and Social Cohesion, 2004). As the 2nd Cohesion Report pointed out3 central regions in this area cover 18 % of the EU-15 territory while accounting for 48% of GDP, 41% of population and 75% of expenditure in R&D. Population density in these regions is 3.7 times higher than in peripheral regions. In all but 10 of the 88 central regions (NUTS 2 level) GDP per head is above the EU-15 average (2001 figures), while all but 22 of the 111 peripheral regions have a level below the average. Average GDP per head in the central regions is more than twice as high as in the peripheral ones and productivity 2.4 times higher. In 2001, expenditure on research and development amounted to 2.2% of GDP in the former as against 0.9% in the latter. In 6 of the 7 ultra-peripheral regions, GDP per head was only around half the EU average.    As underlined in the 2nd. Cohesion Report, the point on R&D is especially pertinent. The structure of production costs of firms has changed considerably in recent years, with the fixed costs of research and development increasing and costs incurred on transport declining. R&D along with other strategic, high-value-added activities tends to be concentrated in central regions where the know-how and specialized infrastructure are located. This is a factor underlying growing polarisation in the EU and the concentration of low value-added activities in peripheral areas.    The transport system is also more developed in the pentagon area. The density of motorways is four times greater in central regions than in peripheral ones, while there are also 40% more railways lines and twice the length of double-track lines. However, there are some signs of the relative positions changing, especially in areas on the periphery where the road system is most developed and is continuing to expand, which are tending to become important access points, such as Lisbon and Centro region (Portugal, Andalucia (Spain) and Attiki (Greece).    The sectoral pattern of employment is too very different in central as opposed to peripheral regions. Although the share of employment in industry and construction is much the same (around 30% of the total), the share of employment in agriculture in peripheral regions is seven times larger than in central areas, whereas employment in services is only 53% of the total as against 69%. This, of course, reflects underlying competitiveness, which helps explain why the employment rate in peripheral regions is under 59% while in central ones, it is just over 67%.    This concentration of economic activity and population in such a restricted area of the Union has adverse effects not only on the peripheral regions but also on the central ones, where it is responsible for traffic congestion and strong pressure on the environment. Whereas transport bottlenecks in peripheral areas are a result of the low standard of infrastructure and a lack of connections, in central regions, they arise from capacity constraints and excessive traffic4. A consequence of this congestion and the concentration of economic activity is that toxic emissions in central areas are 2.3 times greater than in peripheral ones.    With the accession of the 12 new countries (10 in 2004 and 2 more in 2007), the EU will include many more areas where the level of development is well below the average (see Table 1). A new eastern continental periphery will be added to the existing Southern, West Atlantic and Northern. As a result, economic activity would surely tend to be even more regionally concentrated than in the US, where activity is more evenly distributed, despite its land area being twice as large as an enlarged EU and its population being much smaller (270 million inhabitants, 44% less than in the Union). 3. What about ´peripherality´ ?    3.1. Centre-periphery. The conventional view    Centre-periphery patterns of the global economy have been investigated by the structuralist school of development studies since 1950s. Yet this approach has not gained much ground in mainstream regional analysis. In fact, the centre-periphery antinomy is usually employed without any explicit linkage to a specific paradigm in the analysis of regional disparities and locational patterns. It simply refers to the spatial concentration of activities and related disparities, leaving patterns of power and dependency aside. At the same time, it is noteworthy that the structuralist school has avoided penetrating the regional level in their analysis. From a narrow and statistic perspective, peripherality boils down to the problem of accessibility. Distances give rise to transport needs which implies real costs to be borne by somebody. This advantage is increased further by the fact that scale economies in production and in the use of infrastructure cannot always be developed and utilised to the same extent in peripheries as in centres of where there are large numbers of potentially mobile people and numerous socio-economic activities. Centre and periphery patterns in the location of activities derive, to a large extent, from the intrinsic characteristics of economic processes5. The spatial concentration of demand and the network nature of transport infrastructure jointly tend to sustain growth processes, which have for some reason been initiated in a certain region. The resulting spatial polarisation is based essentially on increasing returns, and the process of making use of these increasing returns represents cumulative causation in the classical sense of Hirschman and Myrdal.    In the centre-periphery literature (see for ex. Rokkan and Unwin, 1987), there is seldom a convincing treatment of transportation systems and their supporting physical and organisational networks. The central node in theory is apparently so dense that transportation is not a problem. ´Peripherality´ generally connotes distance, difference and dependency instead. A typical periphery is geographically remote, economically lagging, dependent upon external political and industrial decision-making, and frequently culturally obsolete. In this vein, the paired antinomy ´centre versus periphery´ is often used to loosely characterise asymmetrical relationships and the regional disparities persistence.    As many as Keeble et al. (1988) have pointed out, ´peripherality´ is also synonymous with relative accessibility and inaccessibility to economic activity. The main "product" of a transport system is accessibility. It determines the locational advantage of an area (i.e. a region, a city or a corridor) relative to all areas, including it self. Indicators of accessibility measure the benefits that the firms and households in an area enjoy from the existence and use of the transport infrastructure relevant for their area. The important role of transport infrastructure for spatial development in its most simplified form implies that areas with better access to the locations of input materials and markets will, ceteris paribus, be more productive, more competitive and hence more successful than more remote and isolated areas6.    This relationship has been taken up in the European Spatial Development Perspective (ESDP)7 which gives high priority to improvements in accessibility as a policy target: "Good accessibility of European regions improves not only their competitive position but also the competitiveness of Europe as a whole".    However, the relationship between transport infrastructure and spatial development is becoming ever more complex. There are of course a number of successful regions in the European cores that confirm the theoretical expectation that location matters. However, there are also centrally located regions suffering from industrial decline and high unemployment. On the other side of the spectrum the poorest regions, as theory would predict, are at the periphery, but there also exist prosperous peripheral regions such as those in the Nordic countries and a few in the Southern Europe. To make things even more difficult, some of the economically fastest growing regions are among the most peripheral ones (Balearic Ils., Canary Ils., North of Scotland….)    The European Commission seems to have a very simple, traditional view on the question of central and peripheral regions in Europe as expressed in the 2nd Cohesion Report. Although based on a detailed study with a very broad range of accessibility indicators for NUTS 3 regions by car and lorry8, the categories describing relative location in Europe are collapsed into three: central, other and peripheral. The 3rd Cohesion Report includes also information about the potential accessibility by road, by rail (see maps in the Annex) and by air. Whether this reflects the reality and needs of the European periphery and whether this is an appropriate starting point for regional policy is thus rather questionable.    3.2. Is transport/accessibility so crucial?    There is a growing awareness, amongst both academics and in policy makers, that the concepts of ´centre´ and ´periphery´ can no longer be satisfactory conceived of in purely geographical terms. Within the present-day globalised economies and societies, there is a need for redefinition of these concepts. Starting from the ´innovative milieu´ approach and some other theories and ideas (on innovation; technology diffusion; the actual role of human capital in growth theories, etc.), certain economists and geographers argue that the degree to which a region can be considered ´peripheral´ is no longer determined solely by its distance to markets, resources and decision-making centres. These aspects are still important, but to assess whether a region can be considered ´peripheral´ it is also necessary to look at different factors that determine to what extent a region is capable of interacting and integrating with global markets, societies, and networks. The exact characteristics and the importance of these factors for individual regions are of course to be determined by specific research, but they are likely to encompass indicators of a region´s degree of entrepreneurship, of its communication infrastructure and networks, of the diversification of its economic structure, of its capacity for developing and implementing innovative products and production methods, etc.    Numerous investigations conducted through various academic disciplines have concluded that relations between the dimensions of ´centrality´ and ´peripherality´ are complex. A country or a region, which is peripheral in one field, can be central in another. A region, which was one a periphery, is not necessarily a periphery forever. Eskelinen and Snickars (1995) claimed that geographical ´peripherality´ is not necessarily a fatal syndrome, implying dependency, backwardness and isolation. Peripheral regions can, to an increasing degree, provide viable platforms for economic interaction and political cooperation. Some of them do actually have relevant handicaps and problems to be solved. But, they can strive to become competitive in the new political and economic order, and they are far from being outstripped in this race.    As pointed out above, in my opinion the concept of ´peripherality´ of a region is not only a geographical notion but mostly an economic one. A peripheral area will be identified by the capability to integrate regions into the global market. Innovation plays an important role in this context. The ´peripherality´ of a region cannot be longer determined by accessibility and geographical reasons only. These aspects are still crucial, but it is also necessary to look at different factors such as innovation efforts, entrepreneurships, receptiveness for modernisation and learning capacity.    Increasingly, innovation is regarded as an evolutionary, non-linear and interactive process between the firm and its environment9.    (i) The concept of non-linearity implies that innovation is stimulated and influenced by many actors and sources of information, both inside and outside the firm. It is not only determined by scientists and engineers working in R&D or the top management. In addition, there are interactions feeding the experience of production, marketing, and of customers into earlier phases of the innovation process.    (ii) The interactivity of the innovation process refers to the internal collaboration between several departments of a company (R&D, production, marketing, distribution, etc.) as well as to external cooperation with other firms (especially with customers and suppliers), knowledge providers (like universities and technology centres), finance, training, and the public administrations (regional, local and national). These are all contributing to the firm´s capacity to innovate and to support the regional innovative process.    It is in this context that the concept of ´innovation systems´ has been introduced. According to Lundvall (1992), an innovation system "is constituted by a number of elements and by the relationships between these elements which interact in the production, diffusion and use of new, and economically useful, knowledge". It is a dynamic social system based on the central activity of learning. Social interaction is, however, not necessarily promoting processes of learning and innovation, but might also lead to constellations blocking such processes.    Industrial economists have demonstrated that the nation is a very important level for innovation systems to be located. Institutions are strongly interlinked within national borders. Innovation occurs along specific trajectories, shaped not only by companies but also by particular research environments systems of education, finance, and regulation of a country10.    But as studies on innovative regions have shown, it is also possible that under certain conditions the innovation process becomes "embedded" in the region11 leading to the formation of regional innovation systems12. The following factors and mechanisms have been identified:    (i) Regions differ in their preconditions for innovation such as qualification of the labour force, education, research institutions, knowledge externalities and spillovers. Many of these factors are immobile, giving some regions advantages over other.    (ii) Industrial clusters often are also localised, giving rise to networks and specific innovation patterns in regions13.    (iii) A common technical culture may develop in local production systems through collective learning leading to innovative milieu14    (iv) University-industry and University-services links and knowledge-spillovers often lead to regional high-tech development15.    (v) Finally, regulations and support by regional development agencies and policy is relevant both for small and large firms in the innovation process16.    Then, elements of a regional innovation system, then, are first of all firms of the main industrial clusters of the region, including their support industries. They constitute various kinds of networks, both within the region and to the outside world (supplier/client, cooperation, and information-networks) through which relevant information flows and interactions occur. Research institutions (R&D organisations, laboratories, and universities) act as knowledge suppliers. They only become effective in the region, however, if their supply fits the demand of the firms.    Furthermore, innovations and their success depend to a high degree on the quality of the labour force. Here, not just R&D expenses and personnel is relevant but also qualifications in production, marketing and management. As a consequence, training organisations are another important element of a regional innovation system. Financial institutions have to be mentioned, providing finance for innovation projects to firms in the region. Last but not least, industrial associations and institutions like business innovation centres, science parks, or technology transfer centres support particular segments of firms (e.g. SMEs of start-ups), helping them to overcome obstacles in the innovation process.    Innovations, understood as new competitive compositions of the value chain primarily originate from the dynamics within and around the value chain itself. Innovative networks are the main drivers of both corporate and regional competitiveness. For regions, innovative networks among the companies of the region, their networks outside the region, the mobility of the labour market, and to some extent the consultancy and academic base, are significant factors of growth and competitiveness17.    And, all this is also true and can be applied to the regions located in the ´periphery´ of a integrated group of nations like the EU, besides taking distance and transport as the most important obstacles to be faced. These last factors are important, of course, and in some cases they are particularly relevant to understand the lack of economic growth of a peripheral area. But, if some other elements are taken into account - like innovation capacity, education, R+D efforts, etc. - it is possible to see both some ´non-peripheral´ regions from a geographic point of view, which economic outcomes are rather weak and, simultaneously, some ´peripheral regions´ from de point of view of transport which are actually rather dynamics, developing competitive activities and achieving higher and steady rates of growth. 4. Approaching the Degree of ´Peripherality´ of Regions in Europe    As it has been pointed out, the European Commission appears to have a very traditional view on the question of central and peripheral regions in Europe. The analysis made was based on a detailed study of accessibility indicators for NUTS 3 regions by car and lorry18, which concludes setting up three regional categories: central, peripheral and other (intermediate). To determine the position of regions in this three categories an integrated accessibility index have been calculated, measuring the time each region needs to get the other weighted by its economic importance. Then, central regions are the ones which accessibility index is higher than 50% of the EU average; peripheral regions do have an accessibility index below 40% of the average; and the intermediate regions have an index among 40% and 150% of the Union average. Peripheral regions appear being located in northern Europe (Sweden and Finland); the western Atlantic (North of Scotland, Ireland, Portugal) and Southern Europe (Spain, the Mediterranean Ils; Southern Italy and Greece; and on the West: a good number of regions in the new Member-states.    Given our interest to analyse not only peripheral regions, but to compare them with some central and intermediate regions using specific indicators to approach economic situation and innovation-related figures, some decisions have been taken about the regions to be considered and the data and indicators to use.    4.1. Regions    According to the main goal of our research, that is to investigate some characteristics of European regions either being "Central", "Intermediate" or "Peripheral" on EU scale, in the pilot phase we use a spreading sample of 10 Central, 10 Intermediate, 12 Peripheral regions of the EU-15, and 9 more peripheral regions from the new State Members and candidates.    The 41 regions selected (see Table 2) have been identified on a systematic and objective base; the majority of them are NUTS-2, that is the level defined by Eurostat on the basis of existing institutional arrangements in the Member State concerned and by agreement with the national authorities. Nevertheless, there are a few cases where regional level is NUTS I due to the lack of suitable and complete information at a lower level for the period 1987-2000/2001. This was the case of the first elected region of Antwerpen (Belgium) where Vlaams Gewest has been taken as a proxy; of Ruhrgebiet (Germany), where Nordrhein-Wetfalen has been taken; Berkshire Bucks Osfordshire and Highlands & Islands (UK), where the South East region and Scotland have been taken in some cases as the reference, respectively. Finally, Ireland and Denmark have been considered in each case as a single region due also to a lack of suitable official statistical information.    Then, there are 32 regions from the EU-15 countries (at least one per country), plus 6 of some of the new countries, and 3 from Bulgaria and Romania, expecting to be members in 2007. If more than one region was selected in a country, they should preferably have different positions within the country (central, intermediate, periphery). It must be pointed out that due to the lack of sufficient and reliable figures on the 9 Eastern regions, their analysis is postponed to the second phase of our research.    4.2. The data    As for the figures and indicators used, a number of European data base have been employed.    First, the indicators and figures produced regularly by Eurostat, within the base REGIO, Newcronos. Specifically, with this is possible to start from 63 different variables, the details of which are shown in Table 3. The data we have taken into account cover the period 1987 to 2000 (or 1999 as proxy) and includes aspects of demography, economy, employment and unemployment, R+D, transport, energy, life conditions, education, etc.    Second, the data included in the 2nd Progress Report on Cohesion (January,2003), updated by the Commission. It includes figures on demographic changes, levels of education, employment by sector, knowledge society and productivity. All of these are important factors explaining disparities among the regions. New data on indicators of technological advance confirm a picture of lower activity relating to technological innovation and the growth of knowledge economy in the Southern member states, compared to the Northern ones. It is important to note that we could not manage to find any relevant statistical data for the Danish region of Fyn, firstly selected as of our sample. However, country-level data for Denmark were available and we incorporated them in our research work. As a consequence of this, we considered Denmark as a whole to be an intermediate region for comparison.    Third, the data from the European Innovation Scoreboard (EIS), developed at the request of the Lisbon European Council in 200019. It focuses on high-tech innovation and provides indicators for tracking the EU´s progress towards the Lisbon goal of becoming the most competitive and dynamic knowledge-based economy in the world within this decade. The EIS contains 17 main indicators, selected to summarise the main drivers and outputs of innovations. These indicators are divided into four groups: Human resources for innovation (5 indicators); the creation of new knowledge (3 indicators, of which one is divided into EPO and USPTO patents); the transmission and application of knowledge (3 indicators); and innovation finance, outputs and markets (6 indicators). Unfortunately the EIS does not provide trend results for these indicators and it does not contain a summary innovation index similar to the one offered in 2001. Subject to the availability of new CIS data, the 2003 EIS is expected to offer an updated composite innovation index and a comparison between the index and average trends for each country20 .    One of the expansions of the 2002 EIS is the development of a Regional Innovation Scoreboard (RIS). The 2002 RIS is limited to those indicators from the EIS for which regional data are available and to a static comparison only.    4.3. Some methodological aspects    As it will be later pointed out, first we have considered as very interesting to compare the evolution and changes of the 32 regions of the EU-15, not only in terms of GDP per capita but approaching the real situation in terms of economic and social cohesion. This last is one of the main objectives of the EU, according to the Treaty and all recent documents, and we maintain the hypothesis that clear differences appear when regional performances in terms of GDPpc are compared to their results and progress in terms of social and economic cohesion (Cuadrado, Garrido and Marcos, 2003).    To reach the objective described up it is necessary to handle simultaneously a large set of variables already mentioned (63) and to interpret the results jointly. In order to facilitate both tasks the grouping of variables is a common practice to extract the essential information. In other words, the usual practice is to calculate the minimum number of factors (groups of observable variables) to explain the greatest quantity of variability; in this case the differences existing among regions. Therefore, we start with the 63 available variables and by applying the Analysis of Principal Components we manage to synthesize 11 factors which cover more than 80% of the differences among the EU-15 regions. These factors, ordered according to importance21, have been iºnterpreted as follows: Ageing of the population, Labor market, Regional Dynamics, Competitiveness , Basic Factors of Development, Residential and economic attraction, Public R+D, Potential of development, Education, Degree of Urbanization,, Tourism Pressure on resident population.    The decision to adopt the multidimensional approach (11 factors that summarize the regional socio-economic characteristics), requires the application of multi-variant techniques. Among these we have chosen to transfer to regional analysis two methods pertaining to the Discrete Multi-Criteria Decision to the regional analysis. On the one hand, the justification for this is that they permit an important flexibility in the characteristics of the incoming information22 and, on the other, they reach the results of classification and ranking (in our case of the regions) considering simultaneously all the factors.    With these techniques, we attempt to study in detail the results produced by and exclusive use of GDP p.c. and that of a ranking and classification or economic and social factors simultaneously. Multi-criteria Decision seeks to provide methods that permit a satisfactory solution of problems of decision in which different and often possibly contradictory perspectives have to be taken into account. The satisfactory solution does not necessarily have to be the best from all points of view. This approach is seen as relevant to the problem at issue in as much as the multidimensionality of ´economic and social cohesion´ is converted into significant advances in some areas (e.g. production level) but with possible reverses in other fields which are equally interesting (employment, infrastructure, health, education among others).    It must also be pointed out that the Electre methods are called non-compensatory methods, that is to say, bad evaluations of a socio-economic factor cannot be compensated for by good evaluations of another socio-economic factor. For this reason, a well positioned region, both in ranking and in classification, is a region that is better in the majority of the socio-economic factors than the rest. Among the different Electre methods11 designed to respond to concrete problems, we have selected Electre TRI, given that it permits us to classify regions in levels of socio-economic development. Electre TRI can provide us with a classification of regions by groups of reference that we can consider as hypothetical regions13 (levels of socio-economic development), in such a way that the profiles are each other totally comparable14. In this case, the regions of reference (profiles) have been defined from quartiles of each one of the socio-economic factors already commented on. We must clarify that this method (Electre TRI) offers two possible procedures of assignation namely optimistic and pessimistic, consisting of comparing each region with the profiles of reference. In this paper we will only take into account the results applying the second procedure as this situates the region in the worst situation. 5. Results    5.1. Regions performances in terms of GDP p.c.    An interesting comparative approach is to analyze the regional behaviour in terms of GDP p.c. and the average growth rates for central, peripheral and intermediate regions selected. To allow this objective we will use data for the 1987-2001 period (the last year of available homogeneous figures). Figures have been taken in PPS values.    Table 4 shows the GDPpc of all the regions in the sample compared to the 1987 and 2001 EU-15 average. Figure 1 shows graphically the position of each regions according to thoses figures.    From this rather simple approach it is possible to point out some remarks:    - Central regions. They all actually have a GDPpc well above the EU mean although the distances between them substantially. For the period considered, six (out of ten) regions have improved their relative position and only four have worsened, particularly Nordrhein-Westfalen and Alsace.    - Intermediate regions. Their level moves around the communitarian average. Among them, Sachsen scored very low in 1987 and remains below the 75% of the communitarian mean. Seven regions are clearly better-off or remain the same (particularly Madrid and Sachsen) and three of them fall in relative terms: Sydsverige, Schleswig-Holstein and Languedoc-Rousillon.    - Peripheral regions. Within this group of the sample, eight regions show a relative improvement against the communitarian average, especially Ireland that wins 54 points and moves to 117.6% above the communitarian mean. Burgenland, Andalucia, Alentejo and even Extremadura also show considerable improvements. On the other hand, four regions in this group are worse-off, especially Mellansverige, Corse and Itä-Suomi.    This first general approach shows that the behaviour of regions within each group is not homogeneous at all. We can include our figures in a process of slow regional convergence of the European regions in GDPpc. However, we must state that there are many changes in the individual positions within them. This topic has already been analyzed in other works (Cuadrado, Mancha, Garrido, 2002) as a fact of major significance. The approximations in terms of convergence show a broad overview of the issue but it hides individual region movements and behaviours that are of major relevance. It´s clear that central regions aren´t better or worse off as a whole, and so are the peripheral although the latter contains a larger number of regions that improve relatively to the communitarian average. Figure 1 shows graphically the movements of the regions in the sample for the period 1987-2001.    Relative growth rates of the regions considered in the sample support the above said. Eight regions show average growth rates above 6 per 100, of which 3 are ´central´ (Luxemburg, South East and Zuid-Nederland), three are intermediate (Madrid, La Rioja, Sachsen) and two are peripheral (Ireland and Burgenland). The rest all lie between 3.7 (Sydsverige) and 5.9 per 100, not allowing identifying trends or homogenous behaviours for the different groups, although several peripheral regions have comparatively low rates.    5.2. Approach in terms of economic and social cohesion    Article 2 of the present Treaty of the EU includes economic and social cohesion and solidarity of the Member States among the main objectives of the Community, issues which are entered into in greater detail in chapter XVII (art.158-162). To improve ´cohesion´ constitutes, without doubt, an intention of a political nature whose operative definition is not easy. The Reports on Economic and Social Cohesion published by the European Commission so far23 make it possible to better understanding the different dimensions of this concept.    Nevertheless, when the evolution of the European regions is analyzed "cohesion" is generally identified with the idea of convergence in income or GDP p.c. In fact, the majority of analyses of regional disparities in Europe - as in other cases- refer to convergence taking as an indicator the evolution of this variable and, sometimes, that of productivity and employment, but not other variables which are without doubt indicative of the degree of development of a region or a country. For instance the level of education reached, the demographic structure, the rate of labor activity and the distribution of employed population by sectors, unemployment rates, life conditions, etc. which are regularly included in the Cohesion Reports.    However, it seems evident that GDP p.c. cannot capture all aspects of the intention covered by the objective of reaching greater economic and social cohesion among European regions. GDP p.c. is undoubtedly a quite significant variable, but it may well happen that the output of a region evolves positively and, at the same time, other variables and indicators do not perform in the same way or with the same intensity. For example, higher unemployment can be compatible with an increase of output and productivity, whereas education, provision of infra-structures and social equipment do not necessarily follow a parallel line of progress. When GDP p.c. is taken as a reference we are using the GDP and population data and this latter can show aspects that are not picked up by the total figures, such as ageing, migrations, low rate of activity, etc. So, the problem we postulate is that we cannot directly equate the changes in GDP p.c. with the improvements in economic and social regional cohesion. Therefore we consider that, to this effect, we must use a broad set of indicators to permit us to show the real evolution of the regions and whether in fact a process of convergence in its widest sense is occurring or not.    As it has been pointed out (section 4), to take into account the 63 variables by regions, we have adopted the multidimensional approach (11 factors summarizing regional socio-economic characteristics), which requires the application of multi-variant techniques. Among these we have chosen Electre TRI, a method of Discrete Multi-Criteria Decision here applied to regional analysis24.    Table 5 shows the results obtained for applying the mentioned method, and allows us to take a much broader look than that of the GDPpc as it takes into account demographical issues, employment levels, regional dynamics, R+D expenditure, education etc. Considering the results obtained, we can underline the following conclusions:    - Central regions. The figures show that not all of the regions in this group keep a ´medium-high´ position in socio-economic terms although all them had a GDP p.c. above 105%. Additionally, three of the central regions have actually downgraded in their socioeconomic score from a comparative point of view (Salzburg, Lusemburg, Île de France).    - Intermediate regions. As we should have expected, the changes in this group have different directions. While some regions have improved their GDP p.c., this is not reflected in their socio-economic positions (Madrid, La Rioja). Other regions in this group are worse-off (Languedoc-Rousillon, together with a relative decrease in its GDPpc; Noord-Nederland although it experiences a slight improvement in its relative GDP p.c.). The rest of the regions do not undergo relevant changes.    - Peripheral regions. In 1987 practically all the regions in this group were included in the ´Low´ category in socio-economic terms with the exceptions of Scotland (with a GDP p.c. well above the rest of the peripheral regions), Campania and Itä-Suomi. In 2000 two regions (Ireland and Norra Mellansverige) jump from the ´Low´ group to the ´Medium-High´; Kentriki Ellada from ´Low´ to ´Medium-Low´. But the rest of the regions in this group do not change (even though some of them substantially improve their GDP p.c. levels) and two (Campania and Itä-Suomi) fall from ´Medium-Low´ to ´Low´), even when the first one experienced a considerable GDP p.c. increase.    Thus, we shall remark what we had already expected: the variations in GDPpc levels (taking into account all the regions in the sample have positive average growth rates for the period 1987-2000) do not always imply improvement in the socio-economic position. There are several cases in which either they stay in a Low position or they even fall from a better position to the bellow one. Thus, when we refer to regional convergence in terms of GDPpc we should consider other variables that pick up such relevant aspects as demographical characteristics, unemployment, education levels, R+D efforts, etc.    In this sense we shall recall some of the characteristics that are common to the three different groups as they give support to some ideas on what characterizes ´peripheral regions´ in addition to the simple approach of accessibility However, we should notice these are only some general features and that a case by case study should be done for a more precise analysis.    The main characteristics of the central regions are: high competitivity and potential development, comparatively high R+D investment in the private sector, high population density but with a high aging level, high levels of labour participation rate, lofty patents and innovation capacity, and better education levels than those of the intermediate and (generally) the peripheral regions.    Intermediate regions are somehow more complex. In fact, the three socioeconomic categories show different components and they are not always coincidental. Anyway, they present average population density, still high levels of labour participation rates, good endowment of basic development factors and a remarkable R+D expenditure, both in the public and private sectors. On the other hand, education levels vary substantially.    Finally, peripheral regions show low population density (although younger) comparatively lower rates of labour participation, low regional dynamics and limited R+D investment (generally essentially public), small number of students above secondary level and a really heterogeneous picture in terms of competitivity. On the other hand, the main characteristic is that almost all regions in this group continue to be located at the ´Low´ level in socioeconomic terms, despite a good number of them have increased substantially their GDP p.c. compared to the EU-15 average. The only exceptions are Ireland and Norra Mellansverige, which have jumped to ´Medium-High´ Level, and Kentriki-Ellada (jumping from ´Low´ to ´Medium-Low´ level). Scotland was an still is located in ´Medium-Low´ position.    5.3. Some specific indicators on innovation in central, intermediate and peripheral regions    The previous broad approach allowed us to point out some of the characteristic components of peripheral regions in contrast with central and most intermediate regions that are far more important than distance or accessibility. Issues like R+D, education, innovation capacity or participation in the labour market, among others, appeared really distant in central regions compared to the rest. Accessibility and transport difficulties might have had its influence but it doesn´t seem to be enough to explain the differences.    For that reason we will now verge on some of the concepts and indicators that significantly pose disparities. The 2002 European Scoreboard (EIS) focuses on high-tech innovation. It contains 17 main indicators, selected to summarise the main drivers and outputs of innovation. Here we will pay attention to some of them, particularly the ones more recently updated (2003).    Tertiary Education and life-long learning    Figure 2 shows the regional spread for population with tertiary education. The leading positions taken by Finland, Sweden and the UK (are reflected by the fact that the three leading EU regions can also be found in these countries: Väli-Suomi (40,01%), South West (39,96%) and Östra Mellansverige (38,70%). Due to discrepancies in educational systems, definitions of tertiary degrees might differ among countries. This is reflected by the fact that for Austria (14,52%), Greece (17,08%), Italy (10,03%) and Portugal (9,12%) all regions score below the EU13 mean (22%), whereas for Belgium (27,82%), Finland (32,47%), Sweden (29,71%) and the UK (28,56%) all regions score above the EU mean. Both Île-de-France (33,32%) and Utrecht (32,60%) show the highest rates among our targeted central regions, whereas tertiary education is below the EU13 average in Lombardia (10,15%) and Salzburg (15,90%). At the intermediate level, Madrid (31,56%) is the leading region and Toscana (9,12%) is the bottom one. In the periphery of Europe, tertiary education is above the EU13 mean in Highlands & Islands (30,64%), Itä-Suomi (28,87%) and Norra Mellansverige (23,96%), and less than half this figure in Alentejo (9,92%), Burgenland (10,16%) and Calabria (10,21%). For participation in life-long learning, Sweden ´stars´ among the EU top-10 with five regions and the UK with four (see Figure 3). Differences in adult education systems favour these two countries: Northern Ireland, for example, is the only UK-region that is not in the EU top-20. The relatively homogeneous situation within countries, for this indicator, should be noted, reflecting the importance of the national context for life-long learning practices. The central regions of Berkshire, Bucks & Oxfordshire (21,61%) and Utrecht (18,95%) both exhibit a rate in life-long learning around twice the EU12 mean (9,72%), while it is only about one third this figure in Île-de-France (3,40%). The intermediate area of Malmö (23,95%) enjoys the highest number of people involved in life-long education with Merseyside following behind (22,98%), whereas Languedoc-Roussillon (2,21%) has the least figure. The northern peripheral regions of Norra Mellansverige (22,46%), Highlands & Islands (20,12%), and Itä-Suomi (17,6%) all have rates well above the EU12 average. On the contrary, the southern periphery shows the lowest figures like in Ipeiros (0,61%) and Corse (1,33%).    Employment in Medium/High-Tech Manufacturing    It is highest in four German, two French and two Italian regions (see Figure 4). In Baden-Württemberg (Stuttgart-Karlsruhe-Freiburg-Tübingen), this employment exceeds 18%, more than 3 percentage points above the second EU region. The automobile industry (Mercedes) is a major contributor to this success. This industry by itself explains to a large extent the ranking of the top EU regions: for example, Piemonte (rank 3: FIAT), Bayern (rank 4: BMW) and Navarra (rank 6: Volkswagen). Discrepancies within countries are generally large, reflecting imbalances between heavily industrialised regions and rural or services-oriented areas. Among our selected central regions, Ruhrgebiet (13,54%), Alsace (12,79%) and Lombardia (11,91%) display an employment rate in medium/high-tech manufacturing above the EU13 average (7,44%), while the lowest figure is registered in Utrecht (2,14%). In the group of intermediate regions, Merseyside (8,19%) is the only place showing a rate above the EU13 mean and Languedoc-Roussillon (2,43%) comes at the bottom of the ranking. In the European periphery, all regions score below the EU13 employment rate average varying from a maximum of 6,80% in Norra Mellansverige to a minimum of 2,41% in the Andalucía.    Employment in High-Tech Services    Highest ratios are in Stockholm (8,41%) and Uusimaa (7,11%) (see Figure 5). In both regions, we see a strong sector (Ericsson and Nokia). Most of the strongest regions are capital-city regions. Statistically, there is no relation between employment in high-tech services and that in medium/high-tech manufacturing, reflecting the relative specification of regions within countries. For most countries, this indicator shows large regional disparities. At the intermediate level, Comunidad de Madrid (5,68%) exhibits the highest number of people occupied in the high-tech service sector followed by Malmö Area (4,58%), Merseyside (4,01%), Languedoc-Roussillon (3,74%), and Groningen (3,72%), whereas both Toscana (2,66%) and Schleswig-Holstein (2,97%) score below the EU13 average (3,58%). The central districts of Utrecht (6,49%), Berkshire, Bucks & Oxfordshire (6,18%), and Île-de-France (6,46%) all register an employment rate in high-tech services above 6%, while it is less than 2% in Alsace (1,87%). The only peripheral region displaying a figure above the EU13 level is Highlands & Islands. In the same category, the Spanish region of Andalucía (1,48%) has the lowest employment rate in high-tech services.    Public R&D Expenditure    This is a good indicator of the presence of voluntary policies directed to specific regions. However, it should be noted that funds can be of regional or national origin, thus this should not be taken as an indicator of the intensity of regional R&D policies. Public R&D expenditures are highest in Flevoland (2,08%), Midi-Pyrénées (2,04%) and Berlin (1,84%) (see Figure 6). Besides Flevoland, four other Dutch regions appear in the EU top-10 ranking, reflecting the overall strong Dutch performance (0,89%). Among these, we find our central district of Utrecht (1,33%) and the intermediate region of Groningen (1,39%). The only central region scoring below the EU13 average (0,64%) is Lombardia (0,33%), while Île-de-France has the highest figure of public R&D expenditure. At the intermediate level, remarkable is the leading position taken by Languedoc-Roussillon (1,54%), whereas La Rioja (0,20%) and Merseyside (0,34%) are at the bottom of the ranking. A region´s public R&D intensity will depend heavily on the presence of both universities and public and non-profit research institutes. Universities (at least the older ones), for example, are mostly located in more densely populated and urbanised regions. In more rural regions, public R&D expenditures are thus expected to be small. Here again, national disparities are large. This is the case for our peripheral regions of Corse (0,23%), Alentejo (0,35%), and Andalucía (0,46%), with the only two exceptions being Highlands & Islands (0,89%) and Itä-Suomi (0,75%).    Business R&D Expenditure    Figures are highest in Västsverige (4,27%), Stockholm (3,88%) and Eastern UK (3,02%) (see Figure 7). Business R&D expenditures are highly concentrated in several countries. For instance, four Swedish regions appear in the EU top-10 ranking, with our peripheral and intermediate areas of Norra Mellansverige and Malmö both scoring 0,95% and 2,81%, respectively. The overall lagging position for Greece, Italy, Portugal and Spain is also witnessed at the regional level. Only for Piemonte (1,42%), the R&D intensity is above the EU13 mean (1,14%). Interestingly, all countries (with the exception of Belgium) include at least one region with performance far below the EU13 average for this indicator, such as the peripheral regions of Calabria (0%), Corse (0,02%), Alentejo (0,05%), Ipeiros (0,07%), Andalucía (0,22%), and Campania (0,29%). The lagging situation of Flevoland (0,35%) contrasts with the leading position this region has for public R&D. In the core of Europe, Île-de-France (2,33%), Oberbayern (2,08%) and Berkshire, Bucks & Oxfordshire (2%) all register business R&D expenditure rates above 2%. The intermediate regions of Merseyside (1,41%) and Groningen (1,25%) both score above the EU13 level, whereas Toscana (0,31%) and Schleswig-Holstein (0,44%) display rates below this figure.    High-Tech Patent Applications    Uusimaa (Suuralue) (187,8), Noord-Brabant (163,4) and Stockholm (150,3) are the three leading regions for High-Tech Patent Applications (see Figure 8). Finland and Sweden, the EU leaders in high-tech patent activity, both have three regions in the EU top-10 ranking. Finland, Sweden and the Netherlands show large regional disparities. For Finland and the Netherlands, this can be explained by the location of the multinationals Nokia in the Helsinki area and Philips in the Eindhoven area. The highest number of high-tech patents are registered in central districts of Oberbayern (88,8), Île-de-France (56), Berkshire, Bucks & Oxfordshire (39,1), and Utrecht (24,6), whereas the least figures appear in Salzburg (9,7), Ruhrgebiet (19), Alsace (19,1) and Lombardia (16,5). The Malmö area (48) is the only intermediate region displaying a number of high-tech applications above the EU13 mean (22,28), whereas patents score almost zero in La Rioja (0), Toscana (2,8) and Languedoc-Roussillon (3,9). In the periphery, patents are close to zero in all selected regions apart from Norra Mellansverige (14,2), Highlands & Islands (11,3), Itä-Suomi (4,3), and Border, Midland & Western (3,1).    5.4. Final comments    As it was previously pointed out, ´accessibility´ is not the most decisive factor of ´peripherality´. It is, o course, a crucial factor which has influenced in many other regional characteristics, both if it is high or low. But, as we have seen, there are other critic factors that characterize most of the peripheral regions (although not all of them and not in the same ´order´) which are indicative of another ´peripherality´: the ones that are specifically related to innovation capacity, absorption and development of new technologies, higher education levels and growth potential and competitivity of the regions in an international scale. At the end, these factors reflect regional capacity of confronting the actual changing economic and productive environment and its capacity of managing such changes.    If there is something obvious, this is that the highest innovation capacity tends to concentrate in certain areas, generally central or intermediate regions in each country, and many times central in a EU scale. Leading innovation regions within each country and ´Local´ EU Innovation Leaders are shown in tables 6 and 7 respectively, and their relationship is certainly significant.    On the other hand, if we regard the indicators used and commented in the previous section we can build a table that shows the leading regions for each indicator (tertiary education; life-long learning; medium/high tech employment; High-tech employment in services; public R+D; business R+D; High-tech patents) (Table 8). As we can observe, these are nearly all ´central´ regions and only some described as intermediate appear sporadically.    Obviously it would be naïve to think the values of the indicators should be more homogeneous and equal among the regions. Moreover, what such indicators represent isn´t the key elements or factors for regional development and increasing welfare. But, any case, it is clear that ´peripherality´ shows signs that aren´t just a matter of distance.    Certainly, the bad score of peripheral regions in the indicators used and the results from our previous analysis of their position in terms of cohesion are examples of much more relevant factors than that of ´distance to the centre´. Apart from improving communications, it is patent that this is the field in which it is necessary to work to smooth the differences between centre and periphery. Some examples of intermediate and peripheral European regions show that, as we stated in section 2, being a peripheral area does not mean an unavoidable doom. 6. Challenges facing the periphery of Europe    European regions, and peripheral regions in particular, have to cope with fresh challenges, globalisation and rapid technological change specially, in order to provide the economic opportunities and quality jobs needed in less favoured areas.    6.1. Actual challenges and risks    Globalisation and Increased Competition from Third Countries    Globalisation is almost a trite saying, but also a reality. It is about new opportunities in ever-larger markets but also about greater competitive pressures from third parties in more transparent market places. The latter holds particularly true for labour intensive firms in traditional sectors, which have been competing on the basis of cost. As Porter (1990) has observed "... Competitive advantage based on factor costs…is rapidly undone... A low-wage country today is quickly replaced by another tomorrow...´ Instead of building comparative advantage on the basis of low cost/low salary production, European regions should be assisted to develop competitive advantage based on innovation.    SMEs in less favoured regions may need assistance in tapping into the necessary resources (related to knowledge, in the form of technology or qualified human capital in particular), to face up to the new forms of competition being developed in the global economy. So, regional innovation policy may help stimulate firms, SMEs. in particular, in less favoured regions to adopt improved production methods25, make new/different products and services26, and exploit new economic opportunities and markets (university spin-offs, new technology-based firms, etc).Thus using their regional innovation potential to the full in order to compete in the global economy.    Accelerated Technological Change    Fast and faster technological changes are a characteristic of our age. Technological progress raises living standards by creating new goods and allowing existing goods and services to be produced at lower cost. ICT reduce also distances, which have potentially positive and negative consequences for peripheral regions. Positive, as they offer opportunities of reducing their ´peripherality´, giving them a way to better diffuse their advantages, their products, their own image, their environment and landscape, etc. Negative, as ICT seems to support much more central places than far away ones. So, local producers and local markets in the last ones may be easier invaded by foreign producers. Advances in information and communications technology have made it feasible for firms to locate call centres abroad. So, a possibility of shifting jobs to other countries appears which combined with increased anxiety about jobs has generated pressures for a political response. Two approaches habe been suggested. One tries to counteract the market forces underlying globalization, while an alternative approach relies on policies that assist market forces, while providing assistance to those adversely affected.    At the end, a whole new way of doing business is in the making. This means in practice that no business, regardless of its location or size, can ignore this revolution. On the contrary, they have new opportunities. Sooner of later they will have to embrace the net if they are to survive and be competitive in the new economy, and it is a matter of fact that far too many SMEs in less favoured regions are still lagging behind in the use of these new technologies.    Risks and Opportunities for Regions. Human capital and intangibles    The sources of wealth creation and economic growth in the new networked economy are information technology, communications (ICT) and knowledge in the form of intellectual capital, much more than natural resources and the efficiency of physical labour as competitive factors. More than ever, human capital (and training) is the key to innovation and competitiveness. Moreover, network economies as they grow, increase the benefits of all those who are ´connected´ giving a ´win/win situation´. This is why, less favoured regions can not afford not to be connected and regional policy has to cope with the so-called ´digital divide´ between those regions that are connected and have the ICT skills, infrastructures and the knowledge access, and those that have not.    The key to success in the old economy was ´costs´. In the new economy ´non-cost´ factor competition is essential. ´Intangibles´ such as speed of response to market demands, reductions in the life cycle of products, quality, design, product differentiation/customisation of products to niche markets, new management methods and business organisation, the capacity to co-operate in inter-firm business networks are the key to competitiveness in the new economy. Regional advantage will go to those places, which can attract and quickly mobilise the best people (´knowledge workers´), resources and capabilities required to turn innovations into new business ideas and commercial products. This is precisely why regional policy should help the less advanced regions to anticipate and prepare for the new economy through a new type of regional policy. A policy response that needs to have innovation promotion at its heart, if regions are to be successful in coping with the challenges posed by the new economy.    But it is not only the new economy, which is a major concern for regional policy. There are new challenges for regional policy to cope with the ´inter-regional technology gap´ which requires a different policy response.    The Regional RTDI Gap versus the Cohesion Gap    In the Union, the RTDI gap is today nearly twice as great as the cohesion gap. Many of the causes of disparities among regions can be traced to disparities in productivity and competitiveness. Education, research, technological development and innovation are vital components of regional competitiveness. So let us look at some examples of the so-called inter-regional R&TDI gap.    In terms of Gross Expenditure on RTD&I. The 25 least developed regions in Europe spend, as a percentage of GDP, less than a quarter of the European Union average (0,5% compared to an EC average of 2% - 1999). On a regional level, business expenditure on RTD&I as a percentage of GDP in the most developed 25 regions is on average 1,9%, while in the 25 least developed regions this figure falls to around 1 %. This difference in financial input also has consequences in terms of innovation outputs. For example, in 1988 there were over 20 times the number of patent applications in Germany alone, than in the four cohesion countries together (Ireland, Greece, Portugal and Spain).    In RTD&I Human Resources. The ´technology gap´ is a particular cause for concern with regard to the human resources, since human capital is increasingly a source of the dynamic comparative advantage, which governs regional potential for innovation. In an increasingly ´knowledge-based´ economy, the only real capital is human capital. In terms of High Technology employment, in the 25 most advanced regions high technology accounts for an average of 14,6% of total employment, compared to just over 4% on average in the 25 least developed regions. This is compared to a community average of around 10,5%. Denmark, with a labour force of around 2.5 million has almost twice the number of RTD&I personnel than Portugal, with a labour force of around 4.5 million. Germany has almost double the number of RTD&I personnel per thousand labour force than Spain, three times more than Greece and four times more than Portugal.    Inter-regional differences within Member States are even greater than at international level. In Greece for example, over half the country´s RTD&I expenditure is taking place around Athens and over two-thirds of business RTD&I is located in the Attiki region. In Spain over three-quarters of business RTD&I is located in three of the seventeen regions (Madrid alone accounting for over 30%).    Regional Disparities in the Efficiency of Regional Innovation Systems    Moreover, the inter-regional RTD&I gap is not only of a quantitative nature but also of a qualitative one. There are a number of characteristics of regional innovation systems in less advance regions, which make them less efficient as the following:    (i) Firms may not be capable of identifying their innovation needs or maybe unaware of the existence of a technical solution.    (ii) There may be poorly developed financial systems in the area with few funds available for risk or seed capital, which are specifically adapted to the terms and risks of the process of innovation in firms.    (iii) There may be a lack of technological intermediaries capable of identifying and ´federating´ local business demand for innovation (and RTD&I) and channelling it towards sources of innovation (and RTD&I) which may be able to respond to these demands.    (iv) Co-operation between the public and private sectors may be weak, and the area may lack an entrepreneurial culture which is open to inter-firm co-operation, leading to an absence of economies of scale and business critical mass which may make certain local innovation efforts profitable.    (v) Traditional industries and small family firms may dominate which have little inclination towards innovation. There may be a low level of participation in international RTD&I networks and a low incidence of large, multinational firms.    6.2. Innovation as a response    Regional policy should increasingly concentrate its efforts on the promotion of innovation to prepare regions for the new economy and close the ´technology gap´ if it is to be successful in creating the conditions for a sustained (and sustainable) economic development process in less favoured regions.    As it has been pointed out by Cuadrado-Roura, Mancha-Navarro and Garrido-Yserte (2002), peripheral areas (particularly in Southern Europe), with a lower degree of specialisation, low markets flexibility and a lower competitive level, are those that run a greater risk in the present context of economic integration. On the other hand, a good number of central zones (mainly located in Germany, Luxembourg, The Netherlands, France and Denmark) are initially more favoured by the integration process and the extension to the East because of having the opposite economic characteristics of those of the peripheral areas.    However, this does not mean that the peripheral territories cannot also take advantage of the opportunities offered by this scenario, but simply that their initial position is less advantageous and that their efforts of adaptation must necessarily be more costly. In short, the existence of two differentiated models of development appear within the EU: one of a more balanced and symmetrical character (central countries/regions), as opposed to another with greater singularities and divergences in the evolution of their economies, which encompasses the greater part of Southern Europe as well as some areas of Ireland, the United Kingdom and Northern countries    Competing with Central regions and learning from the winner regions    There are two issues to which are given little attention when analysing convergence at the EU-level (Cuadrado, 2001). On one hand, the role played by the ´state factor´ in relation to the behaviour shown by regions integrating in the same country, and, on the other, the fact that - within overall non-convergence - significant ´movements´ take place in regional evolution. Movements which are neither linked to clichés nor to pre-qualifications as being underdeveloped or very developed, central or peripheral regions. The search for an explanation to this fact has led to an examination of new growth theories and to a group of factors and characteristics in which a series of analysed regional cases coincide. It is not so important to pay attention to regional ´convergence´ (and its global trend) than to the causes, which can explain the different behaviour of individual regions, whether these are underdeveloped, recovering or very advanced, independently from their geographical position. Divergences observed between those regions that could be qualified as winners and losers (at least in comparative terms and at a European level) lead to an examination of the possible explanatory elements.    Experience shows that the analysis of common patterns does not allow clear generalisations and that only individual case studies provide sufficiently solid and reliable explanatory elements. However, some of the features characterising a series of European regions that were more dynamic over the 1977-2000. From that analysis it was possible to remove some common attributes that seem to have been very decisive in their success. These regions´ success is not always fulfilled to the highest level but among those attributes that are always related to better performance there were five particularly interesting for peripheral regions:    (i) The availability of human resources (Manpower + Human capital), with a stable supply of qualified labour force and a medium to high educational level thus make up a factor in which all the analysed "winning" regions coincide. If, in addition, relative labour costs - both wage and unearned ones - are moderate, the region offers an additional advantage. However, in an increasingly integrated market as the EU, the importance of this advantage has clearly decreased and will tend to become even less relevant in the future. On the contrary, the presence of prestigious educational and research centres, with the corresponding human capital back up, are much less volatile factors.    (ii) The accessibility of the region also represents a key factor. Accessibility is viewed from a wide perspective and not only from a physical one; that is to say: accessibility of the region and its agents to international markets; accessibility to the persons who take political and economic decisions in the country; quick access to innovations and technological developments, etc. Some instruments, such as information centres, technological institutes, scientific and technological centres act as channels and networks, which facilitate access. At the same time, it is necessary to respond - in terms of receptivity - to the signals, which come from outside the region. All of the analysed regions (Cuadrado, 2001) offered - according to an analysis ordered by the European Commission - a medium or high level of access and receptivity to ideas and innovating investments.    (iii) The availability of advanced production services (nor an easy access to them) also appears as a positive condition for territorial competitiveness (Daniels, 1993; Rubalcaba, 1999; Cuadrado and Rubalcaba, 2000). We are referring to services such as: strategic planning, technological consulting, design, commercialisation and exports, R&D, specialised financial services, etc. Although these services can be acquired from outside the region, it is true that close relationships, frequent contact and adaptation to requirements all represent a very positive factor, both for their use and real efficiency.    (iv) The institutional aspects of each region are not only important but also seem to have been very important in some of the analysed cases. De-centralization constitutes a good way to solve some problems. The existence of a regional authority or government with wide competencies and autonomy in relation to the Central Administration, as well as a system of regular co-operation between the different (national, regional, municipal) authorities, on one hand, and between these and civil organisations, on the other, (chambers of commerce, entrepreneurial and social organisations) represent a specially decisive factor. A competitive territory - whether it is a region or city - requires a good organisational infrastructure for putting into practice support strategies. The most autonomous regional governments - although not all of them - have been able to apply and develop initiatives, which clearly surpass the possibilities which central authorities would have. The achievement of a good exterior image has been, in almost all of the cases, a consequence of co-ordinated actions and therefore leadership from the local area itself.    (v) In all of the case-studies analysed, the more (or less) attractive image of a region to attract foreign investments or to activate local ones is related to the existence there of a pacific social climate (low labour conflicts) and a co-operative atmosphere between the different regional, public and private institutions and organisations. Normalisation and stable dialogue with trade unions and lack of conflicts are necessary requirements for a territory to be more competitive than others are. 7. Some final remarks and future perspectives    In some peripheral areas regional development policy (and so the use of Structural Funds) have been mainly directed towards creating the physical infrastructures (Rodriguez-Pose, 1999 and 2000) which are a necessary pre-condition for sustaining a process of economic development: roads, airports, water-treatment plants, energy, etc. However, intangibles are gradually becoming a priority for regional policy in those less-developed regions that are successfully overcoming their shortage of infrastructures. Therefore, the emphasis must increasingly to be placed on those conditions which will most directly and immediately affect the capacity of businesses - SMEs in particular - to develop new job-creating activities. This necessarily involves more regionalised policies for the promotion of innovation, as intended by some Commission programmes and projects (Regional Innovation Strategies; Regional Innovation and Technology Transfer Initiatives, the Innovating Regions in Europe network…). These policies go well beyond tax incentives, training programmes, and aid for basic research of the provision of R&TD physical infrastructures. They require new policy delivery systems, including financial engineering, and should be based on close cooperation with the private sector through new forms of public-private partnership. They are fundamentally aimed at increasing the capacity of business to innovate as a principal source of regional competitiveness. And, of course, they contrast with public subsidies to individual businesses through horizontal and automatic programmes of public aid. In this sense, they should help reduce the barriers that (new) businesses have to overcome to enter markets, stimulating and supporting (existing and potential) entrepreneurs in all phases of the business creation process.    Action in the Peripheral Areas    The Second Progress Report on Economic and Social Cohesion (2003) confirms both the unprecedented increase in the disparities within the enlarged Union and the long-term nature of the efforts that will be needed to reduce them. There is a widespread agreement by the Commission around the need to continue to concentrate resources on the less developed regions, and especially on those in the new Member Sates. On how to define the less favoured regions, the contributions to the debate have not seriously put into question continued use of the present eligibility criteria based on the NUTS II geographical level and per capita GDP - which have the merit of being simple and transparent - even if some contributions have called for other criteria to be added. Nevertheless, the suggestion has been raised that the GDP criteria need to be complemented by others, such as employment and unemployment rates, productivity, peripherality, population decline, or the level of financial execution27.   There are several important challenges facing the Union as a whole. In particular, the issue of competitiveness, sustainable development, and economic and social restructuring are relevant in all Member States. This wide range of challenges also highlights the need to concentrate assistance, and to focus on qualitative, systematic elements in order to increase Community added value. Within the debate that actions of this nature outside the Objective 1 regions should be abandoned entirely by the Union and responsibilities returned to the Member States (´renationalised´) do not appear to have gained ground, and the need to have the means to achieve major European priorities has been recognised.    In this respect, the Union, particularly at the European Councils in Lisbon, has set itself a strategic goal for the decade: to become the most competitive and dynamic knowledge-based economy in the world, capable of sustainable economic growth with more and better jobs and greater social cohesion. This strategy is designed to enable the Union both to regain the conditions for full employment, growth and social cohesion, and to strengthen regional cohesion. A sustainable development strategy for the European Union was decided by the European Council in Göteborg. A large number of contributions stress that the Member States and regions do not possess the same strengths for achieving these goals. However, these objectives could be reflected through emphasis on the factors contributing to competitiveness such as accessibility, the diversification of productive structure, the knowledge society, innovation, R&D, the environment, employment, social integration, and education and life-long learning. In addition, a policy intended to meet the various challenges facing the Union, that recognises and involves the regional level, is consistent with the spirit of the Commission White Paper on Governance, and especially with the obligations arising under the Treaty on cohesion, which are to ´promote its overall harmonious development´ and ´reducing disparities between the levels of development of the various regions ad the backwardness of the least favoured regions or islands, including rural areas´. These major European priorities have, to a certain extent, been tackled already during the present programming period under Objetive 1 (lagging regions), Objective 2 (regions undergoing restructuring), Objective 3 (human resources), Community Initiatives and Innovative Actions.    Criticisms to European Regional Policy. Towards new strategies and Interregional cooperation    Anyway, current policies and instruments have been object of criticism on the basis of arguments that they lack sufficient added value, sometimes require an excessive administrative input in relation to the outputs achieved and fail to devolve sufficient responsibility to the Member States consistent with the principle of subsidiarity. The majority of contributions stressed the importance of factors for competitiveness: the sectoral distribution of employment, innovation capacity and the links between firms, research and universities, education levels, accessibility, the impact of globalisation on the economy, and ´peripherality´. In sum, policy priorities and instruments for regions lagging behind as well as for other regions in the periphery would need to be reformulated, in order both to address the present shortcomings and to construct a new policy capable of making a greater contribution to economic and social cohesion.    There is also a broad acknowledgement concerning the need to continue actions to promote cooperation across frontiers, between regions (central and peripheral) and within regions. This is in recognition that the successful implementation of such actions, which are particularly important for European territorial integration, requires organisation at supranational level. The current period has demonstrated the difficulties inherent in organising coherent programmes involving authorities from different national administrative and legal/cultural traditions.     Tables                   Figures                     Notas a pié de página 1 See Piore and Sabel, 1984.  2 The report was titled: "Thematic Study on the Situation of Peripheral Areas in Europe" and it has been promoted by the Malaga County Council, co-financed by the European Commission. Authors: B.Cozzari, J.R.Cuadrado-Roura and R.Dressen (Nov-Dec. 2003, polic.).  3 Second Report on Economic and Social Cohesion; Brussels, 2001. This section is mainly taken from this report, pp. 30-31.  4 See Map A.5 and Map A.6 on ´Trans-European Transport Network Outline Plan: Rail and Road Bottlenecks´ in the Annex of the Second Report on Economic and Social Cohesion, Volume II, January 2001.  5 See Krugman, 1991 and 1992.  6 See Linneker, 1997.  7 See section 2.3.3 on Inadequate Accessibility in the EU in the ESPD, European Commission, 1999.  8 See Schürmann and Talaat, 2000.  9 See Kline and Rosenberg, 1986 and Dosi, 1988.  10 See Lundvall, 1992; Nelson, 1993; and Edquist, 1997.  11 See Aydalot and Keeble, 1988; Camagni, 1991; Grabher, 1993; Tödtling, 1994a; and Storper, 1995.  12 See Cooke, 1997.  13 See Grabher, 1993; Saxenian, 1994; and Enright, 1995.  14 See Camagni, 1991; Maillat, 1991.  15 See Castells and Hall, 1994.  16 See Malecki and Tödtling, 1995; Sternberg, 1995; and Hassink, 1996.  17 See Porter, 1990; Scott, 1990; and Storper & Walker, 1989.  18 See Schürmann and Talaat, 2000.  19 A first Provisional EIS was published in September 2000: COM (2000) 567. The first full version of the EIS was published in October 2001: SEC (2001) 1414.  20 The EIS is complemented by six technical papers: (i) Member States and Associate Countries; (ii) Candidate Countries; (iii) EU Regions; (iv) Indicators and Definitions; (v) Thematic Scoreboard "Lifelong Learning for Innovation"; and (vi) Methodological Report.  21Percentage of variability among regions explained for each one.  22 It is possible to combine both qualitative and quantitative information.  23 The First one in 1996; the Second in 2001; the Third in 2004.  24 See: Cuadrado-Roura and Marcos, 2003; paper presented at the North-American Meetings RSAI held in Philadelphia, nov. 2003.  25 e.g.: quality and evironmentally friendly processes, introduction of technological developments and innovation management methods, etc.  26 E.g.: design, customisation, selling by e-mail, etc.  27 See notably the Coreper report on the first progress report (Council document 8026/02). References *(not completed) Aydalot P. and D. Keeble (1998). High Technology Industry and Innovative Environments: The European Experience, Routledge, London. ______ (1986). Milieux Innovateurs en Europe, GREMI, Paris. Camagni, R. (1991). Local Millieux, Uncertainty and Innovation Networks: Towards a New Dynamic Theory of Economic Space, in Camagni R. (1991), Innovative Networks: Spatial Perspectives. GREMI. Belhaven Press, London and New York. Castells, M.(1998). The End of the Millennium, The Information Age: Economy, Society and Culture, Vol. III, Cambridge, Massachussets; Blackwell: Oxford, UK. Castells, M and P. Himanen (2002). The Information Society and the Welfare State: The Finnish Model, Oxford University Press, Oxford. Coccossis, H. and P. Nijkamp (1995). Overcoming Isolation: Information and Transportation Networks in Development Strategies for Peripheral Areas, Springer, Berlin. Cooke (1997): Vusiness Process in Regional Innovation Systems in the EU. Cuadrado-Roura, J. R. (2001). Regional Convergence in the EU: from Hypothesis to the Actual Trends, The Annals of Regional Science, 35: 333-356. Cuadrado-Roura, J. R. et al. (1999). Regional Convergence in Productivity and Productive Structure: the Spanish Case, International Regional Science Review, 22,1: 35-53. Cuadrado, J.R. and Rubalcaba, L. (2000): Los servicios en la industria española; Ed. IEE, Madrid. Cuadrado Roura, J. R. and M. Parellada (2002). Regional Convergence in the European Union, Springer-Verlag, Heidelberg and New York. Cuadrado-Roura, J.R., Mancha, T. and R. Garrido-Yserte (2002). Regional Dynamics in the European Union: Winners and Losers, in Cuadrado-Roura, J. R. and M. Parellada, Springer-Verlag, Heidelberg and New York, 23-52. Cuadrado-Roura, JR and M.A. Marcos (2003): "Economic and Social Cohesion versus regional convergence in GDP p.c. in the EU"; paper presented at the North Amercian Meetings of RSAI, Philadelphia, November. Disi, G. (1988), Technical Change and Economic Theory, Pinter, London. Edquist, C. (1997), "Systems of Innoavtion Approaches - Their Emergence and Characteristics", in E.Edquist, C. (ed.) Systems of Innovation - Technologies, Institutions and Organisations. Cassel Academic, Londo. Eskelinen, H. and F. Snickars (1995). Competitive European Peripheries: Advances in Spatial Science, Springer, Berlin. European Commission (2003). Second Progress Report on Economic and Social Cohesion: Unity, Solidarity, Diversity for Europe, its People and its Territory, 30 January, Brussels. ______ (2001). Second Report on Economic and Social Cohesion: Unity, Solidarity, Diversity for Europe, its People and its Territory, Volume 2, Office for Official Publications of the European Communities, Luxembourg. ______ (1999). Towards Balanced and Sustainable Development of the Territory of the European Union. European Spatial Development Perspective (ESDP), Office for Official Publications of the European Communities, Luxembourg. European Commission, Directorate-General for Enterprise (2002). Member States and Associate Countries, European Trend Chart on Innovation, European Innovation Scoreboard: Technical Paper No.1, 4 December, Brussels. European Commission, Directorate-General for Enterprise (2002). EU Regions, European Trend Chart on Innovation, European Innovation Scoreboard: Technical Paper No. 3, 28 November, Brussels. ______ (2002). Structural Policies and European Territory: Cooperation Without Frontiers, Office for Official Publications of the European Communities, Luxembourg. _______(2004), Third Report on Economic and Social Cohesion, A New Partnership for Cohesion. Brussels. Fisher, M., Suarez-Villa, L. and M. Steiner (1999). Innovations, Networks and Localities, Springer, Berlin. Hassink, WHJ. And J.C-. van Ours (1996), "Job Turnover and Labor Turnover; a Taxonomy of Employment Dynamics"; Annales d´Economie et de Statistique. Grabher, B. (1993), Rediscovering the social in the economics of interfirm relations, in G.Grabher (eds): The Embedded Firm, Routledge, London, 1-32. Keeble, D., Owens, P.L. and Thompson, C. (1982). Regional Accessibility and Economic Potential in the European Community, Regional Studies 16: 419-432. Keeble D., Offord, J. and Walker, S. (1988). Peripheral Regions in a Community of Twelve Member States, Commission of the European Communities. Kline, S.J., Rosemberg, N. (1986) "An Overview of Innovation", in Landau and Rosemberg (eds.): The Positive Sum Strategy Krugman, P. (1992): "A Dynamic Spatial Model", NBER W.P. 4219, NBER Inc. Linneker, B. (1997), Transport Infraestructure and Regional EconomicDevelopment in Euroepa: A review of Theoretical and Methodological Approaches, TRP 133. Dpt. Of Town and Regional Planning, University of Sheffield, UK. Lundvall, B.-A, (ed.) 1992), Natinal Systems o Innovation. Towards a Theory of Innvation and Interactive Learning, Pinter Pbls., London. Maillat, D. (1991): "The Innovation Process and the Role of the Miliey", in E.N.Bergman, G.Maier and F.Tötdling (eds.): Regions Reconsidered: Economic Networks, Innovation and Local Development in Industrialized Countries, Mansell, Londos, 103-118. Piore, M. and C.Sabel (1984), The Second Industrial Divide, Basic Books, N.York Porter, M.E. (1990) The Competitive Advantage of Natins; The Free Press, N.York. Rodriguez-Pose, A. (1999) Convergence or divergene) Types of Regional Responses to Socioeconomic Change. Tijdschrift voor Economische en Sociale Geografie 90: 363-378. ----------- (2000): Economic convergence and regional development strategies in Spain: the case of Galicia and Navarre. EIB Papers, 5, 1: 89-115. Rokkan, S. and D. W. Urwin (1987). Centre-Periphery Structures in Europe, Campus, Frakfurt-Main. Roy,B. (1985). Méthodologie Multicritère d´Aide à la Décision. Economica. Paris Roy, B. and Bertier, P. (1971). "La méthode ELECTRE II: une méthode de classement en présence de critères multíples". SEMA (Metra International), Direction scientifique, Nota de travail Nº 142, Paris, p. 25 Roy, B. (1978): "ELECTRE III: un algorithme de classement fondé sur une représentation flone des préférences en présence de critères multiples", Cahiers du CERO, vol. 20, nº 1, p. 3-24. Rubalcaba, L (1999), Business Services in the European Economy, European Commission, Brussels. Schürmann, C. and A. Talaat (2000). Towards a European Peripherality Index. Report for DG REGIO of the European Commission, Berichte aus dem Institut für Raumplanung 39, IRPUD, Dortmund. Spiekermann, K. and J. Neubauer (2002). European Accessibility and Peripherality: Concepts, Models and Indicators, Working Paper 2002:9. Nordregio, Stockholm. Storper, M. 1995), "Regional Technology coalitions: an essential dimension of national technology policy", Regional Policy, 24, 895-911 Tondl, G. (2001). Convergence after Divergence: Regional Growth in Europe, Springer, Vienna/New York. Storper, M. and R.Walker (1989): The Capitalist Imperative: Territory, Technology and Industrial Growth, B.Blackwell, Oxford. Yu, W. (1992b). "ELECTRE TRI. Aspects mèthodologiques et manuel d`utilisation". Université Paris-Danphine, Document du Lamsade, nº 74, p. 80 About the Author Autor: Juan R. Cuadrado-Roura Dirección: Department of Applied Economics. University of Alcalá. Madrid. Correo electrónico: jr.cuadrado@uah.es   DOCUMENTOS DE TRABAJO EN ANÁLISIS ECONÓMICO (EAWP) Derechos reservados 2002. El permiso para reproducir algún artículo está garantizado si Documentos de Trabajo en Análisis Económico lo acredita, las copias no son vendidas y es en acto de mayor difusión del documento. Editor: Fernando González-Laxe. (Universidade da Coruña) Director: Venancio Salcines. (Universidade da Coruña) Subdirector: Andrés Blancas. Instituto de Investigaciones Económicas (UNAM) Editor Asociado para America Latina: Luis Miguel Galindo. Facultad de Ecomomía (UNAM)