KNOWLEDGE ECONOMY Challenges for Measurement

Home , Economic statistics, Knowledge economy

PRIMA_WEB.pdf 23-06-2006 11:15:47 N O I T I D E 6 0 0 2

Conference on KNOWLEDGE ECONOMY Challenges for Measurement

Y Luxembourg, 8 - 9 December 2005

CMY

THEME E U R O P E A N Economy C O M M I S S I O N and finance WORKING PAPERS AND STUDIES Europe Direct is a service to help you find answers to your questions about the European Union Freephone number (*): 00 800 6 7 8 9 10 11 (*) Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.

A great deal of additional information on the European Union is available on the Internet. It can be accessed through the Europa server (http://europa.eu).

Luxembourg: Office for Official Publications of the European Communities, 2006

ISBN 92-79-02207-5

“KNOWLEDGE ECONOMY - Challenges for Measurement”

8 – 9 December 2005

organised by Eurostat, the Statistical Office of the European Communities

Venue:

Jean Monnet Building Meeting Room M6 Rue Alcide de Gasperi L-2920 Luxembourg 1st day – Thursday, 8 December 2005

Registration

Opening Session

Mr Günther Hanreich, Director General, Eurostat

Key Note Speeches

Mr Anthony Arundel, Professor and Senior Researcher, University of Maastricht Mr Patrick de Smedt, Chairman, Microsoft Europe, Middle East and Africa

Parallel Workshops

R & D and Innovation Chair: Mr Fred Gault, Director, Statistics Canada Discussant: Mr Giulio Perani, Senior Researcher and Head of Statistics on R & D and Technological Innovation, Italian National Institute of Statistics Speakers: Mr David White, Director, Directorate General for Enterprise and Industry, European Commission Mr Xabier Goenaga Beldarrain, Head of Unit, Open Coordination of Research Policies, Directorate General for Research, European Commission Mr Erich Ruetsche, Manager, IBM Mr August Götzfried, Head of Section, Eurostat Mr Carter Bloch, Assistant Professor, The Danish Centre for Studies in Research and Research Policy Mr Peter Kaiser, Consultant / Mr Olaf Arndt, Senior Consultant, Prognos AG

ICT impact on the knowledge based society Chair: Mr Michel Glaude, Director, Eurostat Discussant: Mr Tony Clayton, Director, Economic Analysis, Office for National Statistics, United Kingdom Speakers: Mr Dirk Pilat, Senior Economist SDT/EAS, OECD Mr George Sciadas, Chief, Information Society Research and Analysis; Science, Innovation and Electronic Information Division, Statistics Canada Mr Christophe Demunter, Eurostat – Information Society Statistics Mr Matthew Dixon, SEMTA Visiting Research Fellow, SKOPE, University of Oxford; Labour Market Adviser to CEPIS

KNOWLEDGE ECONOMY Human Capital Chair: Mr Radek Maly, Head of Unit A-1, Directorate General for Employment, Social Affairs and Equal Opportunities, European Commission Discussant: Mr Tom Healy, Senior Statistician, Department of Education and Science, Ireland Speakers: Mr Scott Murray, Director, UNESCO Institute for Statistics Mr Albert Tuijnman, Senior Economist, European Investment Bank Mr George Psacharopoulos, European Experts Network on the Economics of Education and Former State MP, Greek Parliament

2nd day – Friday, 9 December 2005

Reports from parallel sessions of 1st day Mr Fred Gault, Director, Statistics Canada, “R & D and Innovation” Mr Michel Glaude, Director, Eurostat, “ICT impact on the knowledge based society” Mr Radek Maly, Head of Unit A-1, Directorate General for Employment, Social Affairs and Equal Opportunities, European Commission, “Human Capital”

Competitiveness and Growth Chair: Mr Enrico Giovannini, Chief Statistician, OECD Speakers: Mr Heikki Salmi, Advisor, Directorate General for Enterprise and Industry, European Commission Mr Bart van Ark, Professor, University of Groningen, Netherlands

Panel Discussion Chair: Ms Marie Bohatá, Deputy Director General, Eurostat Key Note Speech: Mr Walter Radermacher, Vice President, Statistisches Bundesamt Panel: Mr Steven Keuning, Director General Statistics, European Central Bank Mr Fred Gault, Director, Statistics Canada Mr Michel Glaude, Director, Eurostat Mr Radek Maly, Head of Unit A-1, Directorate General for Employment, Social Affairs and Equal Opportunities, European Commission Mr Enrico Giovannini, Chief Statistician, OECD

End of Conference

* * * * *

KNOWLEDGE ECONOMY Conference on “Knowledge economy - Challenges for Measurement”

8-9 December 2005, Luxembourg

organised by Eurostat, the Statistical Office of the European Communities

The aim of this conference is threefold:

• To give an overview of the state of the art in measurement of the knowledge economy worldwide; • To identify major challenges for statistics with a special focus on the adjusted Lisbon agenda and gaps in the European Statistical System; • To brainstorm on possible solutions.

For this conference, a diversified group of speakers with a broad international background will present their views on the Knowledge Economy and the challenges ahead. Participants in the conference will be invited to take part in discussions and in three different workshops.

All information and working documents related to the conference will be available on:

http://forum.europa.eu.int/Public/irc/dsis/knowledgeeconomy/library

For any further information concerning the conference, please contact the secretariat:

Sheena Blair [email protected] Tel: + 352 4301 37517 Nicole Lauwerijs [email protected] Tel: + 352 4301 33564 Anne-Marie Winding [email protected] Tel: + 352 4301 34934

KNOWLEDGE ECONOMY Contents

Opening Session

Mr Günther Hanreich, Director General, Eurostat...... 9

Key Note Speeches

Mr Anthony Arundel, Professor and Senior Researcher, University of Maastricht...... 14 Mr Patrick de Smedt, Chairman, Microsoft Europe, Middle East and Africa Microsoft Europe have opted to publish their White Paper on ‘Supporting the Lisbon 2010 goals instead of the actual presentation given by Mr. de Smedt...... 24

R&D and Innovation

Report of the Chair: Mr Fred Gault, Director, Statistics Canada...... 42

Mr David White, Director, Directorate General for Enterprise and Industry, European Commission...... 44 Mr Xabier Goenaga Beldarrain, Head of Unit, Open Coordination of Research Policies, Directorate General for Research, European Commission...... 48 Mr August Götzfried, Head of Section, Eurostat...... 51 Mr Carter Bloch, Assistant Professor, The Danish Centre for Studies in Research and Research Policy...... 62 Mr Peter Kaiser, Consultant / Mr Olaf Arndt, Senior Consultant, Prognos AG...... 75

ICT impact on the knowledge based society

Report of the Chair: Mr Michel Glaude, Director, Eurostat...... 82

Mr Dirk Pilat, Senior Economist SDT/EAS, OECD...... 84 Mr George Sciadas, Chief, Information Society Research and Analysis; Science, Innovation and Electronic Information Division, Statistics Canada...... 100 Mr Christophe Demunter, Eurostat – Information Society Statistics...... 123 Mr Matthew Dixon, SEMTA Visiting Research Fellow, SKOPE, University of Oxford; Labour Market Adviser to CEPI...... 142

Human Capital

Report of the Chair: Mr Radek Maly, Head of Unit A-1, Directorate General for Employment, Social Affairs and Equal Opportunities, European Commission...... 163

Mr Scott Murray, Director, UNESCO Institute for Statistics...... 165

KNOWLEDGE ECONOMY Mr Albert Tuijnman, Senior Economist, European Investment Bank...... 192 Mr George Psacharopoulos, European Experts Network on the Economics of Education and Former State MP, Greek Parliament...... 204

Competitiveness and Growth

Report of the Chair: Mr Enrico Giovannini, Chief Statistician, OECD...... 211

Mr Heikki Salmi, Advisor, Directorate General for Enterprise and Industry, European Commission...... 213 Mr Bart van Ark, Professor, University of Groningen, Netherlands...... 223

Panel Discussion

Key Note Speech: Mr Walter Radermacher, Vice President, Statistisches Bundesamt...... 245

Summary of Panel Discussion

Ms M. Bohatá, Deputy Director General, Eurostat Mr Steven Keuning, Director General Statistics, European Central Bank Mr Fred Gault, Director, Statistics Canada Mr Michel Glaude, Director, Eurostat Mr Radek Maly, Head of Unit A-1, Directorate General for Employment, Social Affairs and Equal Opportunities, European Commission Mr Enrico Giovannini, Chief Statistician, OECD...... 257

List of Participants...... 259

KNOWLEDGE ECONOMY Opening Speech

Günther Hanreich Director General of Eurostat

The context • Various observers describe today’s global economy as one in transition At the European Summit in Lisbon in March to a “knowledge economy” or an 2000, European leaders committed the EU to “information society.” As part of this become “the most dynamic and competitive transformation process, the rules and knowledge- based economy in the world capable practices that determined success in the of sustainable economic growth with more and industrial economy of the 20th century better jobs and greater social cohesion, and need rewriting where knowledge becomes respect for the environment “by 2010. The the most critical economic resource. Lisbon strategy, as it has come to be known, was designed as a comprehensive series of Given the importance of knowledge for economic, reforms. social and environmental development, economists are trying to develop formal models In February 2005 the Commission has proposed to explain the role of knowledge in modern a new start for the Lisbon Strategy, focusing the economy. Statisticians, on the other hand, are European Union’s efforts on two principal tasks concerned with the measurement, be it the extent – delivering stronger, lasting growth and more to what the economy is knowledge based, or and better jobs. impacts of crucial factors related to knowledge on competitiveness and growth. The spring European Council of March, as well as the European Parliament and the European social This conference should give an overview how partners, gave full support to the Commission’s good we are at measurement of the knowledge proposal. The Mid-Term Review made clear that, economy not only in the EU but worldwide, what without swift action, the Lisbon targets would be are the main challenges and suggest also some missed. solutions. It will discuss also the Lisbon strategy, competitiveness of the EU and what to measure The concept of a knowledge-based economy till 2010 to meet European policy needs. occupies a prominent place in the knowledge economy. So what are we talking about? I quick One aim of the conference is to relate statistical search about the term “knowledge economy” work with public and private decision making. showed me that the different definitions share We thus tried to attract not only professional two essential elements: statisticians and economists but also policy makers, researchers and businesspeople. I am • The knowledge economy refers to the use of very happy to say that we succeeded and there is knowledge to produce economic benefits. a wide spectrum of participants. We may expect It is thus of immediate importance to policy interesting and stimulating discussions which, I makers, researchers and citizens alike. am sure all of us look forward to.

KNOWLEDGE ECONOMY: OPENING SESSION The workshop sessions of the conference Communication on R&D and innovation. Secondly the business perspective will be Economic science does not yet offer a single and added with questions such as: how does a big universally accepted theoretical model explaining multi-national translate R&D into innovation the role of knowledge and assessing its impact and innovative products? Is the R&D produced on society. However, several crucial knowledge- in-house or bought on the market? related factors have been identified, which have a strong influence on growth and competitiveness Then the data producer perspective comes with a of modern societies. Among them are first and look on the progress made on European and world- foremost R&D and innovation, Information and wide R&D statistics as well as on innovation Communication Technology, Human Capital and statistics where new types of innovation such entrepreneurship. The four workshop sessions as marketing and organisational innovation are of the conference will focus on these factors. defined in the new Eurostat/OECD Oslo Manual In the following, I will allow myself to outline 2005. The measuring and recording of the the objectives and the background to each internationalisation of R&D – a phenomena being workshop. more and more present around the world - needs surely more attention in this context. Finally, data users will have another word in presenting Workshop R& D and Innovation national and international R&D and innovation R&D and innovation – being the topic of the landscapes and scoreboards based on official and first parallel workshop – are in the centre of the also non-official data. Lisbon strategy and in general considered as key factors for economic development and growth. Workshop Information and Communication The European Council Summit in Barcelona of Technology (ICT°) 2002 made this very clear in setting the strategic goal of getting R&D expenditure in the EU to a I already mentioned that this year’s a review of level of 3% of GDP by 2010. the Lisbon strategy led to a major revision of its goals until 2010. The Commission set out the Both R&D and innovation remained also new i2010 strategy which will be the basis for very high on the political agenda within the benchmarking eEurope over the next five years. mid-term review of the Lisbon agenda, focussing This sets the context for the workshop on ‘ICT on a renewed partnership for growth and jobs. impact on the knowledge based society’. The Knowledge creation via R&D and innovation is aim of this session will be to critically review the seen there as a critical factor with which “Europe monitoring of the information society over recent can ensure competitiveness in a global world years with a view to revising measurements in the where others compete with cheap labour or past and discussing new challenges in the future. primary resources”. Presentations will be given on three key areas for statistical measurement in the future: The economic The session on R&D and innovation aims to impacts of ICT, e-skills and their measurement and contribute to this reinforced political context the shift of behavioural patterns in digital times. I from various perspectives. First, it will address am expecting a vivid discussion. the policy context based on views expressed by the European Commission in. The information To benchmark the ICT-driven development, the needs in general and also for monitoring and Commission established Information Society evaluation purposes on R&D and innovation will surveys on ICT usage in enterprises and households be presented, also related to the new Commission in 2002. Eurostat is providing the model

10 KNOWLEDGE ECONOMY: OPENING SESSION questionnaires for these annual Community surveys ‘have-nots’. In 2004 there were several reasons for that have been developed in close cooperation with such a ‘digital divide’, including the availability the National Statistical Institutes (NSIs) and the of technology, the costs of access and equipment OECD. At the outset, they concentrated on access as well as age and education. In conclusion: The and connectivity. Later, annual adaptations to fit user benchmarking of the digital divide will continue needs have been made, such as on specific sectors to be high on the agenda for the future Information like the financial sector, or specific areas such as Society surveys, with emphasis on addressing e-government, broadband access and the use of key areas such as digital literacy and e-skills. e-skills. In April 2004 the European Parliament and the Council adopted Regulation (EC) Another challenge for measuring the Information 808/2004 covering these surveys, which will Society is the economic impact of ICT. Such ensure harmonizes data for all 25 Member States. impact measurement is a complex process, which It is a framework regulation which allows certain will go beyond the current Information Society flexibility, so the surveys can be adapted annually data collection within the Community. Among to encompass newly evolving needs by users and others it should include ICT investment, and it decision makers. should address the growing need to identify how ICT adoption affects business behaviour and The Information Society statistics have become performance and productivity. successful markers for the development in the Community. The co-operative data With regard to the measurement of ICT collection guarantees comparability, easily investment within the ESS Eurostat launched accessible through Eurostat’s database free of a project in March this year. The pilot phase is charge. These data indicate the potential for aimed at developing a methodology for future improving productivity and the quality of life data collections. 10 members of the ESS are due to the developments of broadband and participating. Results of that project are expected e-commerce: Citizens have more convenient for autumn next year. access to information and communication tools. Businesses are benefiting through the take up of In addition, Eurostat launched a project on ICTs to make efficiency gains as well as reach a developing a work process for ICT impact wider customer base and boost competitiveness. assessment within the ESS last month 89% of EU-25 enterprises actively used the (November 2005). The aim is twofold: To Internet in 2004, with 65% of connected provide additional detailed data to the current enterprises having a website. 47% of individuals Information Society statistics without creating had recently used the Internet, mainly looking additional burden on respondents. The current for information and online services and for Community Information Society statistics are communication, preferably by e-mail. About restricted to aggregate data provided by the NSIs. one in three Internet connected households had The project aims at developing a work process a broadband connection. In Denmark, Germany, to link existing micro-data of the ESS that is Estonia and Finland around three quarters of relevant for identifying the impact of ICT on Internet users who were unemployed at the time enterprises’ outcome. Such micro-data are based of the 2004 survey were looking for a job or sent on the enterprises as respondents. They refer to a job application over the Internet. a number of data sets held at Eurostat, such as the Information Society survey on enterprises, However: A gap remains between the Internet the Community Innovation survey, the Structural users and non-users, or between ‘haves’ and Business statistics and others.

KNOWLEDGE ECONOMY: OPENING SESSION 11 Workshop Human Capital • The shift from an initial formal education towards a lifelong learning paradigm: The objective of the session on Human Capital data on the education systems, which will be to better identify appropriate strategies vary substantially from a country or sub- to improve the measurement of the contribution country to another and are continuously of human capital to growth. This is of immediate evolving, present already challenges for relevance to the revised Lisbon strategy strongly comparison. Data on lifelong learning emphasises the relationship between growth, pose even more difficulties due to the competitiveness and human capital. Member multiple forms of leaning and the various states are strongly urged for reforms. They are actors involved. required to improve (and not only increase) • The shift from managing the logistics of their investment in knowledge and in human the education systems to improving its capital through the provision of better education quality: data are required on the education and skills to their citizens in a lifelong learning process itself and on its main actors, i.e. perspective. The rapidly evolving knowledge the teachers and trainers. economy imposes the continuous updating and • The shift from analysing the output of the renewal of skills, so as to adapt to changes. Low education system (graduates) to analysing level of qualifications and outdated knowledge and the outcomes of lifelong learning (skills, skills of workers, inequity in access to education employability): data are additionally and lifelong learning, lack of efficiency of the required on skill levels of the individuals, education systems and under-developed lifelong but also on skills needs of different jobs learning strategies are obstacles to the realisation and sectors in the labour market. of the Lisbon objectives. • The shift from analysing returns on education to analysing returns on lifelong Nowadays, 100 million of Europeans are learning: data are required to support registered in the education systems. Another this analysis and also answer questions 100 million of adults declared to have had some about accumulation and obsolescence of learning activities during the last 12 months. knowledge. However, only 22 percent of the working age population has a tertiary graduation in Europe, In this context, the role of public policy is also compared to 38% in the US. evolving from focusing only on the provision of education to facilitating the development of Production of statistics on human capital are the lifelong learning offer and the access of its definitely challenging, even when only examining citizens to it. The demand of policy makers for the impact of human capital on employment and data on efficiency and equity remains therefore growth. However, its benefits for the individual crucial, even if its focus is broadened. As said (in terms of social condition, health and personal above, the session on human capital will focus on development) as well as for the society (in terms these challenges, and upon their new statistical of citizenship, crime, environment, etc) should data requirements. not be overlooked. Workshop Competitiveness and growth Several recent (and less recent) shifts in the policy issues that are in the centre of interest in The mid-term review of the Lisbon strategy Europe and worldwide, require improved or new revealed, that the overall performance of the statistical data; just to name a few: European economy has been disappointing. The

12 KNOWLEDGE ECONOMY: OPENING SESSION realisation of knowledge economy, completion of Framework Programme for Research in order to the internal market and promotion of competition, create a comprehensive productivity database. the establishment of a favourable climate to Mr. van Ark, one of tomorrow’s speakers, is businesses and an adaptable and inclusive labour leading this work. Eurostat and the Economic market are identified among the key issues for and Financial Affairs Directorate General as well increasing the economic growth and higher as several Member States are involved. Efforts productivity. Sound macroeconomic conditions have also been made during the last years to make the crucial framework for success. increase availability and detail of service sector statistics and data on business demography and Europe is facing both internal and external entrepreneurship and on cross-border activities. challenges. On the one hand, Europe must One important example is of course the Structural address the challenge of ageing populations and Indicators set up in a timely manner by Eurostat the need to increase labour productivity. On the together with the Member States and other other hand, the increased pace of globalisation stakeholders. has exposed the EU economy to mounting competition from abroad. The range of economic The session on competitiveness and growth will activities exposed to external competition enlarge and build on the themes covered by the first has widened, now including the production day’s parallel sessions by providing a summary of both high-tech and labour intensive goods view of competitiveness in the macro context. It and services. To respond to these challenges, will study how competitive the European Union a renewed Lisbon strategy has been designed is at present and in what dimensions. The role focusing the European Union’s efforts on two of statistics in light of the challenges on the path principal tasks – delivering stronger, lasting towards 2010 will be discussed, by examining growth and more and better jobs. Action plans what is, what should and what can be measured in both by the individual Member States and statistical terms with an emphasis on guidance for Commission have been drawn up. future actions for the European Statistical System.

Statistical measurement plays a paramount Conclusion role in policy creation and monitoring. For that purpose, many efforts have been made to This conference should start a series of regular enable better measurement of productivity and events of this type and I am sure we’ll come up competitiveness. For example, one major focus with other interesting and policy relevant topics. of work has been to improve the availability Your interest in the work of official statisticians, of labour input measures in terms of hours demonstrated by the high number of participants worked. Another example I would like to give is not only rewarding but also motivating for you is a project that was launched under the 6th Eurostat.

KNOWLEDGE ECONOMY: OPENING SESSION 13 From the 19th to the 21st century: Indicators for the Knowledge Economy

Anthony Arundel MERIT

The term the ‘knowledge economy’ is becoming evaluated four main characteristics of a knowledge an inevitable part of modern life – frequently economy that have been widely discussed in the turning up in the popular media, in policy economic and innovation literature. I think two documents, and in academic journals. It is also of his characteristics are closely intertwined, so I frequently combined with globalization, with the have reduced them to three. knowledge economy both driving the process of globalization and offered as a solution to the First, knowledge is seen as quantitatively and problems that globalization creates. We have all qualitatively more important than in the past as an heard that Europe must develop into a knowledge input into production. More formally, knowledge economy in order to compete not only with the is viewed as an additional input to labour and United States, but in the future with China and capital. Smith notes that this particular claim to India. The proposed solution usually involves being a ‘unique’ feature of a knowledge economy both working for longer hours and an acceptance does not survive scrutiny as all economies are of greater income inequality in order to provide based on knowledge and technology. The intensive an incentive for our best scientists, entrepreneurs development of the chemical and electrical and engineers to remain in Europe, rather than sectors in the late 19th century made far reaching migrating en masse to California. changes to the economies of the time – possibly The idea of a knowledge economy has also, even more so than the application of Information unfortunately, become intricately connected with and Communication Technologies (ICT) today. another concept – that of R&D. The ‘knowledge Furthermore, the idea that knowledge is a separate economy’ conjures up images of an elite group input from labour and capital is peculiar – either of scientists and PhD holders, or computer whiz knowledge is part of labour – held in the minds kids who spend their lives hunched over their of researchers, technicians and other staff, or it is keyboards. embodied in capital as new equipment, buildings, and infrastructure. But what is this ‘knowledge economy’? Is there anything unique about it that would require a Second, knowledge is often viewed today as a shift in European policy to promote new types product that can be bought and sold, for instance of knowledge? And, would this require the through licensing patents, whereas before development of new indicators? knowledge was presumably freely available for all. This second claim for a unique characteristic A few years ago, Keith Smith (2002) wrote an of knowledge also does not survive scrutiny insightful evaluation of the idea of the knowledge because knowledge has always been bought economy. As part of an exploration of relevant policy and sold – as embodied knowledge contained in responses, he asked the question: What, if anything products. The licensing of knowledge contained is unique about the knowledge economy? He in patents is only a very small share of total

14 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES investment, although it is more important in their earnings, but at a cost of time spent in some sectors with high political visibility, such monitoring and learning how to apply IPM. In as pharmaceuticals and biotechnology. the late 1990s IPM was largely replaced with genetically modified Bt cotton, which produces The third characteristic is the rise of ICT and its an insecticide that kills many of the cotton insect adoption by every economic sector. This has created pests. In this case, the location of the necessary enormous improvements in our ability to manage knowledge to counteract insect pests shifted from and disseminate information around the globe. an embodied form in insecticides, to the minds of Even though the late 19th century also introduced farmers, and then to the cotton plant itself. the telegraph, which opened up a first phase of globalization, ICT has made the process much Keith Smith, while remaining unconvinced faster and created enormous advances in the amount about many of the claimed unique characteristics of information that can be transferred at next to no about the knowledge economy, added an cost. Without ICT, Europe would not be facing additional feature that he thought is perhaps competition from India in many service sectors. the defining characteristic. I personally do not see this additional feature as unique, but as an A common claim is that the generic features intensification of a process that has been going of ICT, influencing all aspects of our lives, on for a long time. Nevertheless, it is a crucial constitutes a unique feature of the knowledge insight that has driven much of Keith Smith’s economy. Yet we have been through this before, research over the past decade. This is the concept with the application of electricity between 1890 of ‘distributed knowledge bases’, or the use of and 1920 to both production and to private increasingly larger circles of advanced knowledge consumption. Conversely, I would suggest that across all economic sectors. the modern ‘knowledge economy’ is part of a long history of technological development. Technology An excellent example of this is the fishing sector, develops in bursts of discovery in which low- which along with agriculture, is one of the least knowledge tools and goods are replaced with R&D intensive sectors in modern economies. In tools and goods with higher information content, fact, the R&D expenditures in fishing are so low as with the continual embedding of greater amounts to be almost unmeasurable. Nevertheless, modern of knowledge. This process applies to production fishing ships bristle with advanced technology machinery – with labour replaced with better to seek, find, and catch fish. They are equipped machines then by automated machinery, and to with satellite communications, global positioning the replacement of the physical movement of systems, sonar technology to find fish, optical goods, ideas, and people with movement via technology to sort fish, computer systems to electronic forms. monitor catches and their preservation at optimal temperatures, quick deep freezers, and many A good example of this process is from agriculture different types of nets. In fishing, innovation occurs – the production of cotton. Cotton is susceptible to through buying technology from a diverse range severe pest infestations. After 1945, insecticides of suppliers rather than developing technology in- were introduced to destroy pests. These were house. In the race to catch the last fish, any new costly and also created environmental problems. technology that provides a competitive advantage In the 1980s Integrated Pest Management (IPM) is quickly adopted. was introduced. Essentially IPM used monitoring and knowledge about pest infestations to reduce To summarize, modern economies – whether pesticide use (Cowan and Gunby, 1996). Farmers called the knowledge economy, the information could save money on pesticides and increase society, the digital era, the learning economy, the

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 15 intangible economy, or the network economy, these indicators is brought into question by do not differ qualitatively from earlier economic three major changes: the ongoing decline systems, but they do differ in terms of intensity. in manufacturing employment, shifts in the Perhaps the deepening of the importance of location of employment, and the rise of China learning, knowledge, and creativity – combined and India as players in innovation. with ICT and globalization, has produced a fundamentally different type of world. Therefore, Almost all countries have experienced an I will use the term ‘knowledge economy’ as a ongoing decline in manufacturing employment, short-hand for all of these changes. replicating the decline in agricultural employment in the 19th and 20th centuries. The decline has As a note of caution, even though the developing been particularly steep in the UK, from almost ‘knowledge economy’ has brought many benefits, 30% of total employment in 1976 to a bit over it has not brought us two important things that 10% today. The declines also hold when using were seen during the 1950s to 1970s as part of absolute numbers of employees, although they our future. The first is a reduction in working are not as steep. Nevertheless, our best indicators hours. With the exception of France, where and research still focuses on R&D and creative shorter working hours had been legislated, the innovation, which reflect conditions in the average hours worked per week has changed very manufacturing sector. This is a problem that we little or increased, as in Germany and the United have been aware of for over a decade, but real States (ILO, 2005). The second thing that has change in terms of evaluating innovation in the not happened is an increase in our general well- economically dominant service sector has been being, as measured by happiness. There has been slow (Tether, 2004). no discernable difference in the UK since 1973, even though incomes have increased 80% (Marks, The second major chan ge is a shift in the location 2005), and a detectable decline in happiness in of employment, including an uncoupling in the the United States since the mid 1950s (Layard, location of research and production activities. The 2003). Perhaps this is partly related to the lack increasing globalization of production, including of a decrease in working hours. I raise these the production of knowledge itself, is a major two cautionary notes because if the knowledge development. Manufacturing jobs have long been economy means working more hours with no moving to lower cost locations, made possible by increase in well-being, it could contradict other modern logistics, software and ICT. The same is important European policy goals to improve the happening to service sector jobs. quality of life. Van Welsum and Vickery (2004) recently estimated Challenges for the Knowledge Economy the percentage of service sector employment that is ‘susceptible’ to off-shoring. A job is defined What are the challenges for the future of the as ‘ susceptible’ if it meets the following four knowledge economy, and what indicators do criteria: it is based on intensive ICT use, output we need to address them? To date, the most can be delivered using ICT, the output has a high widely used indicators cover investment in information or knowledge content, and face- the production of scientific and technical to-face contact is not needed. The percentage knowledge, including R&D, patents, and the of such jobs that could conceivably be done at flows and stocks of scientists and engineers. much lower cost in India is approximately 18% All are indicators of creative innovation, or to 19%. Of note, many of these jobs are the well- the development and commercialization of paid knowledge intensive jobs that are supposed inventions. The relevance and reliability of to be replacing manufacturing employment.

16 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES These shifts in the location of employment are of whom are drawn from China, peaked in 1996 partly the result of near negligible communication at around 5,300 and has been gradually declining costs and vastly improved logistics. They also since then. Conversely, the number of doctoral create serious problems for our indicators, which degrees awarded in China took off in the mid are collected and refer to the national context. 1990s, increasing from approximately 2,000 in As an example, our indicators of national R&D 1995 to 11,000 in 2000 (Wyckoff and Schaaper, intensity are based on the total R&D performed 2004). The two phenomena are probably linked in a country divided by total production in the – increasing educational opportunities at home country. But what if a large percentage of the might have led to a decline in the number of R&D is performed domestically, but an increasing Chinese seeking education abroad. This should share of the production is performed abroad? create concerns about the possible long-term Or, a large share of the R&D is also performed success of policy initiatives that depend on abroad? We have yet to develop indicators that the immigration of the highly-skilled to solve can fully account for these effects of globalization shortages within Europe, or to provide foreign on national statistics. students willing to pay high fees to subsidize tertiary education, as in the UK. The third major challenge is the rise in inventive activity, proxied by R&D, in India and China. The policy response The OECD (2005) estimates that total R&D in China increased 350% between 1996 and The major response on the part of the European 2002 and UNESCO data (2005) indicates that Union to the perceived challenges created by the total R&D doubled between 1996 and 2000 in knowledge economy and globalization was the India. Although the absolute estimates of R&D 2000 Lisbon European Council’s goal for Europe expenditures in both countries are probably to become, by 2010, ‘the most competitive and subject to large margins of error, there is little dynamic knowledge-based economy in the world, doubt that R&D has been growing rapidly in both capable of sustained economic growth with more countries. The implication is that both China and and better jobs and greater social cohesion’. A key India are increasingly likely to become major element of this goal was R&D. The conclusions players in the development of new technology of the Lisbon Council also mention ‘innovation’ and products and to compete with the developed several times, but the context in how this term world through innovation. was used strongly suggests that the Council interpreted innovation as almost identical to A related phenomenon is increasing educational creative, R&D based activities. It should therefore and economic opportunities at home for Chinese be no surprise that the 2002 Barcelona European and Indian students, researchers and entrepreneurs, Council focused on R&D, setting a target of a which could also mean a decline in the numbers three percent R&D intensity for Europe by 2010. of highly skilled scientists, engineers and students in these fields that seek either employment or The Barcelona target is based on one-third of education abroad. As an example, the pool of R&D from public spending and two-thirds foreign graduate students has been indispensable from business expenditures on R&D, or BERD. to the American ICT sector and it is well known that MERIT recently estimated the conditions that the tightening of visa regulations after September would be required for Europe to achieve a goal 2001 has led to a fall in the supply. However, this of 2% BERD (Arundel and Hollanders, 2005). decline began well before September 2001. The Since R&D intensities vary enormously by number of East Asian science and engineering sector, the study used the OECD’s ANBERD PhD graduates from American universities, most database to calculate sector-specific trends in

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 17 R&D performance in each country and to estimate a predicted BERD intensity of 1% in 2015, future R&D intensities within each sector for 13 and Spain with a predicted BERD intensity of European countries for which ANBERD data are about 0.9%. The reason for this is differences available. The 13 countries account for 95.4% of in industrial structure – countries with little total BERD among the EU 25 countries in 2002 activity today in R&D intensive sectors cannot and 93% of GDP. Therefore, the ability of the possibly reach a 2% target. This could only be EU to reach the 2% Barcelona target for BERD feasible with massive and difficult changes to the almost entirely depends on business R&D in industrial distribution of these countries. these 13 countries. Going beyond R&D In 2002, the aggregated business R&D intensity for these 13 countries was 1.22%. An The Barcelona goal of a three percent R&D extrapolation to 2015, using sector and country intensity is neither realistic nor achievable, specific growth rates in R&D, estimates an but the more serious question is if this goal is aggregated business R&D intensity across all even necessary or desirable. The assumption sectors of 1.35% - far short of the 2% target. is that all of the good things that we want

��

Only the most optimistic and implausible – including economic competitiveness and assumptions would result in reaching the goal growing productivity – depends on high levels by 2015. This estimate assumes the best case – it of R&D being performed within Europe. These applies the highest observed growth rate in R&D assumptions are probably not valid, suggesting in each specific sector in any of the 13 countries that the emphasis of the Lisbon and Barcelona to the same sector in all countries. The results Councils on R&D has created five lost years in the are given in Figure 1. development of appropriate policy and indicators. The Councils overemphasis on R&D could have Even in the best-case estimates, several of the distorted policy actions and, ironically, failed to 13 European countries would not achieve the account for some of the main challenges posed 2% BERD intensity, particularly Italy, with by a knowledge economy.

18 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES R&D, both as a concept and as an indicator, knowledge bases derive their economic impacts was developed to measure the last great burst from the diffusion and application of knowledge of economic change driven by technological across all sectors. discovery – the period between the 1880s and the 1920s in which the chemical, automobile An example of the limitations of R&D is its and electrical industries took off. Formalized effect on productivity, which is one of the main research labs were an organizational innovation benefits that is supposed to flow from higher that first began in the German chemical industry investment in R&D. Careful research using in the 19th century and spread to other large micro data at the firm level does find a positive firms in the electrical equipment and automobile correlation between R&D levels and greater manufacturing sectors. All three of these sectors productivity. But the relationship between R&D were at the leading edge of technology for their and productivity at the national level is not time. R&D as a concept therefore belongs not simple, as shown in Figure 2, with no correlation just to the last century, but to the one before at all between labour productivity and lagged that. The first experiments with measuring R&D national R&D intensity for countries with a per date back to 1917. It took over 60 years, until capita income above 20,000 USD (PPP) per 1981, before R&D became a reasonably reliable, year. The lack of a simple correlation between and internationally comparable, Science and R&D and labour productivity is probably due Technology indicator. to many different factors, including investment in productivity enhancing technology by firms This is just around the time when innovation that do not perform R&D – as in the fishing scholars realized that indicators for R&D alone industry. could only tell part of the story. The problem is that R&D only measures a narrow set of Innovation involves not only creative, R&D- innovative activities, telling us very little about based activities, but spans everything from buying the formation of ‘distributed knowledge bases’ new technology off the shelf to a billion Euro, that were identified by Keith Smith as the in-house research project, and all permutations in essential feature of a knowledge economy. These between. We have had empirical confirmation of

��

Data source: Eurostat Structural Business Statistics

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 19 this since the first innovation surveys in the early everything we can to ensure that the over 2 1980s, and better empirical evidence for this since billion people of China and India, including the first European Community Innovation Survey many of the most brilliant minds on the planet, (CIS) from 1993, which found that a very large have the opportunities to develop new products, percentage of firms innovate without performing processes, and organisational methods that we R&D. Yet a focus on R&D indicators leads us to can then acquire and all benefit from. forget this fact. Community Innovation Survey indicators At the same time that the Lisbon Council was focussing on R&D, Porter et al (2000) were I would now like to return to the Community sensibly pointing to the range of options: Innovation Survey, which provides us with a very valuable set of new indicators. The CIS is the “What is important is that a best available source of indicators that capture country participates in the newest the full range of innovative activities in the technologies and innovations, not development and use of products and processes whether it innovates [creatively] and of indicators for the flow of knowledge itself. To raise GDP through across distributed knowledge bases. In the future, technology-related activities, the CIS should also provide us with valuable a country needs to achieve indicators for organisational innovation – an value-added at some stage of area where indicators are lacking, even though the process, not necessarily the research has consistently found substantial inventive stage”. productivity benefits from combining ICT use and organisational change.

There are terms of trade advantages in being The first CIS confirmed that many firms a creative innovator rather than adopting innovated without performing R&D. The second technologies developed elsewhere. Creative, CIS, implemented in 1997 and covering the R&D-based innovation can have large economic years 1994 to 1996 inclusive, added a question payoffs. Yet, by definition, most firms, individuals on how each firm developed its innovations. and countries will acquire the majority of their This information has been used by Bruce Tether innovations from somewhere else. With a few of CRIC in Manchester and others to look more rare exceptions of process innovations that can be deeply into how firms innovate. The question was exploited by a single firm – Pilkington’s float glass also included in CIS 3 and CIS 4. method comes to mind – the economic benefits of creative innovation depend on widespread As part of an experimental analysis by Eurostat diffusion and adoption. and MERIT, we used the CIS-3, including the question on how firms innovate, to assign firms We can see the problems with the focus of much to one of four modes of innovative capabilities. current policy thinking in both Europe and the Experimental results are given in Figure 3 for United States on creative innovation with China Finland and Spain. The vertical axis gives the and India appearing above the horizon as major percentage of all firms that perform R&D on a R&D performers. This is seen in many policy permanent or occasional basis. The ‘strategic’ circles as a threat to domestic competitiveness. innovators include R&D as a core activity of This is absurd, and requires viewing innovation their firm, whereas the intermittent innovators as a zero sum game, which it most emphatically only perform R&D as necessary - the technical is not. Instead, we should welcome and do advance of their products and processes is likely to

20 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES be much less than that of the strategic innovators. statistic such as the percentage of firms that The horizontal axis measures innovation that is innovate, or national R&D intensity. They also largely based on the diffusion of technology. The show that Spain is weak on in-house creative adopters only innovate through purchasing new innovative capabilities. This could spark a policy technology – for example off the shelf, as in our interest in moving ‘upwards’ the innovative example of modern fishing. The ‘modifiers’ bring capabilities of Spanish firms, for instance by

��

� � ��

��

Data source: Eurostat, CIS-3

in new technology developed outside their firm, assisting firms to move from adopters to modifiers. but they have some internal capabilities to adjust However, it is important to remember that there technology to their own requirements or make is no best distribution of how firms innovate. incremental improvements. Technology adoption, as in fishing, could be the most cost-effective method of improving value- The shaded central area in Figure 3 gives the added and productivity. average percentage for 22 European countries. The sum of the percentage of strategic, The CIS is an exceedingly rich data set that intermittent, modifiers and adopters equals the provides many opportunities for both developing percentage of all firms that innovate, with the indicators and for more detailed econometric remainder equal to the share of non-innovative analysis of the effect of innovation on productivity firms. This is 55% of Finnish firms and 67% of and the factors that influence knowledge flows. Spanish firms. The distribution of innovative The value of the CIS was slow to be recognized Finnish firms is squeezed on the vertical axis, by the academic community, with few papers showing national strengths in creative, R&D using CIS data until 2001, when there was a rapid based innovation. Conversely, the distribution increase in its use by academics. Unfortunately, of innovative Spanish firms is squeezed on the even though the primary function of the CIS is horizontal axis, showing strengths for innovation to provide information of value to policy makers, diffusion, particularly through the adoption of it has not yet fully played this role. CIS data are new technology developed outside the firm. rarely used in major national policy documents and the results to date have seldom been used These graphs give a much more realistic picture to develop policy. This is partly the fault of the of national innovative capabilities than a simple academic community – about 80% of academic

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 21 papers based on the CIS make no reference at all appropriate policy instruments for innovation to the policy relevance of their results. wherever it occurs – in all sectors, and by all of the different methods that firms use to innovate. Another factor that has limited the impact of the CIS is that the results were hard to find and Conclusions not widely disseminated. This has also begun to change, partly due to the same Lisbon Council that We need to develop indicators that are appropriate gave far too much stress on R&D. The Council for the demands of the 21st century, rather than requested the European Trend Chart project to only relying on indicators that reflect methods of develop indicators to chart progress towards innovating that have their origins in the 19th century. a knowledge economy. Trend Chart includes Indicators for R&D are certainly very valuable several non-R&D based measures of innovation and will provide, for the foreseeable future, an that were derived from the CIS. CIS indicators are now also publicly available from Eurostat’s essential indicator for policy development, but online database NewCronos, which should also R&D indicators alone are incapable of capturing help encourage their use. the myriad variety of ways in which firms innovate or how knowledge diffuses to firms. In June 2005, as part of a series of interviews with the European policy community on their use In order to help the policy community – and the of CIS data, one of the respondents told me this business community – adapt to new challenges, we about the CIS: need indicators that are relevant to a knowledge economy. We need indicators that can measure the flow of embedded knowledge (as illustrated by “The CIS delivers data that prove the fishing sector), value-added from knowledge that R&D and innovation are not and inventive activity, and the productivity the same, but the CIS has had improvements from innovation. We particularly little impact on specific policy need to have better measures of organisational instruments. It has had a major innovation, which combines with ICT to create effect on developing innovation global supply chains, new logistics, and many policy in general and influenced a other production methods with substantial major programme that explicitly benefits for productivity. We also need to tackle focuses on innovation and not on the effect of globalisation on many indicators. R&D. This is a breakthrough and R&D, skilled people, networks for knowledge the most significant recognition to date of the value of the CIS”. flows, and production now move across borders, creating serious problems for indicators based on nation states. So, the CIS and indicators derived from it are having an influence at last. The programme that As a final comment, I would like to note that is referred to here is the new Competitiveness and recognizing the fact that knowledge economies Innovation Framework Programme, which I think encompass a range of innovative behaviours and is a major step forward compared to the policy strategies, all of which carry benefits, provides us recommendations of the Lisbon and Barcelona with a more inclusive view of society. Learning Councils. This new programme, as noted in this and absorptive capacity are at the centre – all of us quote, takes a much broader view of innovation can participate in this world and contribute. The that is not limited to R&D in high technology knowledge economy is not only the playground of sectors. It opens up the promise of developing scientists, engineers, and R&D performing firms.

22 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES References Porter M, Sachs JD, Warner AM, Cornelius PK, Levinson M, and Schwab, K. The Global Arundel A. and Hollanders H. Policy, indicators Competitiveness Report 2000, Oxford University and targets: Measuring the impacts of innovation Press, Oxford, 2000. policies. European Trend Chart on Innovation, European Commission, December 2005. Smith, K. What is the knowledge economy? http://trendchart.cordis.lu/scoreboards/score- Knowledge intensity and distributed knowledge board2005/pdf/EIS%202005%20Policy%20and bases. INTECH, Discussion Paper Series %20Targets.pdf, 2002-6, United Nations University, Maastricht, 2002. Cowan R. and Gunby P. Sprayed to death: path dependence, lock-in and pest control strategies, Tether, B. Do services innovate (differently)? The Economic Journal 106: 521-542, 1996. CRIC Discussion Paper 66, CRIC (Centre for Research on Innovation and Competition), ILO (International Labour Organisation), Manchester, November 2004. LABORSTA Labour Statistics Database, Geneva, 2005. UNESCO. Institute for Statistics, stats.uis.unesco. org, last accessed December 2, 2005. Layard, R. Happiness: Has social science a clue? Lionel Robbins Memorial Lectures, March 2003. van Welsum D and Vickery G. Potential offshoring of ICT-intensive using occupations, Marks, N. Measuring what matters: Assessing DSTI/ICCP/IE(2004)19, OECD, Paris, December people’s quality of life and well-being across Europe. 2004. New Economic Foundation, London, 2005. Wyckoff A. and Schaaper M. The changing OECD. Main Science and Technology Indicators dynamics of the global market for the highly- (MSTI), OECD, Paris, 2005. skilled. OECD, table 4 and figure 8, 2004.

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 23 Microsoft White Paper Supporting the Lisbon 2010 goals

Autumn 2004

The road to Lisbon 2010 – the role of ICT eEurope+ Action Plan played an important role in stimulating economic dynamism and innovation In March 2000 the European Council assembled in in the preparations for EU accession and continues Lisbon set an ambitious strategic goal: to become, to help drive integration in the new EU. by 2010, “the most competitive and dynamic knowledge-based economy in the world capable ICT plays an important role in improving of sustainable economic growth with more and productivity which in turn is the main driver of better jobs and greater social cohesion.” economic growth. By enabling creativity and efficiency in information processing, sharing and At the mid-way point to 2010, it is becoming utilisation, ICT tools and solutions encourage the critical to move the reform agenda faster and development of widespread and consistent high deeper if the ambitious growth targets set are to performance. be reached. This is all the more urgent in light of the historic accession in 2004 of ten new countries One of the groundbreaking aspects of the to the European Union and the framework for Lisbon agenda was the appeal by the heads of negotiations on accession by additional new European governments to businesses’ new sense Member States in the coming years. of corporate social responsibility as an asset for Europe’s competitiveness goals, particularly The achievement of the Lisbon goals requires through lifelong learning, enabling opportunity the enabling of individuals, enterprises and and social inclusion. Microsoft’s Corporate communities to become more competitive and Citizenship activities are built on four key areas innovative across the European Union. It also which reflect the importance of this call: requires increased commitment and collaboration • Internet Safety and Policy Leadership to by both the public and private sectors to ensure that address key societal challenges in the ICT Europe, in its broadest sense, makes sustainable sector such as online child safety, privacy, progress towards an inclusive Information Society security and spam; for all. • Responsible Business Practices to ensure integrity and transparency in how we conduct Like the EU, Microsoft sees the period 2000- our business and to provide a healthy workplace 2010 as the ‘Digital Decade’ in which technology environment to our employees; acts as a key driver in enabling individuals, businesses, governments and communities to • Economic Opportunity to strengthen realise their full potential. local economic development, growth, competitiveness and innovation, a priority The role of Information and Communication which is at the heart of the Lisbon Agenda; Technologies (ICT) in helping to deliver the • Digital Inclusion and Education to enable Lisbon vision is undisputed; the eEurope people, communities and nations to access the Action Plan provides the roadmap to leverage benefits of technology tools, skills and solutions technology to meet Europe’s shared goals. The through lifelong learning and education.

24 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES The Lisbon agenda extends into almost every new economy. Microsoft is contributing through aspect of European and Member State policy, four main R&D facilities in Europe. from the approach to employment and social affairs to the protection of intellectual property Microsoft Research (MSR) Cambridge, UK rights and innovation; it affects the ways in which Bringing together Europe’s most creative minds, businesses perform and citizens work and live. MSR-Cambridge has over 80 full-time staff of While this paper does not attempt to address some 16 different nationalities. The facility aims every aspect of the Lisbon process, it sets out the to accelerate the next generation of software primary ways in which Microsoft is working to innovation, driven by fundamental challenges support the Lisbon goals and contribute to the and long term vision, not by today’s market information society through partnerships with demands. A key part of its mission involves industry, government and communities across the research to make computers easier to use and new European Union. more cost-effective, and to make developers more productive. Research areas include operating Building a competitive and dynamic systems, networks & distributed computing; knowledge-based economy machine learning & perception; programming principles & theory; and interactive systems. The The Lisbon Agenda sets a key marker for facility has produced significant and tangible increasing Europe’s innovation investment – results in developing Tablet PCs and i2i, a new from 1.9 percent of GDP to 3 percent by 2010, technology enabling cameras to follow movement with the private sector contributing two-thirds of during Web-based video conversation. the increased investment. Achieving this involves several fundamental competitiveness challenges: European Microsoft Innovation Centre (EMIC), • Keeping pace with the research spending of Aachen, Germany key trading partners EMIC provides a focal point for Microsoft’s col- • Retaining and attracting the best researchers laborative efforts with industry and academia in • Improving the foundations for commercialisa- Europe on applied research projects such as those tion of research. sponsored by the European Commission and national research programmes. The Centre partici- Research and development are key to enabling the pates in EC co-funded projects involving Web IT sector to keep pace with the demands of the services for eLearning, eHealth, security and privacy and networking Innovation Centres in EMEA technologies, and is a member of the Integrated Projects and Networks of Excellence selected by the European Com- mission for the first call of the 6th Framework Programme. EMIC is developing strong relationships with the University of Aachen, and has academic partnerships with a number of other educational institutions.

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 25 Microsoft Business Solutions Centre, Vedbaek, to deliver more and better jobs and quality of life Denmark for Europe’s citizens. Although six million new Beyond pure and applied research, the Micro- jobs have been created since the launch of the soft campus at Vedbaek employs more than 700 Lisbon process, unemployment across many parts people and is the company’s largest development of Europe remains too high and EU governments facility outside of the main campus in Redmond recognise that more needs to be done. (US). It focuses on supply chain management One requirement is to speed up the flow of strategy, Microsoft Business Framework, and information across the EU regarding local or Project Green – building the next generation of regional job vacancies and, importantly, to use Microsoft Business Solutions. The Microsoft data mining to identify ways in which to create Business Solutions division works with Europe’s similar job opportunities elsewhere in the EU. developer community to produce solutions to We and our partners have developed solutions help foster growth in the region’s small and me- that can help. dium enterprise sector. In addition, Microsoft helps create sustainable European Product Development Centre (EPDC), jobs through our software development network Dublin, Ireland – today companies that have built their businesses The EPDC is charged with ensuring that Microsoft on sales or development on the Microsoft products are available in different versions across technology platform employ over 1.6 million Europe, reflecting local languages and culture and people in Europe. ensuring communities have access to IT in their own language. The EPDC employs linguistic Recent studies, including one by the Economist specialists to localise software into over 35 languages Intelligence Unit sponsored by Microsoft, and dialects spoken across the region – including confirms the direct link between ICT and regional languages such as Basque and Catalan. economic growth and job creation, once a minimum threshold of ICT development is Microsoft Technology Centres (MTCs), Munich, Germany and Reading, UK MTCs support local IT industries in their objectives to compete in the global market. They work side-by-side with customers’ architects and developers to rapidly find solutions to their technology challenges. To assist in this, the MTCs have formed alliances with industry leaders that provide comprehensive resources such as hardware, software and services to MTC customers from local IT industries which create a sound environment for development.

More and better jobs in the knowledge- based economy

One of the principal reasons for the Lisbon Agenda focus on growth and competitiveness is

26 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES reached. Importantly, ICT uptake and job At Microsoft, we have a long-standing creation rely upon a number of factors beyond the commitment to partnership with NGOs and availability of technology, particularly investment disability charities across Europe, providing in skills, innovation, competition and an enabling IT access and training for people with special business environment, ICT in the public sector, needs. We actively supported the 2003 European and invigorating R&D. Year of People with Disabilities through internal and external awareness-raising and outreach As part of this, Microsoft is participating in the to policy-makers to support and contribute to EU’s ProLearn network of excellence which aims policy and standards. As a follow up, we are inter alia to increase the transferability of training supporting the Business & Disability Network, so that trainees find it easier to apply their new which works to raise awareness of the business knowledge. case for disability, promotes disability inclusion initiatives, and encourages the exchange of ideas Enabling inclusion in the information amongst business, political actors and people society with disabilities.

The potential for everyone in society to Microsoft fully supports the EU follow-up contribute underpins the Lisbon vision. It is a Action Plan and will continue to play our part vision that Microsoft shares. As we have grown in helping increase the level of awareness to become a global industry leader, we recognise regarding the rights of people with disabilities that our responsibilities as a corporate citizen amongst businesses, governments, NGOs and and a responsible industry player have grown in individuals. parallel. We have extended our original business goal – a PC on every desk and in every home – to helping people and businesses everywhere realise their full potential through information technology through multiple platforms.

Accessibility plays a major role in enabling social inclusion. For more than fifteen years, we have been building features into our products to enable people with mobility or sensory impairments and other disabilities to access work, communities and information online. This means ensuring that the needs of people with disabilities are taken into account in the early stages of design and planning of our software. Initiatives range from increasing text sizes and simplifying interfaces and icons to speech recognition and support for specialised hardware.

Accessibility is also key to competitiveness, and assistive technologies can help bolster productivity by helping many disabled people to remain and re-integrate in the workforce.

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 27 Greater social cohesion and Digital and providing IT skills training and lifelong Inclusion learning opportunities for under-served groups and communities through our Unlimited Throughout Europe, there are disparities in Potential (UP) programme. productivity and employment and pockets of marginalisation in urban areas and in rural The goal of both programmes is to help eliminate communities. Bridging the technology gap can technological illiteracy and exclusion by help to re-dress the imbalance of different groups combining enhanced IT access with support for advancing at different speeds. ICT can provide teachers and schools, and for community centres people, communities, regions and countries through IT skills training, technology grants, with the tools to connect to opportunities across curriculum development, employee volunteering, Europe and realise their true potential. and software donations.

At Microsoft, we can mobilise resources and expertise to share in partnerships aimed at bringing IT to the communities that need it most. For the past twenty years we have been an active supporter of thousands of community programmes in Europe and around the world.

Our Digital Inclusion initiative brings together most of our community projects, focusing on classroom education and teacher training through our Partners in Learning programme;

Our programmes are global in scope, adapted to local needs. We recognise that to achieve the shared societal goals of education and lifelong learning, and economic growth and regeneration, partnerships to share and multiply resources are needed to ensure that the opportunities and benefits are available to all.

28 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES Sustainable economic growth people live and work in a more productive, open and inclusive society. Through the eEurope+ Converting innovation into sustained economic 2003 Action Plan, this process is extended to the success is one of the key goals of governments EU candidate countries. across the European Union. Towards this aim, many governments are looking to work in eEurope is focused on the areas of: partnership with industry to define and shape the • eGovernment, to bring citizens, business and optimal enabling framework for innovation in governments closer together, heightening the respect to both financial investment and human democratic process and enabling individual creativity. access to vital public services as and when needed. In order for society to benefit from the investment • eLearning, to enable people to benefit from in innovation, individuals, companies and new innovative approaches to education and universities need to know that intellectual training, producing a highly skilled generation endeavour will be rewarded. Intellectual Property of workers tuned into the modern economy protection is therefore important to promote the and enabling anyone, anywhere to continue growth and development of companies that invest learning. in innovation and the overall knowledge base. • eHealth, to provide access online health data and benefit from a range of improved patient Sharing knowledge skills and expertise is a vital care. prerequisite for sustained economic growth. In • eBusiness, to stimulate the economy and pro- this regard, Microsoft views collaboration equal vide a supportive environment for entrepre- in importance to competition in terms of fostering neurs and facilitate the provision of services sustained economic growth. Through our partner and products to consumers. model we work with thousands of companies to provide the building blocks – platforms, solutions The potential offered by information technology and services – they need to create and market will only be fully realised if people feel secure their own products and services. in the online environment, and if the services are widely available at low cost through fast broadband connections. Microsoft works with 200,000 local partners, including 25,000 independent software vendors, across Europe, Middle East and Africa. It is this ecosystem of resellers, service providers and independent software vendors that provides the basis of sustainable economic growth, driven by continued innovation in the ICT industry.

The key building blocks of eEurope

The eEurope 2005 Action Plan provides the means by which technology will be brought to the centre of people’s lives, acting as an enabler and helping

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 29 Also underpinning the achievement of horizontal (Schools Interoperability Framework). We are eEurope objectives is interoperability. Enabling also a board member of IMS Global Learning, are competing systems and devices to exchange on the Executive of the Brussels-based eLearning information is a challenge for governments in Industry Group, and are the main 2005 sponsor of their move to bring public services online. And in the European ARIADNE foundation. the internal market, the ability to easily access and exchange data and information across national eGovernment borders is increasingly essential for government, industry and the public alike. eGovernment lies at the heart of the drive to build a modern, inclusive Europe, delivering better, Broad public up-take of the services being more efficient public services and transforming developed and offered depends upon the ease the relationship between citizens, businesses and with which they can be accessed via multiple their governments. Member States are reporting platforms, from PCs and digital television sets to solid progress in the growth of online service mobile phones. delivery, Public Internet Access Points (PIAPs), broadband connections for administrations, and Industry has shown that it is willing to take the e-procurement. lead on interoperability, through the formation of the WS-I (Web Services Interoperability) Interactive public e-services organisation, and participation in standards bodies. At Microsoft, we are working with Enabling citizens and businesses to interact industry partners, with national governments and with public departments drives transparency with the EU to help ensure that the solutions we and efficiency; moving from an era of queues to develop meet the demands of eEurope and the requirements of the internal market.

The success of eEurope is dependent upon competing systems and devices being able to exchange information. Industry has long- recognised this need, and is steadily improving the interoperability of systems. This is still evolving, and there is a need to ensure that standards and other means to achieve interoperability are developed to keep pace with the rise of new distribution means.

Microsoft has developed an XML-based Information Bridge Framework to help administrations, businesses and citizens to quickly gain access to the information they need, across multiple data stores. And in the area of education, Microsoft is a major supporter of public-private initiatives to establish open standards for eLearning and lifelong learning, such as SIF

30 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES one in which personalised public services can be There is also increasing emphasis on the ability accessible 24 hours a day. for information to be rapidly exchanged across borders and between administrations. To this end, At Microsoft, we have been working with Microsoft is actively working with administrations the EU and with national, regional and local in their efforts to improve data exchange using governments across Europe to help them open standards-based XML solutions. Microsoft develop interactive public services and provide has collaborated with the EU’s Interchange access to the range of government departments of Data between Administrations (IDA) in through a single contact point. Developing user- evaluating open document formats. friendly services underpins the broad public up-take needed to make eGovernment a success.

Achieving Interoperability Interoperability is the key to enabling seamless access to the services of different departments and linked organisations.

Microsoft’s approach in all its eGovernment work across the region has been to supply a set of standard tools that can be customised to the needs of specific administrations, but which maintain their standards based on functionality and interfaces. This allows for ease of integration with other applications, and ease of use for governments and the public.

Cost-effective, flexible, standards based solutions will help administrations across Europe further integrate and help new Member States move as quickly as possible to a position of parity with existing members.

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 31 All government information systems need to While connectivity to ICT infrastructure is rising be secure, protecting vital data and ensuring in the education and training sector, there is still privacy. And they should also be ‘future-proof’, a need to ensure that eLearning opportunities are developed on the basis of solutions which allow equally available across geographic boundaries for the future addition of new services. and societal sectors.

ICT in the classroom Integrating the use of information technologies in education and training curricula means knowing how to make best use of the new technologies. For educational establishments with IT connectivity, teacher training, and re-training, is often a necessary step.

With many schools and educational establishments lacking the technical and financial resources to participate in the eLearning revolution, this means a focus on cost-effective ICT solutions.

Microsoft has a long-standing commitment to working with Education Ministries and learning establishments to help equip them for the future, including initiatives to: • connect educators through our Innovative Teachers portals enabling best practice and content sharing • develop students’ IT skills and delivering recognised certification through our IT Academy programme which allows educational establishments to offer IT industry-recognised qualifications • expand access to IT through the provision eLearning of software grants and PC refurbishment eLearning underpins the realisation of Europe’s programmes Lisbon goals to lay the foundations for long-term To facilitate the spread of eLearning across competitiveness and to equip everyone in society the EU, Microsoft has developed localisable to participate and contribute to a dynamic and curriculum resource management tools based on growing economy. open standards.

The virtual classroom precedes the virtual office, In 2003, our classroom support initiatives were familiarising students at an early age with the gathered under the banner of the Microsoft demands of the online workplace. And by enabling Partners in Learning (PiL) Programme. new, innovative approaches to education and training, eLearning encourages lifelong learning PiL is a five-year initiative with three key goals: to help people re-enter the workforce or start new • To empower schools to significantly raise the businesses by acquiring new skills. level of ICT literacy amongst their staff

32 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES • To support teachers and schools in developing an internal culture of innovation • To work with schools in preparing students for the digital workplace

At Microsoft, we are working with governments, educational establishments and NGOs to help ensure that IT skills training is developed broadly, and the benefits of the ICT revolution open to all.

Global in scope, but local in implementation, Partners in Learning works with national governments, educators and partners to ensure programme components adapt to local educational needs and challenges. Partners in Learning is already being implemented in partnership with many European Education Ministries, including those in Austria, Bulgaria, Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Italy, Lithuania, Norway, Poland, Portugal, Romania, Slovakia, Spain and Sweden.

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 33 ICT in the Community Microsoft’s Education Solutions Group to provide eLearning needs to look beyond the classroom a ‘connected learning community’ with a holistic, setting to lifelong learning, enabling people to sustainable and forward looking approach. re-skill and retrain throughout their working lives, upgrading their skills in a specific area or The focus of UP projects is as varied as the learning new skills to give them access to new needs of each community, and includes projects job opportunities, improving their employability to support IT learning for children, unemployed and overall quality of life. Lifelong learning calls youth, the homeless, people with disabilities, for partnership for ensuring IT access and skills women, entrepreneurs, elderly people, and training is brought to the broader community. refugees. Today in the Europe, Middle East and Africa (EMEA) region we are supporting over Microsoft Unlimited Potential (UP) is a global 87 projects in 45 countries, involving over 130 programme that focuses on improving lifelong partners and over 300 CTLCs. learning by providing technology skills to As part of our commitment to digital inclusion, people through community-based organisations. the Microsoft Authorised Refurbisher (MAR) Microsoft provides resources to launch or sustain programme was launched to facilitate access and IT skills training programmes, including training lower the environmental footprint across the value technology instructors and expanding course chain. MAR enables authorised PC refurbishers offerings to reach a broader base of community in 153 countries to re-install Microsoft operating members. Microsoft has also developed curricula systems, with only a very low administrative that emphasise real-world technology applications fee, into donated pre-used PCs destined for which will be available in multiple languages. schools, charities, non-profit organisations and The programme works collaboratively with community centres. MAR also responds to the growing need to facilitate the extension of IT products’ lifecycle, which is part of our wider environmental responsibility.

34 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES eHealth technology is reliant upon one fundamental tenet – patient trust. Patients need to have the Healthcare has always demanded the best reassurance that personal medical records can technology, and has pushed the boundaries only be accessed by the appropriate clinicians. in terms of service expectation. Information Microsoft is actively engaged in research to technology is now beginning to occupy a central further define Public Key Infrastructure (PKI) position in the treatment of patients, akin to needs in this area. that of clinician and treatment. It has moved beyond a tool for basic functionality, and now has the potential to integrate care from across the healthcare applications.

At Microsoft, we have been working in partnership with governments and healthcare providers from across Europe to define healthcare needs and to develop tailored healthcare solutions.

Efficiency, transparency and interoperability One of the key developments in healthcare computing has been the establishment of electronic patient records. These have the dual purpose of providing faster, more accurate information for clinicians as well as facilitating greater patient information. This means clinicians can make more informed diagnosis and patients can have a fuller picture of the treatment process.

Healthcare computing is often complicated by the multitude of non-compatible information systems, many of which are based on legacy mainframe systems. At Microsoft, we have been working with hospitals and healthcare organisations to help them migrate towards the optimised solution.

Whether in relation to patient records or in terms of basic infrastructure, hospitals and care professionals need to know that capacity will not be lost, and that it will actually be improved upon with the new, tailored solution.

Privacy and security The application of patient records and the paradigm shift towards the broader use of

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 35 Creating a dynamic eBusiness environment makes a net contribution of €7.50 for every €1 that is spent on Microsoft products in the Continued innovation in information technology European economy. Companies that have built is dependent upon partnership and understanding their businesses on sales or development of the between industry and government. The role Microsoft technology platform employ over 1.6 of both players should not be underestimated million people in Europe alone. – government has a crucial role as regulator, partner and customer of the IT industry.

Information technology is reliant upon new ideas and the rewards that stem from intellectual endeavour. This, in turn, requires a legislative framework that supports intellectual property rights.

Consumers and companies alike also need a legislative bearing from government in respect to the development of e-commerce. Support for e-commerce is sometimes taken for granted, but often legal frameworks fail to keep pace with rapidly evolving developments in technology.

Developing Small and Medium sized Enterprises across Europe Small and Medium-sized Enterprises (SMEs) play a major role in Europe’s business economy, accounting for approximately two-thirds of employment. For Microsoft, SMEs are a focal point for our activity in Europe, not only as a customer, but also as a partner.

Three things are required to stimulate a healthy IT ecosystem: • venture capital to nurture innovative new SMEs • IT skills training to help SMEs integrate technology effectively into their business operations • access to scalable and cost-effective IT solutions for SMEs.

Microsoft’s partner ecosystem, encompassing independent software vendors and developers,

36 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES Microsoft has a range of services to help small businesses set-up as Microsoft partners or to implement technology that can help them make their businesses more efficient and grow faster. • Microsoft Business Solutions offers business applications and services designed to help companies become more connected with customers, employees, partners and suppliers. More than 3,600 independent local partners in EMEA are able to customize, implement and support Microsoft Business Solutions applications, whilst providing a thorough understanding of the local business environment and knowledge of specific industries. • The Microsoft Small Business Centre website is a central hub of information that supports many small business markets in Europe and across the world. It provides tailored business information and advice on how to run a business more effectively, customer support options and information about Microsoft and third parties’ products and solutions. www.microsoft.com/smallbusiness

Broadband adoption

Access to broadband Internet connections is a pre- European Information Communication and requisite for the public up-take of online services Technologies Association (EICTA), Digital Video and content, and is the basis for a successful Broadcasting project (DVB), Third Generation information society. Europe as well as the US Partnership Project (3GPP), Open Mobile Alliance lag behind Korea and Japan, where broadband is (OMA) and the World Forum for Digital Audio far faster and cheaper thanks to focused national Broadcasting (WORLDDAB) are just some of the policy. bodies in which we actively participate.

Developing open standards Driving the development of alternative broadband Microsoft is actively working with European platforms standards bodies to develop the next generation Adoption of broadband cable and DSL is of systems that will get broadband to European increasing in Europe, but many other access citizens at an affordable cost, and will enable a channels need to be explored to make broadband variety of eGovernment, eHealth and eLearning truly ubiquitous. Microsoft is working in its services to be received on Digital Set Top Boxes MSR-Cambridge facility to consider new uses and Mobile Phones, as well as via the PC. for digital radio, digital TV and mobile networks. These networks are being modelled as alternative The European Telecommunications Standards vehicles for the distribution of ‘eContent’ to the Institute (ETSI), European Information Systems, citizens of Europe.

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 37 We are working with industry partners towards many countries regulated the legal recognition of the goal of making access to every service and electronic signatures and Public Key Infrastructure application possible from any device, anywhere, (PKI) based authentication schemes. anytime. The situation has evolved, with the EU adopting a more active security policy. A Directive For European citizens to have harmonising the legal recognition of electronic broadband connectivity that is signatures introduced the concept of Advanced affordable, providing access to quality and Qualified Electronic Signatures. In response, content, applications and multimedia Microsoft is producing a Tutorial which gives services for all regardless of location guidance to architects and developers on how or device, we need strong government to implement the European requirements into and private sector focus. the Windows Platform. Microsoft has also been active in supporting the development of Electronic Signals Standards Initiative (ESSI) standards in Security and Privacy this field. The great strength of the Internet is its ubiquity, but a weakness is its security. This is a reflection of the origins of the Internet. Because it was not originally conceived for business and private communication, its resilience did not extend to protecting users against computer viruses. Increasingly, however, businesses, governments, and citizens rely on information systems to raise productivity, deliver services, communicate, and access information and entertainment.

As a leader in the computing industry, Microsoft recognises that it has a key role to facilitate everybody to work, communicate and transact securely. To that end, Microsoft is focused on delivering improved security across all of its platforms and products. Security-related research has also been of However security alone is not enough. Users want increasing importance since the launch of the 5th privacy, which means the right to be left alone and Framework Programme. The European Microsoft the right to be in control of their personal data. Innovation Centre in Aachen (Germany) has participated in several projects, and Microsoft Microsoft’s security approach in the EU context also supports university research in security- The European Union has had an active Internet related areas such as threat modelling or the security policy since the mid 1990s. The initial development of security curricula. debate was around encryption and digital signatures, with many governments fearing The European Council has issued several policy the wide use of encryption could endanger the statements outlining the directions all Member effectiveness of lawful interception. As a result, States should take in computer security, and

38 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES the eEurope initiative has also been used to Protecting privacy push awareness raising, creation of computer Privacy is one of the key concerns of customers, emergency response systems, recognition of industry and government across Europe. In respect common criteria, and stimulation of electronic of issues of consent, access, security and enforce- signatures. The European Network and ment, data integrity and onward transfer, all parties Information Security Agency (ENISA) will serve concerned need to know that personal information as a centre of expertise for the EU institutions is being used appropriately. on matters related to network and information security. This not only requires the development of new technology, but it also necessitates working with Security is Microsoft’s number one priority. All industry and government to determine ongoing new products and service packs are now released standards development and implementation. following a rigorous security process to significantly reduce the number of vulnerabilities. Unsolicited email, or spam, is a privacy issue Microsoft has also been improving its software of international and cross-industry concern. As update process step by step and through vari- an ISP and provider of email, as a builder of ous means; for instance through a regular update enterprise-level email clients and as a commercial cycle. The company is also providing tools and marketer, Microsoft is fully aware of the effect guidance to ensure that users are informed. of spam on consumers and businesses worldwide and is dedicated to finding new and more efficient The proliferation of computer viruses and ways to combat its growth. worms in recent years has been the key reason why computer users do not trust information Since 2003, we have been working with Yahoo technology in the same way as other essential and AOL to develop platform-neutral technical services. Whether working directly with solutions to protect consumers and to eliminate the customers through our Protect your PC campaign, facility to create fraudulent email accounts in bulk. or with industry partners in the Virus Information Alliance, our focus is always building security Microsoft also works with national governments software and services. to support the implementation of EU rules on Electronic Communications Data Protection, We have made significant investments to develop which stipulate that commercial emails may Windows XP as a security enhancing product. only be sent to users who have given their prior Service Pack 2 for XP will be another step towards consent. We also collaborate with law enforcement our vision of making computing trustworthy. This agencies across Europe, and internationally, release is predominantly about security and will to identify and take action against persistent provide additional protection for users against spammers. malicious Web sites, dangerous email attachments and other common Internet based attacks. The protection of minors We are also working with law enforcement A particular concern is the potential for ‘spammers’ agencies on a global basis to deter hackers to target less sophisticated email users, such as from software sabotage. Microsoft’s Anti-Virus children. This is not the only threat to children Rewards Program offers rewards for information posed by the abuse of information technology, leading to the arrest and conviction of those and Microsoft supports EU policies to tackle responsible for unleashing viruses and worms. illegal and harmful content on the internet.

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 39 We have participated in the EU Safer Internet the broader Europe, which risk being left behind Action Plan and support the follow-up programme. in the competitive race. We are working with law enforcement authorities, governments and non-governmental organisations The private sector also has key activities, and, in at national and international level to help address this, the IT industry plays an important role. New the broader issues surrounding the protection of technologies have become increasingly central to minors online. the way in which we live, work and communicate. They increase productivity, competitiveness Consumer confidence and safety are prerequisites and deliver a new means of interaction between for growing and sustaining computer usage governments, citizens and businesses. across Europe. It is a complex issue, and one that Microsoft will continue to pursue in order The eEurope Action Plan lays out the areas in to alleviate citizens and customers concern and which the public and private sectors can and thereby provide a safe computing experience must work together to ensure the widespread for all. availability of essential content and services delivered on a fast and secure platform. Our involvement in the fight against But eEurope goes beyond this, acting as the spam goes beyond advancing technology stimulus through which to leverage the potential solutions. We have committed to working of ICT in building the new Europe, familiarizing alongside industry and government through people with the online environment and driving technology, industry self-regulation, new, modern working and learning practices. legislation and enforcement, as well as consumer education. Building on this stimulus and driving widespread up-take and acceptance of the new technologies Conclusion relies on collaboration between the public and private sectors. It is through this collaboration At the mid-way point of the Lisbon agenda, that Europe will build and sustain the momentum there is broad agreement that faster progress is to deliver the Lisbon 2010 goals. needed if Europe is to meet its goal of becoming, by 2010, ‘the most competitive and dynamic At Microsoft we are fully committed knowledge-based economy in the world, capable to playing our part in ensuring that the of sustainable economic growth, with more and real potential of ICT to act as a force for better jobs and greater social cohesion’. economic, social and politcal ends is fully realized. We are committed to continuing to Part of these activities lie with the public sector, act as IT partner and regulatory interlocutor in introducing and implementing the legislative with the EU and national governments and framework that encourages innovation and to working with partners across the broader creates more inclusive employment practices. European region to help ensure that this Public sector activities also include providing potential is open to all. incentives for the regions and areas throughout

40 KNOWLEDGE ECONOMY: KEY NOTE SPEECHES Further information

To find out more about Microsoft in the EMEA region please visit –www.microsoft.com/emea

You might find these sites of specific interest: Microsoft EMEA Citizenship – www.microsoft.com/emea/citizenship Microsoft EMEA in Education – www.microsoft.com/emea/education Microsoft EMEA In The Community – www.microsoft.com/emea/inthecommunity

For global and local case studies, visit – www.microsoft.com/resources/casestudies

KNOWLEDGE ECONOMY: KEY NOTE SPEECHES 41 Report on the session: R&D and Innovation

Chair: Fred Gault, Statistics Canada

There were six presentations dealing with the The discussant and the discussion gave rise to production of R&D and innovation statistics and suggestions for better statistics. Quality was a their application. Applications included regional recurring theme and this included timeliness and comparisons of both countries and NUTS 2 relevance. Linking the statistics on R&D and and 3 regions. We had public and private sector innovation to business statistics was proposed as providers of statistics, which demonstrated that a way of enhancing the analytical value of the there was a place for both. data sets. However, this also made the point that for the data sets to be of value, researchers have One of our presenters came from an IBM lab and to be able to gain access to them. Micro data he made some of the same points as Patrick de analysis is better than macro data analysis. Smedt, from Microsoft Europe, the earlier keynote speaker, that business is complex, involves many There were other observations on statistics and players that share the risk of innovation, and the their use. Innovation is managed locally, even if players have to be managed coherently. He also it involves players globally. This makes regional noted that innovation is more than just R&D. This statistics important and size of the firm, in this is a point that I would emphasize, that the activity context, is a key analytical variable. So, statistics of innovation, of getting products to the market must be able to support understanding, policy in new ways, or presenting new products to the development, and monitoring, in regions. But, market, is indeed more than R&D. It includes, for the R&D and innovation are also global and the example, training, knowledge acquisition, and statistics have to be able to measure the linkages capital investment, all brought together to create in the system and that is something that needs value from knowledge. That knowledge does not work. There are policies, statistics, and papers on have to be created by in-house R&D. In fact, in activities, such as R&D and innovation, but not many firms it is not. so many on cooperation, intangible knowledge flows, and skilled labour flows. This does not This raised some questions about policies around means that these statistics do not exist, but we the Lisbon process, mentioned also by Anthony are not seeing high level policy statements on Arundel in his keynote speech. While we making these linkages work. It is not easy to talk understand that innovation is more than R&D, about these things simply. the policies appear to equate the two. There was also the question of policy coherence for R&D, The globalization of R&D requires a better innovation, commercialization, entrepreneurship, understanding of not just the production of and skill development policies. Separate policies R&D data for firms in countries, but also of the for these areas could all act in the same direction, production of R&D by foreign controlled firms (A or not. result of incoming foreign direct investment, FDI – in) and domestic controlled firms, of purchases The Session saw the first results presented from and sales of R&D services (part of technological the fourth round of the Community Innovation balance of payments), and of the R&D performed Survey (CIS.4), which suggested that there were by affiliate firms abroad (A result of FDI out). The many policy relevant findings to follow. R&D statistics collected will also be influenced by

42 KNOWLEDGE ECONOMY: R&D AND INNOVATION the decision to capitalize R&D in the System of for OECD and the joint OECD/Eurostat manuals. National Accounts, the importance of which is just This topic is also being discussed in the African beginning to dawn upon official statisticians. Union by Ministers of S&T.

Our discussant raised the need, in Europe, for a Other suggestions included collecting R&D European Frascati and, I assume, Oslo Manual. data at project level rather than at firm level, the This would reflect the implementation of the production of score boards using official statistics, laws governing statistics on each activity and more attention to knowledge management in would lead to better comparisons and use of the firms, and finally, a suggestion that R&D data be statistics within Europe. A counter view would collected only from firms that do innovation. An be to keep these manuals, which are deliberately interesting thought? high-level guidelines and principles, and build a community of practice through the Eurostat There were criticisms and disagreements, but it meetings on R&D and innovation. This working was a very rich discussion that will provide our knowledge could be codified and, when it colleagues at Eurostat with many things to think suggested a change to the manuals, this could be about as the R&D and innovation programmes addressed through the on-going revision process develop.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 43 Knowledge Economy – Challenges for Measurement

David White Director, Enterprise and Industry Directorate General European Commission

Innovation is crucial for Europe and Lisbon based on outdated data can be of interest objectives. However disappointing performances for them. are observed in many Member States. The patterns Indicators we use are based on these statistics and followed by each Member State are different and selected for the information they provide, which in order to propose the best policies, we need to has to be both rich and parsimonious. In order identify the innovation challenges. We need to to provide robust support to policy analysis, they monitor progress. must provide a manageable picture and be based on a solid methodology. To do so, innovation statistics are needed. The Commission and Member States efficiency will With this objective, the Commission has partly rely on them. These statistics have to developed the European Innovation Scoreboard respect the 3-R rule. (EIS) which is now a well known instrument, 5 • They have to be Relevant, which is year-age, which aims at measuring the innovation particularly the case for the CIS and performance of EU Member States, Candidate implies that questionnaires have to be Countries, Associated Countries plus US and prepared in collaboration with the final Japan. users, DG ENTR for example. The EIS indicators are extracted from several • They have to be Reliable, meaning sources, Eurostat, OECD, OHIM, EPO and that same robust methodology must be specific innovation surveys, the Community applied in all MS. When this is not the Innovation Survey and the Innobarometer. The case, data can become useless. A recent EIS 2005 will come with 26 indicators, covering example can be extracted from the last the main dimensions of innovation. It is necessary innovation survey carried out in 2005, to have a large number of indicators to fully where MS used different methodologies embrace the innovation phenomenon; however in some cases. Portugal for instance this raises other questions such as: carried out a survey with a sample where companies were taken within • How to make sense of so much data? the group of firms employing 5 to 500 • How to build a coherent picture of employees, whereas other countries innovation performance? covered the category 10-500 employees. • How to draw valid conclusions? The consequence is that Portugal results are not comparable to other EU results. • How to stimulate policy discussion (and not statistical disputes)? • And finally statistics have to be Received on time. It is extremely difficult to The 26 indicators of the EIS are used to rank the convince policy makers that analysis Member States depending on their innovation

44 KNOWLEDGE ECONOMY: R&D AND INNOVATION performance. This results in a global picture, Iceland, Estonia, Lithuania, Cyprus and Norway where we observe that Nordic countries plus are examples of countries showing much lower Germany are the EU innovation leaders. The new performance on outputs than on inputs. Member States are either embarked in a catching up process or are in a more preoccupying position One possible explanation for these observed where they are loosing ground. And finally, most differences might be the receptiveness of a of the old Member States are in a larger group of country’s population to new products and average performing countries. services, as it has been measured by the Innobarometer 2005. Among the 10 European You certainly observe that this classification is countries which have the highest share of based on a composite indicator. When combining population attracted by innovative products these various elements we can have a “at glance” or services, 9 have an above average output/ view of relative innovation performance of all input rate. Conversely, 7 countries among the EU countries, and compare them with our main 10 where population readiness for innovation partners. is the lowest are below average output/input rate. This first analysis should lead to further I know that composite indicators are a sensitive considerations on the importance of developing issue, especially when discussed in a meeting innovation readiness in the MS. like this one, with many statisticians. However, we still believe that these composite indicators Not all countries perform on the same level in can be useful instruments when used properly each of the innovation dimensions. And some and when they are based on robust methodology. countries may even prove to be especially weak in one or several dimensions. The EIS brings In this respect, the EIS comes with a strong first evidences that an even performance on all assessment of the composite indicator. We have dimensions (5 EIS dimensions + Governance + carried out 300 simulations to recalculate the Demand, the two last dimensions are extracted Summary Innovation Index, each time proposing from a thematic report of the EIS) fosters a different weighting scheme. The results are that innovative performance, countries which show the relative position of countries with regard to a below average performance on one of these their innovation performance is not sensitive to dimensions as compared to the country’s overall the weighting scheme. This illustrates that the performance, might be in danger of hampered composite indicator we use is a robust one. future innovative performance.

Thanks to such a robust methodology, it is possible The EIS identifies 5 main dimensions of to develop new analyses. We have introduced innovation performance. Each category is based in 2005 an input/output approach where it is on 5 or 6 component indicators. Innovation possible to analyse how countries transform drivers measure the structural conditions required their innovation assets (inputs = education, for innovation potential, Knowledge creation R&D investment, cooperation, firm investments measures the investments in R&D activities, in innovation, etc) into innovation results Innovation & entrepreneurship measures the (outputs = turnover from new products, efforts towards innovation at the firm level, employment in high tech, IPR, etc). Switzerland, Application measures the performance expressed Germany, Luxembourg, Ireland and Malta are in terms of labour and business activities and their examples of countries showing much better value added in innovative sectors, and Intellectual performance on outputs, therefore successfully property measures the achieved results in terms transforming their assets into innovation success. of successful know-how.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 45 These 5 dimensions allow for a rapid high in Hungary and highly directed to high tech understanding of the innovation profile of countries. sector. However, Hungarian firms are much less The example of Germany is characteristic. innovative than EU firms. This may then raise the Indeed, Germany is among the leading countries questions: “How can we increase the spill-over of and it can be seen that Germany performs well is innovative management techniques or innovation 4 dimensions of innovation as measured by the applications from the high-tech, foreign-lead EIS. However, Germany proves to be only an companies to the rest of the economy?” average performer with regard to the innovation driver category, which characterises a relative We also conduct innovation statistical analysis per weak position in the basic long term assets for sector. The EIS comes every year with a sectoral innovation (education, science and engineering thematic paper, the Sectoral Innovation Scoreboard. graduates, lifelong learning). This may raise The sectoral scoreboard is based on the CIS data and questions about the capacity of Germany to explores innovation performance in 25 sectors for maintain long term strong performance. the edition 2005. The results are not dramatically surprising, with Optical and Chemistry being more However, we should not put too much emphasis innovative that textile or mining. on the composite indicators, event if they attract a lot of attention from policy makers or However, it is important to have this kind of journalists. The EIS is mainly instrumental in data to understand the relative performance of the identification of innovation challenges in Member States in sectors. Innovation can follow the Member States. Therefore the selection of different patterns from one sector to another and component indicators is critical. In this respect refined conclusions can only be drawn if statistics the EIS 2005 comes with new indicators which give a more detailed insight. capture new dimensions of the innovation performance (public-private cooperation, To develop further analysis in sectors, the sectoral marketing innovation, foreign trade, etc). These watch initiative will enter into force in 2006. indicators allow for in-depth analyses on different Experts will explore the barriers to innovation in aspects of the innovation performance. several key sectors in Europe, and will directly rely on statistics to complete and orientate their In this respect, the EIS has been turned into a strong policy analyses. policy instrument. The EIS is one of the bases of the innovation policy review carried out every The EIS is the instrument we use to rank year in each Member State, within the Trendchart performance of countries, identify key challenges initiative. Key challenges are identified thanks to and open discussions. It is definitively one of our the EIS, which allows focusing our attention on the key policy instruments in the field of innovation innovation relevant issues for each country. This policy. instrument is therefore of direct use for working together with Member States to identify priorities All these analyses are only possible because data within the revamp of the Lisbon Strategy. exist, and we strongly support the development of innovation statistics. Our sources are mainly Example of Hungary is in this sense very Eurostat structural indicators and the Community interesting. EIS component indicators show that Innovation Survey (CIS). The CIS is the only high-tech sector is dynamic in this country. This source of statistics specifically dedicated to can be crossed with other information showing innovation at European level. This source of that the Foreign Direct Investments are very information is absolutely key for us. Therefore

46 KNOWLEDGE ECONOMY: R&D AND INNOVATION we are very satisfied that this survey is now Policy decisions can only be as good as our compulsory with a two-year period. We will pay sources of information. This implies that great attention to the 2006 edition. innovation surveys are carried on a regular basis, with harmonised methodologies for all member We, in Europe, are in the lead with regard to robust States, and with a robust quality check. statistics on innovation, thanks to the Community Innovation Survey. This allows to also having The EIS 2005 is now available on the lead in innovation policy analysis. We must www.trendchart.org. pursue in this direction.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 47 “S&T indicators: importance for research policy-making and areas for improvement”

Xabier Goenaga European Commission, DG Research

Executive Summary Member States have generally presented a well- founded analysis of the strengths and weaknesses 1) Policy context of their R&D systems and put forward a variety of measures to address them. Overall the NRPs To achieve sustainable global competitiveness, reflect a greater awareness of the need to have a the EU has no choice but to become a vibrant coherent policy mix to support R&D. However, knowledge economy and since 2000 this has a stronger commitment from those Member been the foundation of the Lisbon strategy. In this States that have set no R&D spending target context, in 2002, the Barcelona European Council, for 2010 combined with a determined emphasis recognizing the central role of R&D investment on implementation and mutual learning by all in the production and use of new knowledge, set Member States would lead to a quantum leap in the goals of raising overall research investment R&D. The Commission believes that there is a in the EU from 1.9% of GDP to 3% by 2010 and real opportunity for a break-through in this area. of increasing the private funding proportion of R&D from 55% to two-thirds. The Commission is also pursuing the implementation of the Community Lisbon But five years later, the results -for the Lisbon Program and has adopted in October 2005 a strategy in general as well as for the Barcelona Communication which for the first time integrates target in particular- are clearly not satisfactory. To in a common approach Research and Innovation remedy this situation and revive the commitment Policies. of the Member States, the Commission has taken the initiative to propose a new start for the Lisbon strategy through the establishment of a new kind 2) R&D Key figures of partnership with the Member States. This is the “Lisbon partnership for growth and jobs”, which Statistical indicators play a key role in the Lisbon was endorsed by the European Council this year. partnership in general and in the area of Research policy in particular. The “Key Figures 2005 for With this new start, the Lisbon strategy will be science, technology and innovation”, published more focused and a simpler cycle of governance by DG Research in September 2005 (http://www. is established, aiming to ensure that the policy cordis.lu/indicators/publications.htm), present areas selected for action receive the necessary the main quantitative information we use to assess political attention. Knowledge and innovation the situation of the EU as a whole as well as that for growth was singled out as one of three main of each Member State. This is a very valuable tool areas for action by both the Commission and the for benchmarking the performances of the EU and Member States. its Member States, monitoring the progress (or lack of progress) towards the Barcelona objective In their National Reform Programs (NRP) and identifying the strengths and weaknesses of submitted to the Commission in October 2005, the EU and the Member States. These analyses

48 KNOWLEDGE ECONOMY: R&D AND INNOVATION underpin our work with the Member States in convergence. Countries like Poland, Slovakia, the framework of the so-called “Open Method of Greece, the Netherlands had decreasing R&D- Coordination”. Taking into account the fact that intensities between 2000 and 2003. many of the policy tools available for progress towards the Barcelona goal are at national level, Another worrying conclusion of the Key Figures the identification of best practices in the Member is that Europe is becoming a less attractive place States and the development of a mutual learning to carry out research. Between 1997 and 2002, process between them is crucial. R&D expenditure by EU companies in the US increased much faster than R&D expenditure by The 2005 Key Figures show that EU R&D intensity US firms in the EU (54% compared to 38%). The growth is close to stagnation since 2000. It only net imbalance in favour of the US increased five- grew by an annual average rate of 0.1% between fold between 1997 and 2002, from about €300m 2000 and 2004 and even decreased in 2003 and in 1997 to almost €2b in 2002. Additionally, US 2004. Europe devotes a much lower share of its investment has been growing at a much greater wealth to R&D than the US and Japan (1.93% of rate in areas outside the EU, for instance at 25% GDP in the EU in 2003, as compared to 2.59% in per year in China compared to about 8% per year the US and 3.15% in Japan). While China has lower in the EU. R&D intensity (1.31% of GDP in 2003) it grew at about 10% per year between 1997 and 2002. If Another edition of the Key Figures for science, these trends in the EU and China continue, China technology and innovation will be produced in will be spending the same percentage of GDP on 2006. research as the EU in 2010. 3) S&T indicators: areas for improvement One of the reasons for this has been a slowdown in business funding of R&D. In 2002, In light of the policy context described above and business funding grew at a slower rate than the data currently available, and without pretend- GDP, though this was compensated for by a ing to be exhaustive, some key areas for improve- slightly higher growth of government funding, ments of the S&T indicators can be identified. as well as growth in R&D financed from abroad. In 2003, business enterprises financed 54.3% of • “Input-side”: new, better R&D expenditures domestic R&D expenditure in the EU, compared statistics: to 63.1% in the US and 74.5% in Japan, and this - Timing of data availability: The most recent share is decreasing. If the trend is not reversed, figures just published in the Key Figures date not only will the EU miss the overall target of back to 2003. For the first time, estimated two-thirds of R&D expenditure financed by the figures for 2004 were issued at the end private sector in 2010, but the situation will have of 2005. This is not sufficient to monitor worsened. recent policy developments. Current work coordinated by Eurostat may allow us to In contrast with the overall situation at the EU get next year data at T+3 months. level, some Member States have been doing well, - With regard to Business R&D expenditure, notably those which are already R&D intensive we need to develop the analysis by countries. Sweden’s R&D expenditure grew by industrial sector. Such analysis requires 11% a year between 2000 and 2003, Denmark’s better sectoral data. by 6% a year. Many of the new Member States - Globalisation of R&D: This is currently an such as Hungary are also catching up rapidly with important phenomenon, with a lot of policy the European average. But there is no general implications. But more data are needed to

KNOWLEDGE ECONOMY: R&D AND INNOVATION 49 correctly assess it, for instance data on the developed. Eurostat and DG Research are R&D performed by European companies collaborating to produce them as from 2007. in the emerging markets. - With regard to Public Research Organisations, including the Higher • Output indicators: Key Figures also includes Education Sector, an interesting area to some output indicators such as publications be further developed is the “Positioning and patents. However, improvements and indicators” by institutions (as studied in better exploitation of this kind of data are the framework of the PRIME network necessary. DG Research is financing several of excellence funded by the European projects in this area, for instance on citations Commission (ENIP project)). in patents, co-publications, etc. • Indicators directly linked to policy tools / • From national aggregates to more public research governance / framework disaggregated data : conditions for R&D, such as taxation - In order to better assess regional innovation and funding methods of Public Research systems, “Regional Key Figures” need to be Organisations.

50 KNOWLEDGE ECONOMY: R&D AND INNOVATION The EU R&D Statistics Progress made and the way forward

August Götzfried Eurostat, Unit F-4

Executive summary Furthermore, the main paper will submit points regarding the output of R&D statistics. These At European level, R&D statistics are one of the statistics also producing main indicators amongst cornerstones of Statistics on Science, Technology the EU Structural Indicators contribute to and Innovation. They provide basic information measuring the progress of the Lisbon/Barcelona for the follow-up of the Lisbon and Barcelona process. Based on extra data collected, the output European Council conclusions of 2000 and 2002 programme (i.e. the tables released on New Cronos and for the subsequent, newly established strategy and the publications) was also broadened. on growth and employment. R&D statistics are becoming increasingly internationally In addition to the international harmonisation harmonised, also with regard to the methodology work which has been undertaken, further progress used (which is the OECD Frascati Manual). needs to follow. More user needs and new challenges on those statistics are forthcoming, At international level, R&D statistics have e.g. with regard to better measurement of made considerable progress in recent years. the internationalisation of R&D or more and An internationally harmonised R&D statistics better data on European regions. An additional questionnaire was established which will be used major challenge will come from the expected by Eurostat and the OECD, and possibly other capitalisation of R&D expenditure in National institutes. This harmonised questionnaire has a Accounts which will create new requirements common module for Eurostat and the OECD, but on the amount of data collected and on the data also two specific modules for responding to the quality itself. This requirement coincides in needs of both organisations. some way with the better measurement of the internationalisation of R&D. Based on this internationally harmonised R&D statistics questionnaire, the production process Finally, timeliness of R&D statistics was always of R&D statistics will also be strengthened. A of the utmost importance for users. This means regular data collection from countries will take that the regular data produced needs more and place twice or even three times per year, with more to be complemented by now-casts and the expectation of producing an output which forecasts. will be of better quality (in particular with regard to completeness and timeliness). In the me- With regard to these new requirements, the dium term, the measuring of the data quality of international organizations involved in R&D R&D statistics itself will also be put in place. statistics will have to progress further with the This implies the drawing up of national and a development and production of this data, based European data quality report, including data on the considerable progress already achieved in quality indicators. the recent years.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 51 The EU R&D Statistics (with the Commission Regulation No 753/2004) which means that Member States are obliged to Progress made and the way forward deliver the data to Eurostat. As international comparability of data beyond the 1. Introduction EU becomes increasingly important and as the resource input for the data production at national Statistics on Science, Technology and Innovation and international level needs to be reduced, are mainly referring to the Lisbon and Barcelona OECD and Eurostat (the two main international European Council conclusions, emphasising organisations active in R&D statistics) agreed to the needs for boosting the overall R&D and work out an internationally harmonised R&D innovation efforts in the EU. With reference questionnaire to be used for data collection from mainly to R&D, a series of subsequent Commission countries by the OECD and Eurostat (and maybe Communications were released: “More research even beyond). The two organisations share a for Europe - towards 3 % of GDP”, “Investing in mutual interest in maintaining a high quality research, an action plan for Europe”, “Women and of processing and dissemination of statistical Science – Mobilising women to enrich European data, making the data widely available, while research” or “Researchers in the European Research optimising the use of limited resources and area: one profession, multiple careers”. minimising the burden on national respondents.

R&D statistics - collected and disseminated by Taking into account the existing OECD and Eurostat since many years - try to answer these ZEurostat questionnaires and the contents of policy questions in producing harmonised R&D the Commission Regulation No 753/2004 on statistics based on the OECD FRASCATI Manual statistics on science and technology, this process (version 2002) which is the internationally was launched with the signing of a Protocol for harmonised methodology for these statistics. Co-operation between Eurostat and the OECD on R&D Statistics. Considerable progress on the production and dissemination of R&D statistics has been Eurostat and the OECD then developed a common achieved in the recent years towards more international core questionnaire. The common international harmonisation, a more stringent core questionnaire which was finally designed is production process and towards more and better the result of a thorough analysis of the existing dissemination. New challenges to these statistics questionnaires and the data needs of the users of come however with user requests for better both organisations. measurement of the internationalisation and the regional R&D as well as with the capitalisation The core questionnaire is however accompanied of R&D in National Accounts. by specific modules for the OECD and Eurostat, containing supplementary tables based on 2. Towards an internationally harmonised additional needs of each organisation. The R&D data collection Eurostat module, for example, covers mainly regional data which is not one of the main interests In the area of R&D statistics, as in most other of the OECD at this stage. statistical areas, Eurostat sets the statistical standards and, based on these standards, – The additional modules can and will also be used organises the regular data collection from more for testing new tables, variables or breakdowns than 30 countries concerned. In the meantime, before those are eventually incorporated into the EU R&D statistics also have their own legal base core questionnaire.

52 KNOWLEDGE ECONOMY: R&D AND INNOVATION Most countries welcomed the harmonisation return: end of October (T+10 months) with efforts of creating the common international R&D data dissemination on the web latest by the questionnaire used by Eurostat and the OECD, end of November (T + 11 months) as it also decreases the national resource input considerably. After a transitional and evaluation • Collection of the full set of R&D data in asking phase, double data sending by countries to the two for an update since the summer data collection international organisations should no longer take – deadline for return: end of December (T+12 place as only one organisation will be responsible months) with data dissemination on the web for each of the countries. This organisation will latest by the end of February (T + 14 months) then receive the data set, treat it and forward it to the other international organisation. This time table for data collection and data production is thoroughly strengthened in This approach however requires that the data comparison to the process which previously treatment methods (e.g. on estimations, rounding, existed. The data collection and production etc.) are also aligned between both organisations. rhythm of the OECD has been aligned to this Otherwise, the quality and comparability of the rhythm. The output produced under this schedule finally produced national data and of the EU should satisfy the user needs (in particular for aggregates would suffer. Further work will also very timely data) much better than before. have to be done on the alignment of the metadata related to R&D statistics. This metadata will also In addition to this regular data production rhythm, comprise the collection and dissemination of the Eurostat works on a now-casting method which national data production methods. would – after agreement with Member States – allow the production and also the possible 3. A more stringent production process release of a number of main variables even 3 with higher frequency from 2005/2006 months after the end of the reference year (T + onwards 3 months). The estimations will have to be done on the base of the R D statistics available and on Based on the legal framework and based on the the base of other short term indicators to be taken OECD/Eurostat protocol mentioned above, the into consideration in the data model. production process of R&D statistics also had to be strengthened. From 2006 onwards, Eurostat is also planning to deepen the work on data quality on R&D Starting in autumn 2005, the regular data statistics with the main aim of producing regular collection and data production rhythm was quality reports and data quality indicators on defined as follows: R&D statistics. The results of these efforts should enable Eurostat on the one hand, to synthesise • Collection of the full set of R&D data – the national reports to an overall quality report deadline for return: end of June (T+18 months on R&D statistics that also should contain after the end of the reference year) with data recommendations on how to improve R&D dissemination on the web latest by the end of statistics. These recommendations will possibly September (T + 21 months) refer to issues such as data comparability and data availability, the improvement of the Commission • Collection of the provisional data for Regulation No. 753/2004, the amendment of the main totals (i.e. R&D personnel and the combined OECD/Eurostat R&D statistics researchers in FTE and R&D expenditure, questionnaire and the measurement of data all by sectors of performance) – deadline for quality as such.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 53 4. The output of R & D statistics: What data • Gross domestic expenditure on R&D and indicators are produced? (GERD) by source of funds - industry • Gross domestic expenditure on 4.1 The Structural Indicator R&D (GERD) by source of funds - government In the Lisbon European Council in March 2000 the • Gross domestic expenditure on R&D European Union set a strategic goal for the next (GERD) by source of funds - abroad decade “of becoming the most competitive and dynamic knowledge-based economy in the world capable of sustainable economic growth with The first indicator is themain flagship indicator more and better jobs and greater social cohesion”. on R&D used for the measurement of the so- The Council also invited the Commission to called Barcelona target which says that the draw up an annual synthesis report on the basis R&D expenditure should reach 3 % of the of Structural Indicators, which provide an GDP at national level and for the European instrument for an objective assessment of the Union as a whole. The statistical results based progress made towards the Lisbon objectives. on the reference year 2003 however show that Europe still considerably lags behind the main The EU R&D statistics contribute the following competitors in the Triad US and Japan. Even indicators to the Structural Indicators: China is catching up considerably and is expected • Gross domestic expenditure on R&D to reach EU level in a couple of years. - See also (GERD) the following Graph 1.

54 KNOWLEDGE ECONOMY: R&D AND INNOVATION Graph 1. R&D intensity (R&D expenditure as a percentage of GDP) in the EU-25, China, Japan and the US in 2003

��

�� Notes:Source: Eurostat, except CN, JP and US: OECD. BE, AT and SI: estimates. IS: forecast

All other Structural Indicators related to R&D The following Table 1 shows the results for the are further breakdowns of the R&D expenditure Business Enterprise (industry) sector with most of by source of funds for the main performing the R&D funding of this institutional sector coming institutional sectors covered. These indicators from the same sector again. Dependency of foreign illustrate financial flows related to R&D within a funds used for nationally performed business R&D country, but also cross border. is particularly high in Belgium, France, Cyprus, Latvia, Hungary, Austria and the United Kingdom.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 55 4.2. the full set of R&D data and indicators disseminated

Table 1. R&D expenditure in the Business enterprise sector by sectors of funding (in percentage) in 2003

GEO BES GOV HES+PNP ABROAD BE 83.8 p 5.9 p - 10.3 p CZ 81.0 12.0 1.6 5.5 DK 86.9 2.3 - 10.8 DE 91.3 e 6.1 e 0.2 e 2.4 e EE 87.0 5.6 0.1 7.4 EL 90.5 1.2 - 8.3 ES 83.4 11.1 0.2 5.2 FR 78.4 11.1 0.1 10.4 IE 87.0 f 3.0 f 0.6 f 9.4 f IT 77.4 12.2 0.1 10.3 CY 87.9 1.8 - 10.3 LV 64.2 16.0 - 19.8 LT 54.2 9.6 - 36.2 LU 89.2 p 2.5 p - 8.3 p HU 71.0 6.4 0.3 22.4 MT 75.0 16.7 - 8.3 NL 81.3 3.8 - 14.9 AT 64.5 5.6 - 29.9 PL 83.0 15.2 0.3 1.5 PT 89.2 5.3 - 5.5 SI 93.2 e 4.9 e 0.1 e 1.8 e SK 75.3 22.1 0.5 2.1 FI 95.8 3.3 0.1 0.8 SE 85.9 5.9 0.2 8.1 UK 63.1 10.9 - 26.0 NO 80.7 10.4 - 8.9 CH 91.4 2.3 0.5 5.8 RU 38.2 51.5 0.2 10.0 BG 98.5 0.2 - 1.4 HR 95.7 1.6 - 2.7 RO 67.1 28.2 0.3 4.4 TR 94.3 2.9 1.1 1.6 Notes: Source: Eurostat

Reference years: IE: 2004; BE, IT, MT, AT, TR and CH: 2002; EL: 2001 BES: Business enterprise sector, GOV: Government sector, HES: Higher education sector, PNP: Private non profit sector

56 KNOWLEDGE ECONOMY: R&D AND INNOVATION Beyond this flagship data used for Structural 4.3. More data and indicators to be disseminated Indicators, much more R&D data and indicators are disseminated on the Eurostat webpage. Based on the existing data collection on R&D Annex 1 to this document shows the variables statistics, Eurostat will however undertake efforts and breakdowns used. to increase the amount of data disseminated and also the data quality (in particular with regard The data disseminated is presented in the to data availability). Particular attention will be following structure: given to

• R&D expenditure, national and • The increase of data availability of the regional data related to R&D personnel broken • R&D personnel, national and down by sex; regional • The data related to R&D personnel • Government budget appropriations broken down by citizenship where or outlays on R&D the data availability has also to be increased; For each of these headings a number of multidimensional tables are on the Eurostat webpage • The data related to the R&D expenditure showing the main R&D variables in various in the Business Enterprise sector breakdowns (also including the Structural broken down by size class and economic Indicators mentioned above). activity.

If a smaller number of countries submitted the All data and indicators as well as respective data or breakdowns only or if the publications in electronic format related to data quality of certain data was judged to be R&D statistics can be found on the Eurostat insufficient, Eurostat did not disseminate this webpage under the heading ‘Science and data on the web. But in general, the difference technology’. Since 2004 all information is between the data collected from countries and disseminated free of charge on the Eurostat the data disseminated on the Eurostat webpage is webpage: rather small. http://epp.eurostat.cec.eu.int/portal/ 24 multi-dimensional tables on R&D statistics page?_pageid=0,1136250,0_45572552&_ (including Gbaord) are disseminated in total dad=portal&_schema=PORTAL which is considerably higher compared to some years back. This also reflects the increase in data 5. New challenges for R&D statistics collection based on European legislation and on increasing international harmonisation. Looking to the years to come, there are several new challenges on the EU R&D statistics Beyond the data itself, regular publications are which will have impact on their structure and issued. These are at least 3 Statistics in Focus contents: publications on R&D statistics (including Gbaord) released in the fourth quarter of each 5.1. More data on regions year. Furthermore, R&D statistics are used for one or several chapters in the annual Pocketbook and R&D data broken down by regions currently Panorama publication on Statistics on Science, only exists for the total R&D expenditure and Technology and Innovation. R&D personnel data for all the four sectors

KNOWLEDGE ECONOMY: R&D AND INNOVATION 57 of performance (Business Enterprise Sector, behalf of the funder. Government sector, Higher Education sector and • Type of R&D carried out by affiliates the sector of private non profit organisations). and target market.

Also linked to the list of R&D Structural Based on these distinctions and concepts, Indicators mentioned above, Eurostat is currently the handbook proposes reference indicators, investigating the possibility of compiling data on supplementary indicators and experimental regional R&D expenditure by source of funds. indicators for better measurement of the The feasibility and quality of the production internationalisation of R&D. The list of the R&D of regional R&D expenditure broken down by indicators proposed by the handbook is added in source of funds is assessed in order to proceed to annex 2 to this document. regular production if possible. In general, the user In 2005, the OECD sent a questionnaire to needs for R&D data broken down by the various countries in which the data availability of financing sources mentioned above (i.e. the a number of R&D indicators related to the institutional sectors and abroad) has increased. measurement of the internationalisation of R&D was asked. The availability of the following Beside this concrete and already ongoing activity, indicators was asked: further user requests for more regionalised R&D data exist. Together with the users, Eurostat • R&D receipts from abroad: aggregates will have to investigate in the years to come if and and breakdowns how more R&D data broken down by regions can • R&D extramural expenditure: payments be produced in sufficient data quality. to abroad, aggregates and breakdowns • Nationality of the owner of the R&D 5.2. Better measurement of the performing enterprise, with some more internationalisation of R&D details on the ownership concepts • R&D performing affiliates under foreign Reference is made in this context to the work control done by the OECD on ‘Handbook on Economic • R&D performing affiliates abroad Globalisation Indicators’ which also contains a • Measuring of R&D expenditures by chapter on R&D and on the OECD Task Force Multi-National companies (where Eurostat is fully involved) led by the US Based on this wish list of R&D indicators for dealing further with this issue. better measurement of the internationalisation of R&D and based on the work done by the OECD The OECD handbook distinguishes the and the Task Force mentioned above, concrete internationalisation of R&D into working steps towards data collection should • Aspects linked to R&D performance: follow now. Therefore some of these indicators establishment of R&D activities in might be included in the regular data collection the host country by foreign-controlled in the years to come. affiliates (inward investment) and setting up of R&D activities abroad (outward 5.3. More data for the input to National Accounts investment). • Aspects relating to R&D financing: The System of National Accounts 1993 (SNA) is overall financing of R&D from abroad, currently undergoing a comprehensive revision financing of R&D destined for abroad which is to be completed by 2008 with the and financing of R&D from abroad on publication of the “SNA 93, Rev.1”. One of the

58 KNOWLEDGE ECONOMY: R&D AND INNOVATION issues for updating the SNA is the treatment of A medium term priority would be to collect R&D in national accounts. more detailed data within R&D statistics which The present SNA does not recognize the output of would allow the identification of transactions R&D as capital formation, but treats it essentially on acquisitions and sales of R&D with change as intermediate consumption of market and non- of ownership from one unit, sector or country to market producers. This might very probably change another. as the responsible groups at international level Eventually, the long term goal would be to modify (Canberra II group) recommended inter alia that some parts of the Frascati manual and the R&D • the 1993 SNA should be changed to statistics to obtain accurate data on R&D output recognise the outputs of R&D as at market prices in current and constant price assets, and the acquisition, disposal and terms, by SNA sector and by use, and ultimately depreciation of R&D fixed assets should also data to produce accurate estimates of R&D be treated in the same way as other fixed capital stocks and depreciation by sector. assets; 5.4. R&D now-casts and forecasts • all R&D output should be treated as an asset, irrespective of its nature or As said above, Eurostat developed an estimation whether it is made freely available; method for nowcasting some of the main data • the definition of R&D given in the and indicators on R&D. The application of such Frascati Manual (FM) should be adopted method and the release of the estimated data in the SNA; would thoroughly improve the availability of timely data, which is highly needed by policy • the R&D statistics based on the Frascati Manual (FM) provide the best source makers. In general, the nowcasts produced would of data for deriving estimates of R&D refer to a delay of only 3 months after the end of statistics, principally gross fixed capital the reference period ( T + 3 months). formation (GFCF). However, there are shortcomings in the current R&D However, this method still has to be improved statistics and the FM should be amended and further tested within the Eurostat production to better support the needs of the SNA. system. In addition, the agreement of countries to the release of the estimated data has also to be This means that based on the recommendations obtained. expressed above, additional data needs are or will be expressed towards improving the In a second step this estimation model could current international R&D statistics. be developed further in order to use it also for forecasting. The production and release of The short term priority is to construct aggregate forecasted data is however out of the scope of the simplified bridge tables that would allow the tasks of Eurostat and should therefore be taken estimation of total capital formation of R&D over by other bodies. within the economy with reasonable accuracy. These aggregate simplified bridge tables would 6. Conclusion cover the main gaps in concepts, definitions and classifications between R&D statistics and national In recent years the EU R&D statistics made accounts by combining existing R&D statistics considerable progress. This momentum has to be with national accounts data. This would not yet kept as further challenges are coming up, based require changes to the harmonised R&D statistics on new and additional user needs, with still more and the Frascati Manual. data and better data quality to be achieved.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 59 ANNEX 1: R&D personnel

Data Dissemination at the National R&D personnel Eurostat webpage under • Total R&D personnel by sectors of performance (employment), occupation ‘Science and Technology’ and sex • Total R&D personnel and Researchers as RESEARCH AND DEVELOPMENT % of labour force and total employment • Total R&D personnel and researchers by Statistics on research and development sectors of performance (employment) and fields of science R&D expenditure • Total R&D personnel and researchers by sectors of performance (employment), National R&D expenditure qualification and sex • Business enterprise R&D personnel by • Total intramural R&D expenditure economic activity, type of occupation (GERD) by sectors of performance and sex Total intramural R&D expenditure • • Share of female researchers by sectors of (GERD) by sectors of performance and performance (employment) fields of science • Researchers by age and sex (HC) in • Total intramural R&D expenditure Government and Higher education sector (GERD) by sectors of performance and • Researchers by citizenship and sex (HC) source of funds in Government and Higher education • Total intramural R&D expenditure sector (GERD) by sectors of performance and • R&D personnel and researchers (FTE) by type of costs size class in Business enterprise sector • Total intramural R&D expenditure (GERD) by sectors of performance and socio-economic objectives Regional R&D personnel • Business enterprise R&D expenditure (BERD) by economic activity and source • Total R&D personnel by sectors of of funds performance (employment) and region • Business enterprise R&D expenditure (BERD) by economic activity Government budget appropriations or outlays on R&D • Key indicators - GERD by source of funds (%) Annual provisional data on GBAORD by • Total R&D expenditure (GERD) by • sectors of performance and type of activity NABS socio-economic objectives at the chapter level • Business enterprise R&D expenditure (BERD) by size class and source of funds • Annual final data on GBAORD by NABS socio-economic objectives at the chapter Regional R&D expenditure level • Annual final data on total GBAORD • Total intramural R&D expenditure as a % of total general government (GERD) by sectors of performance and expenditure, for total NABS socio- region economic objectives only

60 KNOWLEDGE ECONOMY: R&D AND INNOVATION ANNEX 2: R&D indicators related to internationalisation

Box 1. R&D indicators concerning multinational enterprises of a compiling country

KNOWLEDGE ECONOMY: R&D AND INNOVATION 61 Innovation measurement: present and future challenges

Carter Bloch *

Paper prepared for the Eurostat Conference, “Knowledge Economy – Challenges for Measurement”

Luxembourg, December 8-9, 2005

promotion or pricing. Organizational innovations Executive Summary are the implementation of new organizational methods in firms’ business practices, workplace Reflecting both developments in the innovation organization or external relations. concept and the changing economic environment, innovation policy has taken on a broader scope, The new Oslo Manual also devotes greater increasing emphasis on ‘non-technical’ forms of attention to knowledge transfer and linkages innovation, market driven innovation, knowledge in the innovation process, where linkages are transfer and firms’ capacity to capture and utilize characterized by their source, cost and level knowledge. Changes in the recently completed of interaction. Three types of linkages are third edition of the Oslo Manual reflect these identified: open information sources, acquisition developments with the aim of providing data of knowledge and technology, and innovation to improve understanding of innovation and to cooperation. inform policymaking. Open information sources provide access to The innovation measurement framework in the knowledge without the purchase of technology or new Oslo Manual includes, in addition to product intellectual property rights, or interaction with the and process innovations, marketing innovation and source. Acquisition of technology and knowledge organizational innovation. Marketing innovations involves the purchase of external knowledge are the implementation of new marketing methods and/or knowledge and technology embodied involving significant changes in product design in capital goods and services, which do not and packaging, product placement, product involve interaction with the source. Innovation

* The Danish Centre for Studies in Research and Research Policy, University of Aarhus, Finlandsgade 4, 8200 Aarhus N., Denmark; Tel. (+45) 8942 2398, E-mail: [email protected].

62 KNOWLEDGE ECONOMY: R&D AND INNOVATION co‑operation involves active participation in joint pace. Innovation surveys are now conducted innovation projects with other organizations. in a broad range of countries, including EU countries, other OECD countries and a growing The new guidelines in the Oslo Manual thus number of non-OECD economies. The first allow for greater coverage of a number of areas Oslo Manual set down guidelines for surveying important for innovation policy, such as the ‘full technological product and process innovation in scope’ of innovation policy and interactions with manufacturing industries, providing a standard other firms and institutions. However, a full framework for internationally comparable assessment of the usefulness of data based on the innovation statistics. This framework has since new guidelines can only be made after the new been expanded to include innovation in service data have been put into use in analysis and policy sectors. development. Here it is important to recognize that policy needs require not only a broader The recently completed third edition of the Oslo coverage of firm innovation but also better use Manual1 has undergone a number of substantial of innovation data in order to understand firm changes, with the aim of keeping innovation innovation. Data on the implementation of measurement abreast of policy needs and changes innovations, on knowledge transfer and from in the innovation concept and the economy. auxiliary questions can be used to compile Among the most important changes to the manual composite indicators that provide valuable are: a broadened definition of innovation to information on how firms innovate. Examples are include marketing innovations and organizational indicators that identify whether firms are leaders innovations; a much expanded coverage of or adopters, whether their innovation is primarily knowledge flows and the role of linkages in the R&D or technology driven or market driven, and innovation process; and an adaptation of the whether firms’ innovations are integrated over manual to reflect the importance of innovation in various firm activities. less R&D-intensive industries, such as services and low-tech manufacturing. Other areas of relevance for innovation policy receive only limited attention due in part to A number of recent insights on innovation are their coverage elsewhere, such as the role of of great relevance for innovation measurement. human capital, intellectual property rights and For example, it is increasingly recognized that environmental innovation, or are outside the innovation is possible without conducting R&D, scope of the Oslo Manual, such as innovation and that the inspiration for many innovations may in the public sector. Work remains to develop be market based. Thus, many innovation projects indicators for public sector innovation, and to may originate from contact with customers and better integrate existing work in areas such as suppliers, or market analysis, as opposed to human capital with that of general innovation new research results or the development of new measurement. technologies. In addition, the role of linkages between firms, institutions and other actors has 1. Introduction grown in importance for innovation processes, as has the role of firms’ ability to access and From their beginnings in the 1980’s, the use external knowledge, i.e. their absorptive or measurement of innovation has grown at a rapid learning capacity.

1 OECD/Eurostat (2005).

KNOWLEDGE ECONOMY: R&D AND INNOVATION 63 Changes in the economic environment have developments that help shape the environment had a large impact on innovation measurement. in which policy is made. Section 3 discusses The service sector has grown dramatically in central EU policy areas related to innovation. economic importance for the EU and other OECD Section 4 outlines the main elements of the recent countries, along with the recognition that a great Oslo Manual revision and section 5 assesses deal of innovation may take place in services. how well these changes meet policy needs. Globalization has increased both the potential for Section 6 discusses some current and potential interaction and pressures on firms of all types to future methods for using innovation data to innovate in order to maintain competitiveness. develop a better understanding of firm innovation. Progress in information and communication Section 7 concludes. technologies (ICT) has dramatically enhanced opportunities for knowledge transfer and for firms 2. Changes in the policy environment in both low and high tech sectors to implement ICT-based product or process innovations. A number of developments in the understanding of innovation and economic changes have set The Lisbon strategy for the EU outlines targets the stage for the Oslo Manual revision, and have for innovation, competitiveness and economic helped form the background for innovation policy performance2. Innovation policy objectives formation. address a number of areas, among them improving the transformation of research into innovations, The limitations of a linear view of innovation regional development, entrepreneurship, public have long been recognized. The transformation sector innovation, and the use and development of research into new products or processes is a of ICTs. They are also closely related to other process of interaction with a number of actors, policy areas, such as environmental, education, often involving feedback and redesign of initial employment and trade policies. Data and analysis innovations (Klein and Rosenberg, 1986). The within all these areas can be of great use for policy concepts of national or regional systems of development and the coordination of different innovation have also received increasing attention policy areas. since their initial development. The systems of innovation approach highlights the influence of This paper discusses the main changes that have external institutions on the innovation activity of been undertaken in the new Oslo Manual, and firms and other actors, and the role of interactive examines how well the new framework addresses processes in the creation, diffusion and application policy needs. The changes to the Oslo Manual of knowledge (Lundvall, 1992; Nelson, 1993). reflect developments in our understanding of innovation processes and in the economy, and Additional insights have appeared in recent years. also the accumulating experience in innovation The role of knowledge transfer and management measurement, allowing the extension of previous is growing in importance. Knowledge has become frameworks to new areas. These changes make increasingly complex and individual firms’ significant progress in meeting many policy knowledge more specialized (Pavitt, 2005). Firms demands for a more comprehensive coverage of are thus more dependent on external knowledge, innovation and knowledge transfer. with information sources and cooperative partners taking a greater role in innovation. This The next section of this paper discusses thus places additional demands on firms’ access recent insights on innovation and economic to external knowledge and also on firms’ ability

2 Commission of European Communities (2003).

64 KNOWLEDGE ECONOMY: R&D AND INNOVATION to utilize external knowledge, i.e. their absorptive Globalization has had a large impact on innovation. or learning capacity. Organizational structures Firms have much greater access to information and methods play a significant role in determining and markets, and are much more easily able firms’ ability to learn and innovate (Lam, 2005). to undertake joint projects with firms in other countries. Globalization has at the same time Innovation has become more market driven than meant increased international competition, making before (OECD, 1999). In searching for new innovation vital for firms in all industries. For ideas for products, processes or other business many firms, competition on price and efficiency is methods, firms may look to a greater extent than not enough. They also need to compete on product earlier towards customers, suppliers or market characteristics and marketing methods. research. This implies a greater emphasis on the later stages of innovation development, but it Advances in ICTs have resulted in a large increase also reflects changes in the flow of knowledge. in opportunities for obtaining and exchanging Customers may provide input to initial work, knowledge. ICT-related innovation can take while public research institutions may assist over place in very wide range of industries. Firms the course of the development process as opposed may implement ICTs to improve organizational to functioning as the initial source of research. efficiency or utilize ICTs in products, production This market orientation may also require greater processes, or other business activities. interaction between business functions and departments in their innovation activities. 3. Policy needs

Developments in our understanding of innovation Parallel to developments in innovation theory and are to an extent due to economic changes that the economy, innovation policy has increasingly have had a large impact on innovation processes taken on a broader scope. While the importance of in recent years, in particular, the growth of the R&D for innovation and economic performance service sector, globalization, and information and has been maintained, greater emphasis has been communication technologies (ICTs). placed on improving the capacity to transform new and existing knowledge into new products A number of recent studies have focused on and processes, with corresponding increases in innovation in services3. A central point is that a employment and growth. great deal of innovation activity takes place in service firms, though innovation in services may The Lisbon strategy for the EU outlines targets often be quite different from that in manufacturing. for innovation, competitiveness and economic The distinction between products and processes performance. In designing innovation policy, in services is often blurred, with production it is essential to understand what drives firm and consumption occurring simultaneously. performance and economic growth and how EU Development of products and processes can economies can best benefit from investments in be more informal for services than for goods, R&D and innovation. While there is a notable with an initial phase consisting of search, idea emphasis on the importance of increasing R&D gathering and commercial evaluation, followed in the EU4, there is also a growing recognition of by implementation. Innovation activity in services the need to focus on the ‘full scope’ of innovation, also tends to be a continuous process, consisting including ‘non-technical’ forms of innovation. of a series of incremental changes in products By developing a more complete understanding and processes. of innovation processes, innovation policy may

3 E.g. Hauknes (1998), de Jong et.al. (2003), and Howells and Tether (2004). 4 Commission of the European Communities (2002).

KNOWLEDGE ECONOMY: R&D AND INNOVATION 65 be better equipped to identify what areas to development of cleaner energy sources, motivate address, and how. collecting data on environmental innovation (Kemp and Arundel, 1998). Measures here EU policy statements highlight the importance of might include regulations, direct support for the transforming research into innovations, elements development of environmental technologies, and of which might be an understanding of customer business service centers. demand and the marketplace and organizing different activities of the firm in order to best Policy for intellectual property rights (IPRs) centers utilize both their own R&D and the knowledge around streamlining the patenting process and and innovations of other firms. Policy here addressing difficult issues concerning whether to can involve a variety of measures, including strengthen certain IPR areas such as software. Data supporting research on market development, here is needed on firms’ use of protection methods value creation and entrepreneurship, education, and how legislation affects firm innovation. regulations that may affect the development of marketing strategies, value-chain and industry- Innovation policy also includes measures to science relations, and business service centers. promote the use and development of ICTs. Policy here can cover a wide array of measures, The recognition that innovation is widespread including education and training, business service, across sectors implies that policy should not be infrastructure and direct support. limited to a few selected sectors. Though, many measures may need to be tailored to the specific 4. The Oslo Manual Revision innovation characteristics of industries or target groups. The revision of the Oslo Manual is in many ways a joint product of policy needs and developments The regional dimension of innovation policy is in the innovation concept and changes in important for a variety of reasons. A number of the economy. These demands on innovation regional factors are central to firms’ performance. measurement are at the same time balanced with These include the presence of related firms previous measurement experience and a trade off (supplier, customers and competitors) and public between looking forward in measuring innovation research institutions, the availability of skills and and maintaining continuity with previous surveys expertise, and an infrastructure that is supportive in order to follow developments over time. This of innovation. Regional innovation policy is also section will briefly discuss the main aspects of an important instrument in promoting growth in the measurement framework in the Oslo Manual lesser developed regions. Regional policy is also and its departures from earlier frameworks. relevant in strengthening the development of industrial clusters in selected regions. Innovations

The public sector can function both as a source of The most central change to the Manual is the use innovation and as a consumer of new products, of a broader definition of innovation. In addition processes and other methods. An important policy to product and process innovations, the definition goal is improving both the efficiency of public of innovation now includes marketing innovations sector operations and the quality of services. and organizational innovations. Operationalizing Data on public sector innovation may thus be the framework for measuring all four types of very useful in determining policy measures. innovations proved to be a challenging task. Among the most difficult issues here were Environmental policy goals, such as pollution delineating types of innovations and adapting control, recycling, energy conservation, and the other topics in the manual to the broadened

66 KNOWLEDGE ECONOMY: R&D AND INNOVATION definition. The overall definition of innovation in Marketing innovations cover new marketing the new Oslo Manual is the following: methods aimed at better addressing customer needs, opening up new markets, or newly positioning a firm’s product on the market, with An innovation is the implementation of a new the objective of increasing the firm’s sales. In or significantly improved product (good or order to ease firms’ understanding of this concept, service), or process, a new marketing method, the definition is based on the well-known 4 P’s or a new organizational method in business model for marketing strategies: Product, Price, practices, workplace organization or external Placement and Promotion. relations. A marketing innovation is the implementation The definitions of product and process innovations of a new marketing method involving were formulated with the aim of maintaining significant changes in product design or continuity with definitions in the previous edition packaging, product placement, product of the Oslo Manual5 and those used in recent promotion or pricing. surveys such as CIS3. Hence, only minimal changes to the definitions of product and process innovations have been made. One noteworthy Product design changes refer to changes in product form and appearance that do not alter change is the removal of the word ‘technological’ the product’s functional or user characteristics. from product and process innovations. While They also include changes in the packaging of product and process innovations still require products such as foods, beverages and detergents, significant improvements in functions or uses, where packaging is the main determinant of the removal of the word ‘technological’ is the product’s appearance. Product placement designed to make these concepts more suitable involves methods used to sell goods and services for less R&D intensive firms, for example in the to customers. Promotion includes concepts for service sector. promoting a firm’s goods and services, such as new advertising methods or new brand symbols. A product innovation is the introduction of Pricing involves the use of pricing strategies to a good or service that is new or significantly market the firm’s goods or services. improved with respect to its characteristics or intended uses. This includes significant Organizational innovations involve new methods improvements in technical specifications, in three areas. Business practices are routines or components and materials, incorporated procedures for the conduct of work. These can software, user friendliness or other functional range from practices for sharing knowledge to characteristics. the sets of procedures involved in management systems. Workplace organization involves A process innovation is the implementation organizational structures and the distribution of a new or significantly improved production of responsibilities and decision making, while or delivery method. This includes significant external relations involve the organization of changes in techniques, equipment and/or software. relations with other firms or public research institutions.

5 OECD/Eurostat (1997)

KNOWLEDGE ECONOMY: R&D AND INNOVATION 67 Linkages in the innovation process An organizational innovation is the implementation of a new organizational The third edition of the Oslo Manual also method in the firm’s business practices, includes a much greater coverage of knowledge workplace organization or external relations. transfers and linkages. While the second edition contained some coverage of linkages in terms An important challenge in defining the four of information sources and a brief discussion of types of innovations was addressing how best R&D cooperation, emphasis on the importance to distinguish between innovation types for of linkages in theory, policy and recent surveys borderline cases. Efforts have been made to motivated expanding on this issue. A separate minimize borderline cases, though it was not chapter has now been devoted to linkages, which considered feasible or desirable to make clear- presents a coherent framework in which linkages cut distinctions among types. Distinctions will are characterized by their source, cost and often depend on the nature of the firm’s business level of interaction. Three types of linkages are and on the specific characteristics of a firm’s identified: open information sources, acquisition innovations. In many cases, innovations may of knowledge and technology, and innovation actually span more than one type. Examples are a cooperation. firm that introduces a new product that requires the development of a new process or that introduces a Open information sources provide access to new marketing method to market a new product. knowledge without the need to pay for the These ‘integrated innovations’ may often involve knowledge itself, although there may be marginal coordination of innovation activities across a fees for access (membership in trade associations, firm’s functions or departments, and thus are of attendance at conferences, subscriptions to great interest. journals). This type of linkage involves the transfer of codified knowledge, though some open The ongoing CIS4 survey of innovation in EU sources, such as attendance at fairs or exhibitions, countries was initiated prior to the completion of can give access to some tacit knowledge through the new Oslo Manual, though much of the new personal interaction with other participants. framework has been incorporated into the survey6. Questions on all four types of innovations are Acquisition of technology and knowledge included, though marketing and organizational involves the purchase of external knowledge and innovation are placed in a separate section at technology without active co‑operation with the the end of the survey. Each type of innovation source. External knowledge can be embodied in is divided into subtypes (e.g. goods and services machinery or equipment, in new employees, or for product innovations; production methods, the in use of contract research and consulting delivery methods, and support services for process services. Disembodied technology or knowledge innovations; product design and sales methods also includes other know-how, patents, licenses, for marketing innovations; and management trademarks and software. systems, work organization and external relations for organizational innovations). This provides Innovation co‑operation involves active greater detail on the types of innovations that participation in joint innovation projects with firms implement. other organizations. Innovation co‑operation

6 In particular, the definitions of marketing and organizational innovations have been modified since the launch of CIS4.

68 KNOWLEDGE ECONOMY: R&D AND INNOVATION allows enterprises to access knowledge and A number of recent innovation surveys have technology that they would be unable to utilize included questions on information sources and on their own. Innovation co‑operation can take innovation cooperation. CIS4 includes a question place along supply chains, involving customers of the relative importance of information sources and suppliers in the joint development of new and on the geographic location of innovation products, processes or other innovations, or it can cooperation partners. There is also a question on involve horizontal collaboration, with enterprises acquisitions of external knowledge and capital working jointly with other enterprises or public goods related to the firm’s innovation activities, research institutions. though this does not specify the type of source the purchases were made from.

Types of linkages: Novelty Open information sources: openly available Novelty is a central concept in examining information that does not require the purchase innovation. The previous Oslo Manual used the of technology or intellectual property rights, or concepts, new to the world, new to the country interaction with the source. and new to the firm. New to the world clearly Acquisition of knowledge and technology: gives the highest degree of novelty, but in many purchases of external knowledge and/or cases, ‘country’ may not adequately describe the knowledge and technology embodied in capital competitive environment that a firm innovates goods (machinery, equipment, software) and in. In order to measure novelty in terms of firms’ services, which do not involve interaction with competitiveness, the concept new to the market is the source. utilized in the manual. This concept has already been used in recent innovation surveys such as Innovation co‑operation: active co‑operation CIS3 and CIS4. with other enterprises or public research institutions for innovation activities (which Innovation activities may include purchases of knowledge and technology). Innovation activities are defined in the Oslo Manual as “all scientific, technological, organizational, financial and commercial steps These three types of linkages are thus which actually, or are intended to, lead to the distinguished by the level of interaction with implementations of innovations”. They include the source, and to an extent also by the costs both activities for ongoing work on innovations involved in accessing the knowledge. The and also R&D that is not directly linked to a framework allows both for an examination of specific innovation. which types of linkages are used by firms in their innovation activities and a comparison The new Oslo Manual expands the coverage of the relative importance of different types of of innovation activities to include activities for linkages and sources. Sources include: market marketing and organizational innovations. There and commercial sources, such as competitors, were a number of issues involved in this. For clients, suppliers and consultants; public sector example, a number of activities, such as R&D and sources such as universities, public research acquisitions of external knowledge or equipment institutes and public support services; and and software, may be relevant for all 4 types of general information sources, such as patent innovations. On the other hand, it may be useful disclosures, conferences, scientific journals, to be able to have information on innovation informal networks and standards. activities for individual types of innovations.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 69 This would also allow greater comparison over Other preparations for product and process time for figures on innovation expenditures. innovations: Other activities related to the development and implementation of product The convention that was adopted in the Oslo and process innovations, such as design, Manual was to maintain the list of innovation planning and testing for new products (goods activities used in the previous edition of the Oslo and services), production processes, and Manual (though in a slightly modified form), and delivery methods that are not already included add separate categories for activities for marketing in R&D. innovations and for organizational innovations. Market preparations for product innovations: Activities aimed at the market introduction Innovation activities: of new or significantly improved goods or Research and experimental development: services. Intramural (in‑house) R&D: Creative work Training: Training (including external training) undertaken on a systematic basis within the linked to the development of product or process enterprise in order to increase the stock of innovations and their implementation. knowledge and use it to devise new applications. Activities for marketing and organisational This comprises all R&D conducted by the innovations: enterprise, including basic research. Preparations for marketing innovations: Acquisition of R&D (extramural R&D): Same Activities related to the development and activities as intramural R&D, but purchased implementation of new marketing methods. from public or private research organisations Includes acquisition of other external or from other enterprises (including other knowledge and other capital goods that is enterprises within the group). specifically related to marketing innovations. Activities for product and process Preparations for organisational innovations: innovations: Activities undertaken for the planning and Acquisition of other external knowledge: implementation of new organisation methods. Acquisition of rights to use patents and Includes acquisition of other external knowledge non‑patented inventions, trademarks, and other capital goods that is specifically know‑how and other types of knowledge related to organisational innovations. from other enterprises and institutions such as universities and government research This classification has the advantage of preserving institutions, other than R&D. continuity over time for questions on innovation activities for product and process innovations. A Acquisition of machinery, equipment and disadvantage is that little information is obtained other capital goods: Acquisition of advanced machinery, equipment, computer hardware or on the specific types of activities engaged in software, and land and buildings (including for marketing and organizational innovations major improvements, modifications and as, with the exception of R&D, all innovation repairs), that are required to implement product activities for each of these two types are included or process innovations. Acquisition of capital in a single category. goods that is included in intramural R&D Innovation surveys can include both qualitative activities is excluded. (yes/no) and quantitative (amount of expenditures)

70 KNOWLEDGE ECONOMY: R&D AND INNOVATION on innovation activities. Information on Data on all 4 types of innovations give a more expenditures can be very useful for measuring complete view of firm innovation that can aid innovative intensity (innovation expenditures both understanding and policy development. as share of revenues) and calculating returns to The broader coverage provides information innovation activities. However, questions on on marketing innovations and organizational innovation expenditures are very difficult to innovations, which allows both an examination answer as firms do not keep financial data on of their role in firm performance and how they expenses for a number of the activities. In light function as support for the success of product or of this, many innovation surveys (among them process innovations. CIS4) may limit expenditures questions to those activities for which firms can be expected to A wide range of policy measures concern the have financial data: intra- and extramural R&D, promotion of knowledge flows and interaction acquisition of machinery, equipment and other among firms and with public research institutions. capital goods, and acquisition of other external The expanded coverage of linkages can provide knowledge. information on a number of related issues, such as linkages to public research institutions, 5. The Oslo Manual and policy needs international knowledge flows, a comparison of the importance of sources of codified information How does the Oslo Manual meet the policy with that of active innovation partners, and on demands described above? The Oslo Manual the relative importance of different sources of has been designed with a view to collecting information. data for multiple users, including researchers, policymakers and other stakeholders. However, Innovation surveys can provide a wide range the focus here will primarily be on uses for of information on firm innovation, though there innovation policy. are also a number of limitations to the type of data that can be collected. First, space limitations This section discusses which areas and to what in surveys mean that not all relevant areas can extent the Oslo Manual covers topics relevant likely be covered in a single survey. Second, it is for innovation policy. However, the usefulness difficult to capture the timing of firm innovation, of this data will depend to a great degree on as innovation activities, their implementation what is done with the data. The next section and subsequent impact may span over several will discuss ways in which the data can be used years. Third, full measures of firms’ expenditures to compile indicators that may provide a more on innovation activities, while of great policy comprehensive view of innovation. relevance, may be very difficult to obtain. Fourth, innovation surveys are not well suited Innovation surveys continue to cover a broad to gain information on the general institutional range of sectors, providing results that are environment, though they are able to examine representative for the entire economy and for how firms experience institutional factors. individual sectors or regions. Information can also be collected on regional linkages, for There are also a number of related areas for which example which regional sources are most used detailed coverage is not feasible in innovation and the importance of regional linkages relative surveys, though where further work could be to national and international sources. Given done in order to better integrate the topics with the considerable interest in forming policies to innovation surveys. support and promote entrepreneurship, there is greater demand for data on small enterprises, There is very limited coverage of human capital in which is also reflected in the new Oslo Manual. the Oslo Manual. Coverage includes questions on

KNOWLEDGE ECONOMY: R&D AND INNOVATION 71 the share of employment with higher education, data, however as the share of innovative firms on training that is related to innovations, and on seeking patents is fairly small, greater coverage barriers concerning lack of qualified personnel of this topic in general innovation surveys would and attitudes to change. The role of human capital be less appropriate. is covered elsewhere, though it may be of value to consider a greater integration of this topic in Through questions on linkages and knowledge innovation surveys. transfer, the Oslo Manual contains a number of guidelines for obtaining information on the The role of ICTs is discussed in detail in the Oslo international aspects of innovation. However, Manual, though few questions are specifically difficult challenges remain in measuring the directed at ICTs (as an objective for innovation: ‘to globalization of innovation and other activities. improve IT capabilities’, and the implementation The main difficulty, which is shared by R&D and of new ICTs as a subtype of process innovations). other statistics, lies in the fact that statistics are Extensive coverage of this topic is outside generally compiled at the national level while the scope of the Oslo Manual, though lesser multinational enterprises’ activities may span extensions could be considered, such as whether several national boundaries. R&D and other innovations contain an ICT component. innovation activities may be conducted in one country, production in another, and sales in still Work on indicators for public sector innovation other countries. Thus, national data may in some is at a very early stage7, and the development of cases fail to provide a link between innovation a framework for measuring innovation activity and performance. in the public sector remains a promising area for future work. 6. Developing indicators for use in policy Coverage of environmental innovation in the analysis Oslo Manual is also limited, consisting of two The guidelines in the new Oslo Manual allow for questions on the objectives of innovations (‘to wider coverage of firm innovations and types of develop environment friendly products’ and ‘to knowledge transfer. However, policy needs require reduce environmental impacts or improve health not only a broader coverage of firm innovation and safety’). A full coverage of environmental but also better use of data in order to understand innovation is outside the scope of the Oslo Manual, firm innovation. Data on the implementation though optional questions could be developed of innovations, knowledge transfer and from for use in innovation surveys, such as a small auxiliary questions can be used to compile set of questions that ask directly whether any composite indicators that provide valuable environmental innovations were introduced (for information on how firms innovate. Examples are example, within: pollution control technologies, indicators that identify whether firms are leaders waste management, clean technologies, recycling, or adopters, whether their innovation is primarily clean products, and clean-up technologies8). R&D or technology driven or market driven, and whether firms’ innovations are integrated over The Oslo Manual contains a section covering various business activities. questions on the appropriability of innovations, which can provide useful information on firms’ Arundel and Hollanders (2005) provide some use of protection methods and their importance. examples of composite indicators. They Policy discussions for IPRs may need additional classify firms in terms of their innovative

7 See e.g work under the Publin project (http://www.step.no/publin/) and Earl (2003).

8 See e.g. Kemp and Arundel (1998).

72 KNOWLEDGE ECONOMY: R&D AND INNOVATION novelty and creative effort using CIS3 data on of innovations. As mentioned above, these technological product and process innovations. types of indicators can provide information on The classifications utilize CIS3 questions how firms innovate and whether combinations on innovations, R&D, novelty and whether of innovations have an important impact on innovations were primarily developed by firms performance, and they could function as indicators themselves or by others. of firms’ innovative capacity and the degree to which innovative activities are integrated across The inclusion of marketing innovation and functions and departments. Furthermore, data organizational innovations in innovation surveys can also be collected on subtypes of innovations, opens up possibilities for developing additional along the lines of the questions used in CIS4. indicators along these lines, such as ‘innovative This allows surveys to identify for example what modes’ that draw on information on all 4 types of types of marketing innovations, or types of other innovations. innovations, were implemented, providing a more detailed view of firms’ innovations. Innovative modes: Data on linkages and innovations can be used Strategic innovators: For these firms, to examine whether innovations are primarily innovation is a core component of their research or technology driven, or whether they competitive strategy. They perform R&D on a are user or market driven. Innovation projects can, continuous basis to develop product or process for example, emerge from new research results innovations. They are the main source of or be initiated by in-house work by engineers innovations that diffuse to other firms. and R&D personnel, or they can originate from Intermittent innovators: These firms perform contacts with suppliers and customers, or from R&D and develop innovations in-house when market research. necessary or favorable, but innovation is not a core strategic activity. For some, their R&D 7. Conclusion efforts focus on adapting new technology developed by other firms to their own needs. In expanding the coverage of innovation from Technology modifiers: These firms modify product and process innovation to a broader their existing products and processes through definition that also includes marketing and non-R&D based activities. Many firms in this organizational innovation, the Oslo Manual group are essentially process innovators that has undergone a very extensive revision. The innovate through production engineering. objective of this and other changes, such as increased coverage of knowledge transfers, is to Technology adopters: These firms primarily innovate by adopting innovations developed address recent economic developments and meet by other firms or organizations. data needs for both policy and research. Source: Arundel and Hollanders (2005). The new Oslo Manual provides information on a wide range of issues in innovation policy, though An additional example is indicators on the there are a number of related areas that are beyond implementation of combinations of innovations, its scope. The real test, however, of the usefulness such as product and process innovations, of the changes to the Oslo Manual will be in the product and marketing innovations, process actual use of the new guidelines and the resulting and organizational innovations, or all 4 types data that is generated from them.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 73 References Kline, S.J. and N. Rosenberg (1986), “An Arundel, A. and H. Hollanders (2005), EXIS: An Overview of Innovation”, in R. Landau Exploratory Approach to Innovation Scorecards, and N. Rosenberg (eds.), The Positive Sum European Commission Enterprise Directorate- Strategy: Harnessing Technology for Economic General. Growth, National Academies Press, Washington D.C. Commission of the European Communities (2002), More Research for Europe. Towards 3% Lam, A. (2005), “Organizational Innovation”, of GDP, COM (2002) 499. Chapter 5 in J. Fagerberg, D. Mowery and R.R. Nelson (eds.), The Oxford Handbook of Commission of the European Communities Innovation, Oxford University Press, Oxford. (2003), Innovation Policy: Updating the Union’s Approach in the Context of the Lisbon Strategy, Lundvall, B.-A. (ed.) (1992), National Systems COM (2003) 112. of Innovation: Towards a Theory of Innovation and Interactive Learning, Pinter Publishers, Earl, L. (2003), “Innovation and Change in London. the Public Sector: A Seeming Oxymoron”, Statistics Canada SIEID Working Paper Series Nelson R. (1993), National Innovation Systems, No. 2002‑01. Oxford UP, Oxford.

Hauknes, J. (1998) Services in Innovation, OECD (1999), Innovation and Economic Innovation in Services, SI4S Final Report, STEP Performance – Developing the Links. Paris. Group, Oslo OECD/Eurostat (1997) Proposed Guidelines Howells, J.R.L. and B.S Tether (2004), “Innovation for Measuring and Interpreting Technological in Services: Issues at Stake and Trends – A Report Innovation Data – The Oslo Manual, Second for the European Commission”, INNO-Studies Edition. 2001: Lot 3 (ENTR-C/2001), Brussels. OECD/Eurostat (2005) Proposed Guidelines for de Jong, J.P.J., A. Bruins, W. Dolfsma and J. Meijaard Measuring and Interpreting Innovation Data (2003), Innovation in Services Firms Explored: – The Oslo Manual, Third Edition. What, How and Why?, EIM Report, Zoetermeer. Pavitt, K. (2005), “The Process of Innovation”, in Kemp, R. and A. Arundel (1998), Survey J. Fagerberg, D. Mowery and R.R. Nelson (eds.), Indicators for Environmental Innovation, Idea The Oxford Handbook of Innovation, Oxford Paper 8, Step Group. University Press, Oxford.

74 KNOWLEDGE ECONOMY: R&D AND INNOVATION Benchmarking European High-Tech Regions

Methodical approach and data restrictions Eurostat 2005 Conference: Knowledge Economy– Challenges for Measurement Peter Kaiser Luxembourg, Dec. 8th 2005

„Nomen est Omen“: Mediumand long term On the Relationship between Technological forecasts Innovation and Economic Growth

The „Report on Germany 2002- • The elements that make a regional economy 2020” A consistent Scenario on Ger- vibrant and prosperous today are fundamentally many’s future — economic growth, la- different from those of the past. bour market, qualifications and social • The new economics of place are driven by policy, future trends in industry, demographic their ability to attract and expand science trends and technology assets and leverage them for economic development. The „Prognos Atlas of Technology • Regional economic performance is determined 2002” Technological potentials and by how effectively its comparative performance of the German Regions advantages are used to create and expand knowledge assets and convert them into The „Prognos Zukunftsatlas 2004- economic value. 2005-2006“ Benchmarking future • A study by the Council of Economic Advisers prospects of the German Regions fo- in the United States recently concluded that cussing on special topics, e.g. regional 50% of the growth in the American economy development (2004), family (2005) econom- over the last 40 years had been due to ic clusters (2006) investments in research and development. The World Reports Short, middle and long term scenarios on economies Challenges to be met by high wage and markets for 90 per cent of the economies world‘s economy

The European Transport Report • Generation and/or absorption of new Data for analysis and scenarios on knowledge, trends in logistics and transport in 22 • Keeping and/or creation of adequate european countries framework conditions and • to maintain/create efficient innovation The Sectoral Report „Entsorgung- processes swirtschaft“ with options and policy • in order to come up with commercially viable proposals new products, attractive and enough new jobs

KNOWLEDGE ECONOMY: R&D AND INNOVATION 75 Regional perspective: Final translation of Challenges for research establishing a set into economic relevant activities knowledge of key indicators in order to… takes place at the regional level. Not all regions are equal • Measure the performance and understand the – Metropolitan and technology regions are underlying structures of innovationbased, the major producers and appliers of frontier growthorientated economies. knowledge • Define and understand the term innovation – Other regions must have a chance to be knowledgebased and its dynamics and able to apply new knowledge generated implications for regional and local economic in the above mentioned regions as fast as development. possible. • Assess the implications for public intervention and determine the linkage between knowledge based and more traditional business models. • Provide a subregional analysis of national´s innovationbased economy. • Generate comparative benchmarks with leading regional economies.

Benchmarking regions by creating an index that encapsulates a comprehensive inventory of technology and economic development.

Technological Potential/Input Economic Performance/Output

• Highly qualified employees in technologyorientated industries • Gross value added • Patent application per mill. • Development of Gross value Labour force added • Employees in knowledge • Level of employment intensive services • Dynamics in employment • Population with tertiary education • Development of these parameters Annotation: • Percentage of R&D employees Bold type: German AND EU-15 • Percentage of engineers analysis • Development of this parameter Bold type in italics: EU-15 analysis • Business startups in technology only orientated industries Standard: German analysis only

76 KNOWLEDGE ECONOMY: R&D AND INNOVATION The technology map 2002 identifies successful and future oriented regions.

• North-South and East-West Deviding: South Germany surpasses West, North and East Germany by a substancial margin

• Technology axis and Technology islands: There are coherent regions with high technological efficiency.

• Regions with low innovation level (Screwdriver regions): Peripheral North and East German regions with structural problems

The technology map 2002 identifies successful and future-oriented regions.

• Strong regional concentration of • Technological capacities of today are an technological capacities: early indicator for economic success of Metropolitan regions do have the tomorrow: Technological capacities are best prerequisites for technological converted largely in economic success. competitiveness.

KNOWLEDGE ECONOMY: R&D AND INNOVATION 77 Interrelation between technological capacity and future prospect

Similar regional pattern of technological success and socio-economic prospects of regional future

78 KNOWLEDGE ECONOMY: R&D AND INNOVATION „European Technology Map 2005” — Overall Index

„European Technology Map 2005” — Dynamics Index (2000-2004)

KNOWLEDGE ECONOMY: R&D AND INNOVATION 79 „European Technology Map 2005” — Strength Index (2004)

What determines the regional What determines the regional technological capacity and technological capacity and competitiveness? competitiveness?

• Balanced industry and/or technologymix, • Existence of competence centres, cluster which also represents a riskmix structures, and an effective networking of the (specialization with simultaneous main regional stakeholders diversification) • A high absolute and relative level of R&D • Integration into the international exchange employees in economy and science of knowledge, innovations and goods with • An economic friendly climate and a leading regions in the world conspicuous image of a high tech region, • Basis on technologically active and which work as self amplifiers during the experienced industries/firms development process. • Application orientated production of • Strategic focusing of the economic and knowledge in R&D institutions technology policy on growth and competence fields and the requirements of the economy.

80 KNOWLEDGE ECONOMY: R&D AND INNOVATION Indicators on Research & Development, condition of efficient spending of public (“strengthen the strengths”). Innovation and Economic Success money – Difficulties & restrictions What is the use of it for regional • Available Indicators for Input and economic development? Output don’t cover the whole process of (technological) innovation • Decision guidance for the changes of EU • Emphasis lies on input indicators structural funds to be under way. • Only a few output indicators are available, • Enabling regions setting up a benchmarking more are needed process in order to learn from the best (as a • Process indicators are also needed, to give kind of regional “Lissabon-process”). a broader view of innovation and economic • Indicator-based analysis as a basis for success regional governance (interrelationship • Time-lag of data-availibility inhibits between technology policy and economic interrelationship to policy-action performance). • Incompletion by regional means • Assessment of effects of structural and • Restricted comparability-Different political innovation policy and its regional impacts. and administrative importance of NUTS- Level 1, 2 and 3 in the member-states Thank you very much for your attention!

Prognos AG What is the use of it for regional Division economic development? „Regions & Innovations“ • Identification of regionalstrength and weaknesses finding out the future challenges. Peter Kaiser • Observation and analysis of positive regional Consultant spill-over-effects. Peter.Kaiser@ • Absorption capacity for structural policy prognos.com and structural aid. • Starting point for measuring the effectiveness Wilhelm-Herbst-Straße 5 D-28359 Bremen of regional cluster policy. Tel. +49 421 20 115 782 • Decision guidance for governance under the Fax +49 211 887 97 8582

KNOWLEDGE ECONOMY: R&D AND INNOVATION 81 Report on the session: ICT Impact on the Knowledge Based Society

Chair: Michel Glaude, Eurostat

Four presentations addressed the influence of ICT performance. Macro and micro data show that on society and economy, and the achievements ICT investment has an impact, with the European and problems to measure it properly. progress falling short compared to the United States (US). Results of research at firm level gave The first presentation gave an overview on our new insight with regard to impact measurement. behavioural changes by the communications However, the comparability of data among technology over the years, and the myths and countries and regions such as Europe and the US reality how ICT changed our lives: Instead of is quite limited due to differences in measurement. paperless offices the demand for paper has been Differences in training of employees and the use growing, the amount of phone calls reached new of ICT for business processes and organisation dimensions, the demand for retail space has seem to be of importance, but additional studies increased, and travelling did not at all decline. By at micro level will be needed. the new technology the world became smaller, because it removed geographical constraints, Over the years the OECD and Eurostat have allowing for new kinds of interactions with developed a strategic partnership on developing new communities that is changing social life. tools and techniques for ICT measurement. However, that technology also changed the way A new chapter will be the use of firm level how we are using our time, and new applications data for assessing the ICT impact on business and sources of information may reach a limit for behaviour and performance: Eurostat recently using it efficiently – a challenge for developing launched a call for proposals on developing appropriate statistics. a work process for measuring that impact at European level by linking official data from A challenge for society is that chances by the new different sources. technologies are related to access and skills, and not all groups of society are benefiting equally. The two final contributions focused on e-skills That is true within Europe, as indicated by the and their measurement, which are crucial for the presentation on Eurostat’s data on e-skills. In effective usage of ICT. addition, the recent World Summit on Information Society in Tunis also concluded that education The presentation on the measurement of ICT and skills to use the new technologies are key skills in Eurostat’s Information Society statistics issues for social and economic development - a informed on the political and legal background, challenge for measurement at both European and the e-skills variables of the two annual global level. Community surveys on Information Society for enterprises and for households, and on current The second presentation addressed the strengths problems and future developments. First results and weaknesses of current Information Society from the 2005 survey confirmed one of the main statistics with regard to firms’ outcome and problems as discussed earlier: Digital literacy

82 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY is an important problem throughout Europe, in The conclusions focussed on how to strengthen particular for persons older than 55 years, lower and improve the ICT impact measurement at educated citizens and, to a lesser extent, women European level in the future. Such measurement and unemployed. with involvement of each of the Member States would be most convincing to give evidence to The final presentation was tackling the challenges national policy makers of what influences social in measurement in relation to e-skills. Though and economic behaviour, and which kind of it is one of the main interests in policy to know programmes would be most promising to enhance about e-skills that are on the one hand needed performance. by employers and on the other hand available on the labour market, there is still a dispute on Lack of skills seems to be the main driver for the definition of e-skills and competences. The the digital divide. Public programmes should European e-skills forum proposed three types: ICT address that issue, in particular when it comes to practitioner skills, ICT user skills and e-business developing e-government platforms. skills. Unfortunately these types are already slightly different from the 3 types proposed With regard to business productivity there seems by the OECD. Studies on practitioner skills in to be no needs for starting new data collections. different Member States indicate different needs Exploitation of current data by linking micro-data and shortages. As the understanding of e-skills at Member States’ level that refer to a number of and competences differ among those countries, Eurostat surveys, such as the Information Society over all conclusions at European level have to be survey on enterprises, the Structural Business drawn with care. For a labour market with about Statistics or the Community Innovation survey 3.2 Mio ICT practitioner jobs in the Community, could provide more detailed information without improvements in measuring needs on e-skills will creating additional burden on respondents. be crucial for the future. In general, an improvement in the analysis of The ensuing discussion revealed the problem of European data at Eurostat level could be reached defining e-skills in terms of proper measurement, by better access to micro-data. The ongoing as formal education is not necessarily the main international cooperation between Eurostat, the way to acquire e-skills. Therefore the development OECD and academia should be strengthened. of specific curricula for formal degrees on e-skills Current data collections should be adapted to the may not be the adequate tool. In addition, these needs, as should be international classifications degrees may quickly be outdated in times of such as ISCO. globalisation with a growing number of specialists coming from outside Europe and an accelerated Finally, policy should be more specific in change of technology which emanating new formulating their needs in measurement in order standards and demands. to get well tailored statistics in due time.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 83 The economic Impacts of ICT – lessons learned and new challenges

Dirk Pilat1 Directorate for Science, Technology and Industry, OECD

Executive Summary its impacts are particularly important in the services sector. Introduction – why we are interested in the economic impacts of ICT. Which statistics do we have to underpin empirical analysis on ICT and economic growth: What we have learned from work over the past • Statistics on ICT investment help 5-10 years: examine the contribution of ICT to • The impacts of ICT come through capital formation. different channels, notably through • Statistics on ICT production help ICT investment, production and use. examine the role of the ICT- • The impacts of ICT investment and producing sector, both as concerns use are sizeable in some countries, ICT manufacturing, as well as such as Australia, United States and telecommunications and computer Sweden, and are relatively small in services. others, such as Germany and Italy. • Statistics on ICT use, which help us The impacts of ICT production are look at the ICT technologies that firms concentrated in a few countries with use, the ways in which they are using a large manufacturing capability, e.g. ICT, and the impacts this has on firm Ireland and Korea. performance. ICT is no panacea, it is one factor in • • Statistics on other firm-level factors, a range of firm-level factors that can such as innovation, organisational improve performance. These changes factors, skills and firm age can be used include innovation, improvements to examine other characteristics of in skills and organisational changes. firms that use ICT. Not all firms succeed in making these changes. New firms can play a particularly important role in introducing ICT and the innovations Which challenges are currently faced by that may accompany it. statisticians and analysts in this field? • While ICT is a general-purpose • ICT is now broadly available to all technology and relevant to all firms, firms – to explain why some firms

1 This paper draws on previous OECD work which was carried out in collaboration with researchers and analysts in several OECD countries (see OECD (2004) and Pilat (2004)). The paper reflects the view of the author and not necessarily those of the OECD or its member countries. Comments are welcome: dirk.pilat@oecd.org

84 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY benefit from ICT and others do not, we of mobile telephony and of broadband networks have to look beyond the technology. We all demonstrate how pervasive this technology need to know about firm behaviour, the has become. But how precisely does ICT affect precise e-business applications of ICT economic growth and the behaviour of firms? What by firms, and complementary factors are the conditions under which ICT can become a if we want to understand impacts. technology that is effective in enhancing growth and productivity? Why have some countries and • Technological changes are affecting our ability to infer use and location regions thus far benefited more from ICT than and complicate classification. National others? To what extent do measurement issues surveys may not cover all service still pose a problem in quantifying the impacts of suppliers in domestic markets. ICT? What have we learned thus far about these Statistics will have to keep pace with questions and what are some of the puzzles that rapid technological changes, including still need to be resolved? new technologies that are coming into the market place. Despite the downturn of the economy over the past few years and the passing of the Internet • Much use of ICT is in the services bubble, these questions remain important to sector, where output measurement is academics, statisticians and policy makers. This poor, and impacts may be intangible. is because ICT has become a fact of (economic) Measuring e-finance or other digital life in all OECD economies. Almost all firms now services is challenging. use computers and most of them have an Internet connection. Moreover, a large share of these firms What can be done, who should do it? use computer networks for economic purposes, such as the buying, selling and outsourcing of • Keep statistical surveys and measurements up to date. But better goods and services. But despite the widespread ICT measurement alone is insufficient diffusion of ICT, questions remain about the impact to improve our understanding. (who: of the technology on economic performance and OECD, Eurostat, NSOs) behaviour. Thus far, only few countries, including Australia, Canada and the United States, have • Improve the ability to link data from clearly seen an upsurge in productivity growth in different surveys in NSOs (who: those sectors of the economy that have invested NSOs) most in the technology, notably services sectors • Use the existing data better, e.g. such as wholesale trade, financial services and by developing indicators that are business services. In many countries, including more meaningful for analysts and much of the European Union, these impacts have policy makers (who: NSOs, OECD, yet to become visible in the productivity statistics. Eurostat) Improving the understanding of the ways in which ICT affects economic behaviour and the factors • Engage in more empirical work that influence the potential impacts of ICT thus with micro data (who: OECD, EC, remains important. academics, NSOs).

Introduction This paper first briefly discusses the measures that are available to examine the economic Information and communications technology (ICT) impacts of ICT. Next, it summarises some of the has proven to be the key technology of the past main findings on the impacts of ICT. It primarily decade. The widespread diffusion of the Internet, focuses on the impacts of ICT on growth and

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 85 productivity, partly since these are particularly particularly problematic since countries vary in hard to measure; it does not discuss other how much total software spending is counted economic impacts of ICT, such as effects on as investment. Measuring software has been the wages, employment or trade. Next, it discusses subject of an OECD/Eurostat Taskforce that has the challenges that are currently being faced to produced a range of recommendations to improve improve our understanding of ICT’s contribution measurement (Lequiller, et al., 2003). to economic growth and makes recommendations on actions that could be taken to make further One other important element for aggregate progress in measurement and analysis. measures of ICT’s contribution to GDP growth is having the appropriate price deflators for ICT investment. These should ideally adjust for What data are available? rapid quality change in ICT products; i.e. the In empirical analysis of economic growth, so-called hedonic deflators (Triplett, 2004). To three effects of ICT are typically distinguished. address problems of international comparability, First, investment in ICT contributes to capital empirical studies often use US hedonic deflators deepening and therefore helps raise labour to represent price changes in other countries. This productivity. Second, rapid technological progress is only a second-best solution as countries should in the production of ICT goods and services may ideally develop deflators that properly account contribute to growth in the efficiency of capital for quality change of ICT products in their own and labour, or multifactor productivity (MFP), in national context. A particularly important area the ICT-producing sector. And third, greater use is hedonic deflators for software investment; of ICT throughout the economy may help firms currently, the United States is one of the few increase their overall efficiency, thus raising MFP. OECD countries to use hedonic deflators for pre- Moreover, greater use of ICT may contribute to packaged software. lower transaction costs and more rapid innovation, which could also improve MFP. The second approach to measuring the impacts of ICT is based on industry level data. It has These impacts can be examined at different levels typically involved analysis of the ICT-producing of analysis, i.e. with macro-economic data, with sector and of industries that are highly intensive industry data or with data at the level of individual in their use of ICT, as any impacts of ICT, such as firms or establishments. The measurement of increases in productivity growth, would possible the economic impacts of ICT at the aggregate first emerge in such industries. The ICT-producing level is relatively straightforward and has been sector is of interest for several countries, as it has outlined in detail in Schreyer et al. (2003). It is been characterised by high rates of productivity based on growth accounting, which involves the growth, providing a considerable contribution estimation of the productive capital stock on the to aggregate performance. The sector has been basis of measures of ICT investment, followed by defined in official statistics (OECD, 2005). the estimation of the capital services flowing from Examining the contribution of this sector to that stock. The method can be applied at both the aggregate productivity growth is relatively macro-economic and industry level, providing straightforward, although it involves some the appropriate data are available. One important measurement challenges that are discussed below. challenge for this approach concerns the basic Several studies have also measured the role of data; measures of ICT investment at the aggregate ICT-intensive sectors to aggregate productivity level are not available for all OECD countries and growth (Pilat and Wölfl, 2004; Van Ark and when they are, they are not necessarily comparable Inklaar, 2005). While there is no uniformly across countries. Data on software investment are accepted way to distinguish ICT-intensive sectors,

86 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY many studies have examined the contribution several cases, some of the firm-level data are of services sectors that are highly intensive in available over long time periods. Several countries their use of ICT, such as finance, insurance and have recently established such databases and business services. centres for analytical studies with firm-level data. Examples include Australia, Canada, Finland, The third source for analysis of the impacts of France, the Netherlands, the United Kingdom ICT concerns firm-level data; such data are and the United States. The data integrated in these now available for many OECD countries. Most firm-level databases differ somewhat between of the early work with firm‑level data on ICT countries, since the underlying sources are not the and productivity was based on private data. same. However, many of the basic elements are For example, Brynjolfsson and Hitt (2003) common. The basic sources for such databases are examined more than 600 large US firms over the typically production surveys or censuses. These 1987‑94 period, partly drawing on the Compustat data typically cover the manufacturing sector, database, while Bresnahan, Brynjolfsson and although firm-level databases increasingly cover Hitt (2002) examined over 300 large US firms (parts of) the service sector as well. from the Fortune‑1000 database. Similar studies with private data exist for other countries. In recent years, such firm-level databases have Studies based on such private data have helped increasingly started to incorporate data on to generate interest in the impacts of ICT on business use of ICT, e.g. those deriving from the productivity and have given an important impetus OECD/Eurostat model surveys on business use to the development of official statistics on ICT. of ICT. Firm-level studies of ICT’s impact on However, private sources suffer from a number economic performance require that researchers of methodological drawbacks. First, private data and statisticians link data for the same firms are often not based on a representative sample of derived from different statistical surveys, e.g. data firms, which may imply that the results of such from a production survey and from a survey on studies are biased. For example, studies based ICT use. Other types of data can be integrated too, on a limited sample of large firms may be biased which is important since empirical studies often since large firms may benefit more from ICT than suggest that the impact of ICT depends on a range small firms. Moreover, studies based on a limited of complementary investments and factors, such sample of firms will tend to ignore dynamic effects, as the availability of skills, organisational factors, such as the entry of new firms or the demise of innovation and competition (OECD, 2003). existing firms, which may accompany the spread of ICT. Second, the quality and comparability of Unlike the analysis of economic impacts of ICT private data are often not known, since the data do at the aggregate and industry level, analysis at the not necessarily confirm with accepted statistical firm-level is characterised by a wide range of data conventions, procedures and definitions. and methods (Table 1). This variety is partly linked to differences in the basic data, but also reflects Over the past decade, the analysis of firm- that a wide range of methods can be applied to level impacts of ICT has benefited from the firm-level data. To some extent, this variety is establishment of firm-level databases in statistical desirable, since the empirical evidence on impacts offices. These databases cover much larger and is stronger when it can be confirmed by different statistically representative samples than private methods. data, which is important given the enormous heterogeneity in plant and firm performance. On the other hand, cross-country comparisons These databases typically include data from require common methods and comparable data. several firm-level surveys that are linked. In Some researchers have recently engaged in

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 87 Table 1. Approaches followed in some recent firm-level studies of ICT and economic performance

Study Countries Survey covering ICT Method Economic Impacts Survey of Swiss business Labour productivity Labour productivity & Arvanitis (2004) Switzerland sector regressions complementarities US Computer Network Labour productivity (United Denmark, Japan, United Usage Survey, Denmark Labour productivity Atrostic, et al. (2004) States, Japan), Multi-factor States survey of ICT use, Japan regressions productivity (Japan) survey of IT workplaces Baldwin and Sabourin Survey of Advanced Labour productivity & Market share, labour Canada (2002) Technology market share regressions productivity Labour productivity and TFP Labour productivity, TFP, Clayton, et al. (2003) United Kingdom ONS e-commerce survey regressions price effects Crepon and Heckel (2000) France BRN employer file Growth accounting Productivity, output Criscuolo and Waldron Annual Respondents Labour productivity United Kingdom Labour productivity (2003) Database regressions IT adoption, e-commerce, De Gregorio (2002) Italy Structural business survey Multivariate analysis organisational aspects De Panniza, et al. (2002) Italy E-commerce survey Principal components Labour productivity Labour productivity and Doms, Jarmins and Klimek Asset and Expenditure Labour productivity, United States establishment growth (2002) Survey establishment growth regressions Business longitudinal survey, Labour productivity Labour productivity, MFP, IT Gretton, et al. (2004) Australia IT Use Survey regressions adoption US Computer Network Labour productivity Haltiwanger, et al. (2003) Germany, United States Usage Survey, German IAB Labour productivity regressions establishment panel Sales, contribution of ICT Regressions based on Hempell (2002) Germany Mannheim innovation panel capital, innovation, labour production function productivity Value added, contribution Innovation surveys, Regressions based on Hempell, et al. (2004) Germany, Netherlands of ICT capital, innovation, structural business statistics production function labour productivity Survey of Swiss business Hollenstein (2004) Switzerlan Rank model of ICT adoption ICT Adoption sector Labour productivity Maliranta and Rouvinen Internet use and E-commerce Labour productivity, Finland regressions, breakdown of (2004) survey productivity decomposition productivity growth Enterprice survey of Malmquist indexes of TFP Milana and Zeli (2004) Italy economic and financial TFP growth growth, TFP correlations accounts Basic survey on business structure and activities Production function, TFP Motohashi (2003) Japan Output, TFP, productivity (BSBSA); ICT Workplace regressions Survey Source: See references and OECD (2003; 2004). cross-country comparisons (e.g. Atrostic, et al., What we have learned about ICT and 2004; Hempell, et al., 2004; Haltiwanger, et al., economic growth 2003), and the methods used in these studies are increasingly also being adopted by other countries. ICT investment has boosted economic growth For example, the approach followed by Atrostic, et al. (2004) was also applied by Criscuolo and Investment in ICT can make an important Waldron (2003), and, to some extent, by Gretton, contribution in labour productivity growth. et al. (2004). Investment expands and renews the existing

88 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY capital stock and enables new technologies of the components of ICT investment are likely to enter the production process. While some to affect these estimates. countries have experienced an overall increase in the contribution of capital to growth over the ICT -using services have only experienced past decade, ICT has typically been the most more rapid growth in some OECD countries dynamic area of investment (e.g. Van Ark, et al., 2003; Schreyer, et al. 2003). Studies Several studies have also been undertaken at the show that ICT investment contributed to GDP industry level (O’Mahony and Van Ark, 2003; growth in most OECD countries in the 1990s, Inklaar, et al., 2003; Pilat and Wölfl, 2004). These accounting for between 0.35 and 0.9 percentage show that the ICT-producing manufacturing sector points of growth in GDP over the period contributed substantially to labour productivity 1995‑2003 (Figure 1).2 In all countries but and MFP growth in certain OECD countries Finland, investment in ICT hardware accounted such as Finland, Ireland and Korea (Figure 2),

Figure 1. The contributions of ICT capital to GDP growth, 1995-20032 In percentage points �

��

�� Source: OECD, OECD Compendium of Productivity Indicators 2005, Paris.

for the bulk of the contribution of ICT capital. and that the United States benefited more from In several countries, notably Denmark, France, the ICT-producing manufacturing sector than the Netherlands, Sweden and the United States, the European Union (O’Mahony and Van Ark, investment in software accounted for one- 2003). They also show that ICT-using services third of the total contribution of ICT capital in the United States and Australia experienced to GDP growth. In Finland, investment in an increase in productivity growth in the second communications equipment was the most half of the 1990s (Figure 2), which seems important component of ICT’s contribution to partially associated with their use of ICT. Few GDP growth. Differences in the measurement other countries have thus far experienced similar

2 A large number of studies of ICT investment and impacts at the industry level are available at the national level. These are not examined here; several are summarised in OECD (2003).

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 89 Figure 2. Contributions to aggregate labour productivity growth1 In percentage points

� ��

��

�

��

1 Annual average contributions to the growth of value added per person employed. The residual reflects adding up differences in aggregating from industry to the aggregate economy level. ICT-producing includes ICT manufacturing and ICT-producing services; ICT-using services include wholesale and retail trade, finance, insurance and business services. Source: OECD STAN database, see Pilat and Wölfl (2004).

productivity gains in ICT-using services (OECD, data may help in understanding why investment 2003). Moreover, most studies show that the in ICT has not yet led to greater productivity European Union lags behind the United States in impacts, as it can point to factors influencing the experiencing an increase in productivity growth impacts of ICT that can not be observed at the in ICT-using services (O’Mahony and Van Ark, aggregate level, e.g. organisational factors or the 2003; OECD, 2004; Van Ark and Inklaar, 2005). availability of skills. Firm-level data can also point to dynamic and competitive behaviour that may accompany the spread of ICT, such as the entry of Firm-level evidence adds important new new firms, the exit of firms that failed, and changes insights in market share of existing firms. Confronting firm-level and more aggregate evidence may thus The aggregate and industry-level evidence enhance our understanding of the ways in which provides helpful insights in the impacts of ICT ICT affects productivity and can contribute to on productivity, but also raises many questions, solving some of the questions that still surround notably as regards the conditions under which the impacts of ICT on productivity. ICT investment becomes effective in enhancing productivity. Moreover, the aggregate and ICT has positive impacts on firm performance industry-level evidence points to very limited productivity impacts of ICT in many countries, While aggregate and industry-level evidence is despite substantial investment in ICT. Firm-level often inconclusive about the impacts of ICT, firm-

90 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY level studies show that ICT use has a positive on productivity growth, but primarily through impact on firm performance in almost all OECD electronic purchasing, not through selling. This countries. For example, a study with Australian result confirms that networks can help firms firm-level data (Gretton et al. 2004) found that improve the management of their supply chain. the use of computers has a positive effect on MFP growth in the mid-1990s, i.e. before the peak Firm-level evidence also demonstrates that in ICT investment, with considerable variation ICT can affect the performance of the services across industries. A series of recent studies for the sector. For example, Doms, Jarmin and Klimek United Kingdom (Clayton, 2005) finds significant (2002) showed that growth in the US retail returns to ICT investment, higher in services than sector involved the displacement of traditional in manufacturing. These studies are confirmed by retailers by sophisticated retailers introducing many others (see OECD, 2003; 2004). For example, new technologies and processes. For Germany, firms using ICT typically pay higher wages. In Hempell (2002) showed significant productivity addition, the studies often show that the use of ICT effects of ICT in firms in the German service does not guarantee success; many of the firms that sector. A comparative study for Germany and the improved performance thanks to their use of ICT Netherlands (Hempell et al. 2004) found that ICT were already experiencing better performance than capital had a significant impact on productivity the average firm. Moreover, the benefits of ICT in the Netherlands’ services sector. For Australia, appear to depend on sector-specific effects and are Gretton et al. (2004) found positive impacts of not found in equal measure in all sectors. ICT use on labour and MFP growth in several services sectors, in both industry- and firm‑level The studies also suggest that some ICT analysis. And for the United Kingdom, Clayton technologies may be more important to strengthen (2005) showed that the productivity impacts of firm performance than others. Computer networks IT investment and communications are larger in may be particularly important, as they allow services than in manufacturing. a firm to outsource certain activities, to work closer with customers and suppliers, and to better The evidence summarised above suggests that integrate activities throughout the value chain. the use of ICT does have positive impacts on These technologies are often considered to be firm performance, but primarily, or only, when associated with network or spill-over effects. In accompanied by other changes and investments. recent years, more data have become available on Some early studies on the rates of return to ICT this technology. For the United States, Atrostic investment suggested that the returns to ICT were and Nguyen (2002) found that average labour relatively high compared to other investments in productivity was higher in plants with networks fixed assets. This is now commonly attributed and that the impact of networks was positive and to the fact that ICT investment is accompanied significant after controlling for several production by other expenditures, which are not necessarily factors and plant characteristics. Networks were counted as investment (Brynjolfsson and Hitt, estimated to increase labour productivity by 2003). This includes expenditure on skills and roughly 5%. Atrostic et al. (2004) also provided organisational change. This is also confirmed evidence for Japan and found that both inter- by many empirical studies that suggest that firm and intra-firm networks were correlated ICT primarily affects firms where skills have with higher MFP levels in firms. Open networks, been improved and organisational changes such as the Internet, as well as EDI networks, have been introduced. Another important factor were particularly important. For the United is innovation, since users often help make Kingdom, Criscuolo and Waldron (2003) found investment in technologies, such as ICT, more that the use of networks had an important impact valuable through their own experimentation and

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 91 invention. Without this process of “co-invention” that the greatest benefits from ICT are realised (Bresnahan and Greenstein, 1996), which often when ICT investment is combined with other has a slower pace than technological invention, organisational changes, such as new strategies, the economic impact of ICT may be limited. new business processes and practices, and new Some further findings on these complementary organisational structures. The common element factors are discussed below. among these practices is that they entail a greater degree of responsibility of individual workers Effective use of ICT requires appropriate skills regarding the content of their work and, to some extent, a greater proximity between management Many firm-level studies confirm the and labour. Because such organisational change complementarity between technology and skills. tends to be firm-specific, empirical studies show A study for Canada (Baldwin, et al., 2004), for on average a positive return to ICT investment, example, found that a management team with a but with a large variation across organisations. focus on improving the quality of its products by adopting an aggressive human-resource Several studies on organisational change are strategy – by continuously improving the skill of its available for European countries. For Germany, workforce through training and recruitment – was Falk (2001) found that the introduction of ICT associated with higher productivity growth. For and the share of training expenditures were Australia, Gretton et al. (2004) found that the important drivers of organisational changes, such positive benefits of ICT use on MFP growth were as the introduction of total quality management, typically linked to the level of human capital lean administration, flatter hierarchies and and the skill base within firms, as well as firms’ delegation of authority. For France, Greenan and experience in innovation, their application of Guellec (1998) found that the use of advanced advanced business practices and the intensity of technologies and the skills of the workforce were organisational change within firms. The data for both positively linked to organisational variables. Australia also showed that the earliest and most Organisations that enabled communication within intensive users of ICTs and the Internet tended to the firm and that innovated at the organisational be large firms with skilled managers and workers. level seemed more successful in the uptake of advanced technologies. Moreover, such For France, the data include details about worker organisational changes also increased the ability characteristics, which allow more detailed of firms to adjust to changing market conditions, analysis. Entorf and Kramarz (1998) found that e.g. through technological innovation and the computer-based technologies are often used by reduction of inventories. workers with higher skills. These workers became more productive when they got more experience Gretton, et al. (2004) on Australia also found in using these technologies. For the United significant interactions between ICT use and Kingdom, Caroli and Van Reenen (1999) found evidence that human capital, technology and complementary organisational variables in organisational change are complementary, and nearly all sectors. The complementary factors that organisational change reduced the demand for which data were available and which were for unskilled workers. found to have significant influence were: human capital, a firm’s experience in innovation, its use Organisational change is key to making ICT of advanced business practices and the intensity work of organisational restructuring. Computer use was also commonly associated with use of Closely linked to human capital is the role of advanced business practices, the incorporation organisational change. Studies typically find of companies and firm reorganisation. Arvanitis

92 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY (2004) found important complementarities for an important factor in driving a firm’s decision to Switzerland. He found that labour productivity is implement B2B electronic commerce. positively correlated with human capital intensity and also with organisational factors such as team- A closely related issue is that of experimentation. work, job rotation and decentralisation of decision This was raised in a recent comparison between making. the United States and Germany (Haltiwanger et al. 2003), that examined the relationship Maliranta and Rouvinen (2004) found some between labour productivity and measures of the evidence of complementarities for Finland, choice of technology. The study distinguished notably for human capital and organisational between different categories of firms according factors. Organisational factors appear important to their total level of investment and their level in Finland since the productivity effects of ICT of investment in ICT. It found that firms in all in the manufacturing sector seem to be much categories of investment had much stronger larger in younger than in older firms. Some productivity growth in the United States than other studies have shown that the productivity in Germany. Moreover, firms with high ICT of capital (primarily non-ICT) tends to be higher investment had stronger productivity growth in older plants, which is possibly due to learning than firms with low or zero ICT investment. The effects. While such learning effects undoubtedly study also found that firms in the United States also exist with ICT, the finding for Finland is had much greater variation in their productivity consistent with a view that it may be even more performance than firms in Germany. important to be able to make complementary organisational adjustments. Such changes are These differences may occur because US firms arguably more easily implemented in young or engage in much more experimentation than their new firms. German counterparts; they take greater risks and opt for potentially higher outcomes (see Competitive effects and the role of Bartelsman, et al., 2003). This may be related to experimentation differences in the business environment between the two regions; the US business environment In a competitive economy, the effective use of permits greater experimentation as barriers to ICT may help efficient firms gain market share at entry and exit are relatively low, in contrast to the cost of less productive firms, raising overall many European countries. Having scope for productivity. For example, Maliranta and Rouvinen experimentation may be an advantage in times of (2004) point to the role of firm selection in Finland. great technological uncertainty, when firms need While most of the increase in ICT use in Finland to learn in the market place about what works and is driven by growth within firms, restructuring what does not. The current period of ICT-driven (the growth of some firms and decline of others) growth might be such a period. also plays an important role. This is notably the case among young firms, where some succeed and ICT use is closely linked to innovation grow, and many others fail. Several studies point to an important link between Several other studies also point to the role of the use of ICT and the ability of a company to competition and of foreign firms. A recent study innovate. The role of innovation was raised by by Clayton (2005) found that subsidiaries of US Bresnahan and Greenstein (1996), who argued multinationals in the United Kingdom received that users help make investment in technologies, the largest productivity gains from the use of such as ICT, more valuable through their own computers. For Germany, Bertschek and Fryges experimentation and invention. Without this (2002) found that international competition was process of “co-invention”, which often has a slower

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 93 pace than technological invention, the economic orientation on high-technology is often the core impact of ICT may be limited. For example, work of a successful firm strategy. The study also finds for Germany, based on innovation surveys found that firms that combined ICT with other advanced that firms that had introduced process innovations technologies do better than firms that only use one in the past were particularly successful in using technology. Furthermore, the results emphasise ICT (Hempell, 2002); the output elasticity of ICT that combinations of technologies that involve capital for these firms was estimated to be about more than just ICT are important. For example, 12%, about four times that of other firms. This adoption of advanced process control technology, suggests that the productive use of ICT is closely by itself, has little effect on the productivity linked to innovation, notably process innovation. growth of a firm, but when combined with ICT and advanced packaging technologies, the effect Hempell, et al. (2004) points to the is significant. Similar effects are evident when complementarity of innovation and ICT for both firm performance is measured by market-share Germany and the Netherlands. They test the growth instead of productivity growth. hypothesis that firms that introduce new products, new processes or adjust their organisational Firm size affects the impact of ICT structure can reap higher benefits from ICT investment than firms that refrain from such A substantial number of studies have looked at the complementary efforts. For both countries, the relationship between ICT and firm size, notably results indicate that ICT is used more productively as regards differences in the uptake of ICT by if it is complemented by a firm’s own efforts to size of firm. This question has been addressed in innovate. These spill-over effects are a particular a large number of studies, most of which find that feature of ICT capital, since no complementarities the adoption of advanced technologies, such as between non-ICT capital and innovation could ICT, increases with the size of firms and plants. be found in the study. The results also show that innovating on a more continuous basis seems to Evidence for the United Kingdom, with 2000 pay off more in terms of ICT productivity than data for a variety of network technologies used innovating occasionally. This effect is found for in different combinations, shows that large firms product innovations (Germany) and non-technical of over 250 employees are more likely to use innovations (Netherlands) and, to a much smaller network technologies such as Intranet, Internet or extent, for process innovations. For Germany, EDI than small firms; they are also more likely to the study also finds evidence for direct benefits have their own Web site (Clayton and Waldron, from product and process innovation in services 2003). However, small firms of between 10 and on multi-factor productivity (MFP). Service firms 49 employees are more likely to use Internet as that innovate permanently show higher MFP their only ICT network technology. Large firms levels. This positive direct effect of innovation are also more likely to use a combination of on productivity, however, cannot be found for the network technologies. For example, over 38% of Netherlands. all large UK firms use Intranet, EDI and Internet, and also have their own Web site, as opposed to Baldwin, et al. (2004) finds that such less than 5% of small firms. Moreover, almost characteristics are also important in Canada. 45% of all large firms already used broadband The innovation strategy of a firm, its business technologies in 2000, as opposed to less than 7% practices, and its human-resource strategies all of small firms. influence the extent to which a firm adopts new advanced technologies. A central theme emerging These differences are partly due to the different from the Canadian evidence is that a strategic uses of the network technologies by large and

94 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY small firms. Large firms may use the technologies further technical developments can set further to redesign information and communication productivity-enhancing processes in motion. flows within the firm, and to integrate these flows throughout the production process. Some small Challenges and the way forward firms only use the Internet for marketing purposes. Moreover, skilled managers and employees often The evidence from the work summarised help in making the technology work in large firms above suggests that turning investment in ICT (Gretton et al. 2004). into stronger business performance is not The impacts of ICT use often only emerge over straightforward. It typically requires time complementary investments and changes, e.g. in human capital, organisational change and Given the time it takes to adapt to ICT, it should innovation. Moreover, ICT-related changes are not be surprising that the benefits of ICT may only part of a process of search and experimentation, emerge over time. This can be seen, for example, where some firms succeed and grow and others in the relationship between the use of ICT and the fail and disappear. Countries with a business year in which firms first adopted ICT. Evidence environment that enables this process of creative for the United Kingdom shows that among the destruction may be better able to seize benefits firms that had already adopted ICT in or before from ICT than countries where such changes are 1995, close to 50% bought using electronic more difficult and slow to occur. commerce in 2000 (Clayton and Waldron, 2003). For firms that only adopted ICT in 2000, less than A number of challenges also emerge from the 20% bought using e-commerce. The evidence current state of work. The greatest challenge presented by Clayton and Waldron suggests is further improving our understanding of the that firms move towards more complex forms economic impacts of ICT. Solid evidence on of electronic activity over time; out of all firms this issue, and on the conditions under which the starting to use ICT prior to 1995, only 3% had not economic impacts of ICT occur, is important in yet moved beyond the straightforward use of ICT underpinning evidence-based policy formulation. in 2000. Most had established an Internet site, or A few steps can be taken in this regard. bought or sold through e-commerce. Out of the firms adopting ICT in 2000, over 20% had not yet Improve the measurement of ICT impacts at gone beyond the simple use of ICT. the aggregate and industry level

The role of lags also emerges from analysis for While aggregate and industry-level measures Australia. Gretton et al. (2004) used firm level provide little insight in the underlying factors that information on productivity growth and the drive ICT’s role in the economy, such measures duration of computer use to examine the dynamics can point to cross-country differences that provide of the impact of the introduction of computers. helpful insights for policy makers. Improving They found that computers had a positive effect measurement at the aggregate level will primarily on MFP growth that varied between industries require more comparable investment data, as and that the positive effect was largest in the well as a greater use of quality-adjusted deflators, earlier years of uptake but appeared to taper off including for software investment. as firms returned to ‘normal’ growth after the productivity boost of the new technology. This Industry-level analysis primarily requires indicates that the ultimate productivity effect from improved output measures for service industries. adoption of ICT is a step up in levels, rather than a This is because much use of ICT is in the services permanent increase in the rate of growth. However, sector, where output measurement is poor, and the

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 95 impacts of ICT may be intangible. There is little from analysis with these surveys are used to agreement, for example, on the output of banking, inform the revision of these surveys. Clayton insurance, medical care and retailing. In addition, (2005), for example, points to several areas where some services are not sold in the market, so it is hard improvements in existing measures can be made. to establish prices. In practice, these constraints The other challenge for firm-level measures of ICT mean that output in some services is measured use is technological change itself.4 ICT continues on the basis of relatively simple indicators. While to be a constantly changing technology. The some new approaches to measurement in these convergence of different ICT technologies, for sectors are being developed, only few countries example, makes it difficult to infer the particular have thus far made substantial changes in their use of technology. Indeed, in many cases users official statistics to improve measurement. may use specific devices in ways that were not intended. Moreover, many services have become Statistical offices are well aware of the need to independent of any particular platform. In other improve output and productivity measurement cases, e.g. telephony, national surveys may no at the aggregate and industry level, as can be longer cover all service suppliers in domestic noted in the involvement of statistical offices in markets. In these – and other cases – it will be several EU countries in the EUKLEMS project important that statistical offices try keep up with on productivity measurement, and in the active the key technological changes. participation of many countries in a recent OECD workshop on Productivity Measurement.3 This 2. Improve the ability to link data from different work is therefore primarily the task for NSOs, surveys. The firm-level studies presented above the OECD, Eurostat as well as key academic all show that ICT is part of a broader set of researchers working in this field. behavioural changes in companies, including changes in skills, organisation and innovation. Strengthen the use of firm-level data for Examining the interaction between ICT and economic analysis such factors therefore requires that other firm- level sources (e.g. innovation surveys or surveys The largest potential for further analysis of ICT’s of organisational change) are analysed in impacts on economic performance and behaviour combination with surveys of ICT use. Engaging lies with firm-level data, however. This is where in such analysis thus requires that statistical many of the key insights regarding ICT’s impact offices are able to link these surveys, which on economic performance and firm behaviour typically requires a common business register. have been gained and where much further Moreover, it will be important to consider whether progress can be made. Making further progress the key complementary factors to ICT, such as in this area will require various steps: innovation, organisational factors and skills, are appropriately reflected in existing firm-level data 1. Keep ICT measurement up to date. To allow sources and that these can be linked to the other meaningful analysis of firm-level measures, it is firm-level data. important to keep firm-level measures up to date. This implies that existing surveys of ICT use 3. Develop more meaningful indicators from ICT should continue to be updated to ensure that they surveys. Firm-level data are currently typically are still relevant. This also implies that findings aggregated to indicators that can be compared

3 See: www.oecd.org/statistics/productivity 4 These points are based on a presentation by Sam Paltridge of the OECD to the Working Party on Indicators for the Information Society, in April 2005.

96 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY across countries, e.g. indicators of ICT use by 5. Engage in more international comparisons. industry or by firm size. Some of these aggregate Cross-country studies on the impact of ICT at indicators may not be as helpful for analysts and the firm level are still relatively scarce, primarily policy makers as they could be. Developing more since comparable data sources are still relatively suitable indicators for international comparisons new. Some studies have recently engaged in from the firm-level data is a challenge that is international comparisons (Atrostic et al., 2004; being examined by several statistical groups in Hempell et al., 2004; Haltiwanger et al., 2003). the OECD. For example, an indicator that was Understanding the reasons for the cross-country constructed by the UK Office of National Statistics differences in the impacts of ICT reported in such for the OECD project on ICT and economic studies would benefit from further work, and could growth helped point to the link between the year lead to helpful insights for policy. The OECD and of ICT adoption and the degree of e-business the European Commission could work together to activity (Clayton and Waldron, 2003). make progress in this area, e.g. by assembling a consortium of interested countries along the lines 4. Address new questions with firm-level data.ICT of a previous OECD effort (OECD, 2004). is now broadly available to all firms – to explain why some firms benefit from ICT and others References do not, it will be necessary to look beyond the technology. More knowledge will be needed about Arvanitis, S. (2004), “Information Technology, firm behaviour, the precise e-business applications Workplace Organisation, Human Capital and Firm of ICT by firms, and complementary factors to Productivity: Evidence for the Swiss Economy”, help understand the impacts ICT. There are also in: OECD (2004), The Economic Impact of ICT several key issues that remain poorly analysed and – Measurement, Evidence and Implications, OECD, that offer scope for progress. For example, further Paris. work with firm-level data could provide greater insights into the contribution of firm dynamics Atrostic, B.K. and S. Nguyen (2002), “Computer to productivity gain from ICTs, e.g. the role of Networks and US Manufacturing Plant Productivity: new firms, the conditions that lead to successful New Evidence from the CNUS Data”, CES Working survival and the factors determining firm exit. Paper 02‑01, Center for Economic Studies, Moreover, the link between innovation and ICT Washington DC. has only been examined for some OECD countries. Atrostic, B.K., P. Boegh‑Nielsen, K. Motohashi and Understanding this link is of great importance as S. Nguyen (2004), “IT, Productivity and Growth in long-term growth largely depends on the future Enterprises: Evidence from New International Micro pace of innovation. In addition, while there is good Data”, in: OECD (2004), The Economic Impact of ICT evidence for some OECD countries that ICT can – Measurement, Evidence and Implications, OECD, help transform the service sector and make it more Paris. innovative and productive, a good understanding of ICT’s impact on the service sector is still lacking, Baldwin, J.R., D. Sabourin and D. Smith (2004), “Firm partly because of the measurement problems Performance in the Canadian Food Processing Sector: outlined above but also due to lack of cross- the Interaction between ICT Advanced Technology country empirical analysis. Analytical work with Use and Human Resource Competencies”, in: OECD (2004), The Economic Impact of ICT – Measurement, firm-level data has been undertaken by researchers Evidence and Implications, OECD, Paris, pp. 153- and statistical offices in many countries, but could 181. be supported by policy makers in this area and could benefit from joint efforts by the OECD and Bartelsman, E.J., S. Scarpetta and F. Schivardi (2003), the European Commission. “Comparative Analysis of Firm Demographics and

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 97 Survival: Micro-Level Evidence for the OECD Falk, M. (2001), “Organizational Change, New Countries”, OECD Economics Department Working Information and Communication Technologies and Paper No. 348, OECD, Paris. the Demand for Labor in Services”, ZEW Discussion Paper No. 01‑25, ZEW, Mannheim. Bertschek, I. and H. Fryges (2002), “The Adoption of Business‑to‑Business E‑Commerce: Empirical Greenan, N. and D. Guellec (1998), “Firm Evidence for German Companies”, ZEW Discussion Organization, Technology and Performance: An Paper No. 02‑05, ZEW, Mannheim. Empirical Study”, Economics of Innovation and New Technology, Vol. 6, No. 4, pp. 313‑347. Bresnahan, T.F. and S. Greenstein (1996), “Technical Progress and Co‑Invention in Computing and the Gretton, Paul, Jyothi Gali and Dean Parham (2004), Use of Computers”, Brookings Papers on Economic “The Effects of ICTs and Complementary Innovations Activity: Microeconomics, pp. 1‑77. on Australian Productivity Growth”, in: OECD (2004), The Economic Impact of ICT – Measurement, Bresnahan, T.F., E. Brynjolfsson, and L.M. Hitt (2002), Evidence and Implications, OECD, Paris. “Information Technology, Workplace Organization and the Demand for Skilled Labor: Firm‑Level Haltiwanger, J., R. Jarmin and T. Schank (2003), Evidence”, Quarterly Journal of Economics, Vol. 117, “Productivity, Investment in ICT and Market February, pp. 339‑376. Experimentation: Micro Evidence from Germany and the United States”, Center for Economic Studies Brynjolfsson, E. and L. Hitt (2003), “Computing Working Paper CES-03-06, US Bureau of the Census, Productivity: Firm-Level Evidence”, Review of Washington, D.C. Economics and Statistics, Hempell, T. (2002), “Does Experience Matter? Caroli, E. and J. Van Reenen (1999), “Organization, Productivity Effects of ICT in the German Service Skills and Technology: Evidence from a Panel of Sector”, Discussion Paper No. 02‑43, Centre for British and French Establishments”, IFS Working European Economic Research, Mannheim. Paper Series W99/23, Institute of Fiscal Studies, August. Hempell, T., G. Van Leeuwen and H. Van Der Wiel (2004), “ICT, Innovation and Business Performance in Services: Clayton, T. (2005), “IT Investment, ICT Use and UK Evidence for Germany and the Netherlands”, in: OECD Firm Productivity”, ONS, London, August. (2004), The Economic Impact of ICT – Measurement, Evidence and Implications, OECD, Paris. Clayton, T. and K. Waldron (2003), “E‑Commerce Adoption and Business Impact, A Progress Report”, Inklaar, R., M. O’Mahony and M. Timmer (2003), Economic Trends, No. 591, February, pp. 33-40. “ICT and Europe’s Productivity Performance – Industry-level Growth Account Comparisons with Criscuolo, C. and K. Waldron (2003), “Computer the United States”, Research Memorandum GD-68, Network Use and Productivity in the United Kingdom”, University of Groningen, December. Centre for Research into Business Activity and Office of National Statistics, mimeo. Lequiller, F., Ahmad, N., S. Varjonen, W. Cave and K.H. Ahn (2003), “Report of the OECD Task Force Doms, M., R. Jarmin and S. Klimek (2002), “IT on Software Measurement in the National Accounts”, Investment and Firm Performance in US Retail Trade”, Statistics Directorate Working Paper 2003/1, OECD, CES Working Paper 02‑14, Center for Economic Paris. Studies, Washington DC. Maliranta, M. and P. Rouvinen (2004), “ICT Entorf, H. and F. Kramarz (1998), “The Impact of and Business Productivity: Finnish Micro-level New Technologies on Wages: Lessons from Matching Evidence”, in: OECD (2004), The Economic Impact Panels on Employees and on their Firms”, Economic of ICT – Measurement, Evidence and Implications, Innovation and New Technology, Vol. 5, pp. 165‑197. OECD, Paris, pp. 213-239.

98 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY OECD (2003), ICT and Economic Growth – Evidence Schreyer, Paul, Pierre-Emmanuel Bignon and from OECD Countries, Industries and Firms, OECD, Julien Dupont (2003), “OECD Capital Services Paris. Estimates: Methodology and A First Set of Results”, OECD Statistics Working Papers 2003/6, OECD (2004), The Economic Impact of ICT – OECD, Paris. Measurement, Evidence and Implications, OECD, Paris.

OECD (2005), Guide to Measuring the Information Triplett, J.E. (2004), “Handbook on Hedonic Indexes Society, DSTI/ICCP/IIS(2005)6/FINAL, OECD, Paris. and Quality Adjustments in Price Indexes: Special Applications to Information Technology Products”, O’Mahony, M. and B. van Ark (2003) (eds.), EU STI Working Paper 2004/9, OECD, Paris. Productivity and Competitiveness: An Industry Perspective – Can Europe Resume the Catching Up Van Ark, B., J. Melka, N. Mulder, M. Timmer and Process?, European Community, Luxembourg. G. Ypma (2003), “ICT Investments and Growth Accounts for the European Union, 1980‑2000”, Pilat, Dirk (2004), “The ICT Productivity Paradox: Research Memorandum GD‑56, Groningen Growth Insights from Micro Data”, OECD Economic Studies, and Development Centre, Groningen. 2004/1, pp. 37-65.

Pilat, Dirk and Anita Wölfl (2004), “ICT production and Van Ark, B. and R. Inklaar (2005), “Catching Up ICT use – what role in aggregate productivity growth?”, or Getting Stuck? Europe’s Trouble to Exploit ICT’s in: OECD (2004), The Economic Impact of ICT – Productivity Potential”, Research Memorandum GD- Measurement, Evidence and Implications, OECD, Paris. 79, University of Groningen.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 99 Our Lives in Digital Times

by George Sciadas Statistics Canada

Excutive Summary This paper uses a variety of data to debunk myths, identify some realities, and arrive at reasonable Life in our times is irrevocably influenced by the inferences about ICT-induced behavioural deluge of ICTs. Over the last two decades or so, change. In the process, it demonstrates how we whether at work or our social lives, computers, cell can now start to use harder quantitative evidence phones and the Internet are gradually changing to separate what really happens from what does daily routines and altering behaviours. The early not happen. focus of policy and research on ICT access gave way to issues of use, as the next logical stage Among the myths perpetrated are: more closely associated with the derivation of benefits. All along, though, the ultimate interest The paperless office: The arrival of the personal has been to assess the outcomes and comprehend computer in the early 1980s and its speedy the longer term impacts associated with ICTs. diffusion, combined with the subsequent arrival of networks that made possible the electronic Efforts to understand the impacts of ICTs on capture, storage, display and transmission of the macroeconomy are well underway and, with documents, gave rise to much talk of a ‘paperless new datasets available, research is taking place office’ and, by extension, a ‘paperless society’. at the level of industry and firm performance. However, data reveal that the production and use While researching economic outcomes is well of paper has increased significantly over the last justified, there are many more outcomes of two decades and now stands at an all-time high. ICTs related to the social and other domains of people’s lives. Economic and social outcomes The death of mail: The arrival of the fax machine are interrelated. led to early arguments concerning the drastic reduction in mail. This was multiplied manifold By their nature, outcomes are subject to in the early to mid-1990s with the arrival of gradual evolution. To gain insights, both direct e-mail (in its commercial incarnation), with quantification and analytical inference through significant repercussions for the post office -a diverse and time-series data are needed. In the fixture in every country. Yet, notwithstanding the absence of data-backed research not only our truly meteoric explosion of faxes, e-mails, SMS understanding is incomplete but may well be and other electronic communications, data on the erroneous. Hypotheses typically postulated on volume of mail deliveries fail to validate the end the basis of little more than euphoric hype fuelled of transporting paper. We are nowhere close to by technological possibilities become allegedly the imminent demise of postal offices, and private ‘predictable’ outcomes, without being necessarily services proliferate. grounded on people’s actual behaviour. Their plausibility influences perceptions and, consciously The end of professional travel: Allegedly, e- or not, infiltrates, and possibly contaminates, mail, videoconferencing, collaborative tools decision-making. over the Web and the like would minimize the

100 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY need, and the expense, of transporting humans gives the semblance of being paperless. While to physical gatherings. Once again, while such the activity of transporting paper is still thriving, ‘telecommuting’ does occur, one would be hard- the operations of postal offices have changed as pressed to think of any other time in history has the composition of mail. With e-commerce, with more events such as business meetings, retail will never be the same, as consumers have conferences, symposia and workshops. never been more empowered. Therefore, ICTs have powerful and lasting influences, albeit The death of retail: In the early days of e- different than the ‘obvious’ ones during early commerce, last decade, much energy was consumed stages of deployment and use. By all accounts, on the detrimental effects of ‘clicks’ over ‘bricks’. we have not seen the end yet. Fears of unemployment, given the huge numbers involved in the retail industry and their relatively Further detailed analysis of relevant datasets lower qualifications, and fears of the disastrous reveals the following real outcomes: consequences entailed in the deflation of real estate assets were repeatedly voiced. At a time when People talk on the phone more than ever in e-commerce is steadily taking hold, grows and history. Over the course of the last twenty years, matures, data show that the number of retailers, usage of wireline networks alone has increased retail space, and employment in the industry have enormously. Both frequency of calls and talking all increased; so have real estate values – at a time have reached historical highs. In addition, growth that is stronger in recent years. we talk more than ever on wireless phones. The Information Society is also a ‘talkative society’. While the above premature announcements of deaths clearly imply causality from ICTs, and Then, we communicate more than ever through even contain grains of logic, quantification at massive amounts of e-mail and SMS, as well this point proves them faulty; they either did as spend considerable amounts of time on not materialize or the opposite has happened. computers, the Internet and other ICTs. The time Presumably, one can always argue that they will spent on such activities, which did not exist in the still happen – they just have not happened yet. A past, is in no way matched by the small decreases better explanation is to acknowledge that we live in the usage of older and more passive ICTs, such in a world of complex interactions, where many as radio and television. forces are at work simultaneously. It may well be that ICT forces are indeed pointing to the alleged Talking on phones and using newer ICTs absorbs directions, but all kinds of other influences, several more hours in daily routines today pointing opposite, dominate. In the very least, compared to twenty years ago. This outcome the ceteris paribus assumption (all other things certainly has consequences as it bumps against equal) does not hold. the inescapable 24-hour constraint. Where is this ‘extra’ time coming from? While specific answers On the other hand, numerous effects, other than await to be uncovered through detailed time-use those predicted, did occur. Beyond any doubt, data, certainly some activities are displaced, ICTs have brought about profound change in all the others curtailed and most likely we economize above areas, manifested in shifting behavioural on time by resorting to ‘multi-tasking’. Thus, patterns with real consequences. Although the added sense of being busier than ever, both at the paperless office is not here, gone are the work and at home. days of hand-writing or dictating to secretaries, a substantial amount of reading is done on the Time spent on the phone reflects both the number screen, and there is even the occasional office that of calls and their duration. Data show clearly

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 101 that calls have increased, presumably to more create new ones in our quest to truly understand people, existing associations or both. Not only ICT outcomes and impacts. This includes all the we communicate more than even, traffic data new ICT-related datasets that are being developed also reveal that within the significantly expanded in some countries and are starting to accumulate total the pattern of communications has changed in the time-series sense, but also extends to all noticeably as well. While all kinds of telephone other existing, seemingly unrelated, data sources. calls have increased drastically, as has time spent on Defining intelligent questions to ask, combined the phone, more telling is the rise of long distance, with imaginative integration of data, can go a long which outpaced local calls and, within that, the way towards serving policy and business needs, as explosion in international calling, both outgoing well as satisfying basic research curiosity. It is still and incoming. It would not be a far stretch to early, though, and we must proceed with caution. surmise that such a pattern would be even more pronounced in e-mail. This does not bode well 1. The context with elaborate theories of modern alienation, the breakdown of the social and community fabric, Life in our times is irrevocably influenced by and the decay in interpersonal relationships. While the deluge of information and communications there are several forces at work in our societies for technologies (ICTs). Over the last two decades some time now that may point to such a direction, or so, whether at work or in our social lives, the story of ICTs is rather different – in the very computers, cell phones and the Internet are least, it has several more angles and nuances that gradually changing daily routines and altering are definitely a big part of the story. behaviours. The early focus of policy and research on ICT access gave way to the next Today, ICTs and their falling prices seem to have logical stage, issues of use, which is more closely liberated unsatisfied urges which now can be associated with the derivation of benefits. All fulfilled. We may not talk to the person next to along, though, the ultimate interest has been to us, but we are talking to someone – who may be assess the outcomes and comprehend the longer next door or miles and time zones away. It is not term impacts associated with ICTs. that we are becoming anti-social, it is that we are SIDE BAR: ICTs represent a very long list of goods becoming differently social. We forge many more and services, including older – albeit revitalized new relations, regardless of location. This does not or transformed – and newer technologies whose necessarily make us increasingly isolated, loners, functionalities increasingly overlap under the remote from our friends and families or socially process of convergence. Those with the most inept – as some argue. On the contrary, it may be impact are relatively few and well-known; others that it enriches us in ways that were not possible are, sometimes literally, of a peripheral nature until now, as we expand the circle of associations and their usage is subsumed by others – for and the ‘proximity’ we have with people. People example, the printer and the computer. make the choice to expand their associations and move from geographically-defined communities Just about twenty-five years ago, the telephone to communities of interest. Data also show that was the most visible communications technology. people are willing to pay for those choices. As Its usage involved, exclusively, plain and simple only one consequence, we may travel more – two-way voice communication, not much different counter to the wisdom that we need to travel less. from its beginnings a century earlier. There were no personal computers, no cell phones, nor the Implicit throughout this exploration is the message Internet in any way that would be recognized that we need all kinds of data to perform reality today. These days, a visitor walking in several checks, explore linkages, test hypotheses and downtown cores will notice that people on the

102 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY street are ostensibly talking to themselves. A Economic and social outcomes are interrelated. closer look will reveal that these are not isolated Moreover, they are subject to gradual evolution. occurrences of eccentric or troubled individuals; To gain insights, both direct quantification and people are talking into microphones and listening analytical inference through diverse and time- through earpieces attached to cell phones, which series data are needed. It is only now, when data have become invisible, tucked away somewhere. have recorded the history of several years, that The visitor will also notice that parties of people, we can start making meaningful comparisons. In presumably close friends, walking together on the absence of data-backed research not only is sidewalks and maneuvering through intersections our understanding incomplete but it may well be with heavy traffic, are busy talking all along – but erroneous. This paper takes a long-term view not among themselves; instead, they are chatting and uses a variety of data sources to arrive at individually with someone else, somewhere. reasonable inferences on selected ICT-induced behavioural changes. Examples of such behaviour precipitate easy statements of the type that we are becoming 2. Myths and realities increasingly isolated, loners, remote from our friends and families, anti-social and the like. In recent years, much has been learned through These are followed by the elaboration of theories the quantification of ICT access and use (see, concerning modern alienation, the breakdown for instance, Statistics Canada 2003a). More of the social and community fabric, the decay competitive marketplaces, infrastructure in interpersonal relationships or the strangeness deployment and falling prices have contributed to of our neighbours, among others. While several high uptake of new ICTs1. As a result, access is no forces are at work in our societies that may point longer the dominant issue – at least in developed to such a direction, the story of ICTs is rather countries2. Usage takes place from many locations different – in the very least, it has several more and, as applications and uptake move in tandem, it angles and nuances that are undoubtedly a big becomes more diversified. E-mail has emerged as part of the story. an indispensable communications medium, while information and entertainment uses proliferate. Since ICTs have permeated the conduct of almost As well, e-commerce is taking hold - whether every economic and social activity imaginable, browsing for product characteristics and prices the outcomes and impacts worthy of examination or actually placing orders online. would make for a very long list. Efforts to understand the impacts of ICTs on the ‘macro’ In the process of such wholesale technological, economy are well underway (OECD 2003) and, economic and social transformations in with new datasets available, research is taking recent years, progress has also been made place at the level of industry and firm performance in both quantifying ICT-related change and (Clayton 2005). While researching economic appreciating the value of such information. outcomes is well justified, there are many more Nevertheless, it takes time to measure change outcomes of ICTs related to the social and other and find analytical approaches to use the data. domains of people’s lives (Kraut et al. 1998, In the meantime, many hypotheses have been Wellman et al. 2004, Nie and Erbring 2000). postulated, frequently passed on as conclusions.

1 In some instances penetration rates start to approach universality levels, especially in some geographically-defined communities. However, ICTs are still evolving and this is not the case for newer technologies (i.e. broadband). 2 We do know that a Digital Divide still exists even among advanced countries, part of which is related to incomes and some to many other factors. As well, we know that the gap between developed and developing countries is huge – but this does not come as a surprise (Orbicom 2003, 2005). Some progress is being made and technological leapfrogging is occurring.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 103 Now, harder quantitative evidence can be used the death of time, the death of traditional retail to see what really happens and what does not. and even the death of history (!) have been However, it is still early and we must proceed proclaimed. The emergence and rapid diffusion with caution. of new ICTs was no exception, but quite typical of such behaviour. It gave rise to several such myths, whose plausibility influences perceptions 2.1. Myths and potentially decision-making. Some of them Quite frequently, the early stages of important are taken on below. new developments are subject to euphoric hype coupled with impatience regarding prospective SIDEBAR: In the context of this paper, a ‘myth’ is possibilities, something that tends to move us not a seemingly plausible but intrinsically wrong ahead of ourselves. One manifestation of this prediction; rather, it is an expected outcome that is associated with premature inferences of is not borne out to be true or supported by factual allegedly ‘predictable’ outcomes, presented as evidence – so far. a ‘fait accompli’ - although not based on hard evidence, which is invariably non-existent at ➢ The paperless office those early stages. Such beliefs may then be engrained into people’s psyche and, consciously The arrival of the personal computer in the early or not, infiltrate, and possibly contaminate, 1980s and its speedy diffusion later in the decade, decision-making3. In their most pompous form, combined with the arrival of networks that made they are typically communicated as premature it possible to electronically capture, store, access, announcements of deaths! The death of distance, display, manipulate and transmit documents, gave

Table 1. Consumption of printing and writing paper

(metric tonnes) Canada 1983 1993 2003 production 1,726,000 4,194,000 6,457,000 imports 185,200 506,166 1,037,123 exports 713,100 2,117,000 4,626,681 consumption 1,198,100 2,583,166 2,867,442

United States 1983 1993 2003 production 15,405,000 21,511,008 20,304,502 imports 1,201,900 2,891,000 7,300,332 exports 167,300 1,017,000 1,286,954 consumption 16,439,600 23,385,008 26,317,880

World 1983 1993 2003 production 45,224,300 71,956,808 97,199,494

Note: Consumption is estimated as production plus imports minus exports. Source: FAOSTAT, Food and Agriculture Organization, 2004.

3 They do not become self-fulfilling prophecies, however, as we shall see.

104 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY rise to much talk of a ‘paperless office’ and, by and there are reports that e-mail alone has extension, a ‘paperless society’. significantly increased printing. Estimates for additional printing by businesses due to the use However, the production and use of paper of the Internet and e-mail range from 30%-35% products is at an all-time high. Data reveal that (Ivey Business Consulting Group 2003) to 40% consumption of paper for printing and writing (Sellen and Harper 2001) depending on enterprise alone has increased significantly over the last two size. If anything, “shred-it” businesses proliferate decades (Table 1). In Canada, consumption more like never before and printers have found their than doubled between 1983 and 2003 – with most way into people’s homes too. At this juncture, of the growth occurring during the first of the the digital era appears thirsty for paper. two decades. As the growth rate of consumption (139.3%) outstripped the rate of growth of the Surely, there are more parts to the story concerning population (23.6%), per capita consumption the relationship between ICTs and paper. There is increased by 93.6% to 91.4 kilograms in 2003. no doubt that ICTs have brought about numerous This is equivalent to almost 20,000 pages4 per behavioural changes – gone are the days of hand- individual, enough to cover an area of almost writing or dictating text, a substantial amount of 1,200 square metres. reading is done on the screen, and there is even the occasional office that gives the semblance of Per capita consumption in the US is comparable being paperless. There are also many reasons to that of Canada; but increased consumption is why paper is thriving, including its versatility and not confined to developed countries. Worldwide its physical properties of tangibility, portability production and consumption of paper also and others. The fact remains, however, that the more than doubled over the last two decades, paperless office is the office that never happened. with especially high growth in emerging Asian This leads to the related issue of transporting economies (particularly China, which absorbs a paper. significant amount of Canadian paper exports). Society’s addiction to paper is expected to ➢ The death of mail intensify; paper consumption is projected to continue to grow, and more so in developing The arrival of the fax machine more than two countries. According to the Forest Products decades ago gave rise to talk about the drastic Association of Canada (2004), growth over the reduction in mail, including the fall to relative next 15 years is forecast at 3.2% annually - 5.5% insignificance of the post office, a fixture in every for developing countries and 2.5% for developed country. This was multiplied manifold in the ones (Forest Products Association of Canada early to mid-1990s with the arrival of e-mail in 2004). its commercial incarnation.

Not only is the notion of a paperless society Indeed, as early as 1998 the International Labor defeated by existing data, but also by a visit to any Organization (ILO) noted: “It is five years since modern office workplace. Printers everywhere the number of international messages sent by spit out massive amounts of paper, paper bins fax took a bigger share of the market than those are full, and many offices look like fire hazards. conveyed by post. In 1996, for the first time, the Jokes concerning the printing of e-mails abound, volume of e-mail in the United States exceeded

4 Estimated on the basis of 216 mm x 279 mm (8 ½ x 11 inches) sheets of 50 lb thickness paper. These figures refer only to paper for printing and writing, not including newsprint and other types of paper products. In Canada, paper for printing and writing represented 21.2% of all paper products in 2003 – up from 20.5% in 2002 (Forest Products Association of Canada, Annual Review, 2003).

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 105 the number of letters delivered by the postal Table 3. Private couriers mail volume service” (p.12). Today, the volume of e-mail is many times higher and still growing enormously. year millions of More recently, electronic messaging through pieces cell phones and other handheld devices is also exploding in several parts of the world. In 1997 260 Finland, for example, four out of five mobile 1999 521 466 phone users (and there are many at 74% of the 2000 2001 485 population aged 15-74 in 2003) report sending 2002 501 text messages weekly, with 1% sending between 2003 489 100 and 200 messages! Nearly all those in the 15-29 age group sent text messages, including picture messages (Nurmela and Sirkiä 2004). Notes: Data refer only to large and medium couriers with annual revenues of $250,000 and up. Yet, despite the truly enormous rise in e-talk, Data for 1997 are not exactly comparable as they refer to carriers with revenues over data fail to validate the demise of transporting $150,000, while no data exist for 1998. paper. The volume of postal deliveries, both Source: Statistics Canada, Surface and Marine by the public and private sectors, has increased, Transport at the same time at which faxes are sent, and Service Bulletin, Catalogue No. 50-002-XIE. e-mails and text messages skyrocket. Canada Couriers and local messengers are proliferating Post’s volume is up over a long period (Table 2) today; so is their employment. Tables 4 and 5 – albeit marginally lately, and down from its show the industry’s firm demographics and peak in the mid 1990s5. Even though the rates of employment, respectively, for selected years growth are slower, or even if there was a decline, - during which Internet usage and e-mail were the trends of mail delivery are nowhere close high in Canada. Again, these are not signs of an to pointing to the imminent demise of postal industry in distress. offices. Considering the additional volume of half a billion pieces moved around annually by Therefore, the business of moving paper around private couriers and local messengers adds to the is alive and well. In large offices, there are still perspective, particularly if the growth of the last people whose job is to physically move paper few years is factored in (Table 3). around from one floor to the next. Once again it must be re-iterated that many things may have changed, including the operations of postal offices Table 2. Canada Post mail volume and the composition of mail, but not the activity

year billions of itself. For instance, personal mail is down, but pieces other types of mail make up for it. Moreover, 1983 6.6 operators have undergone serious reorganization 1993 10.4 and re-allocation of their activities. This includes 2003 10.7 the increasing use of franchising, and the fact Note: Data for 1983 and 1993 refer to fiscal years that they were forced into the provision of (April-March), while data for 2003 refer to the express mail services6 or e-mail (something more calendar year. Source: Canada Post Corporation, Annual Reports. pronounced in countries where there was already a split between post and telecommunications). As

5 While it increased every year until 95/96, it decreased in the next couple of years hitting a low in 97/98, but has since rebounded. 6 In 2003, next-day and overnight delivery services accounted for more than half of couriers’ revenues; 91% of local messengers’ revenues and pieces were for same-day delivery services (Statistics Canada, Surface and Marine Transport Service Bulletin, Catalogue No. 50-002-XIE).

106 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Table 4. Couriers and local messengers - demographics

1998 1999 2000 2001 2002 2003 number of firms Couriers 1,200 1,644 1,782 2,003 2,353 2,624 Local messengers and local delivery 10,121 16,276 16,357 16,960 17,339 17,888 Total 11,321 17,920 18,139 18,963 19,691 20,512

Source: Statistics Canada, Survey of the Couriers and Local Messengers Industry. Table 5. Couriers and local messengers - employment

1991 1993 1997 2000 2003 number of employees Couriers 28,892 30,494 33,433 33,532 34,770 Local messengers and local delivery 4,158 4,410 4,727 4,739 6,306 Total 33,050 34,904 38,160 38,271 41,076

Source: Statistics Canada, Survey of Employment, Payroll and Hours (SEPH). well, junk mail makes up some of the lost ground, desire to see colleagues and other places to take in parallel with the existence of spam and other one’s mind off the daily routine. The fact remains telemarketing activities. that ICTs have more likely increased rather than decreased such travel. ➢ The end of professional travel Even though there are definite advantages to The case of human transport, in conjunction with distant forms of communications and the need for physical gatherings, such as business teleconferencing, videoconferencing and Web meetings, conferences, workshops, symposia casting are indeed on the rise, they are still and the like is also worth exploring. E-mail, the small-scale, somewhat eclectic and seem to take Web and videoconferencing were said to lead to place in parallel rather than replacing physical the death of such movement as it would be more gatherings. practical and economical to tele-commute. This has not been the case. Although comprehensive ➢ The death of traditional retail and reliable data do not exist, the only dent in people’s transportation seems to have come Then we have the case of retail trade that, for from 9/11. Consider the anecdotal evidence many, was a cause for concern in the early days of non-stop executive travel and the volume of of e-commerce. Much has been written about the organized events today, not to mention comments potential detrimental effects on retail of the new way by professionals about the need to be cloned to to do commerce, including fears of unemployment, keep up with the demands for appearances, and given the huge numbers of workers involved in the the importance of personal interaction. All of industry and their relatively lower qualifications, as this makes it abundantly clear that we have never well as fears concerning the negative and profound had more movement, which probably reflects the consequences of possible deflation of real estate need to meet, and perhaps even more than before. assets, particularly in downtown cores. The reasons for such increased mobility may have changed; for example, they may not reflect It is true that although e-commerce started small the need for scientific substance, but rather the it is growing steadily and at a healthy rate – just

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 107 as retail is. Table 6 shows comparative data over Parallel developments during the 1990s, like recent years for which the value of e-commerce the ’big box’ phenomenon, “sprouted across the sales is measured in Canada. Total private sector country, dramatically changing the face of retail” sales over the Internet more than quadrupled (Lussier et al. 2003). At the same time, numerous between 1999 and 2003 (an average annual stores got into the one-stop shopping philosophy, compounded rate of growth of 45%), to account expanding both their physical size and selection for 0.8% of total sales. Similarly, e-sales in the of merchandise. There have also been many shifts retail industry more than tripled (36%), accounting in the composition of sales, such as the increased for 0.6% of the total (Statistics Canada 2004a). share of health and personal care products and While e-commerce takes hold and begins to mature, automotive products at the expense of food and data show that the number of retailers, retail space beverage, and clothing.

Table 6. E-commerce sales - share and growth average compound 1999 2003 annual growth rate 99/03 billions $ % of sales billions $ % of sales % Retail trade industry 611 0.3 2,113 0.6 36 Total private sector 4,180 0.2 18,598 0.8 45

Source: Statistics Canada, Survey of Electronic Commerce and Technology.

Table 7. Retail trends

1983 1993 1998 Number of stores 19,776 22,989 24,784 Total floor area (m)2 11,051,190 12,430,885 15,471,815 Employment (,000) - 1,624 1,889

Source: Statistics Canada, Annual Retail Chain and Retail Store Surveys. and employment in the industry have all increased 2.1.2. A time of change – at stronger rates in recent years (Table 7). These are some of the many myths surrounding ICTs that have made the news over the last While clicks did not destroy the bricks, e- several years. While all of them contain implied commerce has permanently stamped its mark on causality, and even grains of logic, so far either the retail – and wholesale - industry, as consumers they did not materialize or the opposite has use the new medium for information, product happened. The paperless society, the end of mail, characteristics, pricing, etc. The Economist (2004) the decline of traditional retail and numerous states that not only do people buy more online other such proclamations have all been grossly but “…they are also increasingly adept at using exaggerated. Quantification at this point proves the Internet to decide where and how to spend them faulty. their money offline” (p. 9). The fact remains that retail does not seem to be in danger. Moreover, Presumably, one can always argue that they will other adjustments are taking place which are not still happen – they just have not happened yet. necessarily linked to e-commerce. Another way to explain why what was supposed to

108 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY happen has not happened is by acknowledging that Years ago, and at a time when the many forces are at work simultaneously, some of telecommunications industry was still dominated which concern people’s behaviour. Then, it may by monopolies, a vast amount of literature was well be that the ICT forces are indeed pointing converging on the low price elasticity of the in the alleged directions, but all kinds of other demand for telecommunications services (meaning influences, pointing in opposite ways, dominate. voice telephony, and especially long distance). Such interactions are quite complex and do not Of course, it’s easy to refute it now, but it could result in straight-line movements for a multitude not have been more wrong. This conventional of reasons. In the very least, the ceteris paribus wisdom came crashing down from the very early assumption (all other things equal) does not hold stages of opening up the markets. People have true between ICT causes and their supposed never spoken from long distances so much. effects. Traffic data reveal that the use of wireline networks In addition to macro factors, ICTs involve alone has increased enormously over the last people’s behaviours, reaction to the new, inertia, twenty years. Both frequency of calls and talking inter-generational attitudes and many more. In time have gone up. In Canada lines increased any event, such a view would manifest itself in from about 11.5 million in 1983 to approach 20 changing composition and patterns of usage, million by 20037, and in the US from 102.2 million and this seems to be the case. It is evident that in 1980 to 188 million by 2001. Notwithstanding ICTs have significant implications on each of the big increase in the number of lines, average the above areas. Although the paperless office calls per line increased, as did average calls is not here, people’s behaviour has changed per capita. More significant was the increase in profoundly and continues to do so. Working lives time spent on the phone8. In the US, the estimated are certainly not what they used to be, and even volume of conversation time increased from family lives have changed - with implications about 1.7 trillion minutes in 1980 to approach for time use and personal interactions. The 5 trillion by 2001, while in Canada, the volume composition of mail has changed, as have the in 2003 was estimated at just short of half operations of the post office, including prices. a billion minutes9 (Table 8). In 2001, this represented Retail will never be the same as consumers have 71 minutes per line per day in the US, up from 45 never been more empowered. The list can go on minutes in 1980. Estimates for Canada are a bit and on. Thus, ICTs have powerful and lasting less, but not by much and, considering the margin influences, albeit different from the “obvious” of error involved (see Technical Box for complete ones at early stages of deployment and use. By methodological explanations), they are of the same all accounts, we have not yet seen the end. order of magnitude. This increase in time spent on the phone is somewhat higher when estimated on a 2.2. Realities per capita basis, and such estimates would represent an additional 26 minutes per line per day. If adjusted Having seen what has not happened, it is for potential under-coverage, the passage of time and instructive to look at what has. demographics (i.e. proportionately fewer children who may not use the phone), this estimate may well Fact 1: People talk on the phone more than be more than half an hour a day compared to twenty ever in history years ago.

7 Refers to voice grade equivalents (VGE) due to ISDN channels. 8 Cannot use average duration with these data (see technical box and later). 9 This is not due to network effects (externalities) as penetration was complete by the early 80s. As well, while network effects may be present with respect to international traffic, this is a tiny proportion of the calling volumes and times for local and domestic long distance.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 109 Table 8. Traﬃc over wireline networks, US and Canada

lines calls minutes calls per day minutes per day (millions) (billions) (billions) per line per capita per line per capita US 1980 102 312 1,734 8.4 3.8 46 21 2001 188 609 4,866 8.9 5.9 71 47 Canada 1983 11.5 29 - 6.9 3.3 - - 1987 12.8 37 - 7.9 3.8 - - 1997 18.4 - 340 - - 51 39 2003 19.5 - 461 - - 65 47

Note: See Technical Box for methodological explanations. Sources: Federal Communications Commission, Statistics Canada and author's estimates (italics).

This increase is even more telling if we factor capita (Table 9). This phenomenon is on the rise in that it happened during the period of the and there is no indication that a ceiling has been introduction of cell phones - which did not exist reached10. These trends may be more pronounced in the early 1980s - and their increased use, in several European countries which are heavier particularly over the last decade. Cell phone users, and elsewhere in the world, where cell traffic follows the same pattern as the diffusion of phones have overtaken fixed lines for some time cell phones - rather slow in the beginning but more now (International Telecommunications Union rapid lately. In Canada, cell phone subscribers (ITU) 2004). went from 98 thousand in 1987 to 13.5 million by 2003, while in the US from 92 thousand in 1984 The combination of wireline and wireless voice to over 140 million by 2002. communications point to the fact that, at this juncture of evolution, the Information Society At the same time, data show that in Canada the is also a “talkative society”. On the basis of number of billed minutes increased by a factor of the Canadian and US figures, the amount of time 20 over the last 10 years, from just over 2 billion in 1993 to approach 40 billion by 2003. Table 9. Traﬃc volume over cell phones, US and Canada

In the US they increased even subscribers billions of minutes more, from almost 27 billion to (millions) minutes per line/day per capita/day more than 720 billion by 2002. US Gradually, this has added an 1993 16 26.9 4.7 0.3 average of 8 minutes a day 2002 140.8 721.3 14.2 6.7 per cell phone subscriber or 4 Canada minutes per capita in Canada. 1993 1.3 2.1 4.4 0.2 In the US this amounted to 14 2003 13.5 39.4 8.2 3.5 minutes per day per mobile subscriber or 7 minutes per Sources: Federal Communications Commission and Statistics Canada.

10 There may be substitutions from fixed lines, but overall this is not het case – more of both are used.

110 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Note to readers: Technical box

Accurate and systematic data on calling volumes for wireline telephone traffic are hard to come by in North America, mostly due to the flat pricing which bundles access and unlimited local calling. In Canada, while the annual telephone survey collected the number of calls for sometime until 1993, the quarterly telecommunication survey collects data on long distance traffic since 1995 (Statistics Canada 2003). In the US, periodic studies are conducted to arrive at such estimates, which are subject to wide margins or error. (These are used, for example, to estimate the proportion of interstate calling to allocate the costs between intrastate and interstate calling among companies).

Several adjustments are made in order to arrive at the figures shown in this paper, which are therefore subject to a margin of error. They are used to provide estimates of the order of magnitude - the exact figures are not critical in the analysis contained in this paper.

The US data in Table 8 are constructed from data contained in FCC (2003). Local minutes refer to dial equipment minutes (DEMs) and 2 minutes are captured for every DEM. In the intrastate and interstate long distance only the domestic portion of the outgoing international calls is captured. Thus, the minutes for international traffic are added to the estimates available for local and domestic long distance (the source of the 1980 incoming traffic is the ITU). International traffic does not really affect the estimated number of minutes per line per day.

To arrive at the data for Canada in Table 8, the following steps took place: starting with the long distance data contained in Cat. No. 56-002, first an average figure is extracted from the quarterly series on VGE equivalent lines and long distance minutes (which includes inbound, outbound and toll free calls). From that number, the outgoing international minutes, reported by the ITU, are subtracted and the remainder is multiplied by two to capture the volume of all conversation minutes within Canada. The factor 56/15 (3.73333) from the US is then applied to arrive at some estimate of local minutes. Obviously, this assumes that the pattern is the same in the two countries. (In the absence of a better estimate, this is not a very heroic assumption. It turns out that the figures are in line with the rough factor of 10 sometimes used in Canada-US comparisons). In international calling, Canada’s proportion is much higher – and has been so historically. Then, the international calls are added (not multiplied by two, as only one of them reflects Canadian conversations).

The wireless billed minutes are those reported by the companies and they are subject to some under-coverage. Among the problems with the estimates are that the cell phones are not adjusted for the two calling parties – since both calling and receiving parties are billed for air time in North America. While this captures the air time for calls between cell phones, to the very likely extent that calls from cell phones go to fixed lines in Canada, only the billed cell phone air time will be captured resulting in an underestimate.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 111 spent talking on the phone went up somewhere in the 64% of Internet-use households in 2003, the range of half-an-hour to 45 minutes per person 95.7% are using e-mail (Statistics Canada 2004b), per day. This is only an average figure – if there and for many people everywhere this is a daily was a distribution to take into account that just activity. Obviously, such usage claims an amount over half of the population has a cell phone, the of time that was devoted elsewhere in pre-e-mail time increases substantially for that part of our times. Moreover, this time represents a fraction society that possesses and uses them. It is as if the of the time devoted to other ICTs. While precise urge to talk was suppressed and is now liberated estimates of such time-use are not available, and it by technology and prices. differs by country, according to several marketing

��

�

��

�

� ��

��

Source: Statistics Canada, General Social Survey, Cycle 14: Access to and Use of Information and Communications Technology.

Fact 2: We communicate more than ever information sources estimates of total on-line use with e-mail and spend more time on ICTs of personal computers is 75 hours a month at work and somewhere between 25-30 hours at home. The extra time spent on the phone as part of daily This order of magnitude is corroborated by data routines today is substantial, but still pales in from Statistics Canada (Chart 1). comparison to the amounts of time absorbed by other ICT activities that did not even exist over On average, individual Internet users (52.8% in two decades ago – and therefore laid no claims 2000) spent 7.4 hrs a week on-line, averaging on our time. Computers and the Internet – and more than one hour per day. Younger people more particularly e-mail - are inextricably linked spend even more time and some are very heavy to images of people in their daily lives, at the users. For instance, 10% of those in the 15-24 age workplace and/or at home. Whether we actually group spent more than 2 hours daily – perhaps a lot type away on the keyboard, use printers and more. There are also indications that such usage scanners, browse the Web or shop electronically, expands over time and would probably be even we do things of the past differently, and we also higher today. For instance, 2004 data for Canada do new things. show that the average time spent online per user ranged from 31 hours per month for the Prairies E-mail has surfaced as the top activity associated to 37 hours in Ontario (Comscore Media Metrix with Internet use. In Canada, for example, from 2004). Time spent on off-line use of computers

112 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY and other ICTs must be undoubtedly added to the radio and television. There is statistical evidence total. For instance, 43.3% of Canadian computer that time spent on television viewing and radio users aged 16-25 years used computers at home listening is declining. In Canada, for instance, for an average of one hour or more per day, while data show a decline of just over two hours per 18.1% used them for 2 hours or more (Veenhof week for television between 1983 and 2002, while et al. 2005). radio listening has actually increased somewhat given its relative revival in the ‘90s (Chart 2). Fact 3: The extra time spent talking on the Similar evidence exists in the US and many other phone, communicating with e-mail and using countries. Aggregate data mask the fact that other ICTs is in no way matched by decreases this decline is more pronounced among Internet in the use of older, more passive and less users than non-users (UCLA 2004). On the other interactive ICTs, such as television. hand, however, such declines barely make up for the extra time spent on wireline phones alone11. Some of this extra time devoted to new ICTs is For instance, it is estimated that on average one taken away from more traditional media, notably hour on the Internet reduces time spent watching

��

Notes: Data for television viewing refer to all persons aged 2 years and over. Data for radio listening refer to all persons aged 12 years and over, except for the 1983-1986 period which refers to all persons aged 7years and over. Source: Statistics Canada, Television Viewing Databank and Radio Listening Databank.

11 There is also the fact that time spent in front of television sets for videos, games and many applications other than television viewing is exploding – even more so than was the case in the early ‘80s.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 113 television by about 10 minutes, and one-half hour market liberalization and the virtual collapse of per day for the user who is online about three prices (tariff re-balancing, etc.). hours a day (Beacham 2005). It seems, therefore, that the time taken away from Table 10. Long distance billed minutes, Canada

the television is a far cry from the extra time spent outbound inbound toll-free total on the newer ICTs. millions of minutes

1995 - - - 19,123 Fact 4: The pattern of communications has 1998 28,530 4,363 5,143 38,037 changed 2000 27,711 7,226 10,466 45,402 2002 38,638 7,307 9,317 55,262

It is therefore well established that today we use Source: Statistics Canada, Telecommunications Statistics. the new means at our disposal and communicate much more than ever before. Data also In Canada, the number of wireline long distance reveal that within our significantly expanded calls increased from about 1.5 billion in 1983 communications, the pattern has changed quite a to almost 3.5 billion calls in 1992, while the bit as well. volume of billed (not ‘talked’) long distance minutes almost tripled in a few short years, from The rise of long distance... 19.1 billion in 1995 to almost 56 billion by 2002 (Table 10). Long distance calling now accounts While the number of all kinds of telephone calls for a much larger proportion of total calling. and time spent on them went up, more telling is the growth associated with long distance. This The situation is similar for cellular telephony. In is a process that has gone on for some time Canada, the proportion of long distance minutes – Chart 3 shows the evolution of local and long from cell phone calls has also increased more distance calls in Canada from 1963 to 1987, data than local minutes (by a factor of 10 compared permitting – but it intensified in the ‘90s with to 7) in only a six-year period from 1997 to 2003,

��

Source: Statistics Canada, Telecommunications Statistics.

114 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Table 11. Wireless billed minutes, Canada it creates a large deficit, in the order of US$3.5 billion). Chart 4 shows in index form the evolution local long distance total of minutes for local, interstate and intrastate long millions of minutes distance, and international calling in the US. Long 1991 - - 150 distance now accounts for a much higher proportion 1997 4,044 332 4,376 of the expanded volume of calling. 2002 28,861 3,199 32,060

Source: Statistics Canada, Telecommunications Statistics. This trend is mostly due to the collapse of long distance prices. For international calls, price per even though it is still much more expensive to use minute dropped from $1.34 US in 1980 to 34 cell phones for long distance calling (Table 11). cents in 2001 (Chart 5). The decrease has been more precipitous in recent years – while prices …and the explosion in international calling decreased by less than one-third between 1980 and 1995, they dropped by almost an additional As well, the phenomenon of increasing long two-thirds between 1995 and 2001. The drop in distance increases with distance. This can be seen average prices per call has been similar.

��

� ��

��

Source: Federal Communications Commission.

by data referring to the situation within a country Moreover, from all indications this process is not and internationally. The significant increase in complete. This pattern is expected to continue and local calling in the US is much smaller than that even intensify due to Voice over Internet Protocol of long distance (intrastate and interstate), which (VoIP), something with profound implications in turn does not even compare to the staggering for the industry. Among other impacts, VoIP is rise in international calling – particularly outgoing. expected to make long distance calls longer. The number of calls from the US to other countries increased from just under 200 million in 1980 to 6.3 Duration and scope billion in 2001, and the number of incoming calls from 165 million to over 2.9 billion (and as outgoing Time spent on the phone reflects both the number traffic has increased much faster than incoming of calls and their duration. As we talk much

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 115 ��

��

� ��

��

� � ��

Source: Federal Communications Commission.

more and more often, additional questions arise volume of calls has increased, something that can related to the circle of our associations and the be explained by both the added time pressures management of the extra time needed. Do we talk and that the cost of calls is not prohibitive for with the same people more, to new acquaintances people to cram a lot into the same conversation or both? Are calls shorter (an issue feeding to the session. However, data on the duration of local broader one of our attention span)? calls, where the bulk still is, show that these calls last longer (Table 12), while that of domestic long Data show clearly that the frequency of calls has distance oscillates. Thus, the existing evidence increased, especially long distance, presumably is inconclusive. The only sure thing is that local to both new and existing associations. One way calls are definitely shorter. to cope with this is that individual calls become shorter, a hypothesis to which existing data provide Additional data from wireless residential calls in only partial proof. More direct and detailed data the US show that in 2002 the average duration of would be needed to confirm this generally, but intrastate calls was 2.9 minutes, while interstate some available data show that the duration of was 6.3 minutes. Moreover, they are generally wireline international calls has decreased as the short. More than half (51.7%) of wireless residential intrastate calls lasted one minute Table 12. Average duration of wireline calls, US or less, while almost three-quarters lasted two minutes or less. Among interstate calls, fewer int'l local LD outgoing incoming than 40% lasted less than one minute. It may be, (minutes) then, that the duration of calls varies inversely 1980 2.6 4.3 7.9 7.0 with the ease of personal encounters. 1985 2.3 6.4 8.4 6.7 1990 2.3 5.1 8.2 5.9 1995 2.3 4.3 5.6 4.6 While there is indirect evidence from existing 2001 3.8 5.2 5.3 4.6 telephone statistics that we have expanded the circle of people we associate with at a given time, Source: Federal Communications Commission.

116 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY we have definitely done so with e-mail, which relations, friends from the past, or others with knows no limits. There are even some statistics similar interests. on the average number of online ‘friends who have never met in person’, which range from 1.1 The point is that people find utility and must derive in Japan, to 2.6 in the US and 7.7 in urban China satisfaction from all this, as it is done through – and the statistics are higher for younger males their own free will. We place value on having (UCLA 2004). an extended network of people to communicate with, and a sense of community and belonging at 3. Outcomes: social interactions and a very different level. Moreover, as we shall see shortly, we are willing to pay for this. digital speed

Wider communities, richer lives? Busy lives?

Based on the facts above, the only inferences Collectively, using ICTs absorbs several extra that can be supported is that we communicate hours in daily routines and, estimating from the more than ever and the pattern of associations data, some of that is not the result of doing through is wider. Whether or not we do so with shorter ICTs activities previously done without, but communication sessions, we definitely do so with represents extra time. This outcome certainly has more frequency. In any event, the theories of consequences as it bumps against the inescapable people closing-in on themselves, social alienation 24-hour constraint. Where is this ‘extra’ time and the like are not supported by evidence. (Surely coming from? there are those who spend all-day in online solitude, but not the society at large). The pattern First, this time reflects ICT usage everywhere, in of communication and interaction has changed. our various capacities in daily lives, at work and at We choose many relations that widen our sense home. Certainly at work ICT usage has replaced of community and, presumably, enrich our lives. other methods of work and ways of doing things. The new ICTs - and their falling prices - seem to We use computers in the place of calculators have liberated unsatisfied urges which now can be and substitute ICTs for manual processes. Not fulfilled. People make the choice to expand their only are there substitutions from non-ICT to ICT associations and move from geographically- methods, but within ICTs too, i.e. due to e-mails defined communities to communities of interest. we may make and receive fewer telephone calls or We may not talk to the person next to us, but we use fewer post-it notes. Depending on the specific are talking to someone – who may be next door context at hand and the familiarity with the or thousands of miles and time zones away. This execution, such substitutions do not necessarily does not necessarily makes us socially inept, but add time to working lives. There are, however, enriches us in other ways as we expand the circle at least two areas for further exploration. One is of associations and the “proximity” we feel with that average time spent at work has increased, and people. As a consequence, we may substitute the other is the frequent reference to the loss of a telephone call or an e-mail for a visit, but we distinction between work and play, which adds to may also visit and travel more as a result of the sense of being busier than before. Definitely, this expanded communication – counter to the the use of ICTs in our capacities as employees wisdom that we do not have to. It is not that does not provide the whole answer to the time we are becoming anti-social, it is that we are issue – we must look at our social lives too. becoming differently social. We forge many more new relations, regardless of location. Such Data show that we have increased ICT usage at associations may be extended family, professional home considerably too. Off-peak telephone calls

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 117 (roughly, calls during non-working hours) are less lives. Again, though, as people engage in all that overall, which explains the pricing incentives of on their own volition it must be that we attach both wireline and wireless operators as peak time value to such transformation and change. poses strains on the capacity of the networks. Indeed, some data from wireless usage in the Consumer spending US indicate that in 2002 outgoing calls between To underscore this, we have evidence from 7 a.m. and 7 p.m accounted for almost 70% of people’s willingness to pay. The outcomes of the total – albeit lower than the 73% in 2000. ICTs in our lives do not stop in our changed But interstate minutes alone, show that off peak behaviour patterns, but they are manifested in talking (7p.m. – 7 a.m.), increased from less than the changing pattern of spending. In the early 30% in 2000 to 41% in 2002. They also indicate ‘80s what would be considered ICT spending that more than 40% of volumes take place in the would be largely confined to telephones and the weekend, up from 31% only in 2000, indicative television. Today, household spending includes of expanded social calls. significant outlays on computers, cell phones, Internet and satellite connections. Not only ICT The second important fact is that time use is spending has increased as new ICTs entered the subject to co-tasking or multi-tasking – handling consumption basket, but its composition has two or more things at the same time. We talk on changed. For instance, household expenditures the cell phone while driving or running errands, or for telephone services accounted for 35.4% of talk on the phone while doing household chores. total ICT spending in 1997, but for 26.9% in 2003. Thus, it is not clear whether and to what extent On the other hand, Internet spending increased exactly this extra chunk of time must cut into from 1.7% to 6.2% over the same period. Yet, other activities. Well-designed time-use surveys nearly 7 out of 10 households reported owning a would be needed to shed more light into these computer in 2003 and about 22% of households issues. At the same time, we must be cognizant reported buying new computer hardware during of the fact that studying time-use can be tricky, the same year, a figure which has risen steadily in as it is subject to the prevailing technological recent years (Statistics Canada 2004c). possibilities, which are rapidly changing. Suffice it to say that one hour spent on the Internet In Canada, average ICT household spending through a dial-up connection may not accomplish increased from $2,118 to $2,780, in just a short as much as a few minutes through broadband. period (between 1997 and 2003) (Statistics Canada 2004c). Not only does this represent a Generally, though, regardless of whether ICTs significant increase in absolute terms in very recent add to co-tasking or lead to the replacement of years, but it also represents an increase in the other activities for which a learning curve must be proportion of total spending from 4.2% to 4.5%. climbed, or they both take place, all this adds to This is remarkable, as it happened over a period a sense of busy-ness. The situation is obviously where ICT prices have plummeted. For instance, much more pronounced among the sizeable computer prices dropped by 10% between 2002 group of ICT users, and even more so among and 2003 alone (Statistics Canada, Computer and the smaller sub-group of heavy users. Although Peripherals Price Indexes). Moreover, as prices these people may feel the strain more than the fell and even as penetration of home computers others, it trickles over to the rest of the society, increased from 39.8% in 1997 to 66.8% in 2003, and this contributes to a great extent to the feeling spending on computer equipment and supplies that we are busier than ever. In that sense, ICTs over the same period grew from an average of can be added to the broader spectrum of time- $299 per household in 1997 to $326 in 2003 saving technologies that ironically lead to busier (Statistics Canada 2004c).

118 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY More detailed analysis reveals that most of the However, ICT spending represents a higher aggregate expenditure on ICTs comes from proportion o,f spending among lower-income people with higher incomes – which links to the households. In 2002, it accounted for 6.3% of issue of the digital divide. In 2002, households average total spending among households at at the top income quintile accounted for one-third the bottom income quintile compared to 3.9% of all spending, and households at the second of average total spending among households at highest income quintile for almost an additional the top income quintile. However, the desire of one-quarter of total spending (Chart 6). people to participate in the Information Society

��

Source: Statistics Canada, Survey of Household Spending, 2002.

Table 13. Household ICT spending by income, Canada 2002

income quintiles bottom fourth third second top all

Number of households 2,331,550 2,331,550 2,331,550 2,331,550 2,331,550 11,657,730 avg. total spending ($) 20,222 35,625 52,633 71,741 120,227 60,090 avg. current spending ($) 18,627 29,769 40,259 51,618 75,754 43,206 avg. ICT spending - all households ($) 1,279 1,976 2,615 3,355 4,663 2,779 avg. ICT spending - reporting households ($) 3,569 4,179 4,639 5,318 6,554 5,107 ICT as % of avg. total spending - all households 6.3 5.5 5.0 4.7 3.9 4.6 ICT as % of avg. current spending - reporting households 19.2 14.0 11.5 10.3 8.7 11.8

Note: Total spending differs from current spending as it includes personal taxes, insurance and pension and gifts of money. Source: Statistics Canada, Survey of Household Spending, 2002.

12 Total spending differs from current spending as it includes personal taxes, insurance and pension and gifts of money

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 119 can be understood more dramatically by the share home. The paper takes a long view to debunk of average current spending12 among households some recent myths and demonstrate some factual at the lowest income quintile reporting ICT changes. It shows that the paperless office is the expenditures (Table 13), which approached 20%! office that never happened, with consumption of paper is at an all-time high; the business of The reason for this is that a good amount of ICT- transporting paper is thriving; professional related expenses are fixed, such as the price of travel has most likely increased in the era of computers, basic telephone and cable service or videoconferencing, and; e-commerce clicks have Internet connections. The discretionary spending, not made a dent on the bricks, and do not justify such as usage of long distance telephone services, recent fears of negative consequences on retail that can be controlled accounts for a smaller employment and deflation of real estate. proportion of total ICT spending. It demonstrates that the Information Society is a 4. Summary ‘talkative society’. People have never spoken more on the telephone before, and this happens at a time Having lived through the proliferation of ICTs, that we also send and receive massive amounts and having satisfied basic curiosity with detailed of e-mails and other electronic communications measures and analyses of their penetration and with no match in history. We also spend large use, the interest is shifting to the understanding amounts of time using ICTs generally. While of their outcomes and impacts. This represents naturally some ICT-related activities replace fertile ground for research that could lead to previous methods of doing things (at work or at knowledge conducive to improved ICT use and home), a certain amount represents extra time future applications. ICT-induced outcomes touch that must come from somewhere. Whether some virtually everything aspect of life, ranging from activities are displaced or through co-tasking, the economic to the social, the political and the this is particularly the case for the sizeable and cultural domains. From how computers affect growing user segment of the population, but it firm-level productivity, to how cell phones help trickles down to everyone adding to the sense of the efficiency of markets, to how the Internet feeling busier than ever. facilitates arranged marriages in India, there is an endless list of matters to examine. The paper stresses that key outcomes of ICTs are manifested in shifting behavioural patterns This paper takes on selective issues and uses everywhere, with real consequences. Moreover, statistical information to address them and draw the pattern of communications has changed, inferences. Implicit throughout this exploration something exemplified by the rise in long distance is the message that integration of data is needed and the explosion in international calling made to perform reality checks, explore linkages, test possible by liberalized markets and falling prices. hypotheses and create new ones in the quest to Such expanded circles of communication have understand ICT outcomes and impacts. This found an even better expression through e-mail that includes all the new ICT-related datasets that knows no boundaries. It is as if people had hidden are being developed in some countries and urges which are now liberated because they can start to accumulate in the time-series sense, but be expressed. People make the choice to expand also extends to all other existing and seemingly their associations and move from geographically- unrelated data sources. defined communities to communities of interest. As well, they are willing to pay for those choices. Economic and social outcomes are interrelated ICT spending is on the rise and, within this higher and affect people in their working lives and at spending, substitutions take place in favour of

120 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY newer ICTs, such as the Internet, and against older Kraut, R., Lundmark, V., Patterson, M., Kiesler, ones, such as the telephone. The willingness of S., Mukopadhyay, T., Scherlis, W. (1998) “Internet people to pay can also be seen by the choice of paradox: A social technology that reduces social many low-income households to spend a relatively involvement and psychological well-being?”, higher proportion of their income on ICTs. American Psychologist, Vol. 53, No. 9, 1017- 1031. References Lussier, R., McDowell, D. and Cryderman, E. Beacham, F. (2005) “Static Begins to Clear on (2003) “A commodity perspective on recent How Internet Affects TV”, TVTechnology.com, trends in retailing”, Canadian Economic Observer, 04/05. Statistics Canada, Catalogue. No. 11-010-XPB, December. Clayton, T. (2005) “IT Investment, ICT Use and UK Firm Productivity”, Office of National Nie, N. and Erbring, L. (2000), “Internet and Statistics. Society”, Stanford Institute for the Quantitative Study of Society. Comscore Media Metrix (2004) “Canada, Demographic Profile Report”, retrieved from Nurmela, J. and Sirkiä, T. (2004) “Almost Statistics: Fast Facts, Public Works and Everybody Sends Text Messages”, Statistics Government Services Canada. Finland, www.stat.fi

FAOSTAT (2004) Food and Agriculture OECD (2003) “ICT and Economic Growth: Organization www.faostat.org Evidence from OECD Countries, Industries and Firms”, Paris. Federal Communication Commission (2003) “Trends in Telephone Service”, Washington. Orbicom (2003) “Monitoring the Digital Divide… and Beyond”, National Research Council of Forest Products Association of Canada (2003) Canada. “Annual Review”, www.cppa.org Orbicom (2005) “From the Digital Divide to Forest Products Association of Canada (2004) Digital Opportunities: Measuring Infostates for “The Future of Paper”, www.cppa.org Development”, National Research Council of Canada. ILO (1998) “Structural and Regulatory Changes and Globalization in Postal and Sellen, A. and Harper, R. (2001) “The Myth of the Paperless Office”, MIT Press. Telecommunications Services: The Human Resources Dimension”, International Labour Statistics Canada (2003a) “Canada’s Journey to an Office, Geneva. Information Society”, Catalogue No. 56-508 XIE. Statistics Canada (2003b) “Quarterly Ivey Business Consulting Group (2003) “Internet, Telecommunications Statistics”, Catalogue No. e-mail Driving Thirst for Paper”, University of 56-002-XIE, 2nd quarter 2003, November, and Western Ontario. various issues. ITU (2004) “World Telecommunications Indicators” database, Geneva, www.itu.org

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 121 Statistics Canada (2004a) “Electronic Commerce Veenhof, B., Sciadas, G., and Y. Clermont. and Technology”, The Daily, April 16, www. “Literacy and Digital Technologies: Linkages and statcan.ca Outcomes,” Statistics Canada, Connectedness Series, Cat. No. 56F0004MPE, No. 12. Statistics Canada (2004b) “Household Internet Use Survey”, The Daily, July 8, www.statcan.ca Wellman, B., Haase, A. Q., Witte, J., Hampton, Statistics Canada (2004c) “Survey of Household K. (2001) “Does the Internet increase, decrease, Spending”, The Daily, December 13, www. or supplement social capital? Social networks, statcan.ca participation, and community commitment”, American Behavioural Scientist, 45, 3, pp. The Economist (2004) May 15. 437-56. UCLA (2004) “World Internet Project”, Los Angeles, California.

122 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY ICT skills measurement in Eurostat’s Information Society Statistics

Christophe Demunter Eurostat, Directorate F “Social statistics and Information Society”

Outline of the presentation o Skills measurement in the household survey (variables and output • Introduction: general overview of indicators,…) Information Society Statistics in Eurostat, o ICT occupations focusing on the ICT usage surveys o Skills measurement in the enterprise (enterprises and households/individuals): survey o Framework (background) o Contents (main variables and • Outlook breakdowns) o Future developments in the household o Methods survey (ad hoc module on digital o Strenghts and weaknesses literacy in the 2007 survey) o Future developments in the enterprise • Measurement of e-skills survey (ad hoc module on e-skills in the o Definition problems 2007 survey) o Scope of Eurostat’s work • First results from the 2005 survey

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 123 Introduction The eEurope 2005 benchmarking exercise has ended and is currently being revised by the i2010 2 In the middle of the nineties, European policy strategy . This five-year strategy to boost the makers started developing integrated programs digital economy was adopted by the European for fostering the so-called information society. Commission in June 2005. While eEurope 2005 The first big initiative,eEurope – an Information focused on access and connectivity and not so Society for all, was launched in 1999 and wanted much on inclusion, one of the three pillars of i2010 to “bring the benefits of the information society is “to promote an inclusive European information to all Europeans”. The objective of eEurope was society”. Indeed, “the information society will ambitious: to bring every citizen, school and be sustainable only if it ensures inclusion and business online and to exploit the potential of broad electronic participation in society (e- the new economy for growth, employment and participation). Tackling all forms of the digital inclusion. divide is therefore a key concern of i2010”. An important condition of this e-participation is that the citizen has the necessary skills or competences In 2000, the Heads of State and Government of to use ICTs, even in its most basic form. the European Union met in Lisbon and launched a series of ambitious reforms at national and The ICT usage surveys carried out within the European level, amongst others to make the EU European Statistical System (the network of “the most dynamic and competitive knowledge- national statistical institutes and Eurostat) based economy in the world” by 2010. This led have been a major source for monitoring the to the firste Europe Action Plan (2000-2002) built digital divide. However, the issue of e-skills or around three aims: a cheaper, faster and more digital literacy has not been studied in detail secure Internet, investment in people and skills until 2005. In parallel with the changing policy and stimulating the use of Internet. priorities, Eurostat is currently developing new questionnaires for the 2007 survey that put further The 2002 eEurope Action Plan’s main objective emphasis on e-skills. was to “to provide a favourable environment for private investment and for the creation of new This paper gives an overview of the work done at jobs, to boost productivity, to modernise public Eurostat on measuring e-skills or digital literacy. services, and to give everyone the opportunity After a general introduction to the Community to participate in the global information society”. surveys on ICT usage by enterprises and For monitoring the eEurope 2005 Action Plan, a households, the questionnaire items related to e- set of benchmarking indicators1 was developed. skills will be presented, including first results of The majority of these 23 indicators have been the 2005 household survey. The final part takes a collected via enterprise and household surveys look into the future, discussing the preparations conducted by the European national statistical for a more detailed module on e-skills to be institutes, coordinated by Eurostat. included in the 2007 surveys.

1 Available at DG INFSO’s website: http://europa.eu.int/information_society/eeurope/i2010/docs/resolution.doc 2 Available at DG INFSO’s website: http://europa.eu.int/information_society/eeurope/i2010/i2010/index_en.htm

124 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Eurostat’s statistics on ICT usage e‑commerce (sales and purchases via the Internet and other computer networks). The data is Since 2001, Eurostat has been conducting collected from enterprises in business sector with a Community-wide survey on ICT usage in at least ten persons employed and can be broken enterprises, the household survey followed one year down by economic activity and size class of the later. The data is collected by the national statistical enterprises. institutes, following a model questionnaire designed by Eurostat. Such model questionnaire is For the 2005 survey, the information refers a major tool to obtaining harmonised data for all to approximately 200 000 enterprises in 25 participating countries. Each statistical office uses European countries, resulting in reliable data even its national experience and expertise to work out for more detailed breakdowns or indicators. the most appropriate survey methodology, which leads to highly reliable data. On the other hand, The household survey methodological recommendations from Eurostat are a guarantee for comparable data. The Community survey on ICT usage and in households and by individuals collects data on In the ‘early years’ of the survey, only a limited households’ ICT equipment (devices, Internet number of countries participated but in the course connection, broadband, etc) and on individuals’ of time, most European countries have joined frequency and location of computer and Internet this exercise. In 2005, nearly all Member States use, the purpose and nature of their activities and candidate countries took part in the ICT on the Internet, ICT security, e-skills as well usage surveys. Full participation is mandatory as barriers to computer or Internet access. The from 2006 onwards following Regulation scope of the survey is limited to individuals 808/2004 of the European Parliament and of aged 16 to 74 (and households with at least one the Council concerning Community statistics3 member in this age segment). The survey has an on the information society and its implementing extensive set of explanatory variables, allowing measures4. for detailed analysis of societal phenomena: age, gender, employment situation, educational An important part of the questionnaires aims level, occupation, type of household (presence to collect the necessary information for the of children, number of persons in the household), Commission’s benchmarking exercises. However, degree of urbanisation, etc. Due to data collection Eurostat also includes other topics of relevance problems or legal restrictions at national level, to the information society or the knowledge the household income is not collected in all economy. A brief overview of the contents of the participating countries. The total sample size survey is presented in the next sections. is about 150 000 households and 220 000 individuals, the survey are in general conducted The enterprise survey using face-to-face or telephone interviews.

The Community survey on ICT usage and e- Strengths and weaknesses commerce in enterprises focuses on computer and related technology usage, Internet access, The harmonised data collection across Europe, e‑government use, e‑security but also on leading to highly comparable information, is

3 Regulation (EC) No 808/2004 of the European Parliament and of the Council of 21 April 2004 concerning Community statistics on the information society, Official Journal of the European Union, 30.04.2004, L143. 4 Commission Regulation (EC) No 1099/2005 of 13 July 2005 implementing Regulation (EC) No 808/2004 of the European Par- liament and of the Council concerning Community statistics on the information society, Official Journal of the European Union, 14.07.2005, L183.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 125 without any doubt the major strength of these communication services. Developing IT skills is surveys. The harmonisation is based on close not just about satisfying the specialist skills needs cooperation between Eurostat and the national of the industry. It is also about building the skills statistical offices in working groups or task every person and business will need to participate forces, on the model questionnaire developed by fully in the society into the future. This is why Eurostat and on methodological coordination by nearly every advanced world economy has made means of a methodological manual summarising IT skills development a priority. E-skills are best practices of the countries and giving important both to the community and to the recommended guidelines for the survey. The work economy. done in the EU in this field has in the meantime been adopted by various other countries inside Computers linked to other computers in homes and outside Europe. This is not only a recognition and offices, schools and universities, shops and of Europe’s competence or expertise, but also a banks, libraries and cafés around the world are basis for enlarging the comparability of the data changing the way we communicate, learn, gather beyond Europe. information, find employment, buy and sell, and conduct businesses. Citizens will become Further, the relatively large sample sizes ensure increasingly dependent on e-skills to access the reliability, accuracy and representativeness information and services, to develop business of the data, even when analysing more detailed opportunities and to go about their daily lives. societal groups or very specific indicators. Past surveys have shown that for instance the lack of skills is one of the main barriers to having However, as with most official statistics, the an Internet connection at home. In this sense, “e- timeliness of the data could be improved. The skills for all” may be the primary condition to “an coordination with statistical institutes and the Information Society for all” - the ambitious title internal procedures (contracts, legislation, etc.) of the 1999 eEurope initiative – and to bridging are time‑consuming and certainly affect the the so-called digital divide. timeliness. The time lag between defining the variables and the publication of the first data is To the economy, the rapid take-up of new about 20 months. An important share of this lag technologies by manufacturing and other is fixed by Regulation 808/2004 stating that the traditional industries means that more people annual regulation has to be drawn up 9 months with technical skills are needed. Many jobs today before the start of the data collection period. require people to use word processors, e-mail and However, the planning of the survey – with data the Internet. In the future, there will be an ongoing collection in the second quarter – makes it possible and increasing demand for complex technical and to publish indicators at the end of November of business skills. The skills that people need for the reference year, which is actually relatively work (but also for training and education) focus satisfying in the context of official statistics. on using applications to discover and manipulate information and exploit opportunities presented by Measurement of e-skills new technologies. Next to this basic to advanced skills needed in the workplace, specialist ICT Relevance5 skills are a cornerstone to the development of the economy in the 21st century. Specialist ICT skills IT or ICT skills cover a wide range of activities extend beyond the conventional perceptions of from simply reading an e-mail to creating software programmers or hardware solderers. and maintaining complex computing and Although these skills are still very important, the

5 Partly based on Skilling People for an Information Society, New South Wales Information and Communication Technology Skills Action Plan (http://www.oit.nsw.gov.au/pdf/3.4.1-ICT-Skills.pdf)

126 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY industry also needs technical people to support digital literacy may rather refer to a basic more networks and applications, business people to general level of competences and not to more develop solutions in e-commerce, and creative advanced ICT specialist skills (cfr. the scope people to design interactive multimedia products used by DG ENTR’s E-Skills Forum). and services.

Definition problem Approach in the household survey

A first problem which occurs when stepping The set-up of the household survey does not allow into the world of measuring e-skills, is the fact for an in-depth exam of the respondent’s e‑skills. that there are no commonly agreed definitions First, this would expand the survey drastically available. However, several projects led to a (while measuring digital literacy is only one of the multitude of approaches. The discussion of this many aspects the survey tries to cover). Second, issue goes beyond the scope of this paper, but the data collection methods used by the national the workshop document submitted by Matthew statistical institutes may be inappropriate for such Dixon gives an overview of the most important type of study. outcomes6. In the context of the Community survey on ICT Although partly influenced by the work done by usage in households and by individuals, the DG Enterprise of the European Commission via scope needs to be restricted to basic skills or the e-Skills Forum7, the design of the questions user skills. The target population of the survey in Eurostat’s ICT usage surveys has no real are all individuals aged 16 to 74, regardless reference point. The questions discussed further employment situation, age or educational level. in this document are an attempt to measure several Hence the need for ‘low entry’ questions, at least dimensions of e-skills without using a commonly as a starting point. agreed conceptual framework. In the past, the household survey has measured In the framework of the specific module which ICT competences using a self-assessment is being developed for the 2007 household approach, i.e. the respondent simply indicates survey, digital literacy is defined as the ability whether he/she is able to carry out specific tasks to use ICTs in general and more specifically related to computer or Internet use, without computers and the Internet (as these are the assessing or testing these skills or actually main ICT topics in the survey). It is more or observing the skills. In the 2005 and 2006 survey, less the ICT era equivalent of the more general two questions collecting the respondent’s self- literacy (ones competences in reading, writing assessment on skills have been included in the and arithmetic). “Digital literacy” is the term model questionnaire, one on computer skills, and used in DG INFSO’s i2010 programme, it is another on Internet related skills. For both topics, however strongly overlapping with the term a set of six items is presented to the respondent “ICT skills” or “e-skills” which has been used who is asked to simply answer whether he or she in the Community surveys up to now, although has already carried out this activity.

6 Dixon, M. (2005). E-skills and their measurement, unpublished paper submitted for the Eurostat Conference “Knowledge Econ- omy – Challenges for Measurement”. 7 For further information on the E-Skills Forum: http://europa.eu.int/comm/enterprise/ict/policy/ict-skills.htm

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 127 Community survey on ICT usage in households and by individuals, 2006 model questionnaire: computer skills.

Which of the following computer related activities have you already carried out? (tick all that apply)

a) Copying or moving a file or folder ………………………………………………….……………………………. b) Using copy and paste tools to duplicate or move information within a document ………………… c) Using basic arithmetic formulas in a spreadsheet ……………………………………………………………. d) Compressing files …………………………………………………………………………………………… e) Connecting and installing new devices, e.g. a printer or a modem ………………………………………………………… f) Writing a computer program using a specialised programming language …………………………………… g) None of the above ……………………………………………………………………………

Community survey on ICT usage in households and by individuals, 2006 model questionnaire: Internet skills.

Which of the following Internet related activities have you already carried out? (tick all that apply)

a) Using a search engine to find information ………………………………………………………………. b) Sending e-mails with attached files (documents, pictures, etc.) …………………………………………………….. c) Posting messages to chatrooms, newsgroups or an online discussion forum ………………………………………. d) Using the Internet to make telephone calls …………………………………………………………………… e) Using peer-to-peer file sharing for exchanging movies, music, etc. …………………………………………..; f) Creating a web page …………………………………………………………………………………………………. g) None of the above …………………………………………………………………………………………..;

A recent study8 published in the Social Science than in Eurostat’s model questionnaire for the Computer Review, indicates that people’s household survey, the conclusion that ‘self- perception of their computer skills is a very good reported ratings of digital literacy items may be indicator of people’s actual abilities as measured used as a proxy for actual skill measures’ may be through observations or survey items that measure a fortiori valid in the case of the questions and users’ actual knowledge of computer and Internet items used in the current e-skills module (see related terms and functions. In this study, the 2005 and 2006 model questionnaire). validity of self-reported scores on digital literacy items was tested by submitting a subset of the The results of the individual items of the questions respondents to multiple choice questions about the above may not seem highly relevant as stand- terms used in the self-assessment. The correlation alone information. Indeed, the set of items only coefficients for the self-reported ratings and the refers to six very specific activities (and approach multiple-choice question results were statistically which was chosen for reasons of clarity of the significant among the majority of the tested questionnaire). However, the items are used measures. Considering that the wording of the to recode the respondents into levels of skills: question in the mentioned study was less strict9 persons who answered positive to 1 or 2 of the

8 Hargittai, E. (2005). Survey Measures of Web-Oriented Digital Literacy. Social Science Computer Review, Vol. 23 No. 3, Fall 2005, p371-379. Downloadable from http://www.ictliteracy.info/rf.pdf/hargittai-SSCORE05.pdf 9 “How familiar are you with the following Internet-related items ? Please choose a number between 1 and 5 (none, little, some, good, full understanding)” (with items such as ‘modem’, ‘html’, ‘spam’, ‘weblog’, ‘preference setting’, ‘newsgroup’, etc.), compared to the Eurostat wording “Which of the following computer [Internet] activities have you already carried out?”

128 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY computer-related items, are coded as ‘low level Next to the specific module on e-skills, the survey of basic computer skills’, persons who ticked offers more indirect questions allowing to get a 3 or 4 items are coded as ‘medium level’ while broader picture of digital literacy. The most recent those who said to already have carried out 5 or computer (or Internet) use or the frequency of all activities are labelled as ‘high level of basic computer (or Internet) use are indeed very basic computer skills’. A similar approach is used for and very indirect indicators, but the intensity or the Internet related items. frequency of one’s ICT usage is expected to be related to the level of digital literacy (though the Although the items are more or less listed in an direction of the causality is difficult to identify). ordinal scale, the assumption of ordinality was dropped as it is, for instance, possible that a high Supporting information can also be obtained skilled programmer answers positive to all items from the questions concerning the type of excepting the one referring to spreadsheets for Internet services the respondent has used in the the simple reason the person has no need to use three months preceding the interview. Of course, such an IT tool. whether or not someone used a particular service Community survey on ICT usage in households and by individuals, 2006 model questionnaire: way of obtaining skills.

Where or how did you obtain the skills to carry out these activities? (tick all that apply)

a) Formalised educational institution (school, college, university) ……………………………………………………………. b) Training courses in adult education center (but not on the initiative of your employer) ………. c) Vocational training courses (on the demand of the employer) ……………………………………………………………. d) Self-study using books, cd-roms, etc. ……………………………………………………………………………. e) Self-study in the sense of learning-by-doing ……………………………………………………………. f) Informal assistance from colleagues, relatives, friends ……………………………………………………………. g) Some other way ……………………………………………………………………………………………………

The above two self-assessment questions are will not only be based on one’s ability or skills to embedded in a broader questionnaire module use advanced services, but rather by the necessity on e-skills. A first question in the module asks to use such service. all respondents who ever used a computer when they have last taken a training course (of at least 3 Further, the items related to barriers for ICT hours) on any aspect of computer use. The module usage can help to identify to what extent people is closed by a question asking the respondents are not using the Internet or specific services such who answered positive to at least one of the as e-commerce just because they feel they lack skills items, where they have obtained their e- the skills to do so. skills. The question is supposed to give an insight Further in this document, some preliminary in the relevance of certain formal and informal results from the 2005 survey are presented. ways to obtain computer and/or Internet skills. As is the case for the entire model questionnaire the answers can be analysed for different socio- ICT occupations in the household survey demographic groups (by age, gender, educational In order to be able to distinguish ICT level, employment situation, etc.) but can also be professionals from other respondents or to linked to the three skills levels obtained from the distinguish manual workers from non-manual self-assessment questions. workers, an additional background variable was

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 129 inserted in the model questionnaire for the 2005 major groups 0 to 5 are considered to be non- survey: the respondent’s occupation according manual workers11, while major groups 6 to 9 are to the International Standard Classification of considered to be manual workers12. Occupations10, a classification recommended by the International Labour Organisation (ILO) and However, for ICT professionals the recoding is used in most of Eurostat’s social statistics. more troublesome as even at the most detailed level of ISCO coding, identifying ICT jobs can Collecting data on ICT usage broken down by be problematic, partly because the classification individual occupations would go beyond the is almost twenty years old (a revision better scope of this survey (and would necessitate very covering knowledge economy occupations is large samples!), therefore the focus is rather on foreseen to be implemented in 2008). Where the groups of occupations. As mentioned above, OECD uses both a broad and a narrow definition two such breakdowns are foreseen, namely of ICT-skilled employment13, Eurostat preferred ICT professionals versus other occupations and to only consider occupational categories manual versus non-manual workers. where most jobs can be supposed as ICT professionals. To be able to make such regroupings, it is necessary to collect (or code) the occupations at At 3 digit level of ISCO, only 2 categories fit this a detailed level – i.e. at least 3 or 4 digits of the very narrow approach: minor groups 213 and classification – as the major (1 digit) or submajor 312, ‘computing professionals’ and ‘computer groups (2 digits) are too general to distinguish associate professionals’ respectively. Where the ICT professionals from other occupations. national statistical institutes have the possibility to collect and code at 4 digit level, a more refined For manual workers versus non-manual workers, selection of eight unit groups14 is suggested for the regrouping is more or less straightforward: regrouping jobs as ICT professionals.

10 For further info, see the ILO site (http://www.ilo.org/public/english/bureau/stat/isco/index.htm) or Eurostat’s classification server (RAMON), http://europa.eu.int/comm/eurostat/ramon. An introductionary note can be found at http://europa.eu.int/comm/eurostat/ ramon/documents/isco_88_com/isco_88_com.zip 11 Following groups are labelled as non-manual occupations: Major group 1: Legislators, senior officials and managers; major group 2: Professionals; major group 3: Technicians and associate professionals; major group 4: Clerks; major group 5: Service workers and shop and market sales workers; major group 0: Armed forces. 12 Following groups are labelled as manual occupations: Major group 6: Skilled agricultural and fishery workers; major group 7: Craft and related trades workers; major group 8: Plant and machine operators and assemblers; major group 9: Elementary occupations. 13 OECD (2005). New Perspectives on ICT Skills and Employment, www.oecd.org/dataoecd/26/35/34769393.pdf 14 These eight ISCO unit groups (i.e. 4 digit level) are: 1236: Computing services managers; 2131: Computer systems designers, analysts and programmers; 2139: Computing professionals not elsewhere classified; 2144: Electronics and telecommunications engineers; 3114: Electronics and telecommunications engineering technicians; 3121: Computer assistants; 3122: Computer equipment operators; 3132: Broadcasting and telecommunications equipment operators.

130 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY On the basis of the European Labour Force From the more general indicators collected in the Survey (LFS), the share of computer occupations household survey, we observe that one in three in total employment in the European Union EU citizens have never used a computer (see – when considering only ISCO minor groups Graph 1), ranging from 8% in Sweden to 65% in 213 and 312 – is estimated at approximately Greece. It is obvious that these people’s e-skills 1,6% (corresponding to about 3,2 million jobs), will not be appropriate for being fully included in ranging from merely 0,5% in Greece to over the information society. 3% in Sweden and the Netherlands. From the 2004 LFS, some easy to predict conclusions can Looking in more detail at the groups of society be noted: computer occupations tend to be taken at risk (see Graph 2), we see – as expected – that by men (the proportion of female employment digital literacy is particularly a problem for Graph 1. Proportion of individuals aged 16-74 never having used a computer (2005), by country

��

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005

in ICT occupations is less than 20%, compared older persons (60% of the people over 55 years to 44% for all occupations) and by younger has never used a computer), lower educated16 workforce (the share of persons aged 15 to 34 in citizens (56% who never used a computer, ICT occupations is almost 50% while only about compared to ‘only’ 26% and 9% for middle 35% in total employment). and higher educated respectively) and – to a lesser extent - women (36%, compared to 31% First results of the 2005 household survey15 for men) and unemployed (36%). Further, the Digital literacy is a problem that concerns an regional dimension can not be ignored: in thinly important part of the population populated areas, 50% more people have never

15 Please note that the statistics presented in this paper are preliminary. This is especially the case for EU aggregates as not all countries had delivered their data at the time of drafting the document (most aggregates are based on all EU countries, excepting Denmark, France, Ireland, Malta and Spain). The results presented here don’t have the ambition to present a full overview of the state of affairs of e-skills in Europe, but rather to illustrate the measurement of e-skills by Eurostat using the 2005 ICT usage household survey. 16 ‘Lower educated’ means no formal education, primary education or lower secondary education (levels 0, 1 or 2 of ISCED-97, the International Standard Classification of Education), ‘middle educated’ refers to upper secondary or post-secondary non-tertiary education (ISCED 3 or 4), ‘higher educated’ refers to tertiary education programmes (ISCED 5 or 6), e.g. university.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 131 Graph 2. Proportion of individuals aged 16-74 never having used a computer (2005), EU-25, by socio‑demographic background

��

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005

used a computer compared to the more urbanised group of persons aged 25 to 54, half belongs to the densely populated areas (42% for thinly and 28% ‘offline population’. As shown on Graph 4, 75% for densely populated areas respectively). When of the people with a lower educational level is not comparing economically prosperous regions with regularly using the Internet, meaning that specific relatively poorer regions17, the phenomenon is even programmes for e-learning or jobsites aiming at clearer as in the latter the share of the population lower educated will reach only one fourth of their never having used a computer is almost double potential target public. compared to the more prosperous regions not covered by Objective 1. Also in groups of society where Internet use is expected to be widespread, such as youngsters Similar conclusions as for computer use can be between 16 and 24 years old, more 30% are drawn when looking at Internet use (see Graphs not regular Internet users. In Sweden, the 3 and 4). More than half of the European citizens Netherlands and Iceland, only 6 to 8% of are not regularly18 using the Internet (56%), youngsters are not regularly using the Internet ranging from about 25% in Sweden and the but in Czech Republic, Greece, Italy and Netherlands to over 80% in Greece. Again, older Cyprus about half or more of the persons in or lower educated persons are significantly more the age group 16 to 24 are not online (data not absent on the Internet. But even in the middle age presented in the graphs).

17 Regions where development is lagging behind and which are eligible for support from the EU’s Structural Funds under Objective 1, i.e. regions where per capita GDP is below 75% of the EU average. In general, these regions are mainly located in the Southern and Central European Member States. 18 Regular Internet users are defined as individuals who use the Internet at least once a week.

132 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Graph 3. Proportion of individuals aged 16-74 not regularly using the Internet (2005), by country

��

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005 Graph 4. Proportion of individuals aged 16-74 not regularly using the Internet (2005), EU-25, by socio‑demographic background

��

�� Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005 When households were asked for the reasons why not interesting) and costs too high (equipment they don’t have an Internet connection at home, costs or access costs), but is more important than the 2005 household survey shows that 22% of the reasons linked to security concerns, concerns non-connected households gave lack of skills as about harmful content on the Internet or simply one of the main reasons. This reason ranks lower having the necessary access elsewhere than at than don’t want Internet (content is not useful, home.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 133 From the above results, it is clear that a Important differences between the European considerable part of Europe has not yet entered countries can be observed. For instance in Iceland, the information society. Although lack of e-skills lower educated persons are more inclined to use to fully participate is not the only explanation, it the Internet than higher educated people in quite is a major issue in extending e-inclusion to all some other countries. Indeed, as Internet use societal groups. The process of inclusion tends comes closer to saturation level, the absolute to be relatively slow: the offline population divide between different groups in society should decreased by only 6 percentage points from 62% become narrower. However, a detailed analysis in 2004 to the earlier mentioned 56% in 2005. of Internet use in Europe or the digital divide is outside the scope of this paper. More than half of the EU citizens accessed the Internet in the three months preceding the Only few people recently attended a computer survey training course From Table 1, we see that 52% of the population between 16 and 74 years old said that they had A first question in the 2005 e‑skills module asked used the Internet in the three months before respondents when they had last taken a course the interview. From the previous section, we (of at least 3 hours) on any aspect of computer remember that a majority uses it regularly (44%), use. Graph 5 below shows that approximately while the remaining 8% of the population who half of the population (aged 16 to 74) have never does use the Internet does this only occasionally taken any course on computer use, without big (less than once a week). differences across societal groups - apart from

Table 1. Proportion of individuals aged 16-74 having used the Internet in the three months before the interview (2005), by country and by socio‑demographic background.

EU25 BE CZ DE EE EL IT CY LV LT LU HU NL AT PL PT SI SK FI SE UK IS

All individuals 52 58 32 65 59 22 34 31 42 34 69 37 79 55 35 32 47 50 73 81 66 86

Men 56 62 35 70 62 26 39 33 43 35 81 37 84 60 37 35 50 54 73 85 71 87 Women 48 53 29 60 57 19 28 29 41 34 58 37 75 50 33 29 44 46 72 78 62 85

Aged 16 to 24 80 83 64 93 89 46 62 58 84 74 91 61 97 84 74 70 84 82 96 97 89 98 Aged 25 to 34 67 74 41 89 74 34 49 47 60 45 73 44 92 76 46 46 71 58 95 95 82 97 Aged 35 to 44 61 69 41 77 70 26 41 31 42 33 79 47 89 65 35 34 55 55 87 93 77 94 Aged 45 to 54 48 57 29 65 57 17 31 21 30 26 71 37 81 52 23 21 36 50 73 85 62 85 Aged 55 to 64 33 36 15 46 33 7 14 8 14 9 56 18 62 26 12 10 u 17 51 74 48 70 Aged 65 to 74 13 12 2 20 10 1 4 4 4 2 26 5 34 8 3 2 u 1 18 27 25 42

Men, 16 to 24 81 83 63 94 93 48 62 56 82 72 93 61 96 85 76 68 86 88 96 98 90 98 Men, 25 to 54 62 69 39 79 66 29 46 35 43 33 85 40 90 68 33 38 56 57 82 92 78 92 Men, 55 to 74 31 32 12 43 25 6 14 9 11 6 63 15 62 26 10 9 u 13 44 64 44 61

Women, 16 to 24 79 83 65 91 84 44 61 61 85 76 89 62 98 83 73 72 83 76 97 96 89 97 Women, 25 to 54 56 63 35 73 68 22 34 31 46 36 65 45 84 60 34 30 52 52 87 90 70 92 Women, 55 to 74 18 19 6 26 21 2 5 3 8 5 24 11 40 12 6 4 u 9 33 48 33 57

Students 88 93 78 96 99 62 77 75 92 93 96 77 99 95 87 95 92 94 98 97 98 100 Employees 66 71 40 77 69 34 46 36 53 43 78 49 91 70 45 39 62 56 84 90 78 91 Self-employed 55 73 u 83 67 23 44 26 30 20 76 51 89 65 28 27 u 69 75 83 71 86 Unemployed 41 45 16 58 47 18 29 38 19 13 51 26 91 47 18 19 u 34 57 87 68 66 Retired, inactive 22 24 6 33 20 4 7 10 12 4 41 12 54 23 7 4 u 8 33 39 37 46

Lower educated 31 38 26 56 52 6 14 12 28 28 53 12 61 32 31 16 21 27 58 68 30 79 Middle educated 56 62 28 65 54 29 53 30 38 24 76 51 86 59 29 77 48 55 73 79 73 87 Higher educated 81 84 73 77 73 57 74 63 71 69 92 79 94 80 72 85 90 81 90 96 89 97

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005 (‘u’ = unreliable estimate)

134 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Graph 5. Most recent training course on computer use, EU‑25, by socio‑demographic background

� � ��

��

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005

higher educated people who may have easier the respondent’s self-assessment on tasks he access to computer training because of their job or she has already carried out. The analytical content. value of the separate items is relatively limited, but Table 2 nevertheless presents the results for Only a minority of about 12% took a course in two selected computer activities: ‘compressing the last year. This last figure includes the 26% files’ and ‘writing programs using a specialised of persons aged 16 to 24 who participated in programming language’. a course during the last year, probably in the framework of a formal education programme 37% of the individuals who at least once used they are enrolled in. a computer, stated that they have already compressed files while 13% stated they have Level of computer skills already used specialised programming languages for writing computer programs. Age, gender and The core of the e-skills module in the 2005 educational level appear to be important factors, survey consists of sets of items related to one’s while the degree of urbanisation or geographical computer and Internet skills, more precisely location are rather insignificant.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 135 Table 2. Proportion of individuals (who ever used a computer) who already carried out specific computer related activities (2005), EU-25, by socio‑demographic background.

Writing a computer program Compressing files using a specialised programming language

All individuals aged 16 to 74 37% 13%

Men 47% 18% Women 26% 8%

Aged 16 to 24 48% 20% Aged 25 to 34 45% 16% Aged 35 to 44 37% 11% Aged 45 to 54 31% 9% Aged 55 to 64 23% 8% Aged 65 to 74 15% 6%

Men, 16 to 24 60% 26% Men, 25 to 54 49% 17% Men, 55 to 74 27% 11%

Women, 16 to 24 35% 15% Women, 25 to 54 27% 7% Women, 55 to 74 12% 3%

Lower educated 28% 10% Middle educated 34% 11% Higher educated 49% 20%

Students 49% 21% Employees 40% 13% Unemployed 29% 12% Retired, other inactive 19% 7%

Living in thinly populated areas 39% 14% Living in intermed. populated areas 34% 12% Living in densely populated areas 34% 11%

Living outside Objective1 regions 38% 14% Living in Objective1 regions 35% 11%

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005

136 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY More relevant from an analytical point of view middle group is relatively stable, but differences is the skill level, calculated by evaluating the are clearer for the groups representing low or number of activities the respondent had already high levels. In Greece, Latvia and Poland, basic carried out. Graph 6 and 7 offer a closer look at computer skills can be labelled as low for more the levels of basic computer skills in Europe19. than one in three persons. At the other end of the spectrum, one can observe that more than half On average (for the countries where data is of the computer users in Luxembourg consider available20), 25% of the persons who ever used a themselves to have high basic computer skills computer are in the category ‘low level of basic – meaning that they are familiar with all or all but computer skills’ – meaning that they are familiar one of the presented computer related activities. with 1 or 2 of the 6 presented activities (Graph 6). A methodological caveat is however that results The medium level represents around 39% and the of self-assessment reporting can be biased by higher skill levels another 31%. The size of the cultural differences in terms of self-confidence.

Graph 6. Distribution of levels of basic computer skills (2005), by country21

��

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005

19 For Graph 6 and 7, no EU-25 could be calculated due to lack of available information. The bar ‘Total’ is based on the data of 17 Member States (DE, EE, EL, CY, LV, LT, LU, HU, NL, AT, PL, PT, SI, SK, FI, SE, UK) representing 58,5% of the EU population. 20 The three levels do not necessarily sum to 100% because people may have indicated none of the items (this could be the case for people who did already use a computer, but a long time ago). 21 The data for Finland is not entirely comparable with the other countries as one item – the simplest activity, namely ‘using a mouse to open programs’ – was omitted from the national questionnaire for 2005.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 137 Where in Table 2, important gender differences Graph 8 on the next page gives the distribution were observed for the individual skill items, no of the persons who already used the Internet significant gender effect is found for the different over the three categories of skills level. skill levels (Graph 7). Also the regional dimension Contrary to the discussion of computer skills, seems to be relatively unimportant. On the other a lot of people only have very basic skills hand, age and educational level turn out to the for Internet use (56%), meaning they have main factors: around 45% of higher educated and carried out only 1 or 2 of the 6 activities they of younger people have high basic skills; more were asked about. In practice (and assuming than 30% of lower educated and of older people some ordinality in the items) this means most only dispose of low basic computer skills. Internet users only know how to use a search Graph 7. Distribution of levels of basic computer skills (2005), aggregate data, by socio-demographic background

��

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005

Level of Internet skills engine and/or how to send e‑mail messages with attachments. Indeed, when looking at the As for computer skills, the information by item is scores on the individual items (not shown in only of a secondary interest, although the results this paper), these two ‘low threshold’ items are are not always as expected, e.g. only 73% of by far the best known. In most countries, the the respondents who already used the Internet scores on the remaining four items are at least (EU-25) said they had already sent e-mails with 40 or 50 percentages points lower, apart from attached files (documents, pictures, etc.), 88% has ‘posting messages to chatrooms, newsgroups used search engines to find information (Google, or online discussion for a’ which seems to be Yahoo!, etc.) and 15% declared already having well-known in the Baltic countries, Poland and created a web page. Hungary.

138 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY This conclusion on the skills level of Internet Table 3 clearly shows that two methods are by far users is also relevant in the i2010 context of the most important, two methods which are also promoting the use of more advanced Internet the most informal or ad-hoc ways to obtain IT services. Providers of such services may need to skills: self-study via learning by doing (relevant pay sufficient attention to the skills level of the for 58% of the persons) and informal assistance target public of new applications. from colleagues, relatives and friends (relevant Graph 8. Distribution of levels of basic Internet skills (2005), aggregate data, by socio-demographic background

��

� � � � � � � � � � ��

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005 From Graph 8, we further see that 29% of the for 60%). Although still the most important ‘tool’, people who already used the Internet are filed in self-study via learning by doing appears to be of the category ‘medium skills level’ and only 7% is a lower importance for older people, compared supposed to have high Internet skills. Not really to other socio-demographic groups. On the other surprising is the observation that the highest hand, this group gives – together with higher skilled Internet users are found in the age group educated persons – relatively more attention 16 to 24. to training courses in adult education centres (whether or not on the demand of the employer). Ways of obtaining e-skills Finally, we have a look at the relation between The question was an optional or pilot question one’s e-skills and the way these skills were in the 2005 household survey, but data is obtained. Table 4 show important differences: nevertheless available for quite some countries22. persons with relatively high skills seem to

22 The data in this section is based on the following countries: DE, EL, CY, LV, LT, LU, HU, NL, PL, PT, SI, SK, FI, SE, UK (i.e. 15 EU Member States representing 56,4% of the EU population) and Iceland.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 139 Table 3. Way of obtaining IT skills (2005), by socio‑demographic background

Gender Age group Educational level Total Men Women 16 to 24 25 to 54 55 to 74 Lower Middle Higher Formalised educational institution 31% 30% 33% 72% 24% 10% 34% 28% 34% (school, college, university, etc.) Training courses in adult education 15% 12% 17% 6% 16% 20% 11% 15% 18% centres, on own initiative Training courses in adult education 23% 22% 25% 4% 27% 31% 12% 23% 33% centres, on demand of employer

Self-study using books, cd-roms, etc. 27% 33% 20% 25% 28% 24% 20% 25% 36%

Self-study (learning by doing) 58% 64% 52% 63% 61% 42% 52% 57% 66%

Informal assistance from colleagues, 60% 60% 61% 59% 62% 54% 59% 60% 61% relatives or friends

Some other way 3% 4% 3% 3% 3% 3% 5% 3% 3%

Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005

Table 4. Way of obtaining IT skills (2005), by level of basic computer skills

Level of basic computer skills Total Low Medium High Formalised educational institution (school, college, university, etc.) 31% 16% 31% 47%

Training courses in adult education centres, on own initiative 15% 7% 21% 22%

Training courses in adult education centres, on demand of employer 23% 11% 41% 23%

Self-study using books, cd-roms, etc. 27% 10% 25% 45%

Self-study (learning by doing) 58% 25% 77% 96%

Informal assistance from colleagues, relatives or friends 60% 51% 99% 48%

Some other way 3% 4% 3% 3% Source: Eurostat, Community survey on ICT usage in households and by individuals, 2005 have obtained these skills mainly in a rather level creates fears or confidence problems to sign independent way, namely via self-study up for such courses. (especially in the sense of learning by doing, but also via books, cd-roms, etc.). From both Table 3 and Table 4, the items in the pilot question seem to adequately cover the Nearly all persons (99%) in the middle skills channels people use to obtain their e-skills as group obtained their skills – amongst other only 3 to 4% of the respondents indicate other ways – via informal assistance from colleagues, ways than those presented by the interviewer. relatives or friends, while this straightforward method is relevant for only half of the low or high skilled computer users. Especially for low The future … skilled computer users, participation in training courses offered by adult education centres is The past years, the focus of policy makers and particularly low. For the data, it is however not analyst has more and more shifted from access possible to identify whether the non‑participation and connectivity to usage and impact of ICT’s on causes the low skill level, or whether the low skill citizens’ life and on business. In this context, the

140 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY need for data on skills or digital literacy can not of questions related to e‑skills or digital literacy be neglected. in the 2007 questionnaire to allow for a better analysis of this subject in the framework of i2010. The model questionnaire for the 2006 enterprise The first discussions with the Member States and survey includes a new question relating to ICT the main users took place in a Task Force at the skills, next to the existing question on what end of November. This brainstorming led to a percentage of employees uses computers at least first list of proposals, ranging from an expansion once a week. In this new question, enterprises of the two self-assessment questions (see above), are asked whether they experienced problems a question on the respondent’s judgement on how in recruiting ICT skilled staff (from basic to suitable his/her skills are for the labour market professional skills) and – if so – whether the or the reasons for not taking computer related problems was/were that i) the personnel with courses, to items or questions relating to e- required skills in the use of ICT applications learning (i.e. the use of the Internet for learning/ is not available or not entirely suitable, ii) education purposes, with or within the framework ICT specialists with the required skills are not of an actual course). As the discussions have available or not entirely suitable or iii) the high just started, the current state is to be considered remuneration costs of ICT specialists. very preliminary … Analogous to the enterprise survey, the final questionnaire will be ready by Currently, the discussions on the 2007 model spring 2006 and first results are scheduled for questionnaire are still on-going, but to serve the autumn 2007. needs for the i2010 strategy, a more detailed module on ICT skills will be included for that Conclusions survey year. The current proposals by Eurostat This paper tried to give an overview of the integrate the above question from the 2006 survey continuous efforts by Eurostat and its partners in a larger set of questions relating to different in the European Statistical System to measure aspects of ICT skills in a business environment: e‑skills. Although the work had to start from whether the enterprise employs ICT specialists, scratch due to a lack of a conceptual framework whether the enterprise offers training to develop or widely agreed standards or definitions, the first or upgrade ICT skills, as well as questions relating results of the 2005 survey presented in this paper to the outsourcing or offshore outsourcing of seem to indicate that the approach used is a valid ICT functions (by type and by ‘geographical one and produces relevant and coherent results. destination’). The module is planned to be finalised by spring 2006, data should be available This strengthens the statisticians working on in autumn 2007. this subject to continue their work on measuring digital literacy or e-skills, as requested by the The 2006 model questionnaire for the household main users in the context of the Commission’s survey will keep the 2005 module on e-skills. But i2010 strategy to boost the information society this survey too will include a more detailed set and more specifically the digital economy.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 141 e-skills and their measurement

Dr. Matthew Dixon Labour Market Adviser to CEPIS1 and SEMTA Visiting Research Fellow SKOPE, University of Oxford

Workshop ‘ICT impact on the knowledge-based Society’ 8 December 14:30 – 17:30

Executive Summary skills – of growing importance in a global ICT market, including the monitoring and impact of The paper reviews the current state-of-the-art outsourcing (both intra- and inter-nationally), of measuring e-skills, both nationally and at and mobility (both physical and virtual). the European level, and flags key hurdles to be The paper then turns to the challenges for skills overcome, if more meaningful tracking of human capital is to be achieved. measurement – both generally and in particular for ICT skills – of which there are a number! In contemplating how these different hurdles It summarises the needs of skills data users, can be overcome, the paper emphasizes the covering both employers and policy analysts, need for approaches that focus on the specific and presents the major classes of e-skills, as (policy or enterprise) needs. In considering viewed from EU Commission-led and OECD-led the crucial importance of agreed occupational perspectives, noting the considerable structure classifications, the paper summarises recent present within both IT User and IT Practitioner (or Professional, or Specialist) types of skill. It developments in the efforts to improve common reports the nature of both general workforce- and understanding and coherence of structure at the specific IT- survey datasets, and presents a case European level (through work on a “European study addressing the development of different ICT Skills Meta-Framework”), and concludes aspects of IT Practitioner Skills, showing what with some key priorities that could help improve can be achieved from general survey data our ability to track the people (not just ICT at the European level and for Member State practitioners) who embody the human capital comparative purposes. The paper touches on which the growing knowledge economy will on certain “geographical” dimensions of e- be built.

1 Council of European Professional Informatics Societies

142 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Introduction population as a whole. EU-level interest arose in particular from the very significant shortages of Few can have failed to recognize the way in which ICT Practitioner skills in most Member States in Information and Communication Technologies the late 1990s, and European level work in this (ICTs) have changed our world over the past few area was initiated, in particular by the European decades. ICTs handle what has become one of Commission (D-G Enterprise), in 2001 with the key resources in our world. As the enabling the establishment of an ICT Skills Monitoring mechanism for the vast majority of information Group of officials from Member State industry flow, ICTs are – in the workplace - acknowledged ministries. There followed: to play a fundamental role in the operation of both public (where there is a need for open, effective • a European e-skills Summit in October flow of information between providers and users) 2002 in Copenhagen under the Danish and private (as the lubricant for the operation of Presidency, markets) sectors. • the establishment of a European e-Skills Forum early in 2003, bringing It is thus unsurprising that our efforts to capture, together Member State ICT/skills harness, share and deploy the information policy representatives and other major resource through ICTs represent an all- stakeholders, pervasive activity in today’s world. And while • the production, in the summer of 2004, we see, and experience every day, many of their of a Synthesis Report of the work of the problems and limitations, as well as the deeper Forum, challenges that arise from their influence, ICTs • a major conference in September 2004 in have established an enormous presence for us Thessaloniki, and all - a presence on which we often depend more than we might care to. The word ‘revolution’ • a number of projects arising from the priorities identified by the e-skills Forum. is therefore not ill-placed for the impact ICTs have had, and most would accept the arrival of an ‘Information Age’. What are e-Skills?

So, despite much ‘hype’, ICTs are with us – It is interesting to note that what might be viewed presumably to stay – and it is necessary for us as a ‘starting point’ for policy development in this to come to terms with them, at the very least by area, namely an agreed set of definitions of what e- understanding enough about them and how they skills actually are, claimed non-trivial amounts of work to be able to explain the basic role they attention from the e-Skills Forum. The outcome have – both in relation to work and to our lives of these discussions was the specification of three more generally – and deploy and use them for our distinct types of e-skills: benefit. 1. ICT Practitioner Skills (sometimes called It is understandable, therefore, that there has ICT Professional or ICT Specialist skills); been growing interest over recent years in the 2. ICT User Skills; and implications of all this on skill requirements. 3. e-Business Skills. The ‘e-skills’ debate has grown steadily, not least at the European level, and since the end of the The detailed definitions of these three types last century there has been increasing interest are given in the e-Skills Forum Synthesis in the policy implications of the growing need Report (EeS-F, 2004), but the key principles of for e-skills among the workforce and indeed the importance are that:

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 143 1) ICT User and Practitioner skills are It is important to note that recent OECD work importantly different: in essence Practitioner in this area has also identified three categories skills involve a person working full time in ICT, of e-skills, but that these are different(!). The for the benefits of others, whileUser skills involve approach has distinguished between: people making use of ICT tools for doing their own jobs (not ICT). A fundamental implication of this • Basic ICT (user) Skills, is that there are ICT Practitioner occupations, but • Applied (or Advanced) ICT (user) Skills, there are not ICT User occupations. The fact that and many experienced (often ‘expert’) users go on • Professional ICT skills (corresponding to to work as ICT practitioners does not mean that Practitioner skills) these two skill-sets are the same, or that there is any ‘overlap’ between the two types (progression and level should not be confused with type). The author’s conviction is that this classification, while appealing in terms of apparent simplicity 2) there is a considerable amount known and ‘linear’ progression, suffers from two about each category of skills – and for Practitioner serious flaws, since it a) misses the fundamental and User skills detailed frameworks have been distinction between user and practitioner types of developed that specify skills/competences across skill, and b) asserts a simplistic division between functional areas and at a number of levels. two ‘levels’ of user skills that does not have much meaning in reality. 3) e-Business skills are (also) about the use of ICTs, but, unlike ICT user skills that describe In particular, as explained in point 2) above, the needs of the individual in working with a within the user skills domain (as well as within specific ICT device (normally a PC), e-Business the practitioner domain), there is a range of skills relate to the effective use (and indeed functional activity to be considered at a number exploitation) of ICTs for an organization, and are of levels, and it is therefore not helpful to attempt therefore a requirement for senior management. to ‘squeeze’ this reality into just two ‘levels’. While such skills are generally taken to relate Table 1 shows the structure of the IT User Skills to the exploitation of the opportunities afforded Framework developed by e-Skills UK for their e- by the internet, it applies equally to take-up of skills Passport /ITQ, and shows that – even for the any type of ICT enabling-technology. With the generic functionality of standard office software growth of other major areas of activity (e.g. e- (in particular the Microsoft Office environment Health, e-Science, e-Learning, etc.), this category many of us are used to) skill/competence exists at of e-skills might perhaps more generically be a number of levels – in this model ranging from viewed as ‘e-Leadership’ skills. ‘Inexperienced’ to ‘Super User’.

144 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Table 1. Structure of an IT User Skills Framework

(e-skills UK: (updated in 2005)) (skill/competence requirements are specified in all shaded ‘cells’) Level:------‡

‚Functional Activity· Foundation Intermediate Advanced Super User Inexperienced Systems: Use, improve and maintain systems: …………………… …………………… …………………… …………………… …………………… Operate a Computer …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… IT Trouble-Shooting for …………………… …………………… …………………… …………………… …………………… Users …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… IT Maintenance …………………… …………………… …………………… …………………… …………………… for Users …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… IT Security …………………… …………………… …………………… …………………… …………………… For Users …………………… …………………… …………………… …………………… ……………………

Communication: Access and share Information …………………… …………………… …………………… …………………… …………………… Internet and Intranets …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… e-mail …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… ……………………

Use Application Software: …………………… …………………… …………………… …………………… …………………… Word Processing …………………… …………………… …………………… …………………… …………………… Software …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… Spreadsheet …………………… …………………… …………………… …………………… …………………… Software …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… Database …………………… …………………… …………………… …………………… …………………… Software …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… Website …………………… …………………… …………………… …………………… …………………… Software …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… Artwork and Imaging …………………… …………………… …………………… …………………… …………………… Software …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… Presentation …………………… …………………… …………………… …………………… …………………… Software …………………… …………………… …………………… …………………… ……………………

Make Effective Use and Evaluate: …………………… …………………… …………………… …………………… …………………… Make Selective Use …………………… …………………… …………………… …………………… …………………… of IT …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… …………………… Evaluate the impact of IT …………………… …………………… …………………… …………………… ……………………

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 145 The same range of competence levels also exists 2005). And if we are to examine seriously and for the use of each of the more specialized types try to measure, monitor and understand e-skills, of IT system, for example Computer Aided we need the same kind of statistical evidence that Design (CAD) software, Engineering Simulation we use in support of policy-making elsewhere. or Design packages (e.g. Finite Element Analysis), or Advanced Statistical Analysis tools Let us now consider the needs of different users – even Manufacturing- or Enterprise- Resource for which statistical data in relation to e-skills is Planning (MRP/ERP) systems. And each of us required. In particular, there are needs from both who use these different types of software/system individual organizations (i.e. employers from the will have greater or lesser skills/competence private and public sectors) and governments (in levels in each. Thus it is both possible and likely relation to policy development): that an experienced design engineer might have considerable expertise in ‘driving’ a CAD system, Examples of questions relating to some significant but rather modest ability in using certain elements business (employer) needs would include: of Microsoft Office. Attempting to track ‘basic’ as opposed to ‘applied’ or ‘advanced’ use of IT systems is therefore neither easy to specify nor - how do I find out enough about ICTs to meaningful to interpret. know in broad terms what I need to do for my business? Even the use of the ‘sector-specific’ concept in - how do I assess the ‘offer’ from ICT relation to identifying a potentially interesting suppliers? set of ICT user skills has real limitations. For - how do I recruit and retain knowledgeable, example, Computer Aided Design is now a dependable experts (who understand core activity for any work involving 2-D or 3-D enough about my business), especially (spatial) design, but (as testified to, for example when there is a shortage of skills supply? by the breadth of vertical-market penetration of - am I paying (offering) enough to get the AutoCAD) this activity transcends engineering, best ICT Practitioners? textiles, petro-chemicals, and many (but not all) - how do I manage my team of ICT other sectors (SEMTA, 2005). It is not clear Practitioners – normally within an ‘IT what aggregate estimates for the skill levels in all department’? these sectors and for the cross-sectoral (albeit not pan-sectoral) generic tools would mean – either - how do I get (all) employees (who need to) to the enterprises involved or for skills policy up to an adequate level of using ICTs? considerations… - how should I assess whether to outsource the ICT function (perhaps even overseas)? e-Skills Statistics: the need Some questions relating to significant policy It is clear that in the Information Age we need to needs would be: be able to use ICT effectively in our work and beyond, and evidence is growing that ICT skills - what is needed to ensure sufficient ICT – both user and practitioner – influence effective skills in the economy, and society as a exploitation of ICTs, which can lead to an increase whole? in productivity, quality improvement and/or - what are those skills, and how well are they competitiveness. Evidence for the impact of ICT being supplied currently in the education investment on productivity is continuing to grow, (supply) system and learning-provision as the recent ONS-led work shows (Clayton, marketplace?

146 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY - what (for the different types) should be the Meta-Framework (CEN/ISSS, 2005) - an effort role of the state in promoting/supporting by a group of key stakeholders with support ICT skill development? from D-G Enterprise to explore whether greater - what additional learning opportunities are coherence and common understanding for these needed with the education system? skills might be possible at the European level. - ‘how are we doing c.f. others’ (other Member States, North America, etc.) in (There is currently little in the way of elaborated relation to ICT skills? frameworks yet available for e-Business/e- Leadership skills, although these are increasingly - what is the experience with policy recognized to be an important element within interventions in this area so far? more general Management and Leadership skill - what role will skills availability play requirements) in influencing inward investment?, and conversely e-Skills Labour Markets - what will be the impact of ‘offshoring’ on demand for ICT skills?, and, of course, In the context of skills analysis and policy-making, - what skills does the citizen need for effective we make liberal use of the term ‘Labour Market’. access to e-Government services? However, it may be that our understanding of the realities of what Labour Markets are, and how they Responding to these questions in a sound way operate, are more limited than our understanding depends on better understanding of certain of the behaviour of product- and service- markets. underlying principles about measuring skills. As A labour market could be viewed as a ‘place’ where indicated, the types of skill required vary with our labour (or skills) supply and labour (or skills) relationship to ICT (in particular as individuals demand are reconciled, and where in principle - whether ICT practitioners or using ICT tools the allocation of skills resources takes place in a within our (non-ICT) jobs - or in relation to its self-regulating way, through the use of the price deployment and exploitation for the organizations mechanism (i.e. the wage or salary). However, we work in - as business managers) there are many pitfalls in the smooth operation of labour markets in comparison with product This is an area of considerable complexity: as markets: monopsony (or more likely oligopsony - already indicated, even defining e-skills/ICT analogous to monopoly and oligopoly in a product skills is not easy, and within each category a or service market) can cause reduced levels of considerable amount is known about the structure competition and so effectiveness of the resource of skill needs. In addition to the IT User skills allocation. And there can be market failures in framework structure shown in Table 1, there are relation to the efficiency or equity of the market. a number of highly developed ICT Practitioner Certainly there are – at least for ICT Practitioners Skills Frameworks, including three major ones – examples where the price mechanism does not in use within larger EU Member States: the appear to have worked effectively – or even very CIGREF2 Nomenclature (France), the Advanced much at all - in the ‘market self-regulating’ way IT Training System (Germany), and the Skills expected. Framework for the Information Age (UK). These, and ICT Frameworks in general, are examined The interests of the agents buying and selling in some detail in the proposed CEN Workshop labour/skills – employers and individuals – are Agreement (‘CWA’) on a European ICT Skills in principle in direct opposition within a labour

2 ICT Club of Large French Companies

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 147 market. However, in reality, there is a range of the set of skills relevant for that occupation. ways in which the relationship between employers This is particularly important to remember in and employees transcends considerably the simple relation to ICT Practitioner skills, since not ‘opposing–interest’ model that may be meaningful all ICT Practitioners work in the ICT (supply) in relation to product trading that proceeds based Industry. Indeed it turns out to be the case in on ‘single-event’ transactions. In particular most (if not all) countries that the number of ICT there is the essence of the ‘extended duration Practitioners who work in User Organisations relationship’ between employer and employee, (e.g. Banks, Manufacturing companies, over which a number of different dimensions of Public Administration, the Retail sector, etc.) the relationship can develop beyond the narrow exceed the numbers working in ICT (supply) financial (‘rent’) element. Perhaps the more Companies. The reality is that the career of an interesting markets to be examined in this context ICT Practitioner would generally include periods are the recruitment market (where there is a single- in both the supply and user sides of the economy, event transaction, in which the labour market and such progression is beneficial both to the (economic) valuation of the - human - resource individual and to the employers on both sides. is most central), and the training market, which (For example, an ICT supply company selling (as a secondary market) operates through the into, and developing systems for, a particular demand and supply of ‘skills-raising’ – assumed user sector – e.g. Manufacturing – would be to be the mechanism by which skills supply can interested in recruiting an ICT Practitioner with be increased for the (primary) labour market. (It expertise and experience in that sector). This is also, of course, necessary to consider the skills- sectoral/occupational separation is particularly raising/supply channel of the formal education important in that many with interest in the skills system, for which governments are in many cases area start off by assuming, incorrectly, that attempting to improve alignment to employer skill skills (and the corresponding labour markets) needs, sometimes using quasi-market incentive are sectorally-based. On reflection it quickly mechanisms in the process.) becomes evident that, from an ICT skills point of view, we are interested in what ICT Practitioners Furthermore, as with product and service markets, do (and need to be able to do) when designing (… the effectiveness of the market mechanism developing, installing, operating, maintaining, may be affected (constrained) by its dynamic etc.) ICT systems wherever they are, and we characteristics – the fact that situations can occur are not interested in the skills of (e.g.) the where the ‘proximity to equilibrium’ assumption accountants, cleaners, and HR people within ICT of the ‘well-behaved’ market does not hold: where companies. These realities all contribute to the large transients can occur which strongly benefit fact that the main area of data availability and one side or the other. statistical interest in e-skills is ICT Practitioner skills, above all since these a) have genuine It is important to recognise that, in particular (meaningful) labour markets operating, and b) from a skills point of view, labour markets are can be tracked using statistics, precisely because occupational. That is to say that employers they correspond to occupations. recruit people to do a job, for which a set of skills are required, and assess the strength of The key sector/occupation relationships within the applicants mainly in relation to whether they structure of the workforce can best be understood possess (or are felt to be able to quickly acquire) visually - as shown in Table 2.

148 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Table 2. Sector <-> Occupation relationships Skills or e-Business (e-Leadership) skills. For within the workforce ICT User skills, a certain level of competence with standard office software is generally taken ICT (Supply) ICT User for granted in applicants for office-based jobs. Thus use of Microsoft Office is fast becoming Companies Organisations an employability requirement, rather than a ICT Practitioner skill for which there is a wage premium. For e- n n Occupations 1 2 Business skills, it is generally a question of senior management briefing themselves – by whatever Other Occupations n n 3 4 means - with the additional understanding about the potential capabilities of current and emerging ICTs, and then creatively relating this to the While the contents of the firstcolumn (n n ) are of 1 + 3 realities of the business and market context in interest to the ICT Supply sector, since they cover which they are operating. the total employment levels in ICT companies, it is the first row (n + n ) that matters for all 1 2 Indicators ICT Practitioner skills matters. As indicated, in most economies n2 > n1, although the trend in There are a number of statistical indicators that many cases has been towards greater growth in throw some light on the state of e-skills within the supply sector employment: it is worth noting an economy. The recent report on the supply and that the effect of each new outsourcing contract demand of e-Skills in Europe carried out for the (within an economy) is generally to ‘shift groups Commission (also D-G Enterprise) by the Rand of ICT Practitioners’ from n1 to n2. corporation (Rand Europe, 2005) singled out the following eight indicators in relation to what As mentioned, it is not possible in the same could – in principle – be measured in the labour way to speak of Labour Markets for ICT User market (for ICT Practitioners).

Table 3. Indicators relevant to ICT Practitioner Skills (Rand Europe, 2005)

Tracking Sources/Details Indicator (an element of): BISER (2004) collected information on number of establishments that experience Unfilled vacancies demand unfilled vacancies for selected European regions Job vacancies demand At EU level no data was found. Replacement demand demand Estimates of 2.5%, 3% and 9% of stock are used

Offshoring activities demand No data sources were found EU LFS data provides information on numbers of unemployed in each occupation. Unemployment supply However, low sample size can make estimates less reliable, and not all unemployed can be considered as constitution real potential supply. 2-digit ISCED EF numbers classify the academic disciplines (the most relevant one is Number of graduates supply for ‘Computing’, and data on stock and flows for Tertiary Education courses are held for each by Eurostat for each Member State. ITAA and CompTIA provide information on what types of courses are taken, which Training and certification indicates relative importance of topics (however most data on commercial volumes is supply volumes confidential). Eurostat CVTS shows number of hours spent on computer courses (all levels) (both) Stock of ICT practitioner Eurostat LFS data is readily available for 3 digit ISCO codes and provides information (lower of supply occupations on industry, gender, age distribution and nationality of labour force and demand)

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 149 It is important to note that several of these But there are other needs for occupational indicators are currently ‘theoretical’ in the sense frameworks for ICT practitioners beyond those that no meaningful, comparable data exists for of national statisticians (as custodians - classifiers them in most Member States. Thus we are quite - of the workforce as a whole). For example, a a way from being able to use them directly at the need for classifying occupations for: international level. • recruitment activity; Occupational Frameworks • salary surveys; • promotional material for work in the area; It is important to recognize that skills can generally • managing technical teams (skills not be measured directly: the most relevant ‘second management) best’ indicator relates to tracking occupations • career development; and (almost all of the above indicators relate to occupations). And this leads directly to the main • assessment of individual skills/ challenge for such statistics, namely that of gaining competences. agreement from the key stakeholders involved on valid and meaningful occupational classification While these may seem a little less relevant to frameworks. When measuring occupational the world of policy, in reality most are, and it statistics, it is necessary for the working people is interesting to note that separate occupational/ being ‘counted’ in surveys to be allocated to an skills/ competence classifications have been appropriate occupational category. For a number developed for each of these, in relation to ICT of reasons, a range of occupational- (or skills-, Practitioners, within the United Kingdom in sometimes competence-) classifications for ICT recent years. Practitioners have been developed (even within a single country) for different purposes. Those As indicated, occupationally-oriented classifications with which statisticians – with the Frameworks specifying ICT Practitioner skills need for a ‘bird’s eye view’ of the economy (in have been developed in different EU Member this case the workforce as a whole) – are generally States, and it is interesting to note that the familiar are, of course, the official occupational structure of these is generally two-dimensional classifications developed in each of our countries. (in the sense of including both ‘horizontal And these have been developed over time within descriptors’ specifying skill/competence each of our cultures, responding in particular to requirements in terms of sets of work activities/ the need to cover the whole range of occupational functions, and ‘vertical descriptors’ specifying activity with a certain, manageable, number of the relevant levels for each set of functional categories. In most cases within the EU Member competences). However, while the three main States this has resulted in the emergence of a set of models in EU Member States purport to cover some 4-10 categories to cover ICT Practitioners. the same overall scope of ICT Practitioner work, Of course trying to track the main categories of ICT there are – as well as certain similarities – a Practitioners has been a ‘moving target’ over recent number of important differences between them. decades, as the structuring of ICT (practitioner) As already mentioned, work over recent months work within organizations has responded to the to better understand this situation in terms of a seemingly relentless waves of innovation and so possible agreed European-level model has led to of new enabling technologies appearing in ICT, a number of recommendations that could help and we cannot hold out much hope that maturity clarify and distil a ‘European ICT Skills Meta- will be reached in the near-term. Framework’ (CEN/ISSS, 2005).

150 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Returning to the ‘official’ classifications of validity of the process for 4 Member States, and occupations used across the EU, the same the results were not encouraging, either for ISCO ‘reconciliation’ challenge applies, in the sense that or in relation to the validity of mapping across most Member States use their own classifications, to a common sectoral classification (NACE) and the ‘mapping across’ between them is often (Huws, 2003). The author is convinced that the not particularly straightforward. All Member imperfections in the coding process to ISCO- State statistical offices now supply data from their 88(COM) consist of two distinct components: national Labour Force Surveys to Eurostat every quarter. This data includes ‘slices’ across the • the inherent challenge of mapping across to workforce by occupation. In order to make the a different classification (which cannot in occupational data comparable, it is necessary to principle be overcome), and cross-map the national occupational frameworks • the practical (human) errors, arising from to a single ‘European-level’ classification. The a combination of imperfect briefing and classification used for this is ISCO-88(COM). administrative processes, and an inherent lack The International Standard Classification of of incentive for commitment to data quality Occupations (ISCO) was developed by the assurance (since this process brings no direct International Labour Organisation. Like all benefit to the national statistical offices that do classifications that are intended to track what the work!). is going on in the market-place, ISCO needs to be updated regularly, since the structure of what The use of ‘Qualification’ as a supply proxy: happens in the world of work inevitably changes what are Skills/Competences? (not least caused by the considerable impact of ICTs on changes in working practices!). Traditionally, for macro-economic purposes, Unfortunately the updating of any international skills have generally been measured from classification system is not something achieved Labour Force Surveys with the use of indicators quickly. Indeed the date implicit in the ISCO of the ‘Highest Qualification Achieved’. This title confirms that the current classification stems is understandable, although it suffers from from 1988 (in fact the years leading up to 1988!). a number of serious flaws, in particular in Although there has been comparatively little relation to ICT skills. It brings an inevitable interest (or resource commitment) over recent emphasis on qualifications awarded within the years it has now been decided that ISCO should formal education system – and this stresses be updated, and the process is underway with the Knowledge component of Skills or overall the intention of the next update being issued in Competence, since qualifications awarded within 2008. It must be recognized that the challenge the formal education system are (inevitably) very of getting agreement on the new update is likely largely based on assessments of knowledge (as to be even greater than that of getting agreement opposed to skills or competence). on such things at the EU level, not least since the spread of occupations within countries While it could be argued that trends towards at different stages of development can show a Knowledge Economy would in principle significant differences. strengthen the importance of the Knowledge component of overall Competence, work on The mapping of (Labour Force Survey) clarifying the underlying concepts of Lifelong data gathered under national occupational Learning within the EU (e.g. CEDEFOP, 2005) classifications across to ISCO is undoubtedly a now recognizes that, in order to prove relevant in problem. The recent STILE project explored the relation to employer needs, qualifications need to

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 151 recognize broader outcomes of learning. These structures relating to learning (and in particular are articulated, in particular within the proposed formal education) supply arrangements, given European Qualifications Framework (EQF) (EU the particularly great diversity of traditions and Commission, 2005), as comprising Knowledge, institutional arrangements for education around Skills and wider Competences3. the European Union.

The emergence during 2005 of proposals for the The approach taken in the CEN Workshop EQF – based on a proposed set of (8) Reference Agreement in response to this challenge was a Levels specified as learning outcomes articulated three-tier contribution: in terms of these ‘components of competence’ – posed an intriguing new challenge for work 1) A ‘first-cut’ interpretation of the generic already underway on a European ICT Skills reference-level descriptors proposed for the Meta-Framework. As mentioned, this work aimed EQF into articulations of the competences at exploring the possibilities for the development required by ICT Practitioners. While not of a European-level two-dimensional framework extending into any ICT-specific ‘technical’ for specifying ICT Practitioner skills. That competencies, the approach nevertheless Framework is assumed to be a) adequately picked up a number of generic competence consistent with the existing ICT Practitioner areas acknowledged (and specified) in Skills/Competence Frameworks already being existing ICT Practitioner Frameworks. These used within certain Member States, and b) included: cognitive competence, functional agreed by the ‘owners’ of these and by other competence, personal competence, social key stakeholders to provide a basis for policy competence and professional or vocational measures to support greater ‘mobility’ of ICT competence. The proposed set of ICT Practitioners around the EU. Practitioner competence-level descriptors is recognized to need considerable scrutiny The injection into this work of the need to and refinement. ‘explore how the proposed EQF could be applied to (or interpreted in) the world of ICT Practitioner 2) A first-cut at a reasonably simple structured work’, posed some interesting questions. In grouping of the main types of existing particular, the focus of the work on the demand- ICT Qualifications, allocating the different side – clarifying employers’ skill/competence types of qualification at likely levels needs – now needed to be extended into the in accordance with the EQF Reference complex world (and politics) of Learning supply. Levels, and separated into three basic types As was recognized in advance in debates within of qualification, depending on the relative the European e-Skills Forum, the possibility importance of Knowledge, Skill and of gaining agreement in the specification of Competence being assessed. The three types employers’ competence needs (in relation to an are: Competence-based Qualifications, already global (ICT) industry and a growing Skill+Knowledge-based Qualifications internationalization of the relevant labour and Knowledge-based Qualifications (in markets) was likely to be considerably greater particular of the type assessed within the than that of gaining agreement on principles and formal education system).

3 It is important to recognize that the major implication of this approach is that Qualifications can (and increasingly do) extend beyond the largely knowledge-based assessments acquired in the early years of life as we go through the formal Education System. It is equally important to recognize that this major paradigm shift is not one that will be quickly digested and adopted (or even, sometimes, accepted) within communities of influence, in particular of course within the tertiary education community. One reason for this is that qualifications involving assessment of more practical – experience-based – skills have traditionally been vocational qualifications – the object generally of lower esteem from these communities, and often rather widely within society.

152 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 3) An attempt to examine in some depth But perhaps the most important insight gained the relationship between the concepts of by this deeper scrutiny was the recognition that Competence and Qualification. It became Competence Frameworks were NOT the same clear from this exploration that: thing as Qualification Frameworks essentially since they attempt to structure things that are on a) the concept of Competence is not opposite sides of the labour market: Competence consistently understood or widely (or Frameworks attempt to specify the capabilities fully) accepted within the communities that employers state they need for effective involved in the supply of learning within performance in the workplace (i.e. specifying different Member States; labour market demand), while Qualification Frameworks position types of assessment and b) the concept of Qualification is not viewed certification that give an indication of what in the same way within all Member the individual understands and can do (almost States; and perhaps most importantly always in a rather broader way), so specifying labour market supply). Where it is meaningfully c) the mutual positioning of the two is seen possible to assess competence directly (and distinctly differently in some Member the UK’s National Vocational Qualification States. arrangements and Germany’s new Advanced IT Training System could be viewed as fully Competence – a term generally arising from competence-based qualifications) this still does articulation of employer needs – is undoubtedly not mean the Qualification is the same thing as growing in recognition within most countries, the Statement of employers competence needs. as representing, in a sense, the ‘target’ at which The author believes that the underlying reality learning supply needs to aim. However, the here is that competence becomes the currency in ‘debate’ between those who view learning as in which the labour market operates. support of the need to strengthen the competence of the learner to do what employers’ need and In terms of existing statistical indicators at the those (often within the Tertiary Education international level, while ISCO brings a measure community) who tend to view Knowledge as the of the occupational competence specification4, most fundamental concept in relation to learning is the best we have at present as an indicator for evidently at different stages in different countries. qualifications is ISCED (the International Standard Classification of Education – latest The examination of the mutual positioning of version 1997). Most commentators show qualification and competence carried out as distinct scepticism of the value of ISCED, not part of the work on a European ICT Skills Meta least since it is probably more validly viewed Framework led to the conclusion that, even where as an attempt to measure – in as internationally- qualifications involved assessment of abilities comparable way as is possible – the main well beyond knowledge, they could never stages of the formal education system (in use involve direct assessment of specific practice/ in industrialized countries). However, even if performance in the working context, and thus it were possible to ignore the general constraint Competence and Qualifications would always be of this perspective restricting analysis to the separate concepts. largely Knowledge-assessment measures that

4 as will be seen, each ISCO category has a short specification of what eoplep in each occupation do – reasonably similar in style to the - top level - competence statements specified, for example, within theNational Occupational Standards developed in the UK for the vast majority of occupations

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 153 come from formal education, this classification from new primary research - specific surveys has an additional – very serious – limitation in the commissioned for a particular purpose – in this case of ICT Practitioner skills. This arises from case in relation to e-skills. As already indicated, the comparative ‘disconnect’ between the areas the fact that ICT practitioner skills are related of understanding needed for work in different directly to occupations means that there is very aspects of ICT Practitioner work and the Body of considerable data available, since certain general Knowledge that has so far developed in relation to surveys include evidence on occupations. For Computer Science – the largest disciplinary area ICT user and e-Business/e-Leadership skills, data of ICT courses in tertiary education by some way. from general surveys is very limited, since these There are a number of reasons for this, but the skills are not directly related to occupations, effect is that many employers in some Member and only represent part of the skill-sets needed States eschew Computer Science graduates when for occupations. Thus data on ICT user and e- recruiting for ICT practitioner work (an insight Business skills is generally only available from into remarkable differences in employer graduate- specific surveys carried out with questions related recruitment practices in different Member States to these included. is given in Steedman et al (2003)). In fact, the vast majority of people working as ICT Practitioners, It is particularly important to recognise that the while in most cases graduates, do not hold degrees comparatively dynamic nature of the labour market in ICT subjects (as well as the disconnect problem, involved for ICT Practitioner skills means that the the other reason for this is the continuing shortage, value of individual specific surveys is seriously until recently, of ICT degree course capacity in limited. With a history of rather significant comparison with demand levels). labour market ‘swings’ over the lifetime of ICTs, and with the relentless innovation both in the Thus however useful ISCED has proved in enabling technologies and the ways we organize relation to statistical efforts thus far to relate Skills ourselves to handle them, the ‘half-life’ of the to Competences at the macro level (Lemaitre evidence arising from a one-off e-skills survey is (2002), explored correlations between ISCED particularly limited. The anecdote that Microsoft and ISCO levels in LFS data), talk about their ability to plan meaningfully being in terms of small numbers of quarters rather than • there is no doubt we need a better small number of years appears to support this. indicator of skills that ISCED (in This reality provides powerful support for the principle something drawing on the desirability of getting all we can from regular principles underlying the EQF might existing surveys, in particular where these are form the basis for something more carried out more frequently (e.g. quarterly). meaningful), and The main general survey source for useful data • there are particular flaws with trying to on ICT practitioners is therefore national Labour use a ‘highest-academic-qualification’- Force Surveys (LFSs). As indicated above, all type of indicator for skills supply in the Member States of the European Union now supply case of ICT Practitioners. data from their national LFSs to Eurostat on a quarterly basis. Labour Force Surveys are regular household surveys with reasonably significant Data Availability sample sizes, which involve a number of questions As with many policy application areas, there are about a range of aspects of the respondent’s work two broad kinds of data source: those arising situation. The questions produce data on such from regular general surveys and those resulting things as the respondent’s:

154 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY • gender, occupational classification (SOC 2000), the • age, following: • income, • ICT Managers • employer (including its number of • IT Strategy and Planning staff employees and its main business activity), • Software Professionals • employment status (e.g. employed-status, • Operations Technicians self-employed, or unemployed, and – if • User Support Technicians employed - part-time or full-time, etc.), • Database Assistants • Highest Educational Achievement (in terms • Telecomms Engineers* of a qualification), and • Line Repairers/Cable jointers* • Occupation. • Computer Engineers

In addition to the questions asked every quarter, Case Study: what can be done there are – once a year - certain additional questions, which include information about It will, perhaps, be helpful for the above principles the respondent’s employment one year before. to be illustrated through the example of the work These can provide valuable information about carried out by the author on IT Practitioner occupational and/or sectoral mobility. Skills in Europe (CEPIS, 2002), building on the seminal Skills99 study in the UK (AISS/ It is therefore possible, in principle, to cross- ITNTO, 1999). The study was essentially an tabulate data on these variables, which – inter alia initial exercise in ‘data-mining’ from the Eurostat – enables a range of interesting characteristics holdings of Member State Labour Force Surveys. to be assembled about those working in any As explained, this involved using the relevant one occupation. Clearly the more detailed the categories from the occupational classification breakdown, the more limited the statistical validity used in this dataset – ISCO-88 (COM). The of the cross-tabulated estimates. However, scope of occupational activity covered by the two ‘data-mining’ within Labour Force Surveys is categories, ISCO 213 and ISCO 312, is specified particularly encouraging for ICT Practitioner as follows: occupations, since the size of the ICT practitioner • Computing Professionals (ISCO 213) labour force within most EU Member States is ‘conduct research, plan, develop and improve now sufficiently large as to allow most st1 -level computer based information systems, software cross-tabulation estimates to be very robust, and related concepts, develop principles and and indeed for 2nd-level cross-tabulations to be operational methods as well as to maintain .. generally valid. systems .. ensuring integrity and security of data’. An example of what is involved here can be seen • Computer Associate Professionals (ISCO 312) from the case of the United Kingdom, where the ‘provide assistance to users …., control and estimated total population of ICT Practitioners is operate computers and peripheral equipment around a million, within a workforce of a little and carry out limited programming tasks over 30 million. ICT Practitioner occupations connected with the installation and maintenance comprise, within the UK’s current national of computer hardware and software’.

* these two occupations would be viewed as from the ‘telecomms’ world, representing the ‘C’ part of ICT, while the others would generally be viewed as ‘IT’ Practitioner occupations.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 155 While at the time the work was carried out quarterly employment levels must, while not constrained data was not available for all Member States, it by, nevertheless be coherent with, recent trends, was nevertheless possible to gather comparable and generally changes will be incremental, data over a number of years. The development of rather than involving ‘step’ (or even ‘dramatic’) employment levels over recent years, as well as the changes. estimates available from cross-tabulations of each occupation against the other variables, provided The Conference presentation shows both the a considerable amount of valuable information recent trend data (the details are available at http:// about what is going on in the ICT Practitioner www.cepis.org/download/cepis_report.pdf) and labour market. Specifically, the work showed a comparison with the actual employment levels – in particular for ISCO 213 - the recent trends in that ensued over the following two years. The relation to breakdowns of gender, size of employer, deviations of the development path from the age-bands, employer sector, employment status four postulated trajectories provide much food (whether employed or self-employed), highest for thought, which should – as with all ‘future- qualification level, status one year before, and gazing’ - be fed back into the next set of ‘foresight’ training received within last month. analyses.

This data enables policy response development Geographical Dimensions of e-skills to draw on: It is impossible to talk about e-skills, or indeed a) (rough) levels of the variables (e.g. possible ICT itself, without a mention of the growing inter- initiatives to attract more women into ICT nationalisation of this activity. It is, of course, practitioner occupations), the very success of ICTs that has in recent years b) quarterly trend information (from which both produced the remarkable global infrastructure the nature of the market movement and, in that can reduce (in some ways eliminate) the principle, the impact of policy response can effect of distance. Sciadas addresses some of be monitored), and the issues around such major impacts here in his c) broad comparison with the labour market paper (Sciadas, 2005). position in other Member States (e.g., including potentially, in terms of the dynamic patterns, In considering these international aspects, we assessing relative labour market flexibility). need to recognize a number of key realities:

The CEPIS report went one step further than - the ICT market is already global, with a showing recent trends. It explored four possible comparatively small number of very large future developments of employment levels using multi-national corporations already operating a range of plausible growth assumptions, starting internationally (as are the ‘majors’ from (as was expected at the time of the study) with a different user sectors); clear downturn from 2001, followed by recovery - that, as with a number of business processes, of employment growth at annual rates of 2%, 5%, organizations are increasingly considering 10% and 15% (taking into account that growth in approaches that involve Outsourcing (both these indicators in the late 1990s had averaged within national boundaries and beyond); over 10% for the EU). - that, with the growing convenience of international travel and general reductions This approach, while not involving scenarios of barriers to entry, there is a corresponding in any detailed sense, did provide a basis growth in the international mobility of ICT for a serious policy look-ahead, since future Practitioners; and

156 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY - that this mobility is both physical and virtual problem existed with the operation, or state, (enabled through a high level of international of the markets, to clarify whether any policy communications infrastructure, both voice response is required. In general, as employer and, increasingly, data). needs are paramount for productivity and/ or global competitiveness, policy should be All these factors can have considerable impact driven from Demand-side considerations! on national employment. In general this impact tends in the first instance to be a negative one in b) The most significant concern that might arise industrialized countries like the majority of EU in relation to a labour market generally relates and OCED Members. The combined effect of to skill shortages. In principle if a labour offshoring (shifting operations overseas, whether market is suffering from serious shortages outsourcing or to subsidiaries in lower-wage of supply, then there might be a case for a economies), and inward migration has threatened policy response of some kind. However, it both the employment benefits of the recovery turns out that many – probably most – sectors from the ‘dot.com bubble’ (there are increasing complain of skill shortages at some time or fears of ‘jobless growth’) and, perhaps even more another(!) The occasions when an employer seriously, the growth in international outsourcing has a vacancy, and – when this is known – beyond ‘low-end’ occupations (like call-centre quickly finds a number of keen applicants, all work) to higher-level activities like software of whom have adequate skills for the job, are development can pose a real threat to the growth indeed rare. Thus government is often told occupations so crucial in countries attempting, of skill shortages from a range of employers, as most industrialized countries are, to respond relating to a range of sectors or occupations. to job losses in waning industries. The need for In this situation, policy-makers need some robust evidence about the state of the national way of distinguishing between those cases labour market (and so the choice of indicators where successful recruitment might not and the validity of the statistics) in this context be instant and those where it really is not is likely to grow in importance, and may finish possible, over an extended period of time, up as the most significant of the policy needs for to find any applicants who ‘fit the bill’. The data on certain indicators – in particular on skill current assumption is that the best indicator shortages (q.v.). of serious skill shortage is the number (degree) of hard-to-fill vacancies reported The evidence needs of policymaking in skills by a significant sample of employers, where the recruitment is believed to arise through In order to clarify the priorities in relation to serious lack of supply (as opposed to certain statistics-gathering, it is first necessary to be other spurious causes5). As hinted above, clear about the kinds of policy consideration of where inward-migration is accepted to be one importance within skills policy generally. way of responding to serious skill shortages in the ‘home’ market, clearly the evidence a) The first requirement that could be considered of serious shortage in the indigenous labour is an ability to monitor the state of the labour market needs to be particularly robust. markets, where it is possible to track (at least in broad terms) both supply and demand. c) As well as serious shortages of skills supply The value of this would be an ability to in the labour market, it is also recognized that assess, with confidence, whether any serious organizations may well be under-performing

5 these would include, for example, recruitment difficulties experienced by employers who do not offer the ‘going rate’ or who have a reputation (for whatever reason) for not being ‘good places to work’.

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 157 due to skill gaps. These are shortfalls between should be can – and increasingly are – viewed the skills that an organisation’s workforce in different ways. On the one hand, there currently have and the skills that would be are challenges to the normal RoI principles needed for all to perform effectively. in the case of training that arise from the ability (in principle) for the person who has d) The differences between skill shortages benefited from such investment choosing and skill gaps immediately bring out the to ‘walk out the door’ before the returns to fundamental question of the responsibilities the employer from the (hopefully) resulting of the three main ‘players’ in relation to higher performance can be captured. On the skills: the role of the state (involving the use other, in many areas – including e-skills – the of ‘tax Euros’), employers and individuals. competition to find, recruit and retain high- In principle, each might have certain performing technical experts can be viewed as responsibilities for addressing skill shortfalls. becoming as important as the competition for Clearly, having received a ‘head start’ of other resources and indeed for product/service some kind from the state through the formal sales… Many forward-looking employers education system, each of us must quickly have responded to this competition with take responsibility for making ourselves innovative approaches to HRM – providing a employable as an individual. Employers, range of attractions/benefits for the individual who pay for the contribution of each of their (and even his/her family) hitherto unknown. staff, have a certain interest in maximizing If public policy becomes too responsive to that contribution to the organisation’s employers’ cries of ‘skill shortage’, such overall performance by making sure that ‘market responses’ may be lost. the individual’s skills are adequate to make it effective. And finally, the state may well f) Specifically, as elsewhere in public policy, accept a responsibility for addressing issues governments are now increasingly thinking that employers would see as ‘nothing to do in terms of responses mainly (or only?) with them’. This might include accepting where there is clear evidence of market responsibility for tackling a shortage of skill failure. The challenge for the economics supply in the marketplace, as well, more and statistics community is therefore to generally, as taking a broader look at the provide such evidence, often in the face of skill needs of the workforce in supporting considerable difficulty. As in other policy greater productivity and prosperity of the areas, governments are seeking to raise their country as a whole. The challenge, of game beyond policy responses involving course, lies in understanding labour market (only) areas where they have (through public behaviour sufficiently well and designing service delivery) particular influence. E-Skills policy responses that are appropriate and will is certainly one area where limiting policy be effective in ‘nudging market behaviour responses, for example, to adjusting what is in a beneficial direction’, rather than simply taught in the formal education system (even making good ‘things that individuals and if this were easy to effectively implement) is employers should be doing anyway’. unlikely to catalyse labour market behaviour in the way it generally needs steering. e) Policy in the skills area needs to reflect carefully on the nature of employer responsibilities. g) However, perhaps one of the greatest challenges While in principle enterprises will behave in in the case of e-skills it that of attempting to a commercial way, the boundaries of what an estimate future needs in a world of ‘disruptive employer’s responsibility in relation to skills technologies’. The lessons painfully learned in

158 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY the 1970s about the pitfalls of manpower planning The Impact of ICTs in relation to e-skills, – in particular that trying to ‘size’ learning and the main challenges for their provision channels based on extrapolations of measurement recent trends can be seriously unwise - are of great importance in relation to trying to forecast The Conference and the Workshop ask contributors the future of e-skills demand. In particular the to consider the impact of ICTs on the world, and realities of what official statistics can provide in to elucidate the challenges for measurement. In relation to (delayed reporting of) recent trends terms of impact in the skills area, it is possible to using reference frames developed some years confirm that: before generally produce strongly negative reactions from industry leaders whose gaze • the remarkable growth in the development, is, understandably, focused on the ‘next hill to deployment and use of ICTs has inevitably climb’. The forthcoming Commission-funded triggered a huge need for the acquisition of project on ‘e-Skills Foresight Scenarios’ will significant new skills – for both practitioners have to find a way of reconciling the necessary and users; creative thinking about the future with the less • this need has in turn caused the creation of, exciting realities of labour force numbers of ICT and corresponding growth in, learning about Practitioners. One important way of doing this ICTs within the formal education system, is to distinguish between (and treat separately) pretty much at all levels; the skill needs for large workforce volumes and • the growth has also produced major new the relatively very small numbers of innovators economic sectors, as well as a remarkable engine whose work will sow the seeds for the major of job creation, resulting in (IT Practitioner) businesses and sectors of the coming decades. workforces in our countries often involving hundreds of thousands of people - workforces h) Recent OECD work on tracking the ‘overall that correspond to sizeable fractions – e.g. 2%, penetration’ of e-skills is worthy of particular 3% of the working population; and mention. While the author may have reservations about the core OECD e-skills • the skill-raising needs of a workforce so definitions, he has nothing but admiration big have ‘pulled through’ a secondary new for the recent work done in exploring the use sector of commercial ICT training (for both of ‘ICT intensive’ occupations and sectors practitioners and users), providing further (OCED, 2004). The general proposition is jobs. both creative and useful – what would be good would be the refinement of the postulated As far as challenges for measurement are allocation of ‘ICT-intensiveness levels’ to concerned, the paper has illustrated a considerable occupations and sectors with more empirical range, including: data. It is possible that some of this could be distilled from data on ECDL take-up, although • the impossibility of measuring skills directly it may be that this data source might suffer (either on the demand or supply side); to a degree from the same confidentiality • the difficulty of getting agreement, constraints as quantitative information on the internationally, on a single occupational more commercial ‘industry and proprietary classification framework for ICT Practitioner certifications’. skills;

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 159 • the difficulty of measuring serious skill set of challenges beyond those for skills in shortages directly; general. • the barrier to clarifying meaningful ‘returns to training’ for employers that arises from the And yet it is not surprising that governments ability of the trained employee to ‘walk out are convinced that future economic prosperity the door’; and depends on the skills of their people, and would like to find ways of supporting the raising of these • the considerable challenge of achieving skills generally. Indeed such considerations have, internationally-comparable statistics on inter alia, led to the strong commitment in some occupationally-based workforce data arising Member States to push for ever-higher Tertiary from problems of coding from national to Education participation rates (even though there international occupational classification are powerful arguments questioning whether this frameworks. is necessarily the best approach to national skill- raising (Keep & Mayhew, 2004)). Should we give up and go home? So are there things that can be done to improve the It is clear from the above that there are many, validity and usefulness of the statistical evidence many limitations in our ability to track, in a we do have, as well - perhaps - as to open up data meaningful and internationally-comparable way, streams for indicators not yet used? The author both the skills that employers require and the would single out the following: skills that our people possess – and perhaps even more difficult to track, quantitatively, the learning Some priorities to improve the situation: experiences (provided by whatever source) that help individuals raise their skills. (It is interesting 1) provision of guidance and support for national to note that many of these difficulties have been statistical office coding to ISCO (for LFS recognized by speakers in the Conference’s data); Human Capital Workshop.) 2) work for sensible update of ISCO in relation to ICT Practitioner occupations; In one sense, it should come as no real surprise that it is particularly difficult to capture and 3) continue to include questions on e-skills and measure in a rigorous, consistent way the (multi- e-skills training in Enterprise (employer) dimensional) ‘chimera’ of human potential. And (and perhaps also Household) surveys where indeed the author believes that there is a risk possible; in focusing all our attention on measures of the 4) recognise the real (substantial) cost of macro (or average) capabilities of our workforces getting better information on skills for policy (or indeed our populations). After all, we have purposes, and seriously consider significant known for years that the innovations that have investment (and strong steering of classification caused major advances in our capabilities, and frameworks) if we mean business in ‘tracking generated new economic activity, have arisen from the knowledge’ within the Knowledge the ingenuity and creativity of either individuals Economy (in particular increase LFS sample or rather small teams. Thus we would almost size and improve question on training); and certainly be well advised in future to find ways of tracking excellence as well as the average. 5) in the meantime, focus investment in new primary research on approaches that will To make matters worse, as is shown in the paper, provide data relevant to specific policy attempts to track e-skills suffer from an additional questions.

160 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY It is clear, from the many insights provided at Learning’; Commission Staff Working Document: the Conference, that mainstream economic and consultation paper, SEC(2005) 957, July 2005 statistical analysis recognizes the strong need for European e-Skills Forum (2004): ‘e-skills for data on skills to be built into work on innovation, Europe: Towards 2010 and beyond’; Synthesis productivity and competitiveness. The author Report, September, 2004 believes that experience in use of occupational statistics can indeed help inform the design of Huws, U. (2003): ‘Classification of Sectors and effective quality-adjusted labour input measures. Occupations in a Knowledge-based Economy’; results of Work Package 3, project on Statistics and Indicators on the Labour Market in the e- References Economy (STILE) AISS/ITNTO (1999): ‘Skills99 – IT Skills Keep, E., and Mayhew, K. (2004): ‘The economic Summary’ for the Department of Trade and and distributional implications of current policies Industry and the Department for Education and on Higher Education’; Oxford Review of Employment; Alliance for Information Systems Economic Policy, Summer 2004, Vol. 20, No. 2 Skills, IT National Training Organisation, July 1999. LeMaitre, G. (2002): ‘Measures of Skill from Labour Force Surveys – An Assessment’; DSTI/ CEN/ISSS (Beier, Y., Dixon, M.) (2005): ‘A OECD, October, 2002 European ICT Skills Meta-Framework: State- of-the-Art Review, Clarification of the Realities, OECD (2004): ‘Information Technology Outlook and Recommendations for Next Steps.’ Proposed 2004’: Chapter 6: ICT Skills and Employment CEN Workshop Agreement

CEDEFOP (Winterton, J. et al) (2005): ‘Typology Rand Europe (2005): ‘The Supply and Demand of e-Skills for Europe’; October, 2005 of knowledge, skills and competencies: clarification of the concept and prototype’ Centre SEMTA (2005): ‘Specialist IT User Skills for for European Research on Employment and Manufacturing, Processing and Logistics’; Project Human Resources Groupe ESC Toulouse Final Report, Sector Skills Council for Science, CEPIS (2002): ‘Information Technology Engineering and Manufacturing Technologies Practitioner Skills in Europe’; Council of (SEMTA), March, 2005 European Professional Informatics Societies, May, 2002 Sciadas, G. (2005): ‘Our Lives in Digital Times’, Proceedings Eurostat Conference on ‘Knowledge Clayton, A. (2005): ‘IT Investment, ICT Use Economy – Challenges for Measurement’; and UK Firm Productivity: Summary of ICT Luxembourg, December, 2005 effects, and measurement conclusions’; Office for National Statistics; October, 2005 Steedman, H., Wagner, K., Foreman, J. (2003): ‘ICT skills in the UK and Germany: How EU Commission (D-G EAC) (2005): ‘Towards a companies adapt and react’; Anglo-German European Qualifications Framework for Lifelong Foundation, September 2003

KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY 161 Abbreviations Association of America ITQ (UK) IT (user) Qualification AutoCAD (Standard PC-based Computer- LFS Labour Force Survey Aided Design software) NACE Nomenclature generale des BISER (Project) Benchmarking the Activites economiques dans Information Society for European la Communaute Europeenne Regions (Nomenclature of economic CEN/ISSS European Committee for activities in the European Standardisation/ Community) Information Society Standardisation System OECD Organisation for Economic CEPIS Council of European Professional Cooperation and Development Informatics Societies ONS (UK) Office for National CVTS EU Continuing Vocational Statistics Training Survey RoI Return on Investment D-G Directorate-General SEMTA (UK) Science, Engineering and DSTI Directorate of Science, Manufacturing Technologies Technology and Industry (at Alliance OECD) (Sector Skills Council) EAC (D-G) Education and Culture SKOPE (Centre for) Skills, Knowledge ECDL European Computer Driving and Organisational Performance Licence (Universities of Oxford and EQF proposed European Qualifications Warwick) Framework HRM Human Resource Management SOC (UK) Standard Occupational ISCED International Standard Classification Classification of Education STILE (project on) STatistics and ISCO International Standard Indicators on the Labour market Classification of Occupations in the e-Economy IT Information Technology 2-D/3-D Two-dimensional/Three- ITAA Information Technology dimensional

162 KNOWLEDGE ECONOMY: ICT IMPACT ON THE KNOWLEDGE BASED SOCIETY Report on the session: Human Capital

Chair: Mr. Radek Maly, Directorate General for Employment, Social Affairs and Equal Opportunities European Commission

In his introduction, the chairperson reminded that the challenges facing the statistical system in a in the context of progress towards a knowledge- life-long learning framework. He showed that the based economy, human capital can be seen as concept of life-long learning has a “life-wide” a key factor linked to economic prosperity, full dimension – referring not only to formal education employment, better quality at work and social provided in institutions but also to non-formal cohesion. Human capital is thus one of the most learning at work and in daily life. Thus, public important dimensions of the EU employment and authorities can no longer be the sole provider of cohesion policy. resources, nor necessarily determine curriculum content, achievement standards or qualifications, Despite the policy efforts that have been undertaken even though governments are still required to so far by the European Commission and the act in a steering capacity and provide a legal Member States to promote human capital, there framework. Life-long learning brings a number of is a need to better identify appropriate strategies large conceptual and methodological challenges to improve the measurement of its contribution of to the statistical system, including not only the human capital to growth, employment and social measurement of cumulative learning across the cohesion. This was the purpose of this session. life-span but also the measurement life-wide learning and the learning processes themselves. Three presentations were proposed on new strategies to improve the measurement of human Mr George Psacharapoulos, European Experts capital. Network on the Economics of Education and Former State MP, Greek Parliament, took a Mr T. Scott Murray, Director at the Unesco critical look at the present state of human capital Institute for Statistics, argues that there is an statistics in the European Union, against the increasing need to focus research efforts on the background of the Lisbon agenda. He identifies middle and bottom end of the skill distribution, several potential weaknesses such as the use i.e. the skills of the average worker. The need to of multiple instruments to measure a single concentrate more efforts on the skills of the average variable, the extreme comprehensiveness of some worker has led in particular to the conduct of the questionnaires, the frustration of missing data for world’s first international comparative assessment many countries, the long time interval between of adult skill and learning – the International Adult surveys and the late availability of the databases. Literacy Survey (IALS) – and, subsequently to It points at the scarcity of indicators between the the development and implementation of the Adult employer and the employee and data on returns on Literacy and Life Skills Survey (ALL). Such education and private expenditures on education. works should be improved in the future. These three presentations were followed by Mr. A. Tuijnman, Senior Economist at the a discussion led by Mr Tom Healy, Senior European Investment Bank, describes some of Statistician at the Department of Education and

KNOWLEDGE ECONOMY: HUMAN CAPITAL 163 Science in Ireland, who stressed that among the By way of practical suggestions, Mr. Healy preceding presentations the common themes underlined the need for more interaction included need for more and better data on HC between statisticians, policy makers and and the fact that measuring HC is more than just researchers in order to identify relevant needs measuring formal education. Mr. Healy also and priorities. pointed out that not only economic, but also social and personal benefits of HC need to be In fact the whole panel on HC was very much in measured. One of the key problems related to line with this suggestion, as it provided a useful the measurement of the HC is the fact that skills framework for discussions between people from are multi-faceted and acquired in different ways academia, policy-making bodies as well as at different stages of lifecycle. statisticians.

164 KNOWLEDGE ECONOMY: HUMAN CAPITAL Aspects of Human Capital and the Knowledge Economy: Challenges for Measurement

T. Scott Murray The UNESCO Institute for Statistics December, 2005

People are the common denominator of progress. So... no improvement is possible with unimproved people, and advance is certain when people are liberated and educated. It would be wrong to dismiss the importance of roads, railroads, power plants, mills, and the other familiar furniture of economic development.... But we are coming to realize... that there is a certain sterility in economic monuments that stand alone in a sea of illiteracy. Conquest of illiteracy comes first. John Kenneth Galbraith, The Affluent Society (1958) US (Canadian-born) administrator & economist (1908 - )

1. Introduction 2. Background

In March 2000, Europe’s heads of state and The OECD has defined human capital as “the government met in Lisbon, Portugal, and knowledge, qualifications, competences and declared their intention to make the European other qualities possessed by individuals that can Union (EU) “the most competitive and dynamic be put to productive use” (OECD, 1998). knowledge-based economy in the world, capable of sustainable economic growth with more and Much of the public policy attention, and the better jobs and greater social cohesion.”2 To associated data development and analysis, related achieve this goal by 2010 they adopted what is to the role of human capital in fostering economic now called the Lisbon Strategy of economic and growth and competitiveness in a knowledge structural reforms. economy has concentrated on the elite end of the skill distribution. Within this context, Eurostat has organized a conference focused around the theme “Knowledge This has involved a focus on rates of participation Economy, Challenges for Measurement”. This in tertiary education systems and on the relative paper has been written as a contribution to the quality of tertiary programs and institutions, session that looks at human capital as a key particularly in the hard sciences at the MA and factor contributing to competitiveness and PhD levels. growth. The paper seeks to help European policy makers, statistical systems and researchers to It has also explored the role of research and better identify appropriate strategies to improve development in the innovation process, including the measurement of the contribution of human a focus on the efficiency and effectiveness capital to growth, competitiveness, employment of knowledge creation, transmission and and social inclusion. management systems.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 165 Finally, policy makers have spent time looking This is partly pragmatic – the globalization of at individual and firm characteristics that are markets for goods and services, for capital and associated with achieving the high rates of for technology, the reduction of barriers to trade, technological, process and organizational the concentration of ownership in multinationals innovation that are felt to be needed in order and the diffusion of ICT’s into manufacturing to realize the productivity gains that will keep processes have reduced the ability of governments OECD economies competitive in the emerging to gain comparative advantage through traditional global knowledge economy and information means. More pointedly, one f the few places society. where public policy can hope to have a marked impact on long term economic performance is in Over the past 15 years the Statistics Canada, the the area of active education policies that seek to US National Center for Education Statistics and increase the skill of adult workers. the OECD have led an international research effort focussed on the middle and bottom end of the It is also a matter of simple arithmetic – most skill distribution. This research has concentrated productivity gains in the coming decade will on profiling the level and distribution ofa come the diffusion of ICT’s into the full range of range of skills believed to be tightly linked to occupations, including those that currently have the productivity of workers and their ability to the lowest knowledge intensity. absorb technological, process and organizational innovation. Initial work has provided data on the The focus on the skills of the average worker is supply of four skills that are know to be tightly also driven by a sense of fairness. Left to their linked to job performance in a broad range of own devices, labour markets appear to work as occupations: prose literacy, document literacy, savagely efficient engines for rewarding those numeracy and, more recently, problem solving. that have skills. While some inequality is useful Recent work has also explored the impact that in that it creates incentives to work and invest the use of ICT skills has on the labour market in further learning, it is clear that too much outcomes of individuals, including wage rates inequality is unfair and risks social discord. and employability. Government policy can play a role in levelling the playing field, by creating opportunities for This work has been motivated by a series of the least skilled to acquire the skills they need to linked considerations. compete in increasing knowledge intense labour markets. Such measures become increasingly First, it has attempted to address concerns important in a period when outsourcing is that inadequacy in the supply of these skills is leading to massive outflows of employment somehow constraining economic growth. from developed countries to the third world. Given the size of the underlying economic Second, it has explored the degree to which incentives to firms, governments can do little differences in skill level among population sub- to Influence these flows. Rather, they should be groups might explain rising social inequality in concentrating their efforts on assuring that the economic outcomes as the knowledge intensity workers who will inevitably be displaced have of jobs rises. the skills required to fill the replacement jobs that are being created. Third, it has reflected the belief that the skills of the average worker have become relatively Finally, the focus on the skills of the average more important to relative economic success and, worker is, in part, a matter of economic hence, to public policy. philosophy. One interpretation of economic

166 KNOWLEDGE ECONOMY: HUMAN CAPITAL theory is that individuals and firms are the – human capital does matter to growth (Becker, most efficient actors in adapting to change and 1964, Schultz, 1967, Mincer, 1974), a fact that fostering productivity growth and maintaining has driven governments in the OECD area to competitiveness. In comparison, government is make massive public investments in education. a relatively inefficient economic actor. Assuming this is the case, then government role should be These investments have translated into rapid restricted to creating the framework conditions increases in participation and completion rates that foster and enable growth and high rates at the elementary, secondary and post-secondary of adjustment to competitive pressures and to level. With the demographics of the post-war interventions designed to correct market failures. baby boom on their side these investments were One of the fundamental assumptions underlying sufficient to balance increases in the demand for the work on adult skills is that adult learning skill that were evident in most OECD economies. systems are woefully inefficient, to the point that government intervention is warranted. By the early 1980’s, however, policy makers were becoming increasingly concerned about This is work that is of central importance to policy skill shortages at all levels in OECD labour makers in the EU because it has sought to measure markets. Notwithstanding the rapid increases in the level and distribution of a range of adult skills educational output, wage inequality continued thought to be important to realizing the Lisbon to grow rapidly, suggesting that something objectives. Known as the International Adult was happening that could not be explained Literacy Survey (IALS) and the Adult Literacy with measures of educational quantity. Most and Life Skills Survey (ALL) these studies have labour market and macroeconomists, however, combined the tools of educational assessment continued to use indirect measures of skill, such and household survey methods to assess the as educational attainment, to explain differences quality of the current stock of skill and upon in labour market outcomes at the individual and the quality various human capital flows – from economy level. Faced with empirical results the initial education system, from adult learning that were not consistent with what theory systems and from worker mobility. What follows would predict labour market economists turned is a more detailed overview of the thinking, and to increasingly elaborate theories, including scientific advance, that lead to the conduct of the segmented labour markets, signalling theories, IALS and ALL studies. racism and discrimination, etc., to explain how the labour market was actually operating. While The importance of human capital to the economic each of these theories was plausible none of performance of nations has long been known. In them yielded conclusive results when tested fact, human capital theory can be traced back to empirically. the 18th Century to authors such as Adam Smith. The first empirical treatment of the subject did By the late 1980’s advances in macroeconomic not, however, appear until the 1960’s when Gary theory started to provide an alternative Becker published his seminal work on the subject explanation. (Becker, 1962). Becker’s theory suggests that economic growth depends on range of inputs, Theories of economic growth, loosely termed including physical capital, financial capital and endogenous growth theories, were developed human capital. that suggested that human capital would become increasingly important as technological Over the past 40 years the key tenets of human innovation created what was termed skill-biased capital theory have been empirically confirmed technical change.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 167 The basic hypothesis underlying skill-biased In most cases the adoption of these technologies technical change is that new technologies are not has precipitated a sharp, short-term increase in only more productive but also demand higher the demand for technical training and a long- skill levels to employ them, and hence offer the term increase in the level of skill required of possibility of both higher economic growth and most workers, a classical example of skill biased higher wages to workers that possess the requisite technical change at work. skills. Thus, public policy became interested in how skill influenced productivity growth, The diffusion of ICT’s also precipitated a particularly through its impact on the rate of significant change in the organization of work technical, process and organizational innovation with many firms adopting less hierarchical, within firms. flatter structures that greatly increase the need for independent problem solving, decision-making The accompanying policy rhetoric suggested that and communication. These organizational changes economically successful nations would have to serve to amplify increases in demand for the same have post-secondary education systems capable range of occupationally generic skills. of sustaining high rates of knowledge generation and skilled, flexible and highly adaptable labour The economic context within which these changes forces. This model implied the need for two in technology of production and the organization complimentary sets of skills: of work have occurred has itself created strong incentives for firms to adapt at a rapid pace. – advanced technical skills and awareness of scientific bodies of knowledge sufficient to The general context is one in which processes support the generation of new knowledge and of globalization and the integration of markets its application in the production process, and has afforded firms enormous opportunities for growth while simultaneously exposing them to – a set of key competences that transcend unprecedented levels of competition. occupations, competences that are needed to master modern modes of production and One feature driving globalization is the emergence organization, and which enable efficient of efficient global capital markets that afford firms learning in adulthood. in all regions access to low cost finance, thereby reducing barriers to market entry and expansion. Up to this point researchers believed that the economic demand for skill prevailing at any given A second feature driving globalization is the point in a country was, in large measure, determined consolidation of ownership and the emergence by the industrial and occupational distribution of of powerful multinationals in virtually every employment in the economy. After the oil shock industry. Their existence has greatly increased the of the 1970’s all OECD economies began to mobility of technology, capital and production, experience a rapid increase in the demand for these a fact that puts additional pressure on firms to latter skills, a demand that was driven by powerful reduce costs and increase productivity. forces beyond the control of policy makers. A related feature is the emergence of global By far the most important development markets for technology, a fact that has further influencing the demand for skill during this reduced barriers to market entry for firms. period was the rapid diffusion of information and communication technologies (ICT’s) into Significant improvements in transportation production systems at the firm level. networks have also contributed to globalization

168 KNOWLEDGE ECONOMY: HUMAN CAPITAL by reducing transportation costs to the point where changes in the demand for skill? What is the they pose much less of a barrier to market entry. quantity and quality of skill flowing out of Combined with steadily falling prices for many the formal education system? To what extent key input commodities, improved transportation is adult learning constrained by low literacy networks have opened markets to a much larger skills? number of firms. – How efficient are markets for skill? Are there barriers to matching that serve to reduce Massive investments in education made in several growth or inhibit growth? developing economies have greatly increased the pool of skilled labour available to multi-national Faced with these questions, the initial firms. Workers in these countries are fully capable statistical response to this concern involved the adopting the efficient technologies of production performance assessment of students in school. and work organization enabled by ICT’s, and Policy makers were interested in the relative are willing to work for lower wages than those quantity and quality of skill being generated in that prevailing in OECD countries. This latter initial education, and on the social and economic fact creates huge incentives for firms to move consequences that were associated with social production to lower cost labour markets. inequity in the distribution of educational outcomes. Finally, political changes have led to a substantial reduction in both tariff and non-tariff barriers to Lead by Torsten Husen, the International trade in goods and services. The Uraguay round Association for the Evaluation of Educational of the GATT, the North American Free Trade Progress (IEA) fielded a series of international Agreement (NAFTA), MERCOSUR and several comparative studies including the First Science other multilateral agreements have opened and Mathematics Study (FIMS), the Second markets. The emergence of the European Union International Mathematics and Science Study has also reduced barriers to trade, particularly (SIMSS), the Third International Mathematics within Europe. and Science Study (TIMSS) and, finally, the IEA Reading Literacy Study. Faced with unprecedented levels of competition firms no longer have anywhere to hide. In a world The IEA studies were the first to demonstrate how where capital, goods and services, technology and variable educational systems across the OECD highly skilled labour flow freely, policy makers were in the quality of their output, for what had were forced to look to the skills of the average previously thought to be equivalent levels of worker, as it is their skills that will allow nations attainment. to compete successfully in the global economy. A few economists have used this new data to The response of governments throughout explain differences in economic growth among the OECD area has been to create elaborate OECD countries. Hanushek, for example, mechanisms to monitor the supply and demand attributes the superior economic performance for skill. In general these systems aim to answer of the USA to their performance on TIMSS several simple questions: (Hanushek, 2000).

– What is the demand for skill and how is the By the late 1980’s, however, it became clear that demand expected to change? data on the quality of recent educational output – What is the supply of skill and what social was no longer sufficient to meet the skill demands institutions can be expected to respond to of most OECD economies.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 169 First, the rapid drop in fertility in most OECD investments in adult learning might serve to reduce countries resulted in relatively small youth social exclusion in all forms, but particularly in cohorts, cohorts that were too small to fill the reducing social inequity in economic outcomes. rapidly rising skill demands. As a result, national policy makers were faced with three options – These seminal assessments launched a trend that improve the quantity and quality of skill flowing eventually lead to the conduct of the world’s out of the initial education system, import the first international comparative assessment of requisite skill or find the means to re-train large adult skill and learning - the International Adult numbers of adults. Literacy Survey (IALS) - and, subsequently to the development and implementation of the Adult Second, evidence began to mount that adult skills Literacy and Life Skills Survey (ALL). were far more variable than could be explained by suspected educational quality in the past, a 3. What the IALS and ALL data reveal fact that suggested that processes of skill loss and gain were having a significant impact on the that has import for achieving the Lisbon available supply of adult skill. Objectives

One can not reflect on the challenges for Third, research began to suggest that skill played measurement that are implied in realizing the a far more important role in mediating the labour Lisbon objectives without first reflecting upon market success of individual workers. what the current generation of assessments reveal that is of interest to policy makers. The following Faced with this evidence and perceiving the chapter of this paper attempts to summarize the increased relative importance of skill to economic key findings that have been gleaned through success, governments in North America began to analysis of the IALS and ALL studies. field direct assessments of adult skill - to profile the actual level and distribution of economically Inter-country differences in adult skills are important skills, to better understand the role much larger than implied by differences in that skill plays in determining labour market educational attainment profiles outcomes and participation in adult learning and to help understand the factors that influence the As shown below, the IALS and ALL studies observed distribution of skill. confirm that there are much larger differences in the literacy, numeracy and problem solving skills At the broadest level, these assessments were of adults in OECD countries than implied by designed to help understand how skill shortages observed differences in educational attainment in might be constraining economic growth and how the same countries.

170 KNOWLEDGE ECONOMY: HUMAN CAPITAL Chart 1. Distribution of literacy scores by country, adult population aged 16-65

��

Source: Literacy Skills for the Information Age: Final report of the International Adult Literacy Survey (Statistics Canada and OECD, 2000)

These differences suggest, in turn, much greater Large shifts in the relative quality of variation in the quality of initial education that educational output have taken place over the conventionally assumed1 or, alternately, large past 40 years inter-country differences in the magnitude of skill As the following chart reveals the relative gain and loss in adulthood. performance of OECD countries, judged by the average literacy skill level of 17 to 25 year olds, has shifted quite dramatically over the past 40 years, largely in response to educational reform.

1 a fact confirmed for the current period by the OECD Program for International Student Assessment (PISA)

KNOWLEDGE ECONOMY: HUMAN CAPITAL 171 Chart 2. Changes in the relative skill quality of 17 – 25 year olds in 14 OECD countries, prose literacy, 1960 – 1995, derived from IALS, 1994

��

�� Source: Statistics Canada and HRSDC, 2004

Adult skill loss erodes the supply of skill problematic in four important respects: available to the economy – it serves to erode the return on what is, in most countries, publicly financed education, and, It is important to note that the size of the observed – it would seem to reduce the long term economic relative shifts over time in inter-country performance growth trajectory far exceed the size of the differences detected by international student assessment systems such as – it also serves to amplify social inequality in the OECD PISA study and the IEA’s TIMSS study. individual economic outcomes This provides strong evidence that adult skill gain – it slows the introduction of more productive and loss lead to significant change in the supply technologies, processes and organizational of skill available to the economy. This finding is structures in the provision of public goods and important in several respects. services such as health and education.

First, it justifies the need to monitor the supply of Patterns of skill gain and loss differ from economically important adult skills over time so country to country that the magnitude of adult skill gain and loss can estimated for important sectors of the economy. The following charts suggest that the patterns Conversely, measuring the skill quality flowing from of skill gain and loss differ greatly from country the secondary education system, while important to country in response to compositional shifts for driving educational reform and improvement, in the supply of labour, to differences in the is not enough to support effective public policies. demand for skill and in the differences in the incentives that these skill demand differences Second, it establishes a need to understand the in demand create for individual skill acquisition processes that lead to skill loss. Skill loss is highly and maintenance.

172 KNOWLEDGE ECONOMY: HUMAN CAPITAL Chart 3. Socio-economic gradients in skill for Canada, Norway and the USA, 1994-2003

��

� � � � ��

��

� � � � ��

Source: Learning a Living: First results of the Adult Literacy and Life Skills Survey, 2003

KNOWLEDGE ECONOMY: HUMAN CAPITAL 173 Synthetic cohort analyses of the IALS and ALL The following chart shows quite dramatically data for those OECD countries with estimates for that the labour markets of the OECD are both periods offers support for this thesis – skill extraordinarily skill-sensitive. It takes over four gain is concentrated in jobs that impose high times as long for half of low skilled adults to exit indices of skill literacy use, that have higher rates unemployment as their more skilled peers – 38 of employer-based education and training and weeks v.s. 9 weeks that involve high levels of ICT use (Willms and Murray, 2006). The following chart highlights the marked impact that literacy skill has on the wage outcomes Skill is an engine for creating inequality: realized individuals. In Canada, over one third differences in average literacy and numeracy of all wage variability can be attributed back to skill have a marked impact on individual social, differences in literacy skill, a level that makes economic, health and educational outcomes difference in skill among different population sub-groups a matter of pressing policy concern. The evidence is clear that these differences in skill matter to individuals. Results presented It is interesting to note, however, that the in a series of related analyses, confirm that influence of skill on wages varies considerably differences in skill have a significant impact upon from country to country with the highest wage individual labour market (Statistics Canada and returns to skill being observed in countries where OECD, 2005; OECD and HRDC, 1997; Statistics skill demand is rising and where skill supply is Canada and HRDC, 2001), educational (Statistics of variable quality. As revealed above, employers Canada, 2002; Desjardins, 2004), health (ETS, in Sweden pay practically nothing for literacy 2004; Murray and Clermont, 2006) and social skill because the majority of the Swedish labour (Statistics Canada and HRDC, 2005) outcomes. force has uniformly high skill levels. This result

Chart 4. Probability of exiting unemployment by skill level

Labour market outcomes and skill: Probability of exiting unemployment by skills levels The probabilities of unemployed adults aged 16 to 65 to exit unemployment over a 52 week period, by low (Levels 1 and 2) and medium to high (Levels 3 and 4/5) skills, document scale, 2003

Probability 1.0 0.9 0.8 0.7 9 0.6 38 0.5 0.4 0.3 0.2 Levels 1 and 2 0.1 Levels 3 and 4/5 0.0 0 4 8 12 16 20 24 28 32 36 40 44 48 52 Weeks

Source : Adult Literacy and Life Skills Survey, 2003.

174 KNOWLEDGE ECONOMY: HUMAN CAPITAL implies that one cannot rely on estimated wage More than an allocative mechanism: The returns to skill as proxies for the economic value estimated impact of average literacy and of skill because observed prices largely reflect the numeracy skill differences on long-term relative conditions of supply and demand (rather economic growth than their intrinsic value). What the individual-level results do not reveal, The overall conclusion remains, however, that however, is whether the observed differences in the institutional actors in the markets for labour, skill contribute to differences in the long term educational goods and services, health goods economic performance of nations. It might be that and services, and that afford access to power and skill plays no role in fostering differential rates influence, identify and reward skill to the point of economic growth –skill might rather serve that skill emerges as the single most important only to allocate scarce resources from a “fixed” source of social inequality in outcomes. This fact economic pie. raises the importance of skill-related policies in reducing levels of social inequality and Recent analyses provide strong evidence that exclusion2. this is not the case. Differences among countries

Chart 5. Earnings and literacy proficiency, controlling for education and labour force experience

...literacy xplains a signiﬁcant fraction of wage variability in Canada, but not in Sweden... Earnings and literacy proﬁciency, controlling for education and labour force experience

Standardised regression weights x 100

0 10 20 30 40 50

Canada Countries are Finland ranked by the United Kingdom Norway magnitude of the New Zeland effect parameter Australia associated with United States educational Ireland Portugal attainment Hungary Denmark Slovenia Netherlands Belgium (Flanders) Sweden Switzerland Chile Czech Republic

Germany Poland

Educational attainment Literacy proficiency Experience

Source : International Adult Literacy Survey, 1994-1998.

2 or at least containing inequality to current levels

KNOWLEDGE ECONOMY: HUMAN CAPITAL 175 in average skill level appear to have a profound Taken together these findings carry two important impact on the size of the economic pie to be implications for policy. divided. Specifically, they explain fully 55% of differences in two key indicators of long term The promise of higher growth and options for macro-economic performance - growth rates meeting rising skill demands of GDP per capita and labour productivity in 14 OECD economies (Statistics Canada and First, they suggest that countries could precipitate HRSDC, 2004) Forthcoming analyses for Canada significant improvements in their economic confirms that average literacy and numeracy performance by finding ways to increase their skill levels have had a strong positive impact average literacy levels. Historically, such on the long term economic growth trajectories improvement was realized by increasing the of Canadian provinces, an impact that comes on duration and quality of the cycle of initial top of a strong positive impact of participation education. rates in tertiary education (Statistics Canada and HRSDC, 2006). EU countries can no longer depend upon this mechanism to solve their skill requirements. Due These are important results because they bring to the low fertility rates that prevail in OECD estimates of the macro-economic returns to skill countries the flow of new graduates will only up to those estimated at the individual level. This change the overall skill profile very slowly. implies that economic policy makers have likely Evidence from the OECD PISA study confirms under-invested in the learning systems that generate that many students leaving the initial cycle of human capital on the mistaken assumption that education in the EU lack the skill levels needed higher returns were to be had elsewhere. to compete in a global knowledge economy.

The mix matters: the distributional effects of This implies a need for the EU to do one of five literacy and numeracy skill upon economic things if it is remain competitive: growth – To reduce the percentage of students leaving The same macro analyses also reveal strong the initial cycle of education with inadequate distributional effects of skill upon economic skills growth. – to invest heavily in adult learning systems. – to import skills from abroad, or, The percentage of adults with high literacy levels appears to have no impact of long term growth – to permit much more labour mobility within trajectories. This finding suggests that there are the newly enlarged EU no supply constraints - that each of the countries – to promote the substitution of technology for studied have sufficient numbers of these highly labour. skilled workers. As noted above, however, it would seem that the proportion of tertiary None of these options is without its problems. participation does have a marked positive impact on long term economic growth. As the following chart reveals significant proportions of students in some EU economies The percentage of workers with very low literacy fail to meet the graduation standard set by the skills - Level 1 in the ALL study - exerts a very Canadian province of British Columbia, standards strong negative drag on long term growth rates. that are thought to represent the minimum

176 KNOWLEDGE ECONOMY: HUMAN CAPITAL Chart 6. Percentage of 15 year olds that meet, exceed and fail to meet British Columbia’s Graduation Standard for Reading Literacy

Signiﬁcant proportions of 15 year olds fail to meet B.C. ’s grade 10 performance standards... Percentage of 15 -yr olds from various jurisdictions attainig B.C. grade 10 reading standards, 2000

Finland Alberta B.C. grade 10 (2000) Quebec B.C. 15-yr-olds Canada Japan Saskatchewan B.C. grade 10 (2001) Manitoba Ontario B.C. grade 10 (2002) Nova Scotia United Kingdom Newfoundland and Labrador Prince Edward Island France United States New Brunswick Italy Germany -60 -40 -20 0 20 40 60 80 100 Percent Does not Meets Exceeds meet standards standards standards 1. All results shown here are for 15-year-olds except for B.C. grade 10 students who are, on average, 6 months older than B.C. 15 year olds. Jurisdictions ordered by the percentage of students meeting or exceeding expectations.

Source: Table 6. requirements for students to take full advantage Investing in adult learning systems is problematic of tertiary education, to be productive workers, at the EU level because of extraordinary variation good citizens and independent lifelong learners. that exists within Europe in the degree to which Member States have funded and encouraged Improving the quality and equity of initial adult learning systems, a fact that is dramatically educational output will, in some EU Member illustrated by the following chart. States at least, require major restructuring of educational systems.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 177 Chart 7. Participation rates in adult education and training, population aged 16-65, 1994-1998

Signiﬁcant proportions of 15 year olds fail to meet B.C.’s grade 10 performance standards... Figure 9 Participation in adult education and training Rate of participation in adult education and training, population aged 25-65, 1994-1998

Percent Percent 60 60

50 50 Legend: 40 40 Countries with participation rate significantly lower than the United States 30 30 Countries with participation rate not 20 20 significantly different from the United States 10 10 Countries with participation rate significantly higher 0 0 than the United States a d k n Italy Chile Poland Irelan Canada Finland Swede Norway Portugal Hungary Sloveni Australia Denmar VERAGE Switzerland Netherlands A New Zealand United States Czech Republic United Kingdom Belgium (Flanders)

C ou n tr ie s ar e r a n k e d b y t h e r ate of p ar tic ip ation . Countries are ranked by the rate of partecipation. NNote: ote :S Statisticalta tistic a di l ffderence iffe re nisc significant e is sign ifatic p a< n.05.t a t p < .05. SoSoururce:c e : InInternationalte r na tio na Adultl A dLiteracyult L ite Surveyra c y, S1994-1998. ur ve y, 1994 - 1998.

Implicit or explicit subsidies for those Member markets for skills. This latter effect will put heavy States that have chosen to under invest in adult pressure on those Member States with heavily learning systems will be politically unpalatable. unionized workforces and labour regulations that restrict labour mobility. Importing skills from abroad is economically attractive because it involves benefiting from human The final option, of promoting labour/technology capital that others paid to create. The costs and substitution is also fraught with difficulty. While it difficulties of integrating immigrants economically is likely that it would lead to significant productivity and socially are, however, well known. growth in the target industries it would inevitably displace large numbers of workers, many of whom Relaxing labour mobility within the EU, lack the basic skills needed to fill the replacement particularly for workers from the new Member jobs that are created. In the absence of high levels States, holds the promise of meeting the EU’s of investment in adult education and training, skill needs. Higher rates of labour mobility will, higher levels wage inequality and of long term however, put severe strain on both the supplying unemployment and social exclusion for the low countries (who will be denied access to the skills skilled is inevitable – even if the proportion of of the mobile workers) and the receiving countries jobs lost to outsourcing is modest. If, as expected, (who will experience much higher levels of wage outsourcing leads to even higher levels of inequality as lower skilled nationals are displaced displacement/job loss, then wage inequality and by immigrants with higher skills or, where labour the concentration of the burden of unemployment markets are rigid, the emergence of large grey in the low skilled will grow rapidly.

178 KNOWLEDGE ECONOMY: HUMAN CAPITAL Chart 8. Likelihood of participation in active modes of adult learning by literacy levels, adults aged 16-65, 2003

Equity and skill ﬂows from adult learning: Likelihood of participation in active modes of informal learning by literacy levels

Adjusted odds ratios 1 showing the likelihood of adults aged 16 to 65 participating in active modes of informal adult learning during the year preceding the interview , by document litera cy levels, 2003

Odds (x times) 6

0 Canada Italy Switzerland United States Norway Bermuda

Level 4/5 Level 3 Level 2 Level 1

Countries are ranked according to the odds of persons who score at Level 4/5.

1. Odds estimates that are not statistically different from one at conventional levels of significance are

reported as one in the figure.

For the actual estimate and its corresponding significance, see Table 4.8 in the annex to this chapter.

Source: Adult Literacy and Life Skills Survey, 2003.

The promise of higher growth and the advanced OECD economies provides further mechanics of markets for employer-based justification for government intervention in adult education and training learning and labour markets. The available macroeconomic evidence suggests that basic skills The second implication for policy, the largest make a significant contribution to key indicators increases in growth would be realized through of long term economic success, including the investments targeted at the lowest skilled. As growth of GDP per capita and productivity illustrated in the following chart, left to their growth. It is clear from the temporal relationship own devices employers allocate most of their that increases in skill lead to higher growth skill enhancing investments to their most skilled rather than the reverse. The presence of skill workers. loss implies that some fraction of employers are pursuing growth strategies that are sub-optimal This suggests that governments must intervene in at the macro level, the most likely mechanism adult learning markets to create incentives for firms being firms seeking productivity gains through to train the least able, if only on equity grounds. the substitution of technology for low skilled labour. If true, this fits the classical definition of Skill loss: a classic example of market failure a market failure of the sort that only government and a reason for government intervention can correct.

The fact that the ALL study identified significant Governments should not impose policies that levels of adult skill loss in some of the most inhibit this process. Rather they have a role in

KNOWLEDGE ECONOMY: HUMAN CAPITAL 179 fostering this process of creative destruction Policy implications of skill hierarchies and the i.e. they must create incentives to speed and amplifying effects of ICT’s enable the process as this is the only way that low skilled industries will remain competitive. Similarly, if skill acquisition in the other skill A range of measures is available - more flexible domains depends upon someone having high labour market regulations and tax incentives for prose literacy levels then the policy response technology investment to name just two. would have to deal, in the first instance, with reducing the proportion of individuals with Second, they need to provide opportunities inadequate prose literacy skills before tackling the for skill upgrading for those workers that will issue of skill development in the other domains. inevitably be displaced. A failure to do so will precipitate rapid increases in wage and income This insight is fundamental to the nature of the inequality, higher social welfare costs, health optimal policy response, to governments being costs and lower rates of economic growth. the “lender of last resort” in an inefficient market for adult skill. Implications of the inter-skill covariance matrix for public policy Table 1 explores how performance on the prose literacy, document literacy and numeracy scales One of the unique features of the ALL study is is related, specifically whether individuals who its ability to explore the inter-skill covariance have skills below level 3 in multiple skill domains. matrix. This information allows one to see how This is an important threshold - analysis of IALS consistent individual performance is across the and ALL data that individuals who fail to achieve skill domains. It also allows one to understand Level 3 skills face higher risks over a range of the nature of any skill dependencies, that is, the degree to which acquisition of high proficiency in outcomes including unemployment, low wages, document literacy, numeracy, problem solving and poor health and difficulty in accessing tertiary high levels of Information and Communication and adult education and training systems. For the Technology use depend upon an individual purposes of this paper individuals were classified having a high level of prose literacy skill. into the following groups: 3 -- At risk in all 3 domains In many respects the scope and nature of the 2 -- At risk in two domains policy response to the skill distributions presented 1 -- At risk in one domain in the IALS and ALL comparative reports will 0 -- Not at risk in any domain depend upon what the data reveal about these Table 1 reveals that significant proportions of relationships. For example, if a relatively small adults in all countries can be judged to be at risk number individuals exhibit low skills in multiple domains the policy response could be targeted - in fact the proportion of adults that are judged to and program triage would be straightforward. be risk free by this standard is quite small ranging Alternatively, if low skill is more randomly from a high of 50.2% in Finland to a low of 10.3% distributed in the population, then the policy in Italy. The proportions of adults who are weak response would have to reach a much larger in only one skill domain is also relatively small population who may or may not share a common ranging from a low of 8.5% in Italy to a high of set of demographic characteristics. 16.7% in French-speaking Switzerland.

180 KNOWLEDGE ECONOMY: HUMAN CAPITAL Table 1 . Distribution of Population 16-65 Across Categories of Multiple Disadvantage

Number of Domains in Which Individuals are at Risk

(Below Level 3) 0 1 2 3

Canada (English) 44.5% 14.5% 9.5% 31.6% 0.007 0.009 0.007 0.009

Canada (French) 33.8% 14.5% 14.0% 37.8% 0.011 0.011 0.014 0.011

Switzerland (German) 39.1% 15.8% 17.3% 27.8% 0.021 0.012 0.018 0.015

Switzerland (French) 30.4% 16.7% 19.0% 33.9% 0.017 0.016 0.017 0.018

Switzerland (Italian) 26.1% 16.6% 21.6% 35.6% 0.017 0.015 0.017 0.014

Italy 10.3% 8.5% 11.7% 69.5% 0.006 0.008 0.009 0.012

Norway (Bokmal) 50.2% 16.0% 10.8% 23.1% 0.011 0.008 0.007 0.010

Bermuda (English) 38.8% 16.0% 13.1% 32.1% 0.014 0.011 0.011 0.013

USA 33.7% 12.3% 10.7% 43.3% 0.015 0.012 0.010 0.012 Note: Standard errors in italics.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 181 The proportions of adults who are judged to availability of low cost reliable assessment tools. be at risk by this standard in either two or Chart 9 shows the proportion of adults who are three skill domains are striking. The fact that weak in all three domains and the proportion of not all individuals who are at risk are weak in adults judged to be at risk in any skill domain three domains suggests that programs designed of the percentage of adults. The chart reveals to improve skill levels need to incorporate that countries differ in both dimensions i.e. how assessment procedures to identify learning needs many people are at risk and how concentrated at the individual level, something that requires he risk levels are.

Chart 9. The proportion of adults 16 to 65 with skills below Level 3 in prose literacy, document literacy and numeracy, by the proportion of the total population that is at risk in any domain

��

Source: ALL, 2003

182 KNOWLEDGE ECONOMY: HUMAN CAPITAL 4. The cost of inaction is high: EU policy work for much lower wages. This fact is likely to makers must act place a strong downward pressure on employment levels and wage rates in the OECD area. This will There is some urgency for OECD governments to act be particularly problematic for some EU Member as the cost of inaction would appear to be too high. States as they struggle to cope with downward wage pressure created by the expansion of the To begin with, the concentration of ownership in European Union. multinational corporations has created a class of economic actors that have the means, and strong Finally, the diffusion of information and economic incentives, to move production to the communication technologies is expected to lowest cost locations rapidly. The rate of adjustment is being driven by globalization of markets for amplify the growth effects of skill.. As revealed goods and services and for every production input, in the following chart the application of these save the skill of the average worker. technologies is highly dependent on workers having advanced literacy skills. At the same time, education reform in the developing world is leading to a rapid increase in A second chart adds to the picture – it reveals the global supply of skilled labour. These workers a strong dependency between high literacy and are able to compete with the best that OECD heavy ICT use and income – itself the joint member states have to offer but are willing to product of wage rates and hours worked.

Chart 10. Likelihood of being a high intensity computer user by literacy skill levels, adults aged 16-65

Labour Market outcomes and skill: Likelihood of being a high-intensity computer user by literacy skill levels

Adjusted odds ratios showing the likelihood of adults aged 16 to 65 of being high-intensity computer users, by prose literacy levels, 2003

Bermuda

Norway

Canada

Italy Level 4/5

Level 3 United States

Levels 1 and 2 Switzerland

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Odds (x times) Countries are ranked by the odds of those who score at Level 4/5.

Source: Adult Literacy and Life Skills Survey, 2003.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 183 Chart 11. Likelihood of being an income earner in the top quartile, by combined literacy and computer use profile, adults aged 16-65

Labour market outcomes and skill: Likelihood of being a top income quartile earnerby combined skill and user proﬁles

Adjusted odds ratios ¹ showing the likelihood of adults aged 16 to 65 of being a top income quartile earning, by combined literacy and computer user proﬁles, 2003

Switzerland

Bermuda

Canada

Norway Group 4 Group 3 Group 2 United States Group 1

Italy

0 1 2 3 4 5 6 7 Odds (x times) Countries are ranked by the odds of those in Group 4. 1. Odds estimates that are not statistically different from one at conventional levels of significance are reported as one in the figure. For the actual estimate and its corresponding significance, see Table 8.12 in the annex to this chapter.

Source: Adult Literacy and Life Skills Survey, 2003.

Note: Group 4 have both high literacy levels and intense computer use

Specifically , given high literacy skills and high Table 2 extends this analysis to explore levels of ICT use, workers have a much higher the impact that multiple disadvantage has probability of finding themselves in high wage, upon the probability that individuals will be stable employment. Thus, firms, and nations, that heavy (top quartile) users of information and achieve a rapid diffusion of these technologies communication technologies (ICT’s). The throughout the production process stand to gain a research literature predicts that the acquistion productivity advantage whereas workers without and use of ICT’s depends upon a range the basic skills to master ICT’s will be left behind. of cognitive skills, including literacy and Rising wage inequality will be the result. numeracy (ETS, 2003).

184 KNOWLEDGE ECONOMY: HUMAN CAPITAL Table 2. Distribution of High-ICT Users Across Categories of Multiple Disadvantage

Number of Domains in Which Individuals are at Risk

(Below Level 3) 0 1 2 3

Canada (English) 63.8% 13.4% 8.5% 14.3% 0.018 0.014 0.010 0.013

Canada (French) 55.6% 16.0% 12.9% 15.5% 0.026 0.023 0.025 0.016

Switzerland (German) 57.1% 17.9% 15.5% 9.5% 0.036 0.023 0.030 0.022

Switzerland (French) 46.7% 21.7% 17.8% 13.8% 0.047 0.035 0.033 0.034

Switzerland (Italian) 45.2% 21.7% 20.5% 12.5% 0.048 0.054 0.057 0.027

Italy 22.3% 14.9% 17.8% 45.0% 0.013 0.016 0.021 0.022

Norway (Bokmal) 70.2% 12.8% 7.5% 9.6% 0.024 0.016 0.012 0.017

Bermuda (English) 64.2% 17.0% 8.7% 10.1% 0.040 0.027 0.016 0.020

USA 55.0% 14.0% 11.0% 20.0% 0.025 0.015 0.014 0.022 Note: Standard errors in italics.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 185 Table 2x. Percentage of individuals aged 16-65 who report high levels of ICT use, by category of multiple disadvantage

Number of Domains in Which Individuals are at Risk

(Below Level 3) 0 1 2 3

Canada (English) 38.5% 24.9% 24.1% 12.2% 0.013 0.021 0.027 0.156

Canada (French) 30.5% 20.5% 17.2% 7.6% 0.016 0.034 0.035 0.161

Switzerland (German) 38.7% 30.5% 24.7% 11.0% 0.017 0.034 0.043 0.241

Switzerland (French) 24.9% 21.2% 15.9% 7.6% 0.030 0.034 0.028 0.167

Switzerland (Italian) 26.8% 20.2% 15.7% 6.1% 0.029 0.047 0.041 0.176

Italy 54.2% 44.4% 38.4% 16.3% 0.028 0.030 0.030 0.344

Norway (Bokmal) 34.9% 20.0% 17.4% 10.4% 0.011 0.025 0.029 0.119

Bermuda (English) 41.4% 26.6% 16.7% 7.9% 0.022 0.034 0.032 0.231

USA 39.9% 27.8% 25.5% 11.9% 0.018 0.026 0.027 0.195 Note: Standard errors in italics.

Chart 12 below confirms that ICT use is indeed to amplify existing levels of wage inequality highly dependent upon skill level, particularly within and between countries. It reveals for individuals with weakness in multiple skill interesting differences among countries in domains. These findings suggest the level and the degree to which multiply disadvantaged distribution of literacy and numeracy skill individuals are afforded access to ICT’s will constrain the rate at which countries and compared to their more skilled peers. The size individuals can reap the productivity and of the gap between these two groups of adults wage gains associated with the application of is smallest in French-speaking Switzerland ICT’s. Thus, differences in ICT skill will tend (17.4%) and largest in Italy (37.9%).

186 KNOWLEDGE ECONOMY: HUMAN CAPITAL Chart 12. Difference in the proportion of adults 16-65 reporting high levels of ICT use, between individuals below level 3 in prose literacy, document literacy and numeracy and individuals with at least level 3 skills in prose literacy, document literacy and numeracy

Italy 37.9 %

Bermuda (English) 33.6 %

USA 27.9 %

Switzerland (German) 27.7 %

Canada (English) 26.3 %

Norway (Bokmal) 24.5 %

Canada (French) 22.9 %

Switzerland (Italian) 20.8 %

Switzerland (French) 17.4 %

0 % 5 % 10 % 15 % 20 % 25 % 30 % 35 % 40 %

Note: All differences are statistically significant at the 5% level.

Chart 13. Size of the skill gap against percent in the country who are high-users of ICTs

40 % ITA 35 % BER 30 % CHD- USA 25 % CDA-F CDA-E

p NOR 20 % CH-I Ga 15 % CH-F 10 %

5 %

0 %

10 % 15 % 20 % 25 % 30 % % of high-ICT users

Chart 13 plots the size of the gap in each country As expected the higher the rate of ICT penetration against the percentage of high users of ICT. the higher the gap, a finding that suggests that the diffusion of ICT’s will drive up inequality in economic outcomes.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 187 5. Implications and challenges for technology to detect meaningful levels of change. measurement Adoption of adaptive, computer-based assessment technologies will drop this latter threshold to five The forgoing analysis holds several implications years but it is unlikely that levels of skill change and challenges for measurement, for the EU, and in the EU area would warrant measurement. more broadly for the OECD economies. 4. Future assessment cycles should include non- 1. Adult skill profiles need to be measured OECD members First, assessing the quality of the flow of skill exiting The initial rounds of adult skill assessment have the initial education system, while necessary, is focussed on OECD member states in order to not a sufficient condition for supporting policy determine if comparative assessment offered analysis in the area of adult skills. Repeated new insights to policy makers. The fact that the measures of adult skills are urgently needed to IALS and ALL studies have documented inter- identify the relative contribution of skill gain and country differences in skill that are much larger loss to net skill change and to provide some sense than identified in equivalent student assessments, of the co-factors that lead to either result. and established that these differences exert a profound influence on individual outcomes and 2. Assessment designs should shed light on the the long term economic performance of nations, processes that underlie skill demand within firms justifies the expense of assessing adult skills. It is important to note in passing that household- based assessments such as IALS and ALL are It is clear that the global supply of these limited in their ability to understand the processes economically productive skills are rising rapidly that lead to skill loss in the EU. Ideally, one would in response to massive increases in educational need a longitudinal nested sample of workers investment and educational reform in non-OECD within firms to profile the heterogeneity of skill countries. Including non-OECD countries in the demand within industries and to understand what next round of assessment becomes a priority so factors condition variation in skill demand and that policy makers can assess the rate at which utilization. This understanding is a prerequisite their skill-based comparative advantage is being to striking an optimal balance between supply- eroded. With this in mind the EU should find a side measures aimed at increasing supply of skill, way to support the Unesco Institute for Statistics’ and demand-side measures that serve to ensure Literacy Assessment and Monitoring Program that the available supply of skill is put to effective (LAMP) that is designed to provide proficiency use. The presence of significant skill loss suggests estimates linked to the IALS and ALL scales. that additions to the skill supply may evaporate as quickly as they are produced if skill demand is 5. Assessment designs must provide information inadequate. Separate samples of the unemployed on the inter-skill covariance structure and not-in-the labour force populations would be Analysis suggests that one needs skill profiles for required to complete the picture. several skills derived from a design that allows one to understand the strength of any underlying 3. The natural cycle for skill assessments is 10 skill dependencies. The ALL study provided years direct assessments for four distinct skill domains A review of the available evidence suggests that – prose literacy, document literacy, numeracy skill assessments should be conducted on a ten and problem solving and one indirect measure year basis. This cycle is dictated by the real rate of ICT use. The fact that each of these domains of change observed in OECD countries and on contributes unique variance in key outcomes, the basic sensitivity of current measurement such as wages and employability, suggests that

188 KNOWLEDGE ECONOMY: HUMAN CAPITAL they are worth measuring in any EU-sponsored skills. It is clear that the rapid adoption of ICT’s assessment. are central to EU member states achieving the objectives set out in the Lisbon strategy 6. While theory and evidence suggests that – to become the most competitive and dynamic additional skills might be assessed in future knowledge-based economy in the world, capable cycles the rationale for government intervention of sustainable economic growth with more and suggests that this is not warranted (with the better jobs and greater social cohesion. It is notable exceptions of language and ICT skills) difficult to imagine what other changes might The argument could be made to argue that future bring the needed productivity gains in the face assessments should expand the range of skills of rapidly increasing competition from the assessed to include some of the soft skills such developing world. Given the strong dependencies as teamwork. between literacy, numeracy and problem solving skill posited by theory, and confirmed by the ALL Given the existence of a hierarchy of skills data, a convincing case can be made to include acquisition, however, an equally convincing a direct assessment of ICT skill use as proposed case can be made to limit assessment to the in the OECD PIAAC assessment. The requisite basic skills that are already being measured in theory and approaches to assessment already ALL. This argument rests on the assumption exist to do so. that government’s role in adult skills should be proscribed – that they should only measure, and A second argument can be made to retain intervene, in cases where there is clear evidence the IALS and ALL skill measures in future of a market failure that neither individuals nor assessment cycles. Many of the macroeconomic firms are able to rectify. effects attributed to differences in the level and distribution of literacy skill depend upon The case is clear for adult literacy, numeracy and the derivation of historical time series of problem solving. These are skills that are difficult skill. Synthetic cohort analyses of repeated to acquire through self-study – they require cross-sections will allow for most of the key considerable instruction to become proficient. assumptions in these derivations to be confirmed, The available evidence suggests that neither but these analyses demand a common metric. firms, nor individuals, are willing to make the requisite investments to acquire these skills. In sharp contrast, if firms want more subordinate workers capable of working in teams they have Evidence from the Canadian Level 1 study, the US the incentives to induce workers to behave as Adult Education and Learning Study and national needed. For their part, workers can easily change longitudinal studies in the US, Canada, Australia, their comportment without significant instruction. New Zealand and the UK suggest that a similar Neither workers nor firms need government argument can be made for oral language because interference, nor can taxpayers afford to finance of its relationship to literacy acquisition. This a program where the sole beneficiaries are the argument is supported by the fact that increased individuals and firms themselves. levels of immigration to, and within, the EU will create a strong need for language training to 7. Future assessment cycles should provide more enable social and economic integration. information on the learning needs of the least skilled The second exception to the argument for By design the IALS and ALL assessments restricting assessment to what is currently attempted to profile the full range of skill in measured in the ALL study, is the case of ICT selected domains and to relate the observed

KNOWLEDGE ECONOMY: HUMAN CAPITAL 189 differences in the level and distribution of skill without precipitating rapid increases in the social to outcomes at various levels. These designs inequality of economic outcomes. served the rhetorical purpose of establishing that In many ways, however, the analysis has suffered large differences in quality of adult labour forces because of the limited sample size available did indeed exist and that these differences were for many countries and due to the limited socially and economically meaningful. They did little, however, to help policy makers decide what heterogeneity of the systems studied. To date the learning needs of the least skilled might be. Canada, the US and Australia are the only nations Studies such as the US AEL , the Canadian Level that have fielded assessments that incorporate the 1 and the Unesco Institute for Statistics LAMP relatively large sample sizes that are needed to provide useful insight into the learning needs of support a nuanced policy analysis. This weakness individual adults, insight that will help policy has led the OECD to consider that the PIAAC makers design a new, more effective and efficient focus on a narrower age range. Adopting this generation of remedial programs. The proposed approach would preclude the analysis of inter- OECD PIAAC program would incorporate key cohort differences in skill and the impact that aspects of these designs. such differences have on the relative success of each cohort. In the end, the data flowing from 8. Future assessment cycles should incorporate international skill assessments is sufficiently larger samples of vulnerable populations (but not important that countries must invest the resources at the expense of covering the entire age range) needed assess large samples from a diverse range It is clear that the conduct of adult skill of countries. Without the information that would assessments is an expensive, operationally and flow from such assessments it is highly unlikely technically undertaking. It is equally clear that that the EU will become the most competitive the studies have offered new insight in matters and dynamic knowledge-based economy in the of pressing policy importance to many countries world, capable of sustainable economic growth of the world, including the dual challenge with more and better jobs and greater social of achieving high rates of economic growth cohesion.

190 KNOWLEDGE ECONOMY: HUMAN CAPITAL 6. References OECD (1998), Human Capital Investment: An International Comparison, Paris ETS (2003), Digital transformation: A framework for ICT Literacy, Princeton, NJ. Statistics Canada and HRSDC (2004), Literacy scores, human capital and growth across fourteen Organization for Economic Cooperation and OECD countries, Coulombe, Tremblay and Development and Human Resources Development Marchand, authors, Ottawa Canada, (1997), Literacy Skills for the Knowledge Society: Further Results of the International Statistics Canada and OECD (2005), Learning Adult Literacy Survey, Paris and Ottawa a Living: First results of the Adult Literacy and Life Skills Survey, Ottawa and Paris Statistics Canada and HRDC (2001), Literacy, numeracy and labour market outcomes in Statistics Canada and HRSDC (2006), Human Canada, No 89-552-MIE, No. 8, Ottawa Capital and Canadian Provincial Standards of Living, Coulombe and Tremblay, authors, Ottawa ETS (2004), Literacy and Health in America, (in press) Rudd, Kirsch and Yamamoto, Princeton Murray and Clermont (2006), Health Literacy in Desjardins (2004), Learning for well being: Studies Canada, Ottawa (in press) using the International Adult Literacy Survey, Stockholm University, Stockholm

KNOWLEDGE ECONOMY: HUMAN CAPITAL 191 Challenges in Measuring Human Capital for the Knowledge Economy

Albert Tuijnman Senior Economist, Human Capital Division European Investment Bank Luxembourg

Introduction The onset of the industrial revolution once introduced the need for mass formal education. The purpose of this paper is to describe some of Developments thereafter continued to exert the challenges the statistical system is faced with pressures and raise expectations and demands in the education sector as a consequence of the on formal education systems. By the early 21st decision taken by the economically advanced century education itself had become an economic countries to adopt a life-long learning framework sector of considerable strategic importance. and strategy in response to the move towards the Policy problems are encountered, however, new global economy. because the scope of change in the global knowledge economy is such that governments can no longer depend solely on the formal school New Economy system for the supply of the human capital needed by the nation. For example, a policy of gradually Globalisation provides a backdrop for analysing expanding enrolments and improving the quality economic and social changes and concomitant of schooling so as to meet the demand for skills changes in the education and learning sector. and competencies generated by the labour market Globalisation is not a new phenomenon but will be found to be wanting, because of the time the speed and scope of change are new, mostly lag involved. as a function of the widespread diffusion of information and communication technologies. Jobs in the knowledge economy require a good Developments have altered the social and deal of formal education. But in addition to good economic policy landscape of countries and entry qualifications they demand continuous placed a new premium on both human capital and learning, flexibility, excellent literacy, numeracy social capital (OECD, 2001a). This is because and problem solving skills and the ability to governments can no longer rely on the same acquire and apply new ideas and use knowledge range of policy instruments they once employed creatively. Knowledge management is an to regulate the industrial welfare state. A global increasingly central element in the organisation economy requires governments to develop a new of work and this puts a premium on personal approach not only to trade and fiscal policies but qualities such as the ability to work flexibly in also to structural policies. As the scope for state teams, initiative, creativity and entrepreneurship. intervention in the economic sphere has become Rather than applying a definition that equates more and more constrained policy makers have human capital with a given stock of educational increasingly had to shift their attention to the qualifications, the knowledge economy demands ‘residual’ factors in the production function, a dynamic view that emphasises actual skills and principally technology and human capital. learning outcomes.

192 KNOWLEDGE ECONOMY: HUMAN CAPITAL Major educational challenges follow in the factors in conditioning the efficient allocation wake of these shifting perspectives, both for of primary production factors such as financial schools and for adult education. A large non- capital and labour. Intangibles, conveniently sub- formal training effort also accompanies the move sumed under the label ‘knowledge’, are entered towards the knowledge economy. Over and above as externalities and confer scale economies to the this, the nature of learning in the workplace other factors in production. Important therefore is is also changing, with informal, self-directed not only the optimisation of primary production and team learning gaining on formal classroom factors but especially their optimal allocation. instruction. Improving the knowledge, skills and Life-long learning is implicated in this analysis competencies of the labour force, as part of a because learning itself lies at the heart of the cap- broader strategy for realising life-long learning for ital allocation and conversion process. As such, all, are central to the policy strategies now being learning outcomes are treated as multi-dimen- pursued in all advanced industrial societies in an sional to capture the direct and indirect effects effort to exploit the new economic environment that learning may have on various outcomes. by strengthening the capacity of labour markets Moreover, to expose the multi-dimensionality of to adjust to change, improve productivity and learning, the different forms of learning which capitalise on technological innovation. In this span the ‘life wide’ spectrum of life-long learning sense, ‘life-long learning’ is no mere slogan or (OECD, 1996) are separated to depict how they empty rhetoric because it defines human life itself can potentially relate to each other and in turn in 21st century societies. what their potential contributions are to different learning outcomes. Life-long Learning The general approach draws an analogy to the re- The new interest in life-long learning reflects source conversion process developed by Coleman changes in theory as much as changes in the pol- (1971; 1990). Coleman based his approach on a icy environment, which were alluded to above. convention commonly used by economists, name- For example, current thinking about the econom- ly an input-output analysis or activity analysis ic role of life-long learning has been influenced method. Figure 1 presents a capital allocation and by so-called ‘new growth’ theories, which assign conversion model. Financial, social and human re- much more prominence to the role of intangible sources available at time t0 are converted into other

Figure 1. A Capital Conversion Model

Source: Desjardins and Tuijnman (in press).

KNOWLEDGE ECONOMY: HUMAN CAPITAL 193 financial, social and human resources at time 1 t . should provide everyone with an adequate Depending on how the conversion process plays foundation for later acquisition of new knowledge out with time, the weight of each resource compo- and skills. Second, education systems should also nent can remain the same, or grow or be reduced. more than in the past facilitate the movement of According to Coleman, the education system is at workers across internal labour markets. Third, the heart of the resource conversion process. Edu- because of their deepening links with labour cation serves as a means for reproducing wealth, markets and the economy, education systems social status and knowledge and skills. would have to respond more flexibly to changes in individual demand. Accordingly, the two central From an economic perspective, skill formation tendencies in education policy responses were to involves the allocation of scarce resources to emphasise the importance of consolidation and a learning process that is designed to produce quality in broad-based foundation learning while competence. This process can be represented promoting diversification in the supply of tertiary in an educational production function, which is and continuing education and training. Thus the a mathematical expression that relates inputs to possibility of acquiring new knowledge, skills outputs. Many factors can be included in such and competencies would no longer depend on a function. For example, financial, human and obtaining a formally prescribed education at any physical capital items are usually included among given age. Instead, learning would become the the input factors, whereas educational attainment tool of the individual -- available to him or her and different skills, values and attitudes are at any age. among the commonly measured outputs. The challenge of making the skill formation process In sum, the concept of life-long learning is best cost-effective thus refers to the search for efficient understood as a philosophical as well as a policy ways of converting or substituting resources. device. It is indeed ‘life-long’ -- referring to a Life-long learning is advocated because it is process of individual learning and development believed to offer a flexible and efficient way of across the entire life span, from cradle to grave. organising the skill formation process and so But the concept also has a ‘life-wide’ dimension -- developing the human, cultural and social capital referring not only to formal education provided in that is needed in pursuing a high-skill, high- institutions but also to non-formal learning at work wage jobs strategy. There are several problems, and informal learning occurring at home and in however, and an important one is that it usually daily life. Furthermore, the emphasis on learning takes a long time before the conversion of rather than education is highly significant because financial capital into intangibles such as human it reduces the preoccupation with social structures or social capital pays off (Coleman, 1988), for and instead focuses on individual demand. example, in terms of increased employment, productivity and economic growth. Life-long The role of government in organising and learning thus represents an insurance policy to managing the education system changes minimise the risks that market failures occur as significantly with the adoption of a life-long a consequence of uncertainty over the costs and learning approach, because this involves a shift benefits of investment in intangibles (Tuijnman, from a supply led to a demand led model. Public 1996; Tuijnman and Schömann, 1996). authorities can no longer be the sole provider of resources, nor necessarily determine curriculum The above considerations have influenced content, achievement standards or qualifications, education policy in multiple ways (see Tuijnman, even though governments are still required to 1999). First, they led to a new appreciation of the act in a steering capacity and provide a legal importance of foundation skills: formal schooling framework. To be able to steer in this new role

194 KNOWLEDGE ECONOMY: HUMAN CAPITAL government will need access to information not schooling and the actual labour market skills of hitherto supplied by the statistical system. workers is expected to decrease with increasing experience. Information Needs In order to monitor progress in life-long learning Many information needs are associated with the governments require new information on idea and principles of life-long learning. Today learning outcomes, information that goes well only some of these are met by the statistical beyond indirect measures based on qualifications systems of the economically advanced countries. conferred by the education system. Information on educational attainment does not suffice as Systematic knowledge about the levels of a measure of human capital stock, given that educational attainment and the population people continue to learn beyond schooling. The distribution of labour force qualifications are knowledge and skills that are acquired at work prerequisites for formulating sound human and elsewhere are not normally reflected in resources development policy. But labour force conventional measures of educational attainment. qualifications do not necessarily square with the In comprehending the full extent of the learning highly aggregate attainment levels commonly efforts made by people of all ages, account must employed in manpower planning and forecasting be taken of the fact that learning is a defining studies. Furthermore, labour force qualifications do characteristic of all human activity. The learning not necessarily correspond closely to the skills and that goes on in schools, colleges and universities, competencies possessed by people and required adult education centres and employer-sponsored for jobs in a fast changing labour market. This lack training is only a part, albeit an important one, of of fit between educational attainment and actual the total learning effort of the population. labour force skills can represent a problem. A further issue is that the skill requirements of jobs In the absence of direct observations on the skills are difficult to define and measure with sufficient and competencies of people in the labour force, accuracy. Part of the problem is that jobs change statisticians and social scientists traditionally use and skill requirements evolve with time. A skill proxy variables such as the number of years of mix deemed sufficient for work in the industrial schooling completed or the highest educational society can be inadequate in the knowledge credential attained. This may be admissible economy. The importance of a given skill to in so far as actually possessed competence is employment and the level of performance in that correlated highly with measures of obtained skill, are therefore relative concepts. Accordingly, schooling. However, a serious threat to validity skill gaps in the adult population cannot be taken arises if competence and schooling are only as prima facie evidence of current educational weakly related. Such a situation can occur for failure, partly because skill requirements continue several reasons. For example, people who enter to rise, and partly because the skill profiles of the labour market with similar educational populations are a product of a multiplicity of qualifications have not necessarily acquired the factors working over a long time span. same level of proficiency in solving problems or managing interpersonal relations. Second, Because people learn on the job and develop adult discrepancies will arise because people do not roles in community and work, relying for labour stop learning upon leaving school. Because market allocation decisions on an imprecise opportunity to learn varies depending on a host measure such as nominal educational attainment of personal, situational and economic variables, is certain to misrepresent the actual stock of skills the strength of the relationship between initial available for the labour market. The inability to

KNOWLEDGE ECONOMY: HUMAN CAPITAL 195 take account of the knowledge and skills people for accreditation and quality assurance -- hold a acquire beyond the school system can result monopoly over assessment and certification. This in price distortions and market failures. Such monopoly served useful policy goals framed in failures occur if the amounts of education or the context of the former industrial welfare state. learning supplied or demanded are socially or individually inefficient, leading to either under- That industrial society was characterised by stable or over-investment in human capital. employment patterns and clear-cut occupational categories. Standardisation of both occupations Market failures are a reality in the education sector and the educational qualifications that legitimised because it is information poor. Because of this access to them was an unassailable logic. But lack of information wrong investment decisions if the knowledge economy requires flexibility are made, resulting in errors of skill allocation and adaptability, as was argued above, then and distribution. The resulting skill mismatches the assessment, recognition and qualification in turn lead to inequity and inefficiency. In the frameworks that are applied by educational interest of enhancing labour market flexibility systems also need to be flexible. This is generally and improving skills, governments will seek to not the case at present. Much economically curb market failure. Doing so depends crucially useful knowledge acquired at work and in on the supply of information about the private other life settings is not reflected in educational and social costs and benefits of human capital qualifications. investment. But this cost-benefit analysis hinges on valuations of human capital, and hence it In comprehending the full extent of the learning requires that knowledge, skills and attitudes efforts made by people of all ages, account must are somehow measured. Whereas the costs of be taken of the fact that learning is a defining many inputs are known, for example in the form characteristic of all human activity. The learning of education and training expenditures, this is that goes on in schools, colleges and universities, not true for the economic value of the resulting adult education centres and employer-sponsored knowledge increase, which cannot readily be training is only a part, albeit an important one, quantified. Poor proxy variables are commonly of the total adult learning effort. Present-day used in the absence of direct output measures, for qualification frameworks are inadequate also for example, aggregate levels of qualifications defined reasons other than content validity; for example, in accordance with the International Standard they can block access to jobs and further education Classification of Education (ISCED). Improving and training, and make labour markets more rigid. the means of identifying, assessing and certifying essential work skills is an important aim of policy The implementation of a life-long learning strategies for reducing market failure. strategy demands a new approach to the definition and selection of competencies necessary for work Not only governments and employers need and life more generally (see Salganik, 2001; better data about the nature and distribution of Weinert, 2001; Gilomen, 2002). In a framework essential work skills in the population. This holds of life-long learning, the task of school is to true also for the institutional suppliers of adult ensure that all students graduate and leave the learning. Difficulties are encountered in practice, system with requisite foundation skills. These however, because the assessment and validation skills will need to be certified in some form. But of knowledge and skills are traditionally other skills will need to be acquired beyond this important functions of the education system. In initial stage, and they will have to be applied in many countries formal institutions -- and more the work setting. Accordingly, in the knowledge particularly, the public authorities responsible economy people will not only have to keep up

196 KNOWLEDGE ECONOMY: HUMAN CAPITAL and renew their foundation skills, they will also Life-long learning thus brings a number of large be called upon to acquire new knowledge and conceptual and methodological challenges to skills, which have to be somehow assessed, the statistical system. Five challenges for data recognised and valued. development are singled out below.

Insight into the distributions and levels of Measuring Innovations in Formal Education essential work force skills is a prerequisite for formulating and implementing targeted and cost- On the surface there have been few moves effective human resources development policies. towards the knowledge economy in the education Unfortunately, the currently available knowledge sector. There has been significant investment in about the nature of such necessary skills, their information technology hardware but there is determinants and interrelationships, is scanty. little evidence that major changes have occurred A competency-based approach should be at the in teaching and learning. It could be that the heart of the statistical system needed for the problem is not merely that the education sector monitoring of progress in life-long learning, and is inherently conservative but also that current this calls for the direct measurement of skills and statistics fail to detect change. This may be the competencies. Much work is needed to develop case because the statistical system is front-loaded; the conceptual frameworks and measurement it is mostly concerned with measures of inputs technologies that are needed to achieve this. into education, such as how much money for how many students, teachers, classrooms, and so forth. The important innovations a life-long learning Implications for the Statistical System strategy is expected to bring to the education sector are not concerned with inputs but rather For statisticians interested in data development with process. A mere glance at compilations of for monitoring and evaluation, the wide-ranging conventional education statistics will show that orientation of life-long learning poses near they have very little to tell about the learning unsurpassable conceptual problems. Because it process. A lot of thought should therefore be given is not tied to any institutional context it requires to the measures that ideally are needed to monitor analysts to take a large, holistic perspective. They changes in pedagogy and in the way individual should consider the whole range of education students are engaged in their individual learning provision extending from pre-schooling and projects. care, through all stages of education at primary, secondary and tertiary levels, to continuing Measuring Life-wide Learning vocational training in educational and labour- market institutions, informal learning on the job, The statistical system has worked for many and self-directed and co-operative learning at years to develop adequate measures to cover large in society. Methodological problems arise the life-long dimension of education, and it because it is not possible to draw a clear boundary has successfully developed a range of finance, between what can be considered learning enrolment and graduation statistics and indicators activities and the range of other experiential and that measure aspects of formal institutional behavioural activities in which people engage. provision extending from kindergartens and early The all-embracing nature of the concept of life- childhood education through universities and long learning, as currently embraced by the OECD colleges. Available statistics on adult education and other international organisations, has certain and the vocational training sector tend to be far drawbacks, among them the risk of dispersion, a fewer and less satisfactory, mainly because they loss of focus and the difficulty of assigning and are derived from institutional or administrative evaluating priorities. sources. In many countries there still is a need

KNOWLEDGE ECONOMY: HUMAN CAPITAL 197 for improving the coverage of statistics on adult Measuring Cumulative Learning Across learning. the Life-span

In addition to this, a large challenge for the statistical The collection of European data on adult learn- system is to extend measurement along the life- ing is expected to fill an important data gap. wide axis of life-long learning. This will require Because the survey will also collect socio-de- moving beyond the formal, institutional setting and mographic background information from the into the unchartered terrain of non-formal learning respondents as well as data on several social, at work and informal learning in daily life. Because economic and labour market outcomes, it will learning pathways are individually defined and offer interesting information for policy analysis. because adults – as opposed to school children But what will still be missing from the picture is -- are a non-captive population, this will require the longitudinal dimension inherent in the con- the development of new measurement instruments cept of life-long learning. Life-long learning is a combined with a labour force or household survey cumulative process that cannot be captured with approach to data collection. a cross-sectional data set. Hence the need is for a longitudinal approach to data collection that Fortunately, in recent years steps have been taken will allow for an investigation of the interactive to develop new instruments for the measurement as well as cumulative effects of formal, non-for- of life-wide learning. An experts group working mal and informal learning activities on indicators within the framework of Network B of the OECD/ of financial, human and social capital. A survey INES project has developed a module, entitled over time is proposed to investigate the links and “Participation in Education and Learning” (PEL). interfaces between different learning contexts, This module, which built on the adult education providers and sectors. This will include informa- and training module that was first used in the tion on transitions and pathways between initial International Adult Literacy Survey (OECD and learning, work and further learning as well as the Statistics Canada, 1995; 2000), has been further complementarity or substitutability among them. developed by an experts group associated with Because these effects can only be captured over EUROSTAT and will this year be fielded as time, the design of such a survey should uniquely part of the European Labour Force Survey. This contribute to policy relevant lifelong learning reis expected to yield comparable data on adult search that other studies are unable to do because learning and continuing vocational training for they only provide a snap shot in time. Figure 2 an important group of countries. shows a conceptual model of the survey design

Figure 2. Lifelong Learning Contexts

198 KNOWLEDGE ECONOMY: HUMAN CAPITAL that is ideally required to measure both life-long diagnosis, selection and accountability; hence the and life-wide learning. necessity to clarify the purposes of measurement. The cognitive theory underlying the application of Measuring Occupational Change statistical models and technology in skill testing must also be clarified. A third difficult problem The International Standard Classification of concerns the definition and use of performance Occupations (ISCO) is based on the notion of levels, recognising that essential skills are not competence, defined as “the ability to carry out something one either has or lacks, but that each the tasks and duties of a given job” (ILO, 1990, skill involves a range of ability distributed along p. 2), with abilities characterised by their level a continuum denoting low to high performance. of complexity and their area of specialisation. However, because of the difficulties encountered At the international level three avenues of work in assessing the competencies required by on the direct measurement of skills have been specific jobs, and because objective measures explored in recent years. The first line of work of task complexity are not available, at least not has been the International Adult Literacy Survey internationally, the ISCO system depends on the (IALS), conducted between 1994 and 1998, International Standard Classification of Education through which data were collected on the literacy (ISCED). That is, even though competence is seen profiles of adult populations aged 16-65 in22 as an ability that can be acquired independently countries (OECD and Statistics Canada, 1995; of educational programmes, in operational terms 2000). The second approach has been to collect educational qualifications linked to type and comparative data on student achievement in duration of programmes provide a yardstick mathematics, science and reading literacy among for determining an occupation’s place in the 10 and 14-year-olds as part of a cycle of surveys hierarchy. This example shows that even though conducted under the auspices of the International skills and competencies are abilities that people Association for the Evaluation of Educational acquire in many settings, and that flexible labour Achievement (IEA; Mullis et al., 1997). New IEA markets ought to assess, recognise and value surveys will be launched in 2003 and 2004. The independently of the structures and programmes third line of work has been to assess the reading of educational systems, in practice the education literacy skills of 15-year-old students as part of and training system, through its assessment and an on-going cycle of surveys implemented by certification procedures linked to programme the OECD under its Programme for International orientation, intensity and duration, defines Student Assessment (PISA; OECD, 2001b). labour market qualifications. This monopoly can introduce an element of rigidity in labour The large challenge now is to build on these markets; it also introduces market failures in efforts. Several OECD countries have expressed education and training. The revision of the ISCED an interest in fielding a comparative survey to classification that was completed in recent years measure the distribution in the adult population of now necessitates a review and revision of the a range of skills thought to be important to social ISCO system. and economic success. The Adult Literacy and Life-skills survey (ALL) is being developed in Measuring Competencies and Skills Directly response to this need. Patterned on the International Adult Literacy Survey (IALS), the ALL will see Key to the problem is how to assess and certify the administration of direct performance tests to the essential skills or key competencies in a valid, representative samples of adults aged 16 to 65. reliable, timely and cost-effective way. Assessment In addition to the assessment of skill domains can be performed for a number of reasons, such as such as prose and document literacy, numeracy

KNOWLEDGE ECONOMY: HUMAN CAPITAL 199 and analytical reasoning as a sub-domain of • understand better the involvement of formal problem solving, the survey will also collect education in the skill formation of its adult information about the use of ICT skills and workforce; teamwork skills, information on the respondent’s • identify labor-market correlates to illuminate work force participation, adult learning, literacy the role of such skills in generating activities, and the actual demand for skill use in economic growth and productivity, economic the workplace. The goal of the survey is to profile opportunities, and social cohesion; and and compare the distribution of skill domains that • compare the similarities and differences in the are part of a larger set of life skills. assessed skill domains among participating countries. Like its predecessor, the ALL survey is developed by an international team of experts sponsored Several participating countries have already by the governments of Canada and the United successfully collected and analysed pilot data. States. International direction and management The main survey will be implemented in 2002 are in the hands of T. Scott Murray of Statistics and 2003 and is expected to yield the first data Canada. The survey is expected to yield a for analysis in 2004. comprehensive set of data and produce reports to profile and compare the distribution of life skills Measuring workers’ skills within firms in the adult population and for various subgroups within countries. The reports are expected to The ALL survey will provide a new and rich data highlight the similarities and differences in the set that will allow analysts to study the interactions skill domains and the interrelationship among and effects of learning and skills on a range of them within and across participating countries. outcome measures. This will leave a fifth piece of the life-long learning puzzle still missing. Non- Government agencies and policymakers will be formal learning at work and continuing vocational able to use these data and reports for international training sponsored by firms and public employers comparisons to inform their policy making for both represent large learning sectors about which the formation and maintenance of adult skills. few official statistics have been collected to date. More specifically, each participating country will To complement to the surveys already described be able to: above, therefore, what is needed is a survey of worker knowledge, skills and learning within • build on the International Adult Literacy firms. What might be gained from such a general Survey (IALS) data to monitor changes over survey of workers’ knowledge, skills and learning time in the literacy skills of its adult population is, first of all, the possibility of directly testing how with state-of-the-art survey frameworks and these are related to productivity. Second, direct data analysis methodologies; evidence about how general and firm-specific • start collecting trend data on an expanded set human capital are formed and how this affects of adult skill measures that are collectively productivity would allow teachers and curriculum identified, internationally, as skills relevant for developers to validate what they ask students to employability; learn. This kind of information would begin to • assess the distributions of those workplace fill a great void in the research to date (Stern skills for the adult population as a whole; and Tuijnman, 1997). Such information would • identify sub-populations of adults within the also begin to create a more solid basis for policy country who are at risk because of their low on work-based education and training. Much skill levels; current policy discussion rests on assumptions

200 KNOWLEDGE ECONOMY: HUMAN CAPITAL about the contribution of training to the economy different environments. Last but not least, such -- assumptions that cannot be tested because the an international collaborative project would yield requisite data are lacking. new information and insights necessary for policy analysis and the improvement of adult learning for the new economy. Conclusions

Market failures currently stand in the way of Given the scope and volume of the learning realising life-long learning for all. Implementing activities that occur in sectors where the life-long learning is widely seen as one element education authorities normally exercise little of a policy strategy aimed at facilitating the control or responsibility for management, it is transition to the knowledge economy, while easing clear that the information infrastructure for life- some of the pressure this transition is expected long learning needs to be diverse, yet inclusive: it to bring for individual citizens and communities. needs to comprise comparable indicators not only Lack of information of various kinds is the root of the contexts, inputs, processes and multiple cause of market failure. Governments therefore outcomes of formal education as well as non- have an interest in improving the knowledge base formal and informal learning across the life span, of education, training and learning, so that better but the information also needs to be collected investment and training choices can be made. and presented at several levels of aggregation. The assessment and validation of skills acquired Pre-schooling, tertiary education for young outside of the school setting are elements of a adults and senior citizens, on-the-job training strategy to improve the quality of information as well as informal learning at home and in the supply. Much work has already been done on community, whether undertaken for investment the conceptualisation and measurement of skills or consumption purposes, all need to have their and competencies during the last decade (for proper place in the information system. an overview see Salganik, 2001). But so far the progress has been insufficient to satisfy the Because of this inclusive perspective, a variety of requirements. Research work on the assessment avenues to data collection must be followed. The of different adult skills therefore needs to be demand is for administrative and institutionally given strong support. based statistics, national accounting data, survey- derived measures of adult learning, as well as Another issue is that the efforts of the researchers longitudinal surveys of living conditions and time working in this field need focusing. One way of and money spent on various learning activities. In achieving this might be to enlarge the on-going addition, there is the need for process information, international Adult Literacy and Life-skills survey which can only be collected in the field. Multiple and encourage a range of additional countries indicators organised in a multilevel framework - to take part in the data collection. The benefits - and hence multiple information sources -- are of such a large-scale project are potentially required for the monitoring of progress towards enormous. It might offer a way of aligning better the implementation of life-long learning for the research agendas of various interested parties, all. Given the current state-of-play of the social improving collaboration among the teams now sciences and in particular of survey practice and working on similar problems but in different indicator measurement, the time when a holistic countries, facilitating the development of theory and comprehensive framework of life-long in cross-cultural settings, and establishing learning statistics and indicators can be proposed the relative importance of essential skills in lies still far in the future.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 201 References OECD and Statistics Canada (1995). Literacy, Economy and Society. Paris and Ottawa: Coleman, J.S. (1971). Resources for Social Authors. Change: Race in the United States. New York: John Wiley & Sons. OECD and Statistics Canada (2000). Literacy in the Information Age: Final Report of the Coleman, J.S. (1988). Social capital in the International Adult Literacy Survey. Paris and creation of human capital, American Journal of Ottawa: Authors. Sociology, Vol. 94(Supplement), S95-S120. Salganik, L.H. (2001). Competencies for life: Coleman, J.S. (1990). Foundations of Social A conceptual and empirical challenge. In D.S. Theory. Cambridge, Massachusetts: The Belknap Rychen and L.H. Salganik (Eds), Defining and Press of Harvard University Press. Selecting Key Competencies (pp. 17-32). Bern: Hogrefe & Huber Publishers. Desjardins, R. and Tuijnman, A.C. (in press). A general approach for using data in the comparative Stern, D., & Tuijnman, A.C. (1997). ”Adult Basic analysis of learning outcomes. Forthcoming Skills: Policy Issues and a Research Agenda”, in in J. Gaskal and K. Rubenson (Eds), Debating A.C. Tuijnman, I.S. Kirsch & D.A. Wagner (Eds), Learning Outcomes. Toronto: University of Adult Basic Skills: Innovations in Measurement and Toronto Press. Policy Analysis (pp. 1-16). New York: Hampton Press. Gilomen, H. (2002). Education in the new economy: A statistical framework for strategic Tuijnman, A.C. (1996). “Economics of adult monitoring and management. Paper presented education and training”. In A.C. Tuijnman at the IAOS conference, Official Statistics and (Ed.), The International Encyclopedia of Adult the New Economy, London, 27-29 August 2002. Education and Training (pp. 124-131). Oxford: International Association of Official Statistics. Pergamon Press.

Mullis, I.V.S., Martin, M.O., Beaton, A.E., Tuijnman, A.C. (1999). “Lifelong learning policies Gonzalez, E.J., Kelly, D.L. and Smith, T.A. (1997). in a new technological era”. In S. Mitter and M.I. Mathematics Achievement in the Primary School Bastos (Eds), Europe and Developing Countries Years. Boston College: TIMSS International in the Globalised Information Economy (pp. Study Center and IEA. 155-167). New York: United Nations University Press. OECD (1996). Lifelong Learning for All. Paris: Organisation for Economic Co-operation and Tuijnman, A.C. and Schömann, K. (1996). Life- Development. long learning and skill formation. In G. Schmid, J. O’Reilly and K. Schömann (Eds), International OECD (2001a). The Well-being of Nations: Handbook of Labour Market Policy and The Role of Human and Social Capital. Paris: Evaluation (pp. 462-488). Cheltenham: Edward Organisation for Economic Co-operation and Elgar. Development. Weinert, F.E. (2001). Concept of competence: A conceptual clarification. In D.S. Rychen and OECD (2001b). Knowledge and Skills for Life: L.H. Salganik (Eds), Defining and Selecting First Results from PISA 2000. Paris: Organisation Key Competencies (pp. 45-65). Bern: Hogrefe & for Economic Co-operation and Development. Huber Publishers.

202 KNOWLEDGE ECONOMY: HUMAN CAPITAL About the Author leads an international research programme on lifelong learning and acts as an education policy Mr. Albert Tuijnman obtained his Ph.D. from advisor in countries as diverse as Sweden, Sri Stockholm University in 1989 and was appointed Lanka and the United States. At present he is to the Chair of Comparative Education in Sweden also assisting in the development of the Adult in April 2001. He is also the Director of the Literacy and Life-skills survey. His main interests Institute of International Education at Stockholm are in comparative education, adult learning and University (www.interped.su.se). He was cognition, the measurement of foundation skills, appointed Honourary Professor of Continuing and education policy analysis. He has written and Education at the University of Nottingham in edited over 25 books and 100 papers in his fields 1998, a position he still occupies. Dr. Tuijnman of interest.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 203 Human Capital Statistics in the EU Present state and possible improvements

by George Psacharopoulos European Expert Network on Economics of Education (EENEE) E-mail: [email protected]

Presented at the Conference Knowledge Economy: Challenges for Measurement Eurostat, Luxembourg 8-9 December 2005

Abstract knowledge to achieve the economic and social objectives. The paper takes a critical look at the present state of human capital statistics in the European Union, As an introduction to the statistics needed against the background of the Lisbon agenda. to support the implementation of the Lisbon It identifies several potential weaknesses such strategy, as modified and enhanced in 2005, I as the use of multiple instruments to measure a start by quoting the relevant paragraphs of the single variable, the extreme comprehensiveness April 2005 communiqué and their implications of some questionnaires, the frustration of missing for statistics (Commission of the European data for many countries, the long time interval Communities, 2005): between surveys and the late availability of the • The realisation of a knowledge society … databases. It points at the scarcity of indicators is key to boost our growth potential .” on some key human variables, such as the sharing of the cost of training between the employer and In order to propose statistics towards this the employee, private expenditures on education, objective, one needs to define what is a and equity measures on the who really pays and “knowledge society”. How is “knowledge” who benefits from public education expenditure. operationalized in the first place? The paper concludes by a series of suggestions • “Making Europe a more attractive on how the indicators could be improved. investment location and spurring investment in knowledge”. I. Introduction This calls for statistics on institutions, e.g. corporate and personal tax rates, remuneration of In March 2000, the European Heads of States professionals and the research environment. and Governments met in Lisbon and agreed to • “Knowledge transfer via researcher make the EU “the most competitive and dynamic mobility”. knowledge-driven economy by 2010” – widely known as the Lisbon strategy. Given slow To achieve this, one needs statistics on relative progress towards this objective, at the 2005 mid- pay and work conditions in competitive countries, term review the European Council enhanced the such as the United States, Canada, Australia, and Lisbon strategy by putting special emphasis on recently India and China. This calls for statistics

204 KNOWLEDGE ECONOMY: HUMAN CAPITAL related to the brain drain and cross-border up to age 65 or 80, and what is the tradeoff movement of workers. between educating a 40 year old vs. a 60 years • “Increase investment in human capital old? through better education and skills.” • “Knowledge-based and service-based economies require different skills from The purpose is laudable, but one has to draw a traditional industries”. line between better and worse education, calling for statistics on school and university quality – a This calls for statistics on the skill and competence very tall agenda. Also, one needs to define what content of new occupational titles such as “genetic “skills” are particularly needed, and what skills engineer”. might be now obsolete. • “Skills …. need updating in the face of • “Europe needs to invest more in human technological change and innovation. capital.” Workers….. need to accumulate and renew skills regularly..” This is a moot point, because there are many different kinds of human capital, spanning from Of course they do, but this calls for statistics on formal education to training on the job, lifelong who will pay for such updating? The worker, learning and health. Since one cannot do all, what the enterprise, the Government, or what share statistics are needed to establish priorities? thereof? • “Too many people fail to enter or remain • “The productivity of enterprises is in the labour market because of a lack of dependent on building and maintaining a skills, or due to skills mismatches”. workforce that can adapt to change.” This is one of the most concrete statements in the This raises again the need for statistics on who will communiqué, calling for statistics on why some pay for the creation of an adaptable workforce? people do not participate in the labor market, and • “Governments need to ensure that how a skill mismatch is defined. educational attainment levels are improved • “To enhance access to employment for all and that young people are equipped with ages, raise productivity levels and quality the necessary key competences, in line with at work, the EU needs higher and more the European Youth Pact”. effective investment in human capital How could Governments ensure this? This calls and lifelong learning for the benefit of for statistics on public and private education individuals, enterprises, the economy and financing. Also on establishing a tradeoff between society.” youth education and lifelong learning. This is another concrete statement calling for • “All stakeholders should be mobilised to statistics to determine the effectiveness of various develop and foster a true culture of lifelong forms of investment in human capital, from the learning from the earliest age.” point of view of the individual, the enterprise, This calls for statistics on why at present and society at large. stakeholders are not mobilized towards these • “Member States are committed to causes? establishing comprehensive lifelong • “To achieve a substantial increase in public learning strategies by 2006”. and private investment in human resources In order to do this we need measures that go per capita, it is important to ensure fair beyond the norm found in statistical yearbooks, and transparent sharing of costs and let alone defining what “lifelong” means, e.g., responsibilities between all actors.”

KNOWLEDGE ECONOMY: HUMAN CAPITAL 205 This calls for data on the public and private This calls for data on the current mobility sharing of the cost of education, as well defining obstacles. the fairness of the distribution. • “Better identification of occupational • “Establishment of … appropriate needs…. and anticipation of future skill incentives and cost-sharing mechanisms requirements”. for enterprises, public authorities and This raises the issue on whether, in this age of individuals”. rapid technological change, required skills could This call for data on the current cost-sharing, be anticipated beyond the training period. It also as well as on the incentives individuals and calls for manpower forecasting, an activity that is enterprises respond to regarding cost sharing. now obsolete (Psacharopoulos 1984). • “Reduce the number of pupils leaving school early” II. Concepts and approach This call for data on understanding why some The above catalogue gives a taste of a very tall pupils drop out of school, linking to opportunities order for any statistical system. Many of the in the labor market. keywords used in Lisbon strategy are not helpful • “Increased access to initial vocational, in translating them to measurable variables. Take secondary and higher education, including for example the central keyword – “knowledge”. apprenticeships and entrepreneurship This can range from recording student enrollment training.” in schools and universities, to on-the-job training in enterprises, to lifelong learning, leading to another How does one force someone’s access to problem on how “lifelong” is defined. It could vocational training? What is the tradeoff between also mean “ideas” used in the new growth theory investment in the various types of training and (Romer 1990). Therefore, before any gathering higher education? This calls for statistics on of data, Lisbon concepts should be translated into the costs and benefits of various investments in concrete variables. This could be a major project human capital. on its own and beyond the scope of this paper. • “Enhanced participation in continuous and workplace training throughout the life- Another consideration is whether one should cycle, especially for the low-skilled and start from the existing statistics on human capital, older workers”. identify their weaknesses, criticize them and propose improvements. Although easy, this road This raises the issue of who will finance such might be the wrong start, as it would be too much investment? It calls for data on the incentives influenced by the state of the art that might not of workers and entrepreneurs to share the cost of be optimal. Or it would be equivalent at looking training. where the light is, rather than where the keys • “Lifelong learning systems must be were lost. affordable”. Although I give below a critique to existing Yes, but how? Who will make it affordable? statistics, I prefer to start from the kind of data This calls for data on the cost of various forms of one would ideally need to serve the Lisbon lifelong learning, including who bears such cost. strategy. This approach is driven by what the • “The remaining obstacles to mobility data would be used for, rather than whether there within the European labor market should are already tons of them in statistical yearbooks be lifted” or electronic databases. Actually, this approach

206 KNOWLEDGE ECONOMY: HUMAN CAPITAL may lead to revising and streamlining traditional ➢ Foregone earnings of those in school/training human capital indicators. ➢ After-formal-schooling, lifelong dimension of the above inputs Stripped down to the bare essentials of its spirit, the Lisbon agenda calls for efficiency and equity. Output measures On the efficiency side it is tacitly assumed that there is an underlying aggregate production ➢ School level - Cognitive achievement function ➢ Tertiary level – Number of Nobel Laureates, publications, citations, patents Y = f(……………, HC)

{ HC = g(…., Y) Ultimate indicators ➢ Incidence of unemployment by education where human capital (HC) plays a major role level/type/training in generating national income (Y), while the production of human capital uses national ➢ Duration of unemployment to first job resources. This macro framework is based on ➢ Earnings by education level/training type, contemporary endogenous growth theory. general or specific

The micro underpinning of the above, is that ➢ Longitudinal earnings data (panel/tracer investment in human capital yields returns ( r ) to studies) the individual and society at large, ➢ Returns to investment in education, private and social, by education level/field of study/ n WW− m ()u s t t = WC+ 1 + r , training type ∑t ∑()s u () t= m +1 ()1 + r t =1 where W refers to the earnings/productivity of Equity measures a more (subscript u) and a less (subscript s) educated person, and C is the cost of education. ➢ Private education expenditure by income (Becker 1993). decile ➢ Public education expenditure appropriated by The equity side of the Lisbon agenda is the income decile incidence of who pays and who benefits from ➢ public education expenditure. On-the-job training cost split between worker- employer To fully understand the contribution of HC to economic and social development, one needs Timing several classes of statistics: ➢ Rapid generation of data/analysis for policy Input measures, such as formulation

➢ Public expenditure by educational level/ Consistency training type, including local government ➢ Private expenditure by educational level/training ➢ Comparable statistics within-country over type, incurred by the trainee and the firm time, as well as cross-country.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 207 III. The existing statistics voluntary, and the comprehensives of the questionnaires cause many missing entries in the At first sight, there already exists a plethora of final tables, as well as a three-year gap between the statistics on the needed indicators. The richest time reference of the statistic and its availability collection might be found in the New Cronos, in the database. “Monitoring progress in the “Education and Training 2010” process” (under “Population Regarding out-of-school training, the Continuing and social conditions-->Education and training- Vocational Training Survey (CVTS) has been in ->Education-->Thematic indicators). And there fact discontinuous (available for reference years are many printed reports covering various sub- 1993, 1999, 2005), and misses workers in firms themes of the human capital statistics, e.g., with less than 10 employees. Eurostat’s “Key data on Education in Europe”, “Lifelong learning in Europe”, “Spending on Information on student cognitive achievement tertiary education in Europe”, “Final report of is based on the International Association for the the Task Force on Adult Education Survey”. Evaluation o f Educational Achievement (IEA) OECD’s “Education as a Glance” contains many and OECD’s Program for International Student selected relevant indicators. Assessment (PISA). Both share the limitation that they are irregular, highly dated by the time On closer scrutiny, however, the existing statistics of data availability, and in addition PISA refers have several limitations. Take for example to a very small segment of the student population the most critical variable on the link between – the 15 years old. education and growth, i.e. earnings by level and type of education. Statistics on adult literacy are spread out between the OECD’s discontinued International Adult The income variable is spread out in many Education Survey (IALS), the Adult Literacy overlapping and time-discontinuous surveys. and Life Skills Survey (ALL) conducted in 2003 The European Community Household Panel - the database becoming available in 2006 - and (ECHP) seems to contain the most critical income the Program for International Assessment of variables, but it was discontinued in 2001. It was Adult Competencies (PIAAC) scheduled to be replaced by the Statistics on Income and Living launched in 2010, i.e. quite late, according to the Conditions (EU-SILC), but no databases or Lisbon agenda. publications are yet available from this survey. Data on some indicators are extremely detailed, e.g. on student enrollment by school type and The general Labor Force Survey (LFS) and its region, whereas in other areas information is special modules raise a wealth of information. scant, e.g. on the sharing of the costs of training However, one of the most critical indicators – the between the employee and the firm, the private transition of young people from school to the cost of education and training, and especially labor market, is scheduled to take place in 2009, on the who pays and who benefits from public while the agreement on the variables it would educational expenditure. contain is planned for 2007. Information is also lacking on the institutional Information on participation in education and framework within which education takes place the public cost of education is raised by means and yields results. For example, plotting student of a Unesco/OECD/Eurostat questionnaire, and a achievement against the degree of centralization supplementary Eurostat education questionnaire. of an educational system gives a negative Since the answer to these questionnaires is relationship. The degree of centralization of an

208 KNOWLEDGE ECONOMY: HUMAN CAPITAL educational system has been dropped from the status, should come from longitudinal/panel recent edition of OECD’s Education as a Glance. data.

At present, there are too many “missing” entries Questionnaire streamlining. Questionnaires in statistical tables, perhaps a result of the should be simpler, asking only for information comprehensiveness of some of the questionnaires, on variables that will be actually used for policy. or the fact that reporting might be voluntary. (Psacharopoulos 1980, 1995). The most critical outcome indicator, the returns to investment in education, is reported only for Coverage. Beyond the EU-25, candidate ten countries in OECD’s Education at a Glance. countries should be included in the surveys. And the methodology on which these rates were computed is not very clear. Timely database availability. Questionnaire filling should be on line, e.g. by entering IV. Need for consolidation, simplification information on a laptop at the household door, or and timeliness directly into the database at the Ministry level.

One conclusion that stems from the above review Reduce legal procedures. Agreement on the is that the existing human capital databases are variables and questionnaires should take one perhaps good for providing historical information week, not two years. on some of the critical variables. However, their coverage and timing leave much to be desired for Engage researchers. Beyond Eurostat and formulating education and training policies. country officials, ask researchers on what variables they need in order to address knowledge The following improvements could be issues in the EU. considered: Make compulsory. Answering the streamlined One instrument per variable. For example, questionnaires should be made compulsory to information on income by level and type countries, as it is now information regarding of education could preferably come from a budget statistics. household, rather than enterprise, survey. Dissemination. After removing personal Continuity of the same instrument. Income and information, all databases should be made labor market information, such as employment immediately publicly available to researchers.

KNOWLEDGE ECONOMY: HUMAN CAPITAL 209 References Psacharopoulos, G., “Assessing Training Priorities in Developing Countries: Current Practice and Becker, G., Human Capital. University of Possible Alternatives.” International Labour Review 123, no. 5 (September-October 1984): Chicago Press, Second edition, 1993. 569-83.

Commission of the European Communities, Psacharopoulos, G., “Tracking the Performance of “Integrated guidelines for growth and jobs,” Education Programs: Evaluation Indicators.” New COM(2005) 141 final 2005/0057 (CNS). Brussels, Directions for Program Evaluation, No. 67 (Fall 12 April 2005. 1995): 93-104.

Psacharopoulos, G., “Questionnaire Surveys in Romer, Paul.. Endogenous Technological Educational Planning.” Comparative Education Change. Journal of Political Economy 98: 1990: 16, no. 2 (June 1980): 159-69. S71-S102.

210 KNOWLEDGE ECONOMY: HUMAN CAPITAL Report on the session: Competitiveness and Growth

Chair: Mr. Enrico Giovannini, OECD

In his opening, the chair set the frame for the the role of ICT and the contribution of capital session by referring to the Lisbon strategy, the deepening and total factor productivity on following need for assessment, comparison labour productivity growth. The speaker saw the and measurement of achievement. The two European Union as a whole lagging behind in presentations addressed the competitiveness of terms of productivity performance in the advanced the European Union and its measurement as a world. However, looking at individual Member basis for policy development. State productivity performance, variation can be found. Some Member States are performing The first presentation emphasised that European even better than the United States. Labour force competitiveness is dependent on many policies participation in many European countries has and that policy development and impact analysis increased, but the employment –productivity should be based on sound statistical basis. trade-offs are only temporary and disappear in the Industrial policy in the European Union is being medium term. The stated key to revive Europe’s revisited to take into account effects of increased growth is to create more and in particular more globalisation, technological change and increased productive jobs. societal demands, in essence, to get the Lisbon process back on track. With regards to ICT, the main problem in Europe seems to be on the user side. Relatively slower A description of the use of statistics and integration of ICT into production processes indicators on competitiveness in the Enterprise and less productive use of new technologies and Directorate General of the Commission was innovations across the economy, especially in made and future needs in light of the new the production of services and goods appear to industrial policy explained. Statistics are needed be the cause. A sectoral analysis into productivity on EU, national, regional, sectoral and enterprise performance reveals that especially market level. Comparability of data within Europe is services have contributed towards higher not enough as comparisons and benchmarking productivity growth in the United States, but only with competitors and emerging economies a limited number of service industries account for are important. Timeliness is very important the USA-EU differential. in particular in areas where development is dynamic. More data on SMEs, on services and It was argued that the European slowdown in growth on outsourcing and relocation are needed. It was is a reflection of an adjustment process towards a pointed out that the next multi-annual statistical new industrial structure. Rapid diffusion of new programme has a key role to play in defining how technology will facilitate the adjustment process to meet increasing data needs, when at the same in the future. Non-technological innovations time, the pressure to cut costs is high. should be considered at least as important as technological ones. In the services sector there The second presentation focussed on the is little room for targeted innovation policies, the competitiveness of Europe from the perspective quality of work force and infrastructure is the of the productivity performance. It addressed key. Reform policies should be comprehensive

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 211 and complementary -many measures are industry Measurement of labour input as a key element specific- and support reallocation of resources in in productivity analysis was discussed. Good the most productive uses. progress is being made within EU on measurement of total hours worked. The increasing number of The role of small enterprises and entrepreneurship flexible working arrangements poses a particular in economic cohesion, growth and competitiveness challenge. The production of harmonised data was brought up in the following discussion. The on total hours worked and on the number of importance of micro-data analysis was seen jobs was found to be most important. The lack increasing in the future. The need for sharing of of harmonisation of enterprise and household micro-data, engaging also private data sources, data on employment was seen as a problem for was raised and better use of existing data stressed. analysis. Many participants highlighted the need for good quality data, some preferred improved quality to The contradiction between an increased need for an increased number of available statistics. statistics and reduction of budgets and pressure to cut administrative burden was mentioned by The role of indicators was discussed. Stand several speakers. For example the source data on alone indicators were not found useful for policy ICT investments, considered an important area for analysis. The current Structural Indicators were development, are missing in many countries, and not considered the most appropriate instrument development is difficult due to lack of funding. for monitoring the Lisbon agenda. A coherent The concern was expressed that policy makers framework, focussing on a few key indicators, may not be aware of the negative effects their which would allow an analysis of the inter- decisions to cut burden and funding cause on the linkages was considered important. availability of monitoring tools.

212 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Measurement of Competitiveness as the basis for policy development

By Heikki Salmi European Commission Enterprise and Industry Directorate-General

Paper for the Conference on Knowledge Economy – Challenges for measurement Eurostat, Luxembourg 8 – 9 December 2005

Abstract 1. Competitiveness and growth - a key policy issue Policy development has to be based on sound factual basis. The objective of the paper is to describe how 1.1. What do we mean by competitiveness? statistics and indicators on competitiveness are used in the Enterprise and Industry Directorate- Competitiveness is a multidimensional concept General in the development of industrial policy and it is measured in many ways. The content of the initiatives. It starts with the description of the concept depends on the respective approach. The Lisbon agenda and the Commission latest policy distinction can be made between competitiveness communications that give substance to it by on a country level, on an industry level and on a means of better regulation, innovations, research firm level. and modern SME policy and industrial policy. Competitiveness of a country is synonymous Special focus is in the brand new industrial with productivity growth. The rationale of this policy Communication called “Implementing approach is that the ultimate objective of an the Community Lisbon Programme: A Policy economy is to increase the standards of living of Framework to Strengthen EU Manufacturing its citizens. The basic mechanism to achieve this – towards a more integrated approach for goal is a steady increase in productivity over time. Industrial Policy”. The Communication is based Increase in productivity, the capacity to produce on in-depth screening of facts, opportunities and more goods and services per person employed, or challenges of 27 sectors of EU manufacturing and more appropriately by hour worked, puts more construction. On the basis of this, the industrial goods and services at the disposal of citizens to policy actions over the coming years have been meet their needs. This is the mechanism that has defined. The paper shows the link between policy underpinned the fundamental changes in modern development and statistics and indicators. It also economies since the industrial revolution: a indicates the increasing need of statistics by dramatic increase in the standard of living and the industry sector and also the problems the users of process of growth, characterized by a fundamental statistics are facing when analysing and assessing structural change. the competitiveness of European industry.

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 213 Industry level competitiveness is clearly a between industrial sectors, between member narrower issue. It indicates the performance of states but more than ever we compare Europe a given industry (e.g. chemical) relative to the to Non-European countries, especially to same industry (chemical). In a nutshell, one those that are competitors of Europe or are sector could be characterized as competitive emerging economies: USA, Japan, China, India, on the basis of its capacity to grow, to innovate Russia, etc. This creates a special demand for and produce more and better (higher quality) the international comparability of statistics. goods and services, and to gain market shares in Comparability within European Statistical international and domestic markets. This makes System is not enough. sectoral competitiveness a multidimensional concept, more elusive and difficult to measure and summarize. A widely used approach to characterise 1.2. Lisbon agenda paves the way to the sectoral competitiveness is to distinguish between competitiveness and the well-being of Europe price (and/or cost) competitiveness, and non-price competitiveness. The former, closely related to Five years ago, Heads of State or Government the production process, is determined by labour committed themselves to making the European cost and productivity, while the latter is the result Union the most dynamic and competitive of a combination of various factors (product knowledge-based economy in the world, capable quality, innovation, marketing, producer-client of sustainable economic growth with more relationships, market structure, etc.). and better jobs and greater social cohesion and respect for the environment. Today we know that Most of the characteristics of sectoral the progress has not been as targeted. A specific competitiveness presented above apply to firm- system was developed to monitor and follow-up level competitiveness. Yet three distinctive Lisbon objectives, so called structural indicators features are worth emphasizing: 1) Typical firm- based on close co-operation between Council, level factors of competitiveness can be identified policy Directorate-Generals of the Commission, (e.g. managerial styles, corporate culture, Eurostat and the National Statistical Institutes organization of the firm), which play a significant (NSI) of Europe. role in the determination of the competitiveness of the firm; 2) A non-competitive firm can go A new start for the Lisbon strategy was launched out of business; 3) Firm-level competitiveness in February this year “Growth and jobs - working is assessed in relation to other firms, within the together for Europe’s Future”. The title of the same sector and country, or to firms in other strategy already indicates that the focus is now sectors and countries. very clearly on growth and jobs. More specifically the policy objectives are the following: All these dimensions of competitiveness mean the need to measure competitiveness both on a Europe a more attractive place to invest and macro and a micro economy level. The policy work objectives of the present Commission, where the • Extend and deepen the internal market Lisbon strategy and the competitiveness of the • Improve European and national European enterprises have a key role, analysis by regulation industrial sector is more and more important. • Ensure open and competitive markets inside and outside Europe One specific element in the competitiveness • Expand and improve European analysis is that we often compare. We compare infrastructure

214 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Knowledge and innovation for growth • Implementing the Community Lisbon • Increase and improve investment in Programme: A Strategy for the research and development simplification of the regulatory • Facilitate innovation, the uptake of ICT environment and the sustainable use of resources • Outcome of the screening of legislative • Contribute to a strong European industrial proposals pending before the Legislator base • Implementing the Community Lisbon Programme: A policy framework to Creating more and better jobs strengthen EU manufacturing – towards • Attract more people into employment and a more integrated approach for industrial modernise social protection systems policy • Improve the adaptability of workers and • Implementing the Community Lisbon enterprises and the flexibility of labour Programme: Modern SME policy for markets growth and employment • Invest more on human capital through • Implementing the Community Lisbon better education and skills Programme: More Research and Innovation – Investing for Growth and All this has to be done through renewed Employment: A Common Approach partnership between the Member States and the Union. Member States have to present their DG ENTR has either full or shared responsibility National Reform Plans in order to improve the on these policy issues. commitments to the Lisbon strategies. Sound macroeconomic conditions are the starting point 2. Competitiveness analysis is essential for for success, enhanced competitiveness can be achieved through productivity growth. The big policy development challenges are globalisation, demographic change and the fast changing technologies. 2.1. The role of DG ENTR

Again, statistics are needed to measure the The mission the Enterprise and Industry progress. Structural indicators keep their Directorate-General is to ensure that EU importance but also lots of other statistics are policies contribute to the competitiveness of EU needed. Those needs are closely related to the enterprises and EU policies facilitate job creation theme of this conference, knowledge economy. and economic growth. Lisbon is about the structural change towards the knowledge society, about the very architecture Particular attention is given to the needs of of our economy for the emerging knowledge manufacturing industry and small and medium society. sized enterprises. The Enterprise and Industry Directorate General works towards the Lisbon The Commission has launched this year several objectives and is a central contributor in the policy initiatives in order to give concrete Lisbon process. It’s main instruments are substance to the Lisbon strategies. The following • Economic analysis, macro and especially initiatives aim to boost the competitiveness and micro economic approach growth of Europe by means of better regulation, • Internal market legislation innovations, research and modern SME policy • Budget to support specific actions and industrial policy: • Open method of co-ordination to

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 215 scrutinize the enterprise policies of the An important cornerstone in the sectoral Member States analysis has been the study and publication of Mary O´Mahony and Bart van Ark in 2003 The responsible commissioner on competitiveness “ EU productivity and competitiveness: An is Mr Verheugen, a Vice-President of the industry perspective. Can Europe resume the Commission. His main policy objectives during catching-up process?” that was commissioned the Commission’s mandate: To contribute the by and prepared for the Enterprise and Industry Lisbon strategies with modern industrial policy, Directorate General. The study on 56 industries better regulation, innovations and new SME demonstrates clearly that there is a wide variation policy. in productivity performance across industries, EU countries and time periods. The study and Policy development, better regulation and the database (Groningen University) behind is decision making need sound factual basis. That an essential source for the analytical work of DG is why the Commission has to have proficient ENTR. analytical capacities. Analysis must be based on good quality statistics, policy indicators and The Enterprise and Industry Directorate research. The development of relevant indicators General has started to publish regularly sectoral is also a task of DG ENTR. competitiveness indicators for policy purposes. “Pocketbook of EU Sectoral Competitiveness 2.2. Reporting on competitiveness Indicators” provides insight into the performance of each industry. The publication is based on a The Enterprise and Industry Directorate General wide variety of indicators using many sources reports regularly on the European performance from Eurostat, OECD, UN and the data base in competitiveness and innovation. Annual of Groningen University. It describes main competitiveness reports focus on specific themes. characteristics of the manufacturing sectors, In the latest publication the themes were China, productivity, external trade and industrial the automotive sector and the role of government interrelations based on input-output tables. policies in influencing competitiveness. DG ENTR is also publishing results from benchmarking DG ENTR is also analysing in-depth the and monitoring and co-ordinating exercises, competitiveness of key industrial sectors. Some comparing the Member States’ innovations and of them will be published or made available on enterprise policies. the web.

The aim of the modern industrial policy is to Policy development in DG ENTR needs several provide right framework conditions for enterprise kinds of statistics. Just to mention some of the development and innovation in order to make the most important ones: Key macro-economic EU an attractive place for industrial investment statistics, short term indicators, structural and job creation. Policy initiatives are essentially business statistics by industrial sector, structural horizontal. However, to ensure that industrial indicators, demography of enterprises, statistics policy is effective, it has to take into account on innovations and R&D, statistics on SMEs, the concrete characteristics of various industrial some environment statistics and statistics on sectors and the particular opportunities and productivity and employment. Sector specific challenges they are facing. This has brought into statistics are also important such as statistics on the picture the need to analyse more in-depth the ICT, biotechnology, mechanical engineering, competitiveness of industrial sectors. textiles etc.

216 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 2.3. More focus on sectoral performance in the competitiveness, energy and environmental economy policies, (5) ensuring full and fair participation in global markets and (6) facilitating social and As mentioned earlier a need for competitiveness economic cohesion (employment and regional analysis by industrial sector has become of crucial dimension). All these are important for sectoral importance as the basis for more focused policy productivity and international competitiveness. actions. The starting point for the analysis was a set of The latest industrial policy communication key economic indicators for all the sectors. “Towards a more integrated approach for industrial policy” was based on an in-depth and There is no unique way to define a common core systematic screening of facts, opportunities and set of indicators that demonstrates the economics challenges for 27 sectors of EU manufacturing of sectors from the competitiveness point of view. and construction. The analysis was both Based on the experiences of O´Mahony and van quantitative and qualitative. On the basis of this, Ark studies, the sectoral indicators published the industrial policy actions over the coming by DG ENTR and the availability of statistics years have been defined. It was a good exercise the following set of indicators were chosen to where statistics had a key role to play and which describe the economics of the sectors: also showed gaps and problems in the available • Value added and employment shares, EU- statistics. 15 and USA, 2001 • Value added distribution by enterprise What was a novelty in the screening was a common size EU-25 framework (template) to compile and present • Value added at constant prises EU 15, information about all the sectors. The screening 1979-2001 was carried out by the DG ENTR staff in close • Employment EU 15, 1979-2001 co-operation with other Commission services • Labour productivity evolution EU 15, per and key stakeholders outside the Commission. hour worked, 1979-2001 • EU-15 Intra-industry trade with partners In the following I will illustrate some by income level, 2002 experiences of DG ENTR when screening the • EU-15 trade in manufacturing products, competitiveness of individual sectors based on a revealed comparative advantage index common framework. (average 2000 – 2002)

2.4. What to measure They aim to signal essential characteristics of the industrial sectors: importance, economy of scale, The basic idea behind the screening is to SME intensity, growth rate, dynamism, social determine to what extent the performance of dimension, competitiveness and comparative those 27 individual sectors could be influenced advantage in trade. by instruments of industrial policy. The Commission will provide a mid-term review The screened opportunities and challenges of the Communication in 2007. The screening of covered the following policy areas (1) ensuring the sectors will be done annually based on the an open and competitive Single Market, above described schema. Data updates and some including competition, (2) providing a supportive improvements are necessary. Key elements of framework for research, innovation and skills, (3) the innovation scoreboard will be involved in the better regulation, (4) ensuring synergies between next screening.

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 217 The screening of opportunities and challenges is They are also important in European industry, based on quantitative and qualitative information. contributing almost 80 % of employment in some Statistical needs concentrate on R&D and industrial sectors, such as textiles, construction innovations, skills and education, concentration of and furniture. manufacturing, emissions, energy, employment, foreign trade and geographical location of 3.2. How competitive is Europe? industry. All this by manufacturing sector, EU-25 aggregates but also very often by country. Manufacturing in general For policy development purposes the results of The sectors chosen were in principle based on the screening of 27 sectors have been grouped into 2-digit level NACE classification. However, four broad categories of industry: food and life the policy sectors do no always match with sciences industries; the machinery and systems NACE because they are based on internal market industries; the fashion and design industries; the legislation or economic reality is different than intermediate product industries. what NACE can offer (e.g. biotechnology, defence industry). In those cases sectors are built A big challenge and opportunity for the EU on more detailed NACE levels and their proper manufacturing is the increasing international aggregations. competition as a location for investment, production and R&D spending. New technologies allow fast The screening results have been published in introduction of new products. Enterprises are also the Staff Working Paper of the Commission facing an increased internationalisation of the (SEC(2005)1216). It aims at giving a clear picture world economy driven by improving transport of the current situation of these 27 industrial linkages, falling communication costs, reduced sectors and their respective key challenges. barriers to trade and investment and stronger competition. 3. How competitive is Europe? There is a well-known productivity growth gap between EU and other industrialised economies, 3.1. Characteristics of manufacturing particularly the US. Partly that is due to the structural differences. For instance the typical The manufacturing sector plays a very growth engine, ICT sector, has a lower share in important role in the EU’s economy. According EU manufacturing. Clear European weakness is to Eurostat statistics it provides a fifth of EU also that although some sectors are performing output and employs some 34 million people. very well, such as mechanical engineering, Manufacturing accounts for three-quarters of chemicals and motor vehicles, EU trade is overall EU exports and over 80% of EU private sector still concentrating in sectors with medium-high R&D expenditures. It is also closely inter-linked technologies and low to intermediate labour with the service industries, providing demand for skills. business services and supplying the key inputs to the service industries. International competition for R&D spending is bigger than ever. EU is not competitive enough It is typical for Europe that small and medium- as a location for research. The US and Japan are size enterprises make up a large part of Europe’s leading the competition but China and India are economy. There are in total some 23 million becoming important location for R&D investments SMEs in the EU, providing around 75 million as well. On top the US is attracting more than EU jobs and accounting for 99% of all enterprises. researchers and highly skilled staff.

218 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH To the West of Europe there is a higher chemicals and rubber industries. Main challenge productivity, more R&D and innovations. To the relates to energy and environment and structural East Europe is losing production but also more changes. and more R&D and innovations. This is the real challenge for Europe. The table in the annex (Source: the Communication) presents the synthesis of Investments to the knowledge economy are the the screening by a sector. It also indicates the right tools for this competition. Europe can not horizontal and sectoral policy actions based on the compete with low cost. Products and services have screening. Some of the policy actions are closely to be innovative and based on high technology related to knowledge based economy: R&D and and high skilled work. innovations, Intellectual Property Rights, skills, ICT. These are also factors that are leading to Challenges across sectors better competitiveness. Other factors that directly The food and life sciences industries make up 20 % influence the competitiveness are better regulation of EU value added and are characterized by and better access to world markets. medium to high growth rates. Challenges to this group relates to knowledge and better regulation. 4. Is there enough statistical information Key knowledge challenges are R&D, protection to substantiate a sound analysis? of intellectual property rights and the financing of innovation for highly innovative SMEs. To get a complete and right picture detailed and good quality statistics play a key role. If we do not The machines and systems industries account know enough or the statistics give wrong signals for about one third of EU manufacturing value- the chosen policy may also be wrong. The data added and are characterized by medium to high sources are numerous in DG ENTR, starting from growth rates with high rates of R&D spending. Eurostat, OECD, UN and relevant studies and The challenges for these sectors therefore research and ending with the data from industrial mainly relate to innovation, intellectual property organisations, private consultants and individual protection and ensuring the availability of highly enterprises. The objective is to use Eurostat data skilled personnel. whenever it is qualified and available.

The fashion and design industries make up only In the following, some requirements and issues 8% of manufacturing value-added and have related to statistics are presented, based on the experienced low or negative output growth and experiences of the screening. relatively low R&D spending over recent years. Structural adjustments are needed for these Geographical coverage. All the 25 Member industries as well as improving innovation, IPR States should be included in all key statistics. protection and skills. Obtaining better access to heavily protected world markets is also a key Time series: Often long time series are needed policy requirement. to measure structural changes. When reforming statistics time series have to be ensured (NACE). The basic and intermediate industries account for some 40 % of EU manufacturing value- Timeliness: Some data is not fresh enough added. These industries are important sources especially in the domains where development for innovation for other sectors as suppliers of is fast like in information and communication key inputs for the rest of EU industry. Growth technology. It is also often difficult to fulfil the has been medium to low except well performing user needs for both timely and reliable data.

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 219 Classifications: Structural Business Statistics economy in many ways. They are the source by NACE does not always provide sufficient for value added, employment, innovations and information. Often detailed data is needed in order entrepreneurship. We should know more about to adapt the data into policy domains and political their structures, economy, competitiveness, reality. Examples: The economic indicators based and financing. The problem is that more data on the same definitions could not be used in the is needed but on the other side data burden and screening in such policy sectors as biotechnology, administrative cost programmes are leading to defence industry, Pharmaceuticals, Cosmetics decreasing data collection from SMEs. What is and ICT. The future NACE revision will partly the solution? More use of administrative data improve the situation. and more targeted questionnaires for micro enterprises? Other examples: In sectors like textiles, furniture and shipbuilding, the quality the analysis depends Confidential data: In some industrial sectors on the aggregation level of the data. Similar users are facing a confidentiality problem of problems occur in such sectors as mining and aggregated data. That is often related to cases construction. In order to give a better picture where data is needed on a detailed level (textiles) more detailed classifications should be used. or where there are not so many producers in Europe (steel). Sometimes this concerns even Comparability: Statistical legislation will ensure EU-level aggregates. Eurostat is working with progressively the quality and the comparability of the Member States on this issue. This long term the data across the 25 Member States and future issue can not be postponed anymore. Therefore, I Member States. In competitiveness analysis it is hope that the task force of Eurostat and the NSIs, important to compare EU to the main competitors which will be created in the next months, will and emerging economies. That is why comparable deliver results in a very near future. data on the US, Russia, China, Japan and India are urgently needed. Professor Van Ark has done Services: Availability of statistics on services enormous work to “put” wide international data is still inadequate, also when studying the into useful and comparable form for research and competitiveness of manufacturing. In order analysis purposes. Also the OECD tries to bridge to better understand what is going on in the different sources and classifications. manufacturing information on interrelations between manufacturing and services is needed. Productivity is the key indicator in the competitiveness analysis on the macro and Industrial sectors: More sectoral approach in micro economics level. Could Eurostat create general is needed in the data production. This a continuously updated data base or should concerns many statistics that are relevant for Groningen University continue to have it? competitiveness analysis.

We should Outsourcing and relocation: 5. Concluding remarks know more about outsourcing and relocation of industrial activities. This is indispensable for a Data needs are increasing. The EU policies meaningful competitiveness and globalisation on growth and employment, internal market analysis and for the basis of a well founded policy and competitiveness have to be based on facts development. reflected by statistics. All these are closely related to knowledge economy. Everything that improves SME dilemma: Small and medium size the statistics on knowledge economy will also enterprises are significant players in the European contribute to better competitiveness analysis and

220 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH policies. Lisbon will be a hot topic on the agenda References of the European Union still for several years which means close involvement of statistics and European Commission: Working together for statistical authorities to the Lisbon process. growth and jobs. A new start for the Lisbon Strategy - COM(2005) 24 Enterprise and Industry Directorate General is increasing its analytical approach. More data European Commission: Implementing the and especially more sectoral data is needed for Community Lisbon Programme: A Policy the basis of the industrial policy development. Framework to Strengthen EU Manufacturing – The challenge for statisticians and for the users towards a more integrated approach for Industrial is how to increase the production of statistics Policy - COM(2005)474 when at the same time implementation of better regulation and less administrative costs means European Commission. Commission Staff less data collection. Working paper: European Industry: A Sectoral Overview - SEC(2005) 1217 Another challenge for the statisticians: How to sell the EU statistical legislation and standards to European Commission. EU sectoral the rest of the world or at least to the competitors competitiveness indicators (2004). Enterprise of Europe. This could ensure good comparability and Industry publications of the data across the world. Absorbing European regulatory framework outside Europe is already Mary O´Mahony and Bart van Ark (2003), EU happening in some other areas. Productivity and Competitiveness: An Industry Perspective. Can Europe resume the catching-up All in all good co-operation and dialogue is process. Enterprise publications. needed between the users, the producers of statistics including Eurostat and international European Competitiveness Reports from years 1999 organisations. to 2004. Commission Staff Working Papers.

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 221 4) HLG (2007) HLG HLG Defence HLG CARS 21 HLG 21 CARS Pharmaceutical Forum Pharmaceutical 2015 LeaderSHIP HLG SECTORAL ACTIONS ACTIONS SECTORAL TaskforceCompetitiveness on ICT EuropeanSpace programme/GMES Dialogue for mechanical engineering mechanical Dialoguefor Mid-Term Review of Strategyofsciences Review "Life Mid-Term and bitoechnology" surveillance Market Access land to logistics costs, feedstock and Energy 8) 9) 10) 5) 8) 9) 6) 7) X X X X X X X SECTOR SECTOR SPECIFITIES X X X X X X X X X X CHANGE STRUCTURAL STRUCTURAL Structural Change Structural X X X X X issues GMO (EUP)Products Using Energy instrument Financial Regulatory 5) 6) 7) X X X X X X Trade dumping distortions subsidies/ TRADE X X X X X X raw raw materials Accessto X X X X X X X X X X X X X X markets Accessto External Aspects of Competitiveness and and Competitiveness Aspectsof External Access Market X X X X X X X Use Energy Intensive 2) X X X X X X X X Air X X X X X X X X Water environment X X X X X X X X X X X X X X X Waste ENVIRONMENT & ENERGY ENERGY & ENVIRONMENT 1) X X X X X X X X X HLG on Competitiveness, energy, and and energy, Competitiveness, on HLG Climate Climate change X X X Techn. Standards X X X X X X and ICT: challenges are sector specific; ICT uptake is a general a general is ICT uptake are sector specific; challenges ICT: Safety Health Many sectors will also be affected by the new legislation framework framework be affectednew the legislation also by sectors will Many Commission the of members involving actions and/or actions legislative Includes 2) (REACH) chemicals for 3) industry. the for challenge 4) Hence the absence of a cross does not therefore not necessarilyHence a cross does the challen denote that the absence of 10)

10) X X X X X X X X X Market Internal BETTER REGULATION BETTER X X X X X Admin. Admin. burden/ regulation of sectoralof Complexity Complexity New Legislative Simplification Simplification Legislative New Programme X X X X X finance for SMEs for Accessto Skills X X X X X X X X X X X X X X X Skills Improving Sectoral Improving X X X X X X X X X X X X X KNOWLEDGE IPR IPR, IPR, feiting Counter- X X X X X X X X X X X X X X X X X X X X X X X vation R&D/Inno R&I Industrial System Monitoring The table indicates with crosses, the cases in which a policy challenge is considered of the highest priority for each sector amongst the many relevant policy challenges. tablea policy indicatescrosses, the cases in which the many with challenge priorityThe the highest each sector is considered for of amongst Industry 3)

Food, drink & tobacco Food,& drink Cosmetics Pharmaceuticals Biotech Medicaldevices ICT Mechanicalengineering engineering Electrical Motorvehicles Aerospace Defenceindustries Shipbuilding Textiles Leatherand leather goods Footwear Furniture Non-energyindustries extractive Non-ferrousmetals and Cement lime Ceramics Glass Woodproducts & woodof paper Pulp, paper & products publishing & Printing Steel rubber, and Chemicals, plastics Construction the under sectors fall all activities energy to belonging installations to regard in With 1) indicated threshold capacity the above is question in installation the provided ETS EC. in sectorsThe marked 2003(87) Directive Trading Emisisons I the of Annex Source: COM(2005)474

Industries

Science Industries Science

Machine and System Industries System and Machine Design Industries Goods Intermediate and Basic

Food and Life and Food Fashion and and Fashion

222 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Europe’s Productivity Gap: Catching Up or Getting Stuck?

by Bart van Ark1 Groningen Growth and Development Centre University of Groningen and The Conference Board Europe

Abstract as a monitoring tool of the Lisbon Agenda and promotes the use of Growth and Productivity This presentation will focus on competitiveness Accounts (such as the forthcoming EU KLEMS of Europe from the perspective of the productivity data) as a useful complementary instrument. performance of the European Union. It is argued that productivity and employment creation are 1. Introduction the key drivers of growth and competitiveness. The presentation will first look at the comparative During the second half of the 1990s the productivity performance of EU member states comparative growth performance of Europe vis- (and the aggregate EU) vis-à-vis the United States. à-vis the United States has undergone a marked In particular it will provide a decomposition of change. For the first time since World War II labour productivity growth into the contributions labour productivity growth in most countries that of capital deepening and total factor productivity are now part of the European Union (EU) fell growth, with special attention to the contribution behind the U.S. for a considerable length of time. of Information and Communication Technology Until the beginning of the 1970s rapid labour (ICT). The talk then focuses on the comparative productivity growth in the EU went together with productivity performance from the perspective a catching-up in terms of GDP per capita levels of individual industries. The discussion of on the U.S.. A first break in this pattern occurred manufacturing deals with the comparative in the mid 1970s. While catching-up in terms of productivity and unit labour cost performance of labour productivity continued, the gap in GDP per ‘old’ EU member states relative to the new member capita levels between the EU and the U.S. did not states, China, India and Mexico. The analysis of narrow any further after 1975 (see Figure 1). This services shows that the productivity slowdown in differential performance reflects the slowdown in Europe (although across the board) is particularly the growth of labour input in Europe, which was strong in service industries that make intensive related to increased unemployment, a decline in use of ICT, whereas the acceleration in the U.S. the labour force participation rates and a fall in is particularly strong in these industries. Some average working hours. The second break, which suggestions will be provided on how Europe can is the focus of this paper, occurred in the mid regain its growth path. Finally the speaker will also 1990s when the catching up in terms of labour comment on the features of Structural Indicators productivity also came to a standstill once the

1 This paper was also presented at the International Symposium on “Productivity, Competitiveness and Globalisation” at the Banque de France, Paris on 4 November 2005. The paper is largely based on earlier work, including Van Ark (2005) , O’Mahony and Van Ark (2003), McGuckin and Van Ark (2005a), Timmer and Van Ark (2005) and Van Ark and Inklaar (2005).

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 223 average EU level reached the U.S. level. In fact their most productive uses. This is partly realised a new productivity gap opened up since 1995. through the labour market, as the substitution of Whereas average annual labour productivity low-skilled for high-skilled labour has proceeded growth in the US accelerated from 1.1% during more smoothly and the restructuring of the the period 1987-1995 to 2.5% during 1995-2004, economy was not hindered. It has also been realised EU productivity growth declined from 2.1% to through product markets, in particular through 1.4%.2 the creation of new opportunities for productive applications of ICT mainly in service industries The urgency of the ‘European’ problem is and service-related activities in manufacturing. underlined by the rapid improvements in economic Finally, the combination of reforms and adoption performance of countries in Central and Eastern of new technologies has supported creativity of Europe and Asia. Average labour productivity firms and entrepreneurs to develop new products in the new EU member states increased at 4.2% and services and to reshape the organisational and from 1995-2004. In China and India, GDP per production processes by which these are brought person employed (i.e. not corrected for changes in to the market. working hours) was 3.9% and 6.1% respectively from 1995-2004. Unfortunately there is much less consensus on The striking acceleration in U.S. output and the causes of the slowdown in Europe. Indeed productivity growth in the mid 1990s has been the reasons for the limited impact of technology, much discussed in the literature. A consensus innovation and structural reforms on economic has emerged that faster growth can at least in growth in Europe are still poorly understood. The part be traced to the effects of the information urgency to better grasp the causes of the problems and communication technology (ICT) revolution is underlined in the recent review by the Kok (Oliner and Sichel 2000, 2002; Jorgenson and Commission of the Lisbon agenda for reform Stiroh 2000; Jorgenson, Ho and Stiroh, 2003), in Europe, which aims to improve Europe’s which in turn has depended on a surge in ICT competitiveness (European Commission, 2004). investment, strong productivity effects from ICT- Indeed, the Kok report strongly argues for a producing industries and a more productive use revival of productivity growth in Europe, in of ICT in the rest of the economy. In addition the particular in the light of demographic trends U.S. economy has also benefited from a greater towards a smaller labour force relative to the total flexibility of markets in allocating resources to population in Europe.

2 Business cycles in the U.S. and the EU are not completely synchronised. However, the divergent trend growth rates are clear.

224 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Figure 1. GDP, GDP per capita and GDP per hour, 1955-2003

140%

GDP

120%

100%

GDP per hour

80% EU-15 as % of the U.S.

GDP per capita

60%

40% 1960 1965 1970 1975 1980 1985 1990 1995 2000

Note: EU refers to 15 EU membership as before 1 May 2004. Source: Groningen Growth and Development Centre (GGDC) & The Conference Board (TCB).

At the same time, however, there is also industry structure. In addition it needs to be considerable diversity in terms of both considered whether these features are common to productivity growth as well as comparative all or just a subset of EU countries. levels between European countries. Comparative growth rates of labour productivity between This paper argues that the European slowdown 1995 and 2004 differ between –0 per cent (for in growth is a reflection of an adjustment process Spain) and 4.7 per cent (for Ireland). And there towards a new industrial structure, which has is a variation of plus 28 percentage points (for developed more slowly in the EU than in the U.S.. Belgium) and minus 49 per cent (for Portugal) Rapid diffusion of new technology will facilitate in terms of each country’s productivity level the adjustment process in the future. However, an relative to the US in 2004. Hence although there institutional environment that slows down change are also some common traces to the European may hold up the structural adjustment process in growth problem, one cannot simply treat the Europe and inhibit the reallocation of resources European area as homogeneous. to their most productive uses. The European economic environment creates too little room The cross-regional diversity in productivity for good firms to excel and for failing firms to performance cannot be fully understood without exit the market so as to free up resources for the adopting an industry perspective to output, much-needed transition. input and productivity performance. Thus there is a need to go beneath the aggregate numbers This paper begins with a brief review of the to ascertain to what extent variations across aggregate estimates of productivity and per countries are largely explained by differences capita income in order to identify the extent to

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 225 which labour market developments rather than member states is higher but also varies much productivity has impacted the comparative between -0.4% (Malta) and 11/5% (Lithuania) performance of the EU relative to other regions and from 1995-2004. countries (Section 2). I then proceed to examine the comparative productivity performance of On average EU labour productivity growth is not the EU and the U.S. from the perspective of the only slower than in the U.S., but also compared contributions of the main growth drivers, which to Japan and other OECD countries (not shown are ICT and ‘other’ capital deepening and total in the table). In terms of GDP per capita growth, factor productivity (TFP). (Section 3). Next I the differences are not as big. Between 1995 and approach the differential labour productivity 2004 EU-25 per capita income growth was only growth performance from a sectoral perspective slightly lower than in the U.S. and substantially (Section 4). I first look at the comparative growth higher than in Japan. Compared to China and performance of the EU manufacturing sector in India, all countries except the Baltic states fall global perspective. Then I discuss the key role short. But it should be stressed that the absolute for market services in understanding Europe’s income levels in these two Asian countries are underperformance relative to the United States. In substantially below those of the advanced nations, the final section, I focus on the question whether suggesting a large ‘catch-up’ bonus which is still the European Union should change or intensify to be realised (see Table 2). its strategies to revive productivity growth (Section 5). I argue that policy mechanisms, such as GDP per capita growth is driven by an increased macroeconomic management, existing innovation input of labour and/or labour productivity and reform policies and some horizontal policy growth. Indeed one can simply show that the measures (in particular education policies) should difference in the growth rates of average per be reconsidered for their effects on the allocation capita income and labour productivity can be of resources and their effects on productivity at accounted for by changes in a range of labour industry and aggregate levels of the economy. market and population indicators (see van Ark and McGuckin, 1999; McGuckin and van Ark, 2. Comparative Productivity and Labour 2005a). First, the growth in income per head of the population (∆O/P) is a function of the change in Market Performance labour productivity (∆O/H) and labour intensity, Table 1 shows the growth rates of per capita expressed as the number of working hours per income (measured as GDP per capita) and head on the population (∆H/P): labour productivity (measured as GDP per hour worked) for major regions in the world economy ∆O/P = ∆O/H * ∆H/P (1) with a breakdown to individual European countries. The table shows a large variation in The change in working hours per person can be per capita income and productivity growth rates decomposed into the change in hours worked per in European countries. Within the ‘old’ EU-15, person employed (H/E) and the change in the share the variation of productivity growth is between of employment in the total population (E/P):3 0% (for Spain) and 4.7% (for Ireland) between 1995 and 2004. Productivity growth in the new ∆H/P = ∆H/E * ∆E/P (2)

3 The change in the employment/population ratio (E/P) can be further broken down into the number of persons employed relative to the total labour force (i.e., employed persons plus registered unemployed persons) (E/L), the ratio of the labour force to all persons aged 15 to 64 (i.e., the working age population) (L/P1564) and the share of the working age population in the total population (P1564/P): ∆E/P = ∆E/L * ∆L/P1564 * ∆P1564/P (see see van Ark and McGuckin, 1999; McGuckin and van Ark, 2005a)

226 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Table 2 shows the breakdown of per capita income It is clear from Table 2 that the comparative into labour market indicators and productivity from levels of labour productivity in the ‘old’ EU- the perspective of comparative levels of European 15 countries were substantially higher relative countries relative to the United States for 2004. The to the United States than the relative per estimates are converted on the basis of purchasing capita income levels. This is mainly due to the power parities, which take account of differences substantially lower number of working hours in relative price levels across countries. In addition per employed person and, in addition, to a lower to Europe, comparative estimates are also shown ratio of employed persons relative to the total for Japan, Mexico, China and India. population.

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 227 Table 1. Growth Rates of Per Capita Income and Labor Productivity Growth, 1987-2004 Table 1: Growth Rates of Per Capita Income and Labor Productivity Growth, 1987-2004 GDP per capita GDP per hour worked 1987-1995 1995-2004 of which 1987-1995 1995-2004 of which 2000-2004 2000-2004

EU-15a 1.8 2.0 1.3 2.3 1.4 1.1 Austria 2.1 2.0 1.0 2.3 2.4 1.3 Belgium 2.1 1.9 1.2 2.3 1.6 1.3 Denmark 1.3 1.7 1.1 2.1 1.7 1.9 Finland 0.3 3.4 2.1 2.8 2.5 2.2 France 1.5 1.8 1.2 1.9 1.8 1.9 Germany 1.8 1.2 0.5 3.1 1.9 1.3 Greece 1.2 3.6 4.0 0.8 2.7 2.8 Ireland 5.1 6.6 4.0 4.0 4.7 3.5 Italy 1.8 1.3 0.8 2.0 0.4 -0.2 Luxembourg 3.9 3.7 1.5 2.6 2.0 1.2 Netherlands 2.0 1.7 0.0 1.6 0.4 0.4 Portugal 3.1 2.0 0.0 2.8 1.4 0.3 Spain 2.5 3.2 2.6 2.1 0.0 0.2 Sweden 0.6 2.5 1.8 1.4 2.4 2.4 U.K. 1.7 2.5 2.0 2.1 2.0 2.0

EU-10, newb -- 3.9 3.6 -- 4.2 4.5 Cyprus -- 2.8 2.4 -- 2.0 1.4 Czech Republic -- 2.3 3.2 -- 3.2 4.4 Estonia -- 6.6 7.0 -- 7.1 6.6 Hungary -- 4.1 3.9 -- 2.7 3.2 Latvia -- 7.3 8.4 -- 6.1 7.3 Lithuania -- 5.9 7.7 -- 7.6 11.5 Malta -- 2.3 -0.4 -- 2.1 -0.4 Poland -- 4.1 2.9 -- 4.8 4.3 Slovakia -- 4.0 4.5 -- 4.2 4.6 Slovenia -- 3.8 3.3 -- 3.1 2.8

EU-25, enlargedc -- 2.1 1.5 -- 1.8 1.6 United States 1.5 2.3 1.6 1.1 2.5 2.9 Japan 2.6 1.0 0.9 2.8 2.1 1.9 Mexicod 0.4 2.2 -0.5 0.6 0.3 0.9 Indiad 3.9 4.5 5.2 3.7 3.9 3.1 d China 5.7 6.6 7.7 4.7 6.1 6.8 a) referring to membership of the European Union until 30 April 2004; b) referring to new membership of the European Union as of 1 May 2004; c) referring to all members of the European Union as of 1 May 2004 (see Table 2): d) productivity in China is in terms of GDP per person employed Source: TCB/GGDC Total Economy Database (www.ggdc.net/dseries), based on OECD National Accounts and Labour Force Statistics

228 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Table 2. Labour Productivity and Income. Differences in Ranking, 2004 Table 2: Labour Productivity and Income: Differences in Ranking, 2004 Productivity (per Effect of Productivity (per Effect of Per Capita Income hour) Working worker) Employment/ US$ %US Hours US$ %US Pop. Ratio US$

EU-15a 40.51 91% -13% 63311 78% -6% 27666 72% Luxembourg 56.84 128% -24% 83959 104% 37% 53993 141% France 50.08 113% -24% 72065 89% -14% 28956 76% Belgium 48.12 109% -13% 76890 95% -18% 29826 78% Ireland 46.26 104% -10% 76274 95% -3% 35021 91% Netherlands 44.48 100% -26% 60278 75% 3% 29766 78% Austria 43.81 99% -17% 65646 81% -2% 30466 79% Germany 43.22 97% -20% 62349 77% -7% 27076 71% Denmark 41.65 94% -17% 62364 77% 3% 30746 80% Finland 39.60 89% -8% 65414 81% -4% 29545 77% U.K. 39.28 89% -10% 63676 79% -1% 29935 78% Italy 39.27 89% -11% 62930 78% -8% 26714 70% Sweden 39.24 88% -12% 61789 77% 0% 29517 77% Spain 32.59 73% -1% 58583 73% -8% 24763 65% Greece 28.14 63% 3% 53978 67% -11% 21326 56% Portugal 22.53 51% -3% 38715 48% 1% 18909 49%

EU-10, newb 18.18 41% 3% 35729 44% -8% 13817 36% Malta 26.76 60% 4% 52124 65% -17% 18105 47% Slovenia 25.65 58% 5% 50812 63% -9% 20592 54% Cyprus 22.72 51% 8% 47836 59% -8% 19814 52% Hungary 22.46 51% 0% 40563 50% -10% 15589 41% Czech Republic 20.55 46% 3% 39430 49% -2% 18027 47% Slovakia 17.62 40% 3% 34508 43% -7% 13805 36% Poland 17.16 39% 3% 34029 42% -10% 12169 32% Lithuania 13.57 31% 6% 29402 36% -6% 11779 31% Estonia 13.12 30% 4% 26895 33% -3% 11521 30% Latvia 10.99 25% 4% 23593 29% 0% 11172 29%

EU-25, enlargedc 36.51 82% -9% 59236 73% -7% 25397 66%

United States 44.34 100% 0% 80660 100% 0% 38345 100% Japan 32.74 74% -3% 57263 71% 3% 28460 74% Mexico 13.46 30% 5% 28400 35% -10% 9598 25% Indiad (2003) 9% -2% 7% d China (2003) 14% 2% 15% a) referring to membership of the European Union until 30 April 2004 (see Table 1) b) referring to new membership of the European Union as of 1 May 2004 (see Table 1) c) referring to all members of the European Union as of 1 May 2004 (see Table 1) d) no productivity (per hour) available. Output converted to US$ at 1990 GK PPPs. Figures refer 2003 Source: TCB/GGDC Total Economy Database (www.ggdc.net/dseries), based on OECD National Accounts and Labour Force Statistics, with GDP converted to US$ at 2002 EKS PPPs.

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 229 The relative high levels of labour productivity in In reality, slow productivity growth due to Europe have been pointed at by various scholars increased participation is primarily a short term as an indication of a “European model” that deals phenomenon. For example, in an extensive differently with the trade-off between labour empirical study for almost all OECD countries, intensity and productivity than the U.S. model. McGuckin and van Ark (2005b) find that the According to, for example, Blanchard (2004) negative productivity response elasticity to a 1% and Gordon (2004) the European preference for rise in participation is less than 0.3 and peters out more leisure would be offset against a lower level in less than five years. A too strong focus on the of per capita income. Moreover, Gordon argues ‘trade-off’ issue can easily lead to the mistaken that a significant portion of higher American view that this is a predetermined reality for GDP per capita is required to create decent living Europe in the coming decades. The main source conditions in a much harsher natural environment of productivity differentials between countries (requiring a greater use of energy for heating in longer run, however, is not due to a lack in and air-conditioning), to fight crime and to terms of work effort but primarily because of travel longer distances across huge metropolitan an underperformance of capital and technology, areas. Prescott (2004) argues that tax systems which is the focus of the next Section. explain most of the differences in labour supply between Europe and the United States making 3. The Differential in Sources of Growth work more costly relative to leisure. Alesina et al. (2005) explain Europe’s preference for leisure between Europe and the U.S. through the effect of worksharing agreements in Labour productivity growth can be decomposed declining industries, which have not created more into the contributions of capital and technology employment but have increased returns to longer using a growth accounting framework (Solow vacations leisure through a social multiplier 1957, Jorgenson 1995). Although such effect. decompositions are only possible on the basis of certain assumptions, cost-minimizing producers, While there may be some truth in all these competitive factor markets, well-measured inputs arguments, one should be cautious not to speak and output, and constant returns to scale (which too easily of one ‘European model’ for the labour are unlikely to be fully satisfied), it provides a market. Firstly, Table 2 shows large differences simple and consistent method which can used as in the effects of working hours and participation a starting point to identify the contributions of the on per capita income. For example, participation source of growth. effects are much more negative in Belgium, France and Greece than in Denmark, the Netherlands In the decomposition below, we focus and Sweden. In contrast, average working hours explicitly on the contributions of Information are much higher in the Southern European states and Communication Technology (ICT) to than, for example, in France, Luxembourg and productivity. As a General Purpose Technology, the Netherlands. Secondly, Sapir (2005) clearly ICT may be expected to have a long-lasting effect indicates that Europe may be characterized by at on productivity growth, and it may therefore be least four social models, distinguished by region a possible source of productivity differentials (Nordic, Anglo-Saxon, Continental European and between countries in the longer run. The Mediterranean). Thirdly, and in line with Table 2 contribution of ICT to productivity can be traced and Sapir’s observations, European countries through three transmission channels, namely have made progress to different degrees in terms through investment in ICT, the production of ICT, of raising labour participation substantially and possible “spillovers” from the use of ICT. In during the 1990s. a neo-classical framework, the contribution from

230 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH ICT investment is well defined: firms will invest where Δ refers to first differences andv ’s denote in ICT up to the point where further output gains the two-period average shares in total f a c t o r are equal to the marginal cost of the investment. income and because of constant returns to scale:

This way the contribution from growth in ICT vL + vN + vICT = 1. By rearranging equation (4) capital per hour worked to labour productivity average labour productivity growth, defined as y growth can be identified. Total factor productivity = Y/L, can be decomposed into the ratio of capital (TFP) growth in ICT producing industries will services to hours worked, k = K/L, and TFP quite naturally contribute to aggregate TFP growth growth. Another useful distinction can be made and hence labour productivity growth. The final between TFP growth originating in manufacturing channel, which is TFP growth due to ICT use, is industries producing ICT goods (Aprod) and that the hardest to identify separately and it also raises from other industries, ‘other’ TFP (Aother) some conceptual issues. The underlying idea (5) of spillovers from ICT, is that ICT enables new organizational models and other innovations in Δ ln y = vN Δln k N + vICT Δln k ICT + Δln Aprod + Δ ln Aother the production process, as well as the production of new goods and services. So although new ICT The estimates on the comparative growth investment goods are standard products, they make performance of the EU-15 and the U.S. presented it possible for firms to innovate and accumulate here are an update to 2001-2004 from earlier firm-specific capital (see e.g. Brynjolfsson and work by Timmer and van Ark (2005). Data Hitt, 2000 and OECD, 2004). on investment, GDP and labour compensation are typically derived from national accounts. Insofar as these innovations yield additional However, substantial additional work was output gains, they may show up as additional required to construct separate investment time total factor productivity growth in ICT using series for three ICT assets (office and computing industries and may be labelled as “spillovers”. equipment, communication equipment, and software) as well as three non-ICT assets Gross domestic product (Y) is produced from (non-ICT equipment, transport equipment and aggregate factor inputs X, consisting of ICT capital non-residential structures). The resulting real investment series are used to derive capital services (KICT), non-ICT capital services (KN) and labour services (L). Total factor productivity (A) service growth rates which, in combination is represented as a Hicks-neutral augmentation of with growth rates on total hours by employees (mainly obtained from labour force surveys), the aggregate inputs. The aggregate production give the growth of capital services per hour function has the following format: worked. The contribution of each capital asset (3) to growth was estimated using the share of capital compensation of each asset in total GDP Y = AX (L, K N , K ICT ) as weights. Aggregate total factor productivity growth (TFP) was derived as a residual from Under the assumption of cost-minimizing labour productivity growth minus the contribution producers, competitive factor markets and of capital deepening to GDP growth. To obtain constant returns to scale, total factor productivity separate TFP estimates for ICT-producing growth is derived as the growth of output minus a industries from ‘other’ TFP, we assume that share weighted growth of inputs: TFP growth rates for each of the three ICT- (4) industries (office, accounting and computing equipment, communication equipment and

Δ ln A = Δ lnY − vL Δln L − vN Δ ln K N − vICT Δln K ICT electronic components manufacturing) in the

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 231 U.S. also apply to the European countries.4 To ICT capital deepening during the late 1990s. measure the ICT industry contributions to total However, this investment boom was mostly factor productivity growth, Domar weights for transitory, with ICT capital deepening returning the individual countries were used.5 to pre-1995 levels after 2000 in both the EU- 15 and the U.S. However, since 2000 U.S. Table 3 presents the results for the EU-15 and labour productivity accelerated further, while the U.S. for the periods 1987-1995, 1995-2000 the EU-15 suffered additional slowdown. This and (the updated period) 2000-2004. The tables divergence between the Europe and America shows a decomposition of labour productivity mainly relates to TFP growth outside the ICT growth into the effects of ICT capital deepening producing sector. In Europe, TFP growth in and TFP growth from ICT-producing industries, outside ICT-production was effectively zero and two other sources of growth, namely non- after 2000, while in the U.S. ‘other’ TFP growth added almost 1.5 percentage points to labour ICT capital deepening and TFP growth other productivity growth.6 than that from ICT production. Our main findings are that the EU-15 as a whole has been On the basis of this evidence, it may be lagging behind the U.S. in terms of ICT capital hypothesized that the faster growth and deepening throughout all periods. Both the EU- acceleration in ‘other’ TFP in the United States 15 and the U.S. show a strong acceleration of may be due to a greater degree of spillovers

TableTable 3. 3: Sources Sources ofof labour labour productivity productivity growth growth in the in EU-15 the E U-15and U.S., and 1987-2004 U.S., 1987-2004 1987-1995 1995-2000 2000-2004* European Union-15 Aggregate Labour Productivity Growth 2.3 1.8 1.1 of which: ICT capital deepening 0.4 0.6 0.3 Non-ICT capital deepening 0.8 0.4 0.5 ICT-production TFP 0.2 0.4 0.2 Other TFP 0.9 0.4 0.0

United States Aggregate Labour Productivity Growth 1.2 2.3 2.8 of which: ICT capital deepening 0.5 1.0 0.6 Non-ICT capital deepening 0.1 0.2 0.5 ICT-production TFP 0.4 0.7 0.3 Other TFP 0.2 0.4 1.4 * 2004 is preliminary estimate based on average share of ICT investment in total investment for 2002 and 2003 Source: Van Ark and Inklaar (2005)

4 Of course one would ideally use capital service measures at the industry level for individual European countries. To date such detailed TFP estimates are only available for the U.S. and a few European countries. We use these more detailed estimates in Section 4. 5 The Domar weight of an industry is defined as the industry’s gross output divided by aggregate value added. In general, these weights sum to more than one. 6 Estimates for individual countries can be obtained from http://www.ggdc.net/dseries/growth-accounting.html. Although there is much variation in TFP not related to ICT, the trend is generally downwards with the exception of Sweden and the United Kingdom.

232 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH created by the use of ICT. However, one has to the on-going globalization of product markets be very cautious in interpreting this evidence. and factor markets, the industry structure is Firstly, there is no strong statistical evidence under continuous pressure from competitive about a positive relationship between ICT forces. It is important to establish how these capital deepening and ‘other’ TFP (Stiroh, 2002; changes have affected the overall performance Van Ark and Inklaar, 2005). Secondly, there are of the economy. Finally, the opportunities for many more differences that affect TFP growth new technological applications may have very differences between countries such as, for different implications for industries. Indeed example, differences in market structure and the absorptive capacity for ICT differs highly flexibility of product, labour and capital markets across industries, and has very different impacts between countries.7 Thirdly, without TFP on output, employment and productivity growth estimates for individual industries there performance. is no good way of identifying such spillovers, For an analysis of productivity growth in as the aggregate TFP residual may include a Europe and the U.S. at industry level, the whole range of unmeasured contributions (or GGDC developed a database which contains detractions) to output growth which are difficult information on value added and employment by to distinguish at aggregate level. Hence the industry (see van Ark et al., 2003a; O’Mahony next Section of this paper focuses on industry and van Ark, 2003). This so-called ‘60-industry estimates of productivity growth. database’ has been updated to the year 2003.8 On the basis of this dataset, measures of labour 4. An Industry Perspective on Productivity productivity growth and the contribution of Growth individual industries to aggregate productivity growth can be calculated. These contributions In this section we look at productivity are calculated using a shift-share approach. Table performance from an industry perspective. 4 summarizes the contributions of three major Although many of the policy issues related to industry groups (ICT-producing industries, other the slowdown of productivity growth in Europe producing industries and other market services) are more of a generic nature rather than industry and a reallocation effect to labour productivity specific, the sector perspective is useful for growth in the market sector of the economy.9 several reasons. Firstly, it is important to pinpoint in which industries or industry groups Table 4 shows that differences in the performance the slowdown occurs and to examine whether of ICT-producing industries (which include ICT- it is confined to a few sectors or whether it is producing manufacturing and services industries) more widespread. Secondly, under the influence explain part of the aggregate productivity growth from both intra-EU economic integration and differential between Europe and the U.S.

7 See, for example, Hall (1988) and Roeger (1995). 8 The updated measures will be released on the GGDC website (www.ggdc.net/dseries/60_Industry.shtml) in November 2005. The main source of this database is the new OECD STAN Database of national accounts, but greater industry detail is provided through the use of industry surveys and censuses. As discussed above, we ultimately would like to have estimates of TFP growth for individual industries, in addition to the aggregate figures presented above. Only then it is possible to see which industries are heavy ICT investors and whether these industries have higher TFP growth. This can help determine whether ICT spillovers are an important source of growth differences between Europe and the United States. At this moment such estimates are only available for four major European countries (France, Germany, the Netherlands, the UK) and the U.S.. See also below (Table 8) and Inklaar et al. (2005) and Van Ark and Inklaar (2005). 9 The ICT producing industries include producers of IT hardware, communication equipment, telecommunications and computer services (including software). The distinction is based on an OECD classification (see OECD 2002).

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 233 Table 4: IndustryTable contributions 4. Industry contributions to market economy to market labour economyproductivity growth, 1987-2003 labour productivity growth, 1987-2003 1987-1995 1995-2000 2000-2003 European Union-15 Market economy labour productivity growth 2.7 2.2 1.1 of which: ICT production* 0.5 0.8 0.5 Production industries** 1.3 0.8 0.6 Market services** 0.8 0.6 0.1 Reallocation 0.2 0.0 -0.1

United States Market economy labour productivity growth 1.4 3.4 3.6 of which: ICT production* 0.8 1.2 1.1 Production industries** 0.3 0.5 0.9 Market services** 0.5 1.8 2.0 Reallocation -0.2 -0.1 -0.3 * Includes ICT manufacturing, telecom and software services ** Excludes ICT producing industries Source: Van Ark and Inklaar (2005)

The larger contribution from ICT production in even more striking since 2000: the contribution the U.S. is primarily due to the greater share of of other market services to labour productivity ICT producing industries in U.S. value added (the growth almost disappeared in the EU-15 whereas ‘between industry’ effect). Although productivity it accelerated further in the U.S. growth rates for ICT producing industries (the ‘within industry’ effect) are roughly the same For Manufacturing, Europe Should Look Towards between the EU-15 and the U.S., 12.6 per cent of the East U.S. value added in the market economy consists of ICT production, including IT hardware, When focusing on manufacturing, however, communication equipment, telecommunications it is not sufficiently informative to focus the and computer services (including software), comparison only on Europe versus the United compared to 5.3 per cent in the EU-15. This equals States. Table 5 looks at comparative productivity 2.7 per cent and 1.5 per cent of value added in the performance in aggregate manufacturing for the manufacturing sector of the U.S. and the EU-15 EU-15, Japan and the United States, and the new respectively EU-10 member states, India and China from 1987-1995 and from 1995-2003. The figures There is no role for other production industries, show a clear dichotomy between the advanced which mainly includes manufacturing (excluding and the emerging economies. The EU-15, Japan ICT production), in explaining the aggregate and the United States show productivity growth growth differential. Instead, most of the labour rates of between 3 and 4 per cent (although the productivity acceleration in the U.S. can be U.S. shows a strong acceleration after 1995) and traced to faster productivity growth in other the new EU-10 countries, India and China which market services. This difference has become all show productivity of between 6 and 8 per

234 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH TableTable 5. 5: Manufacturing Manufacturing value value added added per per hour hour worked,worked, annual annual average average growthgrowth rates Advanced economies EU-15 USA Japan 1987-1995 4.0 2.9 3.9 1995-2003 3.2 4.7 3.8 Emerging economies new EU-10a Chinab Indiab 1987-1995 6.5 5.7 1995-2003 6.5 8.2 6.1 a) Average for Czech Republic, Hungary, Poland and Slovakia; b) per person employed, 1987-1994 and 1994- 2002 Source: TCB/GGDC and OECD STAN database

Table 6. 6: Manufacturing Manufacturing Value Value Added Added per per Hour Hour WorkedWorked and and Unit Unit Labour Labour Cost, Cost, annual annual average average growth ratesgrowth (USA=1.000) rates (USA=1.000) Value added per Unit Labour Cost hour worked (exchange rate (PPP adjusted) adjusted)

Advanced economies EU15 0.788 0.905 Japan 0.661 1.195 USA 1.000 1.000

Emerging economies New EU-10a 0.205 0.724 Chinab 0.043 0.495 Indiab 0.053 a) Average for Czech Republic, Hungary, Poland and Slovakia; b) productivity (per person employed) Source: TCB/GGDC and OECD STAN database cent. Hence it is not only wage competition but about 20 per cent of the U.S. manufacturing also ‘productivity competition’ that the advanced productivity level which equals 26 per cent of countries are faced with. the EU-15 productivity level. In India and China, productivity in manufacturing is a fraction of that Table 6 compares relative levels of manufacturing in the advanced world, i.e. 2 per cent of the U.S. productivity for the same three advanced and level in India and 5 per cent of the U.S. level in three emerging economies as in Table 5. The table China. shows that productivity levels in the emerging economies are considerably lower than in the However, as manufacturing goods are primarily advanced economies: the three new EU member tradeables it is useful to compare not only states in Central & Eastern Europe perform at productivity but also the cost of inputs in the

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 235 production process. A well-known measure of ICT, pharmaceuticals, etc., against a higher share international competitiveness combines labour in medium-tech industries, such as machinery cost and productivity into a single measure of and transport equipment (O’Mahony and van labour cost per unit output. Unit labour cost is Ark, 2003). defined as the cost of labour required to produce one unit of output. As wage cost in the emerging For Services, Europe Should Look Towards the economies are also lower than in the advanced West countries, the differences in terms of unit labour cost are much smaller than for productivity. Table Table 4 has shown that market services account 6 shows that lower labour compensation more than for the largest part of the EU15-U.S. productivity offsets lower productivity levels in the emerging growth gap since 1995. It is therefore important economies. As a result, the manufacturing cost to get a better understanding of the sources of the competitiveness in the advanced economies much faster productivity growth in market services is considerably worse than in the emerging in the United States relative to the EU-15. economies. To get a clearer view on this, we first look Indeed it is questionable whether advanced in some more detail at the contribution from countries can ever compete solely on costs. Hence individual market services industries to the the call for an acceleration of R&D investment aggregate EU15-U.S. productivity growth gap (for example, the 3% R&D intensity target for the in the market economy. Here one can distinguish EU) and for more innovation in general seems to again between a ‘within industry’ effect due to be the obvious way forward for manufacturing faster productivity growth in the U.S. than in the activity in advanced countries. However, even EU, and a ‘between industry’ effect which relates in this area advanced countries face increased to a higher share of rapidly growing industries in competition from emerging economies. Recent the U.S.. OECD figures on the number of researchers, for example, show that China already has almost 900 Table 7 shows that most of the difference in thousand researchers as compared to 1.3 million market services productivity growth between researchers in the U.S., 1 million in the EU-15 and 1995 and 2003 can be traced to six industries, 650 thousand in Japan. The share of researchers concentrated in trade and finance industries. Part in total employment in still highest in Japan and of the difference can be explained by the fact that the U.S., but the Russian Federation, Korea and wholesale trade, retail trade and securities trade Taiwan already show a higher researcher intensity are larger sectors in the U.S. than in Europe, than the EU-15. The share of business enterprise but faster productivity growth within each researchers in the total number of researchers industry is by far the most important factor. For is highest in the U.S., Japan and Korea, but Europe, there is only a limited compensation comparable between China, the EU-15 and the due to faster productivity growth in European Russian Federation (OECD, 2005). telecommunication industries and construction. Furthermore despite faster productivity growth In sum, manufacturing competition from in U.S. banking, the somewhat lower share of emerging economies is not exclusively a cost this sector in Europe partly offsets this effect. matter, but also related to the capabilities of Since 2000 (not shown separately in the table) economies to generate innovation and raise the contribution of business services to aggregate R&D. In this respect, Europe is in a somewhat productivity growth has also improved in the U.S.. disadvantageous position relative to other In contrast, in European countries these service advanced economies, because it has a lower industries mostly show a productivity slowdown value added share in high-tech activities, such as – or at best stability – since 2000.

236 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Table 7. Percentage Point Contribution of Market Service Industries to Productivity Growth Gap between EU15 and the United States, 1995-2003

%-point of which contribution to within industry between productivity ('productivity') industry gap effect ('share') effect Wholesale trade 0.387 0.315 0.073 Retail trade 0.296 0.269 0.027 Securities trade 0.361 0.244 0.117 Banking 0.181 0.230 -0.049 Other business services 0.113 0.113 0.000 Motor vehicle trade 0.108 0.085 0.023 Professional services 0.068 0.067 0.001 Hotels & catering 0.051 0.052 -0.001 Transport services 0.032 0.051 -0.020 Air transport 0.048 0.037 0.010 Renting of mach. & eq. 0.017 0.032 -0.015 Computer services 0.022 0.016 0.006 R&D 0.000 0.000 0.000 Social & personal services -0.006 -0.005 -0.001 Inland transport -0.024 -0.014 -0.010 Water transport -0.030 -0.018 -0.012 Insurance -0.037 -0.019 -0.019 Communications -0.014 -0.059 0.045 Construction -0.068 -0.064 -0.004 Source: Van Ark and Inklaar (2005)

Unfortunately our knowledge about why these and productivity differs highly across industries large differences in productivity growth between and between countries. Griliches (1994) showed the EU15 and the U.S. arise is still limited. Van a striking difference between the acceleration Ark (2005) investigates the validity of a number of labour productivity growth in ‘measurable’ of explanations including (1) problems with sectors of the U.S. economy (agriculture, mining, the measurement of service performance, (2) a manufacturing, transport and communication, genuine shortfall in innovative capacity of service and public utilities) and the slowdown in industries in Europe, and (3) a lack of reforms ‘immeasurable’ sectors (like construction, trade, to exploit the productivity potential of service the financial sector, ‘other’ market services and innovation. Below follows a brief summary of government) over past decades. Apart from an these sources of uncertainty. increase in measurement error at the aggregate level due to shift towards the immeasurable Ad 1) Measurement problems in services sectors of the economy, one may also observe In the past few years there have been increasing an increase in measurement problems in the concerns about whether the macroeconomic ‘immeasurable’ sector itself. This component of statistics correctly trace the changes at industry the rise in measurement problems may – at least level. In practice, the quality of measures of output in part – be related to the increased use of ICT.

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 237 In practice the largest measurement problems is related to a lack of innovation. However, there relate to the measurement of output in the service is little direct evidence to substantiate this claim. sector. The current methodology of splitting the change in output value into a quantity component As documented in Section 3, ICT investment is an and a price component is difficult to apply to important enabler of innovation and productivity many service activities, as often no clear quantity growth. When focusing on market services it component can be distinguished. Moreover, is clear that the U.S. has been more successful possible changes in the quality of services are in obtaining productivity effects from ICT also difficult to measure. These problems are investment in services than EU. Table 8 shows not new, and improvement in measurement of growth accounting results for five countries service output has been a topic on the agenda of (namely France, Germany, The Netherlands, the statisticians and academics for a long time.10 In UK and the U.S.), for which the contribution of many service industries information on inputs market services to aggregate productivity growth (such as labour income) was and still is used can be measured (Inklaar et al., 2005; Van Ark as a proxy for output. However, the increased and Inklaar, 2005). The results show that faster importance of ICT may have accelerated quality labour productivity growth in U.S. market services changes in services and raised the potential for is partly due to a faster growth in ICT capital productivity growth in services, which was deepening in the U.S., but much more so due to an previously not envisaged.11 However, to include improvement in TFP growth. Since 1995 TFP has those quality aspects in the output measure, contributed as much to labour productivity growth multiple dimensions of a service need to be taken as ICT capital deepening. ICT capital contributes into account, for example, the service concept, the much less to productivity growth in market services type of client interface and the service delivery in all European countries, and TFP growth is even system (den Hertog and Bilderbeek, 1999). negative with the exception of the UK. This implies that the real output of a particular service cannot be measured on the basis of one But what the TFP residual essentially represents single quantity indicator. New measurement has remained somewhat unclear. Clearly it might methods make use of various volume measures include the positive effects from unmeasured in, for example, financial services (e.g, in the factor inputs in services, notably the effects from Netherlands and in the United States) and health non-technological innovation and intangible services and other government services (e.g., in investments in human capital, organizational the United Kingdom). Even though such changes capital and knowledge creation. Indeed the in measurement methods have not exclusively led productive use of ICT investment in services is to upward adjustments of real output, on balance strongly dependent on various dimensions of the bias is probably towards an understatement non-technological innovations. of the growth in real service output (Triplett and Bosworth, 2004). There is no evidence, however, There are different ways to go about measuring that this bias is in any way bigger in Europe than non-technological innovation and its impact on in the U.S.. productivity growth. For example, van Ark (2005) looks at ways to organize the industry data on Ad 2) A lack of innovation in services? the basis of type of innovation in the industry. A It is sometimes claimed in the literature that crucial consideration for such a service innovation slower productivity growth in services in Europe typology is the way in which suppliers of inputs

10 See, for example, Griliches (1992), Wölfl (2003) and Triplett and Bosworth (2004). 11 See, for example, Baumol (2004) and Triplett and Bosworth (2002).

238 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH TableTable 8. Contributions 8: Contributions of ofMarket Market Services Services and and UnderlyingUnderlying Sources Sources to to Market Market Economy Economy Labour Productivity Growth,Labour Productivity1987-2003 Growth, 1987-2003 France Germany Nether- United United lands Kingdom States

1987-1995 Market Economy Labour Productivity Growth 2.4 2.6 1.7 3.0 1.4 Contribution of market services 0.5 0.9 0.5 1.0 0.5 of which: ICT capital deepening 0.2 0.3 0.3 0.3 0.4 Non-ICT capital deepening 0.2 0.3 0.2 0.5 0.1 Labour quality growth 0.1 0.1 0.1 0.4 0.2 Total factor productivity growth 0.0 0.2 -0.2 -0.2 -0.1

1995-2003 Market Economy Labour Productivity Growth 1.8 2.1 1.4 2.6 3.5 Contribution of market services 0.1 0.3 0.6 1.3 2.0 of which: ICT capital deepening 0.3 0.4 0.6 0.5 0.8 Non-ICT capital deepening 0.0 0.1 0.3 0.4 0.3 Labour quality growth 0.1 0.0 0.1 0.1 0.1 Total factor productivity growth -0.4 -0.2 -0.3 0.2 0.8 Source: Van Ark and Inklaar (2005)

(machines, computers, and human capital), the has also gone together with strong productivity service company and its customers (consumers growth in wholesale trade, which explains the of intermediary users) interact. U.S. advantage in client led services. These industries benefited from the supply of ICT, but Using the service innovation typology, Van Ark have also undergone significant organizational (2005) showed that the innovation process in innovations. Indeed in industries that are primarily services is strongly dependent on innovations characterized by organizational innovations, by suppliers and users in the value chain. For U.S. performance has also strongly improved, in example, the estimates for the U.S. show a strong acceleration in productivity growth in particular in banking. those services which depend most strongly on innovation by their suppliers. For example, Within the EU, the experiences in service the retail industry has benefited strongly from productivity growth are mixed across industries productive exploitation of ICT. For example, the and countries (Van Ark et al., 2003a). Although introduction of barcode scanning allowed for services will be an important engine for future more efficient check-out systems and enabled productivity improvements, the exploitation a reorganization of the supply chain and the of the potential for productivity growth will be introduction of new shopping concepts. ICT also strongly dependent on national circumstances, supported the introduction of complementary including the nature of the innovation system and technologies (such as RFID, transportation the working of product and labour markets technology) and organizational change (new shopping concepts, adjustment in the logistic Ad 3) A lack of market reform in services? chain of supplying the shops more frequently, There has been much discussion in the literature etc.). The strong improvement in U.S. retail trade about the link between, on the one hand, the

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 239 performance of product and labour markets and, There are three categories of regulation that can on the other hand, innovation and productivity. be logically associated with slow productivity The basic argument has been that regulation growth in European retailing—store opening restricts competition to a much greater extent in hours, land usage restrictions (especially on large Europe than in the United States. Quantifying stores), and labour laws. these differences is difficult, but a wide variety of evidence suggests that regulation does indeed But the situation in Europe is changing rapidly. matter in reducing productivity growth.12 Product market regulations have been eased in many countries, and competitive incentives for However, explaining sluggish productivity growth change are increasing. Some of the slow TFP in Europe by broadly casting it as overregulated growth of the late 1990s may be due to the actual and uncompetitive is not very useful analytically. adjustments being made. As many European There is much variety and subtlety in the way by countries quickly increase their IT infrastructures, which regulation affects service productivity and they will be better positioned to exploit the innovation. It is essential to understand if and efficiencies of the new retail business models how regulation constrains productivity. Instead of once the effects of deregulation kick in. giving an overall view of the interaction, it may be preferable to focus on specific industries. In sum, while the overall picture points in the directions of regulations hampering productivity For example, McGuckin et al. (2005) provide a growth in services in Europe, there are many detailed discussion of productivity, innovation subtleties in how it exactly impacts on productivity and regulation in retail trade. The study shows growth. There are large differences between that U.S. retailers and wholesalers have been able EU countries. In fact the lack of a harmonised to boost their overall operational effectiveness regulation system in itself is often cited as a major in a way that firms in many European countries difficulty in building cross-border operations have not. U.S. retailing was transformed from within Europe. It should also be stressed that a low-technology sector to one of the most complete deregulation is not always the best way intense users of information and communication to raise productivity growth. Moreover, there technologies (ICT). The technologies used in this is a substantial time lag in reforms impacting sector rewarded scale and scope, enabling large on productivity. In this respect, it remains centralised chains and big stores to expand rapidly. an important question whether the European U.S. firms, which were relatively unaffected by slowdown is just a reflection of a lagged reform regulation and custom, have taken advantage of process, or that rigid institutions and regulations the opportunity to combine new technologies hamper the adjustment process. and organisational change to generate rapid productivity growth. 5. Policy issues European retailers and wholesalers have also On balance, this paper suggests that the European been investing in ICT capital at similar rates to slowdown in productivity growth is a reflection U.S. firms in recent years. But the IT share of of an adjustment process towards a new industrial overall investment is still considerably lower structure, which has developed more slowly in than in the United States. A likely reason is that the EU than in the U.S.. But with some delay, the incentives to invest in ICT are lower given the rapid diffusion of new technology may ultimately burdensome regulatory environment in Europe. facilitate the adjustment process towards a faster

12 See, for example, Nicoletti and Scarpetta (2003).

240 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH growth track in Europe. After all, the United particular industries or technology areas easily States has also gone through a phase of slow raises questions on whether governments are productivity growth during the 1980s.13 However, able to make the right choices. Moreover, the an institutional environment that slows down scope to directly influence innovation activities change may hold up the structural adjustment in services is limited, as most innovations arise process in Europe and inhibit the reallocation of in the value chain through market interaction resources to their most productive uses. between suppliers and clients. Nevertheless it is clear that governments have a responsibility In a market economy the main way for public for creating the ‘rules of the game’ concerning policy makers to promote and support faster technology creation and diffusion. ‘Technology productivity growth is to try and encourage creating’ measures are of particular importance private enterprises to move in a productivity- for moving the productivity frontier and enhancing direction. Governments can use a mix improving best practices, and include measures of four main policy mechanisms, which are only such as R&D policy and the creation of effective partly directly targeted towards productivity- patent systems. ‘Technology diffusing’ measures enhancing measures. play a major role in reducing the productivity gap between average and best practice firms, The first mechanism concerns macro-economic including best practices abroad. They involve management, which influences the relative prices the facilitating of training programmes, support of capital and labour inputs and hence determines of innovation platforms and other ways of co- the choice of technology. It may be argued that operation between government and business. wage moderation policies and active labour market policies (which have been applied in a The investment decisions concerning tangible different mix and intensity in European countries) and intangible capital, and the (re)allocations of have lowered the price of labour relative to capital these inputs to business processes, are taken by in Europe. Although conclusive evidence on the firms in an environment, governed by markets in precise relationship is still lacking, the relative which supply and demand for factor inputs (labour decline in the price of labour may have impacted and capital markets) and product and services the slowdown in the growth of the capital-labour (product markets) are matched. Governments ratio during the 1990s. For many European play an important role in setting the ‘rules of the countries this slowdown can be clearly observed game’ (or institutions) of these markets, which and is an important source for the slower growth is the third main policy mechanism. In the past in labour productivity. many existing institutional settings or regulatory arrangements have originally been set up with An important explanation for the slowdown the motivation to smooth the functioning of the in Europe comes from slower growth in total markets, by streamlining rules on competition, factor productivity, i.e. productivity growth business conduct, labour markets, consumer corrected for the change in capital-labour ratios protection, public safety, health and so on. (Timmer and van Ark, 2005). Slow TFP growth However, regulations may have become a drag may therefore be related to failing innovation. to the extent that they limit the efficiency of The second policy mechanism, which includes market functioning, reduce entry of new firms measures directed to support technological and delay exits. There has been an increasing change and innovation, is therefore very popular awareness of the need for an innovation-specific with governments. However, direct support of focus on (de)regulation and its impact on growth

13 See, for example, Dertouzos et al. (1989).

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 241 and productivity performance in the knowledge (local) market. Of course, rapid restructuring economy. The opportunities to exploit new through cuts has been propagated as the recipe technologies are to a large extent determined for the recovery of U.S. firms and global firms by the regulatory environment. There is much in general. The fundamental difference is that evidence that higher entry and exit rates of when such a strategy is pursued in a market firm within industries are supportive of faster environment that is more flexible, it may help productivity growth (OECD, 2003). to reposition the firm, activate the resources and realize the potential. Another difference between Finally, ‘horizontal policies’, which represent the the EU and the U.S. is that when entry and exit of fourth main policy mechanism, concern policies firms is speeded up, the reallocation of resources that are not directly related to innovation, are at to its most productive uses is strengthened. least as important to improve service innovation Hence in a more flexible market environment activity. As human capital is a key input in the the strategy towards restructuring can be more innovation process, there is a clear role for the easily aligned with exploiting the potential for government to provide an adequate formal growth and reducing the gap between average education system. More specifically governments and best practice through maximizing the returns should support a higher education system that on investments in high performing capital goods has the flexibility to train excellent researchers, and intangibles. to support their mobility, and to allow business to tap into the knowledge of universities and other References higher education institutions for commercial purposes. As the evidence from recent OECD Alesina, A., E.Glaeser and B. Sacerdote (2005), statistics shows, emerging economies are “Work and Leisure in the U.S. and Europe: Why becoming important challengers in terms of the so Different?”, Discussion Paper Number 2068, competition for talent. Harvard Institute of Economic Research. The optimal mix of these four main policy Ark, B. van (2005), “Does the European Union mechanisms is difficult to determine. It depends Need to Revive Productivity Growth?”, Research on such factors as the distance relative to the Memorandum GD-75, Groningen Growth and world technology and/or productivity frontier, Development Centre, May (downloadable from which may differ between industries. It may also depend on the state of institutional reform http://www.ggdc.net/pub/gd75.pdf) in particular markets. Finally, the nature of the political reality implies that all public policy Ark, B. van, R. Inklaar and R.H. McGuckin interventions are likely to involve costs as well (2003a), “ “Changing Gear: Productivity, ICT as benefits. and Service Industries in Europe and the United States”, in J.F. Christensen and P. Maskell, eds., The key to productivity improvements, however, The Industrial Dynamics of the New Digital is with business itself. For business there is Economy, Edward Elgar, pp. 56-99 a choice between a strategy focused on cost reductions through scrapping and postponement Ark, B. van, L. Broersma and P. den Hertog of investments in new capital goods and (2003b), “Services Innovation, Performance intangibles, or by restructuring through and Policy: A Review”, Research Series No. 6, upgrading the resources and overcoming the Strategy, Research & International Co-operation bottlenecks which account for the difference Department, Ministry of Economic Affairs, The between average and best practice in a given Hague.

242 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Ark, B. van and R.H. McGuckin (1999), Hall, Robert E. (1988), The Relation Between “International Labor productivity and per capita Price and Marginal Cost in U.S. Industry. Journal income,” Monthly Labor Review, July, pp. 33-41. of Political Economy, 96(5), 921-47.

Ark, B. van and R.C. Inklaar (2005), ‘Catching Hertog, den P. and R. Bilderbeek (1999), Up or Getting Stuck? Europe’s Troubles to “Conceptualising Service Innovation and Exploit ICT’s Productivity Potential’, Research Service Innovation Patterns”, mimeographed, Memorandum GD-79, Groningen Growth and DIALOGIC, Utrecht (http://www.eco.rug.nl/ Development Centre, September (downloadable GGDC/dseries/SIID_frontpage.shtml) from http://www.ggdc.net/pub/gd79.pdf) Inklaar, R.C., M. O’Mahony and M. P. Timmer Baumol, W.J. (2004), “Four Sources of (2005), ‘ICT and Europe’s Productivity Innovation and Stimulation of Growth in the Performance; Industry-level Growth Account Dutch Economy”, De Economist, vol. 152, pp. Comparisons with the United States’, Review of 321 – 351. Income and Wealth, 51(4), pp. 505-536

Blanchard, O. (2004), “The Economic Future of Jorgenson, D.W. (1995), Productivity, Volume Europe”, The Journal of Economic Perspectives, 1: Postwar US Economic Growth, MIT Press: vol. 18, pp. 3-26. Cambridge

Brynjolfsson, Erik and Lorin Hitt (2000), Jorgenson, D.W. and K. Stiroh (2000), ‘Raising “Beyond computation: information technology, the Speed Limit: U.S. Economic Growth in the organizational transformation and business Information Age’, Brookings Papers on Economic performance”, Journal of Economic Perspectives, Activity, vol. 1, pp. 125-211. vol. 14 no. 4, pp. 23-48, Fall. Jorgenson, D. W., M. Ho and K. J. Stiroh (2003), Dertouzos, M.L., R.K. Lester and R.M. Solow Growth of U.S. Industries and Investments in (1989), Made In America. Regaining the Information Technology and Higher Education. Productive Edge, MIT Press, 3rd edition. Forthcoming in Corrado, Haltiwanger, and Sichel (eds.) Measuring Capital in a New Economy, European Commission (2004), Facing the University of Chicago Press: Chicago challenge. The Lisbon strategy for growth and Employment, Report from the High Level Group McGuckin, R.H.. and B. van Ark (2005a), chaired by Wim Kok, Brussels. Performance 2005: Productivity, Employment and Income in the World’s Economies, Research Gordon, R.J. (2004), “Two Centuries of Economic Report R-1364-05-RR, The Conference Board, Growth: Europe Chasing the American Frontier”, New York CEPR Discussion Paper No. 4415, London. McGuckin, R.H.. and B. van Ark (2005b), Griliches, Z., ed. (1992), Output Measurement “Productivity and Participation: An International in the Service Sectors, Studies in Income and Comparison”, Research Memorandum GD-78, Wealth, Volume 56, National Bureau of Economic Groningen Growth and Development Centre, Research, Chicago University Press. August (downloadable from http://www.ggdc. net/pub/gd78.pdf) Griliches, Z. (1994), ‘Productivity, R&D, and the Data Constraint,’ American Economic Review, McGuckin, R.H., M. Spiegelman and B. van Ark vol. 84, no. 1, pp. 1-23. (2005), The Retail Revolution. Can Europe Match

KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH 243 U.S. Productivity Performance?, Perspectives on Federal Reserve Bank of Minneapolis Quarterly a Global Economy, Research Report R-1358-05- Review, Vol. 28, No. 1, July 2004, pp. 2–13. RR, The Conference Board, New York. Roeger, W. (1995), ‘Can Imperfect Competition Nicoletti, G. and S. Scarpetta (2003), “Regulation, Explain the Difference between Primal and Productivity And Growth: OECD Evidence: Dual Productivity Measures? Estimates for U.S. OECD Evidence”, Economic Policy, Issue 36, Manufacturing,’ Journal of Political Economy, pp. 9-72. 103(2), 316-30.

OECD (2002), Measuring the Information Sapir, A. (2005), “Globalisation and the Reform of Economy, OECD, Paris European Social Models”, background document for the presentation at ECOFIN Informal Meeting OECD (2003), The Sources of Economic Growth in Manchester, Bruegel Institute, Brussels. in OECD Countries, Paris. Solow, R.M. (1957), “Technical Change and OECD (2004), The Economic Impact of ICT, the Aggregate Production Function”, Review of Measurement, evidence and implications, OECD: Economics and Statistics, vol. 39 no. 3 pp. 312- Paris. 320. Timmer, M.P. and B, van Ark (2005), “IT in OECD (2005), OECD Science, Technology and the European Union: A driver of productivity Industry Scoreboard, OECD, Paris. divergence?”, Oxford Economic Papers, vol. 57 no. 4, pp. 693-716. Oliner, S. D. and D. E. Sichel (2000), ‘The Resurgence of Growth in the Late 1990s: Is Stiroh, K.J. (2002b), “Information Technology Information Technology the Story?’, Journal of and the U.S. Productivity Revival: What Do the Economic Perspectives, vol. 14, pp. 3-22. Industry Data Say?” American Economic Review, vol. 92, no. 5, pp. 1559-1576, December. Oliner, S. D. and D. E. Sichel (2002), “Information Technology and Productivity: Where Are We Now Triplett, J. and B.W. Bosworth (2002), “ “Baumol’s and Where Are We Going?”, Federal Reserve Disease” has been Cured: IT and Multifactor Bank of Atlanta Economic Review, 3rd Quarter Productivity in U.S. Services Industries”, 2002, 87(3), pp. 15-44 Brookings Institution, mimeographed. O’Mahony, M. and B. van Ark, eds. (2003), Triplett, J. and B. Bosworth (2004), Productivity EU Productivity and Competitiveness: An Industry Perspective. Can Europe Resume the in the U.S. Services Sector. New Sources of Catching-up Process?, DG Enterprise, European Economic Growth, The Brookings Institution, Union, Luxembourg (downloadable from http:// Washington D.C www.ggdc.net/pub/EU_productivity_and_ competitiveness.pdf). Wölfl, A. (2003), “Productivity growth in service industries – an assessment of recent patterns and Prescott, E. C. (2004) “Why Do Americans Work the role of measurement,” STI-Working Paper So Much More Than Europeans?” 2003-07, (Paris: OECD)

244 KNOWLEDGE ECONOMY: COMPETITIVENESS AND GROWTH Keynote Speech: Knowledge Economy – Challenges for Official Statistics

Walter Radermacher/ Susanne Schnorr-Bäcker Federal Statistical Office, Wiesbaden

Executive Summary all empirical researchers and interested citizens with basic data they can use to continue their The bodies of official statistics are the largest work. provider of statistical data. Due to the digitalisation of information and the development and In the United Europe, national official statistics is application of new, network-based information and essentially determined by the requirements of and communication technologies (ICT), information by European Community institutions. The focus is of much greater importance today than it was of new developments is defined by the spirit of only a few years ago. This fact is illustrated also the times. Especially in the fields of research and by the expression “information society” used as innovations, ICT, human capital, competitiveness an umbrella term for the structural change which and growth there are numerous approaches to could be observed for some years now in the improve the supply of statistical data. society and the economy, indicating a departure from the industrialised society. Especially in Drawing upon examples, this paper treats selected highly developed economic systems with only measures of European official statistics and limited natural resources at their disposal has outlines their significance from the viewpoint of information become an important factor of a national statistical institute. production. Only if these economies manage to get still further ahead in science and research Numerous synergy effects can be achieved in an and translate that into innovations and human increasingly integrated system of official statistics capital can they secure and consolidate their organised along the lines of work distribution, positions, and particularly economic growth and but there also is the danger of developments in employment. Knowledge will then successfully the wrong direction. Optimised co-operation is replace the former, limiting production factors of an indispensable prerequisite for official statistics work, nature and capital. This holds true not only to secure its leading position and maintain it in for individual nations like Germany, but also for the future. large economic areas such as the European Union or North America. In a globalised economy, all 1. Preliminary notes actors are permanently competing with each other. The term “knowledge economy“ names a phenomenon which for about ten years has been What part is played by official statistics in discussed in all spheres of society, economy and globalised economies? It provides important politics under various headings such as “new information required for planning and for economy“, “information society“ or “knowledge preparing decision-making processes in politics based society“. Through the digitalisation and the economy. Furthermore, it supplies above of information and the development of new,

KNOWLEDGE ECONOMY: PANEL DISCUSSION 245 network-based information and communication information technology was used above all for technologies (ICT), in particular the establishment rationalisation and automation purposes until of the internet, the significance of information the end of the eighties, it was the introduction of changed all of a sudden. An undreamt-of wealth personal computers which provided access to these of information is accessible to a far greater group new technologies for a large group of persons. of persons today than ever before. General and The focus of demand shifted from hardware even up-to-date information may be called up to software. Since the mid-nineties – when a almost without delay and nearly everywhere. worldwide deregulation of networks began – an How rapidly these developments have progressed increasing integration of information processing, is illustrated by the internet1: within a time span telecommunication, software, consumer of only ten years, the number of websites rose electronics, information services and media can be from 100 000 in 1995 to about ten billion today. observed. The world market is probably ready for The private internet surfer surfs the net 650 times informatisation. Digital networks will to a growing faster today than in 1995. While in 1995 internet extent define and determine the relations between users paid some five Euros for being online for producers and users, suppliers and customers, one hour, these days they can use the internet in citizens and the state. Germany for one month for the same amount, paid as a flat rate. Wireless LAN and modern Today the new information and communication radio transmission standards make it possible to technologies enable everyone to get informed go online anytime and anywhere. about the present status quo of a certain range of topics rapidly and almost worldwide, to discuss So far, work, nature and capital have been that information with others and to communicate regarded as the essential determinants of growth. his own insights. As a consequence, however, the This perspective determined also economic half life of up-to-date information is declining statistics and even more national accounts of steadily. For example, the knowledge once post-war times. For some time now, this approach obtained will not suffice to exercise a profession is being abandoned or at least expanded. Closely for a lifetime. To avoid misjudgement, information related with the term “knowledge economy“ is relevant to decision-making has to be checked the name of Paul Romer, an economist who was continuously as to its currentness. Enormous counted among the 25 most influential Americans demands are made on human capacities. by the Time Magazine in 1997 already: For more than twenty years he has pondered the In the following, the role of official statistics in a following question: “By what can economic united Europe will be treated from the viewpoint growth be achieved and ensured in times of of a national statistical institute. For the individual declining marginal utility of physical goods and subject areas of this seminar such as their scarcity?“ As an answer to that question • research and innovation he introduced knowledge as another factor of • ICT production and growth into his “new growth theory“ in addition to the factors mentioned • human capital above. • competitiveness and growth, aspects selected as examples will be pinpointed Information, above all in digitalised form, and used to illustrate the significance for official has become an important input factor.2 While statistics.

1 See www.wdr.de/themen/computer/internet/web 2 See also the explanations in Newodow, L.A.: Der sechste Kondratieff, Sankt Augustin 1996, p. 94 et. seq.

246 KNOWLEDGE ECONOMY: PANEL DISCUSSION 2. Knowledge and official statistics statistics: So, on the one hand, it is demanded that legal norms be abolished. On the other, the Official statistics is an indispensable part ofa bodies of official statistics are allowed to operate democratic society.3 It provides the statistical in the long run only if a detailed legal provision information required for the development of has been adopted for that very purpose. an informed opinion and decision-making processes in politics, the economy and society The bodies of official statistics are the largest while ensuring neutrality, objectivity, scientific supplier of statistical information – and that will independence and confidentiality regarding the not be the case only in Germany. The Federal microdata it has been entrusted with. Statistical Office in Wiesbaden and the 15 statistical offices of the Länder have an enormous It is thus the mission of official statistics to reflect treasure of statistical data. German federalism, the societal, economic and ecological phenomena, a historical feature, is reflected by the structure trends and relations within an economic area, and organisation of federal statistics and does not be it a national state or a confederation like the always make it easy for all those involved to act European Union, as precisely as possible. Not as efficiently as would theoretically be possible only relevance, but also currentness and accuracy and maybe also necessary from a practical are important quality aspects, whose significance viewpoint; this will probably hold true also for may however differ depending on the specific the European Statistical System. Therefore, a information requirements or purpose of a given reform programme was initiated in Germany in user or user group. Users’ demand for data from 2003, the “Master Plan for a Reform of German official statistics is extraordinarily great and Official Statistics“.4 That master plan centres on heterogeneous. Furthermore, all information an optimised co-operation especially where tasks should if possible be problem-oriented, free of are performed jointly (by the Länder). The aim is charge and rapidly available. to improve the efficiency of co-operation between the statistical offices and thus the efficiency of the Even though the bodies of official statistics system of official statistics in Germany. Another can in many places be managed like (business) goal is a still stronger strategic orientation of the enterprises and compete with numerous private German statistical programme5 as has already institutions, their integration into the public sector been realised in several member states of the does not make it easy for them these days to fulfil European Union. their tasks. Efforts of the state to save money result in considerable cuts in the field of official These days, the major part of the national statistical statistics. Nevertheless, complaints about gaps programme is determined by Europe: more than in the statistical programme or unsatisfactory four fifths of the primary surveys and two thirds delivery conditions come in particular from users of the secondary surveys of the programme of of that sphere. At least in Germany, the obligation federal statistics are based on EU legislation. of the state to deregulate and disburden the citizens The focus of the statistical programme is defined and the economy from publicly induced burdens by the spirit of the times. After the end of World conflicts with the principle of legality of official War II, setting up a system of industrial statistics

3 See also “Fundamental Principles of Official Statistics”, passed in 1992 by the United Nations Economic Commission for Europe (UNECE) and adopted in 1994 by the United Nations Statistical Commission 4 See Masterplan zur Reform der amtlichen Statistik, Fortschrittsbericht, September 2004, edited by the statistical offices of the Fed- eration and the Länder 5 See Federal Statistical Office, Strategie- und Programmplan für die Jahre 2005 und 2006, Wiesbaden, April 2005

KNOWLEDGE ECONOMY: PANEL DISCUSSION 247 was of prime importance. Then the establishment especially the necessary indicators including and extension of integrated systems followed, operational definitions. first of all of national accounts. In the eighties, the state of the environment began to be covered The introduction and wide distribution of new more intensively. The more recent history is ICT is probably important in that context, which characterised by an increased coverage of services in the United States of America was accompanied and knowledge. by growth, a high level of employment and a relatively low rate of inflation from the end of the Statistical monitoring and coverage of the eighties to the year 2000. “New Economy“ was the “knowledge economy“ have become a new catchword for that era, which, although not with challenge into which the general economic and the same intensity and with certain delays, could societal developments are embedded. The bodies be observed in Europe, too.6 From 2000, a general of official statistics have been asked questions economic slowdown started worldwide from like: which the economy seems not to have recovered completely and permanently until today. ➢ Where are the drivers of growth and how are input, output and productivity related? In addition, an increasing globalisation started ➢ How can one get rid of “old fetters“, above manifesting itself above all in a growing volume all regarding work processes and (natural) of world trade and an intensified international resources? distribution of labour which in Germany is ➢ What does this mean for structurally referred to by terms like “outsourcing“, “off- changing, i.e. aging societies? shoring“ or “bazaar economy“. ➢ How are these trends accompanied by globalisation? These developments are reflected at least ➢ What does all that mean for official statistics’ rudimentarily in the European Statistical ability to adapt and its adjustment speed? Programme of the recent past and today already result in certain delivery requirements to be met by the national statistical institutes. These include, 3. Knowledge economy: Latest among others, developments from the viewpoint of official • the delivery of data for the so-called statistics structural indicators7

It is true that the fundamental ideas of the • the passing of EU regulations on the 8 “knowledge economy“ were formulated more than information society , on science and 9 10 twenty years ago. Nevertheless, at least in official technology and on innovation , statistics there is no generally acknowledged • the regulation presently treated in the consensus to date about its components and European Parliament and Council on the

6 See also Schnorr-Bäcker, S.: Neue Ökonomie und Amtliche Statistik, in Wirtschaft und Statistik 3/2001, p. 1 et seq. 7 For details see: Jörger, N.: Strukturindikatoren – Messung der Fortschritte im Rahmen der Lissabonner Strategie, in Wirtschaft und Statistik 12/2003, p. 1083 et seq. 8 Regulation (EC) No 808/2004 of the European Parliament and the Council of 21 April 2004 concerning Community statistics on the information society, OJEC No L 143 p. 49 et seq., quoted as Reg. (EC) information society 9 Regulation (EC) No 753/2004 of the Commission of 22 April 2004 implementing Decision No 1608/2003/EC of the European Parliament and of the Council as regards statistics on science and technology, quoted as Reg. (EC) science 10 Regulation (EC) No 1450/2004 of the Commission of 13 August 2004 implementing Decision No 1608/2003/EC of the European Parliament and of the Council concerning the production and development of Community statistics on innovation, OJEC L 118, p. 23 et seq., quoted as Reg. (EC) innovation

248 KNOWLEDGE ECONOMY: PANEL DISCUSSION activities of foreign subsidiaries and on on the input (personnel, investments, running the external activities of parent companies expenses, etc.) and on the output (in value and/or belonging to domestic groups of affiliated volume terms) should be of special interest here. companies.11 That is the only way to obtain indications – at least initially and implicitly – of how efficient Important impulses are provided also by the underlying processes are. These conditions the statistical bodies and seminars of other seem imperative for assessing the effects of R&D international and supranational institutions such and innovations on employment and growth as the OECD or the United Nations e.g. as regards continuously and in international comparison. the application and use of modern information and communication technologies (ICT) in business The following quotation, which is basically a and society. The bodies of national official footnote from the so-called Kok Report13, shows statistics also intensively treat all these recent the significance of statistical data and the context developments initiated by Eurostat and other in which official statistics operates: “…The EU supranational and international organisations. is not good enough at disseminating innovation. While ICT spending has risen significantly since the mid-1990s in the EU, this increase does not 3.1. Research and development, innovations seem to have brought about a similar or uniform boost to productivity across the EU economy, as Research and development as well as has been the case in the US.10“ This sentence is innovations are the basis of economic growth accompanied by a footnote stating: “10 Over the and employment. The bodies of official statistics period 1995-2002, the annual productivity growth in Germany publish rather selective information in ICT-using industries has been 1.6 % in the EU on research and development activities (R&D) and 4.7 % in the US. ICT spending amounts to by institutions and economic sectors12, some about 4.2 % of GDP in the EU in 2002, compared of which are moreover collected by various to 5.3 % in the US.“14 research institutions. However, these data do not provide a complete picture. To this end it would These remarks illustrate the fact that the European be necessary to include all statistical information Union is seen as permanently competing with relevant to R&D or innovation into the overall the United States. From the viewpoint of an system of official statistics and to collect and official statistician, the following question arises analyse them according to the standards applying instantly: Where do these indicators come from? to official statistics. At least for Germany this Parts as for instance GDP are calculated only means that a concept of a modular partial reporting by the bodies of official statistics. And are these system on R&D/innovation has to be prepared statistical results really fully comparable? taking into account the data presently available. The system should make evaluations possible It seems plausible that there are qualitative which are relevant to R&D and innovation both differences between the European Union and institutionally (especially by public and private the United States of America. But it is at least scientific institutions and enterprises by economic questionable if the interpretation of the figures sectors and size classes) and functionally. Data quoted is realistic, implying that productivity

11 Proposal for a regulation (EC) of the European Parliament and of the Council on Community Statistics on the structure and the activity of Foreign Affiliates (FATS) 12 See also Brugger, P.; Hetmeier, H.-W.: Wissenschafts- und Technologiestatistiken in Deutschland, in Wirtschaft und Statistik 3/1999, p. 197 et seq. 13 See also the Report of the Employment Task Force chaired by W. Kok of November 2003, http://www.iue.it/RSCAS 14 http://www.die-rente.info/download/Jobs-Jobs-Jobs_-_Employment_Taskforce_Bericht_de.pdf, p. 24, footnote 10

KNOWLEDGE ECONOMY: PANEL DISCUSSION 249 growth in the US was about three times as high norms is mentioned and, explicitly, the work of as in the EU over the period mentioned and that the OECD. It is undisputed that the Organisation ICT spending in terms of GDP was only 20 % for Economic Co-operation and Development lower in the EU in 2002. (OECD) has done valuable preparatory work in this field. Although not legally binding, this Statisticians always have to check in detail (as for work is generally acknowledged and makes large instance with the above variable “ICT spending”) synergic effects possible for all involved. From if and in how far comparisons on an international a national viewpoint, we welcome in particular basis are acceptable. This can only be done, that the development workload is thus kept as however, if there is great compatibility as regards small as possible and that the burden on all data subject matter and methodology. If there are suppliers is reduced to a minimum. The close co- significant differences it would be desirable – operation which has existed between Eurostat, despite the principle of subsidiarity of member international and supranational institutions and states – to achieve still greater methodological the national statistical institutes in several areas harmonisation of the content and the procedures. partly for a long time (as e.g. with the OECD in research and innovation) should be extended and That is not supposed to mean, however, that the intensified. In view of ever scarcer resources in bodies of European official statistics should merely official statistics, strict priorities must however adopt the procedures and concepts of other large be set. economic areas, instead, it has to be examined for every individual case if and to what extent It is moreover important and trend-setting that such indicators reflect the actual phenomena to a especially research and innovation activities sufficient degree and provide reliable statistical are covered statistically in all areas. This means results for the intended purpose. President that not only the public sector is studied, which Hahlen, Head of the Federal Statistical Office, made a sizeable contribution to the promotion of just recently put it into the following words at an research in the past, but also trade and industry. informative event for delegates of the European Most innovation activities, in particular, are Parliament: “Reliable and comparable statistical probably realised in that area today. This requires data are important instruments in the process at least more co-ordination on the part of official of enlargement and integration. The individual statistics as far as they hold on to the classical member states also need reliable data material to sectoral breakdown. define their own position.“15 While it was sufficient for the staff of official Research and innovation is a field which appears statistics some years ago to have detailed trend-setting also from another aspect. Here knowledge of their respective field of work, two EU regulations were passed in the past interdepartmental knowledge is also needed year, one of them on innovation statistics16 and these days. The close link-up of all areas made the other on science and technology17. I should possible by the new media today has to be like to stress the following aspects of both implemented also at the personal level where initiatives: on the one hand, in the reasons for this has not yet been done. The nearly unlimited consideration, among other things, the coherence capacities of technical means contrast with only of Community statistics with other international limited cognitive abilities of human beings.

15 Hahlen, J.: Zur Lage der deutschen und europäischen Statistik, paper held at an informative event for the German delegates of the European Parliament and their scientific staff in Brussels, printed in Wirtschaft und Statistik 7/2005, p. 665 et seq., here p. 666 16 See Reg. (EC) innovation (l.c.) 17 See Reg. (EC) science (l.c.)

250 KNOWLEDGE ECONOMY: PANEL DISCUSSION The importance of co-ordination and thinking 3.2. ICT in networks is growing because otherwise – if Even though the new information and real phenomena are regarded only in an isolated communication technologies are rooted way back manner – there is a great danger that there will be in the past, there has been a fundamental structural inadequate and untenable burdens on respondents. change since the beginning of the nineties. Their being overburdened combined with an Digitalisation and net-based technologies, above increasing discontent with the state directly all the introduction of the internet, were decisive affects the quality of statistical results and thus determinants of that change.19 the acceptance of official statistics. Such a control circuit is extraordinarily sensitive. All fields of society, economy and politics have been penetrated with a rapidness that is unique In the past, primary surveys were a major data in history. The bodies of official statistics have source of official statistics. It is doubtful whether monitored these developments right from the this will continue to be the case as developments start. The statistical coverage of ICT in official in several countries show. At least the various statistics over the years is a good example of how instruments used to obtain data (primary and to proceed in general to cover new developments secondary surveys, use of other sources and statistically as they develop. estimation procedures) must be subtly geared towards each other and combined with a view to Initial indications of recent developments such as purpose. the importance of ICT for society and economy are usually provided by goods-related statistics. A first step towards using secondary statistical For instance, personal computers were covered data more frequently than in the past, i.e. material for the first time in Germany by the multi-annual available at other, mostly public institutions, is sample survey of income and expenditure of explicitly mentioned in the Regulation on research 1993 and in the annual statistics of continuous and technology18. In Section 3 of the Annex to the household budget surveys. Several non-recurring Regulation, statistical fields such as statistics on additional questions on teleworking were asked human resources, on patents, on high-technology in the 2000 microcensus already.20 industries and knowledge-based services as well as on science and technology are listed, data which Since 2002, ICT has been covered statistically and have so far not been collected for this purpose with more detail at enterprises and at households at least in German official statistics. In practice, and individuals by means of EU-wide pilot the linkage between the various data producers studies. The procedure chosen was exemplary as and suppliers is intensified by such provisions, regards the approach. First, the feasibility of such new forms of co-operation may be found such a project was tested extensively by means of pilot as the public private partnership (PPP), by which studies. From the aspect of content it is noticeable the pioneering role of official statistics can be that the variables covered in the pilot surveys were reinforced and strengthened. closely related with the eEurope benchmarking

18 See Regulation (EC) No 753/2004 (l.c.), here Annex, Section 3 19 See also the comments in Section 3 20 See statement of President Johann Hahlen “Leben und Arbeiten in Deutschland – Ergebnisse des Mikrozensus 2000” of 19 April 2001

KNOWLEDGE ECONOMY: PANEL DISCUSSION 251 indicators of the eEurope Action Plans 2002 and 3.3. Human capital 2005. Eurostat has managed to strengthen the role of European statistics in a field which before had Research and development as well as successful been dominated by various private data suppliers. innovation activities require qualified specialised The major statistical indicators named in the staff. Especially countries like Germany which respective resolution of the Council21 have since are not rich in natural resources depend on human been supplied by the bodies of official statistics. capital to maintain their competitiveness over the A similar approach is intended for the so-called medium and the long term. i2010 strategy22 of the European Commission. For official statistics, human capital is a vast Finally, an EU framework regulation was passed area. There merely are different starting points in 2004 concerning statistics on the information especially with regard to institutional and society23, whose details are settled by comitology functional aspects; however, they do not provide decisions. an overall picture. At the institutional level, various actors can be distinguished, e.g. That Regulation contains several components which are trend-setting despite the related • public institutions additional burden on the bodies of official • enterprises statistics. It should be mentioned in particular • every individual citizen. that In general, education and training take place • the data provision programmes specified in phases. They usually start with pre-school in the annexes to the framework regulation education; then school education follows. After 24 will apply for a limited duration only that, professional training is aimed for. With • the programme of characteristics can be the start into working life, a phase of advanced adjusted annually under the so-called vocational training and continuing education comitology often begins, too. The breakdown into phases • adjustment measures are determined does not necessarily imply, however, that the “… taking into consideration Member stages will be passed in the above order. States’ resources and the burden on respondents ….“.25 In relation with human capital, specific training courses are of particular interest; the focus may This makes it possible to monitor statistically shift over time as the following example shows: also recent trends which are subject to rapid in the past, the German economy profited from its change and whose developments can scarcely excellent training and culture of skilled workers. be anticipated. The hint at the availability of These days, in a time when knowledge increasingly resources is important – even though it can not forms the basis, a good (university) education be determined precisely beforehand, but only especially in the natural sciences and technical be estimated – to ensure the acceptance of these occupations (in the field of engineering)26 is of surveys and their efficiency with all involved. particular importance.

21 See Council Resolution of 18 February 2003 on the implementation of the eEurope 2005 Action Plan, Official Journal of the Euro- pean Union C 48/2, p. 2 et seq. 22 COM (2005) 141 of 12 April 2005 23 Regulation (EC) No 808/2004 of the European Parliament and of the Council of 21 April 2004 concerning Community statistics on the information society, quoted as Reg. (EC) information society 24 See Reg. (EC) information society (l.c.), Annexes 1 and 2, Letter c, respectively 25 See Reg. (EC), information society (l.c.), Article 8, Paragraph 2 26 See also e.g. Egeln, J., Heine, C.: Indikatoren zur Ausbildung im Hochschulbereich, study commissioned by the Federal Ministry for Education and Research (BMBF), Mannheim, March 2005, p. 119

252 KNOWLEDGE ECONOMY: PANEL DISCUSSION In German official statistics, there have for a long particular significance for the statistical coverage time been statistics on the status of education of growth and competitiveness. At its special of the population and, based on the current summit in March 2000, the Lisbon European statistics of education, on school leavers and Council set the goal for the Union “ … to become university graduates by subjects. There also is the most competitive and dynamic knowledge- statistical information on degrees obtained and based economy in the world capable of sustainable on apprentices. Furthermore, statistical data is economic growth with more and better jobs and available on academic staff and selected research greater social cohesion.“27 Nobody could have institutions. Since in Germany these areas are anticipated that in the year 2000 two contrary mainly covered by public institutions in the developments would occur, namely the successes broader sense or controlled by the state, it is rather of the European Union (from the completion of easy for the bodies of official statistics to draw the Single Market and Economic and Monetary upon existing data stocks. The input-oriented Union to the introduction of the Euro) on the one side of (university) education and (vocational) hand and the start of a longer period of general training is thus covered relatively well. economic slowdown in the large industrialised nations on the other. There will be a general What has largely been missing in official consensus about the fact that the goals specified statistics is information on in-plant and individual in Lisbon will quite probably be achieved only measures and expenditure on (advanced) training partly until the end of the decade. and (continuing) education. Major progress was recently achieved by adopting Regulation (EC) With the help of the structural indicators, six No 1552/2005 of the European Parliament and subject areas are covered, including the general of the Council of 7 th September on statistics economic situation, employment, research relating to vocational training in enterprises. It is and development. As the data catalogue often also deplored that there is no information meanwhile comprises a multitude of indicators on the efficiency of (advanced) training and (about 50 indicators with almost 150 partial (continuing) education measures. That so-called indicators) which have to be adjusted to the latest developments every year, the European output-oriented approach is pursued, among Commission and the Council agreed to focus others, at OECD level. As especially company on 14 selected indicators, the so-called shortlist, advanced training and continuing education in the spring report 2004.28 Without treating are increasingly transferred to third parties and individual indicators in detail here, in particular bought from providers of such services, business their relevance and the potentials and limits of statistics supply starting points for statistical their informative value, such a set of indicators coverage. makes high demands on users. What significance do the individual indicators have? How do 3.4. Competitiveness and growth they have to be regarded in combination? How do contrary developments in various member Competitiveness and growth are economic states have to be assessed? These and many indicators which are of fundamental importance other questions arise when one puts oneself in for both every single enterprise and the economy the users’ position. In my opinion, a step in the as a whole. At the macroeconomic level, the right direction was the decision the Council took “structural indicators” initiative should be of together with the Commission to prepare quality

27 European Council: Conclusions of the Presidency, Lisbon, March 2000, paragraph 5 28 See also Radermacher, W., Köhler, S.: Amtliche Statistik in der erweiterten Europäischen Union, in Frankfurter Statistische Berichte, 4/2004, p. 180 et seq.

KNOWLEDGE ECONOMY: PANEL DISCUSSION 253 profiles for the individual indicators which are to standard with physical goods. Such information meet the following criteria29: on so-called metadata is absolutely necessary from the viewpoint of professional ethics. ➢ easy to read and understand ➢ policy relevant It has to be ensured as well that official statistics are ➢ mutually consistent compiled and disseminated according to common ➢ timely available standards in all of Europe. It has to be guaranteed ➢ comparable across Member States and as far that principles like impartiality, reliability, as possible with other countries objectivity, scientific independence, feasibility and statistical confidentiality apply to official ➢ selected from reliable sources statistics and above all that they are generally ➢ should not impose too large a burden on known and accepted. The recommendations Member States and respondents. of the Commission regarding independence, integrity and accountability of the national and With an average length of three pages, the Community statistical authorities30 are of specific individual quality profiles have been kept rather importance here. brief. Each quality profile is made up of seven partial aspects. It starts with a description of the purpose of the indicator and a gives the 4. Outlook reasons for its relevance with regard to the goals Globalisation, economic integration and set by the European Council in Lisbon. Rather technological innovation change life and work detailed comments on availability, accuracy and in Europe. There is a vast variety of measures comparability follow. Finally, the possibilities of and initiatives at the level of European official further development are treated. This enables not statistics to cover recent developments only statisticians but also scientists and empirical statistically, monitor and analyse them. For researchers to work on the possible informative official statistics in the member states this means value of specific statistical indicators and critically two things. On the one hand, these measures have examine their practical relevance not only at the contributed to providing the statistical programme macroeconomic but also the microeconomic with a multitude of impulses. On the other, these level. A recently taken important step in federal programme enlargements have imposed an statistics was to create especially the technical enormous burden on the statistical institutes and and organisational prerequisites enabling also partly also respondents. In a situation of generally empirical researchers and scientists to use de low budgets, the additional effort required from facto anonymised microdata of official statistics. the statistical institutes can scarcely be handled anymore. Quality reports or – in more general terms – the provision of metadata are indispensable in the The challenges to official statistics are enormous. age of the knowledge society. Here some kind of An ever more rapidly changing demand for instruction for use is given which provides the information and increasingly complex inquiries user with a guideline regarding the potentials and accompanied by shrinking resources cannot limits of using statistical information. This is not be managed without an integral, interlinked yet common with services but has long been a approach for official statistics with the

29 Communication from the Commission COM (2000) 594, 27 September 2000, p. 11 30 See Communication from the Commission to the European Parliament and to the Council on the independence, integrity and accountability of the national and Community statistical authorities, COM (2005) 217 of 25 May 2005, quoted as Commission Com- munication on independence

254 KNOWLEDGE ECONOMY: PANEL DISCUSSION continuously increasing number of “players” in To enable the bodies of official statistics to the growing system of information providers. meet requirements optimally also in the future, The necessary technical prerequisites are it is necessary - after these illustrating remarks available: the storage capacities are nearly – to pay attention in particular to the following unlimited, the processing time extraordinarily aspects: short, networking generally in a fairly advanced (1) constant review of tasks stage. Official statistics has always been familiar (2) efficient work flows with the required methodological and subject matter-related components such as classifications (3) modern and user-friendly information and registers. Also with a view to the future, it supply. has to be made sure that the individual groups/ Re (1) Constant review of tasks classes are formed so as to achieve minimum The comments have shown that the national difference within a class and maximum difference authorities of official statistics have continuously between classes. Especially in the case of recent accompanied European Unification and the developments such as the inclusion of ICT into integration process and supported them as best the classifications of economic activities, these they could. Enormous demands were made criteria should be examined thoroughly. Clear on their programmes, which these days are and elaborate structures for all elements relevant determined largely by the European Community. to statistics are an indispensable prerequisite for It is undisputed that the bodies of official statistics being able to act with maximum flexibility and constantly need to develop their product portfolio rapidness. Furthermore – as is shown by the further and adjust it on the basis of the dialogue experience gathered in other countries – registers between users and producers.31 At the present e.g. of economic units or the population are of time, Germany considers it quite unlikely that the essential importance. On the one hand, registers operational capacities in the statistical field will are an indispensable prerequisite for up-to-date be strengthened, as explained in the reasons for and rapid analyses of more than one area. On the consideration of the Commission communication other, they provide the chance of using current on independence32. It was only recently that data available elsewhere without significant Eurostat put up so-called negative priorities for extra efforts in official statistics. As regards discussion, and in a rather selective manner. An economic statistics, respondents could be relieved area-wide and complete review of the European considerably without causing notable reductions statistical programme should become a permanent of the statistical data supply if e.g. a comprehensive task, as it has been in Germany for many years set of statistical data obtained from business now.33 accounting was combined with a thoroughly designed survey programme of primary statistics. Re (2) Efficient work flows Such an approach may even imply the chance for The bodies of official statistics continuously the bodies of official statistics to extend their have to deal with the question how the statistical programme, i.e. integrate new variables into their information required can be obtained at still programme provided that the standards applying lower cost and with a further reduced burden to official statistics are met. on enterprises and households. Here the new

31 See also Radermacher, W., Weisbrod, J., Asef, D.: Bedarf, Qualität, Belastung: Optimierung als Interessensausgleich, in Wirtschaft und Statistik 11/2004, p. 1237 et seq. 32 Compare Commission Communication on independence (l.c.), No 7 33 See also the contribution of Hahlen, J.: Amtliche Statistik zwischen „Schlankem Staat“ und „Informationsgesellschaft“, in Wirtschaft und Statistik 2/1998 p. 97 et seq.

KNOWLEDGE ECONOMY: PANEL DISCUSSION 255 information and communication technologies probably be content if he gets the requested data provide useful starting points for a data flow rapidly and without problems. If the results are without switches between media: starting with the presented after individual surveys – as has been data supplier, the statistical data shall be processed the practice in the last decades – or if a rather automatically and passed on electronically thematic focus should be opted for remains to be between the different processing stages. These shown by future developments. preconditions apply both to data deliveries of the national statistical institutes to Eurostat and to Due to subject areas such as information society the relations between respondents and national or knowledge economy, statisticians increasingly statistical institutes. prepare tables which aggregate and analyse results of several sources of surveys as for Re (3) Modern and user-friendly information instance in the case of structural indicators. The supply publication of cross section products points in In a time of ever shorter product and innovation the same direction and meets with great interest cycles, information becomes obsolete more and on the part of users. Thus Statistics Finland and more rapidly. It is therefore particularly important the Federal Statistical Office issued cross section for the bodies of official statistics in their role publications several years ago already which as information providers in an increasingly compile information on the information society interlinked economy to improve the topicality from all areas of statistics.35 Eurostat and the of their data supply. In the common system of OECD also issued publications on the subject United Europe, official statistics are the result which cover all areas.36 of a co-operation between the national statistical institutes and the European Union based on It is however questionable if the information the division of labour. For European statistical available is sufficient for the casual user or if the results to be up-to-date, all data suppliers have hurried customer can be expected to use it in the to report their results to Eurostat in time and age of an increasingly personalised information with the agreed quality. An important instrument supply. The basic data supply provided free of to identify timing difficulties making it possible charge should be supplemented by individualised to initiate appropriate countermeasures at the advisory services. This is the only way to avoid the earliest time is an adequate early-warning possible overloading of a user, to recognize and system.34 anticipate the danger of wrong decisions. These measures which in the end build acceptance will Users have access to a multitude of statistical contribute considerably to official statistics being information they may obtain mostly free of charge able to secure its leading position and extend it in or against a small fee. The experienced user will the future.

34 Comp. (l.c.) p. 100 35 See Statistics Finland (1997): “On the road to the Finnish Information Society”, Statistics Finland (1999): “On the road to the Finnish Information Society II”, Federal Statistical Office “Im Blickpunkt - Die Informationsgesellschaft” 2002; a comprehensive overview is contained also in the publication of the Federal Statistical Office: IKT in Deutschland, 1995 bis 2003, Wiesbaden, September 2004 36 See, among others, Eurostat (2001): “Information Society – Statistics Pocketbook“, OECD (2000): “Measuring the ICT Sector”

256 KNOWLEDGE ECONOMY: PANEL DISCUSSION Report on the Panel Discussion

Chair: Marie Bohatá, Eurostat

Walter Radermacher gave a national statistical of human capital as a factor input. He called for institute’s perspective of the requirements needed Eurostat and NSIs to compile and publish integrated to adapt a statistical system in order to measure national accounts and labour accounts. The data the Knowledge Economy effectively. He felt that should already be available so there is no need for major changes to statistical infrastructure will be extra surveys. As a producer, he sees potential for necessary, mainly concerning classifications and improvements due to ICT advances, particularly registers. Concerning the former, many of the with regard to dissemination. He cited as an example aspects of the Knowledge Economy are cross- the fact that central banks and ECB access the same sectional (e.g. ICT, innovation, globalisation) database for their own respective web sites. that do not fit into the traditional NACE structure. In addition there needs to be a new structure Enrico Giovannini challenged the idea that the of business registers: statistics on this topic statistical community is actively participating in must be integrated into mainstream business the Knowledge Economy: this can only happen statistics. In addition an important requirement if information it provides changes the knowledge will be for micro-data for use in longitudinal of the user. He cited a study in the United States analysis. The extra work will require that other which showed that more people use ideology as programmes should be dropped. Progress can be a basis for making decisions than evidence. He achieved through centres of excellence, quality proposed a joint Eurostat-OECD study on what management and the use of new technology. people know about their countries and how official statistics influences such a knowledge. Michel Glaude saw no great difference between the requirements from the NSI’s point of view Fred Gault remarked that whilst there are and Eurostat’s, in fact there are similarities satisfactory frameworks to measure the between Germany’s federal statistical system and characteristics of the actors, more needs to be done the European Statistical System. He noted major concerning the linkages and interactions between progress in this domain over the last 15 years. them. Frameworks such as the SNA and the Oslo Nevertheless, there will be the need for some input manual can be used to provide information to harmonisation. The remaining obstacles faced measure any economic theory. Since financing by the European Statistical System to produce remains a problem, some NSIs are turning to high quality and timely statistics are resource funds from outside traditional sources. constraints and barriers due to infrastructure and historical differences. Concerning the former, he Radek Maly stated that the measurement of social noted the possibility for Eurostat to finance NSIs capital has received a great deal of attention through grants for pilot studies. over the past decade, but the development of the concepts is still at an early stage. This has a Steven Keuning stated that the ECB is in a special consequent knock-on effect on the measurement. position as both a user and producer of statistics. It is difficult to compare across countries because He felt that the current framework of national of cultural and historical differences. He called accounting standards lacks the important element for more research in this area.

KNOWLEDGE ECONOMY: SUMMARY OF PANEL DISCUSSION 257 In response to questions, it was agreed that it is Commission DGs are reflecting on an assessment difficult to measure the effects of the Knowledge of adult skills in cooperation with OECD, but the Economy in industries where such costs are precise objectives of the project have still to be fixed. made clear.

Following a suggestion that more research The need to develop further work on social capital be undertaken on the movements of people was stressed. It had been discussed extensively with PhDs, it was noted that OECD, Eurostat in the Sienna group, but there had been no real and UNESCO currently organise a survey on operational conclusions, maybe now is the time the career path of PhDs. Eurostat and other for tying to implement it.

258 KNOWLEDGE ECONOMY: SUMMARY OF PANEL DISCUSSION Participants List ÅKERBLOM Mikael BOHLIN Anders DEMUNTER Christophe Helsinki, Finland Luxembourg Luxembourg

AKTAS Mehmet BØHN Birgitta DESCY Pascaline Ankara, Turkey Oslo, Norway Thessaloniki, Greece

ALAJÄÄSKÖ Pekka BROOKE Matthew DESURMONT Arnaud Luxembourg Brussels, Belgium Brussels, Belgium

AL KAFRI Saleh BUCHOW Hartmut DÍAZ MUÑOZ Pedro Ramallah, Palestine Luxembourg Luxembourg

ALLEGREZZA Serge BUJNOWSKA Aleksandra DICKEN Sandra Luxembourg Luxembourg Berlin, Germany

ANDRÉN Birgitta BURJA Janja DIXON Matthew Luxembourg Ljubljana, Slovenia London, United Kingdom

ARUNDEL Anthony CALIZZANI Cristina DROP Jurand Maastricht, Netherlands Luxembourg Brussels, Belgium

AUFRANT Marc CASALI Simone DUPONT Franck Paris, France Luxembourg Brussels, Belgium

AWAD Ola CAUNE Agris EKELAND Anders Ramallah, Palestine Riga, Latvia Oslo, Norway

BADESCU Mircea CLAYTON Tony EVERAERS Pieter Ispra, Italy London, United Kingdom Luxembourg

BAIGORRI Antonio COSTA Pierre-Valentin FANDEL Jean-Claude Luxembourg Brussels, Belgium Luxembourg

BALEA Virginia CZESANY Slavoj FELIX Bernard Bucharest, Romania Prague, Czech Republic Luxembourg

BECK Marta DANIELS Martin FERREIRA M. Luisa Luxembourg Stockholm, Sweden Luxembourg

BEHMANE Maranda DAUTEL Vincent FINNBJÖRNSSON Thorvald Riga, Latvia Differdange, Luxembourg Reykjavik, Iceland

BLAIR Sheena DE PANIZZA Andrea FISCHER Jakub Luxembourg Rome, Italy Prague, Czech Republic

BLOCH Carter DE SMEDT Patrick FISCHER-KOTTENSTEDE Jens Aarhus, Denmark Brussels, Belgium Wiesbaden, Germany

BLOMQVIST Irja DEFAYS Daniel FOYN Frank Helsinki, Finland Luxembourg Oslo, Norway

BOHATÁ Marie DEISS Richard FRANK Simona Luxembourg Brussels, Belgium Luxembourg

KNOWLEDGE ECONOMY 259 FREY Luigi HEALY Thomas KIMMEL Silja Rome, Italy Dublin, Ireland Tartu, Estonia

GALASSO Giovanna HEINLO Aavo KNAUTH Bettina Rome, Italy Tallinn, Estonia Luxembourg

GAULT Fred HELLSING Eric KORTE Werner B. Ottawa, Canada Stockholm, Sweden Bonn, Germany

GHERGUT Dan Ion HISCOCK Val LANGERS Jean Bucharest, Romania London, Great Britain Luxembourg

GIEHL Peter HODNE Tor Eigil LAPENIENE Vilija Frankfurt am Main, Germany Brussels, Belgium Vilnius, Lithuania

GIOVANNINI Enrico HÖLSCHER Michael LAUWERIJS Nicole Paris, France Wittenberg, Germany Luxembourg

GIUSTA Paolo HRAJNOHA Ludeˇk LEDOUX Laure Brussels, Belgium Prague, Czech Republic Luxembourg

GLAUDE Michel HULIK Vladimir LEIB Karolina Luxembourg Prague, Czech Republic Brussels, Belgium

GOENAGA BELDARRAIN Xabier HULT Merja LEIHARDT Mikkel Brussels, Belgium Luxembourg Copenhagen, Denmark

GÖTZFRIED August IBÁÑEZ MILLA Jesús LIEPINA Ilga Luxembourg Madrid, Spain Riga, Latvia

GRAF PÜCKLER Botho JANKOVIC´ Miroslav LEMAN Stephen Köln, Germany Belgrade, Serbia and Montenegro Sheffield, United Kingdom

GRUBER Harald JEAVONS Mark LIPINSKA Patrycja Luxembourg Luxembourg Thessaloniki, Greece

GUARDA-RAUCHS Alexandra JOHANSEN Jens LIPOVŠEK Brigita Luxembourg Torino, Italy Ljubljana, Slovenia

HAGÉN Hans-Olof KÁDÁR-FÜLÖP Judit LUCTKENS Evelyne Stockholm, Sweden Budapest, Hungary Windhof, Luxembourg

HAGSTEN Eva KADERÁBKOVÁ Anna LUMIO Martti Stockholm, Sweden Ústí nad Labem, Czech Republic Luxembourg

HAHN Martina KAISER Peter LUNDØ Martin Luxembourg Bremen, Germany København, Denmark

HANREICH Günther KALENGA Roger MALÝ Radek Luxembourg Brussels, Belgium Brussels, Belgium

HANSEN Wendy KETTNER Anja MANCINI Chiara Maastricht, Netherlands Nürnberg, Germany Bologna, Italy

HASSAN Emmanuel KEUNING Steven MÅNSSON Helle Brussels, Belgium Frankfurt am Main, Germany København, Denmark

260 KNOWLEDGE ECONOMY MERCY Jean-Louis PARJO Lea ROBSON Stephanie Luxembourg Helsinki, Finland London, United Kingdom

MESSMANN Karl PAULUSSEN Steve ROGERS Beatrice Vienna, Austria Ghent, Belgium London, United Kingdom

MICALI Aurea PEARE Derek ROSE Norman Rome, Italy Lintgen, Luxembourg London, United Kingdom

MOLLING Victor PEKALSKI Tomasz RUETSCHE Erich Luxembourg Warsaw, Poland Rueschlikon, Switzerland

MOORE Sarah PERANI Guilio SALEIRO Emília Ispra, Italy Rome, Italy Lisbon, Portugal

MURRAY Scott PETKOVA Reni SALMI Heikki Montreal, Canada Sofia, Bulgaria Brussels, Belgium

NESTLER Katja PETTERSSON Ingrid SANDARS Patrick Thessaloniki, Greece Stockholm, Sweden Frankfurt, Germany

NILSSON Rolf PILAT Dirk SANTOS Maria João Stockholm, Sweden Paris, France Luxembourg

NÓBREGA Lourenço PILOS Spyridon SARAZIN LOVRECIC Ines Luxembourg Gasperich, Luxembourg Ljubljana, Slovenia

NOIRET Tina POLLARD Mark SCHÄFER Günter Luxembourg Newport, United Kingdom Luxembourg

NOVOTNÁ Edita PSACHAROPOULOS George, SCHIEFER Andreas Bratislava, Slovak Republic Athens, Greece Vienna, Austria

NOWOZYNSKA Anna PRINTSIPAS Athnasios SCHMIDT Pascal Warszawa, Poland Athens, Greece Luxembourg

OBERHAUSEN Josefine PUDLOWSKI Tomasz SCHNORR-BÄCKER Susanne Luxembourg Warsaw, Poland Wiesbaden, Germany

öHMAN Inger RADERMACHER Walter SCHOORENS Yolande Luxembourg Wiesbaden, Germany Strassen, Luxembourg

OSIMO David RAGNARSØN Richard SCHREINER Alette Sevilla, Spain Luxembourg Oslo, Norway

PAASI Marianne RECKTENWALD Joachim SCHUELLER Nathalie Brussels, Belgium Luxembourg Luxembourg

PAJOT Michaël REIS Fernando SCIADAS George Luxembourg Luxembourg Ottawa, Canada

PANGERC PAHERNIK Zvonka REITER Veronika SENAJ Matus Ljubljana, Slovenia Vienna, Austria Bratislava, Slovak Republic

PAPPADÀ Gabriella RICHARDS Adam John SERRADOR Ana Rome, Italy London, United Kingdom Bruxelles, Belgium

KNOWLEDGE ECONOMY 261 SEYBERT Heidi THORBJÖRNSSON Haraldur WAECHTER Gerhard Luxembourg Reykjavík, Iceland Luxembourg

SINDONI Giuseppe TITZ Harald WALDEN Günter Luxembourg Vienna, Austria Bonn, Germany

SKALIOTIS Michail TOENNISSEN Frank WASSINK Eric Luxembourg Bonn, Germany Voorburg, Netherlands

SKARBNIEKS Ilmars TREACY Joe WHITE David Riga, Latvia Cork, Ireland Belgium, Brussels

SKODVIN Ole-Jacob TUIJNMAN Albert WHITWORTH James Oslo, Norway Luxembourg Luxembourg

SLOAN Brian UTVIK Knut WILÉN Håkan Brussels, Belgium Luxembourg Luxembourg

SOLHEIM Leiv VAGUER Christine WILKENING Soren Kongsvinger, Norway Luxembourg Berlin, Germany

SPROGE Linda VALENTE ROSA Maria João WINDING Anne-Marie Luxembourg Lisbon, Portugal Luxembourg

STAUFFER Philippe VAN ARK Bart WINNETT Adrian Neuchâtel, Switzerland Groningen, Netherlands Bath, Great Britain

STENDAHL MORTENSEN Peter VAN RIJN Jacob WINQVIST Karin Aarhus, Denmark The Hague, Netherlands Brussels, Belgium

STOCKMAN Andrew VAN STEEN Jan WRÓBEL Katarzyna Dublin, Ireland Den Haag, Netherlands Warsaw, Poland

STRATH Annelie VERSCHAEREN Frank WURM Nikolaus Stockholm, Sweden Brussels, Belgium Luxembourg

SVENSEN Elin VERVLIET Greta ZIARKO Ward Brussels, Belgium Brussels, Belgium Brussels, Belgium

THIELEN Pierre VIERTL Reinhard ZIDAN Inaya Luxembourg Vienna, Austria Ramallah, Palestine

THILL-DITSCH Germaine VIKSTRÖM Peter ZIGURE Aija Luxembourg Östersund, Sweden Riga, Latvia

THOMASEN Jens VOZAR Ondrej ZUKERSTEINOVÁ Alena København, Denmark Luxembourg Thessaloniki, Greece

262 KNOWLEDGE ECONOMY