Statistics for policymaking: Europe 2020 Charlemagne Building, Brussels 10 & 11 March 2011
1st day – morning (10/03/2011)
Opening of the conference • Walter Radermacher, Director General of Eurostat, European Commission • José Manuel Barroso, President of the European Commission • Janez Poto čnik, Commissioner for Environment, European Commission • Pier Carlo Padoan, Deputy Secretary-General and Chief Economist of the OECD • Elena Flores Gual, Director of “Policy strategy and coordination”, Economic and Financial Affairs DG, European Commission
Plenary session – Europe 2020: statistics for the next decade Chair: Walter Radermacher, Director General of Eurostat, European Commission Key note speeches on Statistics for Europe 2020: Statistics for evidence-based decision making • Didier Reynders, Vice Prime Minister and Minister of Finance of Belgium • Mihály Varga, Parliamentary Secretary, Prime Ministry of Hungary
National experiences with statistics for policy making in Europe • Bertholt Leeftink, Deputy Secretary General, Dutch Ministry of Economic Affairs, Agriculture and Innovation
Statistical agenda for the next decade with regard to using statistics for policymaking outside Europe Chair: Pieter Everaers, Director of “External cooperation, communication and key indicators”, Eurostat, European Commission • John Steven Landefeld, Director of the Bureau of Economic Analysis, US Department of Commerce, USA • Jeni Klugman, Director of Human Development Report Office, United Nations Development Programme (UNDP) - Human Development Index (HDI), USA
1st day – afternoon (10/03/2011)
Sets of indicators for policy making Chair: Marie Bohatá, Deputy Director General of Eurostat, European Commission • Hans Stol, Consultant in ICT-Business Administration-Policy Making, Research Associate University of Tilburg, the Netherlands • Aurel Schubert, Director for General Statistics, European Central Bank, Germany • Andrea Saltelli, Head of Unit of "Econometrics and Applied Statistics", Joint Research Centre, European Commission • Ales Capek, Head of Unit of “Key indicators for European policies”, Eurostat, European Commission
3 parallel sessions on statistics for: Smart growth: innovation indicators, education, digital Europe Chair: Antonio Argüeso, Director of Social and Demographic Statistics, National Statistical Institute (INE), Spain • Mikael Jern, Professor, National Center for Visual Analytics (NCVA) at Linköping University, Sweden • Lucilla Sioli & Michail Skaliotis, respectively Head of Unit of “Economic & statistical analyses”, Information Society and Media DG and Head of Unit of “Information society: tourism”, Eurostat, European Commission • Anders Hoffmann, Director of Entrepreneurship and Innovation Policy, Danish Ministry of Economic and Business Affairs • Matthieu Delescluse, Unit of "Economic analysis and indicators", Research and Innovation DG, European Commission Sustainable growth: indicators for green growth Chair: Pedro Díaz Muñoz, Director of “Sectoral and regional statistics”, Eurostat, European Commission • Robin Miège, Director of “Strategy” , Environment DG, European Commission • Lucio Gussetti, Director of "Consultative works", Committee of the Regions • Catherine Larrieu, Head of the Sustainable Development Delegation, Office of the Commissioner General for Sustainable Development, Ministry of Ecology, Sustainable Development, Transportation and Housing, France • Eric De Brabanter, Department of Environment, Ministry of Sustainable Development and Infrastructures, Luxembourg and Chairman of the OECD Working Party on Environmental Information, Luxembourg Inclusive growth: indicators on poverty and social/economic exclusion Chair: François Lequiller, Director of “National and European accounts”, Eurostat, European Commission • Antonia Carparelli, Head of Unit of “Inclusion, social policy aspects of migration, streamlining of social policies”, Employment DG, European Commission • Sir Anthony B. Atkinson, Professor, Nuffield College, Oxford, United Kingdom • Inna Steinbuka, Director of “Social and information society statistics”, Eurostat, European Commission
2nd day – morning (11/03/2011)
10.50-12.00h Roundtable Chair: Walter Radermacher, Director General of Eurostat, European Commission • Dirk Ahner, Director General of “Regional Policy” DG, European Commission • Marie Bohatá, Deputy Director General of Eurostat, European Commission • Sharon Bowles, Member of the European Parliament for South East England; Chair of the Economic and Monetary Affairs Committee • Enrico Giovannini, President of the National Statistical Institute (ISTAT), Italy • Thomas Wieser, Director General for Economic Policy and Financial Markets, Ministry of Finance, Austria
Conclusions • Walter Radermacher, Director General of Eurostat, European Commission • Olli Rehn, Commissioner for Economic and Monetary Affairs, European Commission
The conference focused on the role of statistics in policymaking, using the Europe 2020 strategy as a case in point. The trends in using statistics for politically highly sensitive purposes and the related complications for statisticians have been discussed.
The aim of the conference was to bring together policymakers and statisticians in a dialogue on how to shape the statistics underpinning policymaking. How should indicators be defined so that they provide the required information? How should policymakers and statisticians cooperate in order to get the best possible outcome on both sides? How can the statisticians ensure that statistics influence policymaking rather than being contaminated by it? How can the current trends in measuring economic, social and environmental developments be reflected in statistics used for policymaking and, in particular, in support of the Europe 2020 strategy? The conference underlined the importance of good interaction and understanding between policy makers and statisticians in order to be able to correctly measure the political targets set.
The participants in the conference were users of these statistics (European and national policymakers, analysts from the European institutions and international organisations, policy advisors responsible for the national coordination of the Europe 2020 strategy, users from media and the academic sphere) and producers of statistics (Eurostat, National Statistical Institutes).
Table of contents
Opening of the conference President José Manuel BARROSO ...... 1 Commissioner Janez POTO ČNIK ...... 6 Pier Carlo PADOAN ...... 10 Elena FLORES ...... 15
Plenary session – Europe 2020: statistics for the next decade Didier REYNDERS ...... 21
National experiences with statistics for policy making in Europe Bertholt LEEFTINK ...... 26
Statistical agenda for the next decade with regard to using statistics for policymaking outside Europe J. Steven LANDEFELD ...... 30 Jeni KLUGMAN ...... 35
Sets of indicators for policy making Hans R. STOL ...... 41 Aurel SCHUBERT ...... 52 Andrea SALTELLI ...... 60 Ales CAPEK ...... 70
Smart growth: innovation indicators, education, digital Europe Mikael JERN, Linnea STENLIDEN ...... 73 Lucilla SIOLI, Michail SKALIOTIS ...... 83 Matthieu DELESCLUSE ...... 95 Antonio ARGÜESO ...... 99
Sustainable growth: indicators for green growth Robin MIEGE ...... 101 Lucio GUSSETTI ...... 104 Catherine LARRIEU ...... 108 Eric DE BRABANTER ...... 112 Pedro DIAZ MUNOZ ...... 118
Inclusive growth: indicators on poverty and social/economic exclusion Antonia CARPARELLI ...... 121 Anthony Barnes ATKINSON ...... 130 Inna STEINBUKA ...... 134 François LEQUILLER ...... 139
Conclusions Commissioner Olli REHN ...... 140
Speech in the opening session President José Manuel BARROSO
Good morning,
It is a pleasure for me to open such an important conference where both policymakers and statisticians meet.
It is indisputable that official statistics do matter. The role of statistics in the social and economic development of our societies has now been officially acknowledged with the first World statistics day celebrated on the 20 th of October 2010, and rightly so.
Statistics also play a key role in communicating our policies. We all enjoy watching diagrams and tables when we read the papers or watch websites. Moreover, modern communication tools, such as blogs and websites, are as hungry for reliable quantitative information as they are hungry for videos: because it captures the essence in a visual and user-friendly way.
Today and tomorrow you will have in-depth discussions on the role of statistics in policy-making using the Europe 2020 strategy as a case in point.
And so intervening at the opening of these series of debates, I thought of just sharing with you a few general remarks from a policymaker's perspective.
First, I would like to say a few words on the current European context .
The economic and financial crisis has wiped out years of economic and social progress. And it has vividly exposed structural vulnerabilities within the economies and societies of individual Member States.
It has highlighted weaknesses in the implementation of our Economic and Monetary Union. And above all it has clearly shown how interdependent our economies became.
Opening Session Page 1
The crisis has amplified the need for strengthened economic coordination and enhanced surveillance in the Euro zone in particular and more generally in the European Union as a whole.
We clearly need an effective framework to deal with macroeconomic imbalances.
We need a common strategic vision on the future of our European social market model in a fast- changing, uncertain and increasingly competitive globalised world.
To address these challenges, the Commission has front-loaded all the measures and put forward new instruments that are to set the agenda for the European policy-making over the coming years: the Europe 2020 strategy for growth and jobs, the reinforced economic governance, including through the European Semester of ex-ante coordination of economic policies in the European Union, - and the regulation and supervision of financial markets.
The Europe 2020 strategy itself is an integrated and coherent approach to support smart, sustainable and inclusive growth rooted in greater coordination of policies at the national and European levels.
The structural reforms that are being implemented as part of the Strategy will contribute to stimulate a broad-based growth, create jobs and allow economies to adjust more readily to changing conditions.
But why are statistics so valuable to policy makers in addressing the crisis and laying sound foundations for an inclusive and sustainable growth?
In a nutshell: implementing the overall response to the crisis will require sound, high quality data and statistical analysis, on which decisions can be based.
As we are increasingly faced with complex questions to answer and decisions to take, we obviously need to know what is going on in our economies and our societies.
The essence of statistics is precisely to concentrate the complexity we face into meaningful and relevant information for the policy makers, but also for the citizens; to ensure transparency of decisions and of decisions making process.
Opening Session Page 2
Transparency and accountability are indeed crucial to foster mutual trust between policy makers, both in Brussels and the capitals of our Member States and the citizens.
There must be not doubt on the European determination to draw a line under the crisis. That is why we need to be providing first a credible assessment of our strengths and weaknesses as well as vulnerabilities. It is only based on such a sound diagnosis that the policy-makers can take decisive actions to address them.
Broader and enhanced surveillance of fiscal policies, but also macroeconomic policies and structural reforms must rely on high quality statistical information.
Statistics are also crucial when setting our targets and using indicators for monitoring and evaluation purposes.
In particular, measuring trends in competitiveness and following the developments in the macroeconomic imbalances within the enhanced economic governance framework will be based on a scoreboard of economic indicators.
The high quality of statistics produced under robust quality management becomes even more important when new enforcement mechanisms are foreseen in case of non-compliance by Member States with agreed targets or policy commitments taken.
Sound statistics are also a condition sine qua non for accurate forecasts and projections that also form an integral part of the recommendations made by the Commission in the framework of the economic policy coordination and fiscal surveillance in the European Union.
Statistical information that the policy-makers receive must be relevant, timely and accurate to best decide on the policy direction and the best course of action to take that should be both ambitious and realistic.
Statistics are, therefore, literally present in all parts of our global response to the crisis. And it is in this overall framework that the Commission has repeated the necessity to grant Eurostat with extended powers in the field of fiscal statistics, which was agreed by the Council in August last year. These powers allow for an in-depth review of upstream data sources used in the excessive deficit procedure.
Opening Session Page 3
But let me tell you here that it is already in 2005 that the Commission has put forward the proposal to strengthen the mandate of Eurostat, which was unfortunately not met with enthusiasm by the Member States at that time. I personally believe that we could have avoided many of the future problems were the extended powers to the European Statistical Authority granted earlier than last year. *** Given that in your discussions you will take the Europe 2020 strategy as a relevant policy case example, let's have now a closer look at it.
Sound statistical analysis underpinned by robust statistics will be key for steering the implementation of the Europe 2020 strategy. As you know, statistics are almost all over the Strategy starting with its headline quantitative targets that embody its overall objectives. Let me recall them for you at the beginning of your conference:
Europe 2020 is about boosting growth and employment through the following objectives at the European level: • raising the employment rate from the current 69% to 75%; • boosting spending on R&D to 3% of GDP from the current 2%; • reducing the school drop out to less than 10% from the current 15% and increase the share of people in their early 30s with a university degree or equivalent from 31% to at least 40%; • achieving the EU's 20/20/20 climate change goal; • lifting at least 20 million people out of the risk of poverty and exclusion.
These targets serve as a benchmark for the necessary policy action and at the same time as a communication device. They are also being translated by the Member States into national targets under each of the headings. This allows citizens to know why and how decisions are taken, and follow their implementation. We will only make a verifiable progress with the many initiatives under Europe 2020 if the related policy assessments are grounded in an indicator- based analysis.
Thus the appropriate choice of indicators is key to boost our understanding of the complexity of our diverse societies within the European Union, to better communicate on it, and to better respond to new policy needs as for example with the " GDP and beyond " initiative to include measurement of well being.
Opening Session Page 4
That being said, certain pre-conditions have to be respected for producing statistics of the highest quality.
Statistics are compiled within the European Statistical System based on an extended cooperation between the National Statistical Institutes across the Union and between the Institutes and Eurostat.
The data used by the Commission reflect both, the Eurostat competence but also the quality of data notified by the Member States. In case the latter were of insufficient quality, this would have a negative impact on the quality of the European statistics but would also put at stake the credibility of our policy-making.
We have seen in the context of the recent economic and financial crisis that the weaknesses in the quality of upstream public accounting data were compounded by the weaknesses in the statistical governance arrangements in place at the national level. This in turn influenced the way we were seeing the economic reality in some Member States at that particular moment.
Further strengthening of the governance of the European Statistical System will be key in this respect. The Commission is currently reflecting on ways to achieve it and we will soon present some concrete proposals in a forthcoming Communication. Obviously designing and monitoring statistical indicators for policy purposes requires close cooperation between policymakers and statisticians at national and at European levels. Policy- makers must provide the right incentives for and guarantee the professional independence of statisticians, strengthening their impartiality and objectivity. This is even more relevant in the European context to ensure trust between the Member States and between the Member States and the Commission.
Given that statistics are expected to assume an even stronger role in our democracies, this is the right moment to reflect on the challenges in the use of statistics in the policy-making.
Your conference, organised in a very timely manner, is the appropriate forum to foster such a discussion.
I wish you very fruitful deliberations and am looking forward to reading your conclusions.
I thank you for your attention.
Opening Session Page 5
Speech in the opening session Commissioner Janez POTO ČNIK
• I would like to welcome Eurostat initiative to organise this conference that brings together EU and Member States politicians and statisticians.
• Europe 2020 strategy has a clear environmental dimension and identifies the new paradigm of green growth. The Strategy and the Flagship Initiative to a Resource-efficient Europe provide the conceptual framework to define the quality of growth. Our objective is to ensure that resource-efficiency considerations are integrated in all policies. This will steer businesses, investors and consumers towards a systemic change which is necessary to achieve green and competitive economy.
• Following the adoption of the Strategy, a new governance cycle has been agreed. It introduces a number of new elements. One of them is a reinforced integrated surveillance, aimed at monitoring Member States efforts in implementing Strategy goals.
• Europe 2020 strategy has defined EU targets which will be monitored by the headline indicators, and flagship initiatives which will be monitored by indicators currently being developed by the Commission services.
• At the same time DG Environment is working on a roadmap on "Resource efficient Europe" which will include indicators and probably targets to monitor progress in resource efficiency. We are working on a set of resource efficiency indicators. They will help us to communicate the need to improve resource efficiency substantially, will track our record against the objectives of the flagship initiative, and will help us to find out where actions towards more efficient and sustainable resources use are taken most effectively.
• In order for the Commission to design, implement and monitor EU policies we need reliable data and information, and a knowledge base approach is fundamental to ensure that we set policies to tackle EU problems and challenges.
• Eurostat and the European Statistical System are one key source of data on which EU policies are built on. When it comes to environmental issues, the European Environment Agency and JRC are also much contributing.
• I'm an economist and was Head of Macroeconomic Institute of Slovenia, so I know about the importance of economic indicators such as Gross Domestic Product, Net Domestic Product, Net National Income. However to achieve truly inclusive and sustainable growth in the
Opening Session Page 6
sense of Europe 2020 we need social and environmental statistics and indicators on the same level with economic statistics, concerning scope, details and timeliness. This is true in particular seeing the increasing dependency of businesses on natural resources, including eco-system services. Therefore I fully support the statement by Mr. Radermacher last year that in the medium to long term there should be as many resources being devoted to environmental and social accounts as classical national accounts. We are working closely with Eurostat and other Commission services to narrow down this knowledge gap.
• Nowadays we have a significant amount of data available, but official statistics cannot always meet the EU data needs. This is because either they do not cover all relevant issues (such as biodiversity) or because they are too old-often 2 or 3 years. Policy makers need fresh information that reflects current problems.
• We should therefore look for additional sources of information and open the door to research institutes business organisations, space agencies and think tanks, and we invite the European Statistical System to also start a dialogue with data sources outside the statistical community.
• I know that resources in the National Statistical Institutes have been reduced e.g. in the wake of the crisis and that, social and environmental data are not considered as essential as economic and financial data. But statistics are an investment in the knowledge base, they are not administrative burden. Good statistics are cheaper than wrong decisions.
• We could think about counting the expenses for statistics together with RTD expenses as they inform us about the state of reality which is the starting point for further research and development.
• In the coming years we need to develop indicators to monitor progress on Green growth, more specifically on the Environmental good and service sector, on Green Public Procurement and on eco-innovation. We also need a common understanding how to measure subsidies that have harmful side effects on the environment and there in the medium to long term on our natural resource base. With the final aim to reshape those subsidies so that the social and economic objectives are met without the harmful side effects. And I am happy to know that some preliminary information exists, even if major progress is needed.
• We need to better measure our natural capital, its eco-system services, in physical terms and – as preliminarily tested by TEEB – in monetary terms. We expect from companies via our accounting rules to set up very detailed asset accounts to be sure that this company is
Opening Session Page 7
not at risk of collapse, that it is actually increasing wealth and not just burning money. However, our economies as a whole we have currently to run without such accounts of our natural capital and man-made wealth. Certainly such accounts would provide very valuable knowledge base for policy making to help reach the sustainable growth objective of Europe 2020.
• To get the prices for natural resources and environmental services right we need to know about extent of the existing externalities – positive and negative ones.
• Figures and indicators on emerging issues need to be good enough, but not perfect. Roughly right is better than no information at all.
• I know about the hesitation on the side of the statistical system to enter the business of valuation or monetisation of environmental damages and eco-system services. However, as the Stern report and TEEB have shown, this is vital information for policy making.
• If the statistical offices are not the right place to provide this vital information we need to discuss who would be best placed to do so.
• With our focus on measuring green growth and resource efficiency we will not forget or neglect measuring the final objectives we strive for: prosperity, quality of life and well-being for citizens while preserving the natural capital of our planet. Measuring these issues in a comprehensive way is a key mandate of the Commission's beyond GDP agenda which I have the pleasure to coordinate together with our host today, my fellow Commissioner Olli Rehn.
• From the adoption of the Communication "GDP and beyond" in 2009 and the announced report on its implementation in 2012, we are now roughly half way through the 3 years. With the first deliverable of the EC/Eurostat on environmental accounts currently debated in EP and Council.
• For designing and assessing policies detailed data are vital. For communication and political debate headline figures are also vital. Indices and small dashboards are very useful when we need to assess complex issues like environmental pressures on our environment, resource use and biodiversity. We have currently no accepted comprehensive overall figure for the environment in the concert of GDP, inflation and unemployment rate. Under the beyond GDP road map, we are developing a composite index on environmental pressures and a Sustainable Development Scoreboard.
Opening Session Page 8
• I'm very glad to see that the initiative coming from the European Commission in 2007 is gaining more and more momentum. International organisations such as OECD and World Bank are fostering measurement of societal progress. There are also important new initiatives in Member States, such as the well-being indicators in UK and a Parliamentary Committee in Germany looking into the relation of growth, prosperity and quality of life. Also with the mandate to develop indicators. I'm looking forward to the results of these efforts.
• I would be glad to co-operate with Eurostat and the European Statistical System, with the EEA and the JRC, with business and civil society organisations, with environmental and space agencies and others in view of producing a regular monitoring system on green growth and resource efficiency and societal progress beyond GDP.
Opening Session Page 9
Better statistics, better policies, better lives Pier Carlo PADOAN
President Barroso, Commissioner Poto čnik, Ms. Flores Gual, Mr. Radermacher,
Ladies and Gentlemen:
I am very pleased to attend this Conference on “Statistics for policymaking: Europe 2020.” Your kind invitation to speak at this opening session attests to the excellent relations between the European Union and the OECD.
Today, 21 EU Member States are also members of the OECD, which now gathers 34 countries overall. This obviously implies close ties between the EU and the OECD. We have common purposes in terms of economic stability, sustainable growth, structural reforms and social cohesion. Our strategic partnership has been developing over the years through constructive discussions, joint initiatives and day-to-day co-operation between the various general Directorates of the Commission and the OECD Secretariat.
This year, the OECD is celebrating its 50 th Anniversary. This is the occasion to restate the Organisation’s mission to develop best standards in public policies in order to improve the lives of citizens in OECD countries and around the world -- in other words to promote better policies for better lives . Our work encompasses nearly all areas of public policy, from macroeconomic to entrepreneurship, from fight against corruption and tax heavens to biotechnology, from health to competition, from education to development aid, from renewable energy to migration, international trade, innovation and corporate governance. In all these areas, we measure and analyse, we compare countries’ performances, we set standards and identify best practices.
Statistics are a central pillar of this policy work: they provide us with the essential foundation for our analyses, forecasting and benchmarking work; they are at the heart of our evidence-based policy recommendations. Today, I would like to highlight the key role of statistics in our response to the crisis and, more broadly, in our work on assessing the well-being of people and fostering the progress of societies.
Enhancing statistics: a major lesson from the crisis
The world is slowly recovering from the worst financial, economic and social crisis of the past seventy years. This is time for us to take stock of a number of important lessons that we have learned from the pernicious dynamics that led the world economy close to the abyss.
Opening Session Page 10
The crisis was not caused by a lack of statistics. Many indicators were available that pointed to stress and unsustainable tensions in our economies but unfortunately these were largely overlooked by policymakers and the general public. Nevertheless, the crisis did reveal some important data gaps and underscored the importance of timely and internationally comparable information. In particular, the crisis has heightened the need to examine closely, systematically and regularly the development of imbalances in the economy, both globally and locally. This includes developing better measures of economic slack (such as the output gap), balance sheets (of households, businesses, the financial and the government sectors), current and capital account positions, prices of financial assets, real estate and commodities, exchange rates and various inflation measures. One main reason why the crisis took us by surprise is precisely because the extent of the imbalances, and their interactions, were underestimated. Stronger surveillance efforts are thus required and this entails a strong demand for relevant statistics.
In 2008, the G-20 leaders requested that the IMF, the Financial Stability Board and other international organisations explore some of the data gaps and provide proposals to address them. The OECD has been actively involved in this effort. Our contribution has focused on the collection and analysis of quarterly sector accounts. These include timely accounts for government, financial and non-financial corporations, and households. While such data are not yet fully developed in all countries, their elaboration is critical to better understand and quickly capture issues such as household and corporation debt, asset revaluation, exposure of sectors to financial risks and the interconnectedness among sectors within an economy and across borders.
The OECD is conducting this work in close collaboration with the European Statistical System (Eurostat and the ECB). I am pleased to see that our two Organisations are increasingly sharing their data, analyses and expertise in the area of key economic and financial statistics.
But today we also observe the social consequences of the crisis: millions of people have lost their jobs, their homes, their pensions. The pace of the global economic recovery is still too slow to reintegrate many of them into the labour market and unemployment remains at unacceptable levels in the face of high public debt levels.
In this context, the massive protests of unemployed youth in some low and middle income countries are a strong warning signal. While the political environment in our countries is of course different, our weak labour markets can be a source of major social conflicts. We are currently putting a lot of efforts to develop statistics to better monitor global economic and financial developments. We also need to improve our statistical base to better monitor the social consequences of the crisis which in many countries will take a long time to unfold. For
Opening Session Page 11
example, a few weeks ago, an OECD report on Jobs for Youth showed that in most of our countries young people have been among the most affected by the crisis at a time when they were already more than twice as likely to be unemployed as the average worker. And an upcoming OECD report will also show that in many countries, households’ poverty rates have increased during the crisis. More detailed micro-data are needed to analyse the situation of at- risk and disadvantaged groups in the population and help design appropriate policies.
Beyond GDP: a stronger focus on people’s well-being
More fundamentally, the crisis has not just highlighted areas where our statistical capacity remains deficient, but it has also undermined the confidence of people in markets, public policies.and official statistics. Restoring confidence requires providing evidence of what matters to people’s everyday lives.
The idea of measuring well-being has been recognized as a priority by top European political Leaders. Among them are President Sarkozy of France, Chancellor Merkel of Germany, Prime Minister Cameron of the UK and President Türk of Slovenia and it also figures highly in the EU2020 agenda. Beyond Europe, President Lee of Korea, and the Prime Minister of Japan, Mr Kan, have also put well-being and progress at the top of their political agendas.
Much of the statistics and indicators that we routinely produce are not looking at what truly matters to people. There is today a consensus that we put too much emphasis on measuring economic production – principally through gross domestic product – and not enough on assessing people’s well-being. Too many important policy decisions are still being taken with GDP per capita as the main measurement rod.
Of course, GDP has to remain at the centre of our statistical systems. But we need to complement it with indicators that measure well-being more broadly. In addressing this challenge, we can build on the rich body of data and indicators that the OECD has produced over the years in such fields as education, health, innovation, the environment and climate change.
We can also rely on the accumulated knowledge and research that has taken place over the past ten years or so. The OECD has in fact led the international reflection on the measurement of progress. We have organised three World Forums in Palermo in 2004, Istanbul in 2007 and Busan in 2009. These events have allowed us to engage in discussions about relevant measures of well-being and progress with policy makers, statisticians, scientists, economic and social actors from more than 130 countries. The 4 th World Forum will take place in New Delhi in October 2012.
Opening Session Page 12
The research agenda has also benefitted greatly from the recommendations made by the Commission on the Measurement of Economic Performance and Social Progress , established by the French President Nicolas Sarkozy, also known as the Stiglitz, Sen, Fitoussi Commission.
Building on all these initiatives, the OECD is now stepping up its statistical work on three areas that matter for well-being and progress: material living conditions, quality of life and sustainability.
Our first effort is directed at better measuring households’ material conditions . What happens at the level of the entire economy does not necessarily tell us what happens to households’ purchasing power. For example, over the last decade, real GDP in my own country Italy grew by about 1.6% per year, whereas the real income of Italian households only rose by half that amount. Large discrepancies can also be observed in many other countries, which points to the importance of understanding and explaining why the benefits from growth have not been passed onto households.
Equally important, standard statistics on economic production do not account for the many important services that households produce at home. These include, for example, child care, cooking, care for the elderly, volunteer activities. Such activities could add between 20% and 40% to our conventional measures of GDP. Taking into account these services would give a very different picture of economic performances across countries and highlight the importance for policies of meeting new challenges, such as that of reconciling work and family life.
Most people would also agree that there is more to life than money. Indicators of quality of life should reflect health conditions, competencies, the time people spent commuting or with their families, their housing conditions, their local environment, their political participation, social connections, and the various risks that shape their feeling of security. Importantly, for each of these dimensions, we need to capture various forms of inequality in the population. We are also working on the development of guidelines on internationally comparable subjective well-being indicators. This is a huge area of work, where gaps between the statistics available and what is needed are even larger than those for economic statistics.
Lastly, can current well-being be sustained over time ? How can we ensure that our well-being today is not achieved at the expense of our children’s well-being? We are now attempting to develop better metrics of how our production and consumption patterns impact on the environment, both domestically and globally, through the development of indicators in the context of the OECD Green Growth Strategy. Taking a broader perspective, the concern on sustainability also requires developing better measures of human and social capital, and of knowledge and innovation.
Opening Session Page 13
As you can see, we have an ambitious statistical work programme ahead of us. We are working in very close collaboration with the EU on this, in particular through joint work on improved measurement of income inequalities and direct participation in the Eurostat/INSEE Sponsorship which is following-up on the recommendations of the Stiglitz-Sen-Fitoussi report at the European level. This will feed directly into the indicators work underpinning the EU2020 agenda.
On the occasion of our 50 th Anniversary this year, we will issue a report entitled How is life? , which will present for the first time a set of comparable indicators on material living conditions and quality of life for OECD countries. These indicators will mainly reflect what data currently exist but over time, as our and others’ work deliver their results, these will be incorporated in future editions of the publication.
Ladies and Gentlemen:
Let me leave you with a key message: developing better statistics is not an end in itself. It is a means to improve policies that affect the functioning of our economic system and the well-being of people living in it.
There are several economic, societal and environmental goals that the EU and the OECD share. These goals determine what we have to measure, how measurements have to be transformed into useful knowledge, and how this knowledge has to inform policy design and decision making.
I hope that this Conference will stimulate ideas on the statistical agenda for the next decade. The OECD stands ready to work closely with the European Commission to enhance the statistical basis that underpins our policies.
Thank you.
Opening Session Page 14
Speech in the opening session – Outline Elena FLORES
Summary
In her introductory speech, Elena will present the main recent policy initiative proposed for the European Union in response to the economic and financial crisis, highlighting their statistical implications.
Europe 2020, the enhanced economic governance and the coordinating framework called "European semester" contains all references to statistical indicators.
These initiatives show how the use of specific statistical indicators helps not only in defining the policy targets, but also in assessing the policies developments and in communicating progress to the general public.
At the same time, the use of statistical indicators in the conception, implementation and monitoring of policies requires data of high quality and, as a consequence, a great effort from statisticians.
Structure
1. Introduction
2. Europe 2020
3. Enhanced economic governance
4. EU semester
5. Conclusions
1. Intro
• The main initiatives proposed by the EU Commission in response to the economic and financial crisis which hit European citizens are: the Europe 2020 strategy and the enhanced economic governance, as well as the enhanced economic policy surveillance within the coordinating framework called "European semester".
• Each of these initiatives makes references to official statistics, in all their phases: conception, implementation, monitoring and assessment of policies.
Opening Session Page 15
• Translating the political objectives into statistical indicators helps to better define them, as well as to identify possible trade-offs.
• Furthermore, statistics constitute a powerful communication tool: they play a crucial role in improving transparency and understanding of the policy performance and promoting European citizens' joint ownership of the initiatives.
• Because of their relevance, statistical indicators used in the policy context need to be of high quality and easily accessible.
2. Europe 2020
• The Europe 2020 strategy put forward by the Commission sets out a vision of Europe's social market economy for the 21st century as a smart, sustainable and inclusive economy delivering high levels of employment, productivity and social cohesion.
• The Union has set five ambitious objectives - on employment, innovation, education, social inclusion and climate/energy - to be reached by 2020. The objectives are defined in very concrete terms, measured by statistical indicators:
1. Employment:
75% of people aged 20 to 64 to be employed
2. Research and development:
3% of EU GDP to be invested in R&D/Innovation (public and private combined)
3. Climate change / energy
20% greenhouse gas emissions lower than 1990
20% of energy from renewables
20% increase in energy efficiency
4. Education
Reducing school drop-out rates below 10%
At least 40% of 30-34 years old completing level education
5. Poverty/social cohesion:
At least 20 million fewer people in or at risk of poverty and social exclusion
• Each Member State adopts its own national targets in each of these areas, reflecting different starting situations and national circumstances.
Opening Session Page 16
• At national level, therefore, the interaction between policy makers and statisticians in the definition of the national targets is necessary.
3. Reinforced EU economic governance: budgetary, macro-economic and structural surveillance
• On 29 September 2011, the Commission presented a legislative package on Economic Governance, whose core elements are:
o the reinforcement of the Stability and Growth Pact (SGP),
o the extension of the surveillance to macroeconomic imbalances,
o the setting up of a range of enforcement mechanisms, including sanctions, which would kick in at an early stage.
• The elements of this approach were outlined in May 2010 and a concrete "toolbox" was presented in June. This was complemented by preparatory work and consultations with a broad range of stakeholders, in particular by the Task Force on Economic Governance chaired by President of the European Council Herman Van Rompuy. The Commission, the European Parliament and the Council have agreed that the Economic Governance Package should be adopted by summer 2011.
• The Economic governance package has several statistical implications, which I would like to briefly highlight here.
3.1 Fiscal side :
• 3 regulations reinforcing the Stability and growth pact.
o One regulation introduces new elements regarding the analysis of medium term debt position, which call for extended statistical information.
o For example: maturity structure and currency denomination of the debt; accumulated reserves and other government assets; guarantees, notably linked to financial sector.
• One directive for adequate national budgetary frameworks
o This directive refers to the quality of "upstream public finance data" which are used to compile EDP statistics. It aims at reforming national public accounting and budgetary processes and making them more harmonised (therefore comparable) among Member States.
Opening Session Page 17
o Improved information on public spending, also at local level, will be available for the surveillance tasks assigned to the Commission.
3.2 Macroeconomic side: the excessive imbalance procedure
• One Regulation on the prevention and correction of macroeconomic imbalances. This regulation introduces an alert mechanism based on a scoreboard composed of a limited number of statistical indicators. For countries whose economic imbalances triggered the alert mechanism, an in-depth analysis of their origins will be carried out.
o Both the alert mechanism and the in-depth analysis will be based on statistical indicators.
o Neither the scoreboard nor the detailed indicators used in the in-depth analysis will be included in legal acts.
o The translation of policy objectives into the economic concept and then to the corresponding statistical indicator is not as straight as one can imagine. An example is given by the indicator used for assessing current external balance, where a set of indicators from different statistical sources can be used.
o The statistical indicators to be included in the scoreboard are being the subject of discussions in several fora, with the participation of economic analysts and statisticians: these discussions have shown the importance of the interaction between policy makers, analysts and statisticians. Such interaction will lead to very well grounded choices.
• One Regulation on enforcement measure to correct excessive macro economic imbalances in the euro area (no statistical implications).
• In both the public finance and macroeconomic excessive imbalances contexts, the use of statistical indicators guarantees transparency of the process and provides objectives grounds for policy decisions.
4. The European semester o The Europe 202 strategy and the budgetary and macro-economic policies are coordinated through a new architecture, called the "European semester" and approved by the Member States in September 2010. o This new cycle has several stages:
Opening Session Page 18
• It starts each year in January with the Annual Growth Survey, in which the Commission provides an analysis of the EU economic challenges and policy responses, on the basis of:
the progress on Europe 2020 targets;
a macro-economic report;
the joint employment report
which is to be discussed by Council formations and the European Parliament ahead of the Spring meeting of the European Council in March.
• At the Spring Council, Heads of State and Government identify the main challenges facing the EU and give strategic advice on policies.
• In April, taking this guidance into account, Member States send to the Commission:
their medium-term budgetary strategies through Stability and Convergence Programmes
the National Reform Programmes, including the national targets under the Europe 2020 strategy.
• Based on the Commission's assessment, the Council will issue country-specific guidance. This will be particularly timely for the finalisation of Member States' draft budgets for the following year.
5. Conclusions o The recent initiatives that I briefly presented have shown once again how statistical indicators constitute a fundamental tool to identify, elaborate, monitor and communicate policies. o As such, the statistics used for policy making should be of high quality: comparable, reliable, timely and easily accessible. o Statistics are a public good which help to:
o Provide objective ground for analysis and policy decisions
o guarantee transparency of surveillance and enforcement actions;
o improve "public appropriation" of policies.
Opening Session Page 19
o This important role requires flexibility and adaptation from statisticians, mainly to respond to new demands, e.g. financial data, statistics on public and private sectors, productivity and competitiveness indicators. o A specific effort should also be made with respect to the publication of both the data and the methodologies applied for their production. This will improve the mutual understanding and cooperation between policy-makers, analysts, statisticians and the citizens.
Opening Session Page 20
Didier REYNDERS
Dear President,
Dear Director-General,
Ladies and Gentlemen,
First of all, I would like to thank you for your invitation to attend this conference, during which the role of statistics as a decisive factor in decision-making related to EU 2020 will be studied. First, I will briefly discuss the role and importance of statistics. Next, I will tackle some questions about the role of the GDP. Finally, I will discuss some questions about the other indicators in the EU 2020 Strategy.
If we want statistics to be useful, we must first set quantified objectives. It is only because the EU 2020 Strategy, elaborated by the President of the European Commission, Mr José Manuel Barroso, has set realistic objectives based on figures, that indicators play a key role in this field. Statistics are therefore an integral part of the Europe 2020 Strategy. The headline indicators measure the progress made by the EU and the Member States towards achieving the five headline targets of the strategy.
We have also seen some months ago, at the triggering of the sovereign crisis in the Euro Zone, how much sound and reliable public statistics were of the utmost importance.
Consequently, our State and - more broadly speaking - all stakeholders in our society need statistic indicators when policies are set, but also when they are tested or subjected to a final evaluation. Since our economies have become even more complex, there has been a significantly greater need for transparency, for the development of IT and the improvement of the population’s education, as well as for reliable and harmonized statistical data.
Indicator-based evaluation of policies has become a democratic requirement to ensure the transparency and objectivity of the results obtained by politicians. Democratically, it’s essential for the citizens to have appropriate statistics at their disposal to evaluate the political decisions and, then, the politicians. Adequate statistics are therefore indispensable. The indicators, as set in the EU 2020 Strategy, enable us to measure and evaluate the social and economic facts.
Decision-making should be based on an efficient, comparable and reliable statistical system. The European Statistical System was built up gradually, on the basis of the national institutes and Eurostat, with the objective of providing comparable statistics at EU level. The European Statistical System is efficient and offers high-quality services. In June 2010, the EU Finance
Opening Session Page 21
Ministers drew their conclusions from the Greek crisis, which had seriously harmed the European Statistical System’s credibility, by granting Eurostat auditing powers.
EU 2020 is the new strategy for the European Union to stimulate a smart, sustainable and inclusive growth, which is interrelated and mutually reinforcing. The European Union has set five EU headline targets, on employment, innovation, education, social inclusion and energy, to be reached by 2020.
In the EU 2020 Strategy, the Gross Domestic Product is an essential criterion to measure economic development, and understandably so. The GDP was introduced after the Great Depression of 1929 and is the sum of the added values of all economic activities of a monetary nature.
The GDP is universally recognized and accepted as a key indicator to make comparisons. It also provides a solid basis for sound economic policy decisions. It is an important advantage that the GDP is objective and therefore not open to interpretation. By increasing our growth, the European Union will meet tomorrow’s challenges, mainly the ageing of the population and the risk of declining. An increase in GDP will contribute to full employment and will enable us to guarantee the sustainability of our economic and social model.
However, no indicator is perfect or sufficient in itself to describe the complexity of our society. The GDP is an indicator of our economic activity, but does fully not measure our population’s well-being. The GDP keeps its relevance, but it would be logical to have a set of indicators at our disposal which would complement the more traditional statistics, without actually replacing them. The GDP does not sufficiently take into account the non-market sector, the distribution of income or the over-use of environmental capital. I will tackle this issue later on.
I would like to quote the example of a car’s dashboard. Initially, the driver only got information on the car’s speed. Nowadays, he sees in one glance extra information, such as the oil level, the engine speed, the petrol level or the temperature. The dashboard has to remain user- friendly and sparing with information and it must meet the stakeholders’ pragmatic needs, without causing administrative or bureaucratic overload to the economic and social actors.
I can conclude by saying that we need more than one indicator, but that an overabundance of indicators should be avoided, because it would prevent us from efficiently monitoring our public policies.
I would like emphasize the role which Eurostat has played in this field by anticipating the improvement of measurement at EU level, as well as the work done by the European Commission in its communication « the GDP and beyond ».
Opening Session Page 22
Consequently, the GDP indicator should be complemented to surpass its limitations, which are from now on well-known in academic circles. These limitations were emphasized in the “Stiglitz- Sen-Fitoussi Report” on the Measurement of Economic Performance and Social Progress, which was published in September 2009.
The GDP does not take into account the loss of environmental capital and the use of non- renewable resources. Intergenerational and intra-generational equity is another issue. The distribution of income is not sufficiently taken into account and the intergenerational equity, for instance in the case of pension financing, is not taken into consideration. However, intergenerational equity, by controlling public expenditure, is essential to the quality of our public policies.
The five headline targets include economic, environmental and social indicators. The chosen indicators have several positive dimensions: feasibility by looking at timeliness and coverage and technical soundness, comprising overall accuracy and comparability (over time and across countries).
- The employment rate is the work-related indicator, with the objective of having 75 % of the population aged 20-64 employed. This traditional and solid indicator is better than the unemployment rate, because it offers a better basis for comparison and is less biased. Contrary to the Lisbon Strategy, the lower limit has been raised to 20 years, to better take into account the reality of the labour market. The upper limit was maintained, although legislative changes have been observed in some EU Member States. Even though senior-specific indicators are no longer included, it is essential to concentrate our efforts on keeping employed persons aged 55 and older.
- Innovation is the second headline target, with the objective of investing 3 % of the EU’s GDP in R&D. This traditional indicator is a major driver of economic growth. However, while implementing policies, the efficiency and effectiveness of the - mainly public - expenses should be thoroughly analyzed. The European deficit in R&D investments, if compared to other large economic entities, leads to insufficient growth and a lack of employment creation (the indicators’ objectives are interrelated). This deficit is mainly due to a lack of private investments in R&D. Would it not be better to include the fiscal expenses contributing to innovation or private research, while at the same time improving the conditions of applied research in Europe, as well as the link between universities and spin-offs?
- There are three environmental indicators, namely the greenhouse gas emissions (base year 1990), the share of renewable in gross final energy consumption and the energy intensity of the economic system. These three indicators are complementary, as they measure at the same
Opening Session Page 23
time the absolute emission threshold, the origin of the energy and the energetic efficiency. We can deduce from these indicators our capacity to maintain our standards of living, by checking whether the environmental capital stock will still be passed on to the next generations.
The economic growth and the environmental objectives are not conflicting. As the EU 2020 Strategy points out: « Resource efficient Europe to help decouple economic growth from the use of resources, support the shift toward a low carbon economy ». What we need, is more growth while consuming fewer resources. An increase in energetic efficiency will even lead to more competitiveness and, thus, to more growth.
A rapid data transmission is crucial in this context. Economic data are provided after a few weeks, while it often takes several years to obtain environmental information. I’m perfectly aware of the fact that this demand is very difficult from a technical point of view.
The two indicators related to education and training are of course essential. The share of early school leavers should be under 10 % and at least 40 % of the persons aged 30-34 should have completed a tertiary or equivalent education.
I would like to make two remarks on this subject:
- By selecting the cohort aged 30-34 in 2020, we include persons who have already begun higher education. The policies which are now being implemented will have effect in 2025, or even in 2030, on the cohort aged 30-34.
- I would also like to emphasize the importance of finding a balance between labour supply and demand. There must be a link between university education programmes and labour market needs. As regards the 40 % educated persons aged 30-34, the share of scientists and engineers has to increase significantly and more generally, the employability of students has to be improved. Once they are educated, they will find in the European Union a sufficiently attractive and competitive environment to keep living there.
Finally, the EU 2020 Strategy aims at preventing 20 million persons from being exposed to the risk of poverty and exclusion. With a view to their efficiency and effectiveness, social policies must be quantified and evaluated on the basis of realistic objectives. The traditional poverty rate indicator rather measured inequality in the distribution of income, than poverty itself. By complementing this indicator with the persons living in households with very low work transfers and severely materially deprived persons, we obtain a composite indicator which better measures the poverty risk.
Opening Session Page 24
However, we must make sure to have a simple, comparable and reliable indicator in the end. This indicator anyhow shows that policies related to work programmes and growth strategy are important in the anti-poverty strategy.
Before I conclude, I would like to make some remarks:
• The indicators of the EU 2020 Strategy have to be analyzed in parallel with the ones related to competitiveness and budget (deficit and debt ratio). They will have to be included in the European Semester and in the six proposals of the European Commission to reinforce economic governance by strengthening the Stability and Growth Pact and by improving macro-economic surveillance in the European Union. The budget and competitiveness indicators have to play a key role in the economic, social and environmental roadmap which is being elaborated.
• As the Sofia Memorandum states: “There is a growing societal and political demand to measure progress, well-being and sustainable development in a more comprehensive way”. I agree that statistics will take better account of the complexity of our societies.
• In a strategy such as the EU 2020 Strategy, we need objective, comparable and reliable indicators. Which should not restrain us from making progress in the well-being indicators, based on the citizens’ subjective perception, to measure the quality of life in a more multidimensional way.
• I will underline the importance of good interaction and understanding between policymakers and statisticians in order to be able to correctly measure the political target set.
I can conclude by stating that analysts are always inclined to measure a society on the basis of economic and social statistics: these data are important and revealing, but other statistics, which are much more difficult to obtain, are as important and reveal as much about a society: does a society have more memories than dreams or more dreams than memories? I’m sure that thanks to the EU 2020 Strategy, which must be read in parallel with the European Semester, the Competitiveness Pact, the Small Business Act and the Single European Act, the European Union has more dreams than memories.
Thank you for your attention.
Opening Session Page 25
National experiences with statistics for policy making in Europe Bertholt LEEFTINK
Plenary Session Page 26
Plenary Session Page 27
Plenary Session Page 28
Plenary Session Page 29
The role of statistics in the United States' Economic Future J. Steven LANDEFELD
Plenary Session Page 30
Plenary Session Page 31
Plenary Session Page 32
Plenary Session Page 33
Plenary Session Page 34
Measures of well-being: The HDI and related indices Jeni KLUGMAN
Plenary Session Page 35
Plenary Session Page 36
Plenary Session Page 37
Plenary Session Page 38
Plenary Session Page 39
Plenary Session Page 40
The contribution of statistics to the improvement of transparency in the policymaking process Hans R. STOL
1. Introduction
I start this paper with three statements. These statements are scientifically proven to be correct in my PhD thesis (Stol, 2009). I mention each of the three statements followed by a short explanation.
Statement 1. The policymaking process (PMP) covers the whole process, from decision making up and until the evaluation of the implementation of the policy.
Policymaking (PM) is the process by which governments translate their political vision into programmes and actions to deliver "outcomes"- desired changes in the real world (Cabinet Office, 1999). This definition of policymaking may suggest that policymaking ends with programmes instead of changes in the real world. Policymaking always intends to change aspects or parts of the reality. 1 Therefore, considering the effectiveness of the policymaking process means that we must take into account (1) the process of setting and directing the course of action pursued by government, (2) the realisation of the actions in the real world, (3) the measurement of the results, and (4) the evaluation of the policy.
Statement 2. Transparency inheres verifiable and measurable observations of changes in reality due to policymaking.
A transparent PMP is a process by which a full, accurate and timely disclosure is given on the content of the process. Verifiability of the PMP is the possibility to ascertain the correctness of all the information about the steps in the PMP. The effects of the PMP should be measurable (if this is not the case: verifiability and transparency are out of question). This means that the following situations in reality should at least be measurable: (1) the reality that is relevant for the realisation of the objectives at t+1 versus the expected relevant reality at t, (2) the implementation of the objectives versus the original objectives, (3) the relation and the expected relation between the target of the policymaking process (at t) and its environment versus the actual relation at t+1. Verifiability as well as measurability are preconditions for full, accurate and timely disclosure on the content of the PMP (=transparency).
Statement 3. Transparency in policymaking improves the quality of the policymaking process
Transparency in policymaking is not only important because of the populism: people "force" nowadays their representatives in government to explain clearly (1) what they want, (2) how they want to realize this, (3) how they want to measure the results of their intentions and (4) how they want to explain the results versus their intention. Transparency is also a characteristic of the policymaking process that is closely related to the quality of the policymaking process (Stol, 2009). The objective of policymaking is achieving the desired changes in the real world. So, the quality of policymaking is as high as possible if all the desired changes in the real world are achieved. The effectiveness of the policymaking process is related to its quality and its transparency.
The following sections implicitly build on these statements.
2. A framework for the policy-making process
1 ) Althaus (2007) formulates this as follows: "Public policy is ultimately about achieving objectives. It is a means to an end. Policy is a course of action by government designed to attain certain results."
Sets of indicators for policy making Page 41
This section shows the framework for the PMP (Stol, 2009) 2. Implementation of the framework enables the transparency of the PMP. A framework for the PMP process should support the quality of the policy-making process. Transparency and verifiability in the PMP are necessary to optimize the quality in decision making. This means in the first place that the policy process should be structured so that all steps can be clearly defined. Although people may argue that every individual policy decision is different because of different situations, objectives, and circumstances (see, e.g., the approach by Davies, 2001), there is always (on a certain level of abstraction) the possibility to structure the process (see, e.g., Easton, 1979; Checkland 1999; de Leeuw, 1982a). This is not only useful to transfer knowledge about politics (Dorey, 2005) but also to make the process transparent and measurable. The framework for PM provides the elements and structure for a PMP, so that individual cases can be judged and analyzed, and, even more, can be designed so that the process gains in transparency and verifiability. In the framework, the Evidence-based Policy Making (EBPM) is considered as a controlling system
ENVIRONMENT
JUSTIFICATION: INTENDED INDICATOR VALUES related to LOBBYING INFLUENCING MEASURED INDICATOR VALUES
EVIDENCE BASED INDICATOR INPUT DECISION VALUES MAKING
OUTPUT RESULTS IN
INTENDED OBJECTIVES/ INTENDED MODELLING / OBJECTIVES/ DERIVING INDICATOR- INPUT INDICATOR- ANALYSING INDICATORS VALUES INDICATORS VALUES
DERIVING INPUT
EBPM MEASURES TO MEASURES TO CHANGE CONTROL SYSTEM: CHANGE MEASURED REALITY MEASURED REALITY VALUES ACCORDING VALUES ACCORDING OBJECTIVES OBJECTIVES
RESULTS IN REPRESENTS
REALISING OBSERVATION REALISING OBJECTIVES OF EFFECTS OBJECTIVES
OBJECT SYSTEM OF INTEREST over the part of reality that is the objective of the PMP. We call the part of reality that is the target of the PMP: the Object System of Interest.
Figure 1 : Framework for Policy Making Process
The EBPM control system has three main relations with its environment.
(1) Actions directed toward its environment in order to obtain the best possible policy decision (high quality and acceptance).
2 ) In Stol (2009) this framework is called "Framework for Evidence-based Policy Making", in order to avoid confusion with the Framework of Easton (1979).
Sets of indicators for policy making Page 42
(2) Receiving reactions and lobbies of interested parties in its environment.
(3) Producing justifications towards its environment for the policy measures achieved.
The EBPM control system itself consists of seven interconnected parts.
(1) The Evidence-based Decision Making (EBDM), in which the policy decision is made after interaction with the environment.
(2) The determination of the objectives and the kinds of indicator to be measured to clarify the results of the policy measures compared with the objectives.
(3) The derivation of the values of the indicators as representation for the achievement of the objectives..
(4) The determination of the measures to change the relevant part of reality according to the objectives.
(5) The measurement of the results of the policy measures.
(6) The analysis of the results of the measurement versus the intended values of the indicators.
(7) Producing indicator values as representation of the results achieved by the implementation of the policy measures.
Evidence-based Decision Making is the part in which the politicians arrive at the policy decision. Although each decision-making process is different and its progression depends on many factors, one may recognise similar elements in all of them. In the context of this paper the following three similar elements are relevant.
(1) The policy decisions refer to an Object System of Interest. The complexity of the Object System of Interest may differ and may lead to differences in (a) the scope of the decision and (b) the organisation of the Change Process.
(2) The kinds of decision process can be categorised varying from rational to more or less random. The most appropriate category for a unique decision process depends on its characteristics (routine decision, complexity Object System of Interest, available information for decision making, see also Stol (2009)).
(3) The content of the policy decision refers to a part of the real world. This means that for each policy decision an interaction exists with the environment that will be influenced by the decision (lobbying, interest groups, etc.). This is a factor that is often used by politicians to escape from EBPM. Their argument is that it is never possible to convince every participant in the Society involved in the decision making. A negative side effect of taking too much time for achieving a most satisfying decision is that the Object System of Interest can be changed and the issue under discussion is no longer actual.
Considering the divers options for a decision, information of former outcomes of PMP's plays a role. According to the literature on EBPM this is considered an important asset to achieving the best possible decision.
Sets of indicators for policy making Page 43
In the framework of the PMP we emphasise the significance of the indicator values that have been produced and analysed as result of former policy decision in the relevant area. These figures may be used for (1) new decisions in the relevant area, (2) continuing the policy measures, and (3) changing the former policy measures.
3. Statistics and transparency in the PMP
This section explains how transparency in policymaking is related to statistics. Two quotes from Wikipedia concerning "Official Statistics" (30-11-2010) are listed below:
(1) "Official statistics are statistics published by government agencies or other public bodies such as international organizations."
(2) "Official statistics should be objective and easily accessible and produced on a continuing basis so that measurement of change is possible. Official statistics: • result from the collection and processing of data into statistical information by the government institution responsible for that subject-matter domain. They are then disseminated to stakeholders and the general public. Statistical information allows users to draw a relevant, reliable and accurate picture of the development of the country, compare differences between countries and changes over time. They enable stakeholders and decision makers to be well informed and develop policies for addressing actual development challenges. • make information on development accessible to the public and therefore assist in the accountability of public decision-making. One use of official statistics is to measure the impact of public policies and highlight the need for development."
The descriptions on Wikipedia show how statistics are related to public policymaking and how statistics need understanding of the reality in order to provide unambiguous information about reality. This section demonstrates the importance of statistics in improving the transparency in public policymaking. Before doing so, we will explain some of the principles of statistics: the meaning of data, meta data and information.
In this paper the following definitions are used for data, message, and information (see, e.g., Stol, 2009, Sundgren, 1975, Stol, 1990, and IFIP, 1998).
Data is a structure intentionally arranged into a medium to represent by this structure a part of reality.
A message is a reference to a part of reality as interpreted by the receiver of the data.
Information is new knowledge derived from (1) the old knowledge and (2) the knowledge implied by the absorption of the message.
In semiotics and linguistics a distinction is made between the syntactical, semantical, and pragmatical levels of communication (see, e.g., http://en.wikipedia.org/wiki/Semiotics, Chomsky, 1975, Nielen, 1976, and, Stamper, 1996 3).
Syntax is the study of principles and rules for constructing sentences in natural language. Syntax is about symbols and a combination of symbols.
3 ) In this context the Semiotic Framework by Stamper (1996) is worth mentioning. In this framework, Stamper distinguishes the IT platform and the human information functions. In the IT platform the framework presents the following three levels (1) physical world, (2)empirics, and (3) syntactics; in the human information function we see (1) semantics, (2) pragmatics, and (3) social world are recognised.
Sets of indicators for policy making Page 44
Semantics is the study of interpretation of signs within particular circumstances and contexts. Semantics is about the references to the real world; it gives the relation between the symbols and the objects to which they refer.
Pragmatics is the study of the ability of the of natural language speakers to communicate more than which is explicitly stated. Pragmatics is about the intended meaning in the communicated language; it gives the relation between the symbols, the use, and the effects.
In Table 1, the above defined concepts are related to each other in a matrix. For most of the cells different sets of metadata are relevant for the combination of the level and concept symbolized by the cell. Each cell is identified by a number and a character.
We will call this matrix: The Metadata Matrix 4. The Metadata Matrix provides a frame of reference for the definition of the kinds of metadata to be used in the communication in the PMP.
A: data B: message C: information concept level
I : syntax IA: Structure put in the IB: Structure in the IC: Structure in the medium to represent brains brains the data
II: semantics IIA: Reference to reality IIB: Reference to part IIC: Reference to part as meant by the sender of reality as of reality in the con text of the data interpreted by the of the frame of receiver reference
III: pragmatics IIIA: Choice of the IIIB: The meaning of IIIC: What is the effect medium and structure the data for the of the new knowledge so that the receiver of receiver on the receiver of the the data can interpret data them
Table1: The Metadata Matrix.
This Section proceeds with using the Metadata Matrix in order to show which kinds of metadata are relevant for (1) the definition of the objectives and the related indicators and indicator values (see parts 2 and 3 of the framework for the PMP), (2) the measurement of the results (part 5), and (3) the dissemination of the intended values of the indicators versus the realised ones (justification, part 6 and 7). When we apply the Metadata Matrix on the output process (i.e. the dissemination of the results) , the user of the output is the “receiver”. The statistical system in its role as producer of the indicators (or better indicator/value combinations) is the sender of the data. Examples of questions to be answered on the metadata are in this consideration: What kind of metadata do we need, to make sure that:
(1) the data refer to the part of reality meant by the Statistical System?
(2) the choice of the medium and structure of the data meet the requirements of the receiver?
4 ) A Dutch version of this Metadata Matrix is published in Stol (1995a) and in English in Stol (2002b). In these publications the physical level was the fourth level referring to the media used in the communication process. Since this level does not contribute to the metadata models presented in the context of this thesis, I have omitted the physical level.
Sets of indicators for policy making Page 45
(3) the meaning of the data for the receiver is the same as meant by the Statistical System 5?
(4) the data once interpreted by the receiver refer to the part of reality as meant by the Statistical System?
Normally speaking we want to provide “unambiguous and objective” indicators, so that we should not bother on the effect of the data (cell IIIC in the Metadata Matrix) 6.
The metadata to be considered for the input process (the process providing the observations in the Object System of Interest) are briefly discussed below, in particular as far as E-Datacapture is concerned. This means in terms of the matrix that: (1) The sender of the data is the respondent 7, (2) the receiver of the data is the (automated) part of the Statistical System that registers the data so that they can be used in the further process, leading to the indicators. In Table 2 we show the relation between the cells defined in the Metadata Matrix and the parts of the E-Datacapture system in which they are addressed.
Cell in the Metadata Matrix Statistical e-Datacapture system
IA : Structure put in medium to Questionnaire form. represent the data
IIA: Reference to reality as Questionnaire responses, link with data in information meant by sender of the data system of data provider.
IIIA: Choice of medium and The software for the e-questionnaires needs to be available structure so that the receiver of for a great part of the respondents; this means that media the data can interpret them used are accessible for all of them, as well as the data formats used. (E.g., Internet and XML.)
IIB : Reference to part of reality The definition of the questions and support files used for as interpreted by the receiver explanation of the expected answer (code systems, nomenclatures).
IIIB : The meaning of the data for Availability of electronic links between the questions and the receiver the content of the business administration, may assure the data receiver that the data filled in are according the bookkeeping.
IIC : Reference to part of reality Question answers received and stored in the environment in the context of the frame of of the data collector. reference
IIIC : What is the effect of the The effect of information for the receiver of the data (the new knowledge on the receiver Indicator System) may lead action from the receiver (e.g. of the data asking for explanation of unexpected differences with former messages).
Table 2: Metadata of data-capture system.
We clarified that the Statistical System plays an important role in the framework for the PMP. Applying the framework is a precondition for obtaining transparency in the PMP. After the policy
5 ) The definition of the meaning of the data is a well know challenge for statisticians: how to add the proper metadata in order to transfer the data as they are meant. 6 ) Transparency requires producing unambiguous data without considering the pragmatical effect of the information. 7 ) In practice, the respondents can be households and enterprises.
Sets of indicators for policy making Page 46
decision making the objectives should be determined as follows: (1) the objectives must refer to a part of reality, (2) the objectives must be measurable, and (3) the objectives must be verifiable. The consequences of these requirements are:
(1) The objectives should be formulated in terms of objects (or object types) and their intended values, after the realisation of the policy. The values are attributes of the objects (or object types) and include a reference to the time on which the values should be achieved.
(2) The values of the objects (including time) must be measurable. This means that before the objectives are determined, the method for measuring the realisation of the objectives can and must be established.
(3) The values of the objects must be verifiable. It must be possible to ascertain the correctness of the information about the realisation of the objectives.
In the Statistical System we use the concept of "indicator" as item to address the progress of situations or processes. It is important to be aware that the concept of indicator is identical to the concept of the formulation of objectives as has been described above. The requirements for the specification of the objectives regard directly the use of indicators. Indicators are the same, or equivalent with, the terms in which the objectives are specified. A statistical indicator is defined as "a data element that represents statistical data for a specified time, place and other characteristics". 8
If the indicators are defined as required, it is not difficult to measure the outcome of the policymaking process. The measurement will be according the basic rules, preferably determined at the moment that the objectives are established. It is clear that this also implies that the population should be known at the moment of the definition of the objectives as well as on the moment of measurement. Depending on the subject of the policy different kinds of measurement methods may be used, varying from survey to the use of administrative data. To avoid administrative burden on the respondents Statistical Offices often use administrative data, provided for other purposes than the policy subject at hand. One should be aware of differences in the metadata by which these administrative data are determined. Quite often the data seem to refer to the same object or object group, but de facto refer to other objects or object groups.
The last step in the statistical process in the dissemination of the results of the PMP. The metadata matrix and the basic rules for transparency tell us that (1) the indicator/value should refer to the part of reality addressed by the policymaking, (2) the choice of the medium and structure of the indicator/value should meet the requirements of the receiver, (3) the meaning of the indicator/value for the receiver should be the same as meant by the Statistical System, and (4) the data once interpreted by the receiver should refer to the part of reality as meant by the Statistical System.
Finally, a few words on the use of data from earlier policymaking processes, to improve the quality of new policy decisions. Literature on Evidence-based Policy Making, show a variety of opinions on the pro's and con's of using data of earlier policymaking experiences for new policymaking (see e.g. Dorey, 2005, Cabinet Office, 1999, Pawson, 2006). In some cases, like health care (McCaughey, 2010) it is clear that results of former experiences are worthwhile to prelude upon. In situation when policymaking addresses a new situation (other circumstances, other issues) it is not always clear how evidence can be obtained. In these cases one should be very careful and not use data that are not in line with the policy that is at hand. The use of EBPM in a broad sense (Stol, 2009) forces politicians to think about their objectives in terms of
8 ) Economic Commission for Europe of the United Nations (UNECE), "Terminology on Statistical Metadata", Conference of European Statisticians Statistical Standards and Studies, No. 53, Geneva, 2000.
Sets of indicators for policy making Page 47
objects and object groups (measurable objectives), and by doing so the quality of their decisions may improve. Since they are related to reality, measurable and transparent.
4. Guidelines for transparent public policymaking
The framework for the PMP provides the means to make the PMP transparent and measurable. However, there are circumstances in which the framework can be applied with more success than in other situations. The EBPM control system must fulfill the following three requirements for effective control of the Object System of Interest ( Stol, 2009, de Leeuw, 1982).
(1) The objectives must be unambiguous and clearly formulated in terms of situations to be obtained in the Object System of Interest.
(2) Full knowledge of the Object System of Interest. This knowledge covers the current situation and the desired situation but also the way to obtain the desired situation and the status of the situation in the Object System of Interest during the Change Process.
(3) Diversity in controlling measures for an effective control of the Object System of Interest:
(a) Adaptive measures to change the organization of the Change Process.
(b) Goal changing measure to change (parts of) the objectives.
In the specification of the requirements for effective control of the Change Process in the Object System of Interest the distinction between the role and responsibilities of the Political System 9 and the government is important. The responsibility for the full control over the Change Process is shared by the Political System and the government. In the division of this task the Political System is in charge as soon as deviations occur in comparison with the original measures defined.
Although special circumstances can be assumed leading to declining evidence-based policymaking, it is hardly possible to raise arguments for denying the fact that consequences of policy decisions can be measured to verify the promises made by the politicians. (Stol 2009, p. 36)
Non-rational behavior in the evidence-based decision making process (EBDM) does not mean that the framework for the PMP is not effective. The reason is that all other parts of the framework can be executed in the flow of the proposed one. Indeed, one could argue that rational decision making leads to a better quality of the decision itself (cf. Cabinet Office, 1999). This may be true, but it does not necessary mean that the execution of the policy measures following the decision is transparent and measurable. The Garbage Can model (March and Olsen, 1979) seams to occur frequently (Koundouraki, 2007). When the decision process takes place according to the Garbage Can model the policy decision is a more or less a coincidental combination of (1) solutions, (2) problems to be solved, (3) people, and (4) choice opportunities.
As the decision process is isolated in the framework the only prerequisite for a successful use, is that the output of the decision process can be transformed in concrete objectives to be realised in an (well defined) Object System of Interest. In the day-to-day practice of public policymaking we miss sometimes this relation between the policy decision and the objective in terms of intended changes in the reality. If this relation cannot be formalized, it is not possible to
9 ) The Political System consists of politicians and the part of the government that may be organized in the same institution as the part of the government that is responsible for the control of the execution of the Change Process.
Sets of indicators for policy making Page 48
measure the outcome of the policy measures. In that case the use of Official Statistics is not is not a serious option, nor the intention of being transparent in the PMP.
We must understand that in the Political System people are chosen based on their position in a political party and the promises (in the context of the persuasiveness of their arguments) of their program. These politicians should be aware of the need of transparency and measurability of the PMP. Sometimes this may mean that they explicitly have to admit that their promises cannot be realised. The reasons for this may be (1) compromising with other politicians, (2) other circumstances than anticipated, or (3) consciously promising more than they can substantiate. When the politicians prefer to hide their deviation of the promises, they will tend to make the PMP less transparent and verifiable so that the deviations are not so visible.
We mention three circumstances that influence the effectiveness of the use of the EBPM framework: the complexity of the Object System of Interest, the complexity of the organization, and unpredictable events.
(1) The complexity of the Object System of Interest and elapsed time from the start of the Change Process until the achievement of the outcome, play an important role in the possibility to control the Change Process in an effective way (Bouchard and Caroll, 2003). Although measures can be taken to lower the complexity of the Object System, this will not always be sufficient. In some cases it is not possible to measure the effect of the policy measures because of the number of variables and uncertain behavior of the Object System in its environment.
(2) The organization in the Political System and the organization of the Change Process may be quite complex cf. Bouchard and Caroll (2003) . There are various measures to adapt the organization of the Policy System and the Change Process to the specific policy issue. Nevertheless these measures may not always work, especially when different political issues require conflicting organizations.
(3) Unpredictable events may happen or unknown effects may occur that change the Object System of Interest and/or its relevant environment during the Change Process. Taleb (2004, 2007) shows many examples of unexpected events and unknown effects of (policy) measures. Contingency planning may be effective to minimise the consequences of these events. 10
An optimal use of Official Statistics and the Statistical System in transparency of the PMP may be obtained if the following conditions are fulfilled:
(1) The Statistical System is involved in the definition of the objectives and the indicators.
(2) The Statistical System uses the same principles for the definition of the indicators mentioned in (1), as it does for the indicators on measurement.
(3) The Statistical System uses unambiguous and identical metadata for the definition of the indicators to be realised as well as for the definition of the outcomes of the PM.
(4) The Statistical System uses metadata models in which all references between reality, addressed by the PM, and the indicators are transparently shown.
10 ) Contingency planning to avoid disturbances in the Object System of Interest may be quite expensive. Compare the organisation of the Olympic Games in China where every possible circumstance has been envisaged in the preparation of the Games. The expenses have been never so high (about 25 billion euro) and probably will not be exceeded in the future. For instance, the budget for the next games in London is 13 billion euro.
Sets of indicators for policy making Page 49
Politicians and statisticians can fruitfull cooperate by using common metadata models on (1) objectives, (2) measurement, (3) dissemination of figures and (4) justification of PM. In Stol (2009, Section 6.3) examples of the metadata models on objectives, measurement and presentation (dessemination and justification) are proposed, to be used as guidelines for the cooperation between politicians and statisticians.
Sets of indicators for policy making Page 50
Literature
Althaus, C., Bridgman, P., and Davis, G. (2007). The Australian Policy handbook. Fourth edition. Allen & Unwin, Crows Nest, Australia. Bouchard, G. and Caroll, B. (2003). One Size Does not Fit All: A Contingency Approach to Policy Making. Paper presented in CPSA meeting, Halifax, Nova Scotia. http://www.cpsa- acsp.ca/paper-2003/bouchard.pdf Cabinet Office (1999). Modernising Government. Cm 4310, London: Stationary Office, UK. Checkland, P. (1999). Systems Thinking, Systems Practice . Wiley, New York, NY. ISBN 0 471 986062. Chomsky, N. (1975). Reflections on language. Panteon Books, New York, NY. Davies, H., Nutley, S. M., and Smith, P. (2001). What works?: Evidence-Based Policy and Practice in Public Services . Policy Press, Bristol, UK. Dorey, P. (2005). Policy-making in Britain . SAGE Publications. London UK, ISBN 0-7619-4903- 8 Easton, D. (1979). A Framework for Political Analysis. University of Chicago Press, Chicago. IFIP (1998). FRISCO A Framework of Information System Concepts (Eds. Falkenberg, E.D., Hesse, W., Lindgren, P., Nilsson, B.E., Han Oei, J.L., Rolland, C., Stamper, R.K., Van Assche, F.J.M., Verrijn- Stuart, A.A., and Voss, K.). International Federation for Information Processing. (ftp://ftp.leidenuniv.nl/pub/rul/fri-full.zip) Koundouraki, E. (2007). The Information Cycle in the European Commission’s policy-making process. Information Research, Vol. 12, No. 4, http://informationr.net/ir/12-4/colis/colis07.html Leeuw, A.C.J. de (1982b). Organisaties: management, Analyse, Ontwerp en verandering ,. Van Gorcum, Assen/Maastricht, the Netherlands. March, J.G., and Olsen, J.P. (1979). Ambiguity and choice in organizations . Universitetsforlaget, Bergen, Norway. McCaughey and Bruning (2010) Rationality versus reality: the challenges of evidence-based decision making for health policy makers , Implementation Science 2010 Nielen, G.C. (1976). De bedoeling van informatie voor mens en organisatie. Samsom, Alphen aan den Rijn, NL. Pawson, R. (2006). Evidence Based Policy, a Realist Perspective. Sage Publications, London, UK. ISBN 10 1 4129 1059 5 Stamper, R.K. (1996). Signs, Information, Norms and Systems, Signs of Work, (Holmqvist, P., Andersen, P.B., Klein, H. and Posner, R. (Eds.), Walter de Gruyter, Berlin, Germany. Stol, H.R. (1990). Informatieplanning in de praktijk . Samsom, Alphen a/d Rijn, Netherlands. ISBN 90 14 03644 2. Stol, H.R. (1995). Silentium en de informatiemaatschappij, Organisatie van de informa- tievoorziening: verschuivende en transparante grenzen , (ed. Mantelaers, dr. P.A.H.M., e.o.), Klu- wer, Deventer, NL, ISBN 90 267 2292 3 Stol, H.R. (2002). Metadata in data capture. Integration of Statistics in the Information Society Proceedings Statistical conference SORS , November 2002, Radenci, Slovenia. Stol, H.R. (2009). A Framework for Evidence-based Policy Making using IT, a Systems Approach. Eburon, Delft, Netherlands, ISBN 978-9059722927 Sundgren, B. (1975). Theory of data bases. Petrocelli/Charter, New York ASA, ISBN 0-88405- 307-5 Taleb, N.N. (2004). Fooled by randomness . Random House, New York, NY. ISBN 0-8129-7521-9. Taleb, N.N. (2007). The Black Swan, “the impact of the highly improbable” . Random House, New York, NY. ISBN 1400063515.
Sets of indicators for policy making Page 51
Setting the right priorities for European statistics: a contribution from a user and producer
Aurel SCHUBERT 1
The European Central Bank (ECB) conducts the euro area monetary policy with the primary objective of maintaining price stability over the medium term. It also contributes to other ESCB tasks such as the conduct of foreign exchange operations, the oversight of payment systems and the stability of the financial system. The performance of all these tasks requires a large amount of high quality statistical information. At the ECB, the statistical information is used to assess economic, monetary and financial developments in the euro area (and increasingly also in its individual Member States), to feed into the ECB macro-economic projection exercises, to assess the transmission of monetary policy decisions of the ECB to the overall economy and to analyse financial integration and financial stability, among others. A major part of the statistics used by the ECB are produced by the ECB itself, with the assistance of the EU national central banks (NCBs), which, together with the ECB, form the European System of Central Banks (ESCB). The ECB also provides the statistical support to the European Systemic Risk Board (ESRB), the newly created independent EU body responsible for the macro-prudential oversight of the overall financial system within the EU.
However, the ECB does not only use its own statistics. For the performance of its tasks, the ECB also relies heavily on statistics produced by the European Statistical System (ESS). This paper recalls some key priorities of the ECB for ESS statistics and how these statistics are used for evidence-based decision making at the ECB. It concludes that the high demand for more quantitative information as a basis for policy decisions represents good news for statisticians but also a major challenge and a major responsibility, as statisticians have to ensure that the demanded figures are “ right for the purpose ”, but also “ right on time ”.
Priority statistics for policy-making: do merits outweigh the cost?
Being a heavy user of ESS statistics, the ECB very much appreciates the current efforts of the ESS to keep European statistics focused on policy needs of highest relevance in Europe. As experience has shown, an early and close dialogue with the key users of ESS statistics is indispensable to get an informed view on the relevance and, hence, on the merits of such statistics . This is in essence what Principle 11 of the European Statistics Code of Practice (Relevance) says. For the ECB, this is one of the most important statistical quality principles.
Sets of indicators for policy making Page 52
Indeed, the ECB is fully reliant on a sound statistical basis in order to properly fulfil its tasks in an increasingly complex economic environment. This aspect will be further elaborated below.
Focusing on statistics to address the policy needs of higher relevance in Europe and re- prioritisation of statistical requirements might, however, not be enough in the current circumstances of statistical offices. An efficient priority-setting in the ESS, focusing on policy relevant statistics, goes hand-in-hand with the so-called “European approach to ESS statistics”, as spelt out in the related communication of the European Commission 2. New production methods of statistics will require investments in a modern statistical function, in particular in IT and human skills, but at the same time they will yield in the medium-term important efficiency gains and cost savings to statistical compilers . These savings are indispensable in particular in periods of resource constraints. The new production methods of statistics should be seen as an investment in the future of statistics and in its essential role in evidence-based policy making.
More modern business processes may also bring very significant gains to respondents . These include a better use of data contained in administrative sources and in business registers that are reported only once, ideally extracted directly from respondents’ systems. The approaches described above, as well as an enhanced storage (e.g. data warehousing) and an enhanced dissemination of data, may have also various advantages from the users’ perspective . For example: i) the reliance on micro-data may accommodate more flexibly and more promptly emerging ( ad hoc and regular) user needs, may reduce the timeliness of new statistical information and may increase the consistency and integration of the various data sets. We will return to this aspect later in the paper. ii) a modern data dissemination increases the accessibility of data and, thereby, increases the users’ awareness of which statistics are available and how they can be best used.
Inevitably, the current circumstances of statistical offices seem to indicate that new and more cost effective solutions alone might not be sufficient to free enough resources to accommodate growing users’ requirements. It might be required to consider, in addition, substantial simplifications and/or discontinuation of statistics that are either very costly to produce or less or no longer relevant for policy-making. Undisputed is that, when setting statistical priorities in the context of a rational use of resources, it is essential to look at the cost and at the merits of
1 The contribution of Alda Morais to an earlier version of this paper is gratefully acknowledged.
Sets of indicators for policy making Page 53
statistics in conjunction, and to give special attention to those statistics whose merits clearly outweigh their cost. And this balance will change over time, thus requiring constant monitoring.
Policy use of statistics: shaping the statistics to further increase their merits
One of the themes of this conference is whether the use of statistics should influence the processes used for disclosing statistics. As with the question on the merits vs. cost above, the answer seems to be affirmative if the relevance of their policy use does justify it. In this way, it might be possible to obtain substantial improvements in the relevance of (existing statistics) by introducing gradual changes in the processes used to produce such statistics and in the standardisation of information.
The forthcoming Excessive Imbalance Procedure (EIP) is a clear example of an important new policy use of (existing) statistics (to monitor sources of macroeconomic imbalances and ensuring appropriate corrective action when necessary). In this case, all EU countries will matter, and not only the euro area countries, and there is a need for individual country data. Already for the preventive part of the EIP, the system will be based on a scoreboard composed of a limited number of statistical indicators. While a final decision on the indicators to be included in the scoreboard is still outstanding (including their detailed definition, source, legal basis, frequency, timeliness, etc.), it is clear that an early and intense discussion between the users of these data and the statisticians will determine the “fitness for purpose” of the scoreboard: i) by choosing the most adequate indicators to meet the policy needs (in terms of their cross- country comparability, length and timeliness, for example); and ii) by ensuring that the indicators are sufficiently robust, i.e. based on a sound statistical framework and compiled in accordance with the principles laid out in the European statistics Code of Practice, rather than on an undue degree of estimation.
To some extent, ensuring the “fitness for purpose” of the scoreboard is an exercise that can be compared with the choice of indicators that have been used by the ECB in the ECB Convergence Reports. Successive waves of enlargement of the euro area (from initially 11 to 17 member countries today) has meant long lead times in the preparatory work for both the ECB and the countries joining the euro area. High quality national statistics are crucial as part of
2 “ Communication from the Commission to the European Parliament and the Council on the production method of EU statistics: a vision for the next decade”, Brussels, 10 August 2009.
Sets of indicators for policy making Page 54
EU Member States obligations to prepare for joining the European and Monetary Union (EMU) because they are also used in the assessment of their economic convergence with the EMU, which is largely based on statistical indicators.
Another example of an important policy use of statistics may be the decision-making at the ECB. Following the first Governing Council meeting in the month (when monetary decisions on the key interest rates are normally taken) the ECB President delivers his “Introductory statement” (published on the ECB website). In these statements, the President gives several examples of how statistics are used by the Governing Council to assess the economic situation and the stance of the monetary policy. From these and other ECB statements and publications, it is notable that the ECB attaches the highest importance to statistics and considers these particularly important for the well-functioning of the Economic and Monetary Union and for the financial stability in the European Union. This is particularly valid at the moment, when the strengthening of the statistical basis is seen as essential in Europe (and globally) to overcome the ongoing economic and financial crisis as soon as possible and in order to avoid a recurrence of this crisis.
Obviously, statistics lose a great part – if not all – of their relevance if they are late for decision making. If data is made available, not when they are needed to underpin the decisions, but only after the policy decisions have already been taken. (This should be without prejudice to the usefulness of quick information, e.g. obtained from opinion surveys to bridge the time needed to compile statistics.) A typical example of the “fitness for purpose” of statistics is the case of the quarterly national accounts. These should ideally be made available by 30, 60 and 90 calendar days after the reference period, in order to feed, in a timely manner, the monetary policy deliberations of the ECB Governing Council, which take place at the beginning of each month.
The important concept here is the term “fitness for purpose”. In the case of those indicators where timeliness is the most important quality feature, the use of more flash estimates could be considered, with various advantages to both users and compilers. For example, GDP flash estimates at t+30 days may be volatile at national level. In this case, however, only the results for the euro area aggregate would be used by the Governing Council, and these aggregated - and timely - figures would be expected to be of sufficiently good quality for the decision making process. Besides, the results of the GDP flash estimates at t+30 days are expected to improve substantially in quality over time, as it has happened with the release of GDP at t+45. Therefore, the ECB is happy to support and contribute, as needed, to the planned ESS study assessing the feasibility of producing such flash estimates.
Sets of indicators for policy making Page 55
The plans to advance the release of integrated euro area accounts from the current timetable (t+120) to around t+90 is yet another example of how statistics can see their merits for policy- making substantially enhanced, if they are “right on time”. These plans will allow advancing the provision of the related briefing by 1 month. In order to achieve this ambitious target, a major joint effort is needed from Member States, not only on the part of the ESS (in securing timely national contributions for quarterly non-financial sector accounts), but also on the side of the ESCB (for timely data on the financial accounts and the euro area balance of payments). Again, the outputs are expected to improve substantially in quality over time, and experience has shown that sharing best practices among Member States is instrumental in achieving such further quality improvements.
The additional effort for advancing the release of integrated euro area accounts is already reflected in the plan of the ESCB Statistics Committee, as agreed at its December 2010 meeting, to achieve a transmission deadline of t+80 for the balance of payments / international investment position by 2019 following a step-wise approach. It is, therefore, essential that the transmission deadlines for the quarterly non-financial sector accounts, quarterly government financial accounts and government debt also move in a synchronised manner. The ECB would consider prudent to make allowances for this synchronisation already in the ESA 2010 Regulation, which will be discussed at the political level within the next few weeks and months.
Last, but not least, a new policy use of (mostly existing) statistics might stem from the European Systemic Risk Board (ESRB). The ESRB will determine and analyse the information relevant and necessary to perform its wide ranging tasks. The specific ESRB data needs are still under examination but most of the new data requirements are expected to relate to financial (stability) statistics and to the detection of imbalances (in the inter-linkages and exposures of sectors and sub-sectors, countries, market segments, etc.). This will be first and foremost in the financial sector (thus, based on data to be provided by supervisory authorities or by national central banks), but also in the other economic sectors.
Without prejudice to the legal framework of the ESS in the field of statistics, ESRB data requirements might also need to encompass some enhancements in the data provided by the ESS, in order to better feed the macro-prudential analysis.
Sets of indicators for policy making Page 56
From the integrated quarterly financial and non-financial accounts, a wide range of indicators derived from the quarterly financial and non-financial accounts 3 are already used in the internal briefing material, the so-called White Book, which is provided to the ESRB on a regular basis by the ECB.
In order to address future ESRB data needs, these accounts might need to be extended, from the euro area to the whole EU and each of its Member States. In the case of financial stability, also the small Member States matter. As the economic and financial crisis has shown, shocks in small countries of the euro area can occasionally have significant effects on the whole area. Therefore, any ESS assessment of priorities should consider that raising the reporting thresholds beyond certain limits may have adverse effects on the availability of country data for national non-financial sector accounts, as this might drastically limit the number of countries actually reporting. This might turn out as saving on the wrong place.
As already mentioned in the context of the use of ESS statistics for monetary policy, in order to remain relevant for ESRB purposes, the timeliness of these accounts might also need to be advanced. Decomposing value changes in the EU financial and non-financial accounts might also bring additional value added to policy-makers, by allowing to distinguish real and price developments. Finally, the list of Principal European Economic Indicators (PEEIs) might need to be extended, namely with respect to property prices indices, in order to provide information not only on residential property, but also on commercial property, a factor that played an important role form the current financial crisis.
The repeated use of policy relevant statistics, e.g. in the ECB Convergence Reports, in the ECB President’s Introductory Statements, in Scoreboards, in ECB speeches and other public statements can be seen as an additional, new “ quasi policy tool ” of monetary policy makers. One could even argue that the traditional “ open market policy ” of central banking has been supplemented by an “ open databank policy ”, using data to influence public opinion and – ultimately – policy makers.
The future of statistics: anticipating future needs
In order to stay relevant, statistics should proactively adapt to the new needs of evolving societies. A case in point in this respect is the plan to improve the linkages between economic, social and environmental statistics in the context of the Commission communication on “GDP
3 Household debt to disposable income; household debt to total financial assets; non-financial corporations’ debt to value added and leverage ratio; household savings rate; household investment rate; non-financial corporations’ debt
Sets of indicators for policy making Page 57
and beyond” and of the Stiglitz-Sen-Fitoussi Report. Many of the initiatives in this domain, in particular as regards the recommendations in the first part of the Report (i.e. under “classical GDP issues”) are already addressed, for some time now, by the ESS and/or the ESCB, for example: i) in the joint compilation the quarterly integrated euro area sector accounts since 2007, ii) in the development of the forthcoming Eurosystem survey to collect data on household finance and consumption. These two examples of important European initiatives show the merits of proactively investing in achieving more integrated statistics, in exploring new data sources, in reutilising existing data and in presenting them in a consistent integrated framework. Priority will continue being given over the coming years to improving the completeness, geographical coverage, consistency and timeliness of the euro area accounts and to further complement them with distributional information on indebtedness, wealth and consumption of the household sector.
The financial and economic crisis confirmed that statistics need to be proactive and not simply reactive to events. The need to monitor and understand economic and financial developments and shocks in almost real time points to the potential of – and actually need for - disaggregated and detailed information. Indeed, the demand for detailed information on all sectors (including on non-financial sectors) and the growing requests for individual EU country data (in addition to the euro area aggregates) are some of the prospective data demands for financial stability purposes and to better inform the policy discussions on shocks and on the transmission mechanism of monetary policy.
The ECB experience has shown that increased use of micro databases and registers maintains the traditional statistical aggregated series but adds much in terms of flexibility for a better statistical response to these new challenges. Moreover, data need to be reported only once and the reporting burden is kept to a minimum, as the same source data can be used – and reused - more flexibly to derive various (micro and macro) statistics to meet new data needs. This presupposes an adequate confidentiality regime and a proper identification and classification of entities and instruments, based on modern data standardisation and matching principles that are a conditio sine qua non to the processing of large amounts of data.
In the case of the ESS, several similar avenues may bring better responses to users, as already hinted in the first part of this paper. These avenues may encompass the setting up of the necessary infrastructure to fully exploit and re-use data contained in administrative sources and in business registers, the integration of the various statistics (via the sharing of data and
to GDP; non-financial corporations’ leverage ratio; non-financial corporations investment share, etc.
Sets of indicators for policy making Page 58
processes) and the extracting of data directly from respondents’ systems, while ensuring that appropriate levels of statistical confidentiality remain.
In sum, standards and common tools for the efficient and secure exchange of data and metadata, be it among the ESS or between the ESS and the ESCB, are highly effective in bringing flexibility into the statistical systems. A case in point is certainly the plan for the interoperability of the Register of Institutions and Assets Database (RIAD) and the EuroGroup Register (EGR), which we expect to deliver promising results.
Concluding remarks: “A little bit late” is simply “too late”
The complexity of the modern world no longer allows conducting policy-making “ by experience or feeling ” but requires instead a sound empirical basis. The final decisions might still require a “judgement call” (“ the art of central banking ”), but this has to be based on models that create a sound and reliable basis for this final “judgement call”.
The demand for more quantitative information as a basis for policy decisions is undoubtedly good news for statisticians. But it also represents a major challenge and a major responsibility, as statisticians have to ensure that the demanded figures are “ right for the purpose ”, but also “right on time ”. Indeed, policy-making requires a timeliness pattern that is much more demanding than the pattern accepted for several other purposes (e.g. for analysis), and “ a little bit late ” is actually “ too late ”, as it makes statistics lose most, if not all, of their policy relevance.
I am convinced that statisticians have many good reasons to be proud of the impressive statistical achievements since the start of the preparation of the EMU, despite the tight statistical resources. The coordination of the two systems (the ESS and the ESCB) has been instrumental to this. A few examples of this success are the joint euro area sector accounts, the joint work on quality frameworks and the coordinated update of the umbrella Council regulations. On this occasion, I would like to conclude by praising in particular the work done in the context of the “Committee on monetary, financial and balance of payments statistics” (CMFB). The CMFB just had its 20 th birthday two weeks ago (the Council Decision establishing the CMFB dates back to 25 February 1991). The achievements of the CMFB in these 2 decades have confirmed its strategic and unique role as a body for the coordination, joint work, discussion and decision between the ESS and the ESCB. This is a solid base on which we can build our common statistical future, a future of relevant and timely statistics, for the respective policy purposes.
Sets of indicators for policy making Page 59
Indicators for EU policies: business as usual?
Andrea SALTELLI
Bruxelles, March 10-11 2011 2
Role of statistics-based knowledge in the making of EU policy
Based on: Saltelli, A➡, D Hombres, B➡, Jesinghaus, J➡, Manca, A➡, Mascherini, M➡, Nardo, M➡, and Saisana, M➡, 2010, Indicators for EU Policies➡ Business as Usual? Social Indicators Research , doidoidoi:doi ::: 10➡1007/s11205-010-9678-4➡
The problem with Lisbon
Bruxelles, March 10-11 2011 3
Too many frameworks Statistical information and OCM scarcely used in the communication of policies Few data based narratives from EU Institutions
But: The Commission is committed to evidence based policy Data-based narratives are routinely created by international organizations, NGO s, civil society, which at time define the political discourse
Sets of indicators for policy making Page 60
The problem with Lisbon
Bruxelles, March 10-11 2011 4
Too many frameworks
The problem with statistics for Lisbon
Bruxelles, March 10-11 2011 5
Few data based narratives from EU Institutions
An ambitious and broad reform agenda needs a clear narrative , in order to be able to communicate effectively about the need for it➡ So that everybody knows why it is being done and can see the validity of the need to implement sometimes painful reforms➡ So that everybody knows who is responsible➡
Wim Kok s report , 2004
Sets of indicators for policy making Page 61
Bruxelles, March 10-11 2011 6
Data-based narratives are routinely created by international organizations, NGO s, civil society
An example, about Mauritius, Economist October 16, 2008
THE 1.3m people of Mauritius love to prove famous people wrong. On independence from Britain in 1968, pundits such as a Nobel prize-winning economist, James Meade, and a novelist, V.S. Naipaul, did not give much of a chance to this tiny, isolated Indian Ocean island 1,800km (1,100 miles) off the coast of east Africa. Its people depended on a sugar economy and enjoyed a GDP per person of only $200. Yet the
island now boasts a GDP per person of $7,000 , and very few of its people live in absolute poverty. It once again ranks
first in the latest annual Mo Ibrahim index , which measures governance in Africa. And it bagged 24th
spot in the World Bank ’s global ranking for ease of doing business —the only African country in the top 30, ahead of countries such as Germany and France. How does it pull it off?
Bruxelles, March 10-11 2011 7
Data-based narratives are routinely created by international organizations, academia, NGO s, civil society
Data sources: EUROSTAT, Integrated Network for Societal Conflict Research , Freedom House, Economic Freedom Network , Transparency International, World Bank , The Economist Intelligence Unit, DG ECFIN and the IMF
Sets of indicators for policy making Page 62
International league tables at times define the political discourse Bruxelles, March 10-11 2011 8
See:See:See: Saisana, M➡, d Hombres, B➡, Saltelli, A➡, Rickety numbers: Volatility of university rankings and policy implications, 2011, Research Policy, 40 165 177➡
Simulated rank range
SJTU rank 1-5 6-10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65 66-70 71-75 76-80 81-85 86-90 91-95 96-100 101-105 106-110 111-115 116-120 121-125 126-130
Harvard UnivBruxelles, March 10-11 2011 100 1 USA9 Stanford Univ 89 11 2 USA Univ California - Berkeley 97 3 3 USA Univ Cambridge 90 10 4 UK Massachusetts Inst Tech (MIT) 74 26 5 USA California Inst Tech 27 53 19 1 6 USA Columbia Univ 23 77 7 USA Princeton Univ 71 9 11 7 1 8 USA Univ Chicago 51 34 13 1 9 USA Univ Oxford 99 1 10 UK Yale Univ 47 53 11 USA Cornell Univ 27 73 12 USA … … … Univ California - San Francisco 14 9 14 3 11 3 7 10 4 3 3 3 6 1 6 1 18 USA … … … Duke Univ 10 6 13 11 6 3 7 6 3 1 3 1 9 9 7 1 3 1 32 USA Rockefeller Univ 4 10 23 26 1 3 3 3 3 3 4 4 6 3 1 1 1 32 USA Univ Colorado - Boulder 19 39 30 11 1 34 USA Univ British Columbia 20 60 20 35 Canada Univ California - Santa Barbara 9 9 10 3 10 6 7 6 11 4 6 3 4 7 1 1 36 USA Univ Maryland - Coll Park 6 37 44 9 4 37 USA … … … Ecole Normale Super Paris 7 9 4 6 7 6 4 9 6 7 4 3 3 4 3 3 1 6 4 73 France Univ Melbourne 1 20 17 31 23 1 6 73 Australia Univ Rochester 1 10 7 16 24 14 10 10 6 1 73 USA Univ Leiden 3 6 9 23 24 13 14 9 76 Netherlands … … … Univ Sheffield 1 21 26 21 9 13 7 1 77 UK Tohoku Univ 4 1 7 1 4 17 19 3 3 3 19 7 3 4 4 79 Japan Univ Utah 4 4 6 1 4 9 6 16 7 13 4 9 6 6 1 79 USA King's Coll London 4 6 9 29 17 14 10 1 6 3 1 81 UK Univ Nottingham 1 6 10 21 21 10 17 7 4 1 82 UK Boston Univ 3 1 6 3 6 11 1 4 3 13 14 10 10 10 83 USA … … … Legend: Frequency lower 15% Frequency between 15 and 30% Frequency between 30 and 50% Frequency greater than 50%
Sets of indicators for policy making Page 63
The problem with statistics for Lisbon
Bruxelles, March 10-11 2011 10
The European Commission needs communicating with statistics: 1➡ To create an identity and a narrative for the goals of the EU policy; 2➡ To engage an increasingly literate constituency (*‶; 3➡ To promote transparency and accountability; 4➡ To fill a space otherwise taken by other actors or stakeholders➡
(*‶ See Stiglitz report
Bruxelles, March 10-11 2011 11
How are we doing with EU2020 ?
Sets of indicators for policy making Page 64
How are we doing with EU 2020
Bruxelles, March 10-11 2011 12
Statistics for policy: three models A rationalrational----positivistpositivist model for the use of indicators and policy (good quality statistics underpin good policies‶ DiscursiveDiscursive----interpretiveinterpretive model (statistics contribute to a process of framing of and focusing on an issue among the many competing for public's attention‶ Strategic model (statistics is used by parties competing for a given constituency‶➡
see Boulanger, P-M➡, Political uses of social indicators: overview and application to sustainable development indicators➡ International Journal of Sustainable Development , 101010 (1,2‶:14-32, 2007➡
How are we doing with EU 2020
Bruxelles, March 10-11 2011 13
But: It is possible to disentangle evidence based policy from policy based evidence?
see Benoît GODIN on eugenics and the birth of R&D stats: The Culture of Numbers: From Science to Innovation, INRS, Montreal, Canada, Communication presented to the Government- University-Industry Research Roundtable (GUIRR‶ US National Academy of Sciences, Washington, May 21, 2010➡
Sets of indicators for policy making Page 65
How are we doing with EU 2020
Bruxelles, March 10-11 2011 14
Improvements in clarity: the five benchmarks eight indicators framework➡
Better methodologies for creating indicators
More analysis at the regional level
Bruxelles, March 10-11 2011 15
+ Improvements in clarity + Benchmarks
Sets of indicators for policy making Page 66
Bruxelles, March 10-11 2011 16
+ Target decomposition into national targets + Education and equity
Bruxelles, March 10-11 2011 17
Better methodologies
Sets of indicators for policy making Page 67
Regional competitivenessBruxelles, March 10-11 2011 18
analysis at the regional level
Bruxelles, March 10-11 2011 19
Methodology from the GCI
Sets of indicators for policy making Page 68
Bruxelles, March 10-11 2011 20
Regional Innovation
Sets of indicators for policy making Page 69
Statistics in the form of sets of indicators
Ales CAPEK
OUTLINE OF PRESENTATION
The place of indicator sets within official statistics Issues and features of indicator sets used for policy making Selected quality aspects Use of indicator sets and borders of official statistics
PYRAMID OF STATISTICAL INFORMATION
Not labelled as Composite indicators „Official Statistics“ Specific Indicator Purpose (sets)
Multi- Accounting systems Purpose
Primary information from surveys, registers etc.
Sets of indicators for policy making Page 70
INDICATOR SETS FOR POLICY MAKING
Policy frameworks as a reason for presenting and disseminating statistics in the form of sets of indicators Relation to the objectives of the policy – targets (Europe 2020, Maastricht criteria) – performance indicators Use of indicators in a surveillance mechanism which includes policy reaction (EDP, EIP) High quality statistics needed to ensure credibility of the policy framework – choice of the indicators, audit-like powers of Eurostat Close cooperation between statisticians and policymakers
INTEGRATED COUNTRY SURVEILLANCE WITHIN EUROPEAN SEMESTER
EUROPE 2020 FIVE HEADLINE TARGETS
Europe 2020 Integrated Guidelines Stability and Growth Pact
Macro-economic Thematic Fiscal surveillance surveillance coordination
Macroeconomic 8 Headline Government finance imbalances and indicators data competitiveness indicators
Sets of indicators for policy making Page 71
SELECTED QUALITY ASPECTS
Coherence – achieved if indicators originate from a standardized data compilation/accounting framework (NA), this is however rare case and horizontal effort is thus needed to ensure coherence Parsimony – how many indicators for the policy framework? Not more than necessary. Europe 2020 – 8 indicators for 5 headline targets. EU SDS – more than 100 indicators
PRESENTING AND COMMUNICATING SETS OF INDICATORS
Specific presentation/communication tools – websites, scoreboards, dashboards – accessibility, user friendliness, meta-data Should statisticians be involved in analyzing the data or evaluating the trends indicated by the data? Where are the borders of official statistics? In practice elements of analysis and evaluation exist Examples: business cycle clock, Monitoring report of the EU Sustainable Development Strategy
Sets of indicators for policy making Page 72
Visual Storytelling applied to educational world statistics
Mikael JERN, Linnea STENLIDEN
Abstract We present innovative visualization that focuses on the most ancient of social rituals: “storytelling” – exemplified through telling a geovisual analytics story about a region’s development over time and shape the measure of economic growth and well-being. Discoveries that more engagingly draw us into reflections about the knowledge on how life is lived - and can be improved – from region to region and in addition let the reader dynamically participate in this interactive process and help advancing research critical to the dissemination of official statistics by means of web-enabled reporting tools. Geovisual analytics is a technique that can help illustrating high-dimensional statistical temporal data which for the eye are hard perceive or interpret. We introduce visual “storytelling” means for the author to import large spatio-temporal statistical data, explore and discern trends, create a story with snapshots and metadata and finally publish understandings and knowledge embedded in web page. The story guides the reader in the directions of both context and discovery through a highly engaging intuitive visual interface “vislet” based on cognitive principles and at the same time follow the analyst’s way of logical content. Value no longer relies solely on the content but also on the ability to access a better understanding of the information. In this paper we focus on the educational setting to improve the teaching in social science and increase the curiosity and learning among the younger generation for global development. We introduce our tool “Statistics eXplorer” here applied to the nations of the world based on a direct database link to the public World dataBank with 420 indicators for 1960-2009 that will help to improve citizens and in particular the young generations knowledge and understanding of a variety in national demographics, healthcare, educational, environment and economic structures and their performances over a longer time period.
Introduction
A well-educated young population is central to the social and economic well-being of regions and individuals. Education plays a key role in providing young people with the knowledge, skills and competencies needed to participate effectively in society. Official statistics is a rich and important source of information and have therefore an important role in education. Official statistics published with geovisual analytics (Andrienko, 2010) may help to improve and even change the terms and structures for learning about our society. What impacts will this have on the young generation?
Official statistics are statistics published by government agencies or other public bodies such as international organisations. They provide quantitative or qualitative information on all major areas of citizens' lives, such as economic and social development, living conditions, health, education and the environment. Official statistics can be found on web sites of national statistical agencies such as Statistics Sweden ( www.scb.se ) and international organisations such as the OECD ( www.oecd.org/statsportal ) and the World Bank (www.worldbank.com ). These massive temporal data are producing what is often called information overload. Statistics have often an unfortunate image of being boring – even though some of us know that they are in fact fascinating and exciting.
Smart growth Page 73
Figure 1: Seamless process from accessing data to publish knowledge and understandings on the Web.
Do teachers know about existing public statistics and its potential for a more engaging education? – Probably not sufficiently. Can they find them? – Not as easily as statisticians tend to think. And if they eventually get to them, do they actually understand them in such a way that they can use them in their educational activities? These are issues that are dealt with in this study. It concentrates on how to give our teachers innovative tools that can make national and regional statistics interactive visually understandable and useable to students.
We introduce innovative web-enabled geovisual analytics (Jern, 2010a) exemplified in our World eXplorer platform for exploring, collaborating and publishing statistics data (figure 1 and 2), facilitating storytelling aimed at producing statistical educational content in support of an automatic authoring process. The author, in this case a teacher, would simply press a button to publish the gained knowledge and understandings from the visual interactive discovery process and let the students read the teacher’s story and at the same time interact with the visualized content. We exploit our latest research focusing on the most ancient of social rituals “storytelling” - telling a story about a region’s development over time and shape the measure of economic growth and well-being. Discoveries that more engagingly draw us into reflections about the knowledge on how life is lived - and can be improved – compare nations and local regions and in addition let the student dynamically participate in this process. The effectiveness of educations rests in many ways upon educators to empower their students to become effective learners and knowledge creators, here through a “Vislet” that is embedded into a web page or blog that now becomes an interactive geoanalytics learning experience: http://www.ncomva.com/?p=978
Smart growth Page 74
Figure 1: World eXplorer with 3 time-linked views showing the world “fertility rates” during 1960-2008; map, scatter plot (fertility rate vs. age 0-14) and time chart; comparing 4 countries Nigeria, South Africa, China and Italy. The story is published to the right side and includes linked snapshots. Students learn that central Africa maintains a high fertility rate (Nigeria), while South Africa starts in 1960 at the same level but then has a reduced trend. The students can read the story (right GUI panel), interact and change indicators to discover reasons behind this trend and knowledge. Interact with the Vislet: http://www.ncomva.com/?page_id=297
A storytelling mechanism is initiated (figure 3) for the teacher to: 1) access statistical data from the World dataBank database through a direct API interface; 2) explore and make discoveries through trends and patterns and derive insight - gained knowledge is the foundation for 3) creating a story that can be 4) shared with colleagues and reach consensus and trust. Visual discoveries are captured into snapshots together with descriptive metadata and hyperlinks in relation to the analytics reasoning. The teacher can get feedback from colleagues then adopts the story and 5) finally publishes “tell-a-story” using a “Vislet” that is embedded in educational blogs or HTML pages providing students with an interactive visual learning experience.
With the ubiquitous availability of geovisual analytics the time has come to explore the possibilities for educators to incorporate these tools into a variety of subject courses and teaching practices (Kinzel & Wright), 2008). The potential for educators to harvest theses powerful tools, to present and explore scientific data sets, ought to be offered and in focus for further investigation.
We consider interactive tools for teaching such as GIS, visualization, computer models and animations allow the educator and the student to manipulate the environment and the outcome of the learning process are effective (Solem, et al. 2009). There are research and usability testing of geovisualization tools, but there is a lack of studies of young students learning processes. A better understanding of how educators and their students can learn by and elicit better user understanding and participation by exploiting these tools is of importance. We are implementing these tools – geovisual analytics – applied in social science – to help and engage educators to communicate progress initiatives, measuring economic, social, educational, health and environmental developments to young students to:
Smart growth Page 75
• Examine the students’ development of knowledge and understanding by using visual analytic storytelling methods in an educational setting • Investigate teachers experiences when using those methods • Contribute to further development of geovisual analytics for educational purposes according to the pedagogical findings.
The study is delimited to investigate this in all public junior high schools in one municipality in Sweden. The tool will be introduced to teachers and their students (age 13 – 15 years) in social science.
Smart growth Page 76
knowledge (Vygotsky, 1986). The learner selects and transforms information, constructs hypotheses, and makes decisions, relying on a cognitive structure to do so. A cognitive structure (i.e., schema, mental models) provides meaning and organization to experiences and allows the individual to "go beyond the information given" (Bruner, 1973). In order to accomplish this, geovisual analytics must try to make connections between knowledge the learner has and the knowledge being taught. An interdisciplinary research in cognition and geovisual analytics includes therefore pressing research questions and theoretical perspectives.
System Implementation
Statistics eXplorer is based on our in-house developed component toolkit adapted for Adobe’s Flash basic graphics (Quan, 2011) and does not require installation of any other software and will run anywhere. The toolkit facilitates innovative methods from information and geographical visualization such as the choropleth map, dynamic histogram, table lens, parallel axes plot “profile plot”, scatter plot, time graph, and pie and time glyphs, flow map (trade and migration) applied and customized for statistics data. Interactive features that support a spatial analytical reasoning process are applied such as tooltips, brushing, highlight, visual inquiry and conditioned statistics filter mechanisms that help detecting outliers. Data are normally preloaded with a set of basic indicators such as demographics, economic indicators, education statistics etc. but the user can also load external data through optional database interfaces such as SDMX, PC-AXIS or other API solutions to be mixed with preloaded data.
In order to detect complex patterns it is convenient to view statistics data through a number of different visual representations simultaneously (figure 2), each of which is best suited to highlight specific features. Any filtering, highlighting or colouring made in one of the linked views is transmitted to all the others.
The conceptual data model for our Statistics eXplorer platform can be seen as a data cube with three dimensions: space, time and indicators (figure 6). The spatial dimension is represented by the regions and the Figure 5: Statistics eXplorer provides an open indicators are various indicators data architecture for flexibility. This is data (GDP growth, elderly dependency interface is based on a programmed API to the rate, etc). Time is the period or point in World dataBank. time to which the data refer.
The general method for finding a value in the cube is by its position (space; time; indicator) or (where; when; what) and fast access time is essential for motion graphs. Space-time- indicator awareness means that the data cube can be analysed and visualized across all three dimensions simultaneously.
Smart growth Page 77
Figure 6: Statistics eXplorer conceptual data model based on (when, what, when) events
Statistics eXplorer performs this task by integrating and time-linking all its motion graphs: choropleth map, scatter plot, dynamic histogram, flow map, table lens, data table, glyphs, time graph, parallel axes plot and regional distribution plot.
World eXplorer applied in Education
The use of geovisual analytics (Andrienko, 2010) have in many ways revolutionized the way we are able to experience and explore our world. A primary target group for our storytelling is the educators and their students. By introducing the use of this tool in their process of learning and knowledge construction they got the opportunity to discover and take advantage of what this technology offer. Our geovisual analytics supplies possibilities for the educator to orchestrate the educational planning and teaching. The World eXplorer platform (World eXplorer, 2011) is customized from our Web-enabled GAV Flash class library (Quan, 2011), programmed in Adobe’s object-oriented language ActionScript and includes a collection of innovative geographic and information visualizations adapted to statistics data handling. The schema shows a seamless integration of a teacher database access, authoring tool, storytelling and publishing interactive education documents for official statistics and it is
based on: • Authoring : data provider (spreadsheet and database); data manager; visualization methods (choropleth map, scatter plot, table lens, histogram, parallel axes plot, time graph, data table); coordinated and linked views; map layers; analytic tools (dynamic query, filter operation); regional categorization; profile plot; highlight regions; motion charts; dynamic colour scale and legend. • Storytelling : snapshots capture mechanism; story editor; metadata with hyperlinks for analytical reasoning; import and export story. • Publisher: Vislet (widget); select visualizations used in a Vislet (map, scatter plot histogram etc), create HTML code; embed HTML code in web pages, wikis or blogs.
The methodological concept offers the educator to:
• Choose educational content : According to what content the teaching deals with statistical indicators and related geographical regions (countries, counties, municipalities etc) are uploaded to the platform. • Use the multiple linked views to simultaneously explore the content and highlight trends and knowledge through:
Smart growth Page 78
o An interactive map: possibilities to different interactive features that support a spatial analytical reasoning process such as tooltips, brushing, highlight visual inquiry and conditioned statistics filter mechanisms. o A motion chart: also offer the interactive possibilities to find patterns, connection and discover outliers among the indicators as well as show time series. o Time series: give opportunity to dynamically show indicators development over time.
• Produce an educational text : the educator can express (in her own language) a descriptive text and point out important spotlights of the content/indicators or even provide questions for students. • Create snapshots : in the educational text the educator is able to highlight different content and a link is created to the interactive map or motion chart.
Empirical Study at public junior in social science classes
With a major interest in studying human understanding and learning within complex technology mediated learning environments our paper builds on a number of analytical concerns and assumptions. The study has a socio-cultural perspective on learning together with perspectives on the significance of visual aspects on learning [5]. An empirical study was carried out at public junior high schools in a municipality in Sweden with altogether 100 students at the age 13 – 15 (grade 6 – 9) involved. The study is divided in different phases. In phase one the teachers were introduced to the tool and made educational plans according to the curricula, organizing the content and the task by involving use of World eXplorer and its storytelling methods for exploring demographics indicators during 1960-2009. This means that educational Vislets were produced based on indicators from the World dataBank and published in their own educational blogs on the internet. In phase two the students worked with the interactive Vislets reading the stories and interacted with a map and scatter plot (figure 2). The teachers were first introduced to the tool; they made educational plans according to the curricula, organized the content and the tasks by involving use of the World eXplorer platform.
Our case study is interdisciplinary and builds upon two different main research areas “technological development of the geovisual analytics tool” and the “educational context of students learning process” when using this tool. Within the later context a Usability Study was carried out to examine the effectiveness, efficiency and user satisfaction. The results show that the storytelling methods are usable within the school. The tool seems to be exiting, understandable and useful even for young students. It is efficient to students, it support their searching and the apprehension of connections between different kinds of statistical indicators. The user satisfaction among students was extensive at least used as brand new tool – the long lasting effects are however unknown (Stenliden & Jern, 2010). The following tasks were investigated: • How do conceptual and perceptual factors interact in learning with different representations? • How does learning differ with presented or constructed representations? • What are the costs and benefits of learning with interactive or dynamic representations? • What are the conditions under which learning is enhanced by combining textual and graphical representations?
Smart growth Page 79
Figure 7: Example of an interactive educational document based on public World dataBank indicators with educational text, map, motion chart, snapshots and time series – the methodological concept. For an interactive experience of the educational tool and teaching material click on the link; http://visletblog.blogspot.com/2011/01/ageing-population-in-europe-and-japan.html .
Conclusion
Within an international perspective our research builds on collaborating work with OECD since 2008 and we have supplied advanced and innovative statistics geovisual analytics technology to this organization http://stats.oecd.org/OECDregionalstatistics/. We have also been involved in the development of the PISA2009 profiles: http://stats.oecd.org/PISA2009Profiles/# and Vislet at: http://www.ncomva.com/?page_id=873 . The national Italian bureau ISTAT provides another very interesting and sophisticated learning material http://noi-italia.istat.it/ produced with our Statistics Publisher about the development and progress in Italian regions in Another user is the European Commission that have used Europe eXplorer for internal analysis of data from Eurostat. The research concerning the learning perspective as presented in this study is highly requested from the international research field of learning and instruction (for example European Association Research of Learning and Instruction). The special interest group (SIG2) of comprehension of text and graphics inside this research field, focuses on how learning is influenced by the form of representation that learners study. Focus has historically been on text and picture comprehension but given the explosion of representations made available since the introduction of graphical interfaces, the field now considers all forms of representation including but not limited to, text, pictures, graphs, diagrams, concept maps, animations, equations, virtual reality, information and scientific visualization, haptics, multimedia, hypermedia, and simulations. Research on learning, when using these aids, is essential. There is research of learning with multimedia environments in different experimental studies but there is hardly any research done of this in real school contexts i.e. in a socio cultural perspective. In Sweden there are hardly any studies considering didactic within social science education in schools.
Smart growth Page 80
All together the results of this study will give valuable contributions to the development of these research fields. The geovisual analytics technique introduced in this paper allows the teacher to communicate with student through interesting and important discoveries captured into snapshots together with descriptive text. Selected indicators and visual representations can be published together with their metadata, thus facilitating the comprehension of statistical information for educational purpose. We believe that this innovative storytelling technology can be useful for a next-generation educational dynamic book for learning about different phenomena in the world, as examples demonstrated in this paper. At the same time, the Vislet technique help developing agile on-line educational publications, which draw the attention on recent trends and inequalities.
The educational World eXplorer presented in this paper can be evaluated at: http://www.ncomva.se/flash/explorer/wbapi/# .
Figure 8: Example of two Vislets 1) ISTAT – the Italian national statistical bureau http://noi- italia.istat.it/ has developed this innovative web site Statistics Publisher providing an interesting example of a sophisticated learning material. 2) Exploring ageing population in Europe 1990-2008: http://visletblog.blogspot.com/2011/02/ageing-population-in-europe- 1990-2009.html .
Smart growth Page 81
Acknowledgements
This research and case studies were carried out by NCVA in close collaboration with OECD and Statistics Sweden who supplied data and comprehensive evaluation of the statistical storytelling system. The research is in part supported by funding from the Swedish Agency for Innovation Systems (VINNOVA), the “Visualization Program” and the “National Graduate School of Childhood, Learning and Didactics” both coordinated by the Swedish Knowledge Foundation. The authors thank the entire research team at NCVA, Linkoping University for valuable contributions, but also to the teachers contributing and the students participating in this study.
References Ainsworth, S., 2006. DeFT: A Conceptual Framework For Considering Learning with Multiple Representations, Learning and Instruction . 16(3), , pp. 183-198. Andrienko G., Andrienko N., Demsar U., Dransch D., Dykes J., Fabrikant S., Jern M., Kraak M., Schumann H., Tominski C. 2010. ”Space, time and visual analytics” , International Journal of geographical information science , 24(10):1577-1600. Bruner, J., 1973. Going Beyond the Information Given , New York: Norton. Gibson, J., 1969 . Våra sinnen som perceptuella system, Stockholm, Beckman, Stockholm. Goldman, S. R., Mertz. D. L., & Pellegrino. J. W., 1989. Individual differences in extended practice functions and solution strategies for basic addition facts. Journal of Educational Psychology , 81(4), pp. 481-496. Jern, M., 2010a. “Educating students in official statistics using embedded geovisual analytics storytelling methods”, Reviewed Proceedings in Eurographics 2010, Norrköping. Jern, M., 2010b. “Explore, Collaborate and Publish Official Statistics for Measuring Regional Progress”, Cooperative Design, Visualization, and Engineering 7th International Conference, CDVE 2010 , Mallorca, Spain, September 19-22, 2010. Lecture Notes in Computer Science, pp 189-198, ISBN-10 3-642-16065-4 Springer Berlin Heidelberg New York Jordan, B., & Henderson, A., 1995. Interaction analysis: Foundation and practice, The Journal of the Learning Sciences 4 (I), pp. 39 – 103. Kinzel, M., Wright, D., 2008. Using Geovisualizations in the Curriculum: Do Multimedia Tools Enhance Geography Education? Paper Number 1290, Environmental Systems Research Institute Education User’s Conference . Quan, H., Lundblad, P., Åström, T., and Jern, M., 2011. A Web-Enabled Visualization Toolkit for Geovisual Analytics Visualization and Data Analysis, Awarded best paper , SPIE: Electronic Imaging Science and Technology, Visualization and Data Analysis, Proceedings of SPIE, San Francisco . Stenliden, L., Jern, M., 2010. Educating official statistics using geovisual analytics storytelling methods, Reviewed proceedings, International Technology, Education and Development Conference INTED, Valencia. Solem, M., Foote, K., Monk, J., 2009. Aspiring Academics: A Resource Book for Graduate Students an Early Career Faculty . Prentice Hall. Vygotsky, L., 1986. Thougt and language . Edited by Kozulin, A., MIT Press, Cambridge. World eXplorer: http://ncomva.com 2011
Smart growth Page 82
Unveiling Europe's Digital Ambitions through Official Statistics: Past, Present and Future
Lucilla SIOLI 14 , Michail SKALIOTIS 15
1. Introduction
The ICT dataset is one of the richest in the Commission. The statistical approach used nowadays to the data collection is summarised in " Benchmarking Digital Europe 2011 – 2015: a conceptual framework ", endorsed by Member States in November 2009. It clarifies the main principles underpinning the Commission's approach to the collection of the information society statistics as well as a list of core indicators to be used for benchmarking.
Historically, Europe's strategic visions and policies for the information society were explicitly expressed in a number of important documents in the 1990's like the Delors Report (Dec. 1993), and the Bangemann Report-(May 1994). The first benchmarking reports were published in the framework of the eEurope initiative, dating back to 2000. With the adoption of the ' eEurope 2005: An information society for all' (May 2002), it was explicitly underlined that '…To improve the quality, measurement of eEurope 2005 indicators should make greater use of official statistics from the National Statistical Institutes and Eurostat….The Commission will propose this legal base before end 2002'. Since then, the European Statistical System has worked in close co-operation with the ICT policy makers and has developed two large scale annual surveys on ICT use (households and enterprises) which constitute a central data source of the annual benchmarking reports. These surveys have been used to measure progress in the following initiatives for the information society, "i2010" and the "Digital Agenda for Europe".
Today, the ambitions of Europe 2020 set the policy framework for the next decade, and the Digital Agenda for Europe reflects the renewed strategic vision of the EU for a digital Europe responding to the expectations of society and stakeholders. There is no doubt that the successful co-operation model of 'ICT policy & Official Statistics' of the recent past will continue to work in the future within the annual Digital Agenda Scoreboard , and that it could be successfully applied to other policy areas . While the focus of this paper is on recent experiences with regard to the interaction of policy and statistics in formulating and monitoring ICT policy targets in a fast changing environment, we also introduce the debate about the impact of ICT on official statistics. The emergence of a largely digital world characterised by an explosion of data and metadata in the form of digital footprints represents new challenges for official statistics.
2. Defining policy targets and monitoring of progress
The Digital Agenda for Europe 16 is one of the seven flagships of Europe 2020. It was launched in May 2010 to define the key enabling role the use of ICT will have to play if Europe wants to succeed in its ambitions for 2020. It consists of a list of 101 actions. Progress is measured in an annual Scoreboard against a set of key performance targets (see Annex). Most of the indicators for the targets were drawn from the Benchmarking Framework 2011-15 17 endorsed by EU Member States in November 2009. As Commissioner
14 European Commission, DG Information Society and Media, Head of Unit 'Economic and Statistical Analysis' 15 European Commission, Eurostat, Head of Unit 'Information Society and Tourism statistics' 16 COM(2010)245, available at http://ec.europa.eu/information_society/digital-agenda/index_en.htm 17 Available at http://ec.europa.eu/information_society/eeurope/i2010/docs/benchmarking/benchmarking_digital_europe_2011- 2015.pdf
Smart growth Page 83
N. Kroes emphasised, "Every European Digital" by 2015 includes broadband for all on the path towards ambitious 2020 broadband indicators (graph below), and a significant reduction in those who have never used the internet, taking into account that the take-up gap is mostly due to lack of skills such as digital literacy.
The 2011-2015 Benchmarking Framework had to allow for sufficient time for the design of the surveys by the National Statistical Institutes (NSIs) and was conceived well in advance of the publication of the Digital Agenda. Built on its predecessor, it foresees further developments though a mid-term review which will take place in 2012. The framework found the consensus of Member States whose discussions involved joint meetings of NSIs and national representatives of ministries in charge of ICT policies.
According to the Framework, the development of ICT and its impact can be described through a supply, use and impact framework:
• efficiency gains in the production of ICTs translate into a growing contribution of the sector to economic growth and into falling prices of ICT goods and services (supply) ; • decreasing prices stimulate investment prompting the take-up of ICTs by individuals, businesses and the public sector: take-up can further be described through readiness and use of ICT services and content applications (use) ; • diffusion of ICT contributes to the sustainable growth of the economy and to jobs, the efficiency of the public sector and the well-being of the population (impact) .
In the 2011-2015 benchmarking framework, Eurostat ICT surveys represent the main source of statistical information, with additional data in the area of telecommunications provided by the Communication Committee 18 . The ICT use surveys are structured around core indicators (to be maintained for tracking development over time) and special modules focusing on different topics each year. When specific policy needs cannot be covered by official statistics,
18 The Communications Committee was established to assist the Commission in carrying out its executive powers under the regulatory framework governing telecoms in the EU. It is composed by representatives of national ministries for telecommunications and coordinates the data collection by the National Regulatory Authorities.
Smart growth Page 84
information is collected through ad-hoc surveys/studies, paying particular attention to the issue of quality and reliability.
As the fast changing world of ICT poses some adaptation issues for official statistics, these studies have become increasingly more important over time. This causes tension as samples in ad-hoc surveys are significantly smaller than in Eurostat, making results less reliable and more open to controversy. Nevertheless, ad-hoc studies enjoy flexibility in the choice of the targeted audience and in the design of the survey. From this point of view they contribute imperfect but often necessary information to the policy design and its implementation. One of the issues that has been extensively discussed in recent fora (comprising policy experts and official statisticians) is the extent to which official statistical agencies can provide some kind of quality labelling of the results of such studies. With the emergence of numerous new 'digital' data sources and future trends in cloud computing and cloud services, the issue of synergies between official statistics and private data sources will gain importance in the next few years.
3. A brief history of the future of ICT statistics
In this section we attempt to outline some medium to long term developments in ICT statistics, which may as well influence other fields of official statistics. Based on the experiences from the recent past in terms of the interaction of ICT policy and official statistics , and an analysis of the main drivers and trends in ICT policy, technology, research and statistics it is possible to tell a 'plausible' story for Official Statistics 2.0 19 .
A user who navigates in Eurostat's website 20 and is looking for Information Society statistics will soon realise that ICT official statistics go hand in hand with EU policy developments. The indicators and tables are structured under thematic names which are almost identical to those of the benchmarking frameworks and EU policies in the ICT fields, namely the eEurope 2005, the i2010 and Benchmarking Digital Europe:
On the same site, users can find a number of recent and past press releases and analyses which focus on ICT policy issues, such as:
19 For a definition of Official Statistics 2.0 see for example the paper 'Timeliness and Accuracy in Official Statistics 2.0' at http://www.nso.gov.mt/docs/MichailSkaliotis%20_Eurostat.pdf 20 http://epp.eurostat.ec.europa.eu/portal/page/portal/information_society/data/database
Smart growth Page 85
Similarly, when looking at policy reports, for example the annual Information Society Reports 21 or the Europe's Digital Competitiveness Reports 22 , we observe that several of the indicators and analyses are derived from the 2 ICT surveys which are carried out annually in the EU, as well as in Norway and Iceland. For example, the following graph, which illustrates the adoption of RFID technologies by enterprises, was extracted from the 2010 Competitiveness Report.
Eurostat's ICT usage surveys represent a very good example of a comprehensive and timely response of official statistics to 'policy needs'. The development and functioning of these surveys is the outcome of intense and continuous collaborative efforts and structures at many fronts: policy, statistical, legislative, administrative and budgetary.
21 http://ec.europa.eu/information_society/eeurope/i2010/docs/annual_report/2007/i2010_ar_2007_en.pdf 22 http://ec.europa.eu/information_society/digital-agenda/documents/edcr.pdf
Smart growth Page 86
3.1 How does this co-operation model work in practice?
The starting point for statisticians in this regard has always been the policy framework, which provides answers as to 'what' should be measured. Then an agreement has to be reached with regard to the 'core' items (to be surveyed every year) and the 'special modules', i.e. ad hoc policy themes to be surveyed at irregular intervals. Examples of special modules are: eSkills, Security and Trust, eGovernment, Mobile internet use, eCommerce and eBusiness processes.
The lead time for preparing a module (and consequently a survey for a given year) is approximately 18 months; as a matter of fact, this (seemingly long) period represents a rather tight schedule because many activities have to be accomplished in an extremely orderly and timely manner: discussions with DG INFSO and other Commission departments, technical meetings with Task Forces and Working Groups, preparation of annual legislation, formal inter-service consultation of other departments of the Commission, agreement of Member States through the European Statistical System (ESS) Committee, approval of the legislative proposal by the Council and the Parliament and adoption by the Commission. It should be underlined that at any step of this process we may encounter difficulties which may delay the whole exercise.
Despite the tight schedule, this model of co-operation has been successful in many respects: (i) the EU ICT surveys constitute a unique harmonised international source in this area which is considered as 'best practice', (ii) it has always been possible to deliver up-to-date results (publication of results within the same year in which the surveys are carried out), (iii) the surveys have managed to cope with the requirements for continuity as well as flexibility (ad hoc policy themes), and (iv) response and administrative burdens have been kept stable at reasonable levels (any new requirements are carefully balanced against other survey-items which are withdrawn).
3.2 Future outlook and short-term challenges 23
There is no doubt that the successful co-operation model of 'ICT policy & Official Statistics' of the recent past will continue to work in the future within the annual Digital Agenda Scoreboard , and that it could be successfully applied to other policy areas . Taking into account the ambitions of the Digital Agenda and the other flagship initiatives of the Europe 2020 strategy, together with a forward looking analysis of other key drivers and in particular trends in technology and society we can identify a number of important challenges that will shape the future of official ICT statistics.
• The quantification of the value of the "Digital Single Market" and its definition, including two aspects :
(i) Although basic indicators/data on the percentage of internet users downloading music/books/film/video are available, data on online content and the media industry's activities based on online distribution are largely missing.
(ii) The source of general indicators about eCommerce (percentage of firms/individuals selling online in the EU, nationally and cross-border, or most popular products) is the Eurostat surveys on ICT. More effort however needs to be made to improve our understanding of these results, in particular when purchases and sales concern digital goods and services, bearing in mind that consumer' surveys are affected by perception and are not always the best instrument to collect this kind of information.
23 The views expressed in this paper are those of the authors and do not necessarily reflect the official opinion of the European Commission.
Smart growth Page 87
• To progress on the definition of indicators for ICT and the low-carbon economy , possibly in an international context (work is ongoing with the OECD) and understand whether the current ICT surveys are effectively appropriate for the collection of this kind of information. • To review current datasets , as the ICT world evolves quickly. Currently efforts are focussed on broadband (to better measure consumer prices, speeds and quality, mobile broadband and next-generation access technological breakdowns) and on eInclusion/digital competences targets. • To gather more information on the use of mobile communications , including mobile access/services/commerce, including online/mobile payments. This will be the theme of a special module in the 2012 Eurostat surveys. • To attempt to use the internet as a source of data so as (i) to have better information on the use of the internet and (ii) to improve the data collection in other fields (health for example). • To strengthen the international comparison with non-European countries. The EU appears more advanced in the data collection in the information society area than many non-EU OECD and ITU member countries. As a result, international comparisons generally concern less sophisticated indicators. • To make all the data collected available in a organized and user-friendly environment on the web . This is a good example of open data (availability of public-sector information) to the benefit of researchers, colleagues and other interested bodies. The upcoming Digital Agenda Scoreboard will be a first step towards this objective. On the policy side, the priorities of Europe2020 and the Digital Agenda in particular constitute a substantial list of topics for the ICT statistical agenda for the next 5 to 10 years: eCommerce, interoperability, security and trust, digital literacy, skills and inclusion, ICT and environment, eGovernment, and cross-border services will certainly be amongst the 'special modules' of the ICT usage surveys. In the past, it had not been possible to pilot in advance alternative 'questions and variables' for the yearly special modules; in the Digital Agenda round of ICT surveys however, it is our intention to pre-test a number of modules in 2012 and 2013 in order to identify the best possible specifications of the corresponding thematic survey units, ensuring therefore high quality standards in this regard. Increased public attention to the monitoring of Europe2020 targets and its flagship initiatives implies that the quality of official statistics used in the benchmarking exercise (e.g. DAE scoreboard) has to be at a high level.
While the political agenda defines the 'what', there are many more drivers influencing statistical decisions as to 'how' ICT statistics should be collected, processed and disseminated in the future. One of the most influential (statistical) drivers is the so called 'vision'—a Commission Communication 24 on the production method of EU statistics: a vision for the next decade —which has become the framework for actions of the ESS (European Statistical System); moreover, most of the National Statistical Institutes are experiencing resource constraints and declining response rates in a number of surveys both for households / individuals and enterprises. Therefore there is a lot of pressure for greater efficiency and rationalisation of the production system of European statistics. On the other hand ICTs offer several new alternatives for statistics, and the underlying trends in the transition of our societies towards greater digitalisation are requiring a completely new professional paradigm for statisticians. We shall try to illustrate with a few examples how ICT statistics will look like in a largely digital world.
24 http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=COM:2009:0404:FIN:EN:PDF
Smart growth Page 88
3.2.1 Using the Internet as a Data Source
There are several reasons that make us investigate the potential of internet for statistics on ICT: First, the internet is becoming an extremely rich data source that can reduce substantially the response burden on individuals and businesses 25 . Second, the internet is the predominant network infrastructure of tomorrow's digital world; internet is currently attracting large funding for cutting edge multi-disciplinary, multi-stakeholders / international research projects. EU policy and legislation are seriously making progress by creating an 'enabling' and stimulating environment for the development of the 'future internet'. Third, we already have some early experiences and pilot initiatives in exploring the use of the internet as a data source for statistical purposes. We can actually build on these experiences and launch more extensive testing, experimentation and evaluation of various alternatives.
It is worth mentioning here the original work carried out by Dialogic on behalf of the Netherlands Ministry of Economic Affairs 26 through a research project entitled 'Internet as a Data Source (IAD)'. The outcome of the project is described in the publication 'Go with the dataflow' and it is downloadable from the website of the ministry http://www.ez.nl/dsresource?objectid=157266&type=PDF . The research questions of the project were (a) to identify new data and indicators derived directly from the Internet and describe new phenomena associated with EDE (emerging digital economy), and (b) to explore and assess the usefulness of the various IaD methods for deriving new, extra and substitute data for the EDE. These questions are addressed through 8 case studies covering both the 'old economy' (established markets & phenomena / low levels of digitalization) and the 'new economy' (emerging markets & phenomena / high levels of digitalization). Various IaD methods are tested and evaluated in terms of their statistical pros and cons, as well as in terms of their usability and disadvantages . The typology of the experimented methods included three strands of measurements: (i) User-centric measurements (spyware, traffic monitoring at operating system level), (ii) Network-centric measurement (Deep packet inspection), and (iii) Site-centric measurement (spiders).
What do these approaches imply for our future statistics on ICT use? At a first glance they look very promising. Think of the current surveys on ICT usage by households and individuals. The main subject is 'what people are doing with ICT, how often they use internet, whether they purchased goods and services via internet and details of those purchases, usage of e-Government services, types of devices used, mobile internet use and ubiquitous connectivity, etc'. In terms of timeliness, direct electronic observation of ICT usage is clearly superior to a traditional household survey. Moreover, instantaneous electronic transmission of data with semantic meaning will enable the automated processing and analysis of data much faster. Accuracy will also be superior because there will be a perfect correspondence between 'what we want to measure' and 'what is actually recorded', for example how long users spend online. In a traditional inquiry, accuracy might suffer from memory effects, proxy responses given by other members of the household, manual recording, and limitations imposed by the small number of details that are usually collected (NSIs try to keep response burden to low levels by reducing the number of questions).
However, the current state of these methods (in terms of statistical robustness) is very primitive and we need to tackle several issues of concern: methodological, regulatory (legal) and privacy being the most important. Nevertheless, these concerns, while taken seriously, should not prevent official statistics from taking a forward looking approach through additional collaborative research and experimentation in this domain. The Dutch IaD study
25 During the last 5 years the European Statistical System has been under severe pressure to reduce the response burden (particularly for enterprises) due to statistical inquiries 26 The main report, its annexes and information about the consultation process can be found on the website of the ministry: http://www.ez.nl
Smart growth Page 89
includes a section on policy recommendations in which there is a 'plea' addressed to NSIs to play the function of a clearing house for Internet-based statistics. In its final words, the report invites publicly funded statistical agencies as well as Eurostat and the OECD to develop a roadmap for innovative methods and innovative statistics in this regard.
A very recent development along these lines is the launching of a study by the European Commission (DG INFSO) 27 entitled Statistical methodologies on Internet as a source of data gathering (SMART 2010/0030) . Eurostat is planning to follow up the results of this study towards the development of an appropriate framework for internet based ICT statistics. It is also expected to have a scientific debate on this topic at the next ISI World Statistics Congress 28 which will take place in Dublin in August 2011; Eurostat and OECD have proposed to co-organise a 'special topic session' on Analysing Internet Traffic Flows and digital footprints for statistical purposes.
3.2.2 Statistics on the use of eServices
During the last years we are all observing and experiencing the fact that a number of public and private services are increasingly being offered via the internet at a high level of sophistication. A good example in this regard is the widely known as eGovernment services (tax declaration, registration of change of address, health services, social security declarations, public procurements, and many other types of advanced interactions of citizens and businesses with public authorities). In properly designed systems, statistics on the use of eGovernment services can be instantaneously produced at any frequency, be this daily, weekly, monthly, quarterly or annually. The majority of web-based systems, even the most simple, include by default in their functionalities the ability to produce basic statistics on the use of the web-services. In other words, statistical functionalities are embedded in the design of the system. A fully-fledged embedded statistical design in the use of eGovernment services however would require the capability of the system to link different data bases through official identifiers (PINs or similar) in order to be able to produce useful analyses by socio-economic, demographic, educational and other characteristics of the users of these services. Today, these statistics are produced on the basis of (costly) household and enterprise surveys, with a delay of 1 year or more, and with very limited details that are not sufficient for proper policy planning; accuracy is further hampered by the fact that respondents--in enterprise surveys--are normally IT specialists who do not necessarily know how the various departments of the enterprise interact with government services 29 . In a future OS 2.0 environment, they will be produced faster, at minimal cost and much higher detail and accuracy. Another issue to consider in this context is the fact that the deployment of new technologies in many activities of every day life is often creating some new statistical demands that did not exist before . For example, thinking of eServices (both government and private), one of the several policy issues relates to the accessibility of the government / private websites for persons with disabilities. In other words, a relevant policy question in this regard would ask whether the ' Web content of government and/or businesses sites is accessible to people with disabilities, and if so to what degree'? The traditional approach to answering this question would be to organise a dedicated sample survey of government and enterprise establishments, and administer such a survey via a postal, telephone, web, face-to-face or mixed-mode interviewing or other approach. However there is an alternative automated way to collect this information in a faster, more accurate, cheaper and lower-burden manner. A Web crawler (similar to the site-centric approach mentioned above) can be designed in a way that automatically examines a large number of web pages and randomly selects and
27 http://ec.europa.eu/information_society/eeurope/i2010/docs/studies/invitation_to_tender_statistical_methodologi es.pdf 28 http://www.isi2011.ie/content/ 29 This statement is based on discussions with survey experts from a number of member states that have carried out cognitive testing in this regard.
Smart growth Page 90
evaluates a sub-sample; compliance of web sites can for example be tested against international standards like the WCAG1.0 standards 30 .
3.2.3 from cloud computing to cloud Statistics?
It is almost inevitable that developments in cloud computing will impact on all kinds of services including official statistics. While the deployment of cloud computing will be gradual, we can already today anticipate its impact on ICT statistics. How will the ICT surveys operate in an environment of cloud statistics within the European Statistical System? First of all, there will be no transmission of data as such, i.e. all the treatment, validation, editing and storage will be entirely distributed at the national data warehouses. Eurostat will offer the services of a 'hub' and 'cloud' at the same time. The 'hub' functionalities will translate and transmit automatically users' demands to the respective national data centres which subsequently will generate (pull) and send the corresponding data to a VDB (Virtual Data Base); the 'cloud' functionalities will provide StaaS (Statistics as a Service)—an intelligent and user-friendly interface—allowing simple and advanced statistical analyses and reporting.
Such a scenario is based on the assumption that internet and cloud computing are becoming GPTs (general purpose technologies) similar to what we have seen happening with energy and electricity. The above presented graph looks simple but it certainly hides several issues that must be resolved at technical and governance levels. Again, we would like to underline that while all these concerns (privacy, security, etc) must to be considered seriously, they should not prevent the international statistical community from preparing the statistical system of tomorrow.
4. Concluding remarks
During the last 10 years, ICT policy and official statistics in the EU have worked hand in hand. The successful co-operation model of 'policy & statistics' will continue to provide the core input for measuring progress with regard to the key performance targets of the Digital
30 For a detailed description of the guidelines see: http://www.w3.org/TR/WCAG10
Smart growth Page 91
Agenda. Topics of particular interest to the Digital Agenda will be studied and measured in depth via 'special thematic modules' in Eurostat's ICT usage surveys.
The fast changing technology and the transition towards a largely digital society in which enormous volumes of data and metadata are produced in the form of 'digital footprints' represents new challenges and responsibilities for official statistics. It is therefore fundamental that the international statistical community exploits the potential of these emerging digital data sources for the purpose of official statistics.
5. References
1. Dialogic: Go with the dataflow! Analysing the Internet as a data source, The Hague, April 2008, available at http://www.ez.nl , last accessed on 01 March, 2011. 2. European Commission: Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, A Digital Agenda for Europe, Brussels, 19.05.2010, COM(2010) 245. 3. European Commission: Communication from the Commission, EUROPE 2020, A strategy for smart, sustainable and inclusive growth, Brussels, 3.3.2010, COM(2010) 2020. 4. European Commission: Communication from the Commission to the European Parliament and the Council on the production method of EU statistics – A vision for the next decade, Brussels, 10.08.2009, COM(209) 404. 5. European Commission: Europe's Digital Competitiveness Report 2010, ISBN 978-92- 79-15829-2, Publications Office of the European Union, Luxembourg, 2010. 6. Skaliotis, M.: Official Statistics in the era of Ubiquitous connectivity and Pervasive Technologies, International Statistical Conference 'Statistics – Investment in the Future 2', Prague, 14-15 September 2009, last accessed on 01 March, 2011 at: http://www.czso.cz/conference2009/proceedings/data/technology/skaliotis_paper.pdf
Smart growth Page 92
Annex: Digital Agenda's Key Performance Targets
These indicators are mainly drawn from the Benchmarking framework 2011-2015 31 endorsed by the EU Member States in November 2009.
1. Broadband targets:
• Basic broadband for all by 2013: basic broadband coverage for 100% of EU citizens. (Baseline: Total DSL coverage (as % of the total EU population) was at 93% in December 2008.)
• Fast broadband by 2020: broadband coverage at 30 Mbps or more for 100% of EU citizens. (Baseline: 23% of broadband subscriptions were with at least 10 Mbps in January 2010.)
• Ultra-fast broadband by 2020: 50% of European households should have subscriptions above 100Mbps. (No baseline)
2. Digital single market:
• Promoting eCommerce: 50% of the population should be buying online by 2015. (Baseline: In 2009, 37 % of the individuals aged 16-74 ordered goods or services for private use in the last 12 months.)
• Cross-border eCommerce: 20% of the population should buy cross border online by 2015. (Baseline: In 2009, 8 % of the individuals aged 16-74 ordered goods or services from sellers from other EU countries in the last 12 months.)
• eCommerce for business: 33% of SMEs should conduct online purchases/sales by 2015. (Baseline: During 2008, 24% and 12% of enterprises was, respectively, purchasing/selling electronically, for an amount equal to or greater than 1% of the turnover/total purchases.
Single market for telecoms services: the difference between roaming and national tariffs should approach zero by 2015. (Baseline: In 2009, the roaming average price per minute was 0.38 cents (call made) and the average price per minute for all calls in the EU was 0.13 cents (roaming included).
3. Digital inclusion:
• Increase regular internet use from 60% to 75% by 2015 and from 41% to 60% for disadvantaged people. (Baseline figures are for 2009).
• Halve the proportion of population that has never used the internet by 2015 (to 15%). (Baseline: In 2009, 30% of individuals aged 16-74 had never used the internet.)
4. Public services:
• eGovernment by 2015: 50% of citizens using eGovernment, with more than half of them returning filled in forms. (Baseline: In 2009, 38% of individuals aged 16-
31 For more information see Benchmarking framework 2011-2015 ; This is a conceptual framework for collection of statistics on the information society as well as a list of core indicators for benchmarking.
Smart growth Page 93
74 had used eGovernment services in the last 12 months, and 47% of them used eGovernment services for sending filled forms.)
• Cross-border public services: by 2015 online availability of all the key cross- border public services contained in the list to be agreed by Member States by 2011. (No baseline)
5. Research & innovation:
• ICT R&D increase: Double public investment to €11 billion. (Baseline: ICT government budget appropriations or outlays on R&D (ICT GBAORD) was 5,7 billion nominal euros in 2007.)
6. Low Carbon Economy:
• Promotion of low energy lighting: By 2020 at least 20% overall reduction in energy use on lighting. (No baseline.)
Smart growth Page 94
Toward a headline innovation indicator for Europe 2020: the economic weight of high-growth enterprises in innovative activities
Matthieu DELESCLUSE
Smart growth Page 95
Smart growth Page 96
Smart growth Page 97
Smart growth Page 98
Summary of the Session
Antonio ARGÜESO
Let me start with something obvious. What do we need to measure smart growth?…smart statistics, and it means a smart use of available information to build relevant statistics and also a smart selection of indicators.
In the session yesterday there were some examples of this. One of the issues was digital Europe. The first thing we can say is that Information and Communications Technology (ICT) surveys go today like clockwork. They have been used to measure progress during the last decade. Today, iniciatives like “Internet for all” or the increase of digital public services drive the change for ICT statistics during the next years. A key source of information is the internet though there are many issues to solve (methodological but even more on the side of privacy…). Internet Traffic measurement can provide even better results than asking to the users not only on the frequency of usage but also on many respects about the type of interaction with the network. This is definitely something we must work in during the coming years.
In the field of R&D and innovation, the “3%” target, based on the indicator on share of GDP has been used for a long time. This is not a perfect indicator of smart growth. As it was said yesterday, it is not so evident that the more an economy spends in R&D the better it performs in terms of growth. A High-Level Panel set up by the Commission Proposed a new indicator: Weight of fast- growing innovative enterprises in the economy
There was an interesting discussion yesterday on this new indicator that can be translated somehow to the selection process of any other indicator used as headline for a given field or policy.
-this indicator has many advantages: • Is a very clear measure of the dynamism of the economy and of its potential to create qualified jobs • It may cover the whole range of innovative economic activities, including non- technological innovation • Focuses on an important driver of structural change
Smart growth Page 99
One of the things that we can say on the other hand is that there isn´t today a definite procedure to calculate such indicator and some ways of calculating it were presented.
But maybe the most important question raised during the debate was not only about how to calculate it but about the relation between innovation and growth. It is clear in principle, we all may think of google, that innovative businesses may grow fast but it is also clear that this is more likely to happen in the -so to say- new sectors than in others. The example of cars production, raised during the debate, was very clever. Innovation is needed in the automobile sector not to grow, but to survive. So a lot of innovation doesn’t necessarily bring high growth, nor even growth at all.
It illustrates the situation in general when thinking of headline indicators as “magic solutions”. As was mentioned many times yesterday it’s useful to visualize political priorities but not perfect to measure the complexity of the real world.
So we may arrive to the conclusion that we must be careful not only in the selection of indicators but also once selected, knowing that no indicator is perfect, in the analysis we make out of them.
There is an important chapter of “smart growth” we didn’t looked at in depth during the session, it is education, so let me say a few words of my own experience. We may agree that this is the most important element in the long term for smart growth, an even for growth in general, because it affects all others. There are initiatives at the European level like “youth on the move”, or academic excellence, that bring new needs in the statistical information on education. I highlight this aspect because for some other policies the statistical infrastructure is already there. In my opinion there’s room for improvement in the European statistics measuring the match between university (or other professional qualifications) and labour market. There are national initiatives in Europe and outside Europe but the European comparability would be a giant step in this field. There is a very concrete and strong demand in this field in Spain and I think it must be the same in other countries, so this is something we could think about in the next years.
Smart growth Page 100
Sustainable growth: indicators for green growth
Robin MIEGE
Sustainable growth Page 101
Sustainable growth Page 102
Sustainable growth Page 103
Lucio GUSSETTI
The CoR has recently considered the issue of statistics in support of local and regional decision making in its opinion on "GDP and beyond," CdR 163/2010 fin, adopted on 5 October 2010.
In this opinion the CoR has noted that GDP is not an accurate measure of the ability of a society to tackle issues such as climate change, resource efficiency, quality of life or social inclusion; therefore, the CoR has proposed that the indicators selected to orient the framing and drafting of policies and public strategies comply with the priorities of the EU 2020 Strategy.
The most obvious examples of such indicators are those for which the European Council has set headline targets to fulfil the goals of the Europe 2020 strategy. They are well know, but worthwhile recalling; employment rate (75% of the 20-64 year-olds to be employed); innovation (3% of the EU's GDP - public and private combined - to be invested in R&D/innovation); greenhouse gas emissions 20% lower than 1990 (or even 30%, if a satisfactory international agreement can be achieved to follow Kyoto); 20% of energy from renewables, 20% increase in energy efficiency; reducing school drop-out rates below 10%, and: at least 40% of 30-34–year-olds completing third level education (or equivalent); Poverty / social exclusion: at least 20 million fewer people in or at risk of poverty and social exclusion.
It would be particularly relevant to note with regard to this workshop that the CoR has in its opinion stressed that it is necessary to improve the methodology used to obtain more up-to-date comprehensive information that better matches reality, allowing for the use of indicators to facilitate the decision-making procedure. The CoR has moreover pointed out that the indices that could be used by local, regional, national and European authorities must be uniform and promote consistency in the adoption of decisions.
The overall message of the CoR in this context is that we welcome the efforts already undertaken by Eurostat in the field of local and regional statistics, and look forward to a close cooperation between the DG Environment, DG Regio and Eurostat to follow up on the Communication on "GDP and beyond".
Eurostat's latest "Yearbook of regional Statistics", released in November last year, is of great interest and value to all those involved in local and regional affairs across the EU.
The CoR deems it particularly useful to disseminate regionally disaggregated economic information on net disposable income, employment and social situation. The CoR is also pleased that statistical data on some variables are also already available at the level of urban and rural areas. We feel that such data could indeed usefully inform the CoR / EU debate on cohesion policy and urban policy.
Such data already provide useful information for the CoR rapporteurs and specialised services.
Sustainable growth Page 104
However, the CoR's opinion clearly points towards a need to do more, and therefore we would like to see in the future a more detailed local and regional breakdown of indicators on growth, jobs and innovation. With this objective in mind, the CoR would like to voice the need for more regionalised statistics and encourage Eurostat to include this concern in its programme for 2013-2017, which, as we understand, is already in preparation, and will be subject to discussions in the Council by the end of 2011.
In the context of our future requirements, I would like to stress the need for harmonised indicators to support the work of the Covenant of Mayors. To develop expertise and favour policy learning processes on climate change policy at the regional and local levels, such statistics should be further developed and tapped to inform better decision making.
Moreover, the CoR agrees in its opinion with the European Commission that a comprehensive environmental indicator should be developed and the quality of life indicators should gain more footholds in statistical analysis and as guidance to decision makers. The CoR has therefore firmly supported the preparation of a pilot project to draw up a comprehensive environmental index covering areas such as greenhouse gas emissions, loss of natural landscapes, air pollution, water use and waste generation.
Actually, Europe 2020 as a whole could be seen as a statistical challenge. The CoR is pleased to see that all European Institutions seem to share its firm belief, that the goals of the new strategy will not be achieved, unless the strategy itself is flexibly adapted to different socio-economic starting points and needs. We think that the Europe 2020 National Reform Programmes, due by end of next month, should be designed – and implemented – in partnership between all tiers of government. In what we see as a key test of multilevel governance, indicators and targets should be set not only at the country level, but also at the regional and possibly sub-regional level.
To respond to this requirement, in a situation when new indicators used (e.g. for environmental accounts) are only being developed at the national level, we insist on the fact that the development of new territorial indicators is taken as a political priority for the EU. On the one hand, this implies that, to identify and build territorial statistics, Eurostat should draw on the needs and experience of local and regional authorities. On the other hand, coherent budgetary appropriations are badly needed, given the sheer amount of financial resources involved in the production of statistics at the territorial level.
This is not a mere matter of "customer satisfaction". Indeed, this approach would bring fresh political value added in terms of increased ownership of the new strategy.
To underline this basic concept – that indicators are useful as a tool to support and communicate political will – a good example is provided by the Commission's intention to develop a new indicator to measure the share of fast-growing innovative companies in the economy. Actually, the old Lisbon Strategy failed to deliver also because it proved unable to speak to all European citizens. Among others, this was
Sustainable growth Page 105
also due to the fact that it seemed to restrict somehow its idea of innovation to world- class research activities, centers of excellence and high tech companies. The Innovation Union flagship initiative was right to see it in a broader way: innovative is any activity leading to increased productivity. The new indicator under construction will help to catch this conceptual and political need.
In its opinion on "GDP and beyond", adopted in October 2010, the CoR stressed that - although the role of local and regional authorities has not been highlighted in the Commission communication on "GDP and beyond" - successful good practice projects suggest that local and regional authorities could play a key role in adopting and disseminating a broader approach to measure societal progress (economic, environmental and social).
For this purpose, however, local and regional authorities should have appropriate capacity and resources including, but not limited to, financial support from the EU or from national sources. Indeed, a case in point to be mentioned here concerns the application of environmental accounting at regional/local level, using indicators which are focused on specific sectors and issues: within the European Interreg IIIC GROW Project, five European regions have started environmental accounting for air emissions. Best practice indicates that the approach of green national accounts can be practised by local and regional authorities, provided that appropriate methodology is in place for the economic valuation of environmental goods.
Of course, numerous examples of local and regional production of new statistics do not diminish the need for local and regional policymakers to refer to Eurostat for assistance when it comes to methodological tools and advice aiming at EU wide comparability of their data.
We would like to discuss more closely at appropriate level possibilities for a better synergy between the needs of local and regional decision makers on one hand and the potential of Eurostat on the other.
The CoR has in its opinion in particular considered that the Sustainable Development Scoreboard, proposed by the European Commission, while ensuring that the statistical systems for each level are compatible with one another – must be a tool that makes it possible to take action, to draft guidelines for the design of sector- specific and regional policies in the EU. The CoR has urged the Commission to present the pilot version of the scoreboard as pledged in the communication on "GDP and beyond."
Most local and regional authorities across the EU attach great importance to availability of statistical data in support of their decision making, and the CoR has taken on board their concern for closer links between statistical data and the reflection of policies in the budget at various levels of governance. The CoR has held in its opinion that it is vital for indicators, to match up with the headline targets of the new EU 2020 strategy and the Financial Perspectives beyond 2013. The CoR has stressed that Community strategies are reflected in the budgets, and these strategies must take future needs into account in order to improve a reality that can only be
Sustainable growth Page 106
based on two sources of information – statistics and public opinion – together with effective leadership on the part of Europe's democratic institutions.
The voice of the CoR has been heard by the other institutions, and key concerns voiced by the CoR opinion have been taken on board.
The EP resolution on the proposal for a regulation of the European Parliament and of the Council on "European environmental economic accounts," adopted on 13 December 2010, proposed to introduce a reference to regional statistics in the first article of the Regulation, along with national statistics, in order "to offer a means of monitoring the pressures exerted by the economy on the environment" which also was a key concern of the CoR opinion.
Moreover, the CoR Rapporteur has met the EC Commissioner for the Environment who agreed with the position of the CoR recognising the limitations of GDP as indicator, since it does not show what kind of resources the economy uses and does not provide sufficient information to decision makers in an economy facing economic, financial, environmental and social pressures. The CoR's views were broadly welcomed by Commissioner Potocnik.
In conclusion, I would like to stress the increasing need of local and regional authorities across the EU to dispose of timely and comparable regional data on key socioeconomic and environmental challenges and the readiness of many local and regional authorities to engage into dialogue on the future development of such data across all the levels of governance.
Sustainable growth Page 107
Catherine LARRIEU
1 - I'll begin with a fast overview of the national framework of our policy in favour of a sustainable and greener economy . This will allow to address how the indicators are a key contribution both to design this policy and to stimulate its implementation.
In July 2010, we adopted a national Sustainable Development Strategy. Endorsed by the Government, this strategy is not restricted to environmental-related issues. Furthermore, it offers a common collective project to be shared by all players in our Nation. The French Sustainable Development Strategy is consistent with our national commitments, particularly those of the so called Grenelle de l’Environment - the Environment round table organised in 2007. But it is also consistent with European and international commitments.
In a context where economic and financial crisis invites us to imagine new models of development, facilitating the transformation of our society will require the involvement of all actors, primarily to the public authorities. Patterns of organization, consumption and production that will allow us within 30 or 40 years, to live with more than 9 billion people on a planet with limited natural resources must be identified and developed at present, as soon as possible.
This national strategy was developed from early 2009, that is after the global economic crisis had started. It was finalized in conjunction with the work at the European level to design the new EU2020 strategy. The "classical" challenges expected in a SDS (such as climate or biodiversity, ...), are of course still part of the strategy, but are no longer given the main focus. Beyond them, the strategy highlights the crosscutting requirements for adaptation to these challenges. Requirements such as shifting towards new patterns of sustainable consumption and production, strengthening education and research, or a new governance. These three crosscutting requirements are given the main focus. The Grenelle de l’environement , our environmental policy implemented since 2007, is obviously not forgotten in the National Sustainable Development Strategy 2010-2013. But the NSDS, while incorporating all inovations of the Grenelle round table, extends the strategy to other conditions for sustainable development. Even though it found its origin in a process that differed very much from the EU2020 strategy, we can still see the strong convergence between the three key challenges of the NSDS that I just mentioned and the 3 axes of the European strategy: smart, sustainable and inclusive growth.
Indicators have been associated to the strategy in order to enhance the visibility and the embodiment of its challenges. Similarly to the strategy itself, their selection has been subject to a broad consultation with the civil society representatives. This led also to strengthen in comparison to the previous period, the socio-economic indicators in the dashboard that you can see currently on the slide. They are visible in the four context indicators and in a number of the nine challenges of the NSDS.
2 - To combine the economic, social and environmental dimensions, and articulate coherently the international, national and local issues, new forms of activity and new
Sustainable growth Page 108
mechanisms are now outlined. The backbone of the 2010-2013 national sustainable development strategy adopted by France is to support our society moving towards a greener and more equitable economy. “Greener” means sober in natural resources, low-energy, and carbon-free . That is not only a low-energy and carbon- free strategy.
But also a more equitable economy : competitive, supportive and respecting future generations. I emphasize here that we do not target only a greener economy. This alone would not be sustainable. The various crises (economic, ecological and social) have increased the urgency of addressing these challenges and strengthened the requirement for greater solidarity both nationally and internationally. Industrial and economic transitions, for instance, must be accompanied, by giving social and intergenerational cohesions their full importance. This can be achieved through targets to reduce inequalities, to fight against unemployment and precarity, targets on training, on prevention of risks, on mastering budget balances, or on governance. To summarize, the metamorphosis of our organizations can not be expected solely on a pure win-win mode for all activities and all stakeholders. These changes will involve all actors in the society and will require an equitable distribution of efforts to achieve it, whether it be between countries on our planet or within each country.
The concept of growth is highly debated, which makes quite sense. Going for Growth is a shared challenge, while it is necessary to clarify what growth we collectively want. In this regard the work on the indicators, following discussions initiated by the Commission on the Measurement of Economic Performance and Social Progress, so-called “Stiglitz-Sen-Fitoussi Commission”, are in the core of current debates. In France, the national conference on sustainable development indicators, organized in January 2010, brought together more than 400 participants, representing the various components of civil society, thus confirming by a large consensus the issues and associated indicators.
Everyone sees now to what extent the changes in our societies and patterns of life will be profound, in order to live in peace on a planet whose natural resources are obviously limited. This will require choices on many levels, for policy makers as to private actors, with consequently an increased need for evaluation and impact assessment of systems and complex changes, and thus the need for indicators available to all.
3 - Indicators should help policymakers to think and choose. In this respect, the work and skills of statisticians to design and make available these indicators are of course necessary.
But a major effort to disseminate these indicators is also needed. Communication plays a key role for each stakeholder to have a perception of changes on major challenges concerning us all today and tomorrow. Communication helps consequently each stakeholder to incorporate these issues and changes when steering and adjusting his own activities and decisions.
There is therefore a real need to communicate these indicators. The perception of these indicators will highly depend upon the limitation of their numbers. It is essential
Sustainable growth Page 109
to highlight a number of indicators small enough so as not to drown the players with a dashboard too complex. For the NSDS, we chose to promote 15 first-level indicators, enlightened by four key indicators of context. The second-level indicators are of course available via internet for those who are interested. But our communication efforts concentrate only on these first level indicators. We don’t try to promote a “magic indicator”, but a small enough set.
And we also prefer - as far as possible - to present the indicators in long data time series, to provide a maximum of information to each player concerning the challenges. This explains why the statistical brochure published to present the SD indicators presents charts with long series, a couple of decades, for example. Furthermore, we also plotted each time possible, on the same graph the targets set by public policies, in 2020 or even later. This enhances the effectiveness of the information, so that everyone can identify issues and possibliy the risks underlying a "business as usual” behaviour.
This broad dissemination of information issue is an important line of work. Indeed, the guidelines of the national sustainable development strategy should not only be translated into policies and decisions of public actors as discussed previously, but should also require the involvement of all stakeholders, including citizens and consumers who can stimulate or inhibit a greener economy in their choices of purchases.
4 - The challenges for sustainable development call for collective or individual, local or national actions. But on many issues, these actions may not have the expected impact unless guidelines and tools are implemented at European or international level. Negotiations on climate or biodiversity, work within the OECD on green growth, European initiatives, discussions within the G8 or in Rio 2012 are essential.
We need more than ever a set of comparable indicators for discussions or negociations at these international levels to make progress. Then, it is up to each country or region to complete or detail the international set of indicators to meet its own national or regional needs.
If we consider our NSDS for instance, we have stated as a principle that this national set of SD indicators would include the European SD indicators. It has been proposed at the very beginning of the consultation with the civil society representatives, and was quite well understood and welcomed.
To account for the complex interdependencies of our socio-economic systems on the planet, we also need global indicators. These are indicators that are not confined to only measure the impact of production activities taking place on the national territory, even if their accuracy is not easy and even if they are compiled at the national level. - The carbon footprint of final domestic demand is an interesting indicator since it calculates the share of global emissions of CO 2 that each country is responsible for through its domestic consumption. - Following the same line, since a green economy cannot only be a low-carbon economy, but shoud also be sober in natural ressources, indicators of material consumption and of material productivity are adjusted of external trade by taking into account the content of materials in products exported and imported.
Sustainable growth Page 110
Indicators of this type, which allow to assess the impact and sustainability of production and consumption patterns, meet the recommendations of the “Stiglitz- Sen-Fitoussi Commission”. Even if they are very sensitive, they may help us to recognize the need to adapt our socio-economic models in respect of future generations. These indicators were thus designed to play their role in international discussions as well as in defining our national policy.
Indeed, in a globalized economy where the localisation of activities and flows are interrelated, the complexity and importance of political, economical and financial issues make it essential to share visions consolidated by the availability of indicators. To move towards sustainable development, we need to convince. Indicators must be simple enough to help us to convince and to modify criteria for choices.
The NSDS and its indicators can be downloaded at : www.developpement-durable.gouv.fr/sndd
Sustainable growth Page 111
Sustainable development and green growth: taking stock of the work at international level
Eric DE BRABANTER
Challenges of today
Two main challenges
1. meeting the demand for better lives and expanded economic opportunities for a global population that might reach 9 bn. by 2050. 2.addressing environmental pressures that could undermine these opportunities.
Sustainable Development Green Growth Strategies Plans or Strategies
Sustainable Development
Definition a development that meets the needs of the present without compromising the ability of future generations to meet their own needs ″.
SOC
external governance dimension institutions ENV ECO
Sustainable growth Page 112
Green Growth (1)
Working definition (Monitoring Report) maximising economic growth and development while avoiding unsustainable pressure on the quality and quantity of natural assets & harnessing the growth potential that arises from transiting towards a green economy.
growth & development - investing in the environment should contribute to new sources of economic growth: new green industries, jobs & technologies; managing the transition for greening the more traditional sectors (brown industry → green industry), hence ...
avoiding unsustainable pressure on the quality (...) of natural assets - ... preventing environmental degradation (air, water, biodiversity, ecosystems), and ...
avoiding unsustainable pressure on the (...) quantity of natural assets - ... increasing environmental efficiency of production & consumption (patterns, decoupling); ... SystemicSystemic RisksRisks
Green Growth (2)
Definition - which growth? but also ...
focusing on households perspective & people ’s (material) well being - (environmental) quality of life, health impacts, consumption patterns. ‘what matters is not production per se but the well-being derived from it ’
proposing elements of green growth for the poorest in developing countries . RioRio OECD multidimensional approach of +20+20 Green Growth Sustainable Development Global Project on Measuring Plans or Strategies the Progress of Societies & Stiglitz-Sen-Fitoussi
Sustainable growth Page 113
Sustainable growth Page 114
Sustainable growth Page 115
Sustainable growth Page 116
Sustainable growth Page 117
Summary of the Session
Pedro DIAZ MUNOZ
1. Set up
Four presentations gave four different perspectives as follows:
European policy perspective
Robin Miege – Director, Strategy, DG Environment, European Commission Main player in the development of the recent Communication: A resource efficient Europe - Flagship initiative under the Europe 2020 strategy
European regional perspective Lucio Gussetti – Director for Consultative Works, European Committee of the Regions
National perspective and its inter links with EU and international perspectives Catherine Larrieu - Head of the Sustainable Development Delegation, Office of the Commissioner General for Sustainable Development, Ministry of Ecology, Sustainable Development, Transportation and Housing, France
International perspective Eric de Brabanter – Ministry of Environment and Sustainable Development, Luxembourg. Chair of OECD Working Party on Environmental Information. To present the Green Growth Strategy (OECD + others) and their link to other international initiatives
2. Main ideas
I would like to report ten main ideas that emerged from the session:
1. Importance of good data to support policies. Several new areas where there is a need of information were mentioned: Resource efficiency – ECO innovation - Ecosystem services - Green procurement - Environmental goods and services
2. For what purposes do policy makers need this information? It was emphasised that not only was the information required for monitoring policies, but also to define them. And, at an earlier stage, it is needed to assess the impact of different policy choices.
3. We can conclude that the phrase of the Conference was:
Good statistics are much cheaper than wrong policy decisions
BUT, it should also be taken into account that while it is easy to measure cost, it is much more difficult to measure benefit.
Sustainable growth Page 118
4. Green growth policies mean choices and trade-offs. There are no win-win solutions for all our activities. In particular, we should be aware that: Green growth does not necessarily imply more jobs Even the concept of growth is debated: Which is the type of growth that we want?
5. The importance of communicating to citizens was stressed. A focused dissemination to citizens is more justified in these policy areas because citizens/consumers are actors in many of these policies. Therefore, informing them is essential for their application of the measures and the success of the policies. As a consequence, it is very important that a simple indicator set is designed in order to transmit with a small number of figures the most important facts.
6. But all levels of government ( national, regional, local ) are also actors. Therefore, they need the information. In particular, some policies are defined and implemented at local level (Urban transport, Waste treatment, City planning...).
7. It is important to ensure a good interaction in the triangle (Territorial dimension – EU2020 – Budget allocation) If there is no good information that will have an impact on budget distribution, financing can go to an activity that damages the environment and further financing will be needed to pay for correction of the damage.
8. A certain confusion has emerged with the variety of concepts linked to green growth and sustainable development which are closely interrelated, but not equivalent. There is a need to clarify.
9. In particular, the current trend may seem to have the effect of diluting the concept of sustainable development into so many other concepts (green growth, greening of the economy, resource efficiency, GDP and Beyond, Stiglitz Report…). There is a feeling that all these initiatives overlap each other.
10. A plea was issued for working together in the international field in an effort to maintain consistency between different concepts and, more importantly, their goals and the indicators related to them. The example of a case where 25 regional indicators had been defined and only 8 of them are internationally comparable was given.
3. Conclusions
These ideas can be summarised in five main conclusions:
1. Communicating to citizens and use by policy makers are different and require different data and different detail.
2. Importance of data for all levels of government where policies are developed and implemented.
3. Need of consistency between indicators emerging from different initiatives.
Sustainable growth Page 119
4. It is true that good statistics are much cheaper than policy decisions! But they cost anyway.
5. Statisticians must move fast and show flexibility to meet changing needs; but policy makers also have the responsibility to make the best of the information that is already offered to them. In this sense, legislative negotiations on the first batch of environmental accounts modules have just been concluded at technical level and the adoption of a regulation could be expected in a few weeks. We trust that this will provide an important wealth of information to our users.
Sustainable growth Page 120
Europe 2020 and the poverty target: how we got there and where to stand
Antonia CARPARELLI
From Lisbon to Europe 2020
March 2000 Lisbon European Council – Presidency Conclusions : “The number of people living below the poverty line and in social exclusion in the Union is unacceptable. Steps must be taken to make a decisive impact on the eradication of poverty by setting adequate targets to be agreed by the Council by the end of the year ” March 2010 European Commission – Europe 2020 “The EU needs to define where it wants to be by 2020… 20 million less people should be at risk of poverty.”
2
Ten Years of Open Method of Coordination
A very evolutionary process Social inclusion, Pension, Health and Long-term care, Streamlining, Renewed Social Agenda, Europe 2020 Work on indicators has been a crucial part of it Original portfolio has considerably evolved with the fine- tuning of initial selection and development of additional indicators Debate on targets as a leitmotiv of this evolution Child poverty report, Debates on working methods in the SPC, Renewed Social Agenda…
3
Inclusive growth Page 121
But poverty and exclusion in Europe have remained stubbornly high…
• In today’s Europe, 16% of the population is at risk of poverty (1 out of 6 six people) and the percentage is even higher for children (19%)
• Even in the years of sustained economic growth, poverty levels have remained unchanged (See SPC Report on Growth, Jobs and Social Progress, 2008)
• The crisis and subsequent fiscal austerity have certainly not improved the situation, although this is not yet reflected in the statistics available
4
2010…a special year for Social Europe
New Lisbon Treaty - A social market economy, highly competitive, aiming at full employment and social progress. - New social clause (Art.9); Protocol on social services; Charter of fundamental rights becomes a binding text. New Commission - President Barroso: “We need to make sure that our values of inclusion, equity and social justice are carried forward into a new approach” (September 2009) European Year 2010 against poverty and exclusion - A far reaching campaign, mobilising EU institutions, national actors and stakeholders. Strong engagement of ES and BE presidencies. - A virtuous competition between institutions, paving the way to the social commitments in Europe 2020 5
Inclusive growth Page 122
Europe 2020 and its social dimension
Three overarching objectives - Smart, sustainable and inclusive growth
Five headline targets - Employment (75 %); R&D (3% of GDP); Climate/energy ("20/20/20“); Education (ESL < 10% and TD > 40%); Poverty (- 20 million AROP)
Seven flagship initiatives - Innovation Union, Youth on the move, A digital agenda, New skills and jobs, Industrial policy, Resource efficiency, Platform against Poverty
6
Measuring poverty: a controversial issue
President Barroso: Reaching consensus on the poverty target has proved one of the most difficult aspects of the negotiations on Europe 2020 Spring European Council 2010: Agrees on the other targets but asks for more work on poverty indicators Member States positions: Reflect very diverse political and institutional concerns, including subsidiarity, but the debate largely focuses on the quality and the pertinence of the reference indicator
7
Inclusive growth Page 123
The headline target on poverty: the Commission original proposal
EU level target: Reducing poverty by one fourth by 2020, lifting some 20 million of people out of poverty Reference aggregate: People at risk of poverty , headline indicator used to quantify and monitor poverty in the EU and in most Member States Definition: People at risk of poverty are those who live with less than 60% of the median income in their country Description: Varies from 9.1% to 25.6% (2008 figures). Poverty line varies from less than 100€ to more than 1000€ across EU countries
The reasons of low income countries…
• The risk-of-poverty rate is a measure of inequality rather than a poverty measure • It underestimates poverty in low income countries. When the poverty line is less than 100 € being above the poverty line is not enough to be out of poverty • The relation with GDP growth is weak and sometimes problematic • It gives a rather implausible picture of the distribution of poverty across Europe
9
Inclusive growth Page 124
…and the reasons of the countries with generous welfare systems
• The risk of poverty, focused on monetary income, does not take into account in kind benefits and access to universal services • The risk of poverty only refers to one dimension of poverty and not necessarily the most relevant one, that is, social exclusion • Participation in the labour market is the most significant indicator of social inclusion • Benefits can lift people above the poverty line, but may create benefit dependency and new forms of exclusion
10
The target adopted by the Council
• EU level target –“Lifting 20 millions people out of poverty or exclusion by 2020” – Based on 3 indicators : • At-risk-of-poverty, • (severe ) material deprivation (adopted in February 2009), • people living in households with low work intensity (“jobless households”) • National targets
– Member States are free to chose the most appropriate indicator to set their national target – Member States to show how they will contribute to meeting the EU level target, in dialogue with the Commission
Inclusive growth Page 125
The (severe) material deprivation indicator
Definition : People whose living conditions are severely constrained by a lack of resources, They experience at least 4 out of 9 deprivations: • people cannot afford i) to pay their rent or utility bills, ii) keep their home adequately warm, iii) face unexpected expenses, iv) eat meat, fish, or a protein equivalent every second day, v) a week of holiday away from home once a year, vi) a car, vii) a washing machine, viii) a colour tv, or ix) a telephone.
Description : Reflects different living standards across the EU (reference items are the same for all EU countries) Concerns 8.1% of the EU population. Varies from 1% (LU) to 42%(BG)
The low work intensity indicator
Definition : People (aged 0-59, not students) living in a family where no one works (or very little) Zero or very low work intensity refers to situation where – all adults in the family are out of work over a whole year, – or they work very little in relation to their total work potential (Work intensity <0.2)
Description : Reflects long-term exclusion from the labour market, for individual workers and the family members who depend on them It concerns approx. 9% of the people aged 0-59, varying from 4% to 20% (provisional calculations)
Inclusive growth Page 126
EU-27: 114 Million people at-risk-of poverty or exclusion (23% of total population)
Low work intensity 9%
Severe material Risk-of deprivation poverty 8.1% 16.3%
Latvia: 0.8 Million people at-risk-of poverty or exclusion (37% of total population)
Low work intensity 6.7%
Risk-of poverty Severe material 25.7% deprivation 21.9%
Inclusive growth Page 127
The Netherlands: 2.5 Million people at-risk-of poverty or exclusion (15% of total population)
Low work intensity 8.3%
Severe material Risk-of deprivation 1.5% poverty 11.1%
National Targets (Draft NRP)
National target Total number of National target Total number of (estimated people at-risk (estimated people at risk contribution*) (2008) contribution*) (2008) AT 235,000 1,530,000 IT 2,200,000 15,100,000 BE 330,000-380,000 2,190,000 LT 170,000 980,000 BG 260,000 (500,000) 3,420,000 LU 3,000 72,000 CY 18,000 174,000 LV 121,000 760,000 CZ 30,000 1,570,000 MT 6,560 79,000 DE 330,000 (660,000) 16,350,000 NL No target in NRP 2,430,000 DK 22,000 890,000 PL 1,500,000-2,000,000 11,490,000 EE 49,500 290,000 PT 200,000 2,760,000 EL 450,000 3,050,000 RO 580,000 9,420,000 ES No target in NRP 10,340,000 SE No quantitative target 1,370,000 FI 150,000 910,000 SI 40,000 360,000 FR 1,600,000 11,240,000 SK 170,000 1,110,000 HU 450,000 - 500,000 2,790,000 UK No target in NRP 14,060,000 IE 186,000 1,050,000 EU 10 Mio - 15 Mio** 116 Mio
* Estimated contribution to the EU target when the indicator used is not the same as the EU target ** EU estimate: lower value without ES, NL, SE and UK – higher value: including an estimate for ES, NL, SE, UK
Inclusive growth Page 128
The challenges ahead…
• Political commitment and ownership still to be built and consolidated • Considerable work is still needed to improve the quality and timely availability of the reference indicators (important appointment in 2015) • It will be important to establish effective monitoring mechanisms, that duly take into account the multidimensional nature of poverty and exclusion
The three challenges are strictly linked…
Measure what you value… Commitment to the target will also ensure continued investment in statistics on poverty and credibility of monitoring And value what you measure… Reliable and timely statistics and effective monitoring will help creating political commitment and accountability
Inclusive growth Page 129
The challenges of monitoring progress towards the Europe 2020 targets
Anthony Barnes ATKINSON
1. A complex monitoring challenge
• Appraisal of national plans Monitoring has to be ex ante as well as ex post. EU has to be in a position to evaluate the relation between the National Reform Programmes of Member States and the target. Complicated by 3 dimensional form of target. Need for policy modelling. Role for a safety first “fall-back” policy?
3
• Criteria for evaluating achieved progress Much effort has gone into design of common indicators; much less attention has been paid to the criteria for monitoring progress. • Monitor all three domains for all Member States; • Agree ex ante on criteria for identifying shortfalls from expected trajectory; • Identification of ex post “success stories” (on course), “cases of policy concern (where sustained shortfall), and “warning cases” (where short-term departure); • Agreement on allowance, if any, to be made for exogenous shocks.
4
Inclusive growth Page 130
• Words as well as numbers Indicators need to be accompanied by Member State reports that cover the wider context, particularly identifying aspects of poverty and social exclusion not covered by the indicators that may have worsened.
5
2. Great progress: remember the 1980s
Country Source for EU poverty estimates Years
Belgium Antwerp Centre for Social Policy Household Survey 1976 (Flanders only) and 1985
Denmark Central taxpayer register (based on tables) 1977
France Enquête Revenu 1975 and 1979
Germany Income and expenditure survey (EVS) (based on detailed 1973 and 1978 tables) Greece Household Budget Survey 1974 and 1981/2
Ireland Household Budget Survey (HBS) and ESRI Household Survey 1973, 1980 (HBS) and 1985 (ESRI) Italy Istat Survey of Consumer Spending 1975 and 1980
Luxembourg Socio-Economic Panel Study of Households 1985
Netherlands Income Distribution of Households (ID) and Housing Demand 1977 (ID) and 1981 and 1985 Survey (HDS) (HDS) Portugal Household Expenditure Survey (HES) and Household Income 1973/4 (HES) and 1980/1 (HIES) and Expenditure Survey (HIES) Spain Encuesta de Presupuestos Familiares 1973 and 1981
UK Family Expenditure Survey 1975, 1980 and 1985 66 O’Higgins and Jenkins, 1990.
Inclusive growth Page 131
But more to do
• Raising the profile of household statistics
New demands: monitoring Europe 2020 headline targets and use of social inclusion indicators in EU policy domains. Competition for reduced resources available to NSIs. Coherence between household data and national accounts. Measuring well-being agenda.
7
Integration within national statistical systems Hierarchy of degrees of standardisation: • common survey instrument (ECHP) • ex ante harmonised framework (EU-SILC) • ex post standardised micro-data (LIS) • ex post customised results (OECD) • meta-analyses of results (Kuznets).
EU-SILC has secured advantages from building on national sources, but should be fully integrated into national statistical systems. • Acceptance of the EU reference source depends on the reconciliation of findings with those from well-established national sources (e.g. German Socio-Economic Panel); • Resource savings can be made by avoiding survey duplication (e.g. UK proposal to use FRS). 8
Inclusive growth Page 132
• Getting more up-to-date statistics European Commission in its 2009 GDP and beyond Communication called for “more timely social indicators”. In part this requires reduced time lags between data collection and data publication; in part it is a matter of survey design and the questions posed. Relying solely on annual income in the previous calendar year may not be sufficient; more use may need to be made of current income in a shorter time period. E.g. DIW Berlin reports measures of inequality/ poverty from the German Socio-Economic Panel on the basis of last year’s income and of current income. The search for greater timeliness may also point to leveraging other sources, as discussed at OECD March 2009 Roundtable. 9
Conclusions: there have been great advances in EU statistics, but we need
• to develop tools for ex-ante policy evaluation; • to develop specific criteria for monitoring progress; • not to focus solely on the numbers; • to give higher priority to household statistics; • to better integrate EU-SILC into national statistical systems; • to get more up-to-date statistics.
10
Inclusive growth Page 133
SILC: a key element to monitor poverty, social exclusion and quality of life
Inna STEINBUKA
HIGHLIGHTS
1. SILC main goal, principles and implementation
2. EU2020 poverty target
3. Complexity of the revision of the EU-SILC legal basis
• Rationalisation • New requirements
4. Challenges of measuring Quality of Life
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 2
SILC MAIN GOAL AND PRINCIPLES (AS IN THE 2003 REGULATION)
Main goal
analysis and monitoring of poverty and social exclusion
Main principles
Output harmonisation of variables. Requirements on methods used (samples, etc) Annual cross sectional and longitudinal data at household and individual level Timely and comparable annual cross sectional data Flexibility in terms of data sources (existing national data, administrative data) Integration into established national statistical systems
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 3
Inclusive growth Page 134
SILC IS IMPLEMENTED
Relatively recent start to data collection (data since 2003-2004)
SILC
Delivers, is accepted at the policy level and extensively used by the research communities Provides rich set of information on social inclusion and quality of life: • Income, housing, deprivation, labour, child care, education, health • In the ad hoc modules: transmission of inequalities, social participation, details on housing and deprivation, …. Allows for the development of the at risk of poverty or social exclusion indicator, based on three sub-components • Relative risk of poverty • Material deprivation • Low work intensity
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 4
EUROPE 2020 HEADLINE INDICATOR ON AT RISK OF POVERTY OR SOCIAL EXCLUSION, EU 27, 2009
At risk of poverty : 80 mio Severe material deprivation: 40 mio
At risk of poverty or social exclusion: 114 mio
Low work intensity: 34 mio Source: EU-SILC
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 5
Inclusive growth Page 135
EUROPE 2020 HEADLINE INDICATOR ON AT RISK OF POVERTY OR SOCIAL EXCLUSION, EU 27
140 At risk of poverty or social exclusion 120
100 At risk of poverty 80
60 Mioindividuals Material deprivation 40 Low work intensity 20
0 2005 2006 2007 2008 2009
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 6
REVISION OF EU-SILC: NEED FOR RATIONALISATION
Modernisation of social statistics:
• Response to the Vision for EU Statistics (COM (2009) 404) • Two-pillar approach and modularisation
Budget constraints in MS
Simplification: review content
Increase the use of administrative data
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 7
Inclusive growth Page 136
REVISION OF EU-SILC: NEW REQUIREMENTS
Better timelines
Higher comparability
Increased relevance, in particular the improvement of: • coverage of the dimensions of quality of life • joint measurement of income, consumption and wealth • coverage of specific sub-populations (migrants) and at regional level
Developing capacity to monitor national situation with respect to at-risk-of-poverty or exclusion
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 8
CHALLENGES OF MEASURING QUALITY OF LIFE
Multidimensional approach vs. single composite indicator Optimal combination of various dimensions, indicators and variables User-friendly way of presentation: Radar chart Multiple data sources • LFS (labour data, education and LLL) • HBS (consumption patterns) • EHIS (health) • AES (adult education) • SASU (safety) • ICT – (access to technology) • + non ESS data sources (ECB wealth survey, European Foundation: European Quality of Life Survey, Dublin Research Network: European Social Survey)
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 9
Inclusive growth Page 137
RADAR CHART
Conditions de vie matérielles
Ins écurité phys ique Contrainte financière
Ins écurité économique S anté Tous Q1
Contacts avec les autres Education
Participation à la vie publique Conditions de travail
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 10
ON GOING REVISION AND OTHER ACTIONS
Revision process started. Competing priorities. Proposals expected in 2013
Ad hoc module on well being in 2013
Other actions • Improved access to administrative data • Feasibility study on data matching - Links with HBS and ECB health survey • Improved links with national accounts • Initiatives to improve timeliness (flash estimates)
Brussels 10 March 2011 Statistics for policymaking: Europe 2020 11
Inclusive growth Page 138
Summary of the Session
François LEQUILLER
Target and indicators
• Target = Lifting 20 million people out of poverty and exclusion in 2020 • Three indicators: – People at risk of poverty (<60% of median income) = 80 million – Severe material deprivation = 40 million – People living in households with low intensity = 34 million – Total population = 114 million
The session
• The ultimate conclusions, given the constituency: – Total priority on households statistics – Social Imbalance Procedure?
• Three presentations: the story/the tool/the challenges
The story of the target + indicators
• Not perfect but a very significant achievement • Good balance between difficulty of defining poverty and realistic policy target • Difficulty in matching national targets with EU target • The tool to measure
• SILC = central tool for monitoring the indicators • Impressive progress compared to 20 years ago • Needs rationalisation and integration into national surveys • Needs more timeliness (flash estimates?)
The challenges to reach the target
• Set an ex ante procedure for monitoring progress towards the target • Need to monitor national targets = regular reports (narratives) • Need for policy makers for more policy modelling: how to reach the target? • NSIs and Eurostat could shift in more analysis of compliance with the target
Priorities for the ESS
• Timeliness (in 2011 only 2007!) • Quality of data on material deprivation • Quality of SILC • Integration of SILC into national survey frameworks • Tools for ex ante evaluation • Integration into national accounts • Importance of regional data = but who pays?
Inclusive growth Page 139
Speech in the opening session Commissioner Olli REHN
Ladies and Gentlemen,
I am glad for the opportunity to make concluding remarks in the closing session of this very topical statistical conference. The annual Eurostat conference has become a valuable tradition, with a visible impact on the development of our statistical agenda and of evidence- based policy-making in Europe. This year's conference has been no exception.
This conference has focused on the Europe 2020 strategy. Yesterday, President Barroso presented you the Commission's approach to deal with economic imbalances and to support smart, sustainable and inclusive growth through. As he said, statistics are found almost all over Europe 2020. To communicate the key policy objectives, we need relevant and reliable statistics.
What is politically and economically at stake?
To reach the EU2020 policy targets, the necessary priority of Europe is to stabilise public finances in the Member States and to safeguard and solidify the ongoing economic recovery. It is essential that we do not allow the relative calm in the financial markets and the somewhat improved macroeconomic outlook to lower the level of ambition, or slow down the completion of the reforms.
Later today, the euro-area leaders will discuss the comprehensive policy package to overcome the latest stage of the financial crisis and to strengthen our economic and financial architecture.
To make the "Competitiveness and Convergence Pact" to work, an unequivocal political commitment by the Member State governments is needed. The new structure of EU economic governance will provide an effective mechanism of policy surveillance and delivery. Needless to say, relevant and reliable statistics provide the very foundation for reinforced economic governance and better competitiveness.
In the context of the EU's comprehensive response, the Portuguese government has just presented significant new commitments to ensure fiscal sustainability, strengthen the financial system and raise the growth potential and flexibility of the Portuguese economy.
Closing session Page 140
I welcome and support this package of far-reaching, concrete measures. It is a major strengthening of the Portuguese macro-economic policies.
The additional immediate fiscal consolidation measures amount to 0.8% of GDP in 2011. The announced measures for 2012 and 2013 amount, respectively, to 2.5% and 1.2% of GDP. This should be sufficient to reach the ambitious deficit targets of 4.6% of GDP in 2011, 3% in 2012 and 2% in 2013.
We also welcome the commitment to far-reaching structural reforms, especially regarding further labour market reforms and the financial sector, which will foster economic growth and correcting imbalances.
The announced package will help Portugal regain control over debt dynamics and put an end to uncertainties. We will continue to closely monitor the situation in the context of enhanced surveillance.
The commitments of the Portuguese government clear an important building-block of the EU's comprehensive response to the sovereign debt crisis, and call for progress concerning the other blocks, especially linked to the reinforcement and increased flexibility of the financial backstops.
Ladies and Gentlemen,
I shall now try to summarise the very rich discussions held during the conference with a couple of conclusions and recommendations.
In selecting the indicators for the Europe 2020 strategy, we recognise the importance of discussions on 'GDP and Beyond' and the results of the so-called ‘’Stiglitz-Sen-Fitoussi Report’’. These initiatives, both published in autumn 2009, have challenged official statisticians to develop new indicators that better describe our environmental challenges, the quality of life of households and the distributional aspects of income.
The European statisticians led by Eurostat and INSEE, the French Statistical Office, have set up a high-level group that will respond to the challenges set by these reports after the summer.
Closing session Page 141
Not all the issues on producing the EU 2020 indicators are clear yet. This can be concluded from the discussions in the opening session yesterday morning. There is a need to discuss the recent EU policy initiatives and the important role statistical indicators will play in benchmarking and monitoring the progress of these policies.
There is also a need to exchange views on the role of official statistics as an essential element of the policymaking infrastructure of modern Europe. I do also believe that we need to promote the interaction between policymakers and statisticians in defining, producing and using statistics for policy purposes, of course without compromising the necessary professional independence of statistical institutes.
As mentioned by President Barroso yesterday, the Commission is currently preparing a communication on the quality management of European statistics, which we will present shortly. We aim at creating the necessary mechanisms to continue to ensure the high quality of statistical indicators.
To ensure quality and reliability, we have to move from correction to prevention, which is exactly the same approach as we reinforced economic governance. At the same time, we have to strengthen the independence of the European Statistical System.
I am looking forward to receive your contributions to this debate in the coming months.
Ladies and Gentlemen,
In this conference it has been widely recognised that sound and timely statistics are essential for making the Europe 2020 Strategy a success. This implies a better use of all the available statistics, as well as developing new statistics and relevant indicators.
Indicators must be accompanied with clear and agreed targets at European and at Member State level. The targets cover economic, social and environmental issues, and they have become part of the economic and policy governance of Europe. They reflect the complexities of the development of modern societies. These indicators act as a common language when we assess policy progress between the Member States. The process of selection and development of indicators is a crucial phase. The users and producers need to interact from early on. The research community has an important role in this work. Statistics for policymaking have to comply with the highest quality standards. Clear quality commitments by the national and international statistical institutions and quality
Closing session Page 142
assurance frameworks are necessary. The credibility of the whole policy framework depends critically on the quality of statistics used.
However, quality has a price. It is important to clarify to users what kind of labelling we use for different statistics. The label "official statistics" should be reserved for those figures which fulfil the requirements of the European Code of Practice for Official Statistics. Model-driven calculations and normative objectives, such as political targets or valuations, do not fall under this category. This difference should be clearly distinguished in communication.
Further, statistics should be developed to react in a more flexible manner to new policy needs. At the same time, increased efficiency would help statisticians to cope with tough external constraints, such as limited resources and requests for the reduction of workload.
European statistics can build on its long tradition of quality work and coordination mechanisms. They have been and continue to be developed within the European Statistical System. We should at the same time reflect on the ways and means to further enhance the quality assurance. The recent granting of audit-type powers for Eurostat in the area of national accounts and public finance statistics is a very important recent example.
Moreover, it is important that indicators are disseminated so that citizens can understand them and their relevance to the decision-making processes. Official statistics should be user- friendly and easy to access, and they should be explained in a more pedagogical way. They can also benefit greatly from modern tools of presentation and visualisation.
Finally, we want to make our policy making ever more efficient and transparent. This is why it is essential that users and policymakers respect and understand the principles on which high quality official statistics are built. Only this way statistics can be used as an indicator in a way that makes policymaking more accountable to the public.
Ladies and Gentlemen,
With these words I would like to conclude and close the conference. Many thanks to all of you for your contribution and participation.
Closing session Page 143