Chapter 1. The continued importance of the “Beyond GDP” Agenda

1.1. Introduction

In January 2008, before the Global Financial Crisis, President Sarkozy of France established a Commission to examine the adequacy of our metrics for assessing economic performance and social progress. He, like many others before him, was worried that too much attention was placed on GDP as an over-arching measure of performance. GDP, as we all know, is a measure of the volume of goods and services produced within a country over a given period of time. It is not – as it is often used – a measure of a country’s success.

This is not a new complaint. Just over 50 years ago, Robert Kennedy gave expression to similar concerns:1

In spite of its shortcoming as a measure of a country’s success, GDP has remained its key proxy. We take measures like GDP for granted – until they fail us. Much of this book is about the discovery of how this metric has failed us when used for purposes that it was not designed to address, and about what can and is being done to create measures that better reflect changes in economic performance and social progress.

It was not until after the Great Depression that governments started to collect the data necessary to measure GDP. Keynesian economics – which explained the level of economic output in terms of the demand from different sectors of the economy and argued that government action could maintain the economy at full employment – made it imperative to have better indicators in order to assess the state of the economy.2 Two economists, Simon Kuznets, of the University of Pennsylvania, and Richard Stone, of Cambridge University, received Nobel Prizes, in part for their contribution in setting up the System of National Accounts (often abbreviated as SNA) which includes GDP, a concept developed by Kuznets.

For a while, economics students had to learn the ins and outs of these metrics, the assumptions that went into them (and why those assumptions were made) and the limitations of these measures and their uses. GDP and other national accounts indicators became part of the tool kit of economists. Careers were built trying to explain the movements of GDP and to show how these movements could help explain that of other indicators. This trend was reinforced by economists’ growing ability to analyse data statistically, as a result of better computers.

But, as time went by, and as the sophistication of analysing the inter-relationship among different data series increased, the attention paid to the data series themselves, and in particular, to the limitations of GDP as a welfare metric declined. The result was that the reliability and relevance of results of analysis in areas such as macro-economics that were heavily reliant on GDP measures may have declined too. The paradox is that those who built the system knew of its shortcomings and were cautious when using it.3 But as the general understanding of these indicators and their construction diminished, their use became more widespread and their limits were forgotten by most users. While GDP had been designed and used to measure market activity, increasingly it became a thermometer used for assessing the general health of societies.

Simon Kuznets warned against this risk more than 50 years ago:

The issues discussed in this book may, at first sight, appear as technical, and addressed to a narrow field of specialists. But they are not only technical: they go to the root of how our democratic systems function. In the words of President Sarkozy, the goal of the Commission was to address:

This gap between the “experts” and the citizen they are supposed to be serving has played an important role in the bitter divisions within society that have been so vividly demonstrated in a number of recent elections. President Sarkozy was thus prescient in drawing attention to the consequences in a democracy of a growing gulf between the statements, assertions and beliefs of the experts and elites, on one side, and the lived experiences of significant numbers of citizens, on the other.

The Commission had as one of its central missions ensuring that our metrics drew our attention to those things that made a difference to people’s lives. One thrust of this book is that, had the recommendations of the Commission been more fully implemented – and some of the issues developed in this volume more fully explored – different policies might have been chosen. There were early warning signs not only of the discontent, but also of the underlying changes in our economy and society that might have been expected to give rise to such discontent. Whether that is so or not, one thing is clear: metrics matter, and in both the financial crisis of 2008 and what some have called the political crisis of 2016, our system of national accounts, centring on GDP, did not give us adequate warning of what was around the corner. Some may say that was not the purpose, that was not what national accounts were designed to do. But surely it would have been good for society to have some well-established and widely used indicators of the major traumas that were about to befall.

1.2. GDP statistics and the Great Recession

Even when focusing on market income, national accounts statistics may sometimes fail to provide the full picture. In September 2008, the United States – and then the world – fell into what has been called the Great Recession, the worst global downturn since the Great Depression 80 years earlier. As the economy was about to slip into recession, leading economists pronounced the economy to be in good health.6 A few years earlier, the long-time head of the US Federal Reserve, Alan Greenspan, dismissed worries about the possibility of a bubble in asset prices by saying that there was just a little “froth”.7 These leading economists had taken the vital signs of the economy and pronounced it fit and healthy.

In retrospect, there were a number of failures. Those in positions of power, who have to decide whether, for instance, to tighten or loosen credit if they judge the economy to be, or about to be, over- or under-heated, rely on a variety of indicators – just like medical doctors take a patient’s pulse, monitor blood pressure, take a blood cell count, look at whether the patient is gaining or losing weight, etc. Indicators that could have provided a warning signal to policy-makers of what was about to happen were, in many cases, available but were not part of a well-established reporting system and were mostly ignored by those who should have noticed. In other cases, these indicators were simply not available – e.g. sectoral accounts and balance sheets, which led (post-crisis) to several initiatives aimed at improving the kind of information available to decision makers.8 In other cases yet, early estimates of GDP failed to provide a sense of the true scale of the recession, and were heavily revised in later periods.9

Most fundamentally, policy-makers ignored these warning signals due to the ideological blinders that prevented them (and their economic advisors) seeing the dangers ahead. They were also reassured by conventional macro-economic models that said that a crisis of such proportions could not happen. They took comfort from past movements in GDP, in the hope that the “Great Moderation” (combining good GDP performance with lack of inflationary pressures) that prevailed before the crisis could extend into the future.

A year later, as GDP began to increase in 2009, President Obama announced that the economy was on the mend, that recovery in the United States had begun. Yet, to the overwhelming majority of Americans, this did not seem to be the case. Again, aggregate economic indicators, such as GDP, seemed out of synch with “the facts on the ground”. What was going on, what was being experienced by most Americans and by most people in the developed world, was not conveyed by the figure that is usually relied upon, GDP.

There were, of course, some simple explanations of this conundrum. One factor was that GDP growth often disproportionally benefitted those at the top of the income scale: in the United States, 91% of all the gains in income in the first three years of recovery (2009-12) is estimated to have gone to the top 1% (Saez, 2016). Thus, in the United States, most households were not experiencing a recovery. There was a similar phenomenon in Europe, more visible in countries most affected by the crisis. In those same three years of supposed recovery, growth in average household income in Europe, as measured in the national accounts, lagged GDP growth;10 while growth in median household income – where 50% of the population is above that level and 50% below – as measured in surveys, lagged average income growth (which was boosted by gains going to a relatively small number of people). No wonder, then, that most people felt that there was no real recovery.

A second factor was that many households lost their home and jobs, and often their hope in the future, in addition to their income. And those who didn’t were fearful of doing so. Income itself does not provide a full summary of the economic anxieties facing individuals.

One clear lesson from the crisis is the need for a broader range of statistics, including more granular data that capture the diverse situations of different groups of the population. GDP, for example, is not constructed to measure the economic situation of individual households. Indeed, if GDP goes up by say 5%, it doesn’t mean that everyone sees their income go up by 5%, nor even that the typical household or person sees his or her income go up by that amount. GDP describes what is happening to total economic production and to the average income generated from this production – whether this income accrues to a few people or many, to residents or foreigners, to households or to firms. Even if the income is going to residents, GDP doesn’t say how this income is distributed among households.

But the US President should have been aware (or been made aware by his economic advisors) that GDP could go up even though the vast majority of people saw no increase in their incomes – if inequality increases enough. And that was precisely what happened in the US “recovery”. The single number, GDP, didn’t adequately summarize what most people were experiencing. So too, when Ireland’s GDP went up 26% in 2015, it wasn’t that the Irish citizens, who had been suffering so much from the euro crisis, suddenly found their worries at an end. Indeed, Irish household disposable income per capita rose by only 2.7% in that year. In this case, the surge in GDP reflected, to a large extent, the transfer of the intellectual property assets of a few multinationals to independent entities in Ireland, which much of the profits generated by these entities transferred abroad rather than benefitting Irish households.11

That GDP didn’t do all that was hoped of it shouldn’t be a surprise: no single number can summarise anything as complex as the economy. Further, an average is just an abstract number for the individuals who are averaged. But there are real consequences of relying on an incomplete set of measures. If the economy is perceived to be well on the road to recovery, based on GDP, one might not take the strong policy measures needed to resuscitate the economy that one would take with metrics that inform on whether most of the population still feels in recession. Nor would one take measures to bolster the safety net and social protection in the absence of metrics on the extent of people’s economic insecurity.

If the measures we rely on are out of synch with how citizens experience their lives, a lack of trust in government will develop. Some would argue this is what happened in the United States and in most other industrialised countries in recent years, when the GDP statistics said the economy was in recovery and yet most people felt otherwise. While other factors might have been at work (Pew Research Center, 2017), the disparity between what was happening on the ground and the announced “recovery” almost surely contributed to the growing lack of trust in governments by so many citizens over this period. Most OECD countries are today facing a “trust crisis”, a crisis that in some cases goes back in time. To give one example, fewer than 20% of Americans trusted their federal government to do what is right most of the time in 2017, as compared to close to 80% in 1964 (Figure 1.1). Data for a broader range of countries show that confidence in national governments, which hovers today around 40% on average across OECD countries, fell by 10 points or more in many of the countries most significantly hit by the crisis (e.g. Greece, Spain and Portugal) while strongly improving in countries that were less affected, such as Germany (OECD, 2017b).12

Figure 1.1. People’s trust in the United States federal government

Percentage of the population

Note: Before 1985, data were not collected on an annual basis.

Source: Pew Research Center (2015), Historic trends of public trust, www.people-press.org/2015/11/23/1-trust-in-government-1958-2015/.

 StatLink https://doi.org/10.1787/888933842014

It was concerns like these that motivated President Sarkozy to establish the Commission. As an elected official, he worried that making claims about how well the economy was doing that did not correspond to what voters were feeling would undermine trust in the government. He worried too about the conflicting signals he was receiving as a public official. He knew that his performance would be graded partly on how well the economy did. But he also knew that voters cared about many other things than GDP – they cared about the quality of their jobs, their housing conditions, the lower opportunities of their children, they cared about pollution, noise and other things that ought to be incorporated in our measures of economic performance and social progress but were not.13

1.3. The Commission and its central recommendations

The Commission, headed by Joseph E. Stiglitz, Amartya Sen and Jean-Paul Fitoussi, consisted of leading economists and other social scientists.14 Their central question was to assess GDP as a measure of performance, to identify its limitations, and to consider how better metrics of well-being and social progress might be constructed. Its recommendations were aimed particularly at statistical agencies and governments, focusing on the need for more data and research. The Commission provided a number of technical suggestions for improvements in our metrics. But at the core of its recommendations was the belief that our measures matter, and that we should move away from over-reliance on GDP when assessing a country’s health, towards a broader dashboard of indicators that would reflect concerns such as the distribution of well-being and sustainability in all of its dimensions.

How large could a dashboard be? As citizens and officials, there are limits to the amount of information that we can absorb and process. The Commission recommended that there should be a relatively small number of indicators (what it called a dashboard), with some of them reflecting how well ordinary citizens were doing (for example, measures of households median income may be a better indicator of the economic well-being of the “typical” person than average income, which can be pulled up by a few people having enormously high incomes) and some measuring sustainability. Economic sustainability is typically captured best by some monetary measure of wealth, but environmental sustainability might be better assessed by using some physical metrics. So too, it made sense to reflect how well society was doing on health by considering physical metrics of longevity and morbidity – not even trying to convert them into some dollar equivalent.

The challenge was to make the dashboard small enough to be easily comprehensible, but large enough to include a summary of what we care about the most.

1.4. An array of problems

As important as metrics are, all metrics are beset by problems. The choice of metrics is part of the debate in many other areas of public and private life. For example, one of the responses to concerns that our schools are not doing what they should be doing is that we need better student assessments. We can then know when we are failing our children. But, putting aside the question of how those assessments are used, the question is what is to be assessed. Standard tests often focus on basic skills, such as reading, writing and arithmetic. However, schooling is about more than that. It teaches social skills – how to get along with others; it teaches basic life skills – to show up on time every morning, to obey orders, to work as a member of a team. And it teaches higher order cognitive skills that may foster creativity. If student testing focuses only on basic skills, teachers will focus on those at the expense of other skills. Overall, schools will perform less well – though they will perform better in the basic skills. Such considerations led to the changes introduced in the well-known OECD Programme for International Student Assessment (PISA) tests. While PISA started in the late 1990s measuring 15 years-old students’ basic skills in reading, science and numeracy, it now also assesses “soft” skills and students’ well-being in school.

1.5. A national dialogue

The Commission report argued that a national dialogue determining what should be included in the dashboard of indicators, and whether existing metrics well reflected what society cares about, is an important part of democratic engagement. Since then, many civil society groups have come forward, urging changes in our measurement system, and many governments have responded through measurement initiatives (see Annex) and by taking steps to anchor well-being indicators in the policy process (see Chapter 4). The response has, in many ways, exceeded what the Commission had anticipated, reflecting growing concerns around the world over issues like income inequality and the environment. The growing participation in the OECD World Fora on “Statistics, Knowledge and Policy”17 which started in 2004 – before the Great Recession and the establishment of the Commission – and continue to be organised periodically by the OECD to promote the “Beyond GDP” agenda, are a testimony to this success.

There is some tension, though, between the desire to have metrics that reflect the particular situation within a country and the need to have metrics that enable cross-country comparisons, i.e. to give a picture of how a country is doing relative to others. Both perspectives are important: we all want to know how well we are doing (in one dimension or another) relative to our past or relative to what is occurring elsewhere. The OECD Better Life Initiative – launched in 201118 in response to the Commission’s recommendations – and its biennial report How’s Life? Measuring Well-being (OECD, 2017a, see Annex), provide a dashboard of well-being indicators that can be compared across OECD member countries. Individual countries have developed their own dashboards, adapting the framework to reflect their specific circumstances.

At the end of the Commission’s term, one of its main recommendations was that more needed to be done at the international level. This message was taken up by the OECD, who took the lead in hosting the follow-on HLEG. But, rather than follow through on the full gamut of issues raised in the Commission report, this Group decided to focus on selected issues that had not been dealt with adequately or fully in the initial report.

Thus, this book and its companion volume (For Good Measure: Advancing Research on Well-being Metrics Beyond GDP), focus on a selected number of topics, many of which are related to each other: the relevance of the “Beyond GDP” agenda for less developed countries; the distribution of household income, consumption and wealth; horizontal inequalities among people sharing common characteristics; inequality of opportunity; subjective well-being; economic insecurity; sustainability; and trust. Some of these issues were not discussed in the Commission’s earlier report, but have since received increasing attention. This is the case, for example, of vulnerability, the risk that an economy (or an individual) could fall into a non-sustainable state, and resilience, the ability of an economy (or an individual) to recover from an adverse shock. Some issues are old but have become an increasing source of concern: GDP growth cannot continue as it is within our planetary boundaries. Global warming and climate change are real and already upon us, with significant well-being impacts from both weather variability and higher average temperatures, and further effects, such as those associated with sea level rise and changing levels of ocean acidity, not far away. Our indicators must account of these realities.

These subjects were chosen in view both of their importance and of the possibility to develop metrics to measure them. Just listing them conveys their richness and importance for the life of people. The fact that standard metrics have not even attempted to capture changes in these variables helps explain our difficulties in understanding our society and its behaviour, and in devising policies that would enhance individual and societal well-being. Several of the topics discussed in this book and the accompanying collection of authored chapters are at the limit of what we can assess with any reasonable degree of confidence. Measurement problems are complex but that is not an excuse: what we measure is a choice, which reflects implicitly our value system. Not measuring something, even if the measurement is imperfect, has consequences, as we have already noted.

One of the objectives of the report of the Commission on the Measurement of Economic Performance and Social Progress was to highlight the role of metrics in policy, and to instigate a more active dialogue between economic theory, statistical practice and economic policy, bringing out more clearly the often implicit assumptions in our statistical practices and their real-world consequences. The same objectives remain central for this book.

1.6. Key changes in the world

The Commission report explained how changes in the economy and society affect both what we want to measure and the adequacy of our metrics. Before the problem of climate change was recognised, there was no reason to measure an economy’s carbon emissions. The report also noted that the changing structure of the economy had led to an increasing reliance on imputations – numbers not directly collected by official statisticians but estimated by them in some indirect way.

Since the publication of the Commission report, there have been several changes that heightened the need for (and shaped thinking concerning) a new report. While the Commission was formed before the Great Recession of 2008, this long and deep recession naturally focused attention on the adequacy of our metric system for assessing the consequences of deep downturns – in some ways, the original mission of the system of national accounts. It also highlighted the importance of metrics of economic sustainability, indicators showing how an economy might be vulnerable to a major shock.

1.6.4. Digitalisation of the economy

One change in the structure of our economy, particularly notable in the last decade, which was not discussed in the Commission report, is the emergence of the digital economy. Google, Facebook, and other digital firms seem among the most innovative in the world, and they have garnered for themselves a large share of the increase in global stock market value. Other companies report large efficiency gains permitted by digital technologies and algorithms. The success of these firms has led many to conclude that we are in an “innovative era”. Yet, aggregate data covering this “innovative era” suggests no reversal in the longer- term slowdown in productivity growth over the last two to three decades in many advanced economies (Figure 1.2). How do we reconcile being in what some call an unprecedented era of innovation with the slow pace of productivity growth? Gordon (2016) suggests that we are not today in a particularly innovative era: the invention of electricity had far more profound effects on our economy and society than the internet did. Indeed, one could argue that much of the profits of these internet giants, based in large part on advertising revenues, merely reflects a displacing effect (away from traditional forms of media) that has arguably only a minimal net impact on societal well-being. Others have argued that the pace of technological progress has not slowed down, but that adoption requires changes in organisational structures and business models and that productivity growth naturally lags innovations (Brynjolfsson and Mac Afee, 2011; Baily, Manyika and Gupta, 2013).

Some argue that our GDP metrics underestimate the value of the new technologies, because some of their key services are provided for free to users. Google’s search engine, for example, has enormous societal value and yet, because access is free, the value of these services does not directly enter GDP (although their advertising revenues do). There are of course good reasons why the value of these services are not directly included in measures of GDP (Ahmad and Schreyer, 2016). As the search is provided for free, its marginal value is zero, and our system of national accounts values goods at their marginal value (the value of the last unit provided), not incorporating “consumer surplus” (the value of all units consumed up to the last). There are other aspects of new technologies that may be of high value, such as the sense of connectedness provided by social media, which are similarly missed by GDP.

On the other hand, critics point out some of the less positive social impacts of digitalisation, such as societal polarisation and a reduction in people’s attention span. Moreover, at the same time that the new technologies provide services of value, there is a reduction in other services of value, such as those that were previously supplied jointly with print media. Consumers have long been accustomed to free newspapers, investigative reporting and television programmes for example, funded by advertising revenues. That being said, current estimates suggest that the impact of this “exclusion”, at least at present, is not significant and cannot explain the slowdown in productivity growth seen during this “innovative era” (Ahmad, Ribarsky and Reinsdorf, 2017), accounting, at best, for 0.1% per annum of US GDP growth.24 Where potential mismeasurement related to digitalisation may be more important is in the area of prices, particularly in relation to services and new technologies where quality changes rapidly. Even here, however, the impact on real GDP growth and on measures of consumer inflation is estimated at most around 0.2% percentage points per annum (Schreyer and Reinsdorf, 2018).

Figure 1.2. Productivity trends

Percentage change at annual rate, total economy

Note: Labour productivity is measured as GDP per hour worked. Trends are estimates using the Hodrick-Prescott filter (OECD, 2018e, OECD Compendium of Productivity Indicators 2018, OECD Publishing, Paris, https://doi.org/10.1787/pdtvy-2018-en).

Source: OECD (2018f), OECD Productivity Statistics (database), https://doi.org/10.1787/pdtvy-data-en.

 StatLink https://doi.org/10.1787/888933842033

1.7. Advances in data availability

We have discussed many of the changes in the economy that call for changes in metrics. An important enabler to the development of these new metrics is the availability of new data sets and advances in technology that allow users to rapidly process and analyse large bodies of data and to integrate different data sets. This is a theme that will be repeatedly noted in this book and even more so in its accompanying volume. For instance, indicators of the income of the very rich could be derived from administrative tax returns, and a better picture of the entire income distribution could be achieved by merging numerous administrative data sets (e.g. tax and social security) and survey results. One of the themes explored by the HLEG was the integration of national accounts with distributional data (Distribution National Accounts, or DINA; see Chapter 6 in the companion volume). One of the topics not explored by the HLEG, however, was harnessing the potential of big data, such as scraping data from the internet or exploiting geo-spatial information to form indicators, sometimes in real time. This subject is likely to become of increasing importance in the future – although care is needed to ensure that such data comply to well-established quality criteria such as ensuring that they are representative of the entire population.

1.8. Policy implications of “Beyond GDP” metrics

As we noted earlier, a major impetus for work going “Beyond GDP” was to improve policy: what you measure affects what you do. If you don’t measure important phenomena, you may not act on them. But distorted metrics can be even worse. When economic metrics don’t include the effects of environmental degradation, we have no way of knowing whether a country’s welfare is increasing when it expands coal-generated energy, for example. The owners of coal mines might argue that GDP is being sacrificed in the name of the environment if the coal mine is shut down, when in reality, correctly measured, total welfare would increase when the value of the lost GDP is more than offset by the environmental benefits of cleaner air and reduced risks of climate change. Similarly, a developing country might allow a foreign company to develop a gold mine within its territory, believing that doing so will increase GDP, while a correct account of the impact of the mine on the well-being of the residents in the country might suggest otherwise. In some (but far from all) countries, better metrics than GDP are produced within the system of national accounts – such as measures of national product and national income, sometimes net of the depreciation of capital goods used in production – that would allow better decisions. But these indicators are not routinely used by those making decisions.

Similarly, some may suggest that privatising social security will increase GDP as a result of the increased efficiencies and associated savings by households, while a correct accounting of the value of social security could show that the private sector may in fact be less efficient than the government in providing these services even when the private sector generates profits, and that privatisation may increase individuals’ economic insecurity.

One of the key examples we develop later is that, because GDP metrics do not fully capture all of the adverse effects of economic downturns on people’s well-being, less may have been done by policy-makers to combat the recession than would have been the case had we had relevant additional or better metrics.

1.9. Conclusions

The Commission report, published in English with the title Mismeasuring Our Lives: Why GDP Doesn’t Add Up (Stiglitz et al., 2009),25 not only spurred important work by National Statistical Agencies and academics, but also helped to found a global movement, which has taken the form of Parliamentary Commissions, statistical initiatives involving civil society, and central and local government initiatives to use “Beyond GDP” metrics in their policy determinations. At the international level, one key set of moments of this global movement is represented by the periodic World Fora on “Statistics, Knowledge and Policy” organised by the OECD, bringing together civil society, academics, national statisticians and government officials to move forward the “Beyond GDP” agenda. There have been conferences in Palermo in 2004, in Istanbul in 2007, in Busan in 2009, in New Delhi in 2012, in Guadalajara in 2015, and in Incheon in 2018 (where we are launching this book), with each meeting gathering more people than the previous one.

A key change that has occurred since the Commission report is the broader acceptance of the need to develop measures of societal welfare, and a stronger appreciation that policies based narrowly on increasing GDP may be misguided.

There were multiple reasons for this increased interest. Environmentalists worried that a focus on GDP resulted in inadequate attention to the environment. And for good reason: current GDP measures pay no attention to resource depletion or environmental degradation. Stiglitz’s experience as a member of the US Council of Economic Advisers under President Clinton is telling. Working with the Under-Secretary of Commerce, he pushed to expand conventional GDP towards a metric of “green GDP”, including resource depletion and environmental degradation. Increased coal production might add to conventionally measured GDP, but that does not take account of the adverse effects on health from the increased air pollution, or on climate change or on the local environment. Perhaps not surprisingly, the coal lobby struck back, and Congress threatened to cut off financing to the federal agencies undertaking this work if they continued to develop these metrics. These responses made it clear that the coal industry itself believed that what a country measures affects what it does. If a country measures environmental degradation, it is more likely it will do something about it. And that is precisely what the politicians from the coal states did not want to happen.

The same applies in other fields. Those who worried about climate change expressed concerns about the absence of measures on sustainability and resilience – and argued that, without such measures, we would pay insufficient attention to these key attributes of a good economic system.

Those who wanted more active government policies – sometimes involving higher public expenditures – also wanted a change in metrics. As the next Chapter will make clear, the costs of recession are not fully captured by GDP. By understating these costs, one is less likely to take strong counter-measures. So too, a focus on government liabilities, without looking at assets, induces excessive conservatism in government borrowing in a recession.

Those who were concerned about social justice, about inequality of outcomes and of opportunity, worried that the focus on GDP had resulted in insufficient attention to “how the pie was sliced”. Countries where GDP was increasing but where most people were not doing well were rated as performing as well as those where the benefits of GDP growth were more equally shared. They should not have been.

All these reasons have translated into large number of measurement initiatives, at both the domestic and international levels. When presenting the Commission report in 2009, President Sarkozy asked INSEE, the French National Statistical Office, to ensure the implementation of the Commission’s recommendations at the national level, and the OECD to do the same at the international one, as well as committing his government to bring the Commission’s arguments to the attention of the whole international community.26 The response by the statistical community has gone beyond our expectations, and explains the presence of several official statisticians in the HLEG. These statistical initiatives (which are described in more detail in the Annex) have followed a double track. On the one side, they aimed at making better use of statistics that already existed (e.g. on household living standards, inequalities, environmental pressures), bringing them together into dashboards aimed at providing a comprehensive picture of current well-being and of its sustainability. On the other side, they aimed at building the foundations for new and better statistics on the many aspects of concern to people and communities that are still inadequately covered by official statistics. Progress in both directions is needed to meet the Commission’s goal to shift the focus of our statistical system from measuring the volume of economic production to measuring the conditions of people “here and now” and assuring their sustainability tomorrow.

The response to the Commission report has also gone beyond the development of new metrics, to explore ways of making systematic use of these metrics directly in the policy process. As highlighted in Chapter 4, having a dashboard of well-being indicators is, in some cases, not enough. What is needed are institutional tools to force public agencies to act based on the evidence provided by these indicators, and to overcome the silo thinking whereby each agency pursues its own narrow goal, without paying attention to how their actions impact on the goal pursued by other government departments and with unintended negative effects being recognised only after the damage has been done. In some countries (e.g. France, Italy), national parliaments, assigned the task of allocating budgets, have pushed for better metrics, because they worried that inadequate metrics would induce wrong budgetary allocations. In others (e.g. New Zealand) it was Treasuries who took the lead in developing broad frameworks – encompassing sustainability, inequalities and risks – as tools to help their officials when designing policy reforms or providing advice to other government departments. In other countries yet, policy use of well-being metrics has focused on aligning priorities and actions across government departments and levels of administration (Scotland) and on identifying “what works best” in achieving specific goals (the United Kingdom), leveraging the initiatives of various public agencies and actors in the pursuit of higher level goals.

While all these initiatives, which are described in more detail in Chapter 4, are fairly recent, they share a common ambition: to anchor in policy practice societal concerns that currently don’t receive the attention they deserve by policy-makers. The hope is that a discussion of measurement would lead to more democratic engagement around the central questions facing all societies. What do we value? What should we measure? What should we pay more attention to?