PREFACE
In an increasingly performance-oriented society, metrics matter. What we measure affects what we do. If we have the wrong metrics, we will strive for the wrong things. In the quest to increase GDP, we may end up with a society in which citizens are worse off.
Too often, we confuse ends with means. One of the criticisms of our economies in the years prior to the crisis is that they did exactly that—a financial sector is a means to a more productive economy, not an end in itself. It is even worse to confuse an improvement in a measurement of well-being with an improvement in well-being itself. Our economy is supposed to increase our well-being. It too is not an end in itself.
The objective of this international Commission was to align better the metrics of well-being with what actually contributes to quality of life, and in doing so, to help all of us to direct efforts to those things that really matter.
The Commission was appointed by Nicholas Sarkozy, President of the Republic of France, in early 2008 in response to increasing concerns about the adequacy of current measures of economic performance, in particular those based on GDP figures, and to broader concerns about the relevance of these figures as measures of societal well-being, as well as measures of economic, environmental and social sustainability.
The Commission was charged with looking at the entire range of issues, and was given full independence in the conduct of its work. The Commission consisted of an international panel of experts, chosen for their expertise in one or another area of enquiry. Its aim was to identify the limits of GDP as an indicator of economic performance and social progress, to consider additional information required for the production of a more relevant picture, to discuss how to present this information in the most appropriate way and to assess the feasibility of alternative measurement tools. The Commission’s work was not focused on France, or on developed countries. It was hoped that the output of the Commission would provide a template for every interested country or group of countries, and would spur further work and discussion on these issues around the world.
The Commission was chaired by Professor Joseph E. Stiglitz, Columbia University. Professor Amartya Sen, Harvard University, served as Chair Adviser. Professor Jean-Paul Fitoussi, Institut d’Etudes Politiques de Paris, President of the Observatoire Français des Conjonctures Economiques (OFCE), was Coordinator of the Commission. Members of the Commission are renowned experts from universities, governmental and intergovernmental organizations in several countries (United States, France, United Kingdom, India). Rapporteurs and secretariats are provided by the French national statistical institute (Insee), OFCE and OECD. The full list of commissioners is provided on p. vii.
The final report of the Commission was publicly presented on September 14, 2009, in a discussion led by President Sarkozy, with contributions from members of the panel and comments from heads of international organizations and French ministers: Angel Gurria, Secretary General of the OECD; Juan Somavia, director general of the International Labour Office; Jacques Barrot, Vice-President of the European Commission; Christine Lagarde, Minister for Economic Affairs, Industry and Employment; Chantal Jouanno, Secretary of State, Minister of Ecology, Energy, Sustainable Development and Sea, in charge of Green Technologies and of Climate Change Negotiations. The IMF and the World Bank were also represented at a high level.
National Income Accounting Goes from the Province of Technicians to a Subject of Public Discourse
It is easy to understand why suddenly a set of issues that have been the province of technicians have become a source of public policy debate. Trying to understand what makes for good performance of a society is central to the social sciences. We see the world through lenses not only shaped by our ideologies and ideas but also shaped by the statistics we use to measure what is going on, the latter being frequently linked to the former. GDP per capita is the commonly used metric; governments are pleased when they can report that GDP per capita has arisen, say, by 5%. But other numbers can tell a very different picture. In Russia, declining life expectancy suggests there are underlying problems, even if GDP per capita is rising. So, too, in the United States, most individuals saw a decline in income, adjusted for inflation, from 1999 through 2008—even though GDP per capita was going up—providing a markedly different picture of performance. Such a disparity may arise when income inequality increases at the same time that income increases.
Metrics Shape Our Beliefs and Inferences
The theories we construct, the hypotheses we test and the beliefs we have are all shaped by our systems of metrics. Social scientists often blithely use easily accessible numbers, like GDP, as a basis of their empirical models, without enquiring sufficiently into the limitations and biases in the metrics. Flawed or biased statistics can lead us to make incorrect inferences. In the years preceding the crisis, many in Europe, focusing on higher GDP growth rates, suggested that they should follow the American model. Had they focused on other metrics (such as median income) or if they had made appropriate corrections for the increased indebtedness of American households and the country as a whole—with the consequent risk of non-sustainability—their enthusiasm might have been more muted.
Inevitably, economists attempt to draw inferences about the desirability of policies by making comparisons over time or between countries, and if the metrics employed are imperfect, there is a risk of biased, distorted and flawed inferences. We worry that if there are systematic measurement errors in, say, the output of the public sector, inferences made about the consequences of a large public sector on overall economic performance may be biased, simply because the larger the sector, the more the distortion. Most of the vast amount of empirical work making such “cross-country” comparisons is insufficiently sensitive to these limitations.
Metrics and Policy
For a political leader, these are matters that are not just academic. Flawed inferences affect economic policy. Metrics that give short shrift to the environment (to air, water or noise pollution) put insufficient weight on something that is of vital and increasing importance to many citizens. A political leader ignores these concerns at his peril.
A political leader attempting to fulfill the wishes of his citizens and promote their well-being is pulled in different directions: he will be graded on economic performance, even though much of that is out of his control. But citizens also care about many dimensions of the quality of life—including the quality of the environment. Current metrics suggest that there may be trade-offs—one can improve the environment only by sacrificing a growth measure. But if we had a comprehensive measure of well-being, perhaps we would see this as a false choice: it could indicate an increase in well-being if we improve the environment, even if conventionally measured output went down.
And it is for the same reason that the work of the Commission has drawn such interest from civil society.
Statistics and Information Theory
There is hardly a decision in modern life that is not colored by our statistics and accounting frameworks. The focus of Stiglitz’s research over many years was on how information affects economic and political decisions. Our statistical and accounting systems provide, as we have noted, an important part of the framework through which we see and analyze the world. They are critical parts of our “information” systems.
Metrics that seem out of synch with individuals’ perceptions are particularly problematic. If GDP is increasing, but most people feel they are worse off, they may worry that governments are manipulating the statistics, in the hope that by telling them that they are better off, they will feel better off. In these cases, confidence in government is eroded, and with the erosion of this confidence, the ability of government to address issues of vital public importance is weakened.
While these public policy concerns were part of the impetus for the work of the Commission, the commissioners decided early on in their work to limit themselves to a focus on our statistical system itself, and not to extend their work to the policy implications that might follow from having a better statistical system. In some cases, these policy implications are obvious, and it would have been easy to achieve consensus among the commissioners. In other cases, reasonable people may come to different policy stances.
On these issues of the reform of our statistical system, what was remarkable was the degree of unanimity among the members of the Commission.
Why Reconsider Metrics Now?
Many of the problems with GDP statistics have been well known. But several factors gave the efforts of our Commission particular relevance. There have been changes in our society and in the structure of our economy that make some of the limitations of GDP accounting of more concern. We mentioned one already: if there is increasing inequality, as there are in many, if not most, countries around the world, there may be an increasing disparity between average income and median income (the income of the representative individual); one may be increasing while the other is declining.
The problems of measurement of government services that are not sold on the market are well known, but these problems become of increasing importance as the share of government expenditures in OECD countries (on average) increases, as it has from about 25% to more than 45% in the last 50 years.
It is obviously more difficult to assess the quantitative importance of quality improvements than to count the quantitative increases in, say, the number of cars; but such quality improvements are of increasing importance. If every family has a car, increased consumption of “automobile services” will take the form of better cars, not more cars. But then we have to have ways to measure these quality differences. This is especially difficult when we have to measure the growth of services.
Globalization itself has meant that the difference between the well-being of the citizens within a country may differ markedly from the output produced within a country. Ironically, the measure focusing on the former, GNP, grew out of fashion, giving way to GDP, which focuses on production, just as globalization was making the difference more important. There are obvious political consequences to the distinction.
When problems of globalization and environmental and resource sustainability are combined, GDP metrics may be especially misleading. A developing country that sells a polluting mining concession with low royalties and inadequate environmental regulation may see GDP increase but well-being decrease.
There are concerns too that a focus on the material aspects of GDP may be especially inappropriate as the world faces the crisis of global warming. Should we “punish” a country—in terms of our measure of performance—if it decides to take some of the fruits of the increase in productivity from the advancement of knowledge in the form of leisure, rather than just consuming more and more goods?
The Crisis as an Opportune Moment
There are several reasons that the timing of President Sarkozy’s initiative was particularly opportune. One was the crisis that overcame the world just as our work began.
The Commission was established before the recession hit; and the scientific work of the Commission, reflecting ongoing research on the underlying problems of the measurement of economic performance and social progress, was not affected by the crisis. But, especially for some members of the Commission, the crisis heightened the importance and relevance of the Commission’s work and underscored certain problems with which it had been grappling. Even before the crisis, Commission members had expressed a belief that a good set of metrics capture the notion of economic and environmental sustainability. It was clear that GDP by itself did not do this. Similarly, before the crisis, Commission members had expressed a concern about the appropriate use of market prices, especially for evaluating long-run sustainability. The crisis has illustrated the importance of both of these concerns. The seemingly strong performance of some countries prior to the crisis (as indicated by GDP) was not sustainable and was based on “bubble” prices that exaggerated profits and output.
Advances in Research
While changes in our economies and societies, including those to which we alluded earlier, have resulted in a sense that the old metrics may be increasingly deficient, especially as measures of well-being, advances in research across a number of disciplines enables us now to develop broader, more encompassing measures of well-being. Some of these dimensions are reflected in traditional statistics but are given more prominence: unemployment has an effect on well-being that goes well beyond the loss of income to which it gives rise.
The timing of the Commission was opportune for another reason. Criticisms of the traditional measures were being expressed in many quarters, including by civil society. Changes in the structure of our economy raised questions about whether some of the assumptions made in the past were still appropriate. Global warming had put issues of sustainability front and center.
For some of these, there may be objective metrics, but for others, replicable subjective assessments may provide the best approach to measurement. Individuals may, for instance, be affected by their sense of security and by their bonds with others. But even the seemingly non-economic factors are affected by economic structures. Reforms in the workplace may lead to increased market efficiency but lower worker job satisfaction and therefore a reduction in their sense of well-being. Some economic reforms in recent years may have increased GDP but may have had adverse effects on important dimensions of quality of life. For instance, one of the criticisms of globalization (in the way it has proceeded) is that it has contributed to the weakening of a sense of community, thereby leading to a decrease in a sense of well-being. It is important to have metrics that would allow us to assess such claims.
This work is just at its beginning stage, and yet the results obtained so far are extremely promising. It has been clearly established that replicable measurement of many of those factors affecting well-being and the quality of life is possible.
A Single Metric or a Dashboard?
The purposes of our statistical systems are multiple, and a metric that is designed for one purpose may be ill-suited to another.
Changing Objectives
National income statistics like GDP and GNP were originally introduced to provide a measure of the level of market-based economic activity (including the public sector but excluding home production). Especially after the development of Keynesian economics and in the aftermath of the Great Depression, as governments took on the responsibility of managing the economy, it became important for them to have statistics that described the state of the economy. To manage an economy without indicators of how well the economy was doing has been described like trying to fly an airplane without instruments. Two of the pioneers in this work, Simon Kuznets and Richard Stone, both received the Nobel Prize, in part for their contributions in creating systems of national income accounts.
Much of the work of national income statisticians in subsequent years has been directed at correcting imperfections in the measure of market activity (for instance, the valuation of housing services or government activity). Some of the work of national income statisticians has been directed at expanding the scope of economic activity, to include, for instance, home production.
But these metrics have increasingly been thought of as measures of societal well-being. Of course, good national income statisticians have warned against these abuses, even as they have worked hard to make our measures better reflect the real level of economic activity and increasingly focused on measures, say, of the real income of households.
Much economic activity occurs within the home—and this can contribute to individual well-being as much or more than market production. A shift in the locus of production may not necessarily be indicative of an improvement in well-being.
Still, this work has centered on attempting to adapt market measures so that they better reflect societal well-being. Today, there is a growing demand for more encompassing measures of social progress and societal well-being, which, while incorporating metrics of market activity, are not limited to such metrics. One contribution of the Commission was to provide further impetus to these attempts.
The Need for Multiple Metrics
There is no single indicator that can capture something as complex as our society. Trying to capture what is going on by using a set of numbers that is too small can be grossly misleading. We might want to know how fast we are driving (55 miles an hour) and how far we can go before we run out of gasoline (250 miles), but a single metric, say formed by adding the two numbers (305) would tell us nothing about either question.
We care about how we are doing “in the aggregate,” but we also care about what is happening to the distribution of income. We care, moreover, not just for how well-off we are today, but for how well-off we will be in the future. If we are borrowing from the future, we at least want to know that our current level of well-being is not sustainable. While there are many dimensions of sustainability, environmental sustainability has taken on increasing importance, especially with the realization that, with global warming, the world is currently on an unsustainable path.
The goal, then, is to construct a simple set of metrics that captures much that is of central concern. That is why we should expect that a revised GDP measure will continue to be used as a measure of market activity; but it will be supplemented by measures that reflect more broadly what is happening to most citizens (measures of median income), what is happening to the poor (measures of poverty), what is happening to the environment (measurements of resource depletion and environmental degradation) and what is happening to economic sustainability (measurements of debt).
Global Resonance and Global and National Dialogues
The concerns raised by President Sarkozy and our Commission have, not surprisingly, struck a global chord. There is resonance throughout the world. Even before the work of the Commission, Bhutan was hard at work creating a measure of GNH, Gross National Happiness, and Thailand was working on its own index.
Because what we choose to measure and how we construct our measures can have such an important role in the decisions that are made, it is important that there be an open and public discussion of our system of metrics.
Hopefully, this report will play a role in this public dialogue. Indeed, we believe that our report may stimulate such a dialogue, as countries come to analyze what is important to them, and whether their systems of metric adequately capture these values. The vitality of the discourse that is already occurring is exemplified by the Third OECD World Forum on “Statistics, Knowledge, and Policy,” held in Busan, Korea, on October 27–30, 2009, where participants considered how better ways of measuring the progress of societies could play an important role not only in charting progress but also in “building visions” and “improving life.”
Unfinished Business
We view our study as neither the beginning nor the end of a journey. The Commission was fortunate to be able to draw from a large body of work on the issues with which we were concerned. The early developers of GDP metrics were clearly far more aware of the assumptions that went into the construction of the index than many of those who have subsequently found the measure of such use. But by reminding modern-day researchers of the limitations and biases in existing measures, we hope that some of our analyses will not only lead to better metrics but will also unleash a flood of studies to help us understand the sensitivity of, for instance, the inferences of the metrics used.
We devoted considerable efforts to thinking about what kinds of reforms in our metrics might lead to better indicators. But often the data required is not available. Thus, much of our report is devoted to recommendations for future work, including that by statistical agencies in gathering data.
This is just the first step in what should be an ongoing effort. Even if we had succeeded in constructing the perfect measure for today, changes in our economy and our society would necessitate constantly revisiting these issues. But we are far from that goal.
The effort to improve our statistics also needs to be a global effort, one in which political leadership is necessary if we are to obtain the requisite momentum. That was one of the reasons that so many of the commissioners, who had devoted so much of their professional lives to these issues, were so enthusiastic about this initiative: President Sarkozy was providing the political impetus.
The reception to our report of September 2009 has been heartening—and reinforced our conviction of the importance of the issues raised by the Commission. Our hope was that the work of the Commission would lead to a broad dialogue on societal objectives, whether commonly used metrics were consistent with those objectives, and whether there were alternatives that would be more consonant with broadly held values.
The timing of our report was, in this sense, fortunate. It coincided with broader skepticism by many citizens about the direction in which society was going. Global warming had become a paramount concern in many parts of the world, and yet no account had been taken of the adverse effects of the increase in materialistic consumption—seemingly “lauded” by GDP measures—and whether the planet could survive this increase.
At the launch of our report in Paris on September 14, 2009, both President Sarkozy and Angel Gurria, the Secretary General of the OECD, committed to carrying forward the work of the commission. At the Third International Knowledge Forum, the work of the Commission helped shape discussions of new approaches to assessing well-being. A number of countries have begun work implementing the ideas of the Commission. And President Sarkozy carried the discussion of these issues to the G-20 meeting in Pittsburgh shortly thereafter.
Personal Notes
As we noted, many members of the Commission had been involved in these and similar issues for much of their professional lives. They have been advocating reforms in our statistical systems and doing the research on the basis of which such reforms could be made.
Amartya Sen had worked, along with Mahbub ul Haq, in the establishment of the Human Development approach of the United Nations, with systematic presentation of information related to the well-being and freedom of people—of which the Human Development Index (the well-known HDI) is the simplest representation. The HDI showed that the ranking of countries using a broader metric, which included health and education, could be markedly different from that focusing just on income. The capabilities approach that Sen had previously developed was very influential in the thinking of the members of the Commission.
We noted earlier that accounting frameworks, at both the corporate and national levels, are an essential part of our information systems. Stiglitz has long been concerned (both in his theoretical work, and in his work as Chairman of President Clinton’s Council of Economic Advisers and Chief Economist of the World Bank) about how we could improve these accounting frameworks and enhance the quality of the information that forms the bases of decision making in our economic and political systems. Some of our proposed reforms were far less ambitious than those proposed here—simply a better accounting of resource depletion and environmental degradation. And yet, political resistance was so great that these initiatives were thwarted. It showed the power of information. There were those who were afraid of the light that better information systems might shed. That key interest groups did not want this kind of information to be publicly disseminated suggested these reforms in our statistical system might have real impacts.
As chief economist of the World Bank, Stiglitz was particularly concerned about how judgments about impacts on GDP could lead to wrong decisions about resource development. Many developing countries were being urged to privatize natural resource extraction, even if meant that much of the profits went abroad. GDP would be increased by the mining activity. But when account was taken of the fact that the profits accrued to those abroad, GNP might not increase. And when further account was taken of the depletion of resources and adverse effects on health and environment, it was even clearer that the citizens of the country might be worse off.
The focus of Fitoussi’s research over many years was to show how changes in the distribution of micro-economic variables affected not only the perception of macro-economic variables but also the conclusions of macro-economic models. Inequality is of the essence, and not taking it appropriately into consideration may lead to the wrong inferences about macro-economic policies.
For so many of the members of the Commission, then, the opportunity to work on this Commission was a chance to push an agenda with which they had long been associated, to give it momentum on a global scale.
This Volume
Though the work of the Commission was written largely by and for social scientists (mostly economists, but the Commission included a prominent political scientist, a prominent sociologist and a prominent economic psychologist), we wanted to reach out more broadly. As we noted, we were interested not just in better metrics, but in using a discussion of metrics to engage a broader dialogue about societal values and objectives. The chapters contained in this book represent the non-technical report of the Commission; the technical chapters are available at the web site of the Commission:
www.stiglitz-sen-fitoussi.fr. But to set the context for the issues under discussion, we begin the volume with a broad overview, written by Fitoussi, Sen and Stiglitz.
The subject of our inquiry was so broad and complex that, in the initial meeting of the Commission, it was decided to divide the work among three working groups. One would focus on the standard but difficult issues of national income accounting: measurement of output of government, adjustments for an open economy, treatment of household production and leisure and “defensive” expenditures—expenditures required just to maintain the status quo, for example, through maintaining security. The second focused on efforts to measure “quality of life,” the sense of well-being. And the third focused on sustainability. Global warming had brought the issue of environmental sustainability front and center; the global financial and economic crisis had raised similar questions for economic sustainability. A cross-cutting focus, affecting the work of all the working groups, concerned distributional issues—how to capture appropriately the diverse situations confronting different individuals. Most of the statistical indicators focus on averages; but when inequality is changing, what happens at the bottom, or even the middle, can differ markedly from what is happening, say, to per-capita GDP.
The initial meeting of the Commission was held in Paris, in April 2009. Subsequent plenary meetings of the Commission and of the working groups were held in New York and in Paris. Preliminary versions of the Report by the Commission were posted on the internet site. Many of the suggestions by commentators are reflected in the final draft of the Commission report.
Acknowledgments
We could not have accomplished what we did were it not for the vast amount of work on the subject that had been done by national income scholars and statisticians over an extended period of time, only some of which could be acknowledged in the individual chapters.
In the years after Kuznets’s and Stone’s pioneering work, vast amounts of work have been done on many of the areas under discussion here: the measurement of output in the public sector or inside the household, the measurement of resource depletion and environmental degradation, the particular problems that arise in “open economies,” dealing with the consequences of inequality or ascertaining sustainability and measuring “happiness.”
We need to recognize the important work of the OECD, not only in helping to give prominence to the issues at hand and for their technical contributions to the subject but also for the support they gave to the work of the Commission, including through Enrico Giovanni, the OECD’s former chief statistician, who spearheaded their efforts in this arena and served as chair on one of the Commission’s three working groups.
We conclude with a word of acknowledgment to all of those whose contributions played such a role in the work of the Commission. We first want to thank the President of the French Republic, Nicholas Sarkozy, for convening the Commission, for the support he has given us, and for the complete freedom he has provided us in the conduct of our work. We also want to thank the other commissioners for the enormous amount of effort they put into the work of the Commission and for their sense of commitment and purpose, which allowed the resolution of even quite disparate positions. Given this hard work, we hesitate to single out any commissioners for attention, but we feel we would be remiss if we did not acknowledge those who served as chairmen of the three working groups—Enrico Giovanni, Geoff Heal and Alan Krueger, who served, respectively, as chairs of the working groups on problems with “standard” national income statistics, on the measurement of sustainability and on metrics of well-being and quality of life. We need also to acknowledge the work of the rapporteurs, headed by Jean-Etienne Chapron, who took to heart all the comments and suggestions made by the commissioners and managed to produce the final report. Finally, we want to thank the French Ministry of Economic Affairs and the French National Statistical Institute (INSEE) for the logistic and intellectual help they provided during the eighteen months of the functioning of the Commission.
—Joseph E. Stiglitz, Amartya Sen, and Jean-Paul Fitoussi
January 2010