In February 2008, French President Nicholas Sarkozy created a committee with a somewhat ungainly name, ‘The Commission on the Measurement of Economic Performance and Social Progress’, headed by Joseph Stiglitz, and advised by Amartya Sen. The aim of the Commission was to ‘identify the limits of GDP as an indicator of economic performance and social progress, including the problems with its measurement; to consider what additional information might be required for the production of more relevant indicators of social progress; to assess the feasibility of alternative measurement tools, and to discuss how to present the statistical information in an appropriate way’ (p. 7).
The Report aims to speak to political leaders, policymakers, the academic community (in particular statisticians) and civil society organisations. The Report itself is lengthy, running to 292 pages. This includes a 12-page executive summary, a ‘short narrative’ of around 60 pages, followed by approximately 200 pages of ‘substantial arguments’. Both the short narrative and the substantial arguments are divided into three chapters, the first dealing with classical GDP issues, the second with the quality of life, and the third with sustainable development and the environment. The first two are of greatest potential interest to economic historians and are covered here.
Given that the Commission was created by the French government, it is dominated to a surprising extent by Americans. Of the 24 members of the Commission, 13 were based at universities or other organisations in the United States, seven in French institutions, three in British universities, and one from an Indian university. That said, the people selected to some extent come from ‘the left’, at least in so far as they are not market fundamentalists. For example, the three British representatives, Tony Atkinson, Andrew Oswald and Nick Stern, are all associated in different ways with challenges to economic orthodoxy.
The Report trots out all the standard lines. What we measure affects what we do. There is a difference between official measurements of key indicators, and widespread perceptions. Aggregates hide inequality. Depreciation and environmental damage are poorly captured. The Report also makes claims that not everyone will accept: ‘we are facing a looming environmental crisis’ (p. 9), although, to be fair, the Report acknowledges that not even every member of the Commission agreed with that statement. It also argues that ‘those attempting to guide the economy and our societies are like pilots trying to steer a course without a reliable compass’ (p. 9).
The Report makes a number of recommendations, and it is on these that it will ultimately be judged. It asks for a greater emphasis on income and consumption, rather than on production, with greater weight on the experience of individual households rather than the economy as a whole, and on wealth as well as income. It argues for greater emphasis on inequality, and the inclusion of nonmarket activities. It argues that we should devote greater effort to looking at issues such as social connections, political voice, and insecurity as other aspects of the quality of life. It argues that both objective and subjective wellbeing are important, and that surveys are a good way of finding out about wellbeing. Finally, it argues that sustainability requires what it describes as a ‘dashboard of indicators’, which would include atmospheric concentrations of greenhouse gases, and depletion of fishing stocks.
At one level this Report is correct. Gross domestic product is the most widely used measure of economic activity, and it has any number of flaws. It does only measure market production, it does use market prices rather than taking into account consumer surplus or externalities. It is poor at coping with quality changes, particularly in services.
And yet in another level, this Report manages to be both correct and pointless. In my time I have met many politicians, many policymakers, many in the academic community including statisticians and any number of representatives of civil society organisations. And in all of that time I have never met a single person who felt that their job was to maximise GDP or GDP per head. Every single person from all of these categories that I have ever met has understood, implicitly as much as explicitly, that GDP is not the be all and end all of success. Indeed, many years ago I was at a political meeting in which someone in the audience accused politicians of being obsessed simply with GDP per head. One of the politicians on the platform – now a member of the British Cabinet – replied that this was self-evident nonsense. He listed a whole range of policies that would raise income per head, but which were completely unthinkable. He pointed out that if we were really dedicated to raising GDP per head we would favour euthanasia for almost all old people, since the vast majority of old people have income lower than the national average. Indeed, he went further and pointed out that since we can predict with a moderate degree of accuracy someone's likely income by the time they start school, any government truly dedicated to maximising GDP per head would start its euthanasia programme at the preschool level. He went on to list a huge range of other policies that we actually enact that clearly reduce measured GDP per head. Planning laws that protect green spaces are one example. Legislation that governs maximum pollution levels from factories is another.
In modern democratic societies rises in income per head are a useful summary measure of economic development. We know that in the last couple of years the fall in GDP per head has measured something very real: a recession. We also know that the rise in GDP in the previous decade also measured something very real: a boom. Anyone with any understanding of the nature of economic development could have predicted that the boom would be followed by a slump, but no one if they are honest could predict with any reliability the exact point at which boom would turn to bust, or the depth or duration of the recession. We have not had the current recession because we failed to follow the sort of advice given by this Commission; we have had the current recession because cyclicality is part of the nature of the economic system. What the Commission emphatically does not make clear is what decisions politicians made in the absence of this Commission's Report, that they would not have made had they read this Report.
Indeed, where the Report makes very specific recommendations it should worry citizens were they to be followed. Let us take the very first recommendation that is made: to emphasise national accounts aggregates other than GDP (section 3, pp. 23–30). In this section the authors make a number of sensible points. They note that we should consider the depreciation of capital goods. But for the vast majority of countries the ratio between net national disposable income – their preferred measure – and gross domestic product is well within the margins of error of being a constant over time. France and the United States are given as examples of exactly this. In the case in Ireland, however, the figures are very different, and the Report is right that the figures for net national disposable income demonstrate that the ‘Celtic Tiger’ was to some extent a fiction. Clearly, for Ireland this paragraph of the Report is worth reading, but for citizens in most countries it will make very little difference.
Second, the Report calls for improving the measurement of the quality of services. And yet, as they note, ‘it is difficult even to define’ things such as the ‘general service level of the staff, choice and presentation of products and so forth’ (p. 26). Too true! Of course, we can point to some aspects in which quality appears to have declined: Ryanair has removed the seat back pocket from its planes, for example, and casual empiricism seems to suggest that for those of us who travel economy class, legroom is in ever shorter supply (for those who travel at the front of the plane, the development of flat beds represents a major step forward on long-haul flights). But taken as a whole it seems at least plausible that the level of service, the presentation of products, and so on and so forth has risen broadly in line with GDP. It would certainly take some evidence to the contrary before I at least would be willing to spend serious amounts of government money working out alternative indicators in this area. Equally, the Report is correct to note that the United States spends more per capita on healthcare than most other countries, and that measuring output by input, as is common in healthcare, does not give a sensible result. But what the Report fails to do is tell us how to better measure the output of government-provided services such as healthcare. Again, it suggests that surveys may be a useful way forward, but I am pretty sceptical of people's ability to rate the quality of medical care on a sort of ‘trip advisor’ scale. My daughter was cured of juvenile idiopathic arthritis, a relatively rare condition. Even though I have read much of the medical literature on juvenile idiopathic arthritis, and even though I attended a conference on child rheumatology, I still have no sensible method of assessing whether the time it took to diagnose my daughter represented a good performance, an average performance, or a poor performance either in the light of UK NHS expenditure in general, UK NHS expenditure on child rheumatology, or in the light of global best practice. And yet if I cannot do it for as straightforward a surgical procedure as this, having read widely on the subject, what hope is there to assess healthcare by consumer surveys internationally?
Third, the Report argues that we should focus on household consumption, rather than total final consumption. They note that this would exclude things such as prisons, military expenditure and the clean-up of oil spills. This is perhaps the most bizarre recommendation of all. It implies that the provision of prisons, military expenditure and the clean-up of oil spills have no beneficial effects on human welfare. A government setting out to maximise household consumption should therefore imprison no one, spend nothing on the military and ignore any oil spills that occur. Writing this review in the summer of 2010, having seen Barack Obama's opinion poll ratings plunge in response to the oil spill in the Gulf of Mexico, I am sceptical that ignoring oil spills is actually what citizens want. I am equally certain that citizens gain some utility from knowing that people who might be dangerous are locked up, and that countries that might attack them are deterred, and can be repulsed if necessary. That is not to say that the specific levels of expenditure on prisons or defence are always appropriate, but to say that expenditure on these things should be excluded from some measure that aims to capture the welfare of citizens is positively bizarre.
Fourth, this section of the Report calls for income to be considered in conjunction with wealth. It argues that ‘wealth is an important indicator of the sustainability of actual consumption’. The problem is that wealth is extremely hard to measure. The Report acknowledges that human capital is hard to measure, and that for many people the skills they have are their single most important asset. This is particularly true for young people, and it is almost impossible to see how this measure of wealth could really be incorporated with any degree of reliability. Furthermore, assets for which there exist markets are likely to be valued at market rates. Here the problem is that we may be in danger of emphasising booms even more prominently. In the 10 years leading up to the most recent recession, for example, shares and particularly house prices rose really rather rapidly in Anglo-Saxon economies. It is easy to dismiss this in retrospect as a bubble (although as British property prices have barely fallen, the evidence for a bubble is easy to contest), but there is little doubt that at the time people felt that these rises in asset prices were securely underpinned and should be seen as legitimate. Were the British and American economies to have followed the recommendations of this Report, therefore, there is a danger that they would have been even more confident five years ago about the sustainability of the economic models that they were then pursuing.
Fifth, the Report urges a greater consideration of the differences between mean and median incomes. There is no doubt that writing an economic history of the postwar world using the median rather than the mean gives a different picture, particularly over the last 30 years. Yet this is surely well understood by all commentators, and above all by politicians who, more than anyone, know that they have to deliver for the average person. Political scientists have median voter models, not mean voter models.
Finally, this section of the Report calls for greater emphasis on including nonmarket activities, including both domestic production and leisure. It presents hypothetical examples to show that moving to a world in which we pay people to do our cleaning, and make our sandwiches, will overstate the rise in national income. Against that, the rise in leisure time in those countries over time means that we are understating the rise in national income. The net direction of bias is not apparent. It notes that these things are very difficult to do, that existing estimates are imprecise, and that ‘only if there is sufficient confidence in extended measures of income will there be a broader take-up by statistical offices’ (p. 38). This is surely correct, and that confidence is unlikely to be forthcoming.
Economic historians are concerned with the quality of life, as much as with material economic progress. This is also an important part of the Commission's Report. The Commission argues that measures such as the human development index should not replace conventional economic indicators, but instead ‘provide an opportunity to enrich policy discussions and to inform people's view of the conditions of the communities where they live’ (p. 41, italics in the original). The Report argues that there are three useful definitions of the quality of life. The first one is subjective: do individuals say that they are satisfied with their life? The second builds on the notion of capabilities, while the third is based on the notion of ‘fair allocations’ (p. 42). This section is characterised by a very sensible discussion of the issues, followed by relatively limited recommendations.
The first recommendation is simply that larger-scale official surveys should be taken of people's self-reported happiness, their hedonic experiences and their priorities. The British government already does this, and in the light of the Report's own acceptance that people in different countries have radically different preferences (p. 48), it is hard to imagine that these sorts of surveys can be standardised internationally in a useful way, in the manner of standardised national accounts that underpin GDP estimates.
The second recommendation is that objective measurement of people's health, education, personal activities, political voice, social connections, environmental conditions and insecurity should be undertaken. Whilst measures of people's health and education are relatively straightforward and already undertaken, it is hard to see how objective measures can be taken of people's personal activities, political voice and social connections. The Report also needs to think through the potential conflict between this recommendation and the previous one. It may well be that our environmental conditions are ‘objectively’ declining, in so far as global warming is a real threat to our standard of living. But it is equally clear that a large number of people think that global warming is a myth, or that in so far as it is not a myth it is a simple natural phenomenon that we should ignore. Subjectively these people derive no disutility from global warming, even if ‘objectively’ they are worse off. Or, were we to find that global warming is a problem that solves itself, or turned out not to exist, then there will have existed people who subjectively felt much worse off, who were scared for the future of our planet, but who ‘objectively’ had nothing to fear. We do not know which of these two scenarios is correct, and nor do we know how we should reconceptualise our understanding of living standards and the quality of life even were we to know which scenario was correct. The balancing of subjective and objective measures in understanding people's wellbeing is very difficult, and the Report of the Commission does not take us very far in understanding this issue.
The Report's third recommendation in the quality of life section is that quality of life indicators should always include inequality of the various measures being looked at. Thus we should not only look at inequality of income, but we should also look at inequality in education, health, social connections, environmental conditions and so on. No evidence is offered that inequality along these dimensions is radically different from income inequality. Furthermore, the Report argues that inequality across these dimensions should be assessed not just across people, as individuals, but also across socio-economic groups, and across generations. It argues, although without a strong philosophical backing, for special attention to inequalities that have arisen more recently, such as those linked to immigration. Why a massively entrenched inequality, which has persisted for generations, should be treated less seriously than one that has recently arisen, it is not clear.
Finally, it calls for statistical offices to release information in such a way that other researchers can aggregate the information in different ways to produce different indices. At one level this request is unobjectionable, but at another it seems to risk undoing the central objective of the Commission. If there are a myriad of different indicators, some of them going up, and some of them going down, some of them showing that Britain is a better place to live in than France, others of them showing the reverse, then it is really unlikely that any of these indicators are going to gain the salience of GDP per head. There are many indicators out there already, and one of the great attractions of GDP per head, for policymakers and civil society, is that it is widely understood and widely respected. It is not perfect, but we know its flaws. In suggesting that we create a myriad of different alternative indices, the Commission risks handing victory back to GDP per head, by default. That is surely not the intention, but it is a risk of this particular recommendation.
Economic historians are busy people. We have research to do, classes to teach, scripts to mark. There are referee's reports to write, books to review, and references on students to produce. Our time is limited, and valuable. The Report by the Commission on the Measurement of Economic Performance and Social Progress is mildly interesting, but offers little by way of value to economic historians. It does not give alternative counterfactuals over time of how different societies have performed according to different indices. It makes no case that, taken as a whole, changes in GDP per head either overestimate or underestimate ‘economic performance and social progress’, or that that measure is more or less accurate for different countries or for different time periods. Instead, it has a wish list of things that individually might well be quite nice, but which collectively do not add up to a research agenda, either for current economists or for economic historians, and which do not collectively add up to any coherent alternative vision of how economic performance and social progress should be measured. Economic historians are likely to have better things to do with their time than to read this Report.
