There are a number of promising lines of research currently underway that aim to shift the focus away from economic indicators such as Gross Domestic Product per capita to broader considerations of human flourishing and related ethical issues. Beyond GDP defends the ‘equivalent income’ approach, which takes individual preferences seriously as developed in the literature on the theory of fair allocation. A significant side benefit of the book is that it provides substantive discussions of four important alternatives to GDP: composite indices, monetary measures, happiness studies and the capability approach. The authors denominate these ‘the four musketeers’, i.e. they are ‘approaches that seek to fight the power of GDP’ (xii). They contend that the construction of a satisfying alternative to GDP does not need to start from scratch. Their characterization and defence of the equivalent-income approach is fuelled by a critical examination of the four musketeers.
The book is authored by two outstanding scholars, Didier Blanchet and Marc Fleurbaey. They were both part of the Stiglitz-Sen-Fitoussi Commission that the French Government tasked in 2005 to propose indicators of social welfare that are alternatives to GDP. In particular, Marc Fleurbaey is an eminent contributor to the theory of the equivalent income approach.
Dissatisfaction with GDP is widespread. Citizens tend to distrust GDP in part because they feel a growing gap between official indicators and their actual experiences. GDP gives a poor account of the quality of life, of environmental problems, and the ability to achieve sustainable growth. Yet one should not expect, neither from this book nor in general, a simple indicator as an uncontroversial alternative to GDP. The problem with GDP is that it captures information on production rather than on welfare. A problem with welfare indicators, however, is that they depend on the relevant conception of welfare that is adopted. The book addresses the ability of the alternative approaches to build indices to take seriously the issue raised by the diversity of welfare theories.
The first among the four musketeers presented in the book stands for the composite indicators, dashboards, and corrected versions of GDP (Ch. 1). A well-known instance and pioneer of the numerous existing composite indicators is the HDI, the Human Development Index developed by the United Nations Development Programme. It is based on the idea that three dimensions are relevant for human development: health, education and material wealth. Health is measured by a longevity indicator; education is captured by mean years of schooling; material wealth is proxied by Gross National Income. HDI is the geometric average of these three normalized indicators. The authors show that the selection of components, the weighting of components, and the type of aggregation (among others) in composite indicators are affected by considerable arbitrariness. Although these choices look technical, they matter normatively. They have a considerable influence on the resulting ranking of countries and they also imply certain trade-offs between components.
Dashboards, meanwhile, provide extensive lists of indicators. For instance, the United Nations Commission on Sustainable Development, since 2007, provides a list of 50 core indicators and 50 others, which capture social, political, demographic, environmental and economic issues. The data are not aggregated at all. We could therefore expect that this takes care of the objection raised against composite indices: since no attempt is made to reduce a wide range of data into one index, arbitrariness should not be an issue. The authors of the book are nevertheless sceptical as the composition of a dashboard is still arbitrary. For instance, inclusion or exclusion of a specific health component in a dashboard will give a different picture overall. The lack of harmonization among the different dashboards in use makes it difficult to compare different states, and the frequent change in their composition makes it difficult to compare one state over time. Moreover, as the information is not reduced to a single indicator, we now face a wide range of information when analysing dashboards, and it is cognitively difficult to attain a transparent overview. Two different persons will not derive the same impression from the same dashboard and there is no way to identify why.
Another ‘musketeer’ is happiness (Ch. 5). Recall the famous Easterlin paradox: while wealth increases over time, subjective evaluations of happiness tend not to. This paradox presents a challenge for GDP-based welfare evaluations and its discovery marks the beginning of the happiness approach to measuring welfare. In this approach, people are asked in questionnaires whether they are happy, how they feel, or to rank social states; for instance, they evaluate their feeling regarding the statement ‘I am very happy’, by entering a number in a predefined scale ranging from 1 (strongly agree) to 6 (strongly disagree). Their answers provide information on their happiness and on which policy is best likely to improve it. The authors strongly criticize this approach. Drawing on the philosophical debate on subjective welfarism, they argue that happiness is the wrong social goal. Among others, they argue that, when well-being is adaptable, there is a risk that the poor may never be helped. They also cast doubt on the use of questionnaires in the happiness approach. It is hard to know which are the relevant dimensions of welfare and hence what the right questions are. But even if this were possible, there is still a calibration problem. When asking individuals to evaluate different options, it has been shown that the phrasing of the questions, the context of the questionnaire, the length or the labels of the evaluation scales influence the ranking of options. In the end, answers to questionnaires are informative as regards a particular framing, but they fail to capture what welfare is without bias. The authors conclude that the happiness approach should be rejected.
The capability approach is the third ‘musketeer’ (Ch. 6). It is based on a comprehensive list of valuable functionings individuals are able to achieve, i.e. their beings and doings, rather than valuing their actual specific achievements. Fleurbaey and Blanchet's main objections to this approach are the following. First, it is a key feature of the capability approach that it focuses not just on achievements, but also on opportunities. The opportunity aspect aims to capture the value of freedom, but it raises a question about responsibility. Taking on board the luck-egalitarian thesis that the distinction between brute luck and option luck matters, the authors object that the capability approach unduly neglects the role of individual responsibility in the evaluation of alternative states of affairs. As these issues are well tackled by the theory of responsibility and equity theory, considering capabilities in such frameworks would improve the analysis. Second, the authors disagree that capabilities should be all that matter. Information about opportunity sets is not more comprehensive than the information about the utilities derived from functioning bundles; these are different and complementary pieces of information and they both matter for welfare. The authors thus argue that an appropriately comprehensive approach to welfare would combine both information about functionings and about the utility derived from these functionings. They also show that, when preferences are diverse, the capability approach imposes a universalist principle that clashes with the principle of respect for personal preferences. Again, there exists a literature on freedom of choice which typically takes these issues seriously, both formally and conceptually. The authors see two lines of development for the capability approach. The first line corresponds to Nussbaum's version. On the basis of a supposedly objective list of valuable functionings, there is a theoretical basis for composite indices such as HDI, although it still faces the drawbacks described above. The second line is distinctive of Sen's defence of the importance of individual values and preferences for welfare. If preferences are to be an ingredient of welfare, the problems described above may be solved by ‘empowering capabilities’ (204) through a full commitment to the theoretical frameworks recently developed in fair allocation theory and the freedom of choice literature.
The fourth ‘musketeer’ stands for monetary measures. Chapters 3 and 4 provide arguments for and against their use for measuring welfare. Fleurbaey and Blanchet argue that neither income nor the market value of total consumption are good proxies for welfare. Such monetary measures do admittedly offer the advantage of providing a reliable trade-off between goods and amenability ‘to convenient mathematical and statistical computations’ (115). But they miss an essential element of welfare, namely, the focus on preferences. The authors on the contrary defend the importance of preferences. Consider a hypothetical situation where an individual can reach a reference social state by paying a certain amount of money; this amount is called her willingness-to-pay. It measures the welfare change between the status quo and the state with project, for parameters describing a reference state – e.g. the project state of the world – on the basis of the individual's preference. This measure of welfare depends on the individual's situation, her preferences, and the choice of the reference point. The advantage of the willingness-to-pay measures is that, unlike composite indices, dashboards, or the capability approach, it permits one to take information on individual preferences seriously. Unfortunately this tool has also been strongly criticized by social choice theorists over the years, who consequently favoured social welfare functions à la Bergson-Samuelson. Among many failures, the standard money-metric approach exposes welfare measures to an ‘arbitrary and capricious ethics’ because the results are dependent on the choice of the reference parameters.
The authors argue that this problem can be remedied by choosing an equivalent-income approach (also in Ch. 4), which they see as a ‘general methodology that can deliver many indices depending on how the reference parameters are chosen’ (245). The equivalent income is the income an individual needs to be indifferent to a reference state. The idea is to measure the distance between the individual's current state and a reference point using her own preferences; this distance in welfare is measured by the individual's willingness-to-pay. As the distance to the reference point is measured in money, it is interpersonally comparable. This feature of the equivalent-income approach allows for policy recommendations that are sensitive to subjective evaluations of individual welfare as in the standard money-metric approach, but it differs from the latter on at least two grounds. Firstly, it provides the missing foundations for the choice of the reference thanks to the advances of the general equivalence approach in fair allocation theory. An axiomatic analysis supports the selection of some reference prices and reference income based on transparent fairness criteria. Secondly, as in the capability approach, the equivalent-income approach is able to consider non-market dimensions of welfare as important. By considering prices for a diversity of dimensions of welfare – e.g. health, leisure, safety, or environmental quality – individual ‘full income’ will cover all these dimensions. Equivalent income hence appears to be a way, the authors argue, ‘to value what has no price’ (115).
A central aim of the book is to overcome Sen's objections to welfarism and resourcism. And while on the whole the approach the authors present is convincing, there is nevertheless an important feature of Sen's idea of justice that is ignored: the centrality of public reasoning. And I believe this point is important to assess the contribution of the book if we value democracy per se.
Firstly, the book is presented as a criticism or a development of Sen's capability approach. But I do not think that Sen is a capability theorist. In Baujard and Gilardone (Reference Baujard and Gilardone2014), we argue that capabilities in Sen's theory do not play a fundamental role in his conception of justice; instead, they serve as a heuristic device. Arguing as the authors do against the capability approach as a conception of justice misses the fundamental role that Sen assigns to the value of public reasoning.
Secondly, the authors show that paternalism is a serious issue for the capability approach, and they trust that respecting individual preferences is the way out. In the Kantian tradition, paternalism is an issue because it violates the principle of equal respect of persons, i.e. it does not treat them as moral agents (e.g. Carter Reference Carter2014: 82–83; Dworkin Reference Dworkin and Zalta2014: 10). Paternalism may be defined through two different perspectives. As many economists including the authors interpret it, paternalism results from a definition of welfare that is independent of individual preferences. For instance, the valuation of functionings is based on an objective account of the good rather than on individual preferences. The authors are right to criticize the capability approach for being paternalist in this sense, while the equivalent income approach they propose is not. According to an alternative definition, however, paternalism can be avoided if agents are involved in the determination of what welfare should be, i.e. if they participate to the decision regarding the way their situation shall be taken into account for social welfare measurements. On this understanding, paternalism only occurs when a theorist takes it upon herself to identify what counts as welfare. This second understanding is better able than the preference-based understanding of paternalism to accommodate equal respect for persons because it not only respects persons as carriers of fixed preferences but also as deliberators about what the content of preferences should be. Sen's idea of justice – at least as long as we do not trap it under the capability tag – copes well with the latter thanks to his insistence on public reasoning. It avoids relying on some external moral authority to define how the definition of welfare should be calibrated in one way or another and discusses the conditions of a genuine involvement of all individuals as being a central issue. Fleurbaey and Blanchet do not consider this alternative understanding of anti-paternalism in their book. While they take seriously the diversity of welfare theories, they do not elaborate on alternative procedures for making decisions about what should count as welfare. I think this is unfortunate.
Thirdly, the authors take for granted that subjective individual preferences are relevant information for social welfare measurements and in deciding public policies. They do not consider individual judgements as an alternative to given subjective preferences. Individual judgements play an important role in deliberative democracy and public reason, as explored in the literature on judgement aggregation (e.g. List and Pettit Reference List and Pettit2002; List Reference List2012). Sen also develops powerful arguments in support of taking individual judgements seriously. Fleurbaey and Blanchet develop a rigorous and convincing argument for how to include individual preferences in social welfare measurement, but one would wish that they had considered the alternatives in greater depth.
There are at least three reasons to read this excellent book. The first reason is that it provides a comprehensive case, considering arguments from both economics and philosophy, for the need to move beyond GDP and beyond the existing alternatives to GDP, i.e. composite index, happiness and capability approaches, and monetary measures. A second reason is that this book offers the only book-length treatment of the equivalent-income approach published so far. Given the different depths at which one may read the book, it is accessible to readers with a very slender technical economic and philosophical background, while also being of serious interest to those with a solid background in microeconomics. A third reason is that this book helps us understand the theoretical underpinnings of proposed alternatives to GDP and renders transparent the normative content of alternative indices. It draws on both economic theories of justice – i.e. welfare economics, social choice theory, the theory of fair allocation, and applied public statistics – and philosophical theories of justice. On that basis, it develops a theoretical framework that identifies the range of admissible alternative indices to GDP. As mentioned, the authors also use it to argue in favour of the equivalent-income approach.
Beyond GDP will fulfil the expectations of academics, policy-makers and anyone else seeking a rigorous discussion of alternatives to GDP and interested in the philosophical issues that arise with the attempt to move beyond GDP.
