Is it time to reboot welfare economics? Overview

The contributions of economists have long included both positive explanations of how economic systems work and normative recommendations for how they could and should work better. In recent decades, economics has taken a strong empirical turn as well as having a greater appreciation of the importance of the complexities of real-world human behaviour, institutions, the strengths and failures of markets, and interlinkages with other systems, including politics, technology, culture, and the environment. This shift has also brought greater relevance and pragmatism to normative economics. While this shift towards evidence and pragmatism has been welcome, it does not in itself answer the core question of what exactly constitutes “better”, and for whom, and how to manage inevitable conflicts and trade-offs in society. These have long been the core concerns of welfare economics. Yet, in the 1980s and 1990s, debates on welfare economics seemed to have become marginalized. The articles in this special symposium of Fiscal Studies engage with the question of how to revive normative questions as a central issue in economic scholarship.

Montesinos-Yufa 2019), as well as having a greater appreciation of the importance of the complexities of real-world human behaviour, institutions, the strengths and failures of markets, and interlinkages with other systems, including politics, technology, culture, and the environment. This shift has also brought greater relevance and pragmatism to normative economics. For example, climate economics is now driven less by abstract debates about carbon prices and discount rates, and more by evidence on what kinds of policies actually lead to effective de-carbonization and the need for complete societal change (Cullenward andVictor 2020, Besley andPersson 2023). Likewise, research on economic inequality has accumulated significant evidence on how differences in policy regimes and institutional arrangements lead to varying outcomes and their normative implications (e.g., Besley andPersson 2011, Nolan 2018).
While this shift towards evidence and pragmatism has been welcome, it does not in itself answer the core question of what exactly constitutes "better", and for whom, and how to manage inevitable conflicts and trade-offs in society. These have long been the core concerns of welfare economics.
Historically, the normative tools of welfare economics co-evolved with the theories and methods of positive economics. The theoretical machinery of utility-maximizing agents and general equilibrium analysis provided tools for making welfare statements, while notions of "efficiency" in the sense of Pareto or the Kaldor-Hicks compensation principle integrated this machinery with a utilitarian moral stance as to what "better" means and implies. This body of work was coherent, tractable, elegant, and enormously influential. It became part of mainstream tradition in public economics, superbly codified in Atkinson and Stiglitz (1980). Its offshoots included widely used tools such as cost-benefit analysis, normative theories about the role of government (e.g., Dreze and Stern 1987), normative theories about the objective of firms (i.e., shareholder value maximization, Friedman 1970), metrics for measuring economic "success" (e.g., GDP), and frameworks for addressing major challenges such as climate change (e.g. Nordhaus 1993, Stern Review 2006. There has been an equally vibrant tradition of critiquing standard welfare economics theories and tools (see Sen (1979) for an older critique and Backhouse et al. (2021) for a recent historical overview). Yet, in the 1980s and 1990s, debates on welfare economics seemed to have become marginalized, prompting Anthony (Tony) Atkinson (2001) to publish his article "The Strange Disappearance of Welfare Economics." While welfare economics never entirely disappeared (the journal Social Choice and Welfare remains lively), it did to some extent go into hibernation, absent from top journals and from the curriculum alike.
This most likely reflected indifference from the mainstream of economists rather than any specific critique of the framework used.
Meanwhile, the policy world often skated over debates concerning normative foundations, casting economic welfare analyses as technocratic and value-free (Fabian and Breunig 2018). The discipline has traditionally drawn its strength from being able to model behavioural phenomena formally and thereby make difficult policy choices tractable. A key tool emerging from welfare economics has been cost-benefit analysis (CBA), widely applied by finance ministries and spending departments as a budgeting and appraisal tool (e.g., the US OMB/OIRA Circular A4 and the UK Treasury Green Book). The ability to put a monetary value on a major decision with seeming rigour has proven attractive to policymakers as a demonstration of 'value for money' and as a means of justifying the choices made -even if that has always ultimately been a value-laden, political choice (Harcourt 2018).

Why does welfare economics need revisiting now?
Underlying tensions and questions about the moral, behavioural, theoretical, and empirical foundations of welfare economics have never gone away. Yet both the efficiency 'turn' described above and developments during the past decades-notably the increasing urgency of the climate crisis, debates over economic inequality, experiences during the pandemic, issues of race and gender, the problems of "left behind" regions, issues of corporate power, and the role of technology in society-have made questions of welfare economics even more salient and consequential than ever. Buchanan (1964) defined economics as the study of exchange relations. He distinguished it from the study of power relations (politics) and moral relations (sociology and anthropology). Contemporary, complex policy challenges generally transcend exchange relations and so the traditional tools of economic analysis, including standard welfare analysis, will in any case miss crucial pieces of the policy puzzle. There is wide awareness of, for example, the fact that market prices are of limited use for anything not traded in markets, that markets are manipulated by advertising and may be overseen by captured regulators, that the modern economy involves products over which people cannot easily form preferences, or that sellers leverage aspects of psychology and culture that economics has traditionally assumed out of its models for reasons of tractability. While perhaps reasonable in the past, such 'non-economic' considerations can no longer be ignored. Two decades into the 21 st century, in the face of substantial current challenges, it is not clear that claims to objectivity (or value-neutrality) for economic analyses will be credible. For these societal challenges reveal that normative choices are intrinsic to policy decisions.
Many economists seem reluctant to acknowledge the normative aspects, in policy areas ranging from the distributional aspects of quantitative easing to the environmental sustainability of investment projects, to the likely detriment of the economics profession's reputation. This is not to say, of course, that economists think distributional issues or climate justice are unimportant; on the contrary. But -although there has been an ebb and flow within the profession -since the 1980s the notion of 'efficiency' has dominated applied and policy economics. Advocated by Chicago economists such as Harberger (e.g., 1954Harberger (e.g., , 1971 and Stigler (1981), this lens implies that policies are desirable as long as monetary gains to losers outweigh losses to losers; and moreover it has become common currency that the efficiency calculus is objective or scientific. This pervasive view is visible in the analogies often made between economics and 'practical' professions from dentistry to plumbing; nobody needs take a normative view about a leaky tap.
The most significant current crisis is climate change and unsustainable exploitation of the environment more broadly (Dasgupta 2021). Economists have long provided many valuable insights into these issues. For example, the Nobel Prize was awarded to Eleanor Ostrom in 2009 for her work on common pool resource management (Ostrom 1990). There are fundamental disagreements within the discipline of economics about how to appraise the climate crisis normatively. Another Nobel laureate, William Nordhaus (2015), utilizing approaches from welfare theory, claimed that 4 degrees of global warming would be economically optimal. Some see this as a manifestation of insularity of economics (Marchionatti and Cedrini 2016), for a focus on efficient pricing as the solution to most environmental challenges betrays naivety regarding human psychology (Kienzler 2018) and human politics (Beinhocker and Farmer 2021). However, that Nordhaus claim has been extensively critiqued not just by climate scientists, who stress the tremendous, even existential, environmental and human toll this would take (Burke et al. 2015), but also within the economics profession itself (for example Stern et al. 2022).
Related controversies around the discount rate used in economic modelling of long-term environmental degradation illustrate the ethical naivety of some economic perspectives (Cole 2008). What discount rate to use is a normative question. The death toll, misery, and intergenerational burden associated with climate change raises deontological issues, so the discount rate question cannot be answered with only the narrowly utilitarian framework upon which a certain type of economic modelling relies. In fact, economic approaches that frame climate as a technocratic cost-benefit problem, to be addressed by finding an appropriate discount rate or shadow prices reflecting the externalities, will fail to rise to the climate challenge. Alternative approaches, framing the issue as one of directed economic, technological, and social transformation in a social and political context (as Ostrom does), and with explicitly and socially determined normative objectives, will be both more congruent with the nature of the problem and more useful for normative policy advice (Beinhocker 2023). Economics is now beginning to explore approaches to address these issues. Instead of working within a narrow framework where behavioural change predominantly through shifting price incentives, there is now greater attention paid to the dynamics of values alongside political change (Besley and Persson 2023). Moreover, this also requires looking at interactions between values, policy, technology adoption and directed technological change (Aghion et al. 2023).
The task of reviving a focus on welfare economics is all the more urgent because the economic approach is rapidly being embedded in artificial intelligence (AI) and machine learning (ML) systems. Maximizing an objective function (or minimizing a regret function), specified in the same way as an economic model, these are being deployed in areas of policy such as criminal justice, firm hiring, or the calculation of welfare benefits -despite the fact that they involve profoundly political and normative judgements (Coyle and Weller 2020).
The task also extends to the teaching of economics, educating future generations of policymakers. This needs to be woven into the core principles of economics taught to students (as Erik Angner discusses in this issue). Current curricula often leave students with the impression that normative questions should be left to philosophers or political scientists. But we need policymakers who can integrate positive and normative economic analysis to address economic challenges. We must provide the intellectual tools they need.
Fortunately, there is no shortage of innovations and insights to draw on in creating a rebooted welfare economics, including empirical work on happiness and broader conceptions of human well-being, the capabilities approach, incentive compatible market design, behavioural welfare economics, and institutional economics. However, there is way to go in integrating such approaches into the kind of mainstream research that appears in leading journals.
One fundamental issue is that most economics begins with the assumption that people have preferences and the problem is to elicit them. But there is a real challenge for people to know their own preferences and much of the policy process is about preference formation not just preference aggregation which is the classical domain of social choice theory.
A rebooted welfare economics also needs a wider take on distributional issues beyond material resources to include differences in status, cognitive capacities and power. There is also greater scope to develop approaches that consider a complex range of range of motivations beyond self-interest. In addition, it is crucial to work with a theory of value that enables the use of metrics beyond what can be learned from market prices and incomes.
Moreover, a rebooted welfare economics is likely to require a wider appreciation of where motivations, values and preferences come from. This rich agenda will require a continuing trend towards integrating approaches in economics with those from disciplines such as political science, psychology, anthropology and sociology.
The papers in this symposium seek to address these issues and outline ways such broader insights might be incorporated into the heart of welfare analysis as well as the implications of a return to welfare analysis for pedagogy. The remainder of this introduction provides an overview of the issues the papers address.

Measurement
One area in which the normative shortcomings of traditional economics are readily apparent is the way we measure progress. Businesses are freely depleting or damaging natural resources, the financial sector enriches the top 1%, the food system is contributing to obesity and promoting antibiotic resistance, pharmaceutical firms rely on people being unwell for the pursuit of profit, and new AI technologies create value for a few technology firms mostly by learning from existing creators without compensation. It is, not surprisingly, widely perceived by citizens that the economic model encourages extraction and exploitation. Yet conventional economic statistics say that society is doing better than ever.
The measurement focus on unidimensional metrics of 'the domain of socially organised production' (Vanoli 2005) calculated using exchange values or market prices is under sustained challenge from a demand to go 'Beyond GDP'.
Although they continue to allocate much of their effort to traditional national accounts data, official statisticians are cognisant of the demand for multidimensional indicators including of non-market production (such as care) and a move away from using only exchange values for the valuation of activities or assets. Similarly, CBA in practice typically remains reliant on conventional income and price data that struggles to capture nonmarket activities or the wedge between market prices and welfare-reflective accounting or shadow prices. But recent reviews of the UK Treasury's Green Book (HM Treasury 2021) guide to CBA have included consideration of 'wider' benefits such as environmental externalities, and measurement of subjective well-being (Frijters and Krekel 2021, OECD 2013, Helliwell et al. 2022. Official statistics and guidance are edging toward more social welfare-reflective measurement, albeit with little advance in the underlying economic scholarship. Paradoxes in the standard Hicks-Kaldor treatment of subjective utility as the source of value were noted as long ago as Scitovsky (1941) and Viner (1937). Yet with a few exceptions -such as Lancaster's 'new approach to consumer theory' (1966), linking utility not to market prices and quantities but to the underlying characteristics of goods and services, extended recently by Hulten and Nakamura to digital goods (2017) -the theory of value per se is little discussed in mainstream economic measurement. Such measurement innovations are being driven more by demand from statisticians, responding to the 'Beyond GDP' imperative in policy, than by supply from economic theorists.
A particular gap is the need to develop theory and methods for accounting for shadow prices of non-market goods. Willingness to pay methodologies in the pricing of environmental goods like biodiversity and national parks have been notoriously ineffective because a) people cannot easily form preferences over whole ecological systems, b) strong preferences require repeated experience to emerge, which is rare in the case of biodiversity loss and climate disasters, c) people cannot easily retract choices that turn out to be deleterious to their utility in the context of environmental damage that compounds over decades, and d) humans struggle to think about willingness to pay for complex ecological systems, such as micro-organisms and soil health, as opposed to individual elements of them that are easily experienced, such as large mammals. Many economists dislike stated preference methods for these and other reasons -such as lack of incentive compatibility and strategic biases (Zawojska and Czajkowski 2017) -yet have not provided an alternative for the many cases of non-market goods where revealed preference methods cannot be applied (Blinder 1991).
An alternative response to the need for innovation in price-based ways of measuring welfare is to look for alternate measurement strategies and even different conceptualisations of welfare, discussed in this issue by Cooper, Fabian and Krekel.
Behavioural economics is developing tools for measuring welfare through massively multidimensional indexes of stated preferences rather than prices and willingness to pay.
The capabilities approach in development economics advocates for a broader conceptualisation of the budget constraint, moving beyond income to also consider items like enfranchisement, mobility, education, health, and the built environment (Alkire 2016).
Capabilities indexes aggregating these items are now widely used in development policy.
These indexes are increasingly developed in partnership with the communities affected by those policies so that they reflect local preferences (Sollis et al. 2022).
Moving beyond preference satisfaction accounts of welfare, happiness economists have made substantial inroads in adapting life satisfaction scale data (a measure of a mental state) for use in cost-benefit analysis (Frijters and Krekel 2021). The new WELLBY (wellbeing adjusted life years) approach builds on earlier learnings from QALYs (quality adjusted life years) and DALYs (disability-adjusted life years) to evaluate spending in terms of its effect on life satisfaction. This approach is controversial within psychology, where life satisfaction and associated scales are a controversial way of conceptualising and operationalising wellbeing.
There were substantial debates in the late 2000s between advocates of 'hedonic' understandings of wellbeing (Kahneman et al. 1999, Diener et al. 2009) as a combination of affective experiences (e.g. happiness, boredom, loneliness) and life evaluations, and advocates of 'eudaimonic' understandings that stress particular ways of living that are congruent with the nature of the human organism (Ryan et al. 2008, Waterman 2008).
These two schools now seem to be integrating, but differences of opinion remain substantial (Martela andSheldon 2019, Fabian 2022). Unfortunately, the multidimensional indexes of psychological wellbeing (see e.g. Marsh et al. 2020) that are often advocated for by critics of life satisfaction are difficult to integrate into cost-benefit analysis.
Perhaps we should not be quick to prioritise the mechanical needs of cost-benefit analysis over concepts and methods that result in an analysis of what actually matters to people. The desire to trade away realism for tractability in service to cost-benefit analyse is one of the most common critiques of 'neoliberal' public administration (Muller 2019). Practitioners and service delivery personnel like teachers, nurses, and social workers complain that the reality of public policy is more complex, fluid, contingent, and human than cost-benefit analysis can typically account for. The application of cost-benefit analysis in these cases, especially by treasury officials who pay for policies but do not implement them, can result in clumsy, wasteful choices (Bason and Austin 2022). Part of rebooting normative economics could be an honest assessment of the limits of cost-benefit analysis in practice.

Behaviour
Behavioural economics has had a substantial impact on economics and policy analysis and birthed a new field of "behavioural welfare economics" (Bernheim 2009). Policymakers have embraced the 'nudge' concept (Thaler and Sunstein 2009), despite the debates in the literature about the implied paternalism (e.g., Saint-Paul 2011) or the inherent assumption that the analyst can identify the optimum even if the individual decision-makers cannot (Sugden 2018a). There is an active agenda in behavioural welfare economics, including Neo-Hayekian social contract perspectives (Sunstein 2023), endogenizing preferences (Fabian and Dold 2022), the social and environmental construction of preferences (Bowles 1998), and more deeply integrating psychology (Rabin 2013).
Behavioural economics in the 'nudge' vein maintains the welfare as preferences stance of traditional welfare economics, including the assumption that people know their preferences (Bernheim & Rangel 2009). Yet it is precisely careful research in behavioural economics that challenges the feasibility of using preference satisfaction as a welfare criterion amid cognitive and behavioural biases and endogenous preferences (Sunstein 2018). Taking a wider view of psychological insights could allow economics to utilise a more holistic and realistic understanding of (a) preference formation and (b) well-being (Fabian and Dold 2022). While 'rational' preferences are a compelling normative standard in some cases, such as retirement savings, rationality is an inappropriate benchmark for many choices. Rational dieting, for example, is very different for an aspiring sumo wrestler or ballet dancer. Literatures in psychology on goal setting, self-actualisation, emotions, motivation, and multiple selves, among others, can shed light on what preferences are tied to organismic wellbeing.
The field of behavioural welfare economics is also increasing returning to ideas in social contract theory and moving towards notions of participatory governance in the realm of economic policy (Gofen et al. 2021). This is a fertile area of scholarship yet to be fully embedded in mainstream economics, and particularly policy analysis, where the simplistic nudge approach remains prevalent. Nudges are prone to technocratic conceit, where detached analysts in central agencies believe local citizens too 'biased' to organise their own affairs. Incorporating participatory mechanisms into the nudge agenda offers citizens their opportunity to assent to being nudged. It also aligns neatly with the emerging 'boost' paradigm in behavioural psychology, which seeks to educate and empower citizens with psychological insights rather than merely steer their behaviour (Fabian and Pykett 2022).
Malte Dold provides a review of these new directions in behavioural welfare economics in his article in this issue, reflecting on the challenges they pose to traditional welfare economics and the opportunities presented by behavioural public policy beyond nudging.

Inequality and power
We noted earlier the distributional issues inherent in a full debate about the economics of climate change. Climate change mitigation efforts occur within a global political and institutional context that sees less developed nations and their impoverished citizens marginalised. This imbalance of power means that the traditional economic way of modelling politics, namely social choice as the aggregation of individual preferences, leads to naïve policy analysis. Justice and power are not accounted for in climate models, nor can they be with this approach to modelling.
Similarly, economics' reluctance to address power and politics also limits how the discipline deals with other complex policy challenges. Take the example of discrimination. The #MeToo and Black Lives Matter social movements have thrown into relief qualitative forms of discrimination experienced by women and minorities in society at large, not just in labour markets. The economic notion of statistical discrimination, while frequently helpful for exposing ineffective anti-discrimination measures like ban the box 5 regulation (Doleac and Hansen 2020), has also prevented economists from appreciating the full complexity of discrimination. Notably, statistical discrimination research, like economics in general, typically applies methodological individualism in its approach, which obscures the role of structural factors like systemic racism and patriarchy in driving behaviour. Some economists (e.g. Loury 2002) have long emphasised the centrality of social influences on economic choices and the consequent importance of group inequality; and these issues are increasingly being acknowledged and analysed in labour economics and some other fields (Francis et al. 2022). But their wider implications for normative economics and social/public theory have remained largely unexplored.
Deindustrialisation, spatial inequalities, and the associated rise in populism and identitarian politics across OECD nations points to another shortcoming of economic theory with regard to the normative implications, namely the absence of the 'social' in economic modelling.
Identity, community, and cooperation are fundamental to our species and its flourishing, most obviously in the role trust and social capital play in providing efficiency and insurance in the absence of complete contracts and perfect information. While economics has made seminal contributions to this literature, especially in evolutionary game theory (Gintis 2016) and institutional approaches (Bowles and Gintis 2002), these themes have remained largely at the margins of the discipline and are very rarely taught in undergraduate courses.
Methodological individualism also prevents these themes, which are all emergent properties of complex social interactions like culture, from entering the economic models typically used in policy analysis (Beinhocker 2020). Ignoring aspects such as identity or complex, non-linear outcomes in favour of profound methodological individualism -and the submerged value judgments it implies -led to an aversion in economics to place-based policies for many decades. The associated misery, lack of hope, and deaths of despair amidst the opioid crisis in 'left behind places' are a tragedy (Case andDeaton 2020, Graham 2023) and the acrimonious political sentiment provoked in affected communities is a major threat to democracy.
The anti-technocratic sentiment that is common to contemporary populist movements further underscores the political awkwardness of economics. The morality of economic agents is not exogenous but a function of how markets are structured (Besley 2019, Carugati andLevi 2021). The deliberative democracy tradition has demonstrated that citizens' values are neither entirely rational nor fixed in the manner assumed by social choice theory.
Instead, citizens both form and maintain their values intersubjectively through political and cultural dialogue (Dryzek and List 2003). This discursive foundation of healthy liberaldemocratic politics and the just institutional management of power is substantially missing from economics, with notable exceptions (e.g. Tirole 1997, Acemoglu andRobinson 2006). This sometimes allows economic analysts to operate with the assumption that their models reflect prevailing preferences effectively, and do not require value judgement.
Economics has long provided sophisticated counterarguments to heavy handed state intervention into people's lives, but it has not sufficiently guarded against technocratic hubris on the part of economists themselves (Sugden 2018b). It has certainly not engaged with the role of civic life in fostering normative deliberation over preferences and forming political communities. As Abba Lerner (1972, p. 259) remarked: "Economics has gained the title of queen of the social sciences by taking solved political problems as its domain." The things that economics has traditionally assumed out of its models in order to achieve tractability are now becoming major threats to the discipline's relevance to contemporary social scientific issues.

Pedagogy
If asked how policy outcomes should be judged, most economics graduates would probably suggest Pareto efficiency, or cost-benefit analysis. They would know how to analyse the effects of a tax or an externality using consumer and producer surplus. Perhaps they would be familiar with a utilitarian social welfare function, hedged with caveats about interpersonal comparability. If they had followed a public economics course they might have studied optimal taxation, encountering Mirrlees' (1971) famous example of an artisan economy with a distribution of abilities, used to explain the problem of incentive compatibility.
But they would have thought little about the associated value judgments and distributional implications in any of these cases. The Mirrlees example illustrates the problem starkly: in the first-best utilitarian optimum with lump-sum taxation, high ability workers are in effect enslaved to feed the others. Yet students are unlikely to have been asked to question whether we really think such an outcome would be "first-best".
As Erik Angner argues in his paper, our students need better guidance than this if they are to make practical judgements, as citizens, policy makers, employers and employees, about the economic challenges facing society. Those teaching economics to decision-makers of the future should provide a framework that encompasses both normative and positive aspects of economic decisions. If we focus only on those where we feel comfortable, we convey the message that other considerations do not matter to us.

Conclusion
The COVID-19 pandemic underlined the salience all of the normative issues outlined above.
Death rates differing by age, class, and ethnicity raised complex deontological questions.
Vaccine rollouts highlighted global power imbalances. Social capital and fellow feeling was critical to combating loneliness, helping the vulnerable, and ensuring compliance with social distancing advice. And yet economics seemed to partake in the policy discourse mostly through its traditional, narrow lens of cost-benefit trade offs denominated in dollar terms, most glaringly in debates over the statistical value of a life. The discipline did not take this opportunity to reflect on the normative assumptions underpinning its models but instead chose to maintain its image as providing dispassionate technical advice to remain close to power. In consequence, economics (and epidemiology) experienced another round of scepticism and critique from the relatively marginalised corners of society.
The articles in this special symposium of Fiscal Studies engage with the question of how to revive normative questions as a central issue in economic scholarship. They draw on recent advances in theory, new insights from empirical studies, and an innovations in economic measurement to propose ways for economics to become more normatively sophisticated while retaining its traditional strengths of tractability, quantification, prediction, and policy relevance.