Taken at face value, human life seems to involve the pursuit an almost limitless variety of plans, projects, goals, pleasures and criteria of any and every sort. People may wish to be pious, healthy, rational or spontaneous. They may hope to help, outdo, gain the respect of or even take retribution on other people. They may aim to run the fastest possible 1500 m, cook the perfect paella, be self-sufficient in vegetables or improve their golf swing, breakdancing or operatic singing.

Yet resources are limited: with finite time, energy, cleverness and money, each individual has to prioritize some projects and downplay or abandon others. The challenge of prioritization arises, moreover, at all scales. Even planning a week of work requires balancing the demands of an inordinate – and continually changing – set of tasks, large and small, and trading off resources that may be devoted to professional and personal objectives. Indeed, even planning to produce a single, conversational remark turns out, when looked at closely, to involve balancing a variety of objectives (brevity, informativeness, truthfulness, relevance, perspicuity and more; Grice, Reference Grice, Cole and Morgan1975).

The same challenge of prioritization arises, in what appears to be an even more formidable form, when we consider public policy. The rules the government sets and the policies it pursues will promote some plans, projects, goals and pleasures and restrict others. A further complication is that the balance of individual goals of the populace has to be weighed alongside a further ominously lengthy list of desiderata for society itself: that it should be efficient, just, wealthy, technologically advanced, culturally vibrant, liberal or theologically sound (according to any extant theologies).

How do we deal with the problem of prioritization? There are two very different approaches to this type of problem.

The first approach is local and bottom-up. According to this approach, we take our vast range of objectives, likings and rules as our starting point and attempt to reconcile the conflicts between them piecemeal, as they arise. This approach is exemplified by the incremental way in which groups of people create rules and norms to govern how they deal with common pool resources (Ostrom, Reference Ostrom1990), the evolution of common law (e.g., Gennaioli & Schliefer, Reference Gennaioli and Shleifer2007), the slow development of institutions of trade and government (e.g., North, Reference North1991), the evolution of cultural norms, customs and conventions of all kinds (Boyd & Richerson, Reference Boyd and Richerson1988; Sugden, Reference Sugden1989) and the decentralized acquisition and application of knowledge in the market economy (Hayek, Reference Hayek1945).

The second approach is global and top-down. Rather than attempting to reconcile the clashes between many different goals one by one, the aim is to identify a single fundamental goal and to prioritize all of the goals purely insofar as they contribute to this goal. The idea is that the apparent plurality of supposedly contradictory objectives is illusory: these objectives are either sub-goals that contribute to achieving the fundamental goal (at least to some degree) or they are spurious and should be rejected. This global approach can be applied on a variety of scales. For example, rather than seeing the various and potentially conflicting interests of stakeholders in a business as being resolved, to some degree at least, by continual negotiation, we may view the firm as having a single overriding goal, perhaps profit maximization (Friedman, Reference Friedman1970) or shareholder value (Lazonick & O'Sullivan, Reference Lazonick and O'Sullivan2000). Rather than viewing an investment firm's actions as emerging from a melee of conflicts, trade-offs and individual incentives within the organization, we might idealize it as maximizing expected return for a given level of risk (Markovitz, Reference Markowitz1952).

Local and global approaches can be applied both to the description of social phenomena and normative questions concerning how they should operate. Indeed, a possible reconciliation, although one I shall reject later, between local and global approaches is that the local approach is most appropriate for considering descriptive questions, but the global approach applies to normative questions.

With this line of thought, for example, the conflicting priorities within a society, a government department, a hospital or a business can be viewed as local constraints distributed across different individuals, roles, professions, teams and groups and applying over a wide range of timescales. The operation of the ‘whole’ emerges through the outcome of the resolutions, deadlocks and continual negotiations concerning these different priorities. A hospital chief executive, for example, may be concerned with balancing a variety of measures of patient care, staff satisfaction and financial stability. Targets and staffing levels may then impact the innumerable trade-offs concerning the management of beds, and this might itself impact multiple questions about patient treatment, the split between in-patients and out-patient operations, demarcations between the work of different professionals and the work–life balance of staff. There will be implications for the split between emergencies and chronic cases, the hours of surgeries, the length and nature of patient appointments (both notionally and in practice) and their interaction with the multiple priorities of patients, carers, wider families and so on. Each of these conflicts is, partially and not always satisfactorily, resolved through a continual process of communication, instruction, occasional dispute and mutual adjustment.

But, one might think, perhaps an organization should ideally be run not by an endless and imperfect process of adjustment and negotiation, but in the service of a single overall objective. For example, suppose, hypothetically, that a company should have just one priority: to maximize profit (Friedman, Reference Friedman1970). If so, the apparently limitless number of priorities throughout a business should be treated as sub-goals of a single larger goal (and, where they do not contribute to the larger goal, priority should be abandoned). Moreover, the aim should be to design the incentives of individual directors, managers and workers with this larger goal (e.g., Eisenhardt, Reference Eisenhardt1989).

What is the single objective that should guide a school, hospital or, indeed, a government? Frijters et al.’s paper ‘A happy choice: wellbeing as the goal of government’ (henceforth WGG) argues that there is a single underlying goal: the maximization of human wellbeing. From this point of view, all other priorities of stakeholders small and large are either relevant to government policy, because they contribute to wellbeing (whether directly or not), or, where they are irrelevant, or even inimical, to the goal of maximizing wellbeing, such priorities should be set aside. From this standpoint, just as a profit-maximizing business may, in theory at least, make individual decisions, set its strategy and design its incentive structures through a process of optimization of profit, shareholder value, return on capital or some similar notion, government should, at least ideally, determine its actions, strategies and structures according to a single, giant cost–benefit analysis, where the currency is not money, but wellbeing.

Now, as we have seen, individuals, businesses, hospitals and governments do not actually work this way. Rather than being driven by a single top-down goal, they are, de facto, driven by the various and conflicting aims and objectives of the individuals, teams, special-interest groups and political parties of which they are composed. The proposal of WGG is normative: that there is a single criterion by which, in principle at least, all conflicting priorities between all parties ultimately could be resolved. Indeed, from this perspective, this approach is a natural, if rather dramatic, extension of the approach adopted in the UK Treasury's Green Book (HM Treasury, 2018) for cost–benefit policy analysis and the evaluation of medical treatments through their cost–effectiveness in terms of quality-adjusted life-years, as used by the UK's National Institute for Health and Care Excellence (Schlander, Reference Schlander2008).

This is an ambitious proposal, and not, of course, one that is easy to realize in practice. First, there is the question of measurement. Wellbeing is inevitably difficult to measure. Moreover, the comparative nature of human judgement (e.g., Stewart et al., Reference Stewart, Chater, Stott and Reimers2003; Vlaev et al., Reference Vlaev, Chater, Stewart and Brown2011; Frank, Reference Frank2012) means that people will assess their wellbeing not in absolute terms, but in relation to their own past experiences and to other people. These factors make it difficult to address even basic questions such as the interpretation of the Easterlin Paradox (Easterlin, Reference Easterlin, David and Reder1974): that the self-reported wellbeing of a country's population does not increase over time with gross domestic product (GDP) per head. Taken at face value, the Easterlin Paradox appears to disconnect material wealth from wellbeing, at least above a certain threshold, but it is also possible that self-reported judgements remain stable because people compare themselves with others, whose lives were also improving in material terms.

Second, there are problems of prediction: the linkage between policy actions and impacts on wellbeing is often complex, indirect and will be difficult to measure directly (Dennett, Reference Dennett1995). So, for example, it may be difficult to predict whether the marginal increase in expected wellbeing is greatest for additional spending in housing, defence or education, or to balance the winners and losers in terms of wellbeing from zero-hours contracts, or allowing Sunday trading.

A third problem concerns accounting for rules: norms, roles, rights, guidelines, laws, social, organizational and governmental life.Footnote ¹ For example, governments appear obligated to follow the rule of law, submit themselves to re-election and deliver, to some approximation, on their manifesto promises. But suppose that, armed with the type of wellbeing analysis that WGG envisages, civil servants and ministers were able to see that violating these rules would enhance wellbeing in certain cases. Should the maximization of wellbeing trump political promises, the rule of law or the will of the people? Indeed, if it turned out that a benevolent dictatorship could be installed and could be relied upon to optimize wellbeing through technocratic means, would this be desirable? These are among the familiar challenges for any consequentialist or utilitarian approach to ethics and political philosophy (e.g., Sen & Williams, Reference Sen and Williams1982).

In this commentary, though, I want to raise a broader objection to the approach that WGG recommends, one that does not rely on difficulties in measurement, prediction or meshing with the rules of conduct. I suggested earlier that we should descriptively understand individuals, organizations and entire societies as complicated networks of local interactions, conflicts, negotiations and agreements through which people attempt to reconcile their conflicting priorities. But, if this is right, what is the justification for taking a single priority, whether it is the maximization of wellbeing or another one, as the ‘master motive’ for government and society to which all other aims, goals, norms and rules should be subjugated? What is the rationale for sweeping aside the manifest plurality of human plans and motives and attempting instead to drive policy by optimizing any single measure?

A first answer might be that people do – or at least should – realize that their innumerable priorities and objectives are really just means to a single ultimate end, and surely this ultimate end must be life satisfaction, wellbeing or some similar notion. Yet, as has often been remarked, this seems to view human motivation backwards. Consider a mountaineer struggling against great odds to conquer a summit (Loewenstein, Reference Loewenstein1999), a monk's devoted life of prayer or a poet hoping to capture a fleeting moment of experience. None of these people conceive of their objectives as aiming to maximize wellbeing. Instead, their various local goals comes first. The mountaineer forms the goal of conquering the summit – and may, potentially, get some satisfaction from the achievement (although see Loewenstein, Reference Loewenstein1999). A devoted life of religious observance may (or may not) have the effect of promoting wellbeing, but this is not its purpose. And surely it is of merely incidental importance whether the poet gains perhaps momentary satisfaction from formulating a beautiful line. Indeed, if we viewed our goals purely as ‘satisfaction promoters’, they would become entirely hollow: their meaning (and tendency to influence satisfaction) would immediately be lost.

Another justification for a single top-down measure to guide policy is that only a single measure will generally provide determinate recommendations. If we have multiple goals, then without some mechanism for trading off these goals, they will conflict and guidance for policy will be ill-defined. But if we do have some mechanism for trading off our goals, this seems tantamount to having a single overarching goal: perhaps a weighted sum – or more complex function – of the individual goals.Footnote ² The weakness with this argument is apparent from observations of how complex real-world decisions are actually made, whether by experts, managers or the public (Phillips, Reference Phillips1984; Dodgson et al., Reference Dodgson, Spackman, Pearman and Phillips2009); for example, concerning the risks and benefits of building a nuclear power plant. The process of forming a decision requires gradually developing criteria for relating different conflicting priorities (regarding costs, jobs, aesthetics, health, any threat of a terrorist attack, contribution to nuclear waste, reduction of carbon emissions), and priorities are typically only made harmonious enough to make the immediate decision of interest, rather than resolving all possible future disputes.

A third suggestion is that the existence of a single underlying objective arises automatically from rational theories of choice in economics and philosophy (Elster, Reference Elster1986). If an individual has preferences that are transitive, complete, etc., then the choices of that individual can be represented by a utility function. Furthermore, if such individuals trade in a market, then a competitive equilibrium will be Pareto efficient, and so, in a specific sense, will optimize welfare (informally, such that there is no reallocation of goods that would be preferred by at least one person and not vetoed by any). But, in conventional economic analysis, there is no concrete ‘good’ that is being optimized; there is instead an attempt to make sense of some notion of maximizing ‘preference satisfaction’, where preferences remain ineliminably various. This preference satisfaction style of explanation, which is at the core of modern welfare economics, is in play when justifying the operation of markets on ‘welfare’ grounds – but the concept of welfare here has no connection to wellbeing or life satisfaction, as described in WGG. In any case, whatever the virtues of this style of explanation regarding individuals (and see, for example, Sen, Reference Sen1977, for innumerable complications), it seems far from clear that even basic rationality assumptions apply to society, as is evident from the well-known paradoxes of social choice (Arrow, Reference Arrow1951). WGG very definitely does not take this as if approach to utility. For WGG, the goal of government is not to maximize some abstract notion of preference satisfaction, but to pursue the concrete goal of maximizing measured wellbeing. Indeed, in contrast to the preference satisfaction approach, WGG allows for the possibility that individuals, and society at large, may often make non-rational choices that do not successfully serve their interests.

In short, then, none of the three possible defences we have outlined appears to be a viable defence of WGG's proposal. But without some such defence, the choice of wellbeing as a gold standard seems to be arbitrary. While some people may value happiness, others prioritize piety, equality, fairness, justice, liberty, education, aesthetic beauty or respect for fundamental human rights (see, e.g., Sen, Reference Sen1985). Indeed, most of us value many of these things in no clear order of importance. And alongside such abstract goals, each of us has a limitless range of plans, objectives and rules to live by, none of which seems any more basic than any other. Indeed, much of human life is orientated around goals that seem inherently arbitrary (sports, games, crosswords, art forms, marathon running, the pursuit of fashion and many more).

From this point of view, the assumption of a specific top-down normative perspective such as that adopted by WGG seems unmotivated. Specifically, WGG simply takes it as immediately apparent that wellbeing is the ultimate goal of life and of government. In this regard, WGG adheres to a form of foundationalism (Hasan & Fumerton, Reference Hasan, Fumerton and Zalta2016): the philosophical view that justification, whether in ethics or political philosophy, mathematics, science or epistemology, must ultimately derive from special self-certifying truths or intuitions (e.g., ethical intuitions, sense data, self-evident axioms or the like).

But there is an alternative perspective: that rather than searching for foundations, which themselves seem unsupported, we should take our variety of plans, choices, intuitions and rules as our starting point, and see the process of reasoning as attempting piecemeal reconciliation between these, whenever they conflict. On this coherentist viewpoint (e.g., Quine & Ullian, Reference Quine and Ullian1978), the local, ad hoc, step-by-step resolution between conflicting priorities that we observe in human relationships, organizations and political processes is not a stumbling towards an ultimate goal (the maximization of wellbeing or some other) that could better be obtained by direct technocratic intervention. Instead, the goal of argument, negotiation, the law and politics is to reconcile conflicting priorities and objectives, step by step, as required. And surely public policy is merely an extension of these wider social and political processes. A well-functioning society is, from this point of view, one that is governed by reason and agreement, where people are allowed collectively to decide their actions, how they should be governed, the mechanisms by which the law is enforced and so on. Societies continually come under criticism from within – and such criticisms are inevitably multifarious. These critiques and debates concern rights for minority groups, concern for the environment, equality, fairness, freedom, the rule of law, the proper oversight of government and many more. They are not directed to a single end: the increase in human wellbeing.

WGG is right to critique GDP as a default objective of government policy, and it is right to highlight the potential value of self-reported measures. Indeed, alongside life satisfaction, we should be interested in people's views of their own health, their sense of liberty, their level of education, trust in fellow citizens and government, fear of crime, job satisfaction and many more. Surveys can be invaluable, and measures of wellbeing no less so. We should see such surveys as reflections of ourselves, which can help us, as citizens (and, by extension, the governments who, in a democratic society, we delegate to make decisions on our behalf), decide how we want our society to develop. Wellbeing, and its measurement, is one among many factors that may (or may not) influence our collective decisions; it is not, though, the one true goal of government.