Discovering ‘the New Science of Happiness’ and Subjective Well-being

Abstract

The ‘new science of happiness’ was not really discovered, but was a coming together of people, publications, projects, politicians, agencies and disciplines around the turn of the twenty-first century. This moment foregrounded the issue of how people feel (subjective well-being), changing how this is understood and measured, driving the ‘second wave’ of well-being. This chapter reviews these interlinked histories to contextualise the ‘new’ well-being data. It presents definitions, theories and methods to help understand what went on behind the scenes and under the bonnet of these data practices. We look at the establishment of the UK’s subjective well-being measures and address the question of what subjective well-being can do that differs from previous well-being measures.

‘The rise of well-being’ in politics and policy-making emerges from developments across intellectual fields, including psychology, social policy, economics and social statistics (Bache and Reardon 2013, 908). In Chap. 2, we also discovered that happiness and well-being are linked, but different, and hard to define, while Chap. 3 offered a brief overview of how well-being data can be collected and analysed. We also discovered that questionnaires can be used in one-to-one interviews and national-level surveys, collecting qualitative and quantitative data.

Subjective well-being data are largely generated using questionnaires. These could be a paper form you may be asked to fill in before entering a weekly therapy session. These data would be looked at in isolation from data on others, are private and confidential, and will be used to track one person over time. Similar questions are increasingly used in national-level surveys, which can generate large-scale datasets to inform national indices. These won’t be traceable back to the individual when analysed and are used to understand how populations and sub-groups are feeling, inviting comparisons between groups of people over time. The latter kind of subjective well-being data are then used to inform important decisions in policy development, monitoring and evaluation, and to promote behaviour change in populations. We are going to look at how these data gained popularity and standing in this chapter by looking at the rise of happiness economics and its impact on well-being data.

Chapter 2 explained that the discipline of economics also has trends over time and sub-disciplines. While people think of economics as primarily financial, it has far broader concerns and also tries to understand the value of things to people. For example, where the nineteenth-century hedonimeter project hoped to measure how people feel about things in a way that was ‘more scientific’, some economists have subsequently tended to focus on understanding what people do in the belief that this indicates what they value, and how they feel, whether subconsciously or consciously.

It is here that happiness plays a role for economics: to understand what makes people happy (in broad terms) at scale. Connectedly, to understand how to best go about measuring and modelling to establish this, and evaluate policy decisions of the past, in order to make better ones in future. This idea is based on the Greatest Happiness principle (Bentham 1996 [1789]), which you may recall from Chap. 2, and is elaborated here. We also spend more time thinking through what is meant by subjective well-being, and how it is defined in relation to happiness, before exploring categories of subjective well-being measures that are used, and what they do, or at least what they claim to. A key thing to keep in mind is that happiness economics measures more than happiness, using the broader (and more complicated) concept of subjective well-being.

We are going to look at the rise of happiness economics, for two main reasons: (1) it is acknowledged as one of the key drivers of the second wave of well-being, and (2) it positioned itself as a new science of happiness, advocating new measures, different data and analyses. This chapter, therefore, looks at how developments in psychology and economics come together to intervene in social statistics and social policy. The introduction argued that well-being data are used to (1) track the health and wealth of society using social statistics and (2) evaluate the success and progress of social projects and policies. Therefore, how all these interventions come together are key to understanding how well-being data work.

4.1 Happiness Economics

People down the ages have agreed that money can’t buy happiness, though this exact form appeared only in the nineteenth century. (Cresswell 2010, 278)

Lord Richard Layard was called the UK’s ‘Happiness Tsar’¹ and his seminal book Happiness: Lessons from a New Science (Layard 2006) consolidates aspects of what Bache and Reardon call the second wave of well-being (Bache and Reardon 2013). The book presents the rationale behind ‘happiness economics’, which this chapter covers as the Greatest Happiness principle, combined with aspects of positive psychology, together with established well-being indicators and newer subjective well-being measures that we will come back to in greater depth throughout this Chapter.

The book Happiness is a call to action to do things differently, in a similar way to the politicians’ statements and reports from international agencies we have already encountered. We are going to begin by looking at Layard’s presentation of knowledge and understanding of well-being and data, as an example from the field of happiness economics. The book opens with the idea that money cannot buy happiness, explaining that this is ‘no old wives’ tale’, but proven by ‘many pieces of scientific research’ (Layard 2006, 3).

The book opens with ‘the Easterlin Paradox’ (Layard 2006, 3) that we encountered in Chap. 2. In short, through looking at subjective well-being data, together with data on income, Easterlin found that while people with higher incomes tend to be happier than those with lower incomes, increased average income has not increased average happiness (Easterlin 1973, 1974). On this basis, Easterlin states that economic growth does not lead to an increase in happiness, at least in countries that are already relatively wealthy (Easterlin 2001). ‘The Easterlin Paradox’ remains a recurrent topic in discussions of well-being data and measurement, even though it has been challenged several times (most notably Stevenson and Wolfers 2008, 2012). Easterlin has nevertheless come out to defend the idea when it has been challenged (i.e. Easterlin et al. 2010) and much work continues to build on this thesis. For example, testing whether it is generalisable (i.e. Grimes and Reinhardt 2019), that is whether the theory works when tested in various ways across countries, contexts and wealth bands. The paradox therefore remains a compelling idea for economists.

The Easterlin paradox is a popular framing narrative to introduce the importance of well-being data and knowledge, especially when it comes to understanding society and policy. If he is not the opening gambit, he’s near the top of the bill (e.g. see Adler 2013, 9; Alexandrova 2017, 4; Allin 2007, 47; Bache and Reardon 2013, 902; Benjamin et al. 2012, 18; Blanchflower 2008, 32). Or, more specifically, his findings published in 1973 are presented as the turning point in understanding the relationship between social progress and societal well-being. For if economic growth does little to improve social welfare, should it be a primary goal of government policy? Layard explains why not, as well as how economics can help us understand how not, with the help of philosophy.

The position taken is that ‘much of the social progress that has occurred in the last two hundred years’ has been driven by ‘the Greatest Happiness principle’ (Layard 2006, 5). The point of Jeremy Bentham’s ‘noble idea of utilitarianism’ for Layard is that:

it is fundamentally egalitarian, because everyone’s happiness is to count equally. It is also fundamentally humane, because it says that what matters ultimately is what people feel.

The best society, therefore, is one where citizens are happiest and therefore the best policy produces the greatest happiness and the most moral action produces the most happiness for those affected (Layard 2006, 5). This is in tension with ideas of happiness maximisation, which is that people will, or should, have the right to pursue or consume or do whatever makes them happy, and they will always want more happiness. We touched on issues associated with individualism in Chap. 2, as fundamental ones of ideology and social justice.

Layard introduces eighteenth-century enlightenment philosopher Jeremy Bentham as a ‘shy kindly man’ who was a great thinker. He argues that Bentham’s ideas had been difficult to apply in practice because ‘so little was known about the nature and causes of happiness’, which ‘left it vulnerable to philosophies that questioned it’ (Layard 2006, 5). The implication being, of course, that this has been resolved because ‘the new science’ means we now have this information. Indeed, the front cover of Layard’s book proudly states using red block capitals in a golden sun-like graphic shape: ‘INSIDE: THE SEVEN CAUSES OF HAPPINESS’.

It is actually rather brave for an academic to announce they know the causes of happiness; doing so asserts a degree of certainty that is infamously evasive. In fact, the influential Sarkozy commission report that also surveys the evidence, in particular from economics, notes:

A general difficulty for the study of the determinants of subjective well-being is to distinguish between causes and correlates. (Stiglitz et al. 2009, 150)

Ironically, it is difficult to identify the ‘seven causes’ in the book, as they are not explicitly presented inside. Instead, on page 62, in a sub-section of a sub-section called Adult Life, and approximately half way through the chapter called ‘So What Does Make Us Happy?’ is a box, much like the ones in this book. Layard’s is called ‘The Big Seven factors affecting happiness’. The box lists family relationships; financial situation; work; community and friends; health; personal freedom; and personal values. It states the first five are in order of importance. Interestingly, they are able to be ordered by a sense of importance using data from the US General Social Survey. Freedom and values are added as ‘two other key factors’ and in a footnote, Layard explains: ‘these last two factors cannot be ranked, but their relevance is shown in the table’ (Layard 2006, 63; 255). It’s not explicit why they cannot be ranked; it is also, therefore, not made clear why they were included to make seven, rather than five.

As you will see in the second half of this book, unequivocal claims that one thing ‘causes’ happiness, or improves well-being, rather than more modest claims, such as ‘contributes to’, ‘is related to’ or ‘affects’ are extremely difficult to substantiate. As with the Easterlin Paradox, which states that increased wealth does not [necessarily] cause increased happiness, it is difficult to claim something is a universal truth. Studies looking at similar relationships with similar data have not resulted in causal claims, and other evidence and theories that are drawn on are contested. Yet, as Chaps. 7 and 8 of this book demonstrate, we often find that useful insights with well-being data become repackaged to make causal claims, which when we look ‘under the bonnet’ should be a bit less assertive or emphatic.

Layard’s ‘big seven’ may read a bit like an ‘objective list’² of what is important to well-being, similar to those OECD and UK Office for National Statistics (ONS ) examples from the last chapter. Coincidentally, Layard informed both of these organisations as an expert on the advisory panels. You will see that just as the categories differ slightly between the ONS and OECD lists, Layard’s own list of categories as to what causes happiness differs slightly, again.

You may note that Layard’s list does not explicitly include personal security and safety, as the domains discussed in the OECD example from Chap. 3. When you think about the concepts of safety and security, for you they may also sit in relationships, financial security or community. To re-cap briefly on Chap. 3, there are no perfect objective lists of the components of well-being, which tend to involve a subjective carving up of societal and personal concerns. In terms of data, objective lists of indicators tend to rely on ‘proxies’, by which we mean proxy measures where the thing we want to understand, say ‘personal safety’ or ‘personal security’ is measured by data that is seen to stand in for it, in some way, but is not exactly the same thing. The OECD example in Chap. 3 demonstrates how personal safety and/or security is difficult to measure directly and so ‘self-reported victimisation’ is used instead as a proxy. These are administrative data from crimes reported in individual countries (which of course is not the same as actual crime, risk or safety). The OECD replaced the proxy metric with ‘feeling safe walking home at night’. Thus, an objective indicator has been replaced with subjective data, as it has come from surveying how someone feels, rather than the administrative data from reporting crimes. However, we can feel safe and secure because of different domains in our life, and we can feel unsafe and insecure across numerous domains as well.

Crucial to the story of data is the moment when well-being is acknowledged to be more than a list of objective indicators, such as crime rate per nation or in a local area. Instead, well-being is understood as how risk of crime is experienced. Even more crucial to this chapter is the delineation between subjective data about well-being and subjective well-being data. Somewhat confusingly, how people feel about crime is subjective data about an objective well-being indicator. Subjective well-being indicators are different again. They are about how we understand our own well-being and how we feel.

Replacing some proxies with subjective data about how people feel about an objective indicator, such as crime, still leaves many questions about personal well-being. To answer questions about personal well-being, we need more rigorous subjective well-being measures that tell us how people feel over time. This was the gap ‘the new science’ aimed to fill and the driving force of the new well-being indices.³ This chapter goes on to unpack the development of subjective well-being measures: how they were decided on; what the different measures capture and what they do not, and so on. It looks under the bonnet of ‘the science’, its: history, theory, politics, data and its methods. First of all, we will return to the Greatest Happiness principle.

4.1.1 The Greatest Happiness? And Other Principles

It is said that Jeremy Bentham himself was not convinced that his political project would work, or indeed, could be proven, and he corrected the Greatest Happiness principle later in his life from ‘the greatest happiness of the greatest number’ to ‘the greatest total sum of happiness’.⁴ Let us briefly consider the limitations to the Greatest Happiness principle. There are pragmatic objections, which we shall deal with first. The principle assumes that happiness can be affected by what we do and what others do; therefore, happiness is a consequence of our own choices and behaviours, as well as those of others.

To apply the Greatest Happiness principle in policy, then, we need to be able to predict how different behaviours and actions affect happiness, so decisions can be made. In turn, this means we need to know what happiness is, and that behaviours, actions and happiness must be measurable. As we already know, agreeing on what either happiness or well-being is has long proved difficult for philosophers and more recently for measurers. We will also discover in Chaps. 6, 7 and 8, that measuring what we do at a large scale is also challenging. This makes it hard to be sure that one action (whether on a personal or policy level) has positively impacted on happiness, or if an alternative would have done better.

Of course, it is here that the new science is presented to best intervene. As Layard indicates, it generates data and the means to analyse them in order to address the pragmatic objections to the happiness principle. Yet, not all believe that happiness can actually really be influenced by targeted actions or changing an individual’s behaviour. In contemporary society we see judgements regarding other’s behaviours being demonised as bad for well-being (as discussed in Chap. 2), and in ‘COVID-19 world’, the endless recommendations that people go for a walk or a run have little consideration as to whether that is available to them (Ryan 2021). So, targeted actions are not universal.

Some argue that it is easier to improve those with better well-being first (Oakley et al. 2013, 23). Relatedly, ‘the utility monster’ was a thought experiment in ethics first developed in the 1970s. It presents a challenge to the Greatest Happiness principle, and to Utilitarianism, more generally. It asks what if a monster could accrue greater happiness from any given resource than anyone else? For example, imagine if being able to attend a concert in a park alone means that the utility monster is happier than all the other audience members in the local community put together. Following utilitarian principles, in order to maximise happiness overall, we’d have to ban everyone except from the utility monster from attending this concert, and potentially any future events ever again. More generally, if the way to maximise utility overall is to make the utility monster as happy as possible, even if this comes at the cost of everyone else’s happiness, are we obliged to do so? While the designer of this thought experiment, Robert Nozick, was proving a point of his own, the issue remains, that achieving the Greatest Happiness principle is not unequivocally fair, or egalitarian.

As such, some argue that instead of focussing on happiness (or well-being), we should focus on social justice and equality. There is an uncomfortable tension in the well-being agenda and those of equality, diversity and inclusion.⁵ We have previously touched on Aristotle’s idea of a good life as dependent on a society supported by slaves. The question remains, at what or whose expense do the good lives of some, who make the ‘good society’ depend?

Returning to the Greatest Happiness principle, the main moral objection holds that it justifies a-moral means. This is owing to its consequentialist ethics: that if the aim is generating the most happiness for the most people, or the greatest total sum of happiness, then many actions may be justifiable. An easy way of imagining how this works is in the distribution of financial resources across a population. If you do something to improve the well-being of the largest number, it is highly possible that those who are marginalised (often the most vulnerable) in society will disproportionally suffer. We will return to this issue in the next chapter when we look at how Big Data and newer data practices disproportionately affect people of colour and the poor, for example. At the more dramatic end, such principles are argued against because they can be used to justify genetic manipulation, mind-control and dictatorship (Veenhoven 2010, 606). A useful example comes from science-fiction. Writer Ursula Le Guin’s (2017) short story The Ones Who Walk Away from Omelas features a thriving, joyful city whose prosperous existence depends on the extreme misery of a single child that lives in a dungeon.

Another issue taken with the ‘Greatest Happiness principle’ emerges from questioning the value of happiness as a goal: is it too focussed on pleasure, or is it just an illusion? Some question whether happiness as a goal fosters irresponsible consumerism and that it makes us less sensitive to the suffering of others. In other words that ‘happiness maximisation’ leads people to pursue an idea of happiness that is fuelled by irresponsible consumption, or to do what makes them happy without considering the consequences. This never-ending pursuit of things ‘to make us happy’ is called ‘the hedonic treadmill⁶’ and never satisfies; people always want more happiness and have been encouraged to seek gratification in the wrong places, to the detriment of their well-being, social well-being and ecological well-being.

You may think this sounds a culturally specific idea of happiness that applies to Western consumerism and you may recall the example from Chap. 2 which points to the dangers of assuming how people value things, comparing a TV to a photo album. You may also be thinking of criticisms of economists’ ideas of ‘preference satisfaction’ from Chap. 3, as well as those who disapprove of applying Western values, and valuation techniques to developing contexts, as we discussed was the case with the Human Development Index (HDI). You may also note that this idea of people as individual consumers seeking personal gratification is at odds with many societies that operate as collectives and, indeed, many of the values of societal well-being that the well-being agenda appeals to. These are not the only contradictions in the well-being agenda and we will continue to explore value judgements of what happiness is, and for who (especially if what we do, or are able to do is a driver of happiness) in further chapters.

John Stuart Mill was Bentham’s godson and another key figure in the story of happiness and economics. He is said to have disagreed with the idea of general happiness as something universally experienced. He believed that happiness from a game of ‘pushpin’ was not comparable to that from poetry; that without the idea of higher and lower forms of happiness, we should have to believe that a dissatisfied Socrates was worse off than a satisfied fool (Layard 2006, 22; 118). These variations in values and value systems are some of the key tensions in the agenda, especially when they inform us of what is good for our well-being.

‘The status race’ between people is seen as a key contributor to unhappiness (Layard 2006, 7) and is one of the behaviours we are encouraged to adopt in our commercialised society. Yet, competition is considered a contributor to progress.⁷ More than that, though, of course, there is competition between policy domains for resources and competition between academic fields to produce the method that gets used, the data that get used and the knowledge that gets used for policy.

There are several discussions surrounding how the well-being agenda addresses competition. On the one hand, it pretends to flatten competition, while on the other, it reinforces it. See OECD (2014) and concepts, such as ‘sustainable competitiveness’ (World Economic Forum 2013⁸). Other influential advocates for the well-being agenda naturalise a desire for ‘success’ and well-being measurements as tools for competition. For example, in a section entitled ‘Why use wellbeing as a measure of progress in society?’ in a report to a think tank, ex-Cabinet Secretary Lord O’Donnell explained:

As individuals we all are keen to know how we are doing: Are we top of the class or in the middle of the pack? So how should we measure success? (O’Donnell et al. 2014, 10)

Layard’s book both sells the Greatest Happiness principle, whilst also embracing some of the critiques, such as how the endless drive for happiness is bad for people and society, and how ideas of competition and success perpetuate this. There is a sense that some advocates of the movement cherry-pick, ignoring contradictions to tell a clear story, and this is familiar in criticism of the movement and its politics.⁹ The focus on meaningful goals (Layard 2006, 197) will always lead to questions of meaningful for who and leading to happiness for who. The focus on individualising happiness as something we can (and should) address for ourselves is linked to prominent positive psychologist, Martin Seligman, and his ideas of ‘authentic happiness’ (2002). Here we move on to consider ‘positive psychology’ for its influence on happiness economists like Layard, and society more broadly.

4.2 Positive Psychology

At this juncture, psychology can play an enormously important role. We can articulate a vision of the good life that is empirically sound and, at the same time, understandable and attractive. We can show the world what actions lead to well-being, to positive individuals, to flourishing communities, and to a just society. (Seligman 1998)

In his speech to the American Psychological Association (APA) in 1998, its new president outlined his hope for a ‘positive psychology’: a psychology which could help everyone as ‘a new science of human strengths’ (Seligman 1998). Positive psychology was more formally launched some two years later in a special issue of the American Psychologist. The editors: Seligman and Csikszentmihalyi framed it as a ‘new science’ for the new millennium (2000, 8).

The authors proposed a move away from psychology’s pathologising tendencies, by which they meant that the academic discipline and practice of psychology typically concentrate on the negative and the abnormal, to instead focus on the ‘positive features that make life worth living’ (Seligman and Csikszentmihalyi 2000, 5). Subsequently , Peterson and Seligman developed a formal classification handbook,¹⁰ Character Strengths and Virtues (2004). There were six virtues: wisdom and knowledge, courage, humanity, temperance, transcendence and a series of ‘character strengths’ (perhaps more traditionally called a trait) that fell under each category. Each of these character strengths is defined behaviourally, and it is recommended that it is measured using psychometric tests.

Having established a person’s strengths, a range of ‘empirically validated interventions’ were proposed to make the most of their positive traits, rather than address their weaknesses (Seligman et al. 2005). This was seen to assist lasting happiness (Seligman et al. 2005). The authors attempted to ‘present a measure of humanist ideals of virtue in an empirical, rigorously scientific manner’ (Peterson and Seligman 2004, back cover). These claims were echoed in reviews at the time in publications such as the American Journal of Psychiatry (e.g. Cloninger 2005,¹¹ 821).

Positive psychology has been lauded (by Seligman and his co-authors) as uniting the dispersed and disparate lines of theory and research about what makes life most worth living (Seligman et al. 2005). In 2000, Seligman and Csikszentmihalyi recognised that ‘positive psychology is not a new idea … and [they] make no claim of originality’ (Seligman and Csikszentmihalyi 2000, 13), instead arguing that they were able to present a ‘cumulative, empirical body of research to ground’ the ideas of ‘distinguished ancestors’.

It is interesting that positive psychology is presented as a ‘new science’ and ‘a cumulative body of research’, as these are also Layard’s claims in his book. These new, but linked, sciences, then, work on several levels as a valuable body of knowledge to claim that happiness can be a new science. The new science asserts that we now know the causes of happiness; that we now know the actions we have undertaken in the name of science, which are wrong; that these can now be measured; and that these measures can overcome philosophical queries via claims to science.

The happiness message here is that knowledge that is both policy-ready and accessible (popular, even ‘pop’) rests on clear and encouraging messaging (positive), innovation (new), authority (science) and morality (philosophy). It also, of course, must be measurable on an individual level that can be aggregated to population level.¹² It is, therefore, entirely dependent on well-being data, in particular the newer subjective well-being data that emerge from developments in positive psychology and economics’ interest in happiness, as an idea that has appeal for policy-makers and the public.

4.3 Establishing a New Science of Happiness

Layard’s (2006) book, Happiness: Lessons from a New Science emerged from a series of public lectures called ‘Happiness: Has Social Science a Clue?’ (Layard 2003). The LSE’s well-being programme was founded as a result of Layard’s public lectures. The website states:

Research from the programme has been devoted to understanding the causes of wellbeing and how wellbeing affects other outcomes that policymakers care about (such as education and physical health). (LSE Centre for Economic Performance n.d.)

The LSE’s well-being programme foregrounds making well-being knowledge popular by way of ‘lessons’, making knowledge ‘that policymakers care about’. These words might imply that the aspects of happiness that policy-makers don’t care about fall outside of the remit of the centre. This is indicative of a general feeling amongst some social policy areas that the work that they do is ‘invisible’ to policy-makers (as with Holden 2012, in the case of culture). Such a feeling is corroborated by academic research (e.g. Stevenson et al. 2010; Gray 2004) and evidence that some domains of social policy hold more sway with policy-makers than others.

Knowledge that policy-makers care about is, therefore, very much a concern. Let’s remember from Chap. 1 that the very idea of using well-being data to inform policy decisions (evidence-based policy) hangs on the idea that policy-makers can make neutral and objective decisions—if fed the right evidence. We have discovered already many indications to the contrary, as with the different interpretations of poverty data to suit political arguments in Chap. 1. We also know that ‘facts’ which reinforce established moral beliefs (or what we feel is right) are attractive to policy-makers and the public (Davies 2018) as confirmation biases. What we see here is the possibilities for the new ‘science[s] of happiness’ to become influential, with some believing the field is dominated by economics’ adaptations of psychology’s tools.¹³ It is easy to see how this might be the case, as a result of their capacity for persuasive arguments that we come to later in this chapter.

Economics (and its sub-disciplines) tend to have much influence with governments and multi-lateral institutions (like the UN, where many countries are represented in the decision-making processes). However, economists have not necessarily presented ideas in accessible ways as a rule. Their relevance to decision-making institutions is also a matter of tradition: they have long-held sway and so are highly represented in the decision-making process. Similarly, decision-makers tend to be literate in the principles of economics and in the UK, there is a trope that all MPs attend the very same course at Oxford or Cambridge universities: PPE (Philosophy, Politics and Economics)—to the extent that it ‘runs Britain’ (Beckett 2017). Decision-making processes are reputedly controlled by Treasury’s economic approaches, including the valuation techniques discussed in Chap. 2. Economics for well-being is an easier message to communicate than economics’ more abstract ideas, and borrowing the language of positive psychology is useful in promoting ideas that governments are, and individuals should be, taking positive action themselves.¹⁴

What we can also see, therefore, is the appeal of happiness in making economics an applied and more relatable discipline. This attraction can be seen in the increase in journal articles on well-being in the EconLit database (EconLit (n.d.) and see Chap. 2). Yet, despite the increase in happiness economics papers and emphasis on the increasingly robust ‘science’ of well-being (O’Donnell et al. 2014; Helliwell et al. 2015; ONS 2015a and 2015b), the lack of conceptual consensus outlined in Chap. 2, and expanded on in Chap. 3, has remained a concern for policy-making (Fleche et al. 2012, 11). Layard himself told a journalist (Rustin 2012) a decade ago that we were a decade away from well-being measures that are good enough for policy to be made using them. Yet numerous policy recommendations have been made on account of these measures over the last decade, as this book can attest to.

In their advisory paper to the ONS’ MNW Programme, Dolan, Metcalfe and Layard explain that any measure of well-being must be ‘empirically rigorous’, by which they mean ‘that the account of wellbeing can be measured in a quantitative way that suggests that it is reliable and valid as an account of wellbeing’ (Dolan and Metcalfe 2012, 411). Although the insistence that any empirically robust account must always be quantitative is preferred practice for certain disciplines, that does not mean it should not be questioned. Measurement of well-being basically wants to understand either change over time or difference between people or groups of people. These data can be captured by qualitative approaches, such as diaries or photographs, as described in Chap. 3, and do not need to actually be quantitative, therefore.

The authors continue by making an important point regarding any measure of well-being: that it should ‘be sensitive to important changes in well-being and insensitive to spurious ones. In practice, distinguishing between the two is quite a challenge and often relies on judgement based on a priori expectations’ (Dolan and Metcalfe 2012, 411). Returning to the well-being data examples we have already come across in Chap. 3, whether the OECD indicators or a small-scale questionnaire, understanding someone’s well-being using data gathered from any questions will have limits.

Recalling our hypothetical example of understanding whether a concert in a local park might improve well-being, how do we understand which aspects of the experience were the contributing factors? How can you disaggregate the contribution of the park, from the music itself, the people you were with, or the quality of the hotdogs for sale or the length of the toilet queue? Let alone understand which contributes to longstanding well-being or momentary happiness? Distinguishing between important changes to well-being and spurious ones is difficult, and therefore well-being data do not always meet Dolan et al.’s (2011a, b) criteria. Evidence of the impacts of particular activities and interventions on well-being is often criticised, as we discovered in Chap. 3: generally, if you ask certain questions because you seek a causal relationship, you are most likely to find it. The same is therefore an issue for well-being research more generally. The theory of confirmation bias is an account of how people tend to respond to causal messages which reinforce what they already believed or which suits their way of living and or thinking.

Thinking of the Facebook posts that have appeared on my feed in recent years, many different accounts, traditions and philosophies (that we have touched on briefly in this book) appear in the posts: we should try harder, we are trying too hard; we should visualise what we want and go for it, we spend too much time living in the future and not enough in the present and so on. All of these memes get shared because they appeal to things the person sharing already believes. Well-being wisdom repackaged is a large part of the wellness industry without any of the concerns with contradictions or evidence against the claims made. It appears that happiness economics may be similarly equipped to package simple ideas and positive psychology with long-held traditions, empirical evidence and call itself a new science.

There are several takeaways from this overview of the new sciences of happiness. First, that happiness economics seems to dominate the social sciences of well-being. Bearing in mind that all social sciences could be argued to be about understanding and improving well-being in some way, it is happiness economics that appears to be at the forefront—and that has certainly seen the largest increase as a discipline. This is because it has gained ‘scientific authority’ based on a couple of factors. First, is the combination of historical examples of moral philosophy, narratives of innovation and claims that the measures are growing increasingly robust. Second, these aspects are presented as simply as possible for media, policy and public audiences. Yet, the multidimensional nature of well-being means that it remains extremely difficult to remove confounders which include philosophical and empirical contradictions. It is, therefore, challenging to make and substantiate simple claims to know ‘the causes of well-being’, for example. Econometric models typically used to analyse subjective well-being data may lay claims to robustness, but are still not economically sound (see Cooper, in McKenzie 2015) and use data collected by questions that do not necessarily translate to the general public (as we shall discover later in the chapter).

These measures are, by the admission of prominent well-being experts, not neutral or objective measures of subjective well-being, but also involve subjective categorisation lists of people’s strengths or moral character (such as that in positive psychology) or a country’s development (as in the Human Development Index), as well as being the result of a process of decision-making when it comes to which data and how to model them. Having looked at the disciplines that have led to this new science of well-being, we will now turn to the data that inspired it and are generated by it. Specifically, we look at the ideas of subjective well-being and the methods that have shaped subjective well-being data and their prominence.

4.4 What Is Subjective Well-being?

Notions of subjective well-being or happiness have a long tradition as central elements of quality of life. (OECD 2013, 10)

4.4.1 How Is This Well-being Measure Subjective?

This portrayal of the ‘new science[s]’ of happiness is (as Seligman hints) not as new as implied, but also results from fundamental theories and indicators of well-being that date back centuries. One important—yet confusing—distinction is that there is the idea of experienced well-being (how we experience well-being or happiness) that gets called subjective well-being and then there are measures of well-being that form objective lists, like the OECD’s, that are based on subjective data.

As we have seen, objective approaches to measuring well-being investigate the objective dimensions of a good life (using largely proxy indicators). However, the subjective approach examines people’s subjective evaluations of aspects of their own lives by collecting numeric data. For example: ‘on a scale of 1–10, how safe do you feel walking home at night?’ This is not the same as how people feel about their well-being.

As we have also seen already, a number of well-being indices that were established around the same time have recognised the importance of taking people’s perceived well-being into consideration alongside objective lists in order to measure overall well-being. Subjective well-being data are generally captured using questions about how people feel they are doing. We are going into more detail about this now, in order to understand how these data can differ, and how they are different from the objective well-being indicators and the qualitative data described at length in the previous chapter. Crucially, it is the subjective well-being data about how we think our own well-being is that are the driving force of happiness economics and the second wave of well-being (Bache and Reardon 2013). As we shall discover, this is largely down to the influence of key advocates, such as Layard, in the well-being agenda.

Let’s consider the UK’s ONS’ subjective well-being data. As we have previously discovered, it uses four questions to understand what it calls ‘personal well-being’. The questions are:

• 1. Overall, how happy did you feel yesterday?

• 2. Overall, how satisfied are you with your life nowadays?

• 3. Overall, to what extent do you feel the things you do in your life are worthwhile?

• 4. Overall, how anxious did you feel yesterday?

How are these data used? The answers to these questions are on a scale of 0–10 and could be traced over time to see how an individual is doing. This is not going to happen in an anonymous national-level survey; instead aggregated data are used to understand population-level well-being over a specific period or to compare population sub-groups by geography or ethnicity, for example.¹⁵ Some of these questions with almost identical wording have been in surveys, and therefore generated data, for decades before ‘the ONS4’ were invented. Therefore, there are baselines to measure change against. The fact that these data have been collected over time can help establish how a major event such as COVID-19 has affected the well-being of the population, as well as more minor events. Chapter 7 runs through an example of how a policy change over ten years affects life satisfaction scores over a decade, for example.

These subjective well-being data can therefore be used to see how a particular event affected anxiety, alongside other social and structural issues, such as, say, poverty. Again, this does not mean that, for example, an individual’s household income is looked at against their anxiety levels, but that average anxiety of everyone who was asked the question (or, as we might say, the population sampled) is measured against the average household income levels. There are two things to remember about samples, the first is that few surveys are completed by a whole population, so the data collected almost always come from a sample; the second is that sampling is cleverly worked out so that if you sample enough of the population, you can make generalisable claims. Therefore, while national-level surveys do not measure nations in their entirety, they can make good estimations using mathematical rules. The other thing to say is that poverty can be measured using whatever indicator has been decided to represent poverty. There are numerous poverty indicators, which could be household income, for example, or the IMD (index of multiple deprivation). As we discovered in Chap. 1, ‘Introducing Well-being Data’, poverty is not one absolute, objective thing when it is discussed in parliament. Politicians cherry-pick from absolute and relative poverty measures and across different timeframes to arrive at the most complimentary statistics for their argument. So, what subjective well-being is measured against can also be subjective, in that the data and their uses are not automatically neutral or without bias, but are indeed chosen.

4.4.2 What Well-being Means to People Is Subjective

While we have covered what subjective well-being means previously, it is important to note that what well-being means for people in their everyday lives is subjective. Recalling the free text field analysis discussed in Chap. 2, when people are asked what is important to their well-being, they present different kinds of answers, about different areas of their life.¹⁶ Similarly, you might look at the aforementioned four questions from the ONS and think, ‘well they don’t capture my well-being!’ You might also think about how your answer to a question about life satisfaction will have fluctuated across a year, or even a day: meanings may not be constant and bad days at work or a bad commute will make it fluctuate, affecting how you might answer the questions on how satisfied and happy you are overall. Alongside these smaller, more everyday interferences to our mood are the major events, such as grief, injury, sudden or long-term unemployment, divorce, or of course, the generalised anxiety caused by an international pandemic. Answers to these questions can reflect a fleeting positive experience, such as attending a concert, or reflect something you are missing out on, on a longer term: good relationships, a stable job, mobility or good mental health. When we come to the different measures, we shall see how these are accounted for—to a degree.

As we shall discover, subjective well-being is complex to capture in a way that can inform behaviour. There are often trade-offs to supposedly positive choices. People who enter into adult education as mature students, for example, gain the pleasure of learning and feeling purpose in their life (Duckworth and Cara 2012), and although the negative effects are less studied (Field 2009), people miss many hedonic aspects of subjective well-being that they were previously used to, because time and energy for social and leisure activities are further compromised (Aldridge and Lavender 2000). The same can be seen in data about parenthood (i.e. Pollmann-Schult 2014): it’s rewarding, but you lose fun, time, money and autonomy; other relationships suffer and it can be unexpectedly lonely (Oman and Edwards 2020). A simpler binary, as found by White and Dolan (2009), is that time spent with children is relatively more rewarding than pleasurable, whereas time spent watching television is relatively more pleasurable than rewarding.

The measurement of well-being aims to capture how life is lived in society so that we can know how people are getting on. But this happens at a scale that means the subjective experience of well-being can be lost. Different people have different opinions on whether this is important to the overall measurements of well-being of populations. Experts who are great with numbers work on the basis that if your unit of analysis is a population (as in population level), and as long as those whose experiences don’t fit the story are outliers, then, it will statistically even out. Therefore, crucially, these measures are not necessarily meant to capture how everyone feels about everything. Instead, they are meant to be able to compare whether particular groups are affected or how things might change over time. The aim of these measures is to do better at measuring how people are doing overall, so that better policy decisions can be made.

Others argue that measuring well-being can obscure ill-being,¹⁷ particularly in already marginalised populations (Ahmed 2012; Tate 2016, 2017). There is concern that people who are already vulnerable are placed at further risk through the way that policy deals with data. For example, an issue which has gained prominence since the #MeToo movement is sexual harassment in universities. These cases can be obscured as they might be considered ‘outliers’, and so not get picked up by data which looks for overall well-being trends (Oman and Bull 2021, forthcoming). Similarly, marginalised experiences of ill-being are generally less visible (Tate 2016; Oman and Bull 2021, forthcoming; Oman et al. 2015). In Chap. 3, we briefly touched on the capacity of the domains and indicators in the OECD index, and how unlikely they would be to find the impact of policy change, like Bogue’s research on the ‘bedroom tax’. Capturing well-being data at scale, therefore, does not always pick up the complexity or subjectivities of ill-being.

The second wave of well-being is distinguished from the first, because it sees the collection of data about how people feel, at scale. For this to be effective, people need to relate to the ideas of well-being they are being asked to think about in the survey questions used. However, people do not always relate to the task at hand, or, even understand the questions asked. In my primary research, people talked about how they felt about the idea of measuring well-being (Oman 2017a), as they did in the ONS’ national consultations (as discussed in Oman 2015a, 2020). In both cases, some said it was a waste of time; that we have more important things to worry about. Others said that they didn’t understand how what is measured reflects their experience, or they didn’t understand the questions (Oman 2015a). As we will discover, the ONS also found this when they trialled the ONS4. So, although subjective well-being measures are thought more democratic (because they are about how people feel), they are—of course—by and large decided by experts and defined by experts, who preside on advisory boards and write influential working papers to the ONS and international agencies. What we see is a tension between ‘robust approaches’ and ‘understandable to everybody’.

4.4.3 Definitions of Subjective Well-being

Subjective well-being encompasses different aspects (cognitive evaluations of one’s life, happiness, satisfaction, positive emotions such as joy and pride, and negative emotions such as pain and worry): each of them should be measured separately to derive a more comprehensive appreciation of people’s lives. (Stiglitz et al. 2009, 16)

Subjective well-being measures aim to capture a number of aspects of how well-being is experienced. This moves the focus from the idea that what matters in a good life is the presence of a specific set of life circumstances or material conditions. Nevertheless, using objective indicators with subjective well-being ones enables estimates of the impact that material conditions (measured with objective indicators) have on how people feel about their life (subjective well-being measures).

Measuring subjective well-being therefore lends itself to analyses of which circumstances and conditions are important for well-being (Kahneman and Krueger 2006). Looking at subjective well-being data also, then, helps to understand the gap between material living conditions and people’s own evaluation of their circumstances (Helliwell 2003). These sorts of relationships are normally tested with a specific research question, for example: ‘how does wealth improve subjective well-being?’ You would pick what variable or data you would like to use to measure wealth: personal income, household income, property value, or identify where someone sits on a scale of poverty and wealth using a marker, such as their postcode. You would then pick how you wanted to measure subjective well-being. Using the ONS4 example, you might want to test the difference between how satisfied someone is with their life nowadays, or overall (life satisfaction) with how happy they say they were yesterday and the relationship between these two and wealth. One such example of this is a paper called ‘High Income Improves Evaluation of Life but Not Emotional Well-Being’ (Kahneman and Deaton 2010).

The OECD which ‘exist[s] to promote policies that will improve the economic and social well-being of people around the world’ (oecd.org) have also reported guidelines on measuring subjective well-being. The OECD propose a relatively broad definition:

Good mental states, including all of the various evaluations, positive and negative, that people make of their lives and the affective reactions of people to their experiences. (OECD 2013, 16)

As this book is not aiming to provide a definition or statement of determinants of well-being, but offer the tools to understand how others use and understand well-being data, we are going to look at an overview of subjective well-being.

The diagram (Fig. 4.1) illustrates the key components of subjective well-being, contextualising them in the theories we have encountered before. You may remember from Chap. 2 that the eudaimonic is based on Aristotelian (c. 330 BC) teachings, and can most simply be understood as purpose or flourishing. The hedonic begins with Epicurious ([341–270 BC] 1994), but is more familiar with the well-being agenda as a utilitarian principle (Bentham 1996 [1789]). It is most simply understood as pleasure, but more accurately means positive feeling.

Fig. 4.1

Accounts and examples of subjective well-being measures. (Adapted from Oman 2017a)

You will see how the divide of pleasure versus purpose is then captured as measurable aspects of life, and how they relate to each other, whether that is in someone’s experience and feeling, their satisfaction or a sense that their life is worthwhile in various ways.¹⁸ Inside each bubble on the right-hand side is the name of the type of subjective well-being measure (i.e. Life Satisfaction), underneath that is an example of the question or method used, and underneath that, a survey in which these questions have been used (the anomaly being ESM, which is not really used in national-level surveys, as I will explain, but is suitable in mobile apps data collection). I found it took me a long time to acclimatise to the idea that all of these measures and approaches are called subjective well-being; that they are related, yet so varied in approach, and use similar language. The next section walks you through this diagram, with examples from each ‘bubble’, to hopefully give you a better idea of how they work together.

4.5 Subjective Well-being Measures for Decision-Making

There have been many attempts to classify the different ways in which subjective well-being can be measured for policy purposes (Kahneman and Riis 2005; Dolan et al. 2011a, b; Waldron 2010). According to the recommendations on measuring well-being to the ONS, there are three uses for any well-being measure in policy: monitoring progress, informing policy design and policy appraisal (Dolan et al. 2011a). There are also three broad types of subjective well-being measure: evaluation (global assessments), experience (feelings over time or at specific times) and eudaimonic (reports of purpose and meaning, and worthwhile things in life). Table 4.1 shows how each of the three ‘types’ of subjective well-being can be used to measure well-being in a way which best informs policy. This section walks you through the array of subjective well-being measures and methods that feature in Fig. 4.1.

Table 4.1 Subjective well-being measures and their uses in policy

4.5.1 Evaluation Measures

Life satisfaction is the most commonly used evaluative measure of well-being (Fleche et al. 2012). Life satisfaction data are collected using questions similar to question 2 in the ONS4, ‘Overall, how satisfied are you with your life nowadays?’ The measure is popular with economists for policy-relevant research for numerous reasons. First, because of its longstanding prevalence in international and national-level surveys, such as Health Survey England, and more recently, the OECD’s high-profile Better Life Index. Second, it is thought to be accessible to policy-makers (Donovan and Halpern 2002). Third, some believe it to be the idea of subjective well-being that overlaps most successfully with how people make decisions in their own lives (Kahneman et al. 1999). However, some evidence suggests that, as a concept, life satisfaction is not understood by all members of the general public, particularly those who are marginalised in some way (Oman 2017a; Ralph et al. 2011). We might also question how universal a measure it is in developing contexts, which calls into question its utility on a global scale.

General happiness has been used as an alternative to life satisfaction and features in many international-level surveys. Key happiness variables seem to impact on general happiness responses in a similar way as life satisfaction (Dolan et al. 2011a, b; Waldron 2010). The measure aims to assess a person’s general happiness, and a popular example of trying to collect data on this concept is Cantril’s (1965) ‘ladder of life’¹⁹ (see Fig. 4.2). The Gallup World Poll uses the principles of Cantril’s ladder, where the questions are asked using a scale. This is a ‘self-anchoring ladder’, which asks respondents to evaluate their current life from 0 (worst possible life) to 10 (best possible life).

Fig. 4.2

Cantril’s ladder. (Adapted from Cantril 1965)

The term ‘general happiness’ can be used in reports (i.e. World Happiness Reports Helliwell et al. 2017, 2019) to mean the general happiness of a nation, or indeed, as John Stuart Mill²⁰ intended, ‘the sum of individual happinesses’ (Mill , cited in Crisp 1997, 78). This can be confusing and is something to be mindful of. It is not always clear if the term general happiness, when used to refer to population happiness, means taking individual-level data from something like Cantril’s ladder and multiplying it to derive a population-level measure, or if it is another measure, such as life satisfaction, used at scale.

Domain satisfaction is an approach which is interested in how people evaluate different features of their life, such as ‘work-life balance’ or ‘relationships’. These different features of our lives are grouped together into domains, which we have seen as a prominent feature in the objective lists approach. With the UK’s national well-being domains, that would be: personal finance, the economy, what we do, health (Physical and mental), education and skills, our relationships, governance, where we live, the environment. In theory you could collect satisfaction data about each domain, and if a person were satisfied with all domains this could demonstrate overall ‘life satisfaction’.

An example of a question to derive domain satisfaction data is from Understanding Society: UK Household Longitudinal Study (University of Essex et al. 2020), in which respondents are asked to rate their satisfaction with their general health on a scale from ‘completely dissatisfied’ to ‘completely satisfied’. Domain satisfaction data can be used to compare the reality of life with various standards of success (Veenhoven 1996, 30). Various domain satisfaction measures have been shown to correlate with numerous socio-demographic characteristics relative to income, health and gender, for example, and this has been replicated across studies (Dolan et al. 2008). Confusingly, sometimes the term ‘domain satisfaction’ is used to describe satisfaction across all domains (van Praag et al. 2003) but it more frequently refers to satisfaction within a specific domain, such as ‘satisfaction with personal relationships’, or ‘satisfaction with health’ which both appear in the UK’s national well-being measures. As with the case in this index, domain satisfaction is most often used in an objective list approach with other administrative data. This means not all the domains are measured using satisfaction data, but with proxy data, such as crime rate or education level.

Affect is a term used to describe the experience of feeling or emotion and is prevalent in psychology. As an aside, the term has recently been taken up in the broader social sciences and humanities to describe emotion and experience in a less medicalised way (Sedgwick and Frank 2003; Thrift 2004; Massumi 2002; Ahmed 2010; Berlant 2011; Wetherell 2012, etc.). While the concept is linked, these theoretical uses of the concept of affect are not really captured by surveys, which is an important distinction that is rarely acknowledged.

General Affect means how people are doing overall and is a concept which is understood in evaluation questions. In psy-sciences,²¹ it is the relative frequency of positive and negative affect that is thought to be key to how we experience well-being. The Affect Balance Scale (Bradburn 1969) and the Positive and Negative Affect Scale, or PANAS (Watson et al. 1988; see Fig. 4.3), involve questionnaires that are designed to gain numerical responses to general statements about different affects. These questions are also used in some large-scale surveys, such as the English Longitudinal Study of Ageing (ELSA n.d.).

Fig. 4.3

PANAS questionnaire. (Adapted from Watson et al. 1988)

Influential psychologists Huppert and Whittington have cautioned for some time that different versions of positive and negative scales are less similar than implied. Also, these scales are susceptible to change and adaptations in surveys. This must be accounted for when considering subjective well-being metrics which use them. Affect is also a key part of experience measures, which want to capture affect at a particular time or context.

4.5.2 Experience Measures

Experience measures aim to capture a person’s feelings at a given, specific time which can be thought of as ‘the amount of affect felt in any moment’ (Dolan et al. 2011a, 7). Measures are constructed with the Benthamite view that certain aspects of life are good or bad, based on their qualities of ‘pleasurableness’ or painfulness (Crisp 2006). How happy, sad or anxious any person is at a particular time is re-conceived as well-being by taking the average balance of pleasure (or enjoyment) over pain, measured over the relevant period. As already pointed out directly above, there is some evidence that positive and negative affect do not directly predict each other and should therefore be measured separately. Heeding Huppert and Whittington’s concerns (2003), positive psychology has more recently begun to conceive of well-being as a continuum (ONS n.d., 3; Diener et al. 2009), rather than something which can be assessed by taking the average of positive and negative measures. The experience approach relied on in surveys will tend to specify a period of time for you to remember how you felt. In the ONS4, this is the only account with two questions, one for happy yesterday and one for anxious yesterday (see also Table 4.1). As well as specifying the exact moment you want someone to recall, other methods capture people’s emotions at multiple points in a day or week, and for that reason, they are not really included in national-level surveys, which would be difficult to administer. However, they are suitable for mobile apps, as we shall discover.

The Day Reconstruction Method (DRM) (Kahneman et al. 2004) is perhaps the most renowned of numerous measures which attempt to capture experienced well-being over time which is called the experience sampling method (ESM). The DRM is a diary-based technique, through which participants reflect on the main episodes that affected them on the previous day and recall the type and intensity of feelings. In other words, it literally takes a sample of feelings from specific days and weeks. Affect is an aspect of subjective well-being that is particularly sensitive to immediate surroundings and activities (Smith and Exton 2013, 230). This is why it is considered suitable for understanding the relationship between what we do and how we feel, as well as situational aspects of life that affect us.

For example, short-term affect data can be collected through DRM approaches to include information about both activities and locations, as well as the affective states accompanying them (Kahneman and Sugden 2005). Such an approach has the potential to capture data on how people spend their time and the ‘experienced utility’ (Kahneman and Sugden 2005) of such activities. For example, 132 teachers in the Netherlands completed a daily diary on three consecutive work days as well as a background questionnaire (Tadić et al. 2013). The researchers found that despite a lack of work-life balance, working hard was not necessarily detrimental to the teachers’ happiness scores. If you take these scores at face value, then if the teachers were ambitious, then striving towards their goals was satisfying, but this motivation was not necessarily constant.

The Ecological Momentary Assessment (EMA) (Stone et al. 1999) is based on self-reports of well-being at specific, but often randomly chosen points in time. Reports explicitly include self-assessments of behaviours and physiological measures, but also the recording of events. In Chap. 5, we discuss how an app alerts its users to record how happy they feel at random moments, allowing the user (and whoever is capturing their data) to track their mood over time and establish what is good for their mood. The researcher who developed ‘mappiness’ has used these data to measure a number of aspects of happiness: that we are most miserable commuting, on the one hand, and that ‘happiness is greater in natural environments’, for example (MacKerron and Mourato 2013; Krekel and MacKerron 2020). These data have also been used (Fujiwara and MacKerron 2015) to compare how happy people feel doing different kinds of activities from birdwatching, to making love; and more specifically, between artforms, such as watching the performing arts or reading alone.

An exploration of the determinants of, and changes to, affect and time-use may offer understandings of how people’s ‘experiences of utility’ vary. Returning to the example of the local, subsidised concert in Chap. 3, again, the questions we asked there can help us understand how people’s responses to the cost, amount of time and effort vary, and how that changed their declaration of how they felt. This may be at odds with the ‘utility’ assumed by ‘the provider’, whether that is the local council, a theatre company or another funder.

However, it is important to remember that people who attended our hypothetical park concert, self-selected to do so. This is one of the key issues with valuing how people experience social and cultural activities: it makes it difficult to say how a particular experience might affect others in the future (Dolan et al. 2011b, 12). Also, people are liable to ‘mind wanderings’, which can mean they are not thinking of what you think they are when you ask them how they are feeling (ibid.: 8). Furthermore, what makes sense, or represents the experience of one person may not manifest in the average of a sample.

These approaches ostensibly measure at different points during the day and they relate to experiences associated with specific activities and time points. However, because in a national-level survey, large population samples are questioned at certain points during the year, it is not feasible to repeatedly survey respondents during a particular day. As an alternative, the rationale with the ONS4 experience measures is to ‘replicate’ or ‘proxy’ ESM approaches by asking respondents for their experiences and feelings relating to a whole day (yesterday).

While there is potential for the measurement of change in affect and time-use longitudinally, questions remain as to whether existing national-level survey data can capture the sensation and emotion of ‘situated experience’ (how it felt, to be there, in that moment) in a meaningful way, and to do so over time. In cultural policy studies, there is often a call for longitudinal measurement of the relationship between cultural participation and aspects of well-being. It is thought that this will solve some of the proclaimed issues with the evidence base (around data and causation, discussed in the latter chapters of the book). However, while longitudinal analysis can help address issues of causal direction in the evidence, they will not address issues related to capturing the duration of the impact of an experience, and this also is not always clearly understood (Oman 2017b).

4.5.3 ‘Eudaimonic’ Measures

Some conceive of eudaimonia as part of subjective well-being (Dolan et al. 2011a, b), while others choose to conceive of subjective well-being as purely hedonic (‘happiness’, ‘life satisfaction’ and ‘affect’). Eudaimonic or ‘eudemonic’ theories conceive of people needing purpose and as having various underlying psychological needs, such as control and connectedness (Ryff 1989). Likewise, that satisfying these needs contributes towards well-being independently of any pleasure they may bring (Hurka 1993). These accounts draw on Aristotle’s ‘eudaimonia’ as what makes for a good life.

4.5.3.1 Psychological Well-being

In the 1960s, Harold Dupuy, psychologist at the National Center for Health Statistics, developed his Psychological General Well-being (PGWB) Schedule, a questionnaire of 68 items to measure the psychological distress of the American population. It was reduced and simplified to 18 items for introduction to a general health survey in the 1970s and then increased to 22 items to become the PGWB Index. One of the case studies in Chap. 7 uses the PGWBI, adapted again for an Italian survey.

Developed by psychologist Carol D. Ryff, the 42-item Psychological Wellbeing (PWB) Scale measures six aspects of well-being and happiness: autonomy, environmental mastery, personal growth, positive relations with others, purpose in life and self-acceptance (Ryff 1989). Again, different versions of the scale have been adapted to suit different contexts, including an 18-item version (Ryff and Keyes 1995). Ryff and Keyes (1995) compared their eudaimonic measures with evaluations of life satisfaction and happiness, finding that self-acceptance and environmental mastery were associated, but that positive relations with others, purpose in life, personal growth and autonomy were less well correlated.

4.5.3.2 Worthwhileness and Overall Evaluation

More simply, eudaimonia is related to ideas of worthwhileness that are connected to the diagnosed psychological needs listed above and, but can also be addressed with one question, as with the ONS in Fig. 4.1. White and Dolan (2009) measured the ‘worthwhileness’ associated with activities using the DRM method. They found some discrepancies between those activities that people find ‘pleasurable’ as compared to ‘rewarding’. The example they used is that spending your time watching telly brings pleasure, but few rewards, while spending time with children is the opposite.

4.5.4 How These Measures Can Be Applied

There are important distinctions when considering how aspects of happiness economics can apply value to what we do. Recalling the photo album versus TV example from Chap. 2, it can be difficult to ascribe value to others’ activities. Ateca-Amestoy has tried to explain the value of leisure as a psychological need for different kinds of experiences, and which impact on how we evaluate our quality of life.

[L]eisure is a human need to be fulfilled by household production and consumption of what we may call ‘leisure experiences’. Those experiences are commodities that fall directly within the individual’s determination and assessment of his/her quality of life. This means that leisure is one of the arguments of the individual’s utility function, one of the instances from which he/she will achieve well-being. (Ateca-Amestoy 2011, 53)

The importance of understanding the different kinds of well-being benefits offered by different types of leisure has been an aim of high-profile research over the isolated periods of COVID-19 lockdowns (https://www.covidsocialstudy.org/). That some activities offer hedonic utility, such as streaming and television watching, and some offer eudaimonic, such as reading (and some both, of course), is being studied (Bu et al. 2020; Mak et al. 2020; Nuffield 2021). However, what people do is often polarised as ‘watching television excessively’ (Bu et al. 2020, 7), with claims that ‘these changes in behaviours and mental health are reflected in people’s assessments of the differences in their lives between this lockdown and that of spring 2020’ (Nuffield 2021). This is slightly misleading: from the evidence presented, we do not know that it is people’s behaviour that has changed people’s assessments of their lives, when policy-making and poor weather in a pandemic are arguably having a greater affect than watching the telly. As you may recall, this is one limit of applying the ‘Greatest Happiness’ principle and can also be the consequence of confirmation bias. For example, the Sarkozy Commission contrasted ‘cultural events’ with ‘poor leisure’²² (Stiglitz et al. 2009, 49) and Layard’s analysis of television’s negative effects was inevitably biased by an idea of good leisure.²³ However, as we have discovered, assumptions as to what qualifies as good leisure and poor leisure are problematic ethically, and will not present universal results.

That pleasure and reward do not map onto each other neatly aligns with Aristotelian thinking. The think tank, New Economics Foundation (NEF), has been highly influential in UK well-being research since the mid-2000s. Its definition of well-being is ‘developing as a person, being fulfilled, and making a contribution to the community’ (Shah and Marks 2004, 2). The report, ‘A Well-Being Manifesto for a Flourishing Society’ (Shah and Marks 2004), called for well-being to be foregrounded and for governments to work towards a ‘flourishing society’ with ‘happy, healthy, capable and engaged’ citizens (Shah and Marks 2004, 2). In 2008, NEF introduced a set of guidelines called the ‘Five Ways to Wellbeing’, based ‘around the themes of social relationships, physical activity, awareness, learning, and giving’ (Aked et al. 2008, 17), summarised as connect, be active, take notice, keep learning and give.

The ‘Five Ways’ have proven successful, and have been adopted in parts of the National Health Service and by organisations such as Mind, the mental health charity,²⁴ as well as many other social policy areas. Individual institutions have chosen to adapt it when offering well-being advice to staff and other members of the institution. The University of Manchester, for example (The University of Manchester n.d.), has adapted it into its ‘six ways to well-being’ which is used to frame its advice to students and staff. The cultural sector has embraced the guidelines, both in arts practices aimed at improving well-being (Dodd and Jones 2014) and as a means of evaluation of eudaimonic and broader well-being aspects of cultural engagement (Daykin and Joss 2016). According to a review of the evidence from international arts and health literature, ‘[t]he benefits from arts programmes resonate strongly with the evidence-based “five ways to wellbeing” model of mental health: connect, take notice, keep learning, be active, give’ (Bidwell 2014, 3).

The success of the ‘Five Ways’ is down to legibility of its framework to many policy sectors, people in the general population and policy-makers. Let us briefly return to the takeaway conclusions from how the new sciences of happiness generate knowledge that is both policy-ready and accessible (popular, even ‘pop’) rests on clear and encouraging messaging (positive), innovation (new), authority (science) and morality (philosophy). The Five Ways to well-being meet all of these criteria, perhaps more than the idea of subjective well-being in and of itself. We will move towards closing, by looking at the ONS4 as a case study to understand the importance of legibility, transparency and understanding, when deciding on how to collect subjective well-being data.

4.6 Case Study: Subjective Well-being, by the Office for National Statistics’ Design

The UK’s national well-being measures are categorised into ten domains. These are as follows: Our Relationships; Health; What we do; Where we live; Personal Finance; Economy; Education and Skills; Governance; Environment; Personal Well-being.²⁵ Each of the ten domains is composed of multiple indicators, just like the OECD’s index that is described in detail in Chap. 3. The subjective well-being domain was named personal well-being, because it was thought to make this domain more understandable to a general audience, which was considered particularly important to the MNW programme.²⁶ This domain comprises ‘the ONS4’.²⁷ Table 4.2 presents the questions, together with their rationale.

Table 4.2 The ONS4 capture different aspects of well-being

‘The ONS4’ were designed to capture three types of subjective well-being: evaluative, eudaimonic and affective experience. The four individual subjective well-being questions ask people to give their answers on a scale of 0 to 10, where 0 is ‘not at all’ and 10 is ‘completely’. The ONS considered consolidating the figure of all four measures to provide a single measure of personal well-being. Just as with the HDI in Chap. 2’s discussion of objective lists, this single number is easier to communicate and is most often discussed in national media and by politicians. It was, however, not considered conceptually robust to do so. Here, again, we see a tension between robust and easy to understand.

The first results from trialling the ONS4 were published in April 2011 (ONS 2011a ). The aim was to gather responses from survey participants which are an ‘assessment of their life overall, as well as providing an indication of their day-to-day emotions’ (ONS 2015a, 5). ‘The ONS4’ gained National Statistics status in September 2014 and, since then, have continued to be introduced to surveys across government. They are, therefore, not necessarily intended to be used by themselves. Table 4.3 shows the variety of these surveys and the sorts of data they capture. The Government Statistical Service has more recently published advice on the harmonisation of the ONS4 (Nickson 2020). This aims to ensure subjective well-being statistics and data are ‘comparable, consistent and coherent’ across government departments and beyond.

Table 4.3 Surveys containing the ONS4

While we know that the ONS4 capture the different aspects of subjective well-being, and there were many reports and working papers from the time, it was quite difficult to find methodological or administrative detail readily available on how the questions themselves were decided on. In particular, the final wording chosen. In parallel to my PhD research, and after much searching, I found a detailed report to the Technical Advisory Group (Ralph et al. 2011) on the findings from 44 interviews.

This report is phase 2 of qualitative findings from testing the ONS4. Notably, not all the responses to the trials were positive in this report. Limitations were found in how able people were to answer the questions. Interestingly, when it came to the life satisfaction question (thought to be the most robust, as you may remember), not everyone thought that being satisfied with life was positive; some believed it neutral and some thought it a negative commentary on their lives (Ralph et al. 2011, 5). With the ‘worthwhile’ question, answers were affected by what was seen as social desirability, leading to inflated scores. This is known as response bias, and meant that certain people (arguably with lower subjective well-being) did not want to appear as if they did not have worthwhile lives to the interviewer (Ralph et al. 2011, 5). A later phase in the cognitive testing also details how, when the questions are administered face to face, people felt uncomfortable giving negative scores in front of loved ones (ONS 2012, 7).

When it comes to understanding the meaning of the questions, the qualitative report also states that:

Where the question was not understood this tended to be by those with lower educational attainment. This group simply did not understand the term ‘worthwhile’. (Ralph et al. 2011, 5)

In some ways, what is more concerning is that:

For the most vulnerable respondents, answering this question was distressing and in some cases respondents became visibly upset. It is recommended that ONS investigate the possibility of creating a flier that interviewers can leave with respondents, which tells them where they can seek help if it is required. (Ralph et al. 2011, 5)

Having a protocol at the end of research interview, should the interview have covered sensitive issues, is standard ethical practice in qualitative research, but less so in survey collection methods. It is not clear whether filers were trialled after asking participants these questions.

In summary, there were a number of issues that the qualitative research in 2011 uncovered with these four questions. These include: how accurately people were able to answer, based on their understanding of the questions; how honestly people felt capable of being when answering sensitive questions; and that arguably these questions could be detrimental for someone who was not experiencing good well-being. These issues revealed by the testing were brought to the attention of the programme’s advisory groups.

The minutes from the Technical Advisory Group in 2011 outline the importance placed on these four questions. Lord Layard refers to these questions as ‘the work of the ONS’ and outlines that it is the status of this work that is the aim of the wider MNW programme, which reiterates the importance of this new subjective well-being data to the broader agenda. Layard also outlined his concerns that the ‘UK is less likely to set international agenda if introducing unnecessary changes’ (ONS 2011b). These minutes might suggest that what was learnt from the trials were unlikely to be able to change the new measures, which we have discovered were built from a synthesis of disciplines and authority.

The Technical Advisory Group (TAG) had disappeared from the ONS publications archive when I was originally undertaking this research to try and ‘follow the data’, and understand the methodological origins of the questions. However, I was able to find a record of the group by way of a fellow researcher. The National Statistician, Jill Matheson, refers to a National Statistician’s Advisory Forum and a Technical Advisory Group. All traceable records of TAG meetings are headed by a list of those present. Only ONS, civil service and academic economists were present at the meetings in the minutes I was able to locate. However, another academic researcher confided to me during my ethnography fieldwork that there was a clear hierarchy in the programme and psychologists were rarely listened to, with the economics experts dominating proceedings. This appears to be substantiated by minutes regarding the development of the SWB measures (ONS 2011a). It also corroborates claims that economists dominate how evidence is presented, acknowledged and applied in these forums. However, it is important to note that these are not impartial accounts, either.

Psychologists reflecting on phase 2 of the testing of the questions advised that they could cause psychological distress in some participants, but this concern is absent from other outputs. Notably the report on phase 3 (ONS 2012) mentions it found no issue of difference in legibility for different people, unlike phase 2 (Ralph et al. 2011). More importantly, however, it does not acknowledge that one phase of research found the ONS4 questions to be detrimental to well-being. As you can see, looking under the bonnet of the data presents questions about how the measures work in practice, how they are decided on and by who, and what evidence of success becomes part of record and what disappears. It also reveals issues with regards to how data collection on well-being can be detrimental to well-being that are rarely considered.

4.7 Summarising What Measuring Subjective Well-being Does

So, as we have discovered, subjective well-being is often characterised as being concerned with happiness alone (OECD 2013, 10). Instead, subjective well-being is a more complex combination of various aspects of the lived experience; it involves several distinct ideas with disciplinary and theoretical histories. While these concepts can sometimes correlate when measured, the evidence for this remains inconclusive (Clark and Senik 2011 in Fleche et al. 2012, 9). Research using secondary subjective well-being data, therefore, should clearly establish the conceptual differences between different components of subjective well-being, to be sure that what is aimed to be measured is what is actually being measured. Furthermore, this could be better communicated.

While subjective well-being has been thought to predict behaviour in meaningful ways (Diener and Tov 2012), the subjective well-being measures we have encountered are thought valuable because they enable an empirical examination of the factors which cause improved or reduced well-being (Fleche et al. 2012, 10). Some economists (such as Layard) believe that these qualities make these approaches an improvement on traditional micro-economics approaches which rely on notions of utility. Utility, as we discovered in Chap. 2, is the idea that satisfaction is experienced by consuming a good or service and that ‘rational choice’ drives consumers to remove dissatisfaction (or discomfort) and to maximise on this satisfaction.

In general, subjective well-being data allow for an assessment of the positive or negative contribution of one factor (such as public libraries) over another, which may seem unrelated (such as being made redundant), to well-being. This therefore allows an appraisal of different factors which can be both monetary and non-monetary (Fleche et al. 2012). However, we must also remember that it can be difficult to separate spurious from essential well-being effects, and doing so often relies on human judgement.

The qualities of these newer measures of subjective well-being have led to influential figures, such as Lord O’Donnell²⁸ arguing for ‘a well-being approach’ to inform decisions that manage COVID-19 (O’Donnell 2020). O’Donnell and other advocates for this type of well-being approach argue that well-being measures should inform ‘trade-offs’ and ‘the true costs of lockdowns’, for example, by declines in mental health and access to healthcare (O’Donnell 2020). It could be a means of deciding the balance between how one policy move related to protecting the economy (which includes people’s jobs) to another, such as healthcare (which includes its own financial considerations and multiple mortality rates). It is also this approach that helps unpick the assumed correlation between having money and attaining happiness that we opened this chapter with.

The different definitions of subjective well-being further complicate issues for those wanting to use well-being data in their research or to understand the research of others. The confusing naming conventions, overlapping definitions and disagreements as to what counts as subjective well-being, objective well-being, personal well-being or societal well-being also don’t help those wanting to understand the ways in which well-being measurement more broadly furthers knowledge of the human experience. There is also work to be done on how the different ideas of subjective well-being overlap with longstanding cross-disciplinary beliefs and assertions regarding the value of different domains of life to well-being that we will encounter later in the book. In short, there is a transparency gap in the discussions of rigour, classifications and measures in the ‘science’ and the legibility of what that means to everyday people, despite the efforts made to do so.

4.8 Conclusion

Looking at the invention of subjective well-being measures in the UK offers context behind the ubiquity of well-being measurement practices. Understanding the recent history behind a specific way of measuring a particular idea of well-being, that is considered robust and universal, is vital to appreciate the limitations of such projects. This chapter’s comprehensive survey and critical lens aims to offer tools to promote better understanding of the power of these well-being data, their capacity to change culture and society, and the limits of their application in areas of social and cultural policy and practice.

In short, 'the new science of happiness' has much to offer understandings of well-being and the human experience more generally. The techniques, whether originating as national-level social survey questions or personal psychological tests, can be adapted and applied to other environments and have been used widely to understand the impacts of COVID-19. Yet, politics, disciplinary and international competition compromise their neutrality. These contexts are vital to understanding the subjective well-being data generated through survey questions and their uses to inform important decisions in policy development, monitoring and evaluation, and the way these, then, promote behaviour change in people.

We have seen evidence that the national well-being measurers want to be top of the class, with possibilities that complexities of the questions in certain contexts were disregarded. This leaves us with questions. Could it be that in the keenness to compete in the new science and the international game of devising the best measures, considering the subjective experience of people answering questions on subjective well-being may have been side-lined? It transpires that less attention is paid to the qualitative trials of questions that end up as ‘robust measures’ than you may imagine, as I also found with some questions long-used to measure class (discussed in Chap. 9). Yet, should it be a great surprise that quantitative researchers and national statistics offices tend to overlook the qualitative aspects of their methodologies? It is hard to say because such evidence is hard to find.²⁹

We have used data on the contexts behind subjective well-being data to understand them better: who collected them, interpreted them, looks after them and uses them. We have seen some trends emerge across people and policy, but found these contextual data have limits to what can be understood, too. It can be hard to find all the archival information we need, and it can be easy to interpret the absence of evidence as some sort of cover-up, when actually in policy-making and public services, institutional memory is often lost through the ‘churn’ of staff and these issues of paper trails. There is, sadly, ‘no culture of a repository of knowledge’ (Hallsworth et al. 2011, 8). Thus, the data we have on how data are made can be as compromised or limiting as the quantitative or qualitative data we have been discussing in these last two chapters.

This chapter has looked at the new sciences of happiness as people, publications, projects, politicians, agencies and disciplines. Easterlin is presented as the turning point in this tale, because he offers a useful narrative device. However, the limitations of how economics was used to understand human flourishing have been known longer—as presented in Chap. 2—and indeed in the introduction to Easterlin’s paper. Discovering the stories behind data in this way, we are able to see how all these different components work together to make the well-being agenda. We can also see that it is the subjective measures, rather than the compiling of objective lists, that are the greater driver of the agenda, and that this is—in part—owing to claims to innovation.

Essentially, however, the new sciences of happiness: the new measures and uses of data from old questions (Allin and Hand 2017), are the driving force behind the well-being agenda. At least what we have referred to as ‘the second wave’ in this book. Without the technological advances and the advocacy for the new measures, we might ask, would we have seen calls for the change in policy? Thus, the terms data-driven decision-making and evidence-based policy-making take on new meaning—where the promise of the possibilities of well-being data changes the policy rhetoric and call for more data to be collected. Data do not only capture social change, but ensure it, and as the next chapter demonstrates, it feels as if Big Data increase this pace of change, but how do they impact on well-being?