Greece and Rome

We can begin with the great Greek philosopher Aristotle (384–322 bc). He, more than anyone, is responsible for the idea that there must be some ultimate end that we should aim at – with other things like wealth, health and good relationships being good because of how they contribute to that end. That is the hierarchical idea we already displayed in Figure I.1. The ultimate end he called ‘eudaimonia’, which he envisaged as a balanced, rational and virtuous state of being. He especially emphasised virtue, which he considered essential if your experience of life is to be truly fulfilling. He did not define virtue exactly as this book does, but he had a realistic modern empiricist view on how it should be acquired – by constant repetition and habit-formation.

Greek philosophers who came after Aristotle emphasised different aspects of his message. Epicurus (341–270 bc) emphasised the importance of a simple life, focused on things you really enjoy, like friendship and family life. Zeno the Stoic (333–264 bc) emphasised civic virtue and the ability to calm your mind whatever adversity befell you. The ideas of these two thinkers spread widely through the Roman world. Eventually, it was Stoicism that became the philosophy of much of the Roman middle class – and of the emperor Marcus Aurelius (ad 121–180), who wrote much about the secrets of happiness. Similar ideas had already been developed in China by Confucius and in India by the Buddha.

However, in the centuries that followed, Christianity brought in a quite different perspective. Happiness was still to be sought but only in the afterlife. In this world, the aim was virtue alone and this would bring happiness after death.Footnote ²

The eighteenth-century Enlightenment

It was only in the late seventeenth century that philosophers dared to re-establish happiness on earth as the goal of life. The great English philosopher John Locke (1632–1704) took it as axiomatic that people wanted to feel good and that to acquire that feeling was their principal motivation. But it was a Scotsman, Francis Hutcheson (1694–1746), who took the first huge step towards establishing happiness as the goal for society, rather than just for the individual. He taught that the moral thing to do was whatever produced ‘the greatest happiness for the greatest number’Footnote ³ – or, more precisely, what produced the greatest total happiness. And he argued that this is what any ‘impartial spectator’ would think.

From the wellbeing point of view, this idea was probably the greatest idea of modern times. It has three key implications.

1. (1) For judgements: If we want to compare two different situations and to say which is better, we should use the happiness of the people as the test. This would be so, for example, if we were asking which country is doing better, or whether the present is better than the past or which is the best way to spend public money.

2. (2) For individual action: If we ask, ‘What is the ethical thing to do?’, the answer is, ‘Whatever produces the greatest overall happiness’.

3. (3) For government policy: When we ask, ‘What should the government be trying to achieve?’, the answer is, ‘The greatest possible happiness for the people’.

As the eighteenth century progressed, these ideas took a firm hold among the educated classes in the English-speaking world, including in North America.Footnote ⁴ Thomas Jefferson, who drafted the US constitution, would write, ‘The life and happiness of the people is the first and only object of good government.’ But the writer who immortalised these ideas was the Englishman Jeremy Bentham (1748–1832) in The Principle of Morals and Legislation. In this book, Bentham argued that all actions should be judged by their consequences – by their impact on the happiness of everyone. In other words, actions should be judged by their ‘utility’. Thus Bentham’s concept of ethics, which is so humane, became known as ‘utilitarianism’, which makes it sound quite the opposite. (Similarly economics, which was founded by Adam Smith (1723–1790) to study the economic conditions for the greatest happiness, chose the word ‘utility’ to describe individual wellbeing rather than the word ‘happiness’).

During the nineteenth century, the principle of the greatest happiness inspired many major social reforms. It was forcefully presented by John Stuart Mill (1806–1873) in his book Utilitarianism and in his essay ‘The Subjection of Women’. However, he also proposed that some pleasures are intrinsically better than others (for example, poetry was better than the game of pushpin). This was in our view a confusion of means with ends. You will certainly be happier in general if you have a clear purpose in life (as we show clearly in Chapter 4). And you will make others happier if you are virtuous. But the ultimate test of a society is how happy people are (irrespective of how they became so).

During the nineteenth century, economists continued to focus on happiness as the ultimate test of economic arrangements, and the economist Edgeworth (1845–1926) foreshadowed the modern science of wellbeing when he talked about the need for a ‘hedonimeter’ – a technique to measure happiness in a cardinal fashion. In the nineteenth century, most economists believed that an extra dollar gives more extra happiness to a poor person than to someone who is better off. Similarly, in psychology, the great psychologist William James was primarily concerned with life as people experienced it from the inside.

Behaviourism

But then in the twentieth century, psychology turned away from the study of wellbeing – to behaviourism. Scholars like Ivan Pavlov and John B. Watson argued that we could not know what occurred inside people or how they felt. One could only study how they behaved and how this was affected by external stimuli. This doctrine soon impacted on economics, and in 1932, the economist Lionel Robbins asserted that we could not compare the happiness of one person with another.Footnote ⁵ Even for the same person, he argued, we could only rank different situations, A, B and C, in order; we could not say whether going from A to B gave more (or less) extra happiness than going from B to C.

The consequences for economics were serious. It became impossible to evaluate any issue involving income distribution. We could not say whether it was better if two people had $50,000 dollars each or if one had $20,000 and the other had $80,000. All that could be claimed was that a change was good if some people gained and no one lost (a so-called Pareto improvement). When, as generally happens, a change involves some gainers and some losers, economists, according to Robbins, could say nothing. But in fact, this did not stop them from pronouncing in such cases, and John Hicks and Nicholas Kaldor quickly provided what they considered a justification. They argued that, even if some people lose, a change is good if the gainers could have compensated the losers – even if they actually didn’t.Footnote ⁶ This was the so-called Hicks–Kaldor criterion for a welfare improvement.

This criterion also justified a second dubious practice: the use of national income (or gross domestic product, GDP) as a measure of national wellbeing. GDP is a measure of the scale of economic activity, the things that happen for which people get paid. In the national income, everybody’s dollars are added up regardless of who has them; and, if the national income increases, that is a Hicks–Kaldor improvement, even if many poor people have become worse off. But GDP was invented for a quite different purpose – by Simon Kuznets – as a way of analysing the business cycle and the fluctuations of unemployment. It was not meant to measure wellbeing. As Kuznets himself said, ‘The welfare of a nation can scarcely be inferred from a measurement of national income.’Footnote ⁷ There are two obvious reasons for this view:

• GDP adds up the dollars of rich and poor as if they are of equal value.

• GDP fails to include any source of wellbeing other than things you can buy. As Robert Kennedy put it, ‘GDP does not allow for the health of our children, the quality of their education, or the joy of their play.’Footnote ⁸

But despite all this, GDP per head became the totem of national success in the post-WWII period.

In recent years, there has, however, been a massive pushback against the idea that GDP is an adequate measure of national progress. In 1974, the economist Richard Easterlin used psychological surveys to show that wellbeing in the United States had not risen since WWII, despite massive economic growth.Footnote ⁹ And subsequent research has shown that economic growth is no guarantee of increased wellbeing (see Chapter 13). So it is not surprising that even before the economic slowdown in the West from 2008 onwards, there was increased public demand for a wider goal than economic growth.

Taking feelings seriously again

But key to the demand for a new approach has been our increased ability to measure wellbeing and to understand its causes. The key figure here has been Edward Diener (1946–2021). Beginning in the early 1980s, he showed that wellbeing could be effectively measured and explained.Footnote ¹⁰ As time went on, more and more psychologists joined this enterprise, including Daniel Kahneman who in 2002 won the Nobel Prize for Economics. Economists also joined in, using large population surveys to throw light on the effect of different experiences upon individual wellbeing.

At the same time, ordinary citizens have become increasingly aware of their own mental states and interested in how to improve their own feelings. Two forces are at work here. The first is cognitive psychology (embodied in cognitive-behavioural therapy) and the second is mind-training techniques imported from the East. We discuss both at length in Chapter 4.

Wellbeing over time and WELLBYs

Of course, most decisions affect wellbeing over a period of time. For example, decisions affecting the climate will affect generations yet unborn. So we need a social welfare function that enables us to find the best path of wellbeing over the future and not just in the current period. So how should we value future wellbeing?

An obvious starting point is that the wellbeing of every human being in every year matters equally. However, since there is major uncertainty about the future, we should slightly discount the wellbeing that could be expected in future years – by multiplying wellbeing in t years hence by ${\left(1-\delta \right)}^t$ where δ is the discount rate per annum. (In Britain, the official value of this ‘pure time preference rate’ is 1.5% per annum).Footnote ¹³ In proceeding in this way, we are assuming that a person’s suffering in any one year is equally important, whether or not the rest of their life is happy.

On this basis, intertemporal social welfare (S*) becomes simply the discounted sum of all future wellbeing, whoever is experiencing it:

$\mathit{\text{Future Social Welfare}}=S^{\ast }=\begin{array}{c}\Sigma \\ i\end{array}\begin{array}{c}\Sigma \\ t\end{array}{\left(1-\delta \right)}^t{W}_{\mathit{it}}$(2)

So if we want to decide whether a policy change is desirable, we should evaluate whether the following expression is positive:

$\Delta S^{\ast }=\begin{array}{c}\Sigma \\ i\end{array}\begin{array}{c}\Sigma \\ t\end{array}{\left(1-\delta \right)}^t\Delta W_{\mathit{it}}$(3)

To make this practical, we have to decide on the length of each period. If we take it as a year, $\Delta W_{\mathit{it}}$ would be the change in wellbeing for person i in year t. In other words, it is a change in Wellbeing-Years (or WELLBYs). So when we come to methods of policy evaluation in Chapter 18, the key issue will be how a policy affects the number of (discounted) WELLBYs.

Sustainability and climate change

Thus the wellbeing approach provides a comprehensive framework for considering the future of our society from the smallest choices to the biggest. Of these, the biggest of all is the future of the planet. A central issue here is the wellbeing of future generations.Footnote ¹⁴

Wellbeing science favours a low discount rate, so that what happens to future generations is really important when we decide what to do now. The only legitimate reason for discounting future wellbeing is ‘pure time preference’, based on uncertainty about the future. Typical rates of pure time preference are 1.5% a year. By contrast economists mainly analyse the future in terms of levels of real income. They generally assume that real incomes will rise steadily, which will reduce the impact of extra future income on future wellbeing. They therefore discount future income by at least 3.5% a year – making the future appear much less important relative to the present.Footnote ¹⁵ So it is not surprising that those who want everything analysed in terms of wellbeing find strong allies among those who want more attention to the future of the planet. For both groups, sustainability is crucial.

Length of life and the birth rate

There is one further issue – the issue of life and death. One way to increase future social welfare (as in equation 2) is to help people live longer. If someone lives longer and has wellbeing greater than zero, that increases social welfare.Footnote ¹⁶

But can we also increase future social welfare by increasing the birth rate? If we could, it would be one of the least expensive ways of increasing future social welfare, and we would therefore choose to do it even if it decreased the number of Wellbeing-Years (or WELLBYs) per person born. John Stuart Mill rejected the idea that social welfare depended on the size of the population, and we shall focus instead on the number of WELLBYs per person born.

Today’s medical policy-makers have adopted the same approach. Their aim is to produce the highest possible number of Quality-Adjusted Life Years (suitably discounted) for each person born.Footnote ¹⁷ We shall follow this approach. So when we use equation (3) to evaluate a policy change, we do not include in our evaluation any effect of the policy change upon the number of people born.

How egalitarian should we be?

We have thus far adopted the stand of classic utilitarianism – that all that matters is average lifetime wellbeing, regardless of how unequally it is distributed. This means that the following distributions of wellbeing are of equal value:

But today many people are more egalitarian than Hutcheson and Bentham were and believe that it is more important to raise the wellbeing of the least happy people than to raise by an equal amount the wellbeing of those who are already quite happy.Footnote ¹⁸ In other words, the social value of i’s wellbeing is not identical to W_i. Instead, it rises with W_i but it rises at a declining rate. So the social value of W_i is a ‘function’ of W_i, f (W_i), which has the property we have just described.Footnote ¹⁹ And social welfare is the sum of f (W_i) added up across all members of society. Thus,

$\mathrm{For}\text{egalitarians}:S=\begin{array}{c}\Sigma \\ i\end{array}f\left(W_i\right)\left(f^{\prime }>0;f″<0\right)$(4)

There are thus two alternative concepts of social welfare that we could use to decide whether one situation is better than another. Which of them should we use? In choosing between the Benthamite (or strictly ‘utilitarian’) view and the more egalitarian (or ‘prioritarian’) view, a good approach is the one pioneered after WWII by Jan Harsanyi and John Rawls.Footnote ²⁰ In this approach, we imagine ourselves in an ‘original position’ behind a ‘veil of ignorance’, not knowing which actual human being we are going to be. We then ask ourselves to choose between different situations, not knowing who we will be in each situation.

So how would we evaluate different situations? Egalitarians who use this framework believe that most individuals have a degree of risk-aversion. They would not therefore evaluate a distribution of possible levels of wellbeing entirely according to its average value. They would also look at the spread of the possible outcomes. And they would prefer prospects where the probability of experiencing low wellbeing was low. In other words, for any given average wellbeing in the population, they would prefer a more equal to a less equal distribution of wellbeing.Footnote ²¹ The simplest way to represent this set of values is by a concept of social welfare of the type we have already described, $\begin{array}{c}\Sigma \\ i\end{array}f\left(W_i\right)$ where the social value of additional W declines as W increases.Footnote ²²

If one accepts this argument, the issue is then ‘How fast is the decline?’Footnote ²³ The most extreme view is one inspired by the work of the Harvard philosopher John Rawls.Footnote ²⁴ In this view, the decline is so sharp that the only thing that really matters is the wellbeing of the least happy person. So the social welfare equals the wellbeing of the least happy person.

$\mathrm{For}\text{followers of Rawls}:S=\mathit{Min}\left(W_1,\dots ,W_n\right)$(5)

A less extreme view is that the only thing that really matters is the number of people below some acceptable level of wellbeing.Footnote ²⁵

Choosing a social welfare function is not a scientific matter – it involves normative considerations, even if we try to solve the issue by ‘positive’ thought experiments or by surveying the population. This is why, however much the science of wellbeing improves, there will always be a spread of views on the exact definition of social welfare.

Therefore, the most practical approach in any choice situation is to begin by first examining the difference in $\begin{array}{c}\Sigma \\ i\end{array}{W}_i$ and then seeing how far the result would be altered by varying assumptions about the form of $f\left(W_i\right)$.Footnote ²⁶ A further practical step is to begin the search for new policies in those areas of life that account for the greatest amount of misery (on which Chapter 8 provides relevant evidence).

Consequentialism and rights: The fat man

The first criticism is that we only take into account the consequences of actions (even though we do include in that the experiences that occur during the action itself). Critics often use the following example.Footnote ²⁷ You are on a railway bridge. You see a train approaching from one side of the bridge, while down the line are five people on the track who will be killed by the train. However there is a fat man sitting on the bridge. Should you push the fat man to his death in order to save five others?

The calculus says there would be a net gain of four lives from this action (5 minus 1). So is it the right thing to do? Critics argue that the wellbeing approach says Yes, and this shows that the wellbeing approach is inadequate. However, how would millions of people in our society feel about their own lives if such actions became acceptable? Clearly, society has to have rules that make its members flourish. The philosophy of utilitarianism thus has two functions.Footnote ²⁸ First, it helps us to choose the rules we should generally follow, and second, it helps us to decide when we should break the rule (for example, when we should lie to protect a Jew in hiding from the Nazis).

Some of the rules will be moral principles we teach our children, while others will be rights enshrined in law. We will teach our children to be kind because that makes other people feel better. But we also need legislated rights, especially for minorities. If we wish to prevent misery, we have to establish many legal rights. But these are deliberate legislative acts designed to promote social welfare; they are not the recognition of some pre-existing ‘natural rights’.Footnote ²⁹ The only natural right is that each individual’s feelings count equally.

The experience machine

But are good feelings the only things that matter? In 1974, the Harvard philosopher Robert Nozick questioned the primacy of feelings by posing the following hypothetical question: ‘Suppose there was an “experience machine” you could link up to, which would make you feel anything you desire and which was equally available to other people. Would you link up to it?’Footnote ³⁰

Many people say No. They say it matters that the experience is real. But what if the real experience was awful and that we reverse the sequence of the question. So imagine you are having a lovely time in the experience machine and are offered instead the real experience of solitary confinement in a rat-infested cell. Which would you choose?Footnote ³¹

A more likely possibility is that scientists come up with a drug that makes everybody feel better and has no bad side effects. This issue was raised in a striking form by Aldous Huxley in his book Brave New World. In it he has people taking soma to make themselves feel better. This was meant to appal the readers. But most people throughout history have used alcohol or other substances to improve their mood. The problem has been that all known substances of this kind also bring bad side effects. But it may be hard to object to a substance that improves mood without any bad side effects.

Adaptation

Then there is the issue that our feelings adapt (see Chapter 3). Most people adapt to hardship to a considerable extent, so that it causes less misery than might be expected.Footnote ³² If we take this fact into account, critics say, we shall do less to reduce hardship than we should.Footnote ³³ But this does not follow, because people who are more fortunate also adapt to their good fortune. So if we take from the privileged and give to the deprived, the privileged will also suffer less than might be expected. Thus the case for redistribution is hardly affected by the fact of adaptation: the poor may gain less extra wellbeing from it than might be expected, but the rich also suffer less than might be expected.Footnote ³⁴ The balance of the argument is thus unaffected.

However, adaptation does have important implications if some experiences are less subject to adaptation than others. It is therefore important to distinguish between hardships that cannot be adapted to and those that can be. Hardships that are hard or impossible to adapt to include mental pain, chronic physical pain, incarceration, torture, indignity and intolerable noise.Footnote ³⁵ These types of hardship need high priority in public policy.

Selfishness is encouraged

Critics often complain that, if we accept wellbeing as the supreme goal, this means that individuals should simply maximise their own wellbeing. Far from it. For each person’s wellbeing depends hugely on the benevolent behaviour of others. This is illustrated in Figure 2.1.

Figure 2.1

How wellbeing is determined by behaviour

In this profoundly important diagram, the outcome we care about is the overall wellbeing of society, represented by the whole right-hand side of the diagram. How can it be maximised? The answer is: only if everyone chooses to behave so as to maximise the overall wellbeing of society. That is the fundamental ethical principle proclaimed by Bentham and by all other supporters of the greatest happiness principle. This aspiration should inform a person’s whole life – both in private matters and also in the job they choose to do. In this view, morality is not just about what you should not do but also what you should do. For example, it is wrong to hurt someone but it is also wrong not to help them.

Some people say this is too ‘demanding’Footnote ³⁶ – it takes over too much of your life. As with all other moral theories, it is hard to specify exactly how far we should sacrifice our individual happiness for the sake of others. But, if we want to reduce suffering, we clearly have to avoid not only hurting people (sins of commission) but also failing to help them (sins of omission).

To encourage good lives, we have two key psychological traits we can build on. The first is the pleasure that people derive from helping others: doing the right thing is not always pleasurable but it often is (see Chapter 3). The second is the impact of norms on habit. If people are expected to behave well from an early age, it just becomes a habit. This was the route that Aristotle stressed.

Both of these routes are at variance with the doctrines of Immanuel Kant, the great philosopher of the German Enlightenment, which provide the main alternative system of ethics to that based on the maximisation of wellbeing.Footnote ³⁷ In the Kantian system, the ethical rightness of a person’s actions are judged by the person’s motives and not by their consequences. A moral action has to be a conscious act of will, and pleasure and habit are considered antithetical to truly moral action. There is, however, one key feature on which both Kantian and Benthamite approaches are agreed: every individual is of equal ultimate importance. This is fundamental to any ethical theory based on reason.

The nanny state

Turning to political theory, critics of the wellbeing approach sometimes claim that, if the government concerns itself with how people feel, this will lead to excessive interference in people’s lives. But wellbeing science itself shows the huge importance of freedom for personal wellbeing (see Chapters 8 and 16). So any government aiming at wellbeing will be constantly restrained by that consideration. Whenever possible, the government will provide opportunities that people can use or not, as they prefer.

It is, however, crucial to realise that many of the things that matter most to people are intensely personal (their mental health, their family relationships and their work situation) and failures in these dimensions are major sources of human misery. So a benevolent state is bound to offer help, if there are cost-effective forms of help that can be made available.

Social justice

A final problem is one that affects all ethical theories – how much weight to give to the interests of the least fortunate. Classic utilitarianism can be criticised because it gives equal value to more happiness whether it accrues to someone who is already happy or to someone who is miserable. But the ‘prioritarian approach’ we advocated earlier avoids this problem, and it provides a stronger basis for the legal creation of rights. It also asserts that income inequality is bad because an extra dollar is worth more to a poor person than a rich one. So a good society establishes rights (as a form of safety net) and it redistributes income.Footnote ³⁸

No ethical theory is without problems. If you find the wellbeing approach problematic, can you think of a better criterion for how to live or how to make policy?

We can end with a different question. Is wellbeing an experience that humans were designed for? The answer from ‘evolutionary psychology’ is ‘Partly yes and partly no’. The basic features of human nature were created by natural selection about 200,000 years ago in Africa. The genes that survived in the struggle for survival were those that maximised our ‘inclusive fitness’ – in other words, the genes that were most likely to produce successive generations carrying the same genes. So how conducive to wellbeing were the genes that got selected? We were certainly constructed to enjoy many of the things necessary for survival and reproduction – sex, food, drink and a capacity to cooperate (see Chapter 3). Moreover, as we have seen, wellbeing is extremely good for many other things that are good for our survival – good for our physical health, for our productivity and for our creativity.Footnote ³⁹

There were, however, some traits that were essential for survival in the Savannah that are not particularly conducive to an enjoyable life. The most obvious of these is anxiety. If lions are about, it is a good thing to be anxious. But life today is a lot more safe, and most people would have a more enjoyable life if they were less anxious (see Chapter 4).

As this book shows, we have the knowledge to improve our own wellbeing and that of others – a noble cause.

Questions for discussion

1. (1) Do you agree that coherent policy-making requires some single overarching criterion? Should this be wellbeing?

2. (2) Should the same idea govern our personal ethical choices?

3. (3) Should the social welfare function be the simple sum of wellbeing across people? Or should it give extra weight to the scale of misery?

4. (4) What is your view on the main criticisms made of the wellbeing approach to ethics and policy choice:

   • Consequentialism, rights and social justice

   • The experience machine.

   • Adaptation

   • Selfishness

   • Nanny state