Most policy-makers have multiple objectives. But in the end, they have to decide whether one policy is better than another. Policy A may have a bigger impact on one objective and Policy B may on another objective. How to choose between the policies? The decision-maker must implicitly weight the importance of one objective compared with another. She must have some view about which objectives are most important. It would be much better to make this explicit by having and applying a single overarching criterion, against which to judge the merit of different outcomes.
That criterion should be the wellbeing of the people. There are of course many good things – health, wealth, freedom and so on. But for each of these goods we can ask ‘Why are they good?’ and expect an answer. For example, health matters because without it people feel lousy. Similarly with wealth, freedom etc. But if we ask ‘Why does it matter how people feel?’, we can give no answer. It self-evidently matters, which is why the happiness of the people is the most obvious candidate for the overarching good.
The great philosophers of the 18th century Anglophone Enlightenment came up with the most obvious answer: the ultimate criterion is the happiness of the people – how they feel about the quality of their life as they themselves experience it . As Thomas Jefferson put it “The care of human life and happiness… is the first and only object of good government.”1 So a government committed to “building back better” would wish to build back happier, and one committed to “levelling up” would want to level up wellbeing and not just economic opportunity.
Many policy-makers have problems with the word happiness and instead prefer the term wellbeing. That is fine, provided we are clear that it is the people’s wellbeing as they themselves judge it- not as some researcher or civil servant evaluates it. In other words, we are talking about “subjective wellbeing”. The typical way of measuring this is to ask “Overall, how satisfied are you with your life these days?” (0 = very dissatisfied, 10 = very satisfied). This question elicits very similar answers to the question “Overall how happy are you with your life these days?” So when we talk about wellbeing, that is what most policy analysts have in mind.
Fortunately, more and more policy-makers in OECD and elsewhere now consider that policy should be targeted at wellbeing. In 2020 the EU Council of Ministers urged EU countries “to put people and their wellbeing at the centre of policy design” [2, 3 p76–79]. In other words the aim of policy must be to create conditions for the greatest possible wellbeing.
But this approach can only be implemented if we know what causes wellbeing. Until recently there was virtually no quantitative information on this subject, which is why the 18th century ideal could only have been implemented crudely. But over the last forty years a whole new science of wellbeing has developed [4, 5], which now tells us enough about the causes of wellbeing for this to become the stated objective of policy.
But why would policy-makers want to maximise the wellbeing of society? Isn’t their aim to be re-elected? Indeed, it is bound to be. But recent research shows that the best way for a government to be re-elected is to maximise the wellbeing of the people. A study of national elections in European countries from 1974 onwards  found that the best predictor of the government’s vote-share in national elections was the life-satisfaction of the people. The decisive factor is not, as a Clinton aide once said, “the economy, stupid”. If we look, country by country, at the variation of life-satisfaction from one election to another, one extra standard deviation of life-satisfaction gives the government an extra 6 percentage points of the popular vote. By contrast, one standard deviation of economic growth gives only 3 extra percentage points of the vote (see Figure 1). So politicians who target the people’s wellbeing increase their chances of remaining in – or gaining – power.
Thus, it is not pie-in-the-sky to advocate (on ethical grounds) that policy should target wellbeing. Not only is it ethical, but it makes political sense, and is operationally practicable thanks to the advances in science.
So how exactly would a policy-maker choose the priorities for spending? We have to assume that the total volume of public expenditure is set by political forces. The task is therefore how to spend this total in the way that produces the most wellbeing. That means choosing those policies which produce the most wellbeing per pound spent. There would be some cut-off value for the cost-effectiveness of policies, with policies only qualifying if their ratio of wellbeing-benefit to cost exceeds the cut-off. Similarly, the redistribution of income would only proceed until further redistribution began to reduce total wellbeing. And regulations would only be introduced when this would increase total wellbeing.
This approach is less revolutionary than it might appear. It has in fact been standard practice in the health field in many countries. Health states are evaluated for their quality-of-life (on a scale of 0–1) and medical treatments are evaluated in terms of their impact on quality-of-life-adjusted life-years (or QALYs). They are only approved if they produce enough QALYs per pound spent.2 The wellbeing approach is essentially an extension of this method.
There are of course important differences. Wellbeing is how people feel about their whole lives, not just their health. And we are looking at the effects of every aspect of policy, not just healthcare.
Some of these effects are economic. So how does the wellbeing approach differ from traditional cost-benefit analysis, where benefits are measured in units of money? Unlike the traditional approach, the wellbeing approach can cover the whole range of public expenditure. By contrast, traditional cost-benefit analysis can only be applied over a narrow range of issues where the benefits either have an actual price, or a value which is implicit in the choices people make. This condition is not satisfied in most of health, social care, child protection, law and order, the environment and redistribution. Indeed, the reason the state is active there is precisely because in these areas market valuations and outcomes would be sub-optimal. So in these areas there is really no alternative to wellbeing (directly measured) as the criterion of benefit.
However, traditional cost-benefit is a totally valid way of measuring benefits in those areas where it can be applied. So the two approaches are complementary and they can be combined by transforming the money measures of benefit (derived from traditional CBA) into wellbeing measures by multiplying them by the marginal impact of money on wellbeing.3 Fortunately the British Treasury’s Green Book manual of policy analysis now endorses ‘social wellbeing’ as the goal and approves the use of direct measures of wellbeing as well as their monetary equivalents [7, 8].
We shall turn to the evidence on wellbeing shortly, but at this point we have to confront a difficult issue. Is total wellbeing really the goal? Or should we not pay more attention to the prevention or the relief of misery? In other words, is it more important to raise the happiness of someone who is miserable than that of someone who is already happy?
Jeremy Bentham, the founder of the wellbeing approach, opted for the total sum of wellbeing as the goal. But many modern thinkers would take a more egalitarian stance . They argue that it is more important to increase the happiness of those who are more miserable than to increase the happiness of those who are already happy.
There are two practical ways of implementing this more egalitarian approach. One is to measure social welfare not by total happiness, but in a way that gives less value to additional happiness the happier a person is.4 So when policies are being analysed, their value would be subject to sensitivity analysis to see how their comparative claims change as the analysis becomes more egalitarian.
Another, more practical approach, is to focus the search for new policies more heavily on those areas of life which account for the greatest amount of misery in society. This, in essence, is what the New Zealand government has done in its wellbeing budget from 2019 onwards.
That brings us directly to the evidence base for selecting the areas for new policy development and for evaluating them [5, 10]. What are the main determinants of wellbeing? And what are the main causes of misery? Our team at LSE recently analysed the findings from major longitudinal surveys in Britain, Germany, Australia and the US. The findings were similar in all these countries, and Figure 2 gives the results for Britain. It shows how each factor – mental health, physical health, employment, quality of work, relationships, income and education – contributes to the inequality of wellbeing, holding the other factors constant. A parallel analysis shows how much each factor contributes to the prevalence of misery, and the ranking of factors is the same for each analysis [5 Table 16.1].
Thus, as Figure 2 shows, more of the misery in our country is due to diagnosed mental illness than to any other factor – and physical illness is also important. Next come human relationships – at work and in the family – and only then comes income. This finding is repeated in country after country. So we need a new, broader concept of deprivation – the inability to enjoy life for whatever reason, rather than just because of poverty.
This view of priorities is found by relating differences in people’s wellbeing to differences in their situation (see Figure 2). An alternative approach to priorities is to simply ask people “How much do you worry about the following issues (0 – never, 10 – a lot)?”. The results for a representative UK sample are shown in Figure 3. They broadly confirm the ranking of priorities shown in Figure 2.
Many of our adult characteristics are laid down in childhood, which makes it pertinent to ask which aspects of child development best predict a satisfying adult life? The answer is that emotional health at 16 is a better predictor of a happy adult life than all the qualifications a person ever gets [5 Fig.1.2]. But how can we influence a child’s emotional health? The evidence is striking: primary and secondary schools (and their teachers) affect the emotional health of children as much as their parents do [5 Fig.1.5(b)].
The policy implication of all this is clear. Policy-makers should give much lower priority to long-term economic growth and much higher priority to the services which sustain mental health, physical health, child development, family life and elderly care. It is the social infrastructure which matters most, not the physical infrastructure (as so many politicians wrongly seem to assume). To level up those areas which are left behind it requires better services more than better economic infrastructure. And our education system should teach more than the skills of earning a living – it should teach the skills of life.
But precisely how should we spend the money for these important aspects of life? We can only know what works by the process of experiment. It is not enough to address the right problem and to have good intentions. The next step in advancing wellbeing is through conducting thousands of experiments to discover which policies make the most difference to wellbeing per pound spent. Let me give a few examples.
In British schools there have been a series of attempts to teach life skills. The last Labour Government introduced a programme called Social and Emotional Aspects of Learning (SEAL). Its impact was evaluated in secondary schools and found to be zero – no effect on emotional, behavioural or academic outcomes . The reason was identified as insufficiently structured materials and a lack of teacher training. By contrast, a more recent 4-year weekly curriculum in secondary schools called Healthy Minds was found to raise the student’s life-satisfaction by 0.4 points (out of 10). If we convert this into a measure of quality-of-life, the cost per extra QALY was only £1,000 – well below the standard criterion of around £25,000 for additional health expenditure [13, 14].
Turning to adults, hundreds of clinical trials of modern psychological therapy show 50% rates of recovery for depression or anxiety disorders after an average of some 10 sessions. They also show that the patients treated will work on average one additional month over the next two years as a result of the treatment. This generates enough additional income to cover the cost of the therapy. On this basis, Layard et al  proposed a programme of Improved Access to Psychological Therapy. When implemented, the results of the trials were repeated in the field [16, 17].
Clearly the same experimental approach should be used by voluntary organisations as by public policy-makers. For example when the Action for Happiness movement’s Exploring What Matters Couse was evaluated, it was found to increase life-satisfaction by over 1 point (out of 10) .
One general point needs to be made about social experiments. Even if the experimenter’s main target is not wellbeing, wellbeing should always be measured throughout the experiment.
It is of course important to evaluate not only new policies but also existing policies in terms of their cost-effectiveness. This is not easy, and I offer Table 1 in the hope that others will improve on it. In it I examine the cost effectiveness of different ways of reducing by one the number of people in misery over a twelve-month period. According to this analysis, the cheapest of the four methods is treating more people for depression and anxiety disorders (this analysis ignores the flow-back of savings). Active labour market policies come second – encouraging evidence of what we might expect from the current government’s Kickstart initiative. Then comes physical health, and finally come income transfers to the poor. We could alternatively do the analysis in terms of the impact on total wellbeing (rather than on misery) but the results would be very similar. Many of the coefficients are already known, but others will need to be determined [10, 19]. And a major exercise will be needed on the costs and cost-savings of each possible policy.
|£k per year|
|Poverty. Raising more people above the poverty line||180|
|Unemployment. Reducing unemployment by active labour market policy||30|
|Physical health. Raising more people from the worst 20% of illness||100|
|Mental health. Treating more people for depression and anxiety||10|
Either way, this analysis is very crude. But, as time passes, it will become possible for Finance Ministers to do a much better job. For this they will need
The second of these is a major project – just as it was in previous times a major project to develop models of the macroeconomy and of the impacts of policy changes upon the finances of households. This new project is equally important and will need significant money to finance it.
One further point on modelling. Many policies affect the length of life. So our measure of the impact of policies needs to take this into account. How? We want people to have lives which are long and full of wellbeing. So the simplest approach is to say that we want for each individual the maximum total wellbeing-years, where we simply add up the wellbeing in each year of their life. A natural acronym for wellbeing-years is WELLBYs, (just as medics talk of QALYs meaning quality-of-life-adjusted life years). So we wish that each life will have the largest possible number of WELLBYs. (If the effects are distant – a pure social time-preference rate of say 1.5% a year needs to be applied).
This approach has huge implications for policy, especially for policies involving safety and the risk of death. At present the value of life in terms of money is derived from one of two methods:
These methods involve major assumptions. By contrast, the wellbeing approach is very simple: it simply examines the change in WELLBYs . And it yields a very different trade-off between money and life-years from that implied by traditional methods. In the wellbeing approach an extra year of life is of equivalent value to up to £750,000 .6 By contrast, existing methods yield values well below £100,000.
Which approach is the more plausible? Traditional values would not justify a lockdown to save lives threatened by COVID-19, while the wellbeing approach would . And public opinion supported the lockdowns. So the wellbeing approach would seem to be in tune with public opinion. Thus it does seem that future policies should give more weight to the preservation of life relative to other objectives – compared with what happened before COVID. This does not mean an increase in public expenditure, which we take as given. But it does mean a rebalancing.7
We have so far focused single-mindedly on wellbeing as the overarching good. We need recruits to this view! But, even if you do not buy that, you should take wellbeing very seriously because of its good effects on other things you value. So here are some important facts.
So, if a government wanted to implement the wellbeing approach, what new procedures would it need? Bids for public money would need to be justified by their effects on wellbeing, and this would apply to current as well as to capital expenditure. Officials would need help in making these estimates. So the Finance Ministry would need to include a Wellbeing Appraisal Group to help train departments in how to make their proposals – and then to vet them when they arrive. At local level there would be Local Wellbeing Agreements between local authorities and local social services about how to raise the wellbeing of residents. There would be an annual report to Parliament on how the nation’s wellbeing had developed – and how government policy had impacted on it.
The wellbeing approach is not new. Its adherents have included William Beveridge, architect of the British welfare state and Sidney and Beatrice Webb, the founders of the London School of Economics.
But now its time has really come. There is an explosion of articles (academic and popular) on both wellbeing and mental health (see Figure 4). More and more universities around the world are teaching the subject, producing a body of trained analysts able to apply these ideas to policy. The OECD have persuaded all member countries to measure the wellbeing of their people . The governments of five countries have formed an alliance called the Wellbeing Economy Governments partnership (WEGo).10 So in future years we shall surely see a major change in policy-making. The common currency will become not money, but wellbeing.
6This depends on the value of the income coefficient in the wellbeing equation. This can be quite low. For example if In a typical wellbeing equation, Wellbeing = 0.3 log Income,. So dIncome/dWellbeing = Income/0.3 = say £100,000. Since the typical life-year provides 7.5 WELLBYs it would therefore be of equivalent value to £750,000.
10Scotland, Iceland, New Zealand, Wales and Finland. https://weall.org/.
The author has no competing interests to declare.
Bentham J. An Introduction to the Principles of Morals and Legislation. 1996 ed. Oxford: Clarendon Press; 1789. DOI: https://doi.org/10.1093/oseo/instance.00077240
Clark AE, Flèche S, Layard R, Powdthavee N, Ward G. The Origins of Happiness: The Science of Wellbeing over the Life Course. Princeton University Press; 2018. DOI: https://doi.org/10.1515/9780691196954
Ward G. Happiness and Voting: Evidence from Four Decades of Elections in Europe. American Journal of Political Science; 2020 (forthcoming). DOI: https://doi.org/10.1111/ajps.12492
Parfit D. Equality and Priority. Ratio. 1997; 10(3): 202–21. DOI: https://doi.org/10.1111/1467-9329.00041
Frijters P, Krekel C. A Handbook for Wellbeing Policy-Making: History, Theory, Measurement, Implementation, and Examples. UK: Oxford University Press. 2021; 464. DOI: https://doi.org/10.1093/oso/9780192896803.001.0001
Humphrey N, Lendrum A, Wigelsworth M. Social and emotional aspects of learning (SEAL) programme in secondary schools: national evaluation. London: Department for Education; 2010. DOI: https://doi.org/10.1037/e607192011-001
Layard R, Bailey L, Coleman J, Judge E. Healthy Minds. A four-year course in secondary schools. In: LSE (ed.). http://wwwhealthymindsinschoolsorg/wp-content/uploads/2017/11/HealthyMinds-Pamphlet-181130pdf. London UK2018.
Layard R, Clark DM, Knapp M, Mayraz G. Cost-benefit analysis of psychological therapy. National Institute Economic Review. 2007; 202: 90–8. DOI: https://doi.org/10.1177/0027950107086171
Clark DM. Realising the Mass Public Benefit of Evidence-Based Psychological Therapies: The IAPT Program. Annual review of clinical psychology; 2018. DOI: https://doi.org/10.1146/annurev-clinpsy-050817-084833
Frijters P, Clark AE, Krekel C, Layard R. A happy choice: wellbeing as the goal of government. Behavioural Public Policy. 2020; 4(2): 126–65. DOI: https://doi.org/10.1017/bpp.2019.39
Dolan P. Thinking about it: thoughts about health and valuing QALYs. Health economics. 2011; 20(12): 1407–16. DOI: https://doi.org/10.1002/hec.1679
Layard R, Oparina E. Living Long and Living Well: The WELLBY Approach. In: De Neve J, Helliwell J, Layard R, Sachs J (eds.), World Happiness Report 2021. New York: UN Sustainable Development Solutions Network; 2021.
Durlak JA, Weissberg RP, Dymnicki AB, Taylor RD, Schellinger KB. The impact of enhancing students’ social and emotional learning: a meta-analysis of school-based universal interventions. Child Development. 2011; 82(1): 405–32. DOI: https://doi.org/10.1111/j.1467-8624.2010.01564.x
Fredrickson BL, Branigan C. Positive emotions broaden the scope of attention and thought-action repertoires. Cognition & emotion. 2005; 19(3): 313–32. DOI: https://doi.org/10.1080/02699930441000238
Steptoe A, Wardle J. Enjoying life and living longer. Archives of Internal Medicine. 2012; 172(3): 273–5. DOI: https://doi.org/10.1001/archinternmed.2011.1028
Lee H, Singh GK. Inequalities in Life Expectancy and All-Cause Mortality in the United States by Levels of Happiness and Life Satisfaction: A Longitudinal Study. International journal of MCH and AIDS. 2020; 9(3): 305–15. DOI: https://doi.org/10.21106/ijma.392
Bellet C, De Neve J-E, Ward G. Does Employee Happiness have an Impact on Productivity? Saïd Business School WP. 2019; WP 2019–13. DOI: https://doi.org/10.2139/ssrn.3470734
Idstad M, Torvik FA, Borren I, Rognmo K, Røysamb E, Tambs KJBPH. Mental distress predicts divorce over 16 years: the HUNT study. BMC Public Health. 2015; 15(1): 1–10. DOI: https://doi.org/10.1186/s12889-015-1662-0