There is a paradox at the heart of our civilisation. Individuals want more income. Yet as society has got richer, people have not become happier. Over the past 50 years we have got better homes, more clothes, longer holidays and, above all, better health. Yet surveys show clearly that happiness has not increased in the US, Japan, Continental Europe or Britain.
By happiness I mean feeling good - enjoying life and feeling it is wonderful. And by unhappiness I mean feeling bad and wishing things were different. Most people find it easy to say how good they are feeling, and in social surveys such questions get 99 per cent response rates - much higher than the average.
What matters is how people feel on average over a longish period. But it is also interesting to study how their moods vary not only from year to year, but also from activity to activity, day to day, even hour to hour. For example, a study of 1,000 working women in Texas discovered how feelings change as the day goes on: as the graph (right) shows, people feel better as time passes, except for a dip after lunch. The study also found which activities the women liked most and which the least (see table overleaf), and confirmed that people are much happier when they are with others than when they are alone.
But can we rely on what people say? Do different people use words in the same way? With the aid of neuroscience, we now know that they do. Positive feelings (for right-handed people) correspond to brain activity in the left side of the pre-frontal cortex, negative feelings to the same place on the right side of the brain. Neuroscientists find that people who have the "happier" brain readings also say they are happier. So the idea, common among cynics, that we cannot know what other people are feeling - and that, unlike national income, happiness cannot be measured - is simply false. It is entirely practicable to make happiness our goal.
It is true that, within any particular society at any particular moment, rich people are on average happier than poorer ones. For example, 41 per cent of people in the top quarter of incomes are "very happy", compared with only 26 per cent of those in the bottom quarter of incomes. The problem is that, over the years, the proportions in each group who are very happy have not changed at all although the real incomes in each group have risen hugely. This is true of all the main western countries.
Is it possible that people have changed the meaning of the words over time? This is unlikely to have happened similarly in every country. We also know that clinical depression, assessed professionally through population surveys, has risen in most countries. A survey from London University's Institute of Education, out this month, shows that as many as 29 per cent of women aged 30 in 2000 reported suffering trouble with nerves or feeling low, depressed or sad; the comparable figure in a similar survey, among those aged 36 in 1982, was just 16 per cent. Researchers disagree over the size of the increase, but nobody believes depression has diminished, despite the much greater ease of our material life.
Further evidence comes from comparisons between different countries. These show that, where average income per person is less than $15,000 a year - in other words, where many people are near the breadline - extra money really does make people happier. But comparing countries where average income is above that level, happiness seems to be independent of income. For example, the average American is much richer than the average Icelander or Dane, but also less happy.
Why is this? Clearly people are comparing their income with some norm and this norm is rising all the time. Thus from 1946-86, the US Gallup poll asked people, "What is the smallest amount of money that a family of four needs to get along in this community?" It turns out that, as actual average incomes rose, so did the income that people felt was needed - and in fact this "needed" income grew in direct proportion to actual income. Likewise, when people were asked "Are you satisfied with your financial position?", the proportion who said they were "pretty well satisfied" fell, despite enormous economic growth.
Two things drive up the norm with which people compare their incomes. One is the income that they themselves have experienced - which habituates them to higher standards of living. And the other is the income that others get, and which they try to rival or outdo.
Habituation is a basic psychological phenomenon. It works both up and down: you adjust to good things and to bad. The clearest evidence that you adjust to income comes from asking people with different levels of actual income what income they would consider satisfactory. Typically, the income that people say is satisfactory rises by almost 50p for every extra pound that they have actually acquired. A whole range of studies shows that people adjust their requirements to their recent experience and that they are constantly surprised by this. People overestimate the extent to which the new house or new car will, once they have got used to it, make them happier.
People also adjust their requirements in response to what other people have: keeping up with (or trying to outdo) the Joneses. In a poor society, a man proves to his wife that he loves her by giving her a rose, but in a rich society he must give a dozen roses.
When people compare their wages, it is generally with people close to themselves, rather than with film stars or paupers. What matters is what happens to your "reference group" because what your reference group gets might have been feasible for you, while what David Beckham gets is not. Much of the most intense rivalry, therefore, is within organisations and within families. In organisations, calm can often be maintained only by keeping people's salaries secret. In families, it has been found that the more your spouse earns, the less satisfied you are with your own job. And among women, if your sister's husband is earning more than your own husband earns, you are more likely to go out to work.
Change the reference group and you may well change levels of happiness. In the eastern part of Germany, the living standards of those in work have soared since 1990, but their level of happiness has plummeted. Why? Because the east Germans now compare themselves with west Germans rather than with other countries in the old Soviet bloc. Again, women's happiness in the US has fallen relative to men's as their pay and opportunities have improved. The most likely explanation is that women now compare themselves more directly with men than they used to, and therefore focus more on the gaps that still exist.
But we do not find the same effects when we ask people about non-monetary benefits. A sample of Harvard graduate students was asked which of two worlds they would prefer: one where they got $50,000 a year and others got $25,000, or one where they got $100,000 and others got $250,000. The majority preferred the first world. They were happier with lower income, provided their relative position improved. Then the students were asked to choose between a world where they got two weeks holiday while others got half that and a world where they got four weeks holiday while others got twice that. Most preferred the second world. In other words, people are much less rivalrous when it comes to leisure. The rat race is for income and when each of us works more and earns more, this imposes a genuine loss of happiness on others. It is a form of pollution.
So now we can see why happiness increases so little when countries get richer. People get hurt as their needs rise in ways they did not foresee - a form of self-pollution. And they get hurt by the extra income that others are earning - pollution by others.
The policy implications are dramatic. For example, if much of the extra income (say 60p in the pound) brings no overall increase in happiness, we should reduce the incentive to acquire it. It would therefore be efficient to have a marginal tax rate of say 60p in the pound - corresponding (on the principle that the polluter should pay) to the 60p worth of pollution caused by the extra pound that is earned. Up to now we have apologised for taxation. The standard economic analysis says that taxation reduces work effort, which is true. But it also says that it is inefficient to reduce work effort, and our analysis shows that, if the aim is to increase human happiness, this is false.
Indeed taxation is one of the most important institutions we have for preserving a sensible balance between work and leisure. We should be proud of it and stand up for it. As it happens, 60 per cent is the typical level of marginal taxation in Europe if you allow for direct and indirect taxes. I suspect that, in some almost unconscious way, the electorate now understands that the scramble to spend more is partially self-defeating and that this explains why people are more favourable to public expenditure. But the time is ripe to make the argument explicit - as a central feature of social democracy or, dare I say it, the Third Way.
There are many other lessons that can be drawn from happiness research. Take performance-related pay. If the output of one individual can be easily measured, it is obviously sensible to pay him for what he produces. The problem arises when the worker's performance cannot easily be monitored, except by himself. Employers and supervisors then start to compare workers with their mates. This can motivate those who are likely to succeed, but demotivate those who are not, by giving constant reminders of how they are not appreciated.
Promotion systems always involve treating people differently at periodic intervals. But promotion is justified by the need for hierarchy. Performance-related pay is different and, being more frequent, it increases the salience of social comparisons. The evidence is that on balance it increases average performance, but reduces average morale - which also matters. Performance-related pay has been spreading in the public sector, and stress in the public sector has grown markedly, according to Eurobarometer surveys.
Status consciousness is deeply ingrained in our genes, as in those of our primate ancestors: remarkably, people who win Oscars live four years longer than people who are nominated but do not win. But we should not turn every human being into a racehorse jockey. The goals we want most people to perform are multifarious and are becoming increasingly so. We should therefore work harder on training and developing professional norms, and less hard on tuning up financial incentives.
And what does happiness research say about equality? It confirms the oldest and most obvious argument for redistribution - that it takes a pound from someone who values it little and gives it to someone who values it more. It is therefore a major way of increasing total happiness - and the poorer the recipient, the more effective it is. So development aid to the third world (if well spent) comes high on the list of public spending, as does child poverty at home.
But if, as the research suggests, money isn't everything, in what other ways can we increase happiness? The research points to six main factors: mental health, satisfying and secure work, a secure and loving private life, a secure community, freedom, and moral values.
It is a complete scandal that we spend so little on mental health. Mental illness causes half of all the measured disability in our society and, even if you add in premature death, mental illness accounts for a quarter of the total impact of disease. Yet only 12 per cent of the NHS budget goes on it. Roughly 25 per cent of us experience serious mental illness during our lives, and about 15 per cent experience major depression. Such depression can in most cases be helped by a combination of drugs and cognitive therapy. Yet only a quarter of people now suffering from depression are being treated, and most of them just get pills from a non-specialist GP. If we really wanted to attack unhappiness, we would totally change all this, and make psychiatry a central, high-prestige part of the NHS.
For most people, however, valued personal relationships with family, colleagues, friends and neighbours are the best guarantee of happiness. That is why unemployment is such a horror. Every happiness study reveals the horrible pain of unemployment, and estimates that the psychic pain from rejection is greater than the pain coming from loss of income. Unemployment offends against a person's basic need to be needed. Therefore we should support the policy of welfare-to-work, even when the job is less than perfect.
But even if you have work, you are much happier if you feel the job is secure. For this reason Continental European laws make it more difficult to sack people than it is in Britain or the US. Is this a terrible mistake, as the apostles of "flexibility" make out? Employment protection laws do not cause higher unemployment. (Employment protection is quite high in many European countries that have lower unemployment than we have, and the high unemployment in France and west Germany is due to a lack of welfare-to-work.) Employment protection may cause lower productivity, but people clearly value it. Rather than focusing on flexibility of employment, we should focus more heavily on wage flexibility, as rigid wages are the main cause of high unemployment in east Germany, southern Italy and southern Spain. If wages were more flexible, work would soon flow to these regions.
This would be a much better policy than to have many more people moving around the country, as is often advocated. Happiness research reveals that geographical mobility has major costs. Crime is directly related to the level of mutual trust in a community, and trust is lower the more people are moving house and the more heterogeneous in origin are the people living in any area. (This is a very uncomfortable fact, but highly relevant to the British debate on immigration.) Similarly, mental illness is more common if you live in an area where your group is in the minority than if you live where your group is in the majority. (This is highly relevant to policies for dispersing migrants.) Migration may also affect the stability of marriage.
Stability is what people like, and it is good for happiness and trust. It should not be denigrated. Yet the Anglo-American elite glorify novelty. Nothing is good unless "innovative". Civil servants gaily reorganise every public service, oblivious of how each reorganisation destroys a major channel of personal security and trust. We have a lot to learn from "old Europe", where the ideology of change has taken less hold.
Finally, there is the question of values. People are happier and better able to function when they feel they can trust other people. Unfortunately, in Britain, there has been a precipitous fall in the proportion of people who say "yes" when asked: "Would you say that most people can be trusted?" (The opposite proposition is "Would you say that you can't be too careful in dealing with people?") The proportion of trusters has fallen steadily from 56 per cent in 1959 to 31 per cent in 1995. A similar fall has occurred in the US, though not in Continental Europe. But even on the Continent, there is evidence of moral decline. People were asked whether they disapproved of those who kept an object they had found. In 1969, only 12 per cent accepted this practice; in 1990, the proportion was 32 per cent. It is said that people always think the world is going to pot. This view is empirically wrong. Americans have been regularly asked whether they think people live "as good lives - honest and moral - as they used to". In 1952, as many thought yes as no; by 1998, only a quarter thought yes. So something has changed: people feel less bad about pursuing their own interests regardless of others.
We do not fully understand why this has happened, but it is vital to counteract it. We should discuss it openly and not apologetically, and in schools we should surely offer a structured education in morals. "Citizenship" can certainly cover such issues as voting, but its focus should be the necessity of helping others and the pleasure that can give. It should also cover the necessity of enjoying whatever you have, without obsessing about what you lack.
Yes, it is difficult to produce a happier society, partly because of the strong role that our genes play in determining our temperament: there is a high correlation in the happiness levels of identical twins, even when reared apart. But there is still plenty of room for nurture as well as nature - just as people are now inches taller than a century ago, despite the strong heritability of height.
The scientific study of happiness is only just beginning. It should become a central topic in social science. But for the moment I would recommend four principles. Don't apologise for taxes; foster security; fight glaring evils such as depression; and discourage social comparison.
Richard Layard is director of the Centre for Economic Performance at the London School of Economics. This article is based on his Lionel Robbins Lectures, to be given at the LSE on 3, 4 and 5 March. The full text will be available from 6 March on http://cep.lse.ac.uk