June 6, 2007
Progressive taxation and the route to happiness
Happiness is fashionable these days. Yet should we accept the common view that the new “science” of happiness has cemented the superiority of Scandinavian social democracy over Anglo-Saxon liberalism? The answer is: No. The results are just as destructive to the pious certainties of “progressives” as to those of their opponents. Richard Layard of the London School of Economics and the UK’s House of Lords produced an elegant, brief and influential exposition of the new doctrine two years ago. That doctrine itself, as he explains, is a modern reincarnation of Jeremy Bentham’s utilitarianism*. What is Professor Layard arguing? First, happiness is the sole goal of human activity. Second, happiness is measurable. Third, we know what makes people happy and unhappy. Finally, policy should aim at achieving the greatest happiness. We will then realise that “there is more to life than prosperity and freedom”. Happiness is the right goal, he argues, “because it is our overall motivational device”. Moreover, “unlike all other goods, it is self-evidently good”. Yet if aggregate happiness could be maximised at the expense of a minority, should we do it?
The remainder of Martin Wolf’s column can be read here (FT.com subscription required). Discussion from our guest economists is free.
Andrew Oswald: Although Martin Wolf says sensible things, at some points his argument is incomplete or wrong.
First, as Martin Wolf shows, reported happiness has run flat in the rich countries over the last few decades. That graph could usefully be pinned up in every prime minister’s and president’s office. Although not explained in the article, in some nations, such as Great Britain and the Netherlands, we know that formal measures of mental health have worsened over time. These findings are primarily due to Richard Easterlin of the University of Southern California, a number of Warwick university economists, and David Blanchflower of Dartmouth College. To my knowledge, there is no dispute about these as facts. But, appropriately, their interpretation is now being debated in various academic disciplines, and more recently in the press. Richard Layard’s popular book has been valuable in disseminating the ideas of happiness economics, but that does not mean that researchers in the field agree with all his views.
Second, Martin Wolf mentions that, to quote, he sees as far from persuasive the philosophical and scientific underpinnings of this line of research. He does not give his reasons. What we do know is that anonymous referees with enough PhDs certificates to wallpaper every floor of the FT building in London have checked the countless ‘happiness’ articles that have appeared recently in economics, psychology and epidemiology journals. The outright rejection rate in these journals is about 90 per cent and it is normal to take five years from beginning on a project to seeing it appear in a refereed journal. All this refereeing does not mean the science is correct; academics often wander down paths that turn out to be muddle-headed. But it does mean the science is the best we have in 2007, and perhaps that should be respected a little more.
Third, almost all researchers would agree with Martin Wolf that the happiness literature has not proved the case for progressive taxation. The best statement of exactly that was presented by David Weisbach last week in a conference at the University of Chicago. The same conference saw a remarkable paper by Betsey Stevenson and Justin Wolfers, both of the Wharton School at the University of Pennsylvania, in which they document an apparent decline across western society in women’s happiness relative to that of men. This fact is important and little-known.
Fourth, the happiness literature should not be treated as something to buttress prior political views. Researchers who see red and blue flashing from their computer screens are not likely to be ones we should trust.
Posted by: Andrew Oswald | June 6th, 2007 at 1:49 pm | Report this commentSimon Milner (guest): As a former academic economist, now turned corporate governance practitioner, the happiness debate has reignited and sustained my interest in academic economics like nothing else. I’m curious as to why Martin Wolf has chosen now to write about my former colleague (strictly, former boss) Richard Layard’s book. Andrew Oswald’s comment demonstrates that the academic debate has moved on consideraly since 2005 and it’s a shame Martin doesn’t acknowledge that and explore those new avenues. Perhaps the FT should invite Andrew Oswald to write about it?
I’m curious as to why the happiness/wellbeing debate hasn’t moved on much from the academic and public policy environment to the world of business. Very few companies seem to assess, publicise or care about the happiness of their employees. Yet there are fragments of information emerging about how employee wellbeing is connected with firm performance - the recent Meaning Inc book by Gurnek Bains et al presents some compelling evidence on this. Moreover, unlike countries where long-term means a decade or more, the culture, structure and processes of companies can be changed relatively quickly in order to improve wellbeing and, potentially, performance. I’d be really interested in knowing whether anyone else is looking into this.
Lastly, Andrew Oswald’s final comment is very telling, and indeed it fits with Martin Wolf’s view that no-one with retrenched political views should find the happiness evidence comfortable. Unfortunately politicians are well versed in picking up prescriptions which fit with their world view, and disregarding those that don’t. A wider, constructive debate on happiness rather than a nit-picking, point scoring one is what’s needed to keep the politicians honest(ish).
Posted by: Simon Milner | June 6th, 2007 at 2:36 pm | Report this commentPaul Seabright: Martin reports - as though it were firmly established - the “most important negative conclusion” of this literature, namely “that, beyond a certain threshold, extra wealth does not make us any happier”. He might have said more about the possible gap between our reported contentment or disgruntlement, and the underlying well-being that these reports are trying to measure.
When we are asked to report how our lives are going most of us don’t have either the information or the habits of comparison required to give an answer that means anything compared to the answer our grandparents might have given. We compare relative to our own aspirations, our own societies, our own social circles, and we might give radically different answers about our current state if we were asked to evaluate it from the perspective of different circumstances. (An interesting experiment by a British television programme asked a family to live in conditions typical of families a century ago, and the subjects were astonished and appalled at the sheer boredom and drudgery involved, and much more appreciative of their modern comforts than they had been before).
The greater prosperity of modern societies has not just meant an end to starvation, but also opportunities for travel, creativity and self-expression that almost certainly do not show up in intertemporal happiness comparison, but the prospect of losing which might make us very unhappy indeed. Arguably, too, the ability freely to express disgruntlement is one of the ingredients of a happy and fulfilled life.
None of this means that reported happiness is uninformative - this literature is full of fascinating findings, like those of women relative to men to which Andrew Oswald referred. It does mean, though, that reported happiness bears a complicated and distorted relationship, about which not enough is understood, to whatever notions of happiness we think really matter. This is, by the way, an argument for more happiness research, not less.
Posted by: Paul Seabright | June 7th, 2007 at 9:29 am | Report this commentTito Boeri: Martin Wolf pays no attention to the issue of measurement of happiness and claims that, according to Richard Layard, “happiness is measurable”. I believe that the issue of measurement comes before any philosophical consideration.
Why bother discussing whether happiness indicators speak in favour of progressive taxation if we have no clue as to what the happiness numbers are telling us? Unsurprisingly, the measurement issue is addressed at the very beginning of Layard’s book as well as of his masterfully written Robbins lectures.
Available happiness measures draw on retrospective subjective well-being scores offered by household surveys. These scores offer, at best, subjective perceptions (or remembered perceptions) of experiences rather than measures of well-being and can hardly be compared across individuals, let alone countries.
Moreover, reported life satisfaction has to do with factors that are (fortunately, in my view) largely unrelated to economic policy. If we are looking for some guidance for the design of economic policy, then it is preferable to focus on subjective income satisfaction measures rather than on life satisfaction. A final word about liberalism. Are we sure that maximising happiness measures regardless of learning processes, social and political interactions is democratic? The idea of maximising happiness, however defined, makes me think of central planning….
Posted by: Tito Boeri | June 7th, 2007 at 12:32 pm | Report this commentGerry Steele (guest): My paper in The Political Quarterly (December 2006) explains how one might be excused for thinking that Richard Layard’s arguments are tailored to fit a conclusion: that happiness is located in socialist ethics. His many inconsistencies and omissions are consistent with that view.
A major theme is that western man has been duped into playing a zero-sum game of trying to do better than everyone else. However, the utilitarian cost-benefit paradigm overlooks the very essence of the human condition: a dynamic competitive spirit that repeatedly sets itself difficult, but potentially achievable, objectives; that accepts a bad outcome as part of a rebuilding process prior to taking on the next challenge. The dynamic is that of entrepreneurship (which is not reducible to mere effort in a cost-benefit trade-off) by which individuals continuously set their wits against events. Unhappiness is notable where individuals are denied opportunities to act entrepreneurially; as under former Soviet systems. Without the exercise of intellectual and/or physical challenges, individuals take on the torpor of caged animals.
Insignificant correlations between happiness and income are surely no surprise given that income is irrelevant to unchanging features of the human condition: love, paternity, bereavement, self-esteem, and the success of your soccer team. Alternatively, if there had been a close association between income and happiness since homo sapiens first emerged 100,000 years ago, it would be hard to imagine how early man could have survived unfathomed depths of depression, misery and pain!
The static utilitarian cost-benefit work/effort-leisure/happiness trade-off is only peripherally relevant to what John Rawls (in A Theory of Justice) refers to as ‘primary social goods’. These include rights and liberties, opportunities and powers, income and wealth and self-esteem. Pleasure and enjoyment arise when we exercise our faculties; and the more enjoyable pleasures arise in engaging with complex problems. So, although Layard might be applauded for seeking to broaden the remit of economics, if moral issues (such as income redistribution) continue to be examined by the yardsticks of utilitarianism, social welfare optimisation and the elusive common good, the engagement is destined to remain Quixotic. The narrow utilitarian preoccupation with activities that make life better or worse against a linear scale, delivers an analysis of an amoral and self-contained agency that reveals virtually nothing of the socio-psychology of the human condition.
G R Steele is a lecturer at Lancaster University Management School
Posted by: G R Steele | June 7th, 2007 at 4:02 pm | Report this commentWillem Buiter: Richard Layard’s book, and the ‘economics of happiness’ school in general scare me. Let me try to explain why.
Replacing the word ‘utility’ by the word ‘happiness’ does not solve any of the problems of utilitarianism either as positive behavioural science or as a normative guide to public policy. It was good PR by the psychologists - and now also the economists -labouring in the happiness vineyard, to appropriate (I would say, mis-label) a word like ‘happiness’ that resonates with everyday meaning, for a very specific and only vaguely related survey response measure, the retrospective subjective well-being scores on which much of the evidence cited in Layard’s book is based. Only ‘Granger-causality’ and ‘Dark matter’ come close as examples of stealing attention by mislabeling. I will in what follows refer to it as self-reported retrospective happinesss or SRRH. SRRH appears to correspond, roughly, to a retrospective warm feeling inside. It has observable correlates in chemical reactions and electrical impulses in various parts of the body, including the brain. The empirical work cited in support of the proposition that most of SRRH can be accounted for by just six observable correlates (divorce, unemployment, the level of trust in other members of society, membership of non-religious organisations, the quality of government, and belief in God) is shot through with selection bias problems. At least two of the factors appear to be proxies for whether people feel life has treated them fairly or justly (trust, quality of government). None of this work addresses the widespread suspicion that security and happiness can be stultifying and may suppress creativity .
Although separate from the general case for or against SRRH, Richard’s trust in cognitive behavioural therapy as an effective treatment for depression, anxiety and related mental illnesses is based on a selective reading of the clinical evidence. To be effective, high demands are placed on the training and quality of the therapist; the patient needs to be persistent and willing and able to work on his/her problems. This combination of prerequisites for effectiveness means that CBT can only ever be effectively implemented for a small minority of people with depression, anxiety and related mental health problems.
One of the more sensible bits of the economics of SRRH is a restatement of the venerable relative income hypothesis. SRRH (or utility) is, once we are above some partly socially constructed concept of subsistence, an increasing function not of our absolute level of income (or consumption) but of our income relative to that of our reference groups. Habit formation (the thrill of great wealth wears off when one gets used to it) may also play a role. If income or consumption is, in rich societies, largely a positional or status good, there will tend to be over-consumption, and too much work effort, because of the negative external effect of each person’s increased consumption and income levels on the happiness of others. We could add to that the environmental externalities of material consumption, which are an issue even when consumption is not a positional good. Richard Layard concludes from this that labour income should be taxed to internalise this negative externality.
As Martin points out, labour income is taxed. My own marginal effective tax rate on labour income is about 57.5 % made up of a 40% marginal income tax rate and a 17.5% VAT rate. Perhaps that is high enough to internalise the envy externality. More importantly, Richard does not pay sufficient attention to how the status and position – seeking instincts and urges of SRRH-seekers will be channelled if additional work and consumption are discouraged through high marginal tax rates or other measures. I get the impression, reading his book, that Richard believes the time diverted from excessive work, would be spent enjoying the company of friends and relatives, writing sonnets and growing bonsai trees. It is rather more likely that the urge for status and prestige would be satisfied by behaviour aimed at achieving power and control over others through non-financial means, by intimidating others, through threats, violence, competitive displays of excessive risk taking, the tribal rivalries of gangs and soccer hooligans etc. Excessive working hours to earn income spent in competitive consumption may be a relatively harmless way to channel the competitive, status seeking urges of men and women. Didn’t the Beatles sing “Happiness is a warm gun”?
Even if people behaved as if they pursued their SRRH, such a positive theory of human behaviour would not make the pursuit of SRRH a normative goal of public policy. All the important philosophical and practical problems of conventional utility-based welfare economics are, of course, still present in the SRRH approach, with one unimportant exception: utility has moved a long way from being an ordinal concept to being a cardinal concept. But whether government should try to maximise, say, the sum of the SRRH of all its citizens (classical utilitarianism) or the SRRH of its least happy citizen (the Rawlesian version of SRRH welfare economics) is as much an open question as it has always been.
And this is on the assumption that the government should pursue utilitarian public policies, that is, that the social welfare function is increasing in citizens’ SRRH. I can find no argument in Richard’s book for the promotion by the government of SRRH other than his belief that SRRH is what people themselves pursue. Because what is does not provide any useful guidelines for what ought to be, we must look for other grounds for SRRH maximisation as the objective of public policy. I am unimpressed with the ethical/moral foundations of any form of utilitarianism. Instead I believe that robust value systems and norms that govern what people ought to do, as distinct from what they actually do, tend to be grounded in religious beliefs or have other deep, non-religious philosophical-ethical foundations. From this perspective, utilitarianim is but one, rather philistine, set of ethical foundations for government behaviour.
Let me illustrate this with the ethical principles I am most familiar with, those of mainstream Dutch/Calvinist Christianity. Christian ethics is based on the two great commandments, quoted by Christ from the Books of Moses: First, you shall love the Lord your God with all your heart, all your soul and all your mind; and, second, you shall love your neighbour as yourself”. Neither ‘love God’ nor the Golden Rule offers much support for utilitarianism in any form. This presumption is reinforced in an often-quoted description of the good life found in the writings of the prophet Micah: “He has told you, O mortal, what is good; and what does the Lord require of you but to do justice, and to love kindness and to walk humbly with your God”. This has a distinctly anti-materialist ring to it, and sits uncomfortably with both classical utilitarianism and its new SRRH version, which would qualify as the self-indulgent, egotistical pursuit of personal comfort and well-being. The pursuit of personal happiness in an unjust world would be a manifestation of moral failure.
In this religious belief system, human beings have the gift of freedom, that is, the ability to make choices and the duty to make moral choices, thrust upon them. People are condemned to be free, and they will be held to account for the choices they make. The task of government therefore is to enhance the domain of choice, that is, to maximise the freedom of all its citizens, so they can make choices and be judged on the nature of these choices and their consequences for themselves and others.
This brief example does not do justice either to Christian ethics as a basis for a liberal economic and political order, or to the infinite complexity of policy design – including the key question of how the freedom of one impacts on that of another. But it does provide at least an outline of one possible alternative to the shallow SRRH utilitarianism promoted by Richard and others.
Richard believes that the government should pursue our happiness and knows how to do it. I perceive in that argument the outline of a nanny-state-on-steroids and the end of freedom. Martin believes happiness is something we have to pursue – and perhaps never find – for ourselves. My inner Calvinist tells me to seek the kingdom of God. God only knows that that means, but it is clear that it does not mean the pursuit of personal happiness, self-reported or other. Happiness is a tale told for children. Grownups should know better.
Posted by: Willem H. Buiter | June 8th, 2007 at 12:18 pm | Report this commentRichard Layard: I am grateful for Martin Wolf’s considered comments on my book Happiness: Lessons from a New Science. But there are two points he has misunderstood.
I did not say that taxes should be higher than they are now. I simply pointed out that taxes have advantages as well as disadvantages, and one advantage is that they discourage the self-defeating rat race. When thinking about optimal taxation, that should be factored in.
Nor did I say that we could abolish the individual’s pursuit of status, which is hard-wired in our genes. But I did say that, through education, we can affect which human qualities confer status. We should, for example, increase the status attached to being kind and considerate.
When English teenagers are asked “Are most of your classmates kind and helpful?” only 43 per cent say Yes, compared with 75 per cent in Scandinavia, where levels of trust have not fallen as they have here and in the US.
So yes, Martin, there are some aspects of modernity that we should ameliorate.
Posted by: Richard Layard | June 8th, 2007 at 4:05 pm | Report this commentAdrian Wood: I recommend to all who have not already read it Daniel Gilbert’s book ‘Stumbling on happiness’, which just won the Royal Society’s annual prize (he’s a Harvard psychologist). Its sub-title is ‘Why we so often fail to know what will make us happy in the future’. Reading it might even prompt Martin to write a follow-up column!
Posted by: Adrian Wood | June 10th, 2007 at 9:01 pm | Report this commentMartin Wolf: This has been a spirited and thought-provoking debate. Let me pick up just a few of the salient points.
First, is the study of happiness scientific? In other words, do its results tell us something useful and important about the world? Andrew says it does and points to refereed journals to prove it. If I had the nerve I would be tempted to say that “garbage in” remains garbage on the way out, however brilliantly analysed.
Does what Willem calls “self-reported retrospective happiness” (SRRH) really tell us anything useful about the states of mind of large populations over periods of time or about those of different populations at the same time? The big chart from Richard’s book, reproduced in my column, shows that Nigeria is “happier” than Italy. Does anybody really believe this? Suppose you opened the borders between the two countries, in which direction would you expect the populations to flow? Does that tell us nothing useful about their relative happiness? Similarly, Mexico is apparently as happy as the US. So why are there some 11m illegal immigrants, most of them Mexicans, in the US? Are they all suffering from false consciousness?
Paul is surely right that we need to understand the difference between what people report and the underlying well-being that researchers are trying to measure. Human beings have limited frames of reference. We cannot imagine the lives of our great great grand-parents. We would certainly be pretty miserable if we were plunged back into the world of 1900. But SRRH is redefined in relation to current expectations and is, consequently, adaptive. In time we would probably adapt to such a shock or at least our grandchildren would. But the conclusion from this is that almost every endeavour, beyond the removal of extreme misery, is pointless, if one considers happiness the objective of human existence, since our happiness adapts to whatever we have.
The second big issue is whether happiness ought to be the objective of human society. As Willem points out in his brilliant comments (with which I largely agree), we cannot derive an “ought” from an “is”. Even if we do individually only pursue something we could credibly call happiness (which I doubt, except tautologically) and even if we could measure happiness both across people and over time reliably (which I also doubt), we would not have proved that society ought to attempt to maximise happiness.
Some of the objections to this philosophy have been made in these comments. As Tito says, it is an inherently undemocratic and (as Willem notes) even scary idea. What does it have to do with democracy or individual liberty? Indeed, it is clear that we would have to suppress people’s freedom wherever the unhappiness they inflicted exceeded the happiness they create. So you want to divorce your wife? Well you can’t because the unhappiness you would inflict upon her and your children exceeds the happiness you would gain for yourself. So you want to vote for someone who would make society unhappy, well, tough, he has been removed from the ballot box. What would be the limits to the actions of the happiness police?
Yet more fundamentally, human beings clearly value many things – effort, creativity and even the endurance of pain and difficulty – over happiness. Unless one says, emptily, that everything is done in pursuit of happiness, by definition, it is untrue that happiness is our sole objective as individuals. Moreover, as a social philosophy it leads us towards the world laid out in the greatest anti-utilitarian tract: Aldous Huxley’s Brave New World. Indeed, there are passages in Richard’s book, particularly his discussion of the use of drugs to cure mental illness, that remind me forcefully of that brilliant dystopia.
The third question is whether we even know how to maximise happiness at all. Here I am comforted by what Andrew says: we do not know how to do so. I note that he reports the decline of women’s happiness as women have become freer, more socially powerful and more economically emancipated than ever before. That underlines how impossible it is to use this idea for social policy.
I hope I have given my reasons for doubting the scientific and philosophical basis for this research. But let me make two final comments.
The first is in response to Willem’s chiding me for my remark about pursuing happiness. This, of course, was taken from the Declaration of Independence. I did so in order to establish firmly the individualistic premises I hold, as did Jefferson himself. He did not say that the newly independent American state should attempt to make its citizens happy, but rather that it should give them the freedom to pursue that goal, if they wished. I accept the freedom of choice. But what people aim to pursue is (largely) up to them. We have no right to impose our own goals on other people.
The second is in response to Richard’s short rejoinder. I agree that kindness is a virtue to be encouraged. It is right that people should take into account the happiness of others. Benevolence is a good. I do not have to accept utilitarianism to believe that.
Posted by: Martin Wolf | June 15th, 2007 at 5:43 pm | Report this comment