Would an hedonimeter help us to choose between push-pin and poetry?

There seems to be enduring interest in the post, ‘Is push-pin as good as poetry’, that I wrote a few years ago. That post was about John Stuart Mill’s rejection of Jeremy Bentham’s assertion that if the game of push-pin gives more pleasure than poetry then it is more valuable than poetry. Mill argued that some pleasures are superior to others and that it is possible for people to become addicted to inferior pleasures. I followed up on the question of whether push-pin might be addictive in a subsequent post.


A logical place to begin this post would be to define an hedonimeter. Before I do that, however, I want to make the point that Mill’s thoughts about higher and lower pleasures seem to have been a side-track that did not lead anywhere in the subsequent development of economics. A few years after publication of Mill’s essay on utilitarianism, the famous economist, William Stanley Jevons, came down strongly in favour of Bentham’s view of utility. According to Jevons:

‘Whatever can produce pleasure or prevent pain may possess utility. ... The food which prevents the pangs of hunger, the clothes which fend off the cold of winter, possess incontestable utility; but we must beware of restricting the meaning of the word by any moral considerations. Anything which an individual is found to desire and to labour for must be assumed to possess for him utility. In the science of Economics we treat men not as they ought to be, but as they are’ (‘The Theory of Political Economy’, 1871, III.2).

Francis Edgeworth, who came with the idea of an hedonimeter, was a strong supporter of Jevons’ view that utility has two relevant dimensions: intensity and time. Edgeworth suggested that we:

‘...imagine an ideally perfect instrument, a psychophysical machine, continually registering the height of pleasure experienced by an individual ... From moment to moment the hedonimeter varies; the delicate index now flickering with the flutter of the passions, now steadied by intellectual activity, low sunk whole hours in the neighbourhood of zero, or momentarily springing up towards infinity’ (‘Mathematical Physics’, 1881).

I think that description tells us, in today’s language, that an hedonimeter might be a sexy idea. It will probably be a long time before you can buy an hedonimeter at your local supermarket but research indicates that levels of various chemicals (e.g. cortisol) in the body and activity in various parts of the brain are related to pleasant and unpleasant experiences. It seems likely that if we were able to conduct surveys using hedonimeters – perhaps one day it might be possible to carry one around like a pedometer – they would give similar results to those obtained by Daniel Kahneman and Alan Krueger using evaluated time use (ETU) techniques. In surveys using ETU techniques respondents are asked to account for time spent on various activities on the preceding day and to rate their feelings for each activity in terms of a range of affective categories e.g. happy, worried/ anxious or angry/hostile. The ETU data suggests that people tend to get most pleasure from sex and socializing and least pleasure from working and commuting.

If anyone was really interested in comparing the pleasures people obtain from push-pin and poetry it is possible to imagine conducting an experiment in which participants played push-pin and attended poetry readings and rated their experiences. If you were concerned that such information might not be relevant to you personally, you might be able to get someone to design an appropriate experiment to enable you to rate and record your emotions during the two experiences.

How would you feel about using the results of an ETU exercise (or an hedonimeter) to decide something that may actually be important to you – for example, whether to play sport A or B or have a holiday at location X or Y - on the basis of the balance of emotions that you have experienced in those activities in the past? I don’t know about you, but I would want to see whether the hedonimeter results are consistent with my memories of the different experiences before I decided whether to use them.

A recent presentation by Daniel Kahneman (on TED) suggests that what my reflective self might feel about my memories of the experiences that I want to choose between could differ substantially from the emotions that my experiencing self has actually felt (or what an accurate hedonimeter might record).

Would you ignore the evidence of an hedonimeter if it conflicted with your memories of the experience? Would you ignore suggestions from your spouse or a friend that your memories of a holiday might be biased by the way you felt about something that happened at the end? If you had a reliable hedonimeter to record your memory of past pleasures would you give less weight to consideration of what experiences it is good to have (or issues such as those raised by J S Mill about the superiority and inferiority of different pleasures) in making your choices? Why do humans have selective memories? Do our selective memories serve a useful function from an evolutionary perspective?
Winton Bates

Postscript 1:
In retrospect I would like to add some further questions: Do the reasons why human have selective memories make any difference to the way you and I should live our lives? If I am told that there are good evolutionary reasons why I should remember how an experience ended (e.g. because it was important to the survival of my ancestors to remember which of their hunting expeditions were successful and unsuccessful) is this relevant to the decisions I should make today?
I am beginning to think that while the evolutionary reasons for cognitive bias may be interesting they may not be particularly relevant to our current decision-making.

Postscript 2:
I obviously became sidetracked while attempting to answer this question. My answer is that whether the hedonimeter would help us to make the choice depends on the criterion we think is most appropriate. If the criterion is pleasure, the hedonimeter might help. If the criterion is what is good for us, then pleasure is only factor that we would take into account and an accurate measure of pleasure might not be particularly useful. We might consider that it is better for us to spend Thursday evenings reading poetry rather than playing pushpin even though we might get more pleasure from playing pushpin.

Should schools and universities teach students how to be happy?

This question arose from the education chapter of ‘The politics of happiness’, by Derek Bok. I wrote some general comments about this book in my last post. Bok’s views about education deserve further consideration, however, not least because he was president of Harvard for a large slice of his professional career.


In reading the education chapter I was particularly interested to see how the author would deal with evidence that education, like economic growth, has little direct effect on happiness in the United States. (Australian evidence suggests that university level qualifications have a negative effect on happiness.) Bok’s views on education are somewhat more positive than his views on economic growth, but he is highly critical of the weight that education policies currently give to acquisition of vocational skills.

The first paragraph of the chapter sets the tone for the discussion that follows:

‘People often misjudge what will bring them enduring happiness or pain. It stands to reason, then, that any serious attempt to increase well-being should give a prominent place to education. Schools and universities are the obvious institutions to assume this responsibility by trying to cultivate interests and supply the knowledge that will help young people make more enlightened choices about how to live their lives’ (p. 156).

I would have liked to see recognition in that paragraph of the responsibilities of parents and of young people (who should be accepting responsibility for their own lives by the time they get to university). But I guess that most parents hope that the schools their children attend will somehow help them learn how to make enlightened choices about how to live their lives.

Does that mean that the results of happiness research should be taught in schools? Bok toys with that idea. He notes that one program based on happiness research, which was presented to school students by Martin Seligman, succeeded in cutting the incidence of depression in half over a two- or three year period. In the end, however, he comes down against teaching happiness in schools on the grounds that teachers who are not adequately prepared to present such material could give students a distorted view based on pop psychology. I suppose that view is wise. I think it would certainly be unwise for time spent on happiness studies to be at the expense of literacy and numeracy skills, and the development of skills in acquiring knowledge.

Most kids probably pick up more than enough pop psychology already from various sources, including their teachers, sports coaches and the self-help section of the local bookshop. If they are lucky they might look on a different shelf at the bookshop and find something more informative, like ‘Stumbling on happiness’ by Daniel Gilbert. When they find difficulty coping with life they might also get some helpful coaching from a competent therapist. To put these thoughts less cynically, while I doubt whether it is appropriate for happiness to be taught in schools, it would be good if children could be mentored by people whose knowledge of positive psychology was not gained solely by reading best sellers on such topics as the power of positive thinking and the secret of becoming an instant success at everything you might like to do.

Professor Bok is more strongly supportive of the study of the science of happiness being included in higher education. He has some reservations, however, about the inclusion of practical exercises to help students bring about ‘a personal transformation’, on the grounds that this ‘may strike some people as uncomfortably close to indoctrination’ (p. 172). It certainly strikes me that way. I think it makes sense to draw a firm line between what is taught as part of a curriculum and practical exercises in self-help that may be offered as an extra-curricula activity.

As noted earlier, Bok is highly critical of the current emphasis in education on acquiring vocational skills. He suggests that more research is required to understand what undergraduate experiences tend to be associated with greater happiness in later life. Our lack of understanding of such matters was highlighted to me in a recent paper by Michael Dockery which indicates that young people in Australia who experience higher education switch from having relatively high to relatively low levels of happiness (compared to others of the same age) at about the time they complete their degrees (LSAY, ‘Education and happiness in the school to work transition’, 2010). A variety of explanations are possible, including the development of unrealistic expectations at universities and a change in reference group following graduation. Andrew Norton's theory is that the graduates become unhappy because they are moving from an intellectually stimulating environment where they have a great deal of freedom to one they have much less discretion over what they are doing.

I think Australian universities should be giving priority to research to shed light on the reasons for this decline in happiness. Some young people may begin to wonder whether the sacrifices they make to attend university are worthwhile when they find out that they will probably end up less happy than their contemporaries who attain less advanced vocational qualifications.
Winton Bates

Postsript:
Since writing this I have been reminded that Martin Seligman has been involved in a major experiment in teaching happiness at Geelong Grammar School. He recently worked and lived with the teaching staff for six months preparing them to teach positive psychology. A transcript of Kerry O'Brien's ABC interview with Seligman in December 2009 can be found here.

Would the politics of happiness be any better than what we have now?

While reading the first few chapters of Derek Bok’s recent book, ‘The politics of happiness’, I thought its main contribution would be to help to highlight findings of happiness research that are relevant to government policy. This book turned out, however, to be more of a treatise arguing that politics should focus on raising average happiness levels of the population. People interested in an explicit discussion of the policy-relevance of happiness research should look elsewhere, for example: Ed Diener et al, ‘Well-being for Public Policy’ (which I discussed here and here and in a review essay for ‘Policy’, Summer 2009-10).


The first five chapters of ‘The politics of happiness’ provide a lucid discussion of broad findings of happiness research, the reliability of this research, philosophical issues relating to its use by policy-makers, and implications of findings for economic growth and equality as policy objectives. There is little discussion of the findings of happiness research in the chapters that follow - relating to the threat of financial hardship, relieving suffering, marriages and families, education and the quality of government. The focus in the second half of the book seems to be more on the policy implications of adopting the objective of raising happiness in the U.S. than on the relevance of happiness research to consideration of particular policy issues.

The main tenets of the politics of happiness as set out by Derek Bok are as follows:
• Americans have not become much happier over the past 60 years despite the dominant place of economic growth on the domestic political agenda and the vast increases in GDP that have occurred.
• The idea of growth ‘has shifted from a means to desired ends to an end in itself, an end with no foreseeable end in sight’ (p.66).
• There is no way that policy-makers could stop the economy from growing ‘without creating problems that would outweigh any hoped for benefits’.
• Policy-makers should re-focus their priorities away from economic growth toward efforts to raise happiness levels - for example through ‘programs to strengthen marriage and family; encourage active forms of leisure; cushion the shock of unemployment; guarantee universal health care and a more secure retirement; improve child care and pre-school education; treat mental illness, sleep disorders and chronic pain more effectively; and focus education policy on a broader set of goals’ (p.208).
• Law-makers should make efforts to build greater confidence in the political process ‘through measures to reduce the influence of money and special interests’ and to curb a range of ‘efforts by politicians to place their own reelection above the general welfare’ (p. 209).

The main problem with this account of the politics of happiness, from my perspective, is that it would substitute a narrower view of well-being than that which is currently used in policy discussions. Derek Bok adopts Ed Diener’s definition that a person is said to be happy who ‘experiences life satisfaction and frequent joy, and only infrequently experiences unpleasant emotions such as sadness or anger’ (p. 9). I accept that is an appropriate definition of happiness – as the term is most commonly used. But self-reports of emotional states cannot reflect the impact on well-being of an expansion of opportunities that cannot be anticipated.

To be more specific, it is unrealistic to expect the benefits of economic growth in high income countries – benefits that are mainly an outcome of technological progress – to be measured by changes in average happiness or life satisfaction. For example, does anyone seriously expect that people living in 1950 could have felt unhappy or dissatisfied - or sad, or angry even - because they did own personal computers or any of the numerous other amenities of modern life that had not then been invented? The fact that we do not feel dissatisfied that we do not yet possess the products of future technological progress does not mean that such products will not enhance our future well-being and that of our descendents. What it means is that our emotional systems fortunately enable us to be satisfied with what life offers us if we can attain a standard of living that is within the bounds of what it is currently possible for humans to attain.

It seems to me that attempting to compare the well-being of succeeding generations of people using happiness survey data is like trying to compare the prowess of succeeding generations of sporting teams by asking the members of the each team to rate their own performance on a scale of 1 to 10. The ratings would be uninformative because each person could be expected to judge his own performance relative to that of his contemporaries rather than to compare his performance with that of counterparts in the other team.

In the case of well-being comparisons, it is relevant to ask survey respondents to compare their own standard of living with that of their parents. Survey responses to that question over the period since 1994 show that between 60 and 70 percent of Americans assess their standard of living to be better than that of their parents at a similar age.

As an outside observer of the U.S. I am surprised by claims that economic growth has had a dominant place on the political agenda and that growth has become an end in itself in that country. As far as I am aware there is not much government intervention in the U.S. economy that is actually directed toward increasing economic growth. Most of the political rhetoric about the benefits of growth seems designed to reduce government interventions that hinder economic growth rather than to raise the growth rate beyond the levels that would otherwise emerge from decentralized decisions of individuals and firms about saving, investment, research, innovation etc.

Apart from its questioning of the contribution of economic growth to well-being, the politics of happiness does not seem to me to provide a particularly innovative agenda for paternalistic politics. The question that is largely overlooked, of course, is whether well-meaning paternalism is more likely to help individuals to flourish or to hinder their flourishing by making them increasingly dependent on governments.

Finally, it is worth noting that just about all politics in democratic countries is about happiness – using the term now in the broader sense to cover all aspects of well-being. Irrespective of the motives of participants who are presenting their particular views an appropriate consideration of just about every public policy issue must involve an assessment of claims about the potential effects of proposed policy changes on various aspect of human well-being. It seems likely that some of the findings of happiness research will add useful information to such assessments and will result in better policy decisions.
Winton Bates

Does identity economics predict happiness in different societies?

The key idea of 'identity economics' - as discussed in the recent book of that name by George Akerlof and Rachel Kranton - is that people gain utility when their actions conform to the norms and ideals of their identity (or social category e.g. gender, race, social class, age group) and lose utility when they do not. We should expect this to be so if we accept that individual behaviour is strongly influenced by socialization and that norms of behaviour can differ between people in different social categories.

This key idea of identity economics suggests that, other things being equal, people are likely to be happier when they live in societies where there are few economic or social pressures for them to act in ways that are contrary to the norms and ideals of their identity. It is important to define what we mean by happiness in this context. I think the appropriate concept in this instance is emotional well-being, as measured in happiness surveys, rather than an all-inclusive concept of well-being explicitly incorporating such factors as wealth, health and education. It seems reasonable to expect that, other things equal, people with high levels of identity utility would be more likely to say that they are happy or satisfied with life than would people with low levels of identity utility.

I get the impression from the methodology chapter of 'Identity Economics' that the authors are not particularly interested in econometric tests of the predictive power of their theory. They suggest that it is difficult to falsify any theory because ‘even the most straight forward test has literally millions of possible specifications’ (p. 115). I think they are exaggerating, but even accepting their point it seems to me that useful knowledge may be gained by considering what predictions might follow from a new theory and how well those predictions describe aspects of the real world.

It seems to me that one prediction we could make on the basis of identity theory is that people in homogeneous societies are likely to be happier than those in societies in which individuals have to interact with people who have different cultural backgrounds. We might expect that this effect would depend on the strength of ties to particular groups in multi-cultural societies, but we would expect homogeneity to be a plus for average happiness.

Is the evidence from happiness research consistent with this prediction? When I did a Google search using the words ‘happiness’ and ‘homogeneity’ the first results I saw seemed consistent with the prediction – people in countries that are homogeneous (as well as affluent) tend to be the happiest. However, delving deeper, I found a more careful study for 65 countries, using multiple regression to control for other variables, which suggests that ethnic homogeneity may actually have a negative effect on average life satisfaction (Douglas Barrett, Kristen Van Rensselaer and Bruce Gordon, ‘Possible effects of national population homogeneity on happiness’, Journal of International Business Research, 6 (1), 2007).

The authors comment as follows:

‘The results suggest that life satisfaction is significantly negatively related to ethnic homogeneity in the presence of our control variables. This is consistent with the “melting pot” concept, as multiple cultural influences create a richer living environment’ (p. 55).

I’m not sure what a richer living environment actually means. Perhaps one way of thinking about it is that a richer living environment enables us to get some of the benefits of foreign travel (e.g. sampling foreign food) without leaving the country. I think the result may also be consistent with identity economics if we view successful multicultural societies as providing learning experiences in which a substantial proportion of the population from all major ethnic groups have come to view themselves as tolerant of diversity. We should expect tolerant people to gain utility from acting in accordance with their ideals.
Winton Bates