Happiness & Public Policy

The truly delicious bits of this new NBER working paper by Di Tella, Haisken-De New, and MacCulloch on adaptation to income and status is the stuff on political leanings:

We study “habituation” to income and to status using individual panel data on the happiness of 7,812 people living in Germany from 1984 to 2000. Specifically, we estimate a “happiness equation” defined over several lags of income and status and compare the long run effects. We can (cannot) reject the hypothesis of no adaptation to income (status) during the four years following an income (status) change. In the short-run (current year) a one standard deviation increase in status and 52% of one standard deviation in income are associated with similar increases in happiness. In the long-run (five year average) a one standard deviation increase in status has a similar effect to an increase of 285% of a standard deviation in income. We also present different estimates of habituation across sub-groups. For example, we find that those on the right (left) of the political spectrum adapt to status (income) but not to income (status).

That is (in case you’re confused), folks on the left get used to money, but not status and the reverse for folks on the right. This is funny, since I’ve been reading a bunch of papers on inequality, mostly by political philosophers on the left, and they are positively obsessed specifically with the status effects of material inequality. It’s pretty amusing if this is just a reflection of a particular personality type. More generally, the fact that the happiness-effects of various things seem to be mediated by ideological leanings seems to basically ruin the prospect of using happiness research as a neutral, scientific way of assessing policy. It may just end up sort-of-usefully reminding us that one group may like a certain policy and another group may not simply because it makes one group feel better and another group feel worse. It doesn’t settle the dispute: it explains why we’re having it. Also, ideological mediation is one more nail in the coffin for the introspective method of normative philosophy. If the effects of this or that on people’s sense of well-being is mediated by their ideological cast, then chances are, our intuitions about real and hypothetical cases are probably already deeply infected with our ideological notions–or with the personality traits that lead us to find those notions attractive–and arguments based on these intuitions simply beg all the interesting questions in a subtle way.

That’s “reliable” in the technical measurement sense of “repeatability” or “consistency.” Alan Krueger and David Schkade are on the case with a new NBER paper, “The Reliability of Subjective Well-Being Measures.”

ABSTRACT

This paper studies the test-retest reliability of a standard self-reported life satisfaction measure and of affect measures collected from a diary method. The sample consists of 229 women who were interviewed on Thursdays, two weeks apart, in Spring 2005. The correlation of net affect (i.e., duration-weighted positive feelings less negative feelings) measured two weeks apart is 0.64, which is slightly higher than the correlation of life satisfaction (r=0.59). Correlations between income, net affect and life satisfaction are presented, and adjusted for attenuation bias due to measurement error. Life satisfaction is found to correlate much more strongly with income than does net affect. Components of affect that are more person-specific are found to have a higher test-retest reliability than components of affect that are more specific to the particular situation. While reliability figures for subjective well-being measures are lower than those typically found for education, income and many other microeconomic variables, they are probably sufficiently high to support much of the research that is currently being undertaken on subjective well-being, particularly in studies where group means are compared (e.g., across activities or demographic groups).

The passage in bold is not exactly a ringing endorsement, and definitely a call for caution, the implication being that the reliability of SWB measures are insufficient for some current research. I look forward to digging in deeper.

[Thanks to Tyler Cowen for the tip.]

In the post below I say I’m skeptical of numbers showing an explosion in rates of depression. Here, in part, is why… From yesterday’s New York Times.

About one in four people who appear to be depressed are in fact struggling with the normal mental fallout from a recent emotional blow, like a ruptured marriage, the loss of a job or the collapse of an investment, a new study suggests. To avoid unnecessary diagnoses and stigma, the standard definition of depression should be redrawn to specifically exclude such cases, the authors argue.

The study, appearing today in The Archives of General Psychiatry, is based on survey data from more than 8,000 Americans; it did not analyze the number of people who had been misdiagnosed.

Psychiatrists and other doctors who take careful medical histories do so precisely to rule out such life blows, as well as the effects of physical illnesses, before making a diagnosis of depression.

But the American Psychiatric Association’s diagnostic manual does not specifically exclude people experiencing deep but normal feelings of sadness, unless they are bereaved by the death of a loved one. And an increasing number of school districts and health clinics use simple depression checklists, which do not take context into account, the authors said.

“Larger and larger numbers of people are reporting symptoms on these checklists, and there’s no way to know whether we’re finding normal sadness responses or real depression,” said Jerome C. Wakefield, a professor of social work at New York University and the study’s lead author.

This is big. As I’ve pointed out before, the depression stats and the happiness stats seem to be in conflict. There has been a stable or shrinking percentage of the population in the bottom happiness category despite alleged huge increases in the incidence of depression. This creates a problem for researchers who lean hard on both sets of data, like Diener and Seligman do in their paper “Beyond Money.” Here’s the dialectic as I see it…

If the depression data is right, the happiness data must be broken for failing to detect any increase in the proportion of the population feeling unwell. You would then have to give up on using the happiness data as evidence that many people are not getting happier, since you’ve already established the unreliability of self-reportinng to track important changes in psychological well-being. Now, you could argue that depression and unhappiness are different and statistically unrelated things. But then you need to convince us which one is more important for well-being. If the depression numbers are right, and depression is a huge deal, but has no relationship to unhappiness, then perhaps self-reported happiness and unhappiness are not very relevant to well-being. But then you don’t get to skip back and forth from one set of data to the other, whenever it is convenient to your argument.

You could argue that the happiness data are right, in which case, you’ll have a problem with the depression data. Again, you could distinguish between unhappiness and depression, and argue that they could vary independently. But, again, you’ll have to take a stand on what matters for well-being. Or argue that they are both important, but incommensurable. Alternatively, you could argue that both sets of data have their problems. This is my view. In this case, for the reasons Wakefield and Horwitz lay out, I am more skeptical of the depression data than the happiness data, which I think is likely to be most accurate at picking up changes at the bottom, since bad feelings are more psychologically salient and available than good ones, and therefore more likely to be accurately reported.

My hunch: much depression is misdiagnosed for failing to distinguish between functional sadness and disordered malaise, as above. And much is more or less intentionally misdiagnosed in order to give non-depressed people legal, insurance-covered access to SSRI’s, as Wakefield and Horwitz describe elsewhere. Importantly, SSRI’s really do make people feel better. Since there are far fewer truly depressed people than we think there are, but prescriptions for SSRIs based on bad depression diagnoses continue to rise, increasing rates of diagnosed depression may actually correlate with an improvement in the average tone of experience. O Brave New World!

Good, but somewhat dated backgrounder in Psychology Today on dopamine:

Dopamine now seems everywhere in the brain: running through four main brain pathways, picked up by five different types of receptors–each with several subtypes, many still just being defined. Suddenly, the neurotransmitter is the target of research into happiness, attention, extroversion, self-confidence, and goal-direction.

“Dopamine, how do I love thee? Let me count the ways,” jokes George Koob, Ph.D., a professor of neuropharmacology at the Scripps Research Institute in La Jolla, California. Excitement about dopamine is now so high, says Koob, that the danger is not underestimating its reach but exaggerating it: “Today’s gig is that dopamine is a kind of everyman’s neurotransmitter because it does everything. And the fact is, it doesn’t.”

I keep stressing the biochemical complexity of good feelings. Serotonin does one thing, dopamine another, etc., and there may be tradeoffs among each kind of good feeling. Except it’s not like each neurotransmitter is responsible for just one kind of feeling. Each actually does a lot of things, and so knowing the levels of certain neurotransmitters tells us less than we might think in the absence of further knowledge about the activity levels of various kinds of receptors. I guess it would be convenient if there was some single substance, “happy juice,” coursing through our veins that could be measured, but there isn’t. The more we know, the more complex it gets.

Economists Andrew Oswald and Andrew Branchflower begin a very interesting new NBER paper [$$$] on the relationship between levels of self-reported happiness and blood levels with this dubious claim:

For effective social and economic policies to be designed, it is necessary for policymakers to be able to measure human well-being.

They better hope they’re wrong, because if they’re right, then effective social and economic policy cannot be designed! Oh no! 

Why? Two reasons:

(1) Human well-being, as opposed to the several dimensions or components of well-being, is pretty much impossible to measure.

Why? Because the specific nature of human well-being is relative to the individual and the components of well-being are diverse and must often be traded against one another. 

What does this mean? Let’s start with the relativity of well-being. The achievement of valued aims (meaningful goals, important personal projects, whatever you’d like to call it) is a component of human well-being if anything is. However, the content of valued aims varies from person to person. It follows pretty straightforwardly that the specific requirements of well-being vary from person to person.

(For those of you on the lookout for the scourge of “post-modernist relativism,” please note that this kind of “relativism” is in fact a kind of relativism, and is also completely innocuous, entailed by the uncontroversial fact that different people have different personalities, different tastes, and different “callings.”)

Next, consider the diversity of the components of well-being and the potential conflicts between them. Health and longevity are components of well-being if anything is. But so is the individual achievement of valued aims. Some people’s perfectly reasonable aims  may be incompatible with maximizing their health and longevity. Imagine a cholesterol-saturated gourmand who would rather die than give up his foie gras, or an adventurer who draws profound meaning from facing down life-threatening challenges. So… how much weight do we give to one component of well-being —  health and longevity, say – relative to another — for example, the achievement of valued aims that conflict with maximal health and longevity? The answer is that there is no answer — no answer science and empirical evidence compels us all to agree on, at any rate.

The upshot, then, is that while we can measure various dimensions or components of well-being — whether it be health and longevity, the experience of pleasure, a sense of self-efficacy and control, the development of basic human capacities, or the achievement of valued aims — we cannot measure well-being as a whole because Mother Nature has nowhere posted a table of exchange rates between the various values that compose individual welfare. It’s simply not out there for the scientist to find.

Now, there may be a rough cultural consensus at any time and place about the relative weight to place on competing individual welfare-constituting values. But this consensus, to the extent that there is one, has to be discovered, and changes as time goes by. So, at this point, we’re not “measuring well-being” so much as attempting to find some bit of overlap if people’s conceptions of well-being. We can use the overlap to base a few general principles of mutually beneficial social interaction almost everyone will be willing to affirm. But the larger and more diverse the society, the smaller and more general the overlap. There are always broad swathes of often heated disagreement in pluralistic societies. And that’s what principles and institutions of liberal neutrality are for: to peacefully accommodate the inevitable lack of consensus about questions of value in open, cosmopolitan societies.

Would you say that a set of policies were “effective” if it peacefully and stably coordinated the behavior of millions of individuals in pursuit of their valued aims, and constantly increased their capacity to to realize them, despite the fact that there are as many conceptions of well-being as there are people?  Would you consider such a set of policies “effective” even if we didn’t know how to measure human well-being scientifically? 

(2) Policymakers have no incentive to accurately measure human well-being – even if it was accurately measurable — or to appoint, or take counsel from, those who do.

Lucky for them, Branchflower and Oswald begin their paper with a monumental falsehood. Their introductory proposition implies, among other things, that effective social and economic policy never has been designed! Their general idea, I take it, is that in order to design something effective, you have to be able to measure “effectiveness.” This may well be, but it

But this is also false. There are many things that work well without most of us knowing how or why — without know the meaning of “well.” And there are ways of “designing” through trial and error that delivers results without delivering knowledge of the mechanisms.  

I think my considerations (1) and (2) imply that not only does effective policy not require that policymakers are able to definitively measure well-being, but that effective policy is much more likely if we fully grasp the indisputable empirical facts that conceptions of well-being (and of “effective”) are plural (and this is so whether or not I am right on the philosophical point that the constitution of well-being for each individual requires trade-offs between different dimensions of well-being) and that policymakers are neither scientists nor reliable consumers of science. 

Maybe it is disappointing to social scientists — frustrating even! – to face up to the fact that no interest or competence in social science whatsoever is required for a hugely successful career as a policymaker, which is to say, as a politician or bureaucrat. This is even the case in places where social science flourishes most! Disappointing as that fact may be, social scientists may want to take it into account when  thinking about the design of effective policy.  

Time to pick up the pace with the happiness blogging. If I’m not doing a lot of thinking, at least I can be doing some linking.

Some Dark Thoughts on Happiness, by Jennifer Senior

More and more psychologists and researchers believe they know what makes people happy. But the question is, does a New Yorker want to be happy?Ab

Positive psychologist Chris Peterson on ABC News piece about measuring happiness, with a link to the survey on the Authentic Happiness site.

Dogs Come Top Of The Human ‘Happiness Index’.

There have been a ton of pieces about the Kahneman, Krueger, et al paper in Science. It’s far less exciting than it looks, and a little bit confused, I think. They have found, basically, that a backrub doesn’t feel any better when you’re rich. But they take this to mean that wealthier people are victim of a kind of “illusion” when they report “very happy” at higher rates. But is the integration of moment utilities have a better claim to “happiness” than global life-satisfaction judgments? If so, why? Argumentum ad Bentham? My take: a cultural constructivist emotional syndrome conception of happiness helps explain both the experience sampling and the life satisfaction survey results. Whoops! Thinking.

Haven’t read it yet, but I’m seeing it all over. The useful thing about the new study is that it compares life-satisfaction survey answers with Kahneman’s day reconstruction method (basically, writing down what you did and how it made you feel at the end of the day. Highlights:

“If people have high income, they think they should be satisfied and reflect that in their answers,” Krueger said. “Income, however, matters very little for moment-to-moment experience.”

And

“Despite the weak relationship between income and global life satisfaction or experienced happiness, many people are highly motivated to increase their income,” the study said. “In some cases, this focusing illusion may lead to a misallocation of time, from accepting lengthy commutes (which are among the worst moments of the day) to sacrificing time spent socializing (which are among the best moments of the day).”

My take: Not surprising. Life satisfaction judgments are going to reflect widely shared cultural assumptions about happiness. Wealth is part of a widely shared conception of the good life. Or maybe people think they ought to feel better about higher relative position, but this doesn’t really enter much into experience. So the self-reported happiness gap between the wealthy and less wealthy will shrink the closer the self-report gets to actual events.

By the way, funny how these things are reported. Why not “Good news! The least wealthy are only 12 percent less happy the wealthiest.” Dueling political inferences: money doesn’t make the rich happy, so it doesn’t hurt them if we take their money vs. not having money doesn’t make the poor unhappy, so it doesn’t help them to give them any. Take your pick.

Kahneman’s Benthamism prevents him from reading happiness-motivation splits as anything but “illusion.” If you simply assume that people value only happiness intrinsically, and desire other things only for the happiness it brings, then working hard to make money, say, will seem like a kind of mistake if money doesn’t maximize happiness. There are a couple alternative interpretations. One is that here we have a revealed preference for something other than happiness. Even if people say they’re trying to be happy, talk is cheap. Action shows us what is genuinely valued—not happiness. Or it may be that there is a kind of “illusion” here, but a good illusion. Our system maybe does this: Identify something valuable. Conclude that it will make us happy, which motivates us to go after it. Go after it. Get it. It makes us happy or not. Whether it does is irrelevant to the system.

Now, if that’s the way the system works, is it really an illusion, exactly? Compared to what? The way the system doesn’t work? Can the fact that we are motivated really be a trick? Maybe. No doubt our Darwinian system “wants” things we don’t. Do I really want pretty women? Do I really want the higher status that will help me get pretty women? If the prospect of happiness is a trick to get us to want stuff that Nature needs us to get, are we really sure we really want happiness after all. We find out that it’s the basis of the motivating trick, and we still want it? Becuase that’s what we’re built to want, even if it jerks us around? What if I can’t stop wanting misery-making pretty women and also can’t stop wanting happiness? Screwed? That’s life?

If the at-a-time gap in happiness between rich and poor is smaller than we thought, is the hedonic effect of relative position smaller than we thought.

By the way, this study involves women only. It will be interesting to see whether there are significant male-female differences with the day reconstruction method.

Economic Policy and the Level of Self-Perceived Well-Being: An International Comparison by Tomi Ovaska and Ryo Takashima, the Journal of Socio-Economics 35 (2006) 308-325.

An excellent econometric paper.

Best part:

Contrary to the findings on political freedom, economic freedom was found to be statistically significant in nearly all estimations, and of the positive sign. Furthermore, the relationship between SWB and economic freedom depicted in Fig. 2 largely held, unlike that for income, after controlling for alternative explanations of well-being. For instance, when the economic freedom index average in the sample rises from 5.76 to 6.34, the happiness levels rise from 3.01 to 3.07. The effect on life satisfaction is identical. The results suggest that people unmistakably care about the degree to which the society where they live provides them opportunities and the freedom to undertake new projects, and make choices based on one’s personal preferences. Compared to the GDP per capita measure, the index of economic freedom – personal choice, freedom to compete and the security of privately owned property as its core components – turned out to be about four times as important, as measured by elasticities. This indicates that the newly found interest of economics and of policymakers in measures of institutional quality is well placed. Based on the regression results, economic freedom holds some promise in serving as one of the policy tools that could be potentially used to increase the SWB of a nation’s population.

The explanation of why political rights (i.e., voting rights) had a negative influence was interesting:

It turned out that though insignificant, political rights had negative sign while civil liberties had a positive sign. This may be a reflection that democracy is not ideal as a collective decision making mechanism. As one can see from the median voter model of public choice theory, only the median voter obtains satisfaction from political rights. In addition, unlike market exchanges, every majority voting decision can potentially create a relatively large number of losers, all those who were in minority. However, the positive sign of civil liberty indicates that regardless of political structure, civil liberties are essential for human beings across society.

They also note that general freedom is likely prone to adaptation. We take it for granted. Sound’s right to me.

The Pew poll mentioned below confirms a longstanding trend: Republicans say they are happier than Democrats. This year, 45% of Republicans said they were “very” happy as opposed to 29% of Democrats. That’s a big gap! Here’s the the 30+ year trendline from Pew:

This stability is interesting in part because, I take it, that the demographic composition of Republican and Democratic voters has changed not insignificantly over the last 30 years. Is that right? Anyway, what accounts for Dem.-Rep. gap? Well, it’s not income. Republicans report themselves as happier at all points on the income distribution, as this Pew graph shows:

So what’s the deal? Here’s the Pew folk

[The regression] analysis shows that the most robust correlations of all those described in this report are health, income, church attendance, being married, and, yes, being a Republican. Indeed, being a Republican is associated not only with happiness, it is also associated with every other trait in the cluster.

Clean-livin’ Christians are more likely to be in good health, go to church, be married, and vote Republican.
What doesn’t the study mean? In today’s Colorado Springs Gazette, hometown paper of the Focus on the Family folk, I am quoted thus:

Does membership in the GOP really make people happy? Probably not, said Will Wilkinson, who studies happiness for the Cato Institute. The bliss is probably connected to some other facet of life that also inclines people to be Republicans, he said.

“People might read that and say, ‘I’d like to be happy, maybe I should be a Republican.’ It definitely doesn’t mean that,” Wilkinson said.

Sorry Dobsonites!

Assuming that the entire Dem.-Rep. difference doesn’t disappear when controlling for demographic variables, what psychological traits would you guess predict both higher self-reports and Republicanism?

The Pew Research Center released a big report today on happiness in the U.S., “Are We Happy Yet.” I’m just now digging in, but I’ll have a lot to say about it tomorrow, I’m sure. For now, let me just leave you with this:

Another great thing about chatting with Carl the other day is the pointer he gave me to the work of Dan Haybron, a philosopher at St. Louis University. Dan has written a couple of the papers that I’ve been trying in vain to find. His web page is a treasure trove. His paper Do We Know How Happy We Are: One Some Limits of Affective Introspection and Recall makes the skeptical case I have been trying to make, based on the same research I have been looking at, much better than I have so far been able to make it. I’m delighted to see this paper in part because it helps me know that I’m not crazy.

ABSTRACT. This paper aims to show that widespread, serious errors in the self-assessment of affect are a genuine possibility—one worth taking very seriously. For we are subject to a variety of errors concerning the character of our present and past affective states, or “affective ignorance.” For example, some affects, particularly moods, can greatly affect the quality of our experience even when we are wholly unaware of them. I note several implications of these arguments. First, we may be less competent pursuers of happiness than is commonly believed, raising difficult questions for political thought. Second, some of the errors discussed ramify for our understanding of consciousness, including Ned Block’s controversial distinction between access consciousness and phenomenal consciousness. Third, empirical results based on self-reports about affect may be systematically misleading in certain ways.

The abstract doesn’t really capture the core of what I’m interested in here, which is the reliability of self-report survey instruments. The paper contains a very trenchant and cogent critique.

Now, I’ve been arguing that the happiness surveys fail to measure increases in average objective happiness. I suppose it reveals my priors to admit that it really hadn’t seriously occurred to me that they could be failing to measure decreases. Haybron seems to think this is a distinct possibility.

Here is a Haybron’s conclusion:

There is a family I know—I will call them the Wilsons—whose members are quite amazingly loud. Wonderful people they are, but the din from their constant shouting, thumping, and crashing about is, for the unseasoned visitor, almost unbearable. Yet they seem to have no idea there’s anything at all unpleasant or odd about it, since it is perfectly normal for them. Those who know them see it differently: however hardened their sensibilities might have become, it’s almost certainly an unpleasant place for the family too. (It must be.) It is worth pondering whether mainstream American society might not be a little like the Wilsons: oblivious, and more or less inured to, a noisy, obnoxious, stressful, and spiritually deflating way of life.

Of course,Haybron’s priors are revealed in the fact that he doesn’t seem to have considered that we might be rather better off than we think. This kind of dispute brings home, I think, the need for a long-term longitudinal physical correlates of happiness study. My guess is that some correlates of unhappiness (stress/cortisol levels, e.g., ) may have gone up, but that some correlates of happiness (some kinds of dopiminergic activity, e.g.) may have also gone up. The multi-dimensional physical constitution of real happiness will complicate efforts to show unambiguous increases or declines, especially since there may be no generally valid way to weigh the disutility of cortisol against the utility of dopamine, or whatever, in terms of real happiness.

I just got off the phone with Carl Craver, a smart philosopher of neuroscience (yes, redundant) at Wash U in St. Louis. I had some vague ideas about brains and happiness and I wanted to talk to somebody who not only understands brains, but philosophy of science, and so forth. In trying to formulate one of my vague ideas to Carl, I think I semi-successfully clarified something worthwhile to myself. It’s not what I was trying to clarify, but I’ll take it! Thanks, Carl!

So . . . here’s a datum that needs explaining:

Self-reported happiness is stable over the past 50 years–the percentages of the population reporting themselves in each category has not shifted significantly.

Here are two hypotheses that account for this fact:

(1) Adaptation, aspiration, and/or social comparison affect the real qualitative feel of subjective states, such that the way people feel now (in the various categories in the distribution) is essentially the same as the way people felt fifty years ago.

(2) Adaptation, aspiration, and/or social comparison affect the way people report
the way they feel, such that the percentages of people who say they feel “very happy,” etc. remain pretty constant, although the real qualitative feel of their subjective states now (in the various categories in the distribution) is not essentially the same as it was fifty years ago. More people are in fact happier now, but the reporting mechanisms keep moving the goal posts.

My gut says strongly that (2) is correct. This is not to say that adaptation, etc. do not at all affect the real quality of our subjective states. I think they do. But not enough to have kept the real quality of happiness totally static over time. (Also, it might turn out that, say, adaptation is a real effect, while social comparison is a reporting effect or vice versa. But I don’t want to get too complicated just now.)

How do you test this? Well, is it really that hard? There is ample reason to believe that self-reports contain real information. However, I suspect that the information they do contain is not not very usefully comparable across time and/or place. Nonetheless, we can say with a high level of certainty—due to various kinds of self-report (there is no other way)—that certain hormones and neurotransmitters, etc. correlate with feeling good, and others correlate with feeling bad. Seratonin, dopamine, oxytocin: good. Cortisol, etc.: bad. Same with certain distinctive patterns of neural activation. My friend Paul Zak takes blood samples and measures oxytocin levels to see how trusting people are. (He doesn’t use self-reports, but real performance in economic games containing an assurance problem. It should also be noted gratuitously that Paul is one of Wired’s 10 Sexiest Geeks for 2005.) It should in principle be possible to measure the quantity of particular substances in people’s system, or the activity levels of certain parts of the brain (generally involving a number of these substances) as a proxy for the way people really feel, as opposed to the way they say they feel.

So here’s the idea: Get a good sized random sample of people in a particular society (or several societies). Measure their happiness-relevant vitals again and again over time—say, twenty years—and see what you get.

My predictions:

(a) There are multiple bases for good and bad self-reports. For example, some “very happy” people may have very consistently low cortisol levels. (Buddhist happy.) Some “very happy” people have very high status-related seratonin and testosterone levels, with a moderately high amount of cortisol. (Big honcho happy.)

(b) Many of the variables that predict high self-reports, such as income, autonomy, sociality, etc., will be shown to correlate with slightly different physical bases of good feelings. Some variables will be more seratonin related. Some variables will be more oxytocin related. Etc.

(c) The composition of the physical basis of high self-reports changes as we age.

(d) Over time, we will see shifting of the distribution of different kinds of happiness (e.g., Buddhist happiness vs. big honcho happiness) within the self-report categories due to changes in cultural, social and economic institutions.

and, finally,

(e) in year twenty (assuming social stability and a continuation of the general trend in economic growth) the percentage of the population having the physical profile(s) that predicted “very happy” in year one will have increased significantly, but the self-reports will not reflect this change.

There’s probably already good evidence for (a)-(c). But let’s really find out.

My intutions here were heavily primed by reading Fogel’s The Escape from Hunger and Premature Death, 1700-2100. Fogel advances a very physical conception of economic productivity in terms of calories consumed and calories spent. I was astonished to see the huge spike in economic productivity with the discovery of the germ theory of disease and the advent of adequate sanitation. Prior to this, almost everyone had some kind of infection almost all the time, and a big portion of the calorie budget went into fighting infection, and not productive labor. If you’re not constantly sick, you have more energy and can work harder longer.

The thing that struck me is that people who were sick all the time cannot have really felt all that well. But people who were sick all the time wouldn’t have a good idea of what it meant to feel not sick all the time, either. That’s just the way things were. And I suspect that had folks in 1880 or whenever answered happiness surveys, they’d mostly say they were doing pretty good, like now. I bet you’d see a upward shift in the self-reports with good public health measures. But I highly doubt the shift would really correspond to the real change in what it felt like on the inside to move from really high to pretty low rates of infection. (I’d like to know the physical correlates for the lousy feelings of bacterial and viral infection. Couldn’t we measure those, too?)

So, there’s a research program for the taking! If you’re a super-rich patron looking to make a big contribution to the science of human well-being, well, you know how to reach me!

Thanks to commenter Conchis for bringing this important paper to my attention.

Heterogeneity in Reported Well-being: Evidence from Twelve European Countries by Andrew Clark, Fabrice Etilé, Fabien Postel-Vinay, Claudia Senik, Karine Van der Straeten

CONCLUSION. This paper modelled the relationship between income and self-reported well-being using random-effect techniques applied to panel data from twelve European countries. We show that people are different, and in more complicated ways than just having different intercepts. We are not able to distinguish between heterogeneity in the utility function (translating income into utility) and heterogeneity in the expression function (turning utility into reported well-being). We can, however, strongly reject the hypothesis that individuals carry out these joint transformations in the same way.

We identify four classes of individuals, and show that the “marginal well-being effect of income” is very different across these classes. In particular one class is satisfied and has a high marginal well-being effect, while another is dissatisfied and has a low marginal wellbeing effect. Descriptive statistics reveal demographic and country patterns between classes. This has at least two important implications. First, in a political economy sense, as the effect of income differs sharply across classes (and classes are not independently distributed between countries), we would expect average opinion regarding economic policies to differ across countries. . . This is a subject for ongoing research.

Perhaps more importantly, our results suggests that aggregating data across diverse populations may be a dangerous practice. Individuals, who seem to fall naturally into a number of different classes, differ in ways that are far more complicated than those picked up by a simple fixed effect. The trend towards comparative research in social science, whereby data from different countries are compared, is laudable. Nonetheless, our results suggest that the blind aggregation of diverse populations risks producing empirical results that are false for everybody.

The econometrics in the happiness literature is clearly edging its way toward adequacy. What we have had is not quite useless. But now, minimally credible statistics that take into account individual variation throws a huge wrench into attempts to ground highly specified policy on happiness studies. Heterogeneity is very problematic for would-be benevolent technocrats. First, they’re stuck with a big information problem. They’d need to know about dynamiclly shifting individual differences at a finer grain of detail than is possible. Second, heterogeneity prevents you from basing policy on the welfare of an “representative” agent. If it is true, and it is true, for example, that income has a big positive effect on some people’s happiness, or that simply keeping very busy has a big positive effect, then tax policy aimed at incentivizing the consumption of leisure against the production of wealth is going to cause a lot of immediate micro-harm, not just the delayed macro-harm of slower growth. “Truces” in income/status “arms races” are very unlikely to be pareto improvements. (Which means the situation is not accurately characterized as an “arms race.”) This may not bother a Benthamite technocrat willing to overlook individual harms if the net utility is positive. But the Benthamite is still stuck with the informational problem. And ignoring the separateness of persons is terribly wrong, which is why Benthamism is false. Morally legitimate policy is left having to take pluralism and the separateness of persons into account, which generally means worrying primarily about the structure of the basic framework of interaction, trying to ensure that it is sufficiently general and neutral between reasonable sets of beliefs and preferences.

Also, back to the paper, note the inability to “distinguish between heterogeneity in the utility function (translating income into utility) and heterogeneity in the expression function (turning utility into reported well-being).” The recognition that there is a distinction between real well-being and self-reported well-being has big implications. We don’t know to what extent talk tracks fact unless we know how talk relates to fact. The recognition of variation in the expression function throws in a further wrinkle. If the expression function was homogenous, discovery of that function would allow us to infer real well-being from talk. But if there is variation in the reporting function, and that variation cannot be accounted for by a general law-like principle, then aren’t we getting pretty close to empirical proof of the near-uselessness of self-reports as a proxy for real well-being?

For Bob, income has a big positive effect on well-being, but, despite growing happiness from growing riches, Bob in his modesty won’t ever put himself in the highest category of reported well-being. Karen is sort of miserable, and higher income means almost nothing to her. But she’s a shallow, ideological, free-market economist, and worries that there must something wrong with her for not being happier given her wealth. So she says she’s “pretty happy” even though she’s not. Bob and Karen report themselves as being in the same category of self-reported well-being. What do we know about well-being? What do we know about the relationship between income and well-being?

Can People Feel Happy and Sad at the Same Time?, by Jeff T. Larsen, A. Peter McGraw, John T. Cacioppo

ABSTRACT. The authors investigated whether people can feel happy and sad at the same time. J. A. Russell and J. M. Carroll’s (1999) circumplex model holds that happiness and sadness are polar opposites and, thus, mutually exclusive. In contrast, the evaluative space model ( J. T. Cacioppo & G. G. Berntson, 1994 ) proposes that positive and negative affect are separable and that mixed feelings of happiness and sadness can co-occur. The authors both replicated and extended past research by showing that whereas most participants surveyed in typical situations felt either happy or sad, many participants surveyed immediately after watching the film Life Is Beautiful , moving out of their dormitories, or graduating from college felt both happy and sad. Results suggest that although affective experience may typically be bipolar, the underlying processes, and occasionally the resulting experience of emotion, are better characterized as bivariate.

This is a cool paper. It suggests lots of interesting hypotheses. They mention that co-activation of positive and negative affect is likely to be dissonant, unpleasant, and short-lived. It strikes me that good literature, for example, may turn on the artful co-activation of negative and positive affect systems. Might an “interesting” life have to do with relatively often occupying complex precincts of the evaluative space? Worthwhile.

This article in the NBER Digest presents a nice summary of Oswald and Branchflower’s paper Happiness and the Human Development Index: The Paradox of Australia in which they criticize the UN’s Human Development Index for failing to incorporate SWB data.

Happiness measures, Blanchflower and Oswald add, “can tell politicians and others how citizens value the different effects upon well-being of diverse influences such as unemployment, the divorce rate, real income, friendship, traffic jams, crime, health, and much else. If we can learn to exploit the power of statistical happiness equations, it should be possible to make public policy choices in a more coherent way than before.”

Branchflower and Oswald seem to make the apparently common error of seeing averages of the way lots of people filled out happiness self-report surveys with “how citizens value” this or that. It just doesn’t follow.

For example, divorce rates. Do people really value the effects of the divorce rate? (”Why so sad, Barry?” “Oh, the divorce rate went up this year.” WTF!?) What people want to know is whether their divorce (or their sister’s or friend’s) will make them and/or their kids happier. And, of course, it depends. For some, divorce is a huge relief, a new chapter, etc. For other’s it’s the end of the world. If we toted up the self-reported SWB of all divorced people to tell whether, on average, divorce makes you happier or not, we’d know almost nothing worth knowing, given the likely high variance in the sample. What’s worth knowing is whether YOU are the sort of person who will gain or lose. We couldn’t care less if the extremes of misery are more acute than the extremes of elation, producing a mildly negative average.

And, anyway, how happy a divorce is going to turn out for you just isn’t that important. Maybe it’ll make you miserable, but unless you get a divorce from the controlling creep, you’ll never become the person you need to become. Maybe the problem is precisely that you like the lack of responsibility that comes from being controlled. The fact that you’re NOT going to like leaving is why you should. Or maybe you’ll be miserable if you stay in the marriage, but you should just suck it up, because the kids need you there, and you made a promise that you should keep. Knowing the happiness averages helps us frame divorce policy how? And what does the self-reported SWB of everybody—divorced, affected by divorce, totally untouched by it—correlated to the divorce rate tell us. Probably nothing, and certainly not how “citizens value” the effects of the divorce rate. Definitely nothing that enhances our ability to make more coherent policy choices.

Will I ever tire of noting that happiness and value can come apart? Probably not. Will I ever tire of noting that self-reported well-being, which is a poor proxy for happiness, may have only an extremely tenuous relationship to value? Probably not. Will you tire of being reminded that poor proxies for things that may or may not be morally relevant in particular cases are unlikely to be extremely useful tools in our efforts to engineer a better society. Probably. But until you do . . .

More from the direction of causation annals. From the Independent:

Researchers from the universities of California, Missouri and Illinois examined connections between desirable characteristics, life success and well-being in more than 275,000 people.

They found that happy individuals were predisposed to seek out new goals in life, leading to success, which also reinforced their already positive emotions.

The psychologists addressed questions such as whether happy people were more successful than unhappy people, and whether happiness came before or after a perceived success.

Writing in Psychological Bulletin, published by the American Psychological Association, they concluded that “chronically happy people” were generally more successful in many areas of life than less happy people.

Good name for an album. Also, supposed to be one of the key pieces of evidence providing objective corroboration for happiness self-reports. Duchenne smiles are “real” smiles, the ones you make when someone cracks a funny joke and your eyes wrinkle and everything, not phony prom picture smiles. Some research has shown that expressing Duchenne smiles is correlated with heightened activity in the font left part of the brain, which is associated with experiences of pleasure. And people who Duchenne smile more report that they are happier.

But what are we to make of this, from Paul Griffiths and Andrea Scarantino, “Emotions in the Wild: The Situated Perspective on Emotion,” forthcoming in the Cambridge Handbook Of Situated Cognition, Robbins, P and Aydede, M (Eds):

One of the most important experimental paradigms for a situated perspective on emotion is the study of ‘audience effects’ – differences in emotional response to a constant stimulus which reflect differences in the expected recipient(s) of the emotion. Amongst the most dramatic effects are those obtained for the production of the so-called ‘Duchenne smile’ – the pattern of movement of mouth and eyes generally accepted as a pan-cultural expression of happiness (Ekman, 1972). Ten-pin bowlers are presumably happiest when they make a full strike, less happy when they knock down a few pins. However, bowlers rarely smile after making a full strike when facing away from their bowling companions and smile very often after knocking down a few pins when they face their companions (Kraut & Johnston, 1979). Spanish soccer fans show a similar pattern in their facial response to goals, and issue Duchenne smiles only when facing one another (Fernández-Dols & Ruiz-Belda, 1997). Fernández-Dols & Ruiz-Belda also demonstrate that at the 1992 Barcelona Olympics, although Gold medalists produced many signs of emotion during the medal ceremony, they produced Duchenne smiles almost exclusively when interacting with the audience and officials.

These results suggest that smiles are not outpourings of happiness which are merely witnessed by other people, but rather affiliative gestures made by one person to another with respect to something good which has occurred. This fits the model of emotions as strategic moves in the context of a social transaction. Obviously, people do smile and produce other classical emotional expressions when they are alone, but studies suggest that they do so far less often than one might expect. Even such apparently reflexive displays as facial expressions produced in response to tastes and smells appear to be facilitated by an appropriate social setting and the same appears to be the case for pain expressions (Russell, Bachorowski, & Fernández-Dols, 2003). Furthermore, it would be a mistake to conclude that audience effects are absent when a physical audience is absent. Solitary subjects who mentally picture taking part in a social interaction produce more emotional facial signals than subjects who focus on the emotional stimulus without an imagined audience. Fridlund has described this as ‘implicit sociality’ and remarked that his experimental subjects display to the ‘audience in their heads’ (Fridlund, 1994; Fridlund et al., 1990).

If Duchenne smiles are subject to audience effects, and especially if they are strategic (which is not to say they are contrived or fake–one of the author’s main points is that emotions are there to strategically manage social interaction), we maybe we shouldn’t make that much of them.

If Griffiths is right, “being happy” may be a longstanding social strategy. Take a different example to get the idea or emotion as strategy. Ever met anyone who cries any time they don’t get their way, or who grows sullen whenever there is conflict. This is, in part, a way of managing other people. Similarly, someone who is “happy” may depend strategically on behaving happily (again, this is not faking it) in order to seem likable, to reduce social friction, or maybe as a way to mask latent malevolence or aggression. We should expect that people for whom happiness plays this kind of strategic social role to both earnestly smile more and to have a self-image as a happy person. This should show up on the happiness survey. And it is plausible that being “a happy person” would require the activation of the neural correlates of positive affect.

However, just as we wouldn’t want to say that the unsmiling Olympic champion is not happy, I don’t think we’d want to say that people for whom happiness is not so strategically central are not equally happy in some deeply important sense, even if they’re not so often getting that tickle in the left frontal lobe that goes along with the Duchenne smile.

The Griffith essay is full of good examples of strategic emotion. People get angrier when their anger can get them restitution. Embarrassment as a signal of one’s realization that one has violated a norm. I like the account of sulking:

Sulking is a behavioral strategy for seeking a better deal in a relationship – an emotional game of ‘chicken’ in which transactions that benefit both parties are rejected until appropriate concessions are obtained. The question confronting an agent deciding whether to become upset in this way is not whether they have been slighted simpliciter, but whether taking what has happened as a slight and withdrawing cooperation will give them leverage. Once again, this strategic appraisal of the situation may be realized by a relatively simple mental mechanism.

I’d very much like to see a strategic account of happiness. Maybe it will turn out that some of our gains in self-reported happiness come from increasingly widespread adoption of the “being happy” social management strategy. And it may be a very good thing to have a society where “being happy” is a good way of getting what you want, even if “being happy” isn’t the same thing as the sense of meaningful well-being we all hope for. Fun stuff!

Last night, I discovered a wonderful passage from Thomas Mann’s short story “Little Herr Friedemann,” in which the title character, crippled from infancy, reconciles himself to the fact that he will never know sensual love, but instead learns to take value in the full range of human emotion. Indeed, he becomes a connoisseur of feeling, sensitive to the subtle notes of value even in unhappiness.

Is not life in and for itself a good, regardless of whether we may call its content “happiness”? Johannes Freidemann felt that it was so, and he loved life. He, who had renounced the greatest joy it can bring us, taught himself with infinite, increidble care to take pleasure in what it still had to offer. A walk in the springtime in the parks surrounding the town; the fragrance of a flower; the song of a bird–might not one feel grateful for such things as these?

And that we need to be taught to enjoy, yes, that our education is always and only equal to our capacity for enjoyment–he knew that too, and he trained himself. Music he loved, and attended all the concerts that were given in the town. He came to play the violin not so badly himseld, no matter what a figure of fun he made when he did it; and took delight in every beautiful soft tone he succeeded in producing. Also, by much reading he came to possess a literary taste the like of which did not exist in the place. He kept up with the new books, even the foreign ones; he knew how to savor the seductive rhythym of a lyric or the ultimate flavour of a subtly told tale–yes, one might even call him a connoisseur.

He learned to understand that to everything belongs its own enjoyment and that it is absurd to distinguish between an experience which is “happy” and one which is not. With a right good will he accepted each emotion as it came, each mood, whether sad or gay. Even he cherished the unfulfilled desires, the longings. He loved them for their own sakes and told himself that with fulfillment the best of them would be past. The vague, sweet, painful yearnings and hope of quiet spring evenings–are they not richer in joy than all the fruition the summer can bring? Yes, he was a connoisseur our little Herr Friedemann.

It’s the last paragraph here that demands our attention. But it requires the preceeding to make full sense. For the point is that the capacity to take full enjoyment in–as opposed to experiencing a superficial, stereotyped positive reaction to–such obviously pleasant things as flowers and birdsong implies the capacity to take pleasure also in frustration and pain. Once our sentimental education attunes us the finer grain of experience, there is no facile distinction between the happy and unhappy.

And so suppose you gave a happiness survey to little Herr Friedemann. What would he say? Let’s look at the next paragraph:

But of course they did not know that, the people whom he met on the street, who bowed to him with the kindly, compassionate air he knew so well. They could not know that this unhappy cripple, strutting comically along in his light overcoat and shiny top hat–strange to say, he was a little vain–they could not know how tenderly he loved the mild flow of his life, charged with no great emotion, it is true, but full of a quiet and tranquil happiness which was his own creation.

So, Herr Friedemann is happy. But his happiness is not the aggregate of happy feelings, but is constitituted even by unhappiness, pain, and frustrated longing. How is this possible? All the flow of life is tranformed–created–into tranquil happiness by his attitude of tender love toward his complete experience. Perhaps the most precious thing is the alchemical education of sensibility that can turn emotional lead to gold. And so shouldn’t we note the flaw in a measurement instrument indifferent between gold seen through shallow waters and gold mined from depths where there is no gold?

An email exchange with Virginia Postrel prompted me to note for the first time that there’s a real problem in trying to use the happiness data and the depression data at the same time. Almost all the “paradox of prosperity” books cite happiness data that is supposed to show that we’re not getting much, if any, happier as we grow wealthier as a society. Alongside, they cite data purporting to show that depression is approaching epidemic proportions. But the happiness and the depression data contradict each other. You can’t in good conscience appeal to both.

The “paradox” books generally show a graph illustrating the stability of the percentage of people reporting that they are “very happy.” But they could also show a similar graph of the stability of people reporting that they are “not too happy” or “not at all happy” (depending on the survey). For example, Branchflower and Oswald, “Well-Being Over Time in Britain and the USA,” show that, in the US, the number of folks reporting that they are “not too happy” (on the three option surveu) dropped from 14% in the 1972-1976 period to 12% in the 1994-1998 period (which is up from the 1988-1993 low of 10%). Similarly, in Britain, the number reporting “not at all” and “not very” (on the four option survey) was 4% and 11% respectively in the 1972-1976 period, and 3% and 10% in the 1994-1998 period.

If depression rates have been rising stratospherically, how come the happiness surveys don’t track this fact at all, but instead show a small decline in the number of dissastified people? On its face, it looks like the happiness data contradicts the depression data. Indeed, the happiness data would seem to support the Horwitz-Wakefield hypothesis that the apparent increase in depression is almost entirely a function of the flaccidity of the diagnostic category.

One way of addressing the apparent contradiction, suggested by Virginia in her review of Gregg Easterbrook’s The Progress Paradox, is to point out that depression and unhappiness are very different things. Virginia points to Julian Simon as an example of a happy person who stuggled with depression. It is indeed plausible that some genuinely depressed people, for whom life is otherwise going very well, will experience their depression as an external malady–an alien force unprompted by events–and so will nevertheless report that they are happy.

However, it seems at least as likely that the depression will cast a shadow over life, and negatively effect one’s assessment of how well things are going. Indeed, it is not at all uncommon for depression to cause major problems in precisely those areas most important for the sense that one’s life is going well, such as work and love. So, if there was in fact a huge increase in depression, we should expect at least a significant fraction of those people to show up in the lowest category of self-reported happiness, even if many others are reporting that they are happy despite their depression.

Additionally, if an increasing number of people are simply functionally sad, as opposed to pathologically depressed, due to a death in the family, a bad break-up, or a lost job, you’d think that they’d show up in the happiness data. This is preciesly the sort of thing the happiness data is supposed to be measuring. It does, in fact, pick up the negative hedonic effects of higher employment rates, for example. There simply is no increase in subjective ill-being, as measured by the surveys.

Therefore, if you’re going to use the happiness data to show that we’re not getting happier as incomes rise, you can’t turn around and use dramatic depression data that, if accurate, ought to show up pretty dramatically in happiness data that, on the contrary, suggests a mild decline in the percentage of the depressed and sad. At the very least, you need some explanation of why this profound epidemic of depressive mental illness has, amazingly, absolutely no effect on the data from which you are drawing most of your conclusions.

I’m sort of amazed that this didn’t occur to me earlier. And I’m really amazed by the fact that the apparent inconsistency just doesn’t come up (I don’t think) in the “paradox” books. When you think about it, if there was really a widening plague of depression AND average happiness was stable, then you’d need need a growing cadre of really happy people to offset the growing legions of the depressed. But all the books make a show out of the fact that there isn’t a growing cadre of really happy people, and they all admit the average is stable. So you’d think the “rising depression” idea would raise a red flag or two.

I think it’s pretty clear that the depression data is junk. And that’s trouble for the “paradox” genre. If you’re unable to honestly say that we’re getting more miserable, you’re left with data that show that we’re slightly better off and no worse off than we’ve ever been, since we’ve been measuring this sort of thing, at least. And here’s a pretty hard argument to sell: We’re happy as ever. Therefore, radical interventions are required!

The Pittsburgh Tribune-Review has published an op-ed by moi about the way the overdiagnosis of depression leads to underestimation of how happy we are, and to books on the "paradox" of misery amid bounty. Check it out.