A bit more on the economics of happiness

I was discussing the economics of happiness with my son, and in particular the Easterlin paradox. Within a given country, people with higher incomes are more likely to report being happy. However, in international comparisons, the average reported level of happiness does not vary much with national income per person, at least for countries with income sufficient to meet basic needs. The same is true over time – average happiness levels don’t change much even as incomes rise.

This is often taken to mean that it’s relative rather than absolute income that determines happiness, so an increase in everyone’s income won’t make anyone happier. Hence, we shouldn’t worry so much about increasing income, but should focus more on factors likely to contribute to happiness. The point that struck me was that, given Easterlin’s data, the paradox is almost certain to apply whatever potential source of happiness we consider, in one form or another.

Let’s consider religious belief, for example and suppose that, in national surveys religious belief is correlated with (self-reported) happiness. But we know that countries differ widely in religiosity and not much in average happiness, so presumably it’s relative religiosity that makes you happy. In Australia, perhaps, going to church at Christmas and Easter is enough, while in the US maybe you have to go every week and pray daily as well. And of course, proselytising is a zero-sum game, since every convert reduces the relative religiosity of existing believers.

Of course, there may be sources of happiness for which there isn’t much variation across countries or over time, so that the generalized Easterlin paradox doesn’t apply. But if something has been observed to be constant across countries and over time (even though people have a very good reason to want more of it), it seems likely that it’s hard to change.

Overall, this leads me to think that the main implication from the happiness research program is one of stoic acceptance of the world as it is. Not being much of a stoic, I tend to the view that we should promote improvements in health, wealth and other good things, and let people take care of their own happiness. As Spike Milligan didn’t quite say “Money can’t buy you happiness, but it does buy a better class of unhappiness”.

32 thoughts on “A bit more on the economics of happiness

  1. John Says:

    Overall, this leads me to think that the main implication from the happiness research program is one of stoic acceptance of the world as it is. Not being much of a stoic, I tend to the view that we should promote improvements in health, wealth and other good things, and let people take care of their own happiness. As Spike Milligan didn’t quite say “Money can’t buy you happiness, but it does buy a better class of unhappiness�.

    That’s one way to read it. Another way to read it is that rich people’s happiness relies on other people’s unhappiness. Nothing new there, it’s just Veblen. 😉

  2. I can think of at least one other plausible mechanism for the effect: a sense of security. That is, relative income could correlate with margin of safety, and it may be that rather than the (also correlated) interpersonal comparisons that is showing up. To test this you’d have to make all sorts of purchasing power adjustments and adjustments for non-cash resources, etc.

  3. The other, I think more plausible, explanation is that people respond to survey questions about happiness in relative terms.

    So what the questions measure are not their absoulte level of happiness (which in a sense may not be measurable) but what they think their own happiness is within the range of possibilities they perceive as available.

    So in that situation, you would expect measures of happiness not to vary between countries and over (long periods of) time.

  4. Billions of human beings have come and gone in this world; all forgotten in time. What is it that gives us significance?

    I don’t know how truly happy were the ones who relied on some external formula or definition of happiness: but I do know that the ones who ‘felt’ happy inside, without a reason – their hearts full because they had opened them to the beauty within – were happy because the heart told them so. Mind and intellect had nothing to do with it; therefore –

    Reasoning, religion, economies, science, relationships etc not required – just the simple act of feeling the feeling within, accessible to each and every human heart. Some have called this knowledge happiness, some peace, some a journey, some Love – all have called it fulfilment!

    Prem Rawat is the best mirror of this message I have had the privilege to know:



  5. Urgh, now that I’ve accidentally posted the above I have to actually say something!!!

    What I was trying to get at is the idea of people are rationally driven to maximise utility or hapiness is overly simplistic.

    I don’t think this truly accounts for the way humans behave. This is the idea Freud was getting at in Beyond the Pleasure Principle. People’s behaviour, economic or otherwise, are not necessarily driven by the desire to maximise hapiness, relative to others or on average.

    In fact one of the ideas Freud picked up upon was the “compulsion to repeat”, and the way in which our unconcious drives us to repeat tasks, that may be pleasurable or unpleasurable for no apparent rational purpose.

  6. Once my basic needs are met with margin to spare, if I have access to books of an academic nature (and novels too!), good coffee, the company of friends and family, and maybe a laptop: I consider myself to be doing well in having all that. The rest is just gravy.

    Most people are driven by a strong sense of status – and that is not meant to imply anything bad – which leads to the `keeping up with the Jone’s’ syndrome. By status I am referring to the need to feel valued by peers and those of higher status. People also try to imply a status-level that they might not have attained; think of the young bloke in a new finance job who buys a flash car to impress the chicks…

    On that relatively unscientific analysis, happiness to whatever extent it can be measured and ordered, has an absolute component (water! food! shelter!) and a relative component (keeping up with the Jone’s, enjoying the benefits of an income, etc).

    Once the absolute component’s elements are satisfied with some margin, then we have a rough lower bound on happiness for people in developed nations.

    One other point with regards to money as a source of happiness. For people desperate to put food on the table a small increase in income may make all the difference, and increase “happiness” significantly. On the other hand, for the very rich, the importance of money is that having enough of it buys you a ticket into the club; increases or decreases of a few million here or there are of no consequence. The money to happiness curve is likely a lot nonlinear.

    On a closing note: Money doesn’t buy happiness, but how many unhappy rich people do you know?

  7. I think the one of the most interesting application of happiness economics would be its implications for income tax. From the literature it seems like there is a log-normal relationship between income and happiness, i.e. happiness initially increases rapidly with income and then the curve becomes steadily flatter. There must be a flexion point, below which a dollar buys more than one unit of happiness and above which it buys less. That would seem like a rational point to kick in the bottom income tax bracket.
    Of course, a progressive income tax could in itself contribute to a flattening happiness curve, as a dollar becomes worth less to those who lose more of it in tax. So there would need to be some study of these recursive effects.

  8. Milligan’s better quote on this (for mine) was “All I ask is the chance to prove that money can’t make me happy”.

  9. Could it be said that while increases in wealth may not lead to increases in happiness, declines in wealth may lead to declines in happiness? We tend to not notice slowly moving up and onwards, but seem to really hate it when we have to take a step back.

    I suppose a simple example might be an adult who is now on a high income after years of ascension might feel a similar level of happiness to when they were a student, but if they were now forced to live on a student income, then their happiness would decline.

    Not sure how to make sense of this, but it sounds intuitively sensible to me.

  10. My observation is that happiness is not linked that much to external events. Some people see the glass half full all their life, some always complain. Think of your own family: there are probably the two kinds of people, and you cannot correlate what happened to them to how they feel. However, there is some evidence that it is easier to be happy after 40 than when you are 18 years old: one learns to deal with life. The correlation of happiness with better circumstances, I think, is linked to a feeling of success or recognition: most of us look for that.

  11. Just four points quickly:
    1. As happiness is the ultimate objective, it is too important to just let people take care of it.

    2. While happiness data are not totally unreliable as some economists seem (or at least seemed) to believe, they are not perfect either. In particular, they may hide some actual improvement over time and do not facilitate interpersonal comparisons well. For details and for ways of improvement, see my paper in Economic Record 2008.

    3. Some factors (e.g. noise) important for happiness may be important absolutely as well as relatively and are not subject to complete adaptation, as partially revealed in recent studies but more have to be done.

    4. Results (e.g. D.T. Lykken) showing the importance of genetic factors in affecting many variables including happiness tend to support John’s point about the difficulty of changing happiness. However, I think that science and technology has brought us to the brink of a quantum leap in happiness. As I explain in the paper mentioned above, brain stimulation and genetic engineering, with sufficient safeguards, will breakaway the constraints of relativity and genetics.

  12. Here’s another mechanism, or perhaps more properly something else that needs to be adjusted for.

    I don’t recall references off hand, but studies were made of how people felt after major injuries like amputation or paraplegia. The initial effect was that they felt worse, but curiously follow ups showed that over time they tended to revert to their own personal norms before the injury.

    To the extent that a similar effect applies to people experiencing upward or downward income mobility, you would expect that upper and lower income groups would be happier/sadder to the extent that they included people who had recently moved there but the reference norm would be largely independent of absolute income. People would indeed be making relative income comparisons – but not with other people, with their own trajectories over time. Again, this would need testing, e.g. for spreads that correlated to mobility.

  13. “In fact one of the ideas Freud picked up upon was the “compulsion to repeatâ€?, and the way in which our unconcious drives us to repeat tasks, that may be pleasurable or unpleasurable for no apparent rational purpose.” (timw@6)

    According to the fMRI people, the brain state for preference is identical to that of familiarity. I think that explains a lot of human activity.

    More generally, the idea that humans are rational maximisers of utility or happiness is just crazy old junk. It’s pretty well on par with any other superstitious belief: it sounds nice. But there’s a plethora of counterexamples.

    The science points elsewhere: we are maximising a condition of our dopamine system. This could involve making money or loosing it, having sex, educating your kids, or sitting quietly in a church thinking about God. Or posting on a blog. Rats with direct access to this system via electrode implants will continue to press a bar to stimulate this system until they actually drop dead from hunger or thirst, rather than actually eat or drink. They’re beating the system. Drugs of addition like alcohol, opiates, etc, plug directly into this system too. People who have good brain function are able to integrate things like cultural values and long term goals into this system. Those whose brain function has been compromised – by things like accident, disease, poor nutrition, poor education and trauma – tend to respond to more basic biological inputs.

    These are generalisations, but there’s a lot of specific hard science being done in this area that is gradually supplanting cute sounding ideas like rational maximisation with much more complex verified theory.

  14. re: #10
    For many people, declines in income, especially if sudden, naturally cause more unhappiness than the increase in happiness caused by equivalent gains.

    Consider a couple making $X, paying a mortgage on a house they’ve lived in for 10 years, with kids in local school, and both spouses in local jobs they like. Suppose they get raises – they can buy a new car, take nicer vacations, maybe save more.

    On the other hand, suppose one loses their job, and their mortgage is no longer payable. They have to sell their house and move somewhere else, which may not be the same town.

    Here’s a good test: suppose you have $Y in total net worth, and you can flip a fair coin with result:
    lose $Y, so you lose everything
    gain $Z: you specify the smallest $Z that will make you take this bet.

    For most people, I’d guess that $Z >> $Y, even though the expected value of the bet is positive for $Z>$Y.

  15. re: #7 (at least, the car)

    Male wealth/power is attractive to human females seeking reproductive partners, for the obvious reason that potential children will be afforded more resources. Hence, for young women, male scions of rich families are attractive, and sports cars are good portable conspicuous displays of wealth.

    But, this general attraction persists in many other circumstances, and at least part of the comparison is *relative* in many human societies, weather the wealth is measured in $Ms or numbers of cattle.

    Hence, there is an obvious reason for wealthier males to be happier.

    Note that within a group, some resources may be purchasable by every member of the group once they reach a given income, i.e., the supply is not particularly limited, and the price won’t change (much) with changes in demand. For these, no one cares whether they make more than their neighbors.

    On the other hand, some important resources are limited, and their prices get bid up. Otherwise identical houses can vary tremendously in price by location, especially as in some places, where you live determines the school system attended by your children, safety, goodness of environment, convenience, etc. Unsurprisingly, not only men, but women rationally care about such things, and are likelier to be happier if they can afford to bid for such resources.

    Here’s a quick check: suppose you are a 21-year-old male at college, and you and the other students all get about $X extra money from your parents.

    Here are 3 cases, in which you always get $2X:

    A. You start getting $2X, no one else does.
    B. All of you start getting $2X.
    C. You get $2X, but everyone else gets $3X.

    Are you equally happy in all 3 cases?

    Likewise, if you are seeking to buy a house, you love A, as you can outbid everyone and choose the house you wish, B is neutral, and C is awful.

    I would not ascribe all of the observed effects to this, but I’d bet at least some of it is, and it’s actually rational for wealthier people to be happier if important resources are limited enough that their prices get bid up.

  16. This topic is as silly as are most of the comments – equating happiness with economics is like measuring marriage with a slide rule.

  17. #17 Mashey, spoken like a true Alpha male. For once I have to agree with rog – what a silly comment (apart from being mostly redundant).

  18. On ‘the economics of happiness’:

    The notion of happiness seems to be quite complex. One of the aspects is the state of being ‘satisfied’. Suppose we concentrate just on this aspect then I can say that the theory of ‘competitive private ownership economies with complete markets’ (‘free market economics’) has the unfortunate implication that happiness, thus defined, is incompatible with ‘free market economics’, thus defined, because without the assumption of ‘non-satiation’ (ie wanting more), it is difficult to show that markets will clear (supply equals demand in micro texts).

    So, without evidence to the contrary, I don’t believe anybody who wants to sell me the belief that ‘free markets’ is a means to achieve happiness and I hand over to psychologists who have a notion which seems to be relevant. The notion is called ‘cognitive dissonance’.

  19. Justin Wolfers has recently put forward a compelling rebuttal to the existence of the Easterlin paradox. Apparently much of the data included comparisons of different questions as to how happy people were filling over time. His analysis of all the data shows a clear correlation over time between average gdp of a country and that country’s average happiness as well as, within a country, income and happiness levels.


  20. ProfQ wrote:

    This is often taken to mean that it’s relative rather than absolute income that determines happiness…

    It seems the absolute/relative income effect applies on the downside as well … its comforting if everyone is in the same boat:

    NY Times: Stunned Icelanders Struggle After Economy’s Fall.

    Aldis Nordfjord, a 53-year-old architect, also lost her job last month. So did all 44 of her co-workers — everyone in the company except its owners. As many as 75 percent of Iceland’s private-sector architects have probably been fired in the past few weeks, she said.

    In a strange way, she said, it is comforting to be one in a crowd. “Everyone is in the same situation,� she said. “If you can imagine, if only 10 out of 40 people had been fired, it would have been different; you would have felt, ‘Why me? Why not him?’ �

  21. Generally my observations of life would support the Easterlin paradox. Another vector, however, is age. Much can be learned by studying the effects of the progressive loadings of people in the full cycle from cradle to grave.

  22. Spot on John Mashey… it’s all relative.

    There are two concepts here – absolute happiness/wealth and relative happiness/wealth.

    There is clearly some absolute level of wealth below which it is very difficult to be happy. If you are struggling for survival you haven’t got time to be happy. Call this the poverty line (and an absolute poverty line not the relative poverty line nonsense that some use).

    Once over the poverty line, happiness can increase with wealth but it’s more relative and there are diminishing returns. If gave you $1m tomorrow that would probably make you happy. If I gave you $2m you’d probably be slightly happier, not twice as happy.

    Once over the poverty line there are far greater factors than wealth that impact happiness which is why it becomes relative. A satisfying job, close family, good friends, sense of self worth – they would all swamp wealth as a happiness predictor once on a good income.

  23. Late comment. Looking for a quantitative description of happiness (a qualitative parameter) will always be difficult. However it seems clear that if one wishes to describe happiness mathematically, one would need to start thinking about derivatives (in the calculus form of the word). I have long held the view that one is happiest when ones measure of wealth, power or s-xual success is, or will soon be, increasing, and one is least happy when the derivative is negative.

  24. Paul @ 22; while there no longer seems to be a paradox, it’s important to note the logarithmic scale of those graphs. A log linear relationship means that, if I am a millionaire, it would take another $9 million to increase my happiness by the same extent as increasing someone else’s wealth from $100 to $1000 would do. If we want to maximise happiness, there’s still an argument for highly redistributive taxation there.

  25. That is a damned good qualifier James Houghton. It puts the correct complexion on these ridiculous CEO salary packages.

  26. Just remember, happiness can’t buy you money.

    As for trying to maximise happiness by progressive taxation, it makes more sense to use regressive taxation whereby only a small group pay taxes at all. It gets the maximum number ahead of the misery point, at the point of making a few very miserable.

  27. P.M. Lawrence, isn’t the normal definition of regressive taxation a tax that affects the poor more than the rich? e.g. the GST (or carbon taxes, for that matter). I don’t understand your point.

  28. Have you read Stevenson and Wolfers (NBER working paper 14282)? I’m not an expert on the literature, but they make a pretty good case that income has a diminishing but positive effect on happiness over the whole range we have data on – crudely happiness = log(income) rather than happiness = a if income > b. They claim that the US postwar experience is an outlier.

  29. JH, if you have a regressive tax you will increase the total misery, but by making a few people very miserable. You have a better chance of a greater number not being miserable.

  30. So if happiness, H is proportional to ln(I) where I is income then the rate of change of happiness is proportional to 1/I. Since your income (which can be negative) is in fact the change in your wealth, is happiness then inversely proportional to your wealth?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s