A simple model of lifetime happiness

From Jeff:

Suppose that what makes a person happy is when their fortunes exceed expectations by a discrete amount (and that falling short of expectations is what makes you unhappy.)  Then simply because of convergence of expectations:

  1. People will have few really happy phases in their lives.
  2. Indeed even if you lived forever you would have only finitely many spells of happiness.
  3. Most of the happy moments will come when you are young.
  4. Happiness will be short-lived.
  5. The biggest cross-sectional variance in happiness will be among the young.
  6. When expectations adjust to the rate at which your fortunes improve, chasing further happiness requires improving your fortunes at an accelerating rate.
  7. If life expectancy is increasing and we simply extrapolate expectations into later stages of life we are likely to be increasingly depressed when we are old.
  8. There could easily be an inverse relationship between intelligence and happiness.


thanks god this assumption doesn't necessarily hold true.

Even if it was, I don't see how all of these points follow from it. Do they really? Even for the sake of argument that'd be a very strong premise then if it led to all these eight far-reaching consequences.

thank god for us maybe, but there'd be a lot less human suffering too if the obverse applied

I think this fails because recent research shows that human memory does not act like a recording machine - even in the most intelligent people. Recall of past has been clearly shown to vary over time and is adjusted to suit ones current situation. This means positive and negative events may be correlated to constructed memories of past expectations to produce a variety of current psychological attitudes.


"Recall of past has been shown to vary over time and is adjusted to suit one's current situation." So artfully demonstrated in the novels of Kazuo Ishiguro.

Explains the devotion to Keynesian economics.

All we can be sure of is uncertainty.


And there is the pithy idea that only when it is lost are you aware that you were happy.

And (in my own life at least) that once you become unconcerned about seeking it, it comes much more easily.

(though maybe that's a corollary to what Jeff is saying with 'expectations').

This is remarkably similar to a Miles Kimball working paper (with Bob Willis) that has been kicking around for a decade or so...they formalize the model and get things like habit formation out of it. It's neat.

Exactly what I thought of when I read the post... Also I would note that life satisfaction and happiness are generally not considered the same thing in that literature. I suspect the bulk of utility is more like life satisfaction and meeting expectations. Not sure about intellectuals, but I bet contrarians are happier than average cause 'it could always be worse.'

I would add that, from personal experience, happiness is a displaced state. Like sadness, it's displaced from neutral, and to me seems to have a higher energy requirement. So it seems a bit closer to a change of state than a state itself.

Contentment or satisfaction, to Claudia's point, seems to me a neutral state. Reading the paper over breakfast on Sunday makes me contented, but not happy. I am happy when something new or unusual happens.

That is reminiscent of the "set-point" theory of happiness: Each of us has an internal, default level of happiness and events cause short-term spikes or valleys but given time we return to our own unique baseline.

#3 falsifies the model for me if falsifying #3 falsifies the model, which also falsifies the set-point hypothesis, for me, which basically says that environment doesn't matter. I couldn't disagree more, for myself, and I've distilled it down to the less of a "school" environment I'm in the happier I've been.

Set point just means that the environment can temporarily perturb you but it can't permanently change you. The young are often protected from environmental shocks by the adults around them, though I suspect fluid intelligence and less overall stress helps younger people process events in a more positive light too.

Andrew' are you sure you aren't just a put upon middle child? My bosses before grad school told me to enjoy it because it would be the best time of my life compared to the 15 or so years after. Five years into the other side and they're still right. Of course I did learn how to sit still and respect my teachers in kindergarten so maybe grad school was easier for me than some.

A large part of that is the grass-greener-on-the-other-side syndrome I'd say. You hate grad-school while in it and then you'll hate the real world when in it too.

Love the armchair psychology.
1) Claudia, you are in economics. As I've said, you jokers have the best racket going.
2) I said schooling. Grad school is a scam, that's a fact. But that's not what I referred to. The jobs were fine. More than fine. Too fine. That's how I realized that for me it was schooling.

"If life expectancy is increasing and we simply extrapolate expectations into later stages of life we are likely to be increasingly depressed when we are old" except that a lot of studies I have seen suggest happiness actually increases with age, here's one reference.


I think that actually a more simple model is that you are happiest when you are surrounded by people who like and respect you, like your family. Possibly as you deepen and make your close relationships more numerous (as happens as you get older) then you can become more and more happy.

"Indeed even if you lived forever you would have only finitely many spells of happiness."

How many spells of happiness would you have, then?

I think that this conclusion (point 2) is incorrect.

Under this conclusion, if an individual lives for an infinite length of time, expectations and reality will converge completely.

It's more reasonable to think of this as a probability that any one expectation will turn out to be correct, and (assume) that this probability will approach 1 with time (ie. convergence of expectations). It is likely that this approach would be asymptotic (assuming perfect memory and a rational actor), with never before experienced situations still occuring, though less frequently and with less difference from the previous most similar situation.

Therefore in an infinite amount of time there will be an infinite number of "spells of happiness". However, they would (under the model) mostly occur earlier in life.

Either way, the premise doesn't necessarily hold true as lff notes above.

There are different numbers of infinity

But only one form of finite.

Presupposing that people actually will feel best about their lives when they are able to maximize only their happiness, when in fact we often feel best looking back on a life in which we have experienced the gamut.

This model of what makes a person happy is hopelessly naive. If it where true, it'd be an argument in favor of being a pessimist. If you are a pessimist, your forecasts have a permanent tendency to be below the actual outcome, thus you continually exceed expectations, yet your pessimsism prevents you from adjusting expectations upward to the degree really warranted.

In the real world, pessimists aren't happier than optimists. Despite the fact that a pessimists expecations are more frequently exceeded. This to me strongly suggests that the model is flawed if not outright wrong.

Yeah, I was thinking the same thing. It seems to endorse the hedonic treadmill, not recognizing that there are cognitive tricks you can use to get off it.

further proof that economists should stick to monetary policy and away from psychology

Studies show unhappiness actually hits its highest point in middle age.

Yep. There goes #3.

Does happiness-research and the many empirical studies support any of those eight conclusions?

I thought that had been debunked recently?

Overcomingbias discusses:


9. Some times that are experienced as "happy", also carry with them some/many unhappy emotions and experiences.

10. Some times that are experienced as "happy" will result in many more unhappy experiences for the rest of your life.

See: Sleeping with your biographer.

Trial over perjury, for instance. Tho perhaps a "blackmail" defense for perjury accusations might exonerate somebody, a top Dem politico perhaps (tho certainly never a Rep).

In three years I bet Petraeus looks back on this period as the best thing that ever happened to him. He'll have learned to divorce his identity from his work and social standing. Either that or he'll become an alcoholic and descend into bitterness and early death.

If you're giving those two outcomes a 50-50 chance the expected value seems similar to status quo.

Happiness is very probably not something we passively experience in light of certain rewards, but is instead an emotion (like any other emotion) meant to regulate our behavior *once we are in a social situation where such rewards are present*. I explain more here: http://sarkology.wordpress.com/2012/07/10/happiness-as-a-behavioral-strategy/. I also humbly solicit any criticism you may have.

Studies show unhappiness reaches a peak in middle age and people actually feel better as they age. This piece seems nothing more than the personal blather of somebody with an economics blog.

Sorry, I thought this comment was lost in the ether.

I have yet to see a definition of happiness that is useful for anything than producing models to write papers about.

I also have yet to see a definition of happiness that makes me want happiness to be my major goal in life.

Confucius said: the wise likes water, the benevolent likes mountain; the wise will be happy, the benevolent will live longer.

I am quite a happy fellow and I can tell you that the model is incorrect. Happiness is an attitude. First, you are happy and then you reinterpret everything that happens to you to find the silver lining.

I have reached this conclusion after spending a lot of time thinking about the difference between me (a happy man) and the people I know who are not happy. I have concluded that, without exception, everyone that I have talked to and is not happy chooses to see life from a negative and pessimistic perspective, which has very little to do with how their life is (some of them are quite successful).

For example, a negative person will complain at the office because he/she has to deal with a easy project that is boring, but then will complain when they are assigned a challenging project because it is too much work. A positive person, under the same circumstances, will be happy to have an easy assignment in the first case, and have the opportunity to learn in the second case.

Fortunately, I think, being happy is a skill that a normal person can learn (I assume that there are people who are chemically imbalanced and unable to see the bright things of life so easily).

I like your opinion. To be happy needs power of will sometimes, at least in the long-run.

I have to agree, I'm pretty much happy 80-90% of the time, and that hasn't changed with how successful I was or what I'm doing. It was a real surprise to me that other successful people, were generally much less happy than I was.

Marty Seligman argues much the same in his book Learned Optimism. I'm happier than I used to be after reading that and applying the ideas.

Over the years, one of the best parts about MR is the number of great books I've discovered from tips in the comments.
Thanks! I'm loving the book; just bought it.

See also The Happiness Hypothesis, which I believe is by one of Seligman's former students.



Did you expect to be able to be reading a how-to happiness book less than two hours after learning it exists?

With kindle it's always a fair chance!

The expectations side of the assumptions seems wrong. Throughout your life expectations shift in different directions and gain nuance.

One more implication of this assumption is that pessimistic people should be very happy since their expectations are low. The fact that this isn't true (or is it?) along with the studies that show that people become happier as they grow older seems to imply that the presumption is wrong.

Might be also because value their fortunes less (due to their pessimistic mood) and that cancels it out

semantic issue, what's meant by happiness here is vague. happiness as a state exceeding expectation sounds reasonably scarce but happiness as prolonged satisfaction definitely sounds possible for most of a lifetime.

Doesn't it also mean that once you adjust your expectations for shorter periods of happiness, your happiness will increase?

Or is that not captured by the model? (didn't read the link)

Doesn't this model mean that you will be happy about half of the time? It does depend on the size of the discrete unit above expectations needed, but so would unhappiness, if it is symmetric.

Perhaps it is not symmetric-- in which case I would guess that avoiding results below expectations is more important than getting results above expectations, so variance is bad.

fortunes exceed expectations
And what about Christmas?
For most people, most of the Christmas season is spent in a variety of unhappy stressful situations trying to do something nice for others, with lots of happy expectations that others will enjoy the presents you give them, as well as some expectation of getting nice presents for oneself.

Haven't there been quite a few posts about how the actual experience of a vacation was less pleasurable than the memories of the good parts of the vacation?

And this is where the "happy optimist" chooses to be happy, by choosing to think about / remember the good parts of vacations or Christmas.

Asymmetrically, many unhappy people are thinking about/ worrying about, negative things they don't fully control. With high expectations of good results, and then unhappiness when less good results are achieved. "Don't worry, be happy" sure seems empirically accurate in many cases.

Finally, too much of this happiness lit seems oriented towards justifying some gov't policy to increase essentially unmeasurable happiness, rather than #s who have running water, a legal place to live/sleep each night, enough food calories, enough clothes.
As absolute poverty goes down in any country, the need for more gov't to reduce poverty should be going down, but it seems in many countries, including the US, there are increasing calls for even more "anti-poverty" programs, despite already much less absolute poverty.

If we have to stick with Eight, why not theEightfold path instead ?


"Suppose that what makes a person happy is when their fortunes exceed expectations by a discrete amount (and that falling short of expectations is what makes you unhappy.) "

Suppose that 1 = 2. Proceed.

I agree here. Our expectations of what will make us happy are not necessarily what make us happy. Often quite to the contrary.

Happy wife...

This is performance art, people. "Economist Deep in Thought."

Ha, sounds like a Buddhist economist.

He could do this for thousands of things. Just come up with a simple, but powerful, idea (Life sucks, lower your expectations, and you'll be a lot happier), and add a little econ jargon (like "convergence of expectations," "cross-sectional variance," "accelerating rates," and "extrapolating expectations").

Yes, exactly. Economists have been reinventing the wheel in other disciplines for a good decade now.

#8 — yes, that's Ecclesiastes 1:18.

Point #2 is a red herring arising from using the wrong kind of infinity. It *seems* to be opposing a finite number of happy periods against an infinite number of unhappy periods, and therefore concluding that the fraction of one's existence that would be happy is effectively zero.

But that can't be right. The kind of infinity associated with living forever is a *potential* infinity, not an *actual* one. That is to say, the claim that someone is going to live forever is equivalent to saying that "There is no number N such that N is the maximum age that this person will reach." It is definitely *not* equivalent to saying that "There will come a time T when this person will have lived infinitely many years."

Therefore it is true to say that at all times during his life an immortal individual will only ever have lived finitely many years. Therefore point #2 is quite consistent with our subject having been happy all his life. So Point #2 is, sadly, vacuous.

As someone who has read more papers on happiness economics than he really wants to have, I can report a stylized fact from the lit is that there is a U-curve relation between happiness and age, ceteris paribus. The lowest point of happiness is middle age, 40s and 50s, when children are whiney adolescents and one may have responsibility for declining parents. Ironically, this is also the age of peak income earnings, further undermining the strict relation between income and happiness that does not exist. Happiness tends to go up after mid-50s or so, although bad health and spousal death tends to knock it down temporarilay at least, although the very latest lit has it dipping down again past 85 or so, when bad health and loneliness tend to overwhelm all that "wisdom of age" and not working hard that seems to be behind the earlier increase.

As I pointed out above, this seems to have been debunked.

Overcomingbias discusses:


Sorry, Finch, the study shows no such thing. Hanson misreads it badly, and the panels are only 10 years long. The minima for the countries shown are still in middle age, varying from around 40 to around mid-50s, with all showing increases in happiness afterwards, peaking around late 60s to early 70s. The main difference with other studies is finding a sharp decline after 75, but I already noted that some find declines a decade later. Does not contradict substantially my summary.

What we really need is wireheading, or hedonic enhancement generally.

Happiness is in the brain, if we could use artificial means to manipulate the brain into feeling happiness, feeling excellent would become a robust pastime, even if life expectations matched life events perfectly.

I think this suggests the managed earnings theory of corporate financial reporting. Each quarter you surprise the analysts by reporting a penny or two over the consensus expectations you have led them to, until the quarter when the accumulated divergence from reality can no longer be hidden, and you take the big bath.

I can be happy by always surprising myself through optimistic interpretation of the world, except every few years when I make a steep downward reassessment that forms the basis of future happiness.

Comments for this post are closed